PDE1B KO Confers Resilience to Acute Stress-Induced Depression-Like Behavior

Home , EHNA, Etazolate, Tofisopam

PDE1B KO confers resilience to acute stress-induced depression-like behavior

A dissertation submitted to the

Graduate School of the University of Cincinnati

in partial fulfillment of the

requirements for the degree of

Doctor of Philosophy

in the Molecular and Developmental Biology Program

of the College of Medicine

Jillian R. Hufgard

B.S. Rose-Hulman Institute of Technology

April 2017

Committee Chair: Charles V. Vorhees, Ph.D.

ABSTRACT

Phosphodiesterases (PDE) regulate secondary messengers such as cyclic adenosine monophosphate (cAMP) and cyclic guanosine monophosphate (cGMP) by hydrolyzing the phosphodiester bond. There are over 100 PDE proteins that are categorized into 11 families.

Each protein family has a unique tissue distribution and binding affinity for cAMP and/or cGMP.

The modulation of different PDEs has been used to treat several disorders: inflammation, erectile dysfunction, and neurological disorders. Recently, PDE inhibitors were implicated for therapeutic benefits in Alzheimer’s disease, depression, Huntington’s disease, Parkinson’s disease, schizophrenia, and substance abuse. PDE1B is found in the caudate-putamen, nucleus accumbens, dentate gyrus, and substantia nigra–areas linked to depression. PDE1B expression is also increased after acute and chronic stress. Two ubiquitous Pde1b knockout

(KO) mouse models, both removing part of the catalytic region, decreased immobility on two acute stress tests associated with depression-like behavior; tail suspension test (TST) and forced swim test (FST). The decreases in immobility suggest resistance to depression-like behavior, and these effects were additive when combined with two current antidepressants, fluoxetine and bupropion. The resistance to induced immobility was seen when PDE1B was knocked down during adolescence or earlier. Expression of Pde1b mRNA and protein showed that Pde1b is localized to dopaminergic and glutamatergic post-synaptic cells. Serotonin transporter, dopamine transporter, and dopamine receptor D1a (Drd1a) Cre drivers were tested and only Pde1b KO in Drd1a specific cells was sufficient to reproduce the immobility phenotype in TST. Intrastriatal injections of dopamine receptor, D1, agonist (SKF38393) and antagonist

(SCH23390) revealed an additive immobility effect of Pde1b KO and dopamine receptor antagonism. The D1 agonist reversed the genotype immobility phenotype. The results suggest that reduction of PDE1B causes resistance to acute stress-induced depression-like behaviors through the dopamine pathways.

iii

ACKNOWLEDGEMENTS

I would like to thank my advisor, Dr. Charles Vorhees, for providing the opportunity to learn and perform research in his laboratory, his encouragement and guidance, and for his dedication to leading his students through the roller coaster that is graduate school. I would also like to thank Dr. Michael Williams and Dr. Matthew Skelton for their constant technical support throughout this entire process. In addition I would like to thank the Vorhees, Williams, and Skelton laboratory members both past and present, especially Chiho Sugimoto for her willingness to always help. The entirety of my committee also deserves significant acknowledgements for their advice and guidance of the following research.

I would like to thank my family for their encouragement and support in this process and in life. An additional thank you is extended to those in Cincinnati, both friends and dogs, that became my family during this process. Their companionship, lovingness and confidence in me provided the strength necessary to complete this process. And lastly, I would like to thank

Shaun Wendel for being my rock during this process and his constant reminders to make sure I enjoy whatever I am doing.

TABLE OF CONTENTS

ABSTRACT ...... ii

ACKNOWLEDGEMENTS ...... iv

TABLE OF CONTENTS ...... 1

LIST OF TABLES AND FIGURES ...... 4

LIST OF SYMBOLS ...... 6

CHAPTER 1

The Role of Phosphodiesterases in Major Depressive Disorder ...... 8

Major Depressive Disorder ...... 9

Brain development in the pathophysiology of depression ...... 10

Current antidepressant treatments ...... 13

Postsynaptic targets for MDD treatment ...... 14

Regulation of secondary messengers by phosphodiesterases ...... 16

PDE1 ...... 17

PDE4 ...... 19

PDE10A ...... 22

Phosphodiesterases in dopaminergic pathways ...... 24

Conclusion ...... 25

References ...... 27

Tables ...... 36

Figures ...... 38

CHAPTER 2

Phosphodiesterase-1b (Pde1b) knockout mice are resistant to forced swim and tail suspension test induced immobility and show upregulation of Pde10a* ...... 41

Abstract ...... 42

Introduction ...... 43

Methods ...... 45

Results ...... 50

Discussion ...... 53

References ...... 58

Tables ...... 66

Figures ...... 69

Acknowledgments ...... 74

CHAPTER 3

Phosphodiesterase-1b deletion confers depression-like behavioral resistance separate from stress-related effects in mice ...... 75

Abstract ...... 76

Introduction ...... 77

Methods ...... 78

Results ...... 85

Discussion ...... 87

References ...... 91

Figures ...... 99

Acknowledgements ...... 113

CHAPTER 4

PDE1B is implicated in dopamine pathways and its deletion in dopamine receptor 1a cells confers an immobility-resistant phenotype in forced swim and tail suspension tests ...... 114

Abstract ...... 115

Introduction ...... 116

Methods ...... 118

Results ...... 122

Discussion ...... 124

References ...... 128

Figures ...... 133

Acknowledgements ...... 144

CHAPTER 5

Discussion ...... 145

Secondary messenger regulation in depression ...... 146

PDE1B regulation in activity and depression-like behaviors ...... 147

PDE1B phenotype differences in male and female mice ...... 150

PDE1B reduction at developmental time points ...... 151

PDE1B regulation in dopaminergic pathways ...... 153

PDE1B in serotonergic and glutamatergic signaling ...... 155

References ...... 160

Figures ...... 165

LIST OF TABLES AND FIGURES

Chapter 1

Table 1. Classes of antidepressants and examples of those currently available ...... 36

Table 2. PDE enzyme family's selectivity, localization, and function ...... 37

Figure 1. Mechanism of action of tricyclic antidepressants ...... 38

Figure 2. The role of PDE1B, PDE4, and PDE10A in dopaminergic signaling ...... 40

Chapter 2

Table 1. Description of mice used in experiments ...... 66

Table 2. Primer sequences ...... 67

Table 3. Chronic Variable Stress Paradigm ...... 68

Figure 1. Pde expression in WT and KO mice striatum and cerebellum ...... 69

Figure 2. Pde1b KO produced resistance to induced immobility in FST and TST compared with

WT littermates ...... 70

Figure 3. Pde1b KO mice have a similar antidepressant-like phenotype as currently marketed antidepressants when compared with WT littermates...... 71

Figure 4. PDE1B is elevated by acute stress...... 72

Figure 5. PDE1B is elevated while PDE10A is reduced in chronically stressed mice...... 73

Chapter 3

Figure 1. Generation of floxed mice and confirmation of global knockout mice (KOCMV) ...... 99

Figure 2. Fluorescent immunohistochemistry analysis of PDE1B in the brain localizes to regions related to stress and depression...... 101

Figure 3. KOCMV mice have decreased immobility time in depression-related tasks compared with control littermates...... 103

Figure 4. Pde1b KOCMV mice are hyperactive at night and when introduced to the novel home cage, although land hyperactivity did not predict water hyperactivity ...... 105

Figure 5. Pde1b KOCMV mice do not differ from WT littermates when exposed to the learned helplessness procedure ...... 107

Figure 6. Analysis of localization and quantity of PDE1B in KOP0, P32, P60 mice ...... 110

Figure 7. When PDE1B is removed prior to sexual maturity KO mice had a reduction in immobility time in the TST and the FST...... 112

Chapter 4

Figure 1. PDE1B co-localizes with DRD1A and NMDA but not DAT or SERT...... 133

Figure 2. Pde1b localizes in the striatum to the cytosolic space of cells also expressing Drd1a,

Drd2, and Nmda...... 134

Figure 3. Pde1b localizes in the hippocampus to the cytosolic space of cells also expressing

Drd1a, Drd2, and Nmda ...... 135

Figure 4. Pde1b localizes in the substantia nigra to the cytosolic space of cells also expressing

Drd1a, Drd2, and Nmda...... 136

Figure 5. Neither Sert nor Dat specific KO of Pde1b induced an antidepression-like phenotype in comparison with WT littermates...... 138

Figure 6. Drd1a targeted KO of Pde1b produced a decrease in immobility in TST when compared with WT littermates...... 140

Figure 7. Pde1b KOCMV mice exhibit a decrease in immobility time in TST and FST that is exacerbated after D1 antagonism but mitigated after D1 agonism ...... 142

Figure 8. D1 cell specific activation and regulation of cAMP...... 143

Chapter 5

Figure 1. The role of PDE1B in AMPA receptor phosphorylation and cell surface regulation... 165

LIST OF SYMBOLS

AD Alzheimer's Disease

AMPA α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid

ANOVA Analysis of variance

BDNF Brain derived neurotrophic factor

CaMKII Ca2+ /calmodulin-dependent protein kinase II cAMP Cyclic adenosine monophosphate

Cdk5 Cyclin-dependent kinase 5 cGMP cyclic guanosine monophosphate

CMS Chronic mild stress

CREB cAMP response element binding

D1 Direct dopamine pathway

D2 Indirect dopamine pathway

DA Dopamine

DARPP-32 Dopamine- and cAMP-regulated phosphoprotein, Mr 32 kDa

DAT Dopamine transporter

DISC1 Disrupted in schizophrenia 1 dKO double knockout

DRD Dopamine receptor dopamine

Epac Exchange factor directly activated by cAMP

FST Forced swim test

GABA gamma-Aminobutyric acid

GLUT Glutamate transporter

HD Huntington's Disease

HET Heterozygous

i.d. In diameter

IHC Immunohistochemical

ITI Intertrial interval

KD Knock down

KO Knockout

LH Learned Helplessness

LS Least Squared

LTP Long term potentiation

MDD Major depressive disorder mRNA Messenger ribonucleic acid

MSN Medium spiny neuron

NE Norepinephrine

NMDA N-methyl-D-aspartate

P Postnatal day

PD Parkinson's Disorder

PDE Phosphodiesterase

PK Protein Kinase

PSD95 Postsynaptic density protein 95

PSMB2 Proteasome Subunit Beta 2 qPCR Quantitative polymerase chain reaction

SEM Standard error of the mean

SERT Serotonin transporter

TST Tail suspension test

WT Wild-type

CHAPTER 1

The Role of Phosphodiesterases in Major Depressive Disorder

Major Depressive Disorder

Major depressive disorder (MDD) is the fourth leading cause of world-wide disability; in fact, in 2010, 298 million people were considered to be clinically depressed (Vos et al., 2013).

Of these, many are women and/or elderly. Women are twice as likely to be diagnosed with

MDD as men. Moreover, MDD is a progressive disorder explaining the high rate of diagnosis in the elderly (Kuehner, 2003, Kulkarni and Dhir, 2009). Whether the higher rates in women and the elderly are due to susceptibility differences or a greater willingness to treatment is unknown.

In recent years, there has also been an increase in the diagnosis of MDD in younger populations. These increases may be attributed to increasing risk from social and environment stress or better awareness, diagnostics, and decreased stigma.

The symptoms of depression widely vary from person to person with hypersomnia or insomnia, increased or decreased appetite, fatigue, or headaches. Patients report inability to experience pleasure and focusing on feelings of helplessness, hopelessness, and worthlessness. Family and friends describe them as lethargic, withdrawn, agitated, and melancholic with poor concentration and memory. These signs and symptoms can affect one’s, work, and social interactions. Aside from affecting interpersonal relations, MDD is the leading cause of suicide in the United States, 2 in 10,000 deaths are reportedly caused by suicide and

35-40% of suicides are linked to MDD (Cassano and Fava, 2002, Kulkarni and Dhir, 2009).

Despite the prevalence and debilitating effects of MDD there are no known causes.

Many hypotheses have been proposed, including: diathesis-stress, damage to the cerebellum, genetic variation of the serotonin transporter, decreased hippocampus size, decreased adult neurogenesis, over-active (hypothalamic-pituitary-adrenal) HPA axis, social and environmental stimuli, developmental perturbations, substance abuse, or combinations of these. The diathesis-stress model suggests that patients become depressed after a stressful life event but only if they had a preexisting vulnerability, such as an over-active HPA axis or a short serotonin transporter allele. When exposed to similar stressors not everyone has the same response.

The diathesis-stress model supports this observation by suggesting both the individual’s previous experiences and genetic predisposition play a role in the response. Interestingly, individuals who claim to be part of a religious community have lower rates of depression, possibly due to a positive social community or more effective coping mechanisms (O. Harrison,

2001). Social interactions can play a large role in the vulnerability to becoming depressed after a stressor. During childhood, children take cues from those surrounding them on how to respond to life events. If they are exposed to family disturbances, poverty, and social isolation they may develop a different understanding of how to interact and cope than someone who is not brought up under these circumstances and as a result may be more prone to depression.

Brain development in the pathophysiology of depression

Childhood stress may cause differing adult responses because these exposures are happening during brain development. During brain development, neuronal connections are continuously strengthening, synaptic pruning is occurring, and new neurons are being generated at specific periods. Perturbations to these processes may lead to altered MDD risk later in life. Both the hippocampus and striatum are involved in mood, motivation, and reward suggesting that changes in development in these areas may affect MDD.

The striatum originates from the telencephalon and prenatal patterning from fibroblast growth factor 8 (fgf8), homeobox transcription factors (dlx1,2,5,8), and basic helix-loop-help transcription factors (mash1) among others that aid in programming the organization and functionality of this region (Jain et al., 2001). The striatum is comprised of medium spiny neurons (MSN), 90%, and interneurons, 10%, receiving input from glutamatergic cortical projections and outputting γ-aminobutyric acid (GABA) projections to the globus pallidus

(indirect) or substantia nigra (direct) (Gerfen et al., 1990, Graybiel, 1990). The MSN of the striatum are subdivided into two compartments: the striosome and matrix, which have differing functions and development. MSNs in the striosome compartment are generated from E12-E17

in rats, and are dependent on dopamine (DA) input from the substantia nigra. By E19, these cells have aggregated into DA fiber islands and project to the substantia nigra pars compacta

(Jain et al., 2001). Those born from E18 to the early postnatal period independent of DA signaling give rise to MSNs in the matrix compartment that projects to the pallidum and substantia nigra pars reticulata (Jain et al., 2001).

During prenatal brain development, neurons are overproduced by approximately 50%; shortly before birth apoptosis reduces the number of neurons substantially (Andersen, 2003).

This overproduction is followed by selective elimination is a process repeated in the periadolescent stage suggesting these two periods may represent vulnerable periods of development (Andersen, 2003). Specifically, DA in rats at birth is approximately 10% of adult levels and originally was thought to increase linearly from birth to periadolescence (Andersen,

2003). DA receptors in the nucleus accumbens increase dramatically from birth until postnatal

(P) day 35 where levels plateau and remain throughout adulthood (Teicher et al., 1995). DA receptor, types D1 and D2, densities in the caudate-putamen increase up to 144% of adult levels, peaking around P40 (Teicher et al., 1995). This rise and then fall in DA receptors is also seen in the human corpus striatum suggesting a transitional state of organization (Seeman et al., 1987). Synaptic pruning deficits resulting in improper putamen reduction have been associated with MDD onset in adolescences and seen in adults with MDD (Whittle et al., 2014).

The hippocampus also originates from the telencephalon and differentiates into the dentate gyrus (DG), cornus ammonis (CA) regions (1,2,3), subiculum, presubiculum, parasubiculum, and entorhinal cortex. The hippocampus and related regions are patterned by expression of genes like wingless-related (Wnt), bone morphogenetic protein (Bmp), and others

(Khalaf-Nazzal and Francis, 2013). Within the hippocampus there are distinct patterns of communication from one region to another. The hippocampus also sends projections to the striatum, prefrontal cortex, and amygdala. During development the timing of synaptogenesis

depends on the maturation of post-synaptic cells. Of these, GABAergic synapses mature prior to glutamatergic cells (Tyzio et al., 1999, Hennou et al., 2002). This progression of development is thought to drive network activity and hippocampal circuit maturity (Khalaf-Nazzal and Francis, 2013). Similar to the striatum, the hippocampus also goes through periadolescent phases of pruning; the infra-pyramidal bundles undergo significant axonal pruning from P20-P30

(Khalaf-Nazzal and Francis, 2013). Serotonin is the earliest expressed monoamine and its expression has been linked to neurogenesis, neuronal removal, dendritic refinement, synaptic remodeling, synaptic maintenance, cell migration, and regulation of dopamine terminal growth

(Whitaker-Azmitia, 2001). Early in developmental serotonin (5-HT) promotes neurogenesis and fetal depletion of 5-HT results in neuronal reduction in the hippocampus and cortex that remains throughout adulthood (Lauder and Krebs, 1976). Shortly thereafter, 5-HT plays a role in dendritic length, spine density, and branching (Whitaker-Azmitia, 2001). Although 5-HT once promoted dendritic development following P21, where 5-HT levels are depleted, it can hinder further dendritic development (Norrholm and Ouimet, 2000).

The hippocampus, particularly the DG, is unique as it is one of the few brain regions where adult neurogenesis occurs. Adult neurogenesis has been linked to the pathophysiology and treatment of psychiatric disorders such as MDD. Stress, as reflected by increased glucocorticoid hormone, inhibits DG proliferation; blocking this increases cell division (Jacobs et al., 2000). Reduction of 5-HT has a similar inhibitory effect on cell proliferation (Brezun and

Daszuta, 1999). Enhanced serotonergic neurotransmission is sufficient to block chronic corticosterone induced depression-like behaviors and increases DG adult neurogenesis, suggesting alternations in adult neurogenesis may affect MDD and its treatment (Hill et al.,

2015).

Current antidepressant treatments

The first antidepressant treatments were discovered through antihistamine and antitubercular drug research. It was observed that these drugs resulted in depression symptom relief (Nestler et al., 2002). This lead to the discovery of monoamine oxidase inhibitors (MAO-I) and tricyclic antidepressants (TCA). Although the mechanism of depression was not known, these antidepressants were effective in some MDD patients and suggested that the pathophysiology may be attributed to dysregulated monoamine neurotransmission. Research originally focused on down regulation of norepinephrine (NE) but was later expanded to other neurotransmitters such as DA and 5-HT, and this extension led to the creation of selective reuptake inhibitors, a class of antidepressants which inhibit monoamine reuptake (Table 1)

(Nestler et al., 2002).

Reuptake inhibitors block the ability for monoamines like 5-HT, DA, and NE from being transported into the presynaptic terminal. This inhibition allows monoamines to bind to receptors longer leading to greater postsynaptic transmitter action (Figure 1). The acute mechanism of reuptake inhibitors is to increase neurotransmitter availability; however, this increase is not believed to directly cause relief of symptoms. Antidepressants take several weeks before they are efficacious suggesting an independent mechanism (Nestler et al., 2002).

The neurotrophic hypothesis suggests that antidepressants work by promoting neurogenesis and plasticity (Nestler et al., 2002). Antidepressants promote neurogenesis and plasticity, specifically they increase neurogenesis of 5-HT neurons and increase DA in pruning.

Despite production of multiple generations of antidepressants we are no closer to understanding the underlying mechanism of their efficacy although many theories have been postulated. This lack of understanding and only moderate efficacy is a factor in the high rate of nonadherence to patients taking these drugs. Of those with MDD that are currently symptomatic, 37% are not undergoing treatment (Béland et al., 2011). An estimated 60% of those being treated are expected to discontinue treatment in under 6 months, and up to 50%

report little to no benefit from taking these drugs (Trivedi et al., 2006). These statistics suggest that current antidepressant treatment insufficient in relieving symptoms for most patients.

Postsynaptic targets for MDD treatment

Changes in monoamine availability may not be the underlying cause of depression, therefore, investigators are looking at neuronal processes such as plasticity, neurogenesis, and apoptosis which have been shown to play a role in both depression and resistance to depression. Second messengers such as cyclic adenosine monophosphate (cAMP) and cyclic guanosine monophosphate (cGMP) are involved in the control of monoamines and the associated neuronal processes. Antidepressant treatment has been associated with increased cAMP and its downstream targets (Marazziti et al., 2009, Reierson et al., 2011). Increased levels of cAMP in dopaminergic neurons have been shown to activate the activity of norepinephrine, conversely increased cGMP in plasma and has been linked to depressive states (Marazziti et al., 2009). Signaling transduction through cAMP plays an important role in neuroplasticity; specifically, in long-term potentiation (LTP) and memory formation. cAMP has been reported to stimulate and maintain synaptic vesicle release of neurotransmitters (Marazziti et al., 2009). cGMP has also been shown to play a role in LTP and neurogenesis (Reierson et al., 2011). In 2002, Edgar et al. showed that long-term treatment with fluoxetine results in increased cGMP in the hippocampus (Edgar et al., 2002). In 2011, Reierson et al. demonstrated that increased levels of cGMP resulted in the induction of proliferation of progenitor cells in the subventricular zone and the DG; suggesting that cGMP plays a role in the induction and alleviation of depression (Reierson et al., 2011).

Secondary messengers participate in an intricate crosstalk with Ca2+/Calmodulin activity.

Ca2+ inhibits adenylyl and guanylyl cyclase activity, and, conversely, their activity through a negative feedback loop regulates the influx of Ca2+ into the cell (Marazziti et al., 2009).

Intracellular Ca2+ is linked to depression through the regulation of LTP; appropriate levels of

Ca2+ facilitate activation of Ca2+/calmodulin-dependent protein kinase II (CaMKII) that is critical for early phase LTP (Reierson et al., 2011). During late phase LTP, Ca2+, CaMKII, and cAMP stimulate a phosphorylation cascade that leads to nuclear cAMP response element binding protein (CREB) activation, neurogenesis, and synaptic structure modifications (Reierson et al.,

2011). Although Ca2+ plays a role in the mechanism that may protect against or ameliorate depression, high levels of intracellular Ca2+ have paradoxically been linked to depression (Guan et al., 2013, Li et al., 2013). The glutamate receptor N-methyl-D-aspartate (NMDA) allows the passage of Ca2+ into the cell, and when blocked with a noncompetitive antagonist such as MK-

801, causes an antidepressant-like effect in mice and rats (Papp and Moryl, 1994, Guan et al.,

2013).

Extracellular Signal-regulated Kinase (ERK) and CREB are part of the mitogen-activated protein (MAP) kinase pathway, regulated by Ca2+ and second messengers. This pathway, has been linked to human neuro-cardio-facial-cutaneous disorders and related syndromes

(Pucilowska et al., 2012). Post-mortem brains of MDD patients who died of suicide show decreased expression and activation of ERK (Guan et al., 2013). Stress has also been linked to the decrease in ERK phosphorylation in the hippocampus as seen in studies that induced depression via chronic corticosterone treatment in rats (Gourley et al., 2008). ERK2-deficient mice show deficits in emotional learning through conditioned fear even after overtraining, while another study showed anxiety in the elevated plus maze and light/dark test in these mice

(Samuels et al., 2008). They attributed the deficits in emotional learning and anxiety to a decrease in the firing rate and less frequent network bursts in the CA1 (Selcher et al., 2003).

Another study looking at the interaction of prenatal stress and the expression of ERK2 and

CREB showed a decrease in mRNA expression in the hippocampus of juvenile rat offspring

(Guan et al., 2013). They also showed an increase in forced swim test (FST) immobility time, suggesting a depression-like phenotype in the offspring (Guan et al., 2013). CREB, through

ERK and other pathways, is also linked to depression because of its role in neurogenesis and

neuroplasticity. Neurogenesis and neuroplasticity are one of the responses to stress and therefore potentially related to the etiology of depression (Guan et al., 2013). Studies of current antidepressants shows an increase in the expression and function of CREB in the hippocampus after treatment (Guan et al., 2013). These studies suggest a role of postsynaptic signaling in

MDD.

Regulation of secondary messengers by phosphodiesterases

Phosphodiesterases (PDEs) are intracellular secondary messenger regulators and are a superfamily divided into 11 families (1-11), each with distinct tissue distribution and functions

(Table 2). Each family is also made up of multiple isoforms resulting in 21 distinct PDE genes, some that have multiple splice variants resulting up to 100 different PDE proteins. PDEs regulate cAMP and/or cGMP by hydrolyzing the phosphodiester bond thereby terminating cAMP or cGMP signaling of downstream targets. The rate of hydrolysis determines the duration of cyclic nucleotide action on effectors such as protein kinase A (PKA) and G (PKG), exchange protein activated by cAMP, and cyclic nucleotide gated channels (Conti and Beavo, 2007). All

11 families share a common catalytic domain, but differ in GAF (PDE 2, 5, 6, 10, and 11), PAS

(PDE8), and anchoring domains (PDE 3, 4, 6, and 7), and in UCR (PDE4), COOH terminals, phosphorylation sites, and calcium/calmodulin binding (PDE1) (Bender and Beavo, 2006).

PDEs also differ in their binding affinities to cAMP and/or cGMP (Table 1). PDE 4, 7, and 8 are cAMP specific while 5, 6, and 9 are cGMP specific and 1, 2, 4, 10, and 11 affect both.

Aside from distinct binding affinities to secondary messengers, PDEs exhibit different tissue specificity and pathway interactions. PDE1, 4, and 7-11 are expressed in the brain; PDE1-4 and 9 in lung; PDE1-3, 5, and 7-8 in heart and muscle (smooth and striated); PDE2-4, 8-9, and

11 in liver; PDE4, 7-9, and 11 in the kidneys; and PDE6 only in eyes. Even within each tissue, there are discrete expression patterns; for example, of the PDEs expressed in the brain, 11A is specific to the hippocampus while 10A is primarily in the striatum. Most PDEs are cytosolic, but

some isoforms are membrane bound, or found in the nucleus or particulate fractions. The wide variety of tissue, pathway, and second messenger specificities indicate their role in diverse pathophysiological functions.

There have been some human studies that suggested an increased risk towards MDD in

Mexican Americans with polymorphisms in Pde9a and Pde11a while those with polymorphisms in Pde1a and Pde11a had better success after antidepressant treatment (Wong et al., 2006).

The expression of PDEs in areas high in 5-HT and DA also suggest a possible link to MDD and its treatment. Specifically PDE1, 4, and 10A can all be found in MSNs of the striatum and each is suggested to play a distinct role in neuronal regulation of mood and cognition (Figure 3).

PDE1

Pde1 is comprised of three genes (a-c) each with multiple splice variants in human tissue. Tissue distribution in humans and rodents are similar with expression in heart, smooth muscle, lungs, and brain (Heckman et al., 2016). Pde1c is primarily found in the olfactory epithelium while 1b is found in the striatum at highest abundance but also in the hippocampus and prefrontal cortex in humans (Bender and Beavo, 2006, Lakics et al., 2010, Kelly et al.,

2014). PDE1A is subcellular while 1B and 1C are cytosolic (Bender and Beavo, 2006,

Heckman et al., 2016). All isoforms preferentially bind cGMP (Km = 0.6-5.9 µM) compared with cAMP (Km = 0.3-124 µM) and have dual substrate specificity (Bender and Beavo, 2006).

PDE1B is abundant in the striatum and regions of dopaminergic enrichment, and it is localized to MSN that are dopamine- and cAMP-regulated neuronal phosphoprotein (DARPP-32) positive, suggesting PDE1B is involved in dopamine regulation (Bollen and Prickaerts, 2012, Heckman et al., 2016).

Vinpocetine®, a PDE1 inhibitor, has been tested in rats for the treatment of fetal alcohol exposure effects and resulted in an improvement in memory retrieval and cognition, increased

LTP, and improved structure of dendritic spines (Filgueiras et al., 2010). More recent research on vinpocetine recapitulated its nootropic properties in a rat model of Huntington’s disease (HD), models using 3-nitropropionic acid exposure. Vinpocetine was capable of reversing body weight, locomotor, grip strength, and cognitive impairments (Gupta and Sharma, 2014).

Biochemically, it reduced striatal oxidative stress, nitrosative stress, acetylcholinesterase activity, inflammation, and mitochondrial dysfunction (Gupta and Sharma, 2014). A different study focused on Parkinson’s disease (PD) using 1-methyl-4-phenyl-1,2,3,6-tretahydropyridine

(MPTP) induced deficits in rats (Sharma and Deshmukh, 2015). Again, vinpocetine ameliorated movement deficits, oxidative and nitrosative stress, and restored cAMP/cGMP and DA signaling

(Sharma and Deshmukh, 2015). Some of these mechanisms have been implicated in MDD suggesting PDE1 inhibitors may have antidepressant effects.

PDE1 inhibition has been hypothesized to partly act similarly to D1 agonists because of its ability to enhance the phosphorylation of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors, GluR1, stimulate glutamatergic frontostriatal signaling (Heckman et al.,

2016). Song et al. showed cross-regulation of PDE1 and PDE2 on GluR1. By decreasing

PDE1, cGMP is elevated that, in turn, stimulates PDE2 and PDE2 facilitates GluR1 at the surface of D1neurons (Song et al., 2016).

PDE1 is characterized as having ‘on-demand’ regulated action. The ‘on demand’ quality of PDE1 driven by Ca2+/calmodulin regulation makes PDE1 a unique target for therapeutic intervention. PDE1 is implicated in mood, movement, and cognitive disorders by genetic and pharmacological studies showing improvements following PDE1 downregulation (Gupta and

Sharma, 2014, Sharma and Deshmukh, 2015, Li et al., 2016, Snyder et al., 2016, Hufgard et al.,

2017). These studies, along with advanced modeling strategies, support PDE1 as a promising drug target.

PDE4

Four genes encode the Pde4 family (a-d) with up to 25 splice variants. PDE4 hydrolyzes cAMP with varying specificity from Km =1.2-10 µM depending on the isoform (Bender and

Beavo, 2006). The subtypes have different expression patterns with 4c primarily in the peripheral nervous system and 4a, 4b, and 4d in the brain (Lakics et al., 2010). Rodent expression closely matches that of humans. Pde4a is decreased in the striatum of aged rats while Pde4d is decreased in cortex and cerebellum (Kelly et al., 2014). PDE4 has been one of the most investigated PDEs with indications in anti-inflammation, antidepressant-like, and reversal of cognitive deficits from depression, schizophrenia, Alzheimer’s disease (AD), and cerebral ischemia (Peng et al., 2014, Wang et al., 2015b). General PDE4 inhibitors were developed and tested in clinical trials for antidepressant treatment, unfortunately these inhibitors, such as rolipram, have significant emetic side effects and therefore cannot be used clinically. The promising nature of PDE4 inhibitors led to the creation of six distinct drug classes: xanthines, catecholdiethers, benzamides, quinozaoline diketones, benzofurans, and aromatic compounds with improved side effect profiles (Peng et al., 2014).

PDE4 is indicated in depression, PD, AD, HD, schizophrenia, anti-inflammation, and substance abuse and the function in the striatum is linked to expression in the indirect dopaminergic pathway (Heckman et al., 2016). PDE4 counteracts hyperdopaminergia by acting like a D2 antagonist or an adenosine A2a agonist (Heckman et al., 2016). This led Heckman et al. to hypothesize that PDE4 is not involved in DA release, but rather inhibition of PDE4 increases dopaminergic tone through an increase in synthesis and metabolism (Heckman et al.,

2016). This may explain the role of PDE4 inhibition in cognitive disorders like AD. RNAi knockdown (KD) of specifically Pde4d was able to block amyloid-β42 (Aβ) induced cAMP, Morris water maze, and novel object recognition deficits (Zhang et al., 2014). These rescues coincided with increased phosphorylated CREB and brain derived neurotrophic factor (BDNF) and

diminished IL-1β, TNF-α, and NK-кB in the hippocampus suggesting long-term KD of Pde4d as a potential treatment in AD (Zhang et al., 2014). The lessor studied Pde4a shows enhanced memory in passive avoidance in knockout (KO) mice but these mice also show an anxiogenic effect in the elevated plus maze (EPM), hole board, light dark, and novelty suppressed feeding tests, and increased corticosterone levels following stress (Hansen et al., 2014), limiting the potential value of this target.

The association between disrupted in schizophrenia 1 (DISC1) and PDE4B has been linked to stress disorders. For example, chronic mild stress resulted in increased hippocampal

Disc1 and Pde4b expression, and chronic treatment with venlafaxine, an antidepressant and a general PDE4 inhibitor, blocked this increase (Zhang et al., 2015). Etazolate, another general

PDE4 inhibitor, rescued immobility in FST and tail suspension test (TST) following chronic unpredictable mild stress in mice (Wang et al., 2014). Wang et al., showed that the reduced immobility time was associated with increased cAMP/pCREB/BDNF/VGF signaling in the hippocampus and prefrontal cortex (Wang et al., 2014). All behavioral and biochemical rescue was abolished by concurrent dosing with a PKA inhibitor, suggesting that general PDE4 inhibition is sufficient for the effect (Wang et al., 2014). Etazolate was also tested for antidepressant-like effects using the olfactory bulbectomy model in mice; etazolate ameliorated changes in FST, sucrose preference, EPM, plasma corticosterone, cAMP/CREB/BDNF signaling, oxidative/nitrosative stress enzymes, and antioxidant enzymes (Jindal et al., 2015a).

The antidepressant-like effects of etazolate were possibly through neuroplasticity changes and suggests a role of oxidative stress reduction in depression treatment. Rolipram was also tested in the olfactory bulbectomy model with similar findings to those of etazolate: normalization of activity, emotionality, and anxiety markers, decreased corticosterone, elevated cAMP, pCREB, and BDNF, diminished oxidative-nitrosative stress markers, and heightened antioxidant enzyme levels (Jindal et al., 2015b). These two olfactory bulbectomy studies strengthen the hypothesis

of PDE4 antidepressant-like activity via control of the hypothalamic-pituitary-adrenal axis, cAMP signaling, and neuroplasticity.

Wang et al. sought to identify the splice variants necessary and sufficient to produce antidepressant-like effects and used RNAi to test their hypothesis. Following KD of Pde4d4 and

4d5 in the prefrontal cortex, mice showed reversal of depressive-like behavior and memory deficits from chronic unpredictable stress (Wang et al., 2015a). The increased cAMP/PKA/CREB and cAMP/ERK1and2/CREB signaling were cited as enhanced neuroplasticity that lead to an antidepressant-like phenotype (Wang et al., 2015a). Effects from

4d4 and 4d5 KD were not additive with concurrent exposure to rolipram suggesting 4d4 and 4d5 are the major splice variants responsible for the antidepressant-like phenotype (Wang et al.,

2015a). The study of general PDE4 inhibitors in depressive-like behavior lead to the investigation of upstream targets such as cylcin-dependent protein kinase-5 (Cdk5). Cdk5 regulates PDE4. When Cdk5 is disrupted in forebrain, ventral striatum, or D1 cells, a reduction of immobility is seen in FST and TST along with antidepressant-like phenotypes in social defeat, baseline, and stress induced sucrose preference test (Plattner et al., 2015). Similar effects were found with Cdk5 peptide inhibitors infused into the ventral striatum (Plattner et al., 2015), further suggesting that regulators of PDE4 have beneficial antidepressant effects. A human randomized double-blind placebo-controlled cross-over study of Zembrin, a general PDE4 inhibitor from the extract of sceletium tortuosum, in healthy subjects improved sleep and mood and enhanced cognitive flexibility and executive function (Chiu et al., 2014). These studies suggest a possible roll for PDE4 in multiple aspects of mood disorders.

PDE4 continues to be one of the mostly studied PDEs with respect to pharmacotherapeutic targets. Genetic and inhibitor based studies showed that PDE4 inhibition produces cognitive, substance abuse, and mood enhancing benefits. Following the failure of rolipram in clinical trials, research has been redirected towards minimizing the emetic effects of

PDE4 inhibition through altered molecular structures, targeting specific isoforms, or indirectly targeting PDE4 through the control of upstream PDE4 regulators.

PDE10A

Pde10 is one gene, a, with four splice variants, 1-4. It is a dual regulator but has a 20- fold higher affinity for cAMP over cGMP (Heckman et al., 2016). Human expression is specific to the caudate, nucleus accumbens, and thymus (Lakics et al., 2010); rodents show similar expression patterns with highest abundance in striatum (Kelly et al., 2014). Within the striatum

Pde10a is found in MSN and localizes to plasma the membranes and dendritic spines (Bender and Beavo, 2006). PDE10A1 and PDE10A3 are cytosolic while PDE10A2 is membrane bound, although it can be translocated to the cytosol (Heckman et al., 2016). This expression pattern suggests that PDE10A may be involved in frontostriatal dysfunctions in striatonigral and striatopallidal pathways. Previous reviews have highlighted that Pde10a mRNA and protein are decreased in the striatum (caudate, putamen, and nucleus accumbens) of HD patients and rodent models (Bender and Beavo, 2006, Bollen and Prickaerts, 2012). This decrease precedes symptom onset (Heckman et al., 2016). Aside from HD, schizophrenic patients were examined for Pde10a mRNA changes but no differences were seen (Heckman et al., 2016).

Polymorphisms in MDD patients were linked to Pde10a variants (Bollen and Prickaerts, 2012).

These associations influenced the development of PDE10A inhibitors: papavarine, MP-10,

THPP-1, and non-selective inhibitors such as ibudilast and tofisopam. These inhibitors attenuate phencyclidine, MK-801, D-amphetamine (AMPH), quinolinic acid, and genetic disruptions in a variety of behaviors (Bender and Beavo, 2006, Heckman et al., 2015, Murthy and Mangot, 2015, Wang et al., 2015b, Heckman et al., 2016).

There are multiple lines of evidence for a role of PDE10A in direct and indirect dopaminergic pathways. Papavarine causes a 6-fold increase in DARPP-32 expression in D2 neurons but only a 2-fold increase in D1 neurons (Heckman et al., 2016). However, others

found that PDE10A inhibition attenuated D1 receptor-stimulated responses (Heckman et al.,

2016). This has led to the hypothesis that PDE10A inhibition results in D1 agonistic and D2 antagonistic effects (Heckman et al., 2016). Despite the positive indications of PDE10A inhibitors there are none marketed. MP-10 underwent clinical trials but failed to have an effect on positive or negative schizophrenia symptoms and caused akathisia and dystonia (Heckman et al., 2015, Heckman et al., 2016).

TP-10, another PDE10A inhibitor, was examined for its role in neuropsychiatric disorders. TP-10 did not alter apomorphine induced deficits in prepulse inhibition of acoustic startle alone but did so in combination with SCH23390; it also blocked deficits induced by quinpirole (Gresack et al., 2014). The antipsychotic effects of TP-10 were only present when

D1 was not activated (Gresack et al., 2014). Aside from antipsychotic treatment, TP-10 is suggested as a treatment for substance abuse, since it reduces alcohol self-administration in stressed, non-stressed, genetically altered to prefer alcohol, alcohol dependent, and non- dependent rats and blocks saccharin self-administration (Logrip et al., 2014). Logrip et al. further showed no alteration in alcohol pharmacokinetics, motor activity, conditioned place aversion, or operant reaction time suggesting differences in substance abuse-specific behaviors

(Logrip et al., 2014). Diminishing a substance abuse phenotype was specific to direct administration of TP-10 in the dorsal lateral striatum and was not seen when the nucleus accumbens was targeted (Logrip et al., 2014).

Aside from inhibitors and ligand targets, the phosphorylation and enzymatic activity at baseline was also investigated. PDE10A at the plasma membrane is in complex with PKA,

NMDA receptors NR1A and B, and post-synaptic density 95 (PSD95) and allows PDE10A to act as a cAMP “gate keeper” (Russwurm et al., 2015). Protein phosphatase 2A was identified as the phosphatase responsible for phosphorylating PDE10A but this does not alter the enzymatic

properties of PD10A, but instead releases it from the complex allowing for cAMP to increase

(Russwurm et al., 2015).

Two unrelated families both presenting with hyperkinetic movement disorders were found to have homozygous mutations in Pde10a (Diggle et al., 2016). These mutations resulted in striatal specific reductions in Pde10a along with recombinant cellular systems (Diggle et al.,

2016). To further study the role of Pde10a in movement disorders Diggle et al. recapitulated one of the mutations in a mouse model that exhibited decreases in striatal Pde10a, motor abnormalities, diminished ability to degrade cAMP, and negligible responses to PDE10A inhibitors (Diggle et al., 2016). This suggests the importance of modulating Pde10a expression in the striatum in the treatment of neuropsychiatric disorders.

Phosphodiesterases in dopaminergic pathways

PDE1s, PDE4s, and PDE10A are expressed in the striatum, in D1 and/or D2 expressing cells. These commonalities indicate that PDE1, 4, and 10A may interact (Figure 2). Nishi et al.

(2008) examined the striatal specific role of PDE4 and PDE10A using genetic, neuroanatomical, and pharmacological techniques and concluded that although expressed in similar areas each protein has a distinct function. PDE4 was found in dopaminergic terminals and MSNs, predominately striatopallidal neurons, whereas PDE10A was found in D1 and D2 MSNs (Nishi et al., 2008). Although PDE4 and PDE10A are both in striatopallidal neurons and function like

D2 antagonists, PDE10A appears to play the primary role while PDE4 regulates tyrosine hydroxylase in dopaminergic terminals (Nishi et al., 2008).

Nishi et al. (2010) also explored the role of PDE1B and PDE10A in MSNs of the striatum concluding that PDE1B is primarily in D1 neurons while PDE10A is in D2 neurons.

Electrophysiological and biochemical studies showed that although PDE10A is present in both

D1 and D2 expressing neurons it has a greater effect in corticostriatal activity and inhibitors

produce antipsychotic effects similar to D2-receptor antagonists (Nishi and Snyder, 2010).

PDE1B has also been localized to DARPP-32 positive MSNs implicating PDE1B in D1 and D2 neurons. Despite the localization in both neuronal types striatal slices from Pde1b KO mice showed potentiation of DARPP-32 and GluR1 phosphorylation following D1 receptor stimulation

(Reed et al., 2002). Pde1b KO mice show a moderate hyperactivity phenotype that is exaggerated following AMPH or MK-801 administration (Reed et al., 2002, Ehrman et al., 2006,

Siuciak et al., 2007). These studies suggest that PDE1B plays a larger role in D1 pathways.

PDE1B, PDE4, and PDE10A function in striatal pathways but it appears that they have distinct roles; however, this does not rule-out some interactions. PDE1B acts in the striatonigral pathway stimulating motor function and may play a role in reward and mood. PDE4 regulates tyrosine hydroxylase at dopaminergic terminals influencing DA synthesis and is in neurological and neuropsychiatric disorders. PDE10A acts in the striatopallidal pathway mimicking D2 antagonists that are used in treating psychotic disorders. Although each PDE has a distinct role in the striatum their presence in overlapping pathways implies at least indirect crosstalk.

Conclusion

MDD is a leading neuropsychiatric disorders. Despite 83.1 billion dollars per year to the

American health care system, current treatments are not sufficient to the problem (Greenberg et al., 2003). Current treatments focus on presynaptic mechanisms by blocking reuptake and take

4-6 weeks to become effective, suggesting that acute increases in monoamine levels are not the primary mechanism behind their effects (Fava and Kendler, 2000). Increased neurogenesis and plasticity are the leading hypotheses for how antidepressants work. These proposed mechanisms mimic developmental processes, therefore, better understanding of striatal and hippocampal brain development may lead to more effective antidepressant treatments. Aside from the lag between treatment initiation and symptom relief only a third of patients find complete relief even after multiple different antidepressants or combination therapies.

Postsynaptic regulation by be targetable for antidepressant treatment. PDEs are second messengers that have been linked to the treatment of other neuropsychiatric disorder through cAMP/CREB/BDNF signaling cascades that increase neurogenesis and plasticity. PDE inhibitors may provide an alternative pathway for the development of greater antidepressant efficacy.

References

Andersen SL (2003) Trajectories of brain development: point of vulnerability or window of

opportunity? Neuroscience & Biobehavioral Reviews 27:3-18.

Béland S-G, Tournier M, Galbaud du Fort G, Crott R, Ducruet T, Pariente A, Moride Y (2011)

Economic impact of nonpersistence with antidepressant treatment in the adult population

of Quebec: a comparative cost-effectiveness approach. Value in Health 14:492-498.

Bender AT, Beavo JA (2006) Cyclic nucleotide phosphodiesterases: molecular regulation to

clinical use. Pharmacological reviews 58:488-520.

Bollen E, Prickaerts J (2012) Phosphodiesterases in neurodegenerative disorders. IUBMB life

64:965-970.

Brezun J, Daszuta A (1999) Depletion in serotonin decreases neurogenesis in the dentate gyrus

and the subventricular zone of adult rats. Neuroscience 89:999-1002.

Cassano P, Fava M (2002) Depression and public health: an overview. Journal of

Psychosomatic Research 53:849-857.

Chiu S, Gericke N, Farina-Woodbury M, Badmaev V, Raheb H, Terpstra K, Antongiorgi J,

Bureau Y, Cernovsky Z, Hou J (2014) Proof-of-concept randomized controlled study of

cognition effects of the proprietary extract Sceletium tortuosum (Zembrin) targeting

phosphodiesterase-4 in cognitively healthy subjects: Implications for Alzheimer’s

Dementia. Evidence-Based Complementary and Alternative Medicine 2014.

Conti M, Beavo J (2007) Biochemistry and physiology of cyclic nucleotide phosphodiesterases:

essential components in cyclic nucleotide signaling. Annu Rev Biochem 76:481-511.

Diggle CP, Rizzo SJS, Popiolek M, Hinttala R, Schülke J-P, Kurian MA, Carr IM, Markham AF,

Bonthron DT, Watson C (2016) Biallelic mutations in PDE10A lead to loss of striatal

PDE10A and a hyperkinetic movement disorder with onset in infancy. The American

Journal of Human Genetics 98:735-743.

Edgar VA, Cremaschi GA, Sterin-Borda L, Genaro AMa (2002) Altered expression of autonomic

neurotransmitter receptors and proliferative responses in lymphocytes from a chronic

mild stress model of depression: effects of fluoxetine. Brain, behavior, and immunity

16:333-350.

Ehrman L, Williams M, Schaefer T, Gudelsky G, Reed T, Fienberg A, Greengard P, Vorhees C

(2006) Phosphodiesterase 1B differentially modulates the effects of methamphetamine

on locomotor activity and spatial learning through DARPP32‐dependent pathways:

evidence from PDE1B‐DARPP32 double‐knockout mice. Genes, Brain and Behavior

5:540-551.

Fava M, Kendler KS (2000) Major depressive disorder. Neuron 28:335-341.

Filgueiras CC, Krahe TE, Medina AE (2010) Phosphodiesterase type 1 inhibition improves

learning in rats exposed to alcohol during the third trimester equivalent of human

gestation. Neuroscience letters 473:202-207.

Gerfen CR, Engber TM, Mahan LC, Susel Z, Chase TN, Monsma FJ, Sibley DR (1990) D1 and

D2 dopamine receptor-regulated gene expression of striatonigral and striatopallidal

neurons. Science 250:1429-1432.

Gourley SL, Wu FJ, Kiraly DD, Ploski JE, Kedves AT, Duman RS, Taylor JR (2008) Regionally

specific regulation of ERK MAP kinase in a model of antidepressant-sensitive chronic

depression. Biological psychiatry 63:353-359.

Graybiel AM (1990) Neurotransmitters and neuromodulators in the basal ganglia. Trends in

neurosciences 13:244-254.

Greenberg PE, Kessler RC, Birnbaum HG, Leong SA, Lowe SW, Berglund PA, Corey-Lisle PK

(2003) The economic burden of depression in the United States: how did it change

between 1990 and 2000? Journal of clinical psychiatry 64:1465-1475.

Gresack JE, Seymour PA, Schmidt CJ, Risbrough VB (2014) Inhibition of phosphodiesterase

10A has differential effects on dopamine D1 and D2 receptor modulation of sensorimotor

gating. Psychopharmacology 231:2189-2197.

Guan L, Jia N, Zhao X, Zhang X, Tang G, Yang L, Sun H, Wang D, Su Q, Song Q (2013) The

involvement of ERK/CREB/Bcl-2 in depression-like behavior in prenatally stressed

offspring rats. Brain research bulletin 99:1-8.

Gupta S, Sharma B (2014) Protective effects of phosphodiesterase-1 (PDE 1) and ATP

sensitive potassium (K ATP) channel modulators against 3-nitropropionic acid induced

s disease. European ׳behavioral and biochemical toxicities in experimental Huntington

journal of pharmacology 732:111-122.

Hansen RT, Conti M, Zhang H-T (2014) Mice deficient in phosphodiesterase-4A display

anxiogenic-like behavior. Psychopharmacology 231:2941-2954.

Heckman P, Wouters C, Prickaerts J (2015) Phosphodiesterase inhibitors as a target for

cognition enhancement in aging and Alzheimer’s disease: a translational overview.

Current pharmaceutical design 21:317-331.

Heckman PR, van Duinen MA, Bollen EP, Nishi A, Wennogle L, Blokland A, Prickaerts J (2016)

Phosphodiesterase inhibition and regulation of dopaminergic frontal and striatal

functioning: clinical implications. International Journal of Neuropsychopharmacology

pyw030.

Hennou S, Khalilov I, Diabira D, Ben‐Ari Y, Gozlan H (2002) Early sequential formation of

functional GABAA and glutamatergic synapses on CA1 interneurons of the rat foetal

hippocampus. European Journal of Neuroscience 16:197-208.

Hill AS, Sahay A, Hen R (2015) Increasing adult hippocampal neurogenesis is sufficient to

reduce anxiety and depression-like behaviors. Neuropsychopharmacology 40:2368-

2378.

Hufgard JR, Williams MT, Skelton MR, Grubisha O, Ferreira FM, Sanger H, Wright ME, Reed-

Kessler TM, Rasmussen K, Duman RS, Vorhees CV (2017) Phosphodiesterase-1b

(Pde1b) knockout mice are resistant to forced swim and tail suspension induced

immobility and show upregulation of Pde10a. Psychopharmacology.

Jacobs B, Van Praag H, Gage F (2000) Adult brain neurogenesis and psychiatry: a novel theory

of depression. Molecular psychiatry 5:262.

Jain M, Armstrong RJ, Barker RA, Rosser AE (2001) Cellular and molecular aspects of striatal

development. Brain research bulletin 55:533-540.

Jindal A, Mahesh R, Bhatt S (2015a) Etazolate, a phosphodiesterase-4 enzyme inhibitor

produces antidepressant-like effects by blocking the behavioral, biochemical,

neurobiological deficits and histological abnormalities in hippocampus region caused by

olfactory bulbectomy. Psychopharmacology 232:623-637.

Jindal A, Mahesh R, Bhatt S (2015b) Type 4 phosphodiesterase enzyme inhibitor, rolipram

rescues behavioral deficits in olfactory bulbectomy models of depression: Involvement of

hypothalamic–pituitary–adrenal axis, cAMP signaling aspects and antioxidant defense

system. Pharmacology Biochemistry and Behavior 132:20-32.

Kelly MP, Adamowicz W, Bove S, Hartman AJ, Mariga A, Pathak G, Reinhart V, Romegialli A,

Kleiman RJ (2014) Select 3′, 5′-cyclic nucleotide phosphodiesterases exhibit altered

expression in the aged rodent brain. Cellular signalling 26:383-397.

Khalaf-Nazzal R, Francis F (2013) Hippocampal development–Old and new findings.

Neuroscience 248:225-242.

Kuehner C (2003) Gender differences in unipolar depression: an update of epidemiological

findings and possible explanations. Acta Psychiatrica Scandinavica 108:163-174.

Kulkarni SK, Dhir A (2009) Current investigational drugs for major depression. Expert opinion on

investigational drugs 18:767-788.

Lakics V, Karran EH, Boess FG (2010) Quantitative comparison of phosphodiesterase mRNA

distribution in human brain and peripheral tissues. Neuropharmacology 59:367-374.

Lauder JM, Krebs H (1976) Effects of p-chlorophenylalanine on time of neuronal origin during

embryo-genesis in the rat. Brain Research 107:638-644.

Li K, Zhou T, Liao L, Yang Z, Wong C, Henn F, Malinow R, Yates JR, Hu H (2013) βCaMKII in

lateral habenula mediates core symptoms of depression. science 341:1016-1020.

Li P, Zheng H, Zhao J, Zhang L, Yao W, Zhu H, Beard JD, Ida K, Lane W, Snell G (2016)

Discovery of potent and selective inhibitors of Phosphodiesterase 1 for the treatment of

cognitive impairment associated with neurodegenerative and neuropsychiatric diseases.

Journal of medicinal chemistry 59:1149-1164.

Logrip ML, Vendruscolo LF, Schlosburg JE, Koob GF, Zorrilla EP (2014) Phosphodiesterase

10A regulates alcohol and saccharin self-administration in rats.

Neuropsychopharmacology 39:1722-1731.

Marazziti D, Dell'Osso MC, Consoli G, Baroni S (2009) Second messenger modulation: a novel

target of future antidepressants? Current medicinal chemistry 16:4679-4690.

Murthy VS, Mangot AG (2015) Psychiatric aspects of phosphodiesterases: An overview. Indian

journal of pharmacology 47:594.

Nestler EJ, Barrot M, DiLeone RJ, Eisch AJ, Gold SJ, Monteggia LM (2002) Neurobiology of

depression. Neuron 34:13-25.

Nishi A, Kuroiwa M, Miller DB, O'Callaghan JP, Bateup HS, Shuto T, Sotogaku N, Fukuda T,

Heintz N, Greengard P (2008) Distinct roles of PDE4 and PDE10A in the regulation of

cAMP/PKA signaling in the striatum. Journal of Neuroscience 28:10460-10471.

Nishi A, Snyder GL (2010) Advanced research on dopamine signaling to develop drugs for the

treatment of mental disorders: biochemical and behavioral profiles of phosphodiesterase

inhibition in dopaminergic neurotransmission. Journal of pharmacological sciences

114:6-16.

Norrholm SD, Ouimet CC (2000) Chronic fluoxetine administration to juvenile rats prevents age-

associated dendritic spine proliferation in hippocampus. Brain research 883:205-215.

O. Harrison HGK, Judith C. Hays, Anedi G. Eme-Akwari, Kenneth I. Pargament, Myleme (2001)

The epidemiology of religious coping: A review of recent literature. International review

of psychiatry 13:86-93.

Papp M, Moryl E (1994) Antidepressant activity of non-competitive and competitive NMDA

receptor antagonists in a chronic mild stress model of depression. European journal of

pharmacology 263:1-7.

Peng S, Sun H, Zhang X, Liu G, Wang G (2014) Effects of selective phosphodiesterases-4

inhibitors on learning and memory: A review of recent research. Cell biochemistry and

biophysics 70:83-85.

Plattner F, Hayashi K, Hernández A, Benavides DR, Tassin TC, Tan C, Day J, Fina MW, Yuen

EY, Yan Z (2015) The role of ventral striatal cAMP signaling in stress-induced behaviors.

Nature neuroscience 18:1094-1100.

Pucilowska J, Puzerey PA, Karlo JC, Galán RF, Landreth GE (2012) Disrupted ERK signaling

during cortical development leads to abnormal progenitor proliferation, neuronal and

network excitability and behavior, modeling human neuro-cardio-facial-cutaneous and

related syndromes. The Journal of Neuroscience 32:8663-8677.

Reed TM, Repaske DR, Snyder GL, Greengard P, Vorhees CV (2002) Phosphodiesterase 1B

knock-out mice exhibit exaggerated locomotor hyperactivity and DARPP-32

phosphorylation in response to dopamine agonists and display impaired spatial learning.

The Journal of neuroscience 22:5188-5197.

Reierson GW, Guo S, Mastronardi C, Licinio J, Wong M-L (2011) cGMP signaling,

phosphodiesterases and major depressive disorder. Current neuropharmacology 9:715.

Russwurm C, Koesling D, Russwurm M (2015) Phosphodiesterase 10A is tethered to a synaptic

signaling complex in striatum. Journal of Biological Chemistry 290:11936-11947.

Samuels IS, Karlo JC, Faruzzi AN, Pickering K, Herrup K, Sweatt JD, Saitta SC, Landreth GE

(2008) Deletion of ERK2 mitogen-activated protein kinase identifies its key roles in

cortical neurogenesis and cognitive function. The Journal of Neuroscience 28:6983-

6995.

Seeman P, Bzowej NH, Guan HC, Bergeron C, Becker LE, Reynolds GP, Bird E, Riederer P,

Jellinger K, Watanabe S (1987) Human brain dopamine receptors in children and aging

adults. Synapse 1:399-404.

Selcher JC, Weeber EJ, Christian J, Nekrasova T, Landreth GE, Sweatt JD (2003) A role for

ERK MAP kinase in physiologic temporal integration in hippocampal area CA1. Learning

& Memory 10:26-39.

Sharma S, Deshmukh R (2015) Vinpocetine attenuates MPTP-induced motor deficit and

biochemical abnormalities in Wistar rats. Neuroscience 286:393-403.

Siuciak JA, McCarthy SA, Chapin DS, Reed T, Vorhees C, Repaske D (2007) Behavioral and

neurochemical characterization of mice deficient in the phosphodiesterase-1B (PDE1B)

enzyme. Neuropharmacology 53:113-124.

Snyder GL, Prickaerts J, Wadenberg M-L, Zhang L, Zheng H, Yao W, Akkerman S, Zhu H,

Hendrick JP, Vanover KE (2016) Preclinical profile of ITI-214, an inhibitor of

phosphodiesterase 1, for enhancement of memory performance in rats.

Psychopharmacology 1-12.

Song RS, Tolentino R, Sobie EA, Neves-Zaph SR (2016) Cross-regulation of

Phosphodiesterase 1 and Phosphodiesterase 2 Activities Controls Dopamine-mediated

Striatal α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic Acid (AMPA) Receptor

Trafficking. Journal of Biological Chemistry 291:23257-23267.

Teicher MH, Andersen SL, Hostetter JC (1995) Evidence for dopamine receptor pruning

between adolescence and adulthood in striatum but not nucleus accumbens.

Developmental Brain Research 89:167-172.

Trivedi MH, Rush AJ, Wisniewski SR, Nierenberg AA, Warden D, Ritz L, Norquist G, Howland

RH, Lebowitz B, McGrath PJ (2006) Evaluation of outcomes with citalopram for

depression using measurement-based care in STAR* D: implications for clinical practice.

American journal of Psychiatry 163:28-40.

Tyzio R, Represa A, Jorquera I, Ben-Ari Y, Gozlan H, Aniksztejn L (1999) The establishment of

GABAergic and glutamatergic synapses on CA1 pyramidal neurons is sequential and

correlates with the development of the apical dendrite. Journal of Neuroscience

19:10372-10382.

Vos T, Flaxman AD, Naghavi M, Lozano R, Michaud C, Ezzati M, Shibuya K, Salomon JA,

Abdalla S, Aboyans V (2013) Years lived with disability (YLDs) for 1160 sequelae of 289

diseases and injuries 1990–2010: a systematic analysis for the Global Burden of

Disease Study 2010. The Lancet 380:2163-2196.

Wang C, Zhang J, Lu Y, Lin P, Pan T, Zhao X, Liu A, Wang Q, Zhou W, Zhang H-T (2014)

Antidepressant-like effects of the phosphodiesterase-4 inhibitor etazolate and

phosphodiesterase-5 inhibitor sildenafil via cyclic AMP or cyclic GMP signaling in mice.

Metabolic brain disease 29:673-682.

Wang Z-Z, Yang W-X, Zhang Y, Zhao N, Zhang Y-Z, Liu Y-Q, Xu Y, Wilson SP, O'Donnell JM,

Zhang H-T (2015a) Phosphodiesterase-4D knock-down in the prefrontal cortex alleviates

chronic unpredictable stress-induced depressive-like behaviors and memory deficits in

mice. Scientific reports 5:11332.

Wang Z-Z, Zhang Y, Zhang H-T, Li Y-F (2015b) Phosphodiesterase: an interface connecting

cognitive deficits to neuropsychiatric and neurodegenerative diseases. Current

pharmaceutical design 21:303-316.

Whitaker-Azmitia PM (2001) Serotonin and brain development: role in human developmental

diseases. Brain research bulletin 56:479-485.

Whittle S, Lichter R, Dennison M, Vijayakumar N, Schwartz O, Byrne ML, Simmons JG, Yücel

M, Pantelis C, McGorry P (2014) Structural brain development and depression onset

during adolescence: a prospective longitudinal study. American Journal of Psychiatry

171:564-571.

Wong M-L, Whelan F, Deloukas P, Whittaker P, Delgado M, Cantor RM, McCann SM, Licinio J

(2006) Phosphodiesterase genes are associated with susceptibility to major depression

and antidepressant treatment response. Proceedings of the National Academy of

Sciences 103:15124-15129.

Zhang C, Cheng Y, Wang H, Wang C, Wilson SP, Xu J, Zhang H-T (2014) RNA interference-

mediated knockdown of long-form phosphodiesterase-4D (PDE4D) enzyme reverses

amyloid-β42-induced memory deficits in mice. Journal of Alzheimer's Disease 38:269-

280.

Zhang X, Li X, Li M, Ren J, Yun K, An Y, Lin L, Zhang H (2015) Venlafaxine increases cell

proliferation and regulates DISC1, PDE4B and NMDA receptor 2B expression in the

hippocampus in chronic mild stress mice. European journal of pharmacology 755:58-65.

Tables

Table 1. Classes of antidepressants and examples of those currently available Class Examples Selective serotonin reuptake inhibitors (SSRI) Fluoxetine Norepinephrine reuptake inhibitors (NRI) Atomoxetine Serotonin-norepinephrine reuptake inhibitors (SNRI) Venlafaxine Norepinephrine-dopamine reuptake inhibitors (NDRI) Bupropion Noradrenergic-specific serotonergic antidepressants (NSSA) NE-5HT Tricyclic antidepressants (TCA) Imipramine Monoamine oxidase inhibitors (MAO-I) Hydrazines

Table 2. PDE enzyme family's selectivity, localization, and function

Km (umol/L) Family Subtype Isoform Substrate Tissue Specificity Indications Inhibitors cAMP/cGMP Vascular smooth A1 73-120 A muscle contraction and 2.6-5 A2 sperm function Brain, heart, lung, narcardipine, nimodipine, 1 cAMP

Cardiac and vascular A . Adipose tissue, heart, 0.02-0.15 smooth muscle control Cilostamide, cilostazol, 3 cAMP>cGMP inflammatory cells, 0.18 Cell cycle regulation milrinone B . liver, lung, platelets and proliferation

A1A A A4B Neurological disorders A5 B1

B2A Immune response and B Brain, inflammatory MK-0952, MEM 1018, MEM 4 B2B cAMP 2.9-10 cells, liver, lung, neurological disorders 1091, Ro 20-1724, kidney, sertoli cells roflumilast, rolipram B3 C . Unknown D1 D D2 Neurological disorders D4-9 Heart, lung, platelets, Vascular smooth Sildenafil, tadalafil, 5 . . cGMP 1-6.2 vascular smooth muscle contraction vardenafil, UK-343664 muscle A . photoresponse signal DMPPO, E4021, Sildenafil, 6 B . cGMP 15-17 Photoreceptor transduction Zaprinast C .

A1 Brain, heart, kidney, A pancreas, skeletal 7 A2 cAMP 0.1-0.2 T-cell activation BRL 50481, ICI242 muscle, T B . lymphocytes

A1 Brain, eye, heart, A kidney, liver, ovary, T-cell activation and 8 A2 cAMP 0.04-0.06 skeletal muscle, sperm function B . testes, T lymphocytes A1 A Brain, heart, liver, 9 A2 cGMP 0.17-0.39 Neurological disorders Bay 73-6691, PF-04447943 lung, kidney B . 0.26 10 A . cAMP>cGMP Brain, testes Neurological disorders MP-10, papaverine, THPP-1 7.2 Hippocampus, liver, kidney, pituitary and 1.04-5.7 11 A . cAMP

Figures

Figure 1. (Image from Lundbeck Institute Image Library). Mechanism of action of tricyclic antidepressants. Tricyclic antidepressants were discovered through antihistamine research and showed efficacy in MDD treatment. The acute mechanism blocks 5-HT and noradrenaline reuptake transporters and also blocks postsynaptic histamine and acetylcholine receptors. By blocking presynaptic reuptake transporters 5-HT and noradrenaline is more readily taken up by receptors on in the postsynaptic membrane leading to increase neurotransmission.

Figure 2. Image from (Heckman et al., 2016). The role of PDE1B, PDE4, and PDE10A in dopaminergic signaling. PDE1B, PDE4, and PDE10A are all found in both D1 and D2 MSN of the striatum. Each PDE has a distinct proposed function in striatal dopaminergic neurons.

PDE1B functions primarily in D1 neurons, PDE4 regulates tyrosine hydroxylase activity in dopaminergic terminals, and PDE10A functions in D2 neurons. The presence of these PDEs in multiple neuronal population also implicates redundant mechanisms.

CHAPTER 2

Phosphodiesterase-1b (Pde1b) knockout mice are resistant to forced swim and tail

suspension test induced immobility and show upregulation of Pde10a*

Jillian R. Hufgard1, Michael T. Williams1, Matthew R. Skelton1, Olivera Grubisha2, Filipa M.

Ferreira2, Helen Sanger2, Mary E. Wright3, Tracy M. Reed-Kessler3, Kurt Rasmussen4, Ronald

S. Duman5, and Charles V. Vorhees1*

1Division of Neurology, Dept. of Pediatrics, Cincinnati Children’s Research Foundation and

University of Cincinnati College of Medicine, Cincinnati, OH, USA

2Neuroscience Research Division, Lilly Research Centre, Eli Lilly & Co. Ltd., Windlesham,

Surrey, UK

3Department of Biology, Mount Saint Joseph University, Cincinnati, OH 45233

4Lilly Research Laboratories, Lilly Corporate Center, Indianapolis, IN 46285

5Dept. of Psychiatry, Yale University School of Medicine, 34 Park St., New Haven, CT 06519-

1109.

*Published in Psychopharmacology (2017)

Abstract

Rationale

Major depressive disorder is a leading cause of suicide and disability. Despite this, current antidepressants provide insufficient efficacy in more than 60% of patients. Most current antidepressants are presynaptic reuptake inhibitors; postsynaptic signal regulation has not received as much attention as potential treatment targets.

Objectives

We examined the effects of disruption of the postsynaptic cyclic nucleotide hydrolyzing enzyme, phosphodiesterase (PDE) 1b, on depressive-like behavior and the effects on PDE1B protein in wild-type (WT) mice following stress.

Methods

Littermate knockout (KO) and WT mice were tested in locomotor activity, tail suspension (TST), and forced swim tests (FST). FST was also used to compare the effects of two antidepressants, fluoxetine and bupropion, in KO versus WT mice. Messenger RNA (mRNA) expression changes were also determined. WT mice underwent acute or chronic stress and markers of stress and PDE1B expression were examined.

Results

Pde1b KO mice exhibited decreased TST and FST immobility. When treated with antidepressants, both WT and KO mice showed decreased FST immobility and the effect was of additive in KO mice. Mice lacking Pde1b had increased striatal Pde10a mRNA expression. In

WT mice, acute and chronic stress upregulated PDE1B expression while PDE10A expression was downregulated after chronic but not acute stress.

Conclusions

PDE1B is a potential therapeutic target for depression treatment because of the antidepressant- like phenotype seen in Pde1b KO mice.

Introduction

Depression is a leading cause of disability, with a lifetime prevalence of 16% (Kennedy,

2013). The Centers for Disease Control and Prevention report that two-thirds of suicides are depression-related (Cassano and Fava, 2002). In the United States, 83.1 billion dollars is spent annually treating depression, yet current treatments are often not effective (Greenberg et al.,

2003).

Most antidepressants target presynaptic neurotransmitter reuptake transporters; postsynaptic targets have received less attention. A potential postsynaptic site for modulating neuronal activity is through influencing the duration of action of second messengers (cyclic adenosine monophosphate (cAMP) and cyclic guanosine monophosphate (cGMP)). Increased levels of cGMP have been associated with antidepressant efficacy by increasing neuronal activity (Reierson et al., 2011). This may contribute to secondary effects, such as promoting progenitor cell proliferation in the subventricular and subgranular zones (, Reierson et al., 2011).

Phosphodiesterases (PDEs) hydrolyze the phosphodiester bond of cAMP and/or cGMP.

There are 11 PDE families composed of 21 isoforms each with a different specificity for cAMP, cGMP, or both. Most PDEs have distinct tissue distributions (Maurice et al., 2014). The rate of hydrolysis determines the duration of cyclic nucleotide signaling on downstream effectors such as protein kinase A (PKA), protein kinase G, exchange protein activated by cAMP, and cyclic nucleotide gated channels (Conti and Beavo, 2007).

Human and animal studies have linked other PDEs (e.g., PDE4) to depression

(O'Donnell and Zhang, 2004). Patients with major depressive disorder have decreased positron emission tomography binding of 11C-(R)-rolipram, a PDE4 inhibitor, (Fujita et al., 2012). Chronic exposure to antidepressants, including rolipram, increase levels of brain-derived neurotropic factor and neurogenesis through the activation of the PKA and phosphorylated cAMP-response binding protein (Duman et al., 1999). Rodents treated with etazolate, another PDE4 inhibitor,

exhibit antidepressive-like changes on tests of locomotor activity, tail suspension test (TST), and forced swim test (FST); it is also effective at blocking the induction of depressive-like behaviors caused by chronic mild stress (CMS) (Jindal et al., 2012, Jindal et al., 2013). RNA interference or knockout (KO) of Pde4d increases cAMP signaling and decreases immobility in the TST and the FST in mice and in the FST in rats (Zhang et al., 2002, Schaefer et al., 2012, Wang et al.,

2013). In humans, PDE4 inhibitors have antidepressant effects, however they also cause unacceptable gastrointestinal side-effects (Hansen and Zhang, 2015).

An alternate PDE target for depression is PDE1. Vinpocetine, a PDE1 inhibitor, produced enhancement of long-term potentiation (LTP) and increased dendritic spine density in rats, suggesting that PDE1 inhibitors have neurotrophic effects (Filgueiras et al., 2010). There are three PDE1 subtypes: A, B, and C. PDE1A is in brain, heart, lung, and testis and is involved in regulating vascular smooth muscle (Kim et al., 2001). PDE1C is found in brain, heart, and testis and promotes arterial smooth muscle cell proliferation and down-regulation of glucose- induced insulin secretion (Han et al., 1999, Rybalkin et al., 2002). PDE1 is a dual substrate for cAMP and cGMP and is found in areas rich in dopamine (DA) (Essayan, 2001), including the caudate-putamen, nucleus accumbens, dentate gyrus, and substania nigra, areas linked to mood and other functions (Polli and Kincaid, 1994, Lakics et al., 2010). The described neurotrophic effects of PDE1 inhibitors and the localization of PDE1B suggests it might be promising in relation to depression. We created a constitutive Pde1b KO mouse (Reed, 2000).

These mice exhibit minor increases in locomotor activity (Reed, 2000), differential responses to stimulants, but in one report, no change in FST behavior (Siuciak et al., 2007). However, in the latter study the mice were on a mixed background, whereas our mice were back-crossed 10 generations. Previously, we crossed Pde1b KO mice with Darpp32 KO mice [dopamine and cyclic-adenosine 5’-phosphate (cAMP)-regulated phosphoprotein, Mr 32 kDa that plays a role in dopaminergic and serotonergic pathways]. Pde1b-Darpp32 double KO (dKO) mice exhibited increased DA turnover in striatum (Ehrman et al., 2006) compared with single KO and WT mice

(Fienberg et al., 1998, Svenningsson et al., 2000, Reed et al., 2002, Svenningsson et al., 2002,

Svenningsson et al., 2003, Svenningsson et al., 2004, Ehrman et al., 2006). These data suggest that PDE1B may be involved in DA signaling, and DA has been implicated in depression (Chaudhury et al., 2013). Accordingly, we hypothesized that Pde1b disruption would result in a stress/depressive-resistant phenotype.

Methods

Animals and Husbandry

Mice used for experiment 1 were congenic C57BL/6N KO mice bred in house from heterozygous (Pde1b+/- x Pde1b+/-) parents to obtain litters containing WT, KO, and heterozygous littermates (Reed et al., 2002). Mice were tested as adults (postnatal day (P) 60 or later) with not more than one mouse per genotype per litter used where possible to control for litter effects. Offspring were housed 2-4 per cage after weaning. All mice were housed in polysulfone cages in a pathogen free vivarium using Modular Animal Caging System

(Alternative Design, Siloam Spring, AR) with HEPA filtered air (Alternative Design, Siloam

Spring, AR) at 30 air changes/h. Water was provided ad libitum using an automated reverse- osmosis filtering system (SE Lab Group, Napa, CA). Cages had ad libitum food, corncob bedding, and cotton nest material. Mice were maintained on a 14 h light-10 h dark cycle (lights on at 600 h) that is standard in our institution’s vivarium. Protocols were approved by the

Institutional Animal Care and Use Committee. The vivarium is accredited by AAALAC

International. Wild-type C57BL/6J male mice used in experiment 2 were purchased from

Jackson Laboratories, randomly assigned to treatment groups, given one week to acclimate before experiments, and housed four per cage. All behavioral testing was done blind to the genotype, and all behavioral testing was done in the Animal Behavioral Core at Cincinnati

Children’s with the exception of those mice tested at Mount St. Joseph University (see below).

Sample sizes for each experiment are given in figure legends and Table 1.

Experiment 1: KO phenotype and mRNA Pde isoform expression

Reverse Transcription-qPCR

RNA was isolated from the striatum and cerebellum of 4 KO and 5 WT mice using the

RNeasy kit (Qiagen) according to manufacturer’s instructions. The striatum was chosen because it is the region of highest Pde1b expression; the cerebellum was chosen as a negative control region. The RNA was treated with TURBO DNase (Ambion), quantified by Nanodrop

(Thermo Scientific), and integrity measured on an Agilent 2100 Bioanalyzer using an RNA Nano

6000 Labchip (Agilent). The RNA integrity number ranged from 8.3 to 9.5. Reverse transcription (RT) reactions were performed using 1 x reaction buffer, 2.5 mM MgCl2, 1 µg of

RNA template, 2.5 μM random hexamers, 0.25 mM of each dNTP, 40 U RNase inhibitor, 150 U

MMLV-RT (Applied Biosystems) in a final volume of 100 μL. Reactions lacking the RT enzyme

(RT-) were used as negative controls. Reactions were carried out in a PCR machine using the following program: 10 min at 25 oC, 60 min at 37 oC, and 5 min at 75 oC. Quantitative PCR

(qPCR) contained 80 ng of cDNA, 300 nM of each primer (forward and reverse), and 1x SYBR

Green Master Mix (Qiagen) in a 40 µL volume. Four 5 µL aliquots of the mix were placed in a

384-well plate and the qPCR was performed on an ABI Prism 7900HT (Applied Biosystems) using the following cycling conditions: 50 oC for 2 min, 95 oC for 10 min, and 40 cycles at 95 oC for 15 s and 60 oC for 1 min. Primers were synthesized by Eurofins Genomics (Ebersberg,

Germany) and selected for this study based on primer efficiency, empirically determined to be

95 - 100%. Mouse primer sequences are listed in Table 2. Negative controls included qPCR with RT- samples or in the absence of template. Ct values were determined by the SDS 2.4 software after manually setting the threshold to 0.5. The denaturation curve showed a single peak, representative of a single PCR product. The average Ct values from quadruplicate repeats were calculated. These were then averaged with values obtained from 2 independent qPCR experiments. Changes in Pde mRNA levels were measured with the ΔΔCt method, using

PSMB2 (proteasome subunit beta type 2) as the housekeeping reference and the Pde1b WT striatum sample as calibrator (set at 100%).

Open-Field Locomotor Activity

One set of mice was given the following tests: Open-field, TST, and FST with sample sizes of 8-25 mice per genotype. Activity was assessed in 40 x 40 cm automated locomotor activity chambers (PAS System, San Diego Instruments, San Diego, CA) as described

(Hautman et al., 2014). Mice were placed in test chambers for 1 h, and data were collected every 5 min. The total number of infrared beam interruptions was analyzed.

Tail Suspension Test

TST followed the method of Cryan et al. (Cryan et al., 2005a). The apparatus allowed the mouse’s tail to be inserted through a hole in a transparent horizontal acrylic plate mounted on four legs. The tail was pulled snuggly against the underneath surface so that no space remained between the base of the tail and the plate. The test was scored manually in 1 min intervals during the 5 min test. Immobility time and latency to the first immobile event were scored. Immobility was defined as the absence of movement except minor paw or nose movements.

Forced Swim Test

Mice were placed in a transparent glass cylindrical vessel 10 cm in diameter (i.d.) and 25 cm tall filled to a depth of 6 cm with 22 ± 1 ºC water. Two procedures were used. The first group was given a single 6 min trial with minutes 2-6 scored for immobility. Later groups were tested using the 2 day method (Porsolt et al., 1979, Cryan et al., 2005b). On day 1, mice were placed in the vessel for 15 min. On day 2, mice were given a second trial for 5 min and scored for immobility, latency to immobility, and active swimming. Immobility was defined as minimal movement sufficient to keep the mouse’s nose above water. The 2 day procedure was used on two sets of mice at two separate institutions to verify the phenotype: Cincinnati Children’s

Research Foundation (Fig. 2D) and Mount St. Joseph University (Fig. 2E).

Antidepressant Treatment

A different set of mice was used for the antidepressant experiment, and these mice also received the FST with sample sizes of 8-14 per group. Antidepressant effects in WT and Pde1b

KO mice were assessed using the FST. Different antidepressants show efficacy with different doses and dosing regimens, therefore, we used procedures previously found to be effective.

We chose one SSRI (fluoxetine) and one non-SSRI (bupropion) for comparison. Drugs were given subcutaneously in a volume of 10 mL/kg. Fluoxetine (20 mg/kg; Sigma-Aldrich, St. Louis,

MO) was administered three times at 23.5, 5, and 1 h prior to day-2 of FST as per (Mason et al.,

2009). Bupropion (20 mg/kg; Toronto Research Chemicals, Toronto, Ontario, Canada) was administered 30 min before day-2 of FST as per (Dhir and Kulkarni, 2008).

Experiment 2: Effects of stress on PDE1B protein

Acute Stress

Adult male WT mice were rehoused four times in random combinations to normalize the gut microbiota between treatment groups (Stappenbeck and Virgin, 2016), and cages were randomly assigned to stress or non-stress groups (8 mice/group for corticosterone and 12 mice/group for Western blots). Mice had a submandibular blood sample taken 48 h before day-

1 of FST or handling in the no-stress group. Mice in both groups had blood drawn after day-1

FST or handling. FST mice were given day-2 of the FST and both groups sacrificed 24 h later.

Mice were decapitated, blood collected, and brain collected (brain was cut along the midline and the striatum, cerebellum, or whole brain collected) and frozen at –80 ºC.

Chronic Mild Stress

Adult male WT mice were housed 4/cage for one week prior to the start of the chronic mild stress (CMS) exposure. Each cage was randomly assigned to CMS or no-stress groups

(12/group). Mice were weighed every third day. No-stress mice were handled daily to control for being removed from cages. Submandibular blood samples were taken before the first and after the last stressor. For the next three days all mice were tested in the morning in TST and in

the afternoon in FST, albeit no differences were detected in the behavior of the stressed and non-stressed groups on any day. Twenty-four hours after the final FST mice were euthanized and blood, thymus, spleen, adrenal, and brain were collected. CMS used two stressors per day for 21 days (Table 3) (Castaneda et al., 2011). Stressors were: Tilted cage (tilted 45° with no bedding); restraint: mice placed in 50 mL conical centrifuge tubes with holes for air circulation; shaker: mice were restrained in 50 mL tubes attached to a shaker plate and rotated at 200 rpm; predator: mice were placed in 50 mL tubes and placed in an F344 male rat’s cage; standing water: 500 mL of water in cage; dirty rat cage: mice placed in a soiled rat cage; grid floor: mice housed in a cage with a wire floor; hypoxia: mice placed in a hypoxia chamber (Biospherix

Lacona, NY) and exposed to 8% oxygen and 92% nitrogen; cold: mice were placed in boxes in a 4 °C room.

Corticosterone Assay

Collected blood was placed in micro-centrifuge tubes with 2% ethylene diamine tetra acetic acid as anticoagulant. Samples were stored on ice and later spun at 610 RCF for 15 min at 4 °C. Plasma was transferred to clean micro-centrifuge tubes and stored at -80 °C. Plasma samples were assayed using a single lot of Enzo Life Sciences® Corticosterone EIA Kits and run in duplicate following the manufacturer’s instructions.

Western Blot

Frozen brain tissue was homogenized in radioimmuno-precipitation assay buffer with protease inhibitors. Protein was quantified using the BCATM Protein Assay Kit and diluted to 3

µg/µL. Western blots were performed using LI-COR Odyssey® procedures. Primary antibodies from Abcam and their dilutions were: rabbit anti-PDE1B C-terminal (Ab170441 or Ab182565) at

1:500 or 1:5000 respectively, rabbit anti-PDE10A (Ab177933) at 1:1000, and mouse anti-actin

(Ab3280) at 1:2000 as a loading control. Odyssey IRDye 680 and 800 secondary antibodies were used at a 1:15,000 dilution. Relative protein levels are quantified using the LI-COR

Odyssey® scanner and Image Studio analysis software that reads fluorescent intensity of the sample normalized to actin.

Data Analysis

Data were analyzed using SAS (v9.3, SAS Institute, Cary, NC), where p ≤ 0.05 was the threshold for significance. To control for litter effects only one mouse per treatment group per genotype per litter was used. T-tests were used when there was only two levels of an independent variable, i.e., genotype (KO vs WT) or stress (stress vs no-stress). In these cases dependent variables were immobility time, protein level, or organ weight. Results from t-tests are presented as ordinary means ± standard error of the mean (SEM). Where there were more than two factors, mixed linear model ANOVAs were used. In these analyses data are presented as least square mean ± SEMs. Two way ANOVAs were used when between subject factors were genotype (KO vs WT) and drug (saline, fluoxetine, or bupropion) or genotype (KO vs HET vs WT) and sex. A three way ANOVA was used when the factors were genotype, gene (Pde

1A, 1B, 1C, 2, 4A, 4B, 4D, 10), and brain region (striatum, cerebellum, or whole brain).

Repeated measure ANOVA was used for body weight, blood samples, and locomotor activity interval. Mixed models used the autoregressive-1 covariance matrix and Kenward-Roger first order adjusted degrees of freedom. Litter was a random factor in ANOVA models.

Results

Experiment 1: KO mice phenotype and mRNA Pde isoform expression

RT-qPCR

Expression of mRNA was assessed by RT-qPCR for eight Pde isoforms in the striatum and cerebellum (control region) of WT and Pde1b KO mice to confirm complete KO of Pde1b and test for compensation by other Pde isoforms. As shown in Fig. 1A, Pde2 and Pde10a expression levels were highest in the striatum, followed by Pde4b and Pde1b. Expression of all

Pde isoforms were low in the cerebellum (Fig. 1B) compared with striatum [F(1,112)=2103.7,

p<0.001]. Pde1b mRNA was abolished in striatum and cerebellum of KO mice [F(1,112)=48.8, p<0.001]. Apart from Pde1b, no differences were seen between WT and Pde1b KO mice for

Pde1a or Pde1c isoforms or for Pde2 or Pde4 mRNA. There was a significant upregulation of

Pde10a in KO mice compared with WT mice in the striatum [F(1,112)=58.1, p<0.001], but this was not observed in the cerebellum.

General characteristics and Locomotor Activity

The appearance and overall behavior of KO mice showed no differences compared with

WT mice. No mortality was observed. KO mice were well groomed and of comparable body weight as WT mice. KO mice were modestly more active in the open-field than WT littermates

(Fig. 2A; [F(1,21.3)=5.1, p<0.05]) but the effect was not overtly observable.

Tail Suspension and Forced Swim Tests

In order to assess acute stress-depressive related behavior, two tests were used: FST and TST. Fig. 2B shows that Pde1b KO mice had reduced immobility in the TST compared with WT mice [t(22)=-4.8, p<0.001]. Similarly, Pde1b KO mice, regardless of FST method, showed reduced immobility using a one-day, Fig. 2C [t(14)= 6.3, p<0.001], or two-day procedure, Fig. 2D [t(13.8)=-3.1, p<0.01], compared with WT mice. This was confirmed by collaborators in which Pde1b KO mice showed reduced immobility compared with heterozygous

(p<0.001) and WT littermates (p<0.001), the heterozygous and WT mice did not differ from one another [main effect: F(2,52)=8.2, p<0.001, Fig. 2E]. In addition, no sex differences (p>0.8) were found, i.e., Pde1b KO females showed similar reductions in immobility as males. This being the case, only males were used in subsequent experiments.

Antidepressant treatment in Pde1b KO mice

In order to determine if Pde1b deletion is efficacious independent of mechanisms of current antidepressants, we tested two drugs from different classes: a selective serotonin reuptake inhibitor, fluoxetine, and a norepinephrine-DA reuptake inhibitor, bupropion. Analysis of time spent immobile in the two day FST following treatment with fluoxetine showed significant

genotype and drug effects [Genotype: F(1,27)=13.1, p<0.01, Drug: F(1,27)=25.0, p<0.001]; the interaction was not significant. KO mice showed a decrease in immobility independent of drug compared with WT littermates (Fig. 3A). Mice given fluoxetine had reduced immobility compared with those given saline. The data suggest that the effects of the KO and fluoxetine were additive but not synergistic.

A similar effect was seen with bupropion. There were significant main effects of genotype and drug on FST immobility [Genotype: F(1,48)= 20.9, p<0.001, Drug: F(1,48)= 102.2, p<0.001] but no interaction. The KO mice had reduced immobility compared with the WT mice regardless of drug treatment: mice treated with bupropion had decreased immobility compared with saline treated mice; hence, the effects were additive but not synergistic (Fig. 3B).

Experiment 2: Effects of acute and chronic stress on PDE1B protein expression

Acute stress

If PDE1B is involved in stress-induced immobility responses, we reasoned that it should change in WT mice subjected to acute stress. Accordingly, we measured corticosterone and

PDE1B following forced swim stress. As shown in Fig. 4A, corticosterone levels did not differ prior to FST. After FST, there was the predicted increase in corticosterone [F(2,30.2)=25.2, p<0.001] compared with non-stressed controls. Corticosterone levels returned to baseline levels 48 h following FST. For PDE1B, there was a stress-induced increase in whole brain

[F(1,61)=6.3, p<0.05] and in striatum [F(1,61)=4.4, p<0.05] but no change in cerebellum as shown in Fig 4B. There was no change in PDE10A (Fig 4C) in these regions.

Chronic mild stress

We next asked, if PDE1B is sensitive to acute stress, would it also be responsive to chronic stress? One common method of inducing chronic stress is CMS. We therefore tested the effect of CMS in WT mice. No differences in body weight were found prior to CMS, however stress decreased body weight after 21 days of CMS [Stress x Day: F(7,139)=11.3, p<0.001] in the CMS stressed mice [23.8 ± 0.6 g] compared with non-stressed WT mice [27.2 ± 0.6 g]. As

expected, CMS-exposed mice had increased corticosterone compared with non-stressed mice

(p<0.001) after the last stressor, but no differences were noted prior to CMS or after 3 days of repeated daily TST and FST testing [CMS X Day: F(2,17.2)=9.5, p<0.01, Fig. 5A]. There was also a decrease in thymus weight in relation to body weight in stressed vs. non-stressed mice

[t(20)=8.0, p<0.001; Thymus: Control=0.128 ± 0.004% Stress=0.060 ± 0.007%] but no change in adrenal (Control=0.024 ± 0.003% Stress=0.022 ± 0.002%) or spleen weight (Control=0.316 ±

0.01% Stress=0.292 ± 0.026%). PDE1B expression was increased after CMS (Fig 5B) whereas PDE10A (Fig 5C) was decreased [PDE1B: t(21)=-3.1, p<0.01, PDE10A: t(19)=2.5, p<0.05]. Hence, acute and chronic stress increased PDE1B whereas acute stress had no effect on PDE10A and chronic stress decreased PDE10A. Since PDE1B is not present in Pde1b KO mice, there could be no PDE1B changes from stress, therefore, we did not test PDE1B KO mice with the CMS procedure.

Discussion

The phenotype of Pde1b deficient mice on a C57BL6/129svj x C57BL/6N F1 mixed hybrid background was reported previously (Reed et al. 2002; Siuciak et al. 2007). Pde1b KO mice showed several effects, including a probe trial deficit in the Morris water maze (Reed et al.,

2002) and modest hyperactivity in an open-field (Siuciak et al., 2007). There were no alterations in conditioned avoidance learning, elevated zero maze, FST, passive avoidance, hot plate, or olfactory orientation (Reed et al., 2002, Siuciak et al., 2007). However, the breeding strategy used in the Siuciak et al. (2007) study was not optimal. By comparison, we used het x het crosses and drew not more than one KO and one WT mouse from any given litter to control genetic background and litter effects, whereas Siuciak et al. used KO x KO and WT x WT mice from separate lines and did not control for litter. In order to ensure that our FST results were sound, we replicated the finding multiple times using different experiments over a period of several years; we also had a collaborator test the mice at another university, and we used the

TST to confirm our FST phenotype. The KO immobility effect is not likely to be attributable to simple activity differences since we previously showed that KO mice are not different in other swimming tests, including the Morris water maze and straight swimming channel. In these tests, KO and WT mice show comparable swim speeds, indicating that KO mice do not differ in swimming even if they do show small open-field activity differences (Reed, 2000, Reed et al.,

2002, Ehrman et al., 2006). This is also in agreement with other studies that show that spontaneous locomotor activity and swimming are not predictive of one another (Cravens,

1974). Moreover, by using the two day FST method we reduced the influence of novelty since the mice habituate to the forced swim environment on day-1 and immobility is assessed on day-

2. For these reasons, we suggest that the present findings are more reliable compared with those of Siuciak et al. (2007). Interestingly though, Siuciak et al. did report increased DA turnover in the striatum, which may be involved in the mechanism behind the TST and FST phenotype (Siuciak et al., 2007).

We show that Pde1b KO mice have complete deletion of DNA and RNA of the catalytic region of the Pde1b gene by Southern and Northern blot analyses (Reed et al., 2002).

Furthermore, Pde1b KO mice have increased DA, DOPAC, and DA utilization compared with

WT mice in striatum and hippocampus and reduced 5-HT in striatum and cerebellum (Siuciak et al., 2007). Here we add a comparison by qPCR of expression of mRNA of Pde isoforms in KO mice relative to WT mice in striatum and cerebellum. The relative expression profile in C57BL mice is similar to that reported for human and BALB/c mice (Lakics et al., 2010, Kelly et al.,

2014). However, exceptions exist in human caudate and nucleus accumbens where PDE1B expression is the highest isoform expressed, whereas it is higher for Pde2, 4b, and 10a in

C57BL mice; BALB/c mice show similar expression patterns of Pde1b and 10a (Lakics et al.,

2010, Kelly et al., 2014). The differences between the expression patterns in different mice may be attributable to genetic strain effects, because of primers targeted to different exon regions, use of different reference genes, or different methods of normalization. Regardless of these

differences, in each case Pde1b and Pde10a are the highest expressing Pde isoforms in the striatum and much lower expression in cerebellum that we used as a control region. As expected, Pde1b mRNA was not detected in KO mice. We found that Pde1b is highly expressed in the striatum of WT mice at levels comparable to Pde4b that is known to be involved with anxiety and depression (Siuciak et al., 2008, Zhang et al., 2008). There were no differences between WT and Pde1b KO mice in the expression of other Pde1 isoforms (Pde1a or 1c). Thus, no other Pde1 isotype showed compensatory changes in the Pde1b KO mice.

There were elevated levels of Pde10a in the striatum of KO mice, a region where Pde10a is expressed at higher levels than Pde1b. Interestingly, Kleiman et al 2010 showed an upregulation in striatal Pde1c after administration of the Pde10a inhibitor TP-10 compared with controls. This suggests an inverse relationship between Pde1 and Pde10a (Kleiman et al.,

2011), but our acute vs. chronic stress data suggest that this relationship depends on the stimulus and its duration.

To date, the phenotype of Pde10a overexpression has not been investigated in relation to depression but Pde10a downregulation has been investigated in relation to schizophrenia.

Genetic or biochemical inhibition of Pde10a in mouse and rat models, respectively, shows a decrease in locomotor activity and in stimulant-induced (phencyclidine, amphetamine, MK-801) hyperactivity (Siuciak et al., 2006b, Schmidt et al., 2008). Other characteristics in Pde10a inhibitor-treated or KO rodents include blockade of apomorphine-induced climbing, inhibited conditioned avoidance (rats and mice), blockade of NMDA antagonist-induced deficits in acoustic startle (rats), improved sensorimotor gating, increased sociability and social odor recognition, reversal of stereotypy, and improved novel object recognition (mice) (Siuciak et al.,

2006a, Siuciak et al., 2006b, Schmidt et al., 2008, Grauer et al., 2009, Höfgen et al., 2010).

Chronic exposure to antipsychotics (haloperidol and clozapine) increases Pde10a (Xu et al.,

2013), suggesting an interaction between Pde10a and the positive symptoms of schizophrenia

(Hebb and Robertson, 2007, Dlaboga et al., 2008, Xu et al., 2011, Natesan et al., 2014).

Siuciak et al. 2006a also showed no differences between Pde10a KO and WT mice in the elevated plus maze.

We also tested FST responses after antidepressant treatment to a prototypical selective serotonin reuptake inhibitor (fluoxetine) and a prototypical norepinephrine-DA reuptake inhibitor

(bupropion). Both WT and KO mice showed reduced immobility from the drugs compared with vehicle-treated mice. Interestingly, the effects of the antidepressants added to the immobility of

Pde1b KO mice. The efficacy of the antidepressants independent of the Pde1b deletion suggests that the mechanism of immobility induced by the drugs and gene deletion are different.

This supports the idea that PDE1B may be a useful target for drug development.

Forced swimming itself increases corticosterone levels, and we used the FST to test for changes in PDE1B protein levels in WT mice. FST caused a significant increase in PDE1B protein in striatum and whole brain. While forced-swim stress causes many changes, the increase in PDE1B is consistent with a role for this enzyme in stress and depression. CMS was used to test chronic stress in WT mice. After 21 days of CMS, mice showed reduced weight gain, elevated plasma corticosterone, and decreased thymus weight all of which are hallmarks of stress; they also showed increased PDE1B levels in whole brain, accompanied by decreased

PDE10A levels. These data suggest that PDE1B responds to acute and chronic stress while

PDE10A responds to chronic but not acute stress. Xu et al. 2013 showed that corticosterone exposure increased Pde2 expression in the hippocampus with a peak 24 h later (Xu et al.,

2013). A similar phenomenon is seen with PDE1B between acute and chronic stress. In this case, the acute stress increases PDE1B expression more significantly than chronic stress, perhaps because of the prolonged exposure to heightened corticosterone.

Our data and those of Kleiman et al. suggest a relationship between Pde1(b and c) and

Pde10a in the striatum (Kleiman et al., 2011). The phenotype of Pde1b KO mice appears specific to the reduction of acute stress-induced depressive-like behavior while the phenotype of

Pde10a KO mice appears to be specific to the reduction of positive symptom-related behaviors

(Höfgen et al., 2010).

We recognize that constitutive KO mice have limitations compared to the use of pharmacological inhibitors. Specifically, constitutive genetic KO models have ablated gene and protein from conception and this can result in compensatory changes during development that can be difficult to estimate. Pharmacological inhibitors can cause changes in neuronal activity when applied, whereas genetic KO models may not induce any change in neuronal function.

Although the current data suggest PDE1B to be a potential target for stress resistant depressive-like effects, further research is necessary to establish this association. Spatially and temporally targeted reductions of PDE1B is another way to estimate the suitability of this model.

Despite the indirect relationship between genetic deletion and a pharmacological treatment, our data suggest that changes to Pde1b may open a new avenue for research into depression.

PDE1 inhibitors have received less consideration for involvement in anxiety and depression; however, there are PDE1 inhibitors in clinical trials for other indications (Li et al., 2016, Snyder et al., 2016). Given that PDE1B is expressed in high abundance in regions known to be involved in anxiety, depression, and other behaviors (Lakics et al., 2010), more research on

PDE1B is warranted.

References

Cassano P, Fava M (2002) Depression and public health: an overview. Journal of

Psychosomatic Research 53: 849-857.

Castaneda T, Nogueiras R, Müller T, Krishna R, Grant E, Jones A, Ottaway N, Ananthakrishnan

G, Pfluger P, Chaudhary N (2011) Decreased glucose tolerance and plasma

adiponectin: resistin ratio in a mouse model of post-traumatic stress disorder.

Diabetologia 54: 900-909.

Chaudhury D, Walsh JJ, Friedman AK, Juarez B, Ku SM, Koo JW, Ferguson D, Tsai H-C,

Pomeranz L, Christoffel DJ (2013) Rapid regulation of depression-related behaviours by

control of midbrain dopamine neurons. Nature 493: 532-536.

Conti M, Beavo J (2007) Biochemistry and physiology of cyclic nucleotide phosphodiesterases:

essential components in cyclic nucleotide signaling. Annu Rev Biochem 76: 481-511.

Cravens RW (1974) Effects of maternal undernutrition on offspring behavior: Incentive value of

a food reward and ability to escape from water. Developmental psychobiology 7: 61-69.

Cryan JF, Mombereau C, Vassout A (2005a) The tail suspension test as a model for assessing

antidepressant activity: review of pharmacological and genetic studies in mice.

Neuroscience & Biobehavioral Reviews 29: 571-625.

Cryan JF, Valentino RJ, Lucki I (2005b) Assessing substrates underlying the behavioral effects

of antidepressants using the modified rat forced swimming test. Neuroscience &

Biobehavioral Reviews 29: 547-569.

Dhir A, Kulkarni S (2008) Possible involvement of sigma‐1 receptors in the anti‐immobility action

of bupropion, a dopamine reuptake inhibitor. Fundamental & clinical pharmacology 22:

387-394.

Dlaboga D, Hajjhussein H, O'Donnell JM (2008) Chronic haloperidol and clozapine produce

different patterns of effects on phosphodiesterase-1B,-4B, and-10A expression in rat

striatum. Neuropharmacology 54: 745-754.

Duman RS, Malberg J, Thome J (1999) Neural plasticity to stress and antidepressant treatment.

Biological psychiatry 46: 1181-1191.

Ehrman L, Williams M, Schaefer T, Gudelsky G, Reed T, Fienberg A, Greengard P, Vorhees C

(2006) Phosphodiesterase 1B differentially modulates the effects of methamphetamine

on locomotor activity and spatial learning through DARPP32‐dependent pathways:

evidence from PDE1B‐DARPP32 double‐knockout mice. Genes, Brain and Behavior 5:

540-551.

Essayan DM (2001) Cyclic nucleotide phosphodiesterases. Journal of Allergy and Clinical

Immunology 108: 671-680.

Fienberg A, Hiroi N, Mermelstein P, Song W-J, Snyder G, Nishi A, Cheramy A, O'callaghan J,

Miller D, Cole D (1998) DARPP-32: regulator of the efficacy of dopaminergic

neurotransmission. Science 281: 838-842.

Filgueiras CC, Krahe TE, Medina AE (2010) Phosphodiesterase type 1 inhibition improves

learning in rats exposed to alcohol during the third trimester equivalent of human

gestation. Neuroscience letters 473: 202-207.

Fujita M, Hines CS, Zoghbi SS, Mallinger AG, Dickstein LP, Liow J-S, Zhang Y, Pike VW,

Drevets WC, Innis RB (2012) Downregulation of Brain Phosphodiesterase Type IV

Measured with 11 C-( R)-Rolipram Positron Emission Tomography in Major Depressive

Disorder. Biological psychiatry 72: 548-554.

Gómez-Pinedo U, Rodrigo R, Cauli O, Herraiz S, Garcia-Verdugo J-M, Pellicer B, Pellicer A,

Felipo V (2010) cGMP modulates stem cells differentiation to neurons in brain in vivo.

Neuroscience 165: 1275-1283.

Grauer SM, Pulito VL, Navarra RL, Kelly MP, Kelley C, Graf R, Langen B, Logue S, Brennan J,

Jiang L (2009) Phosphodiesterase 10A inhibitor activity in preclinical models of the

positive, cognitive, and negative symptoms of schizophrenia. Journal of Pharmacology

and Experimental Therapeutics 331: 574-590.

Greenberg PE, Kessler RC, Birnbaum HG, Leong SA, Lowe SW, Berglund PA, Corey-Lisle PK

(2003) The economic burden of depression in the United States: how did it change

between 1990 and 2000? Journal of clinical psychiatry 64: 1465-1475.

Han P, Werber J, Surana M, Fleischer N, Michaeli T (1999) The calcium/calmodulin-dependent

phosphodiesterase PDE1C down-regulates glucose-induced insulin secretion. Journal of

Biological Chemistry 274: 22337-22344.

Hansen R, Zhang H-T (2015) Phosphodiesterase-4 modulation as a potential therapeutic for

cognitive loss in pathological and non-pathological aging: possibilities and pitfalls.

Current pharmaceutical design 21: 291-302.

Hautman ER, Kokenge AN, Udobi KC, Williams MT, Vorhees CV, Skelton MR (2014) Female

mice heterozygous for creatine transporter deficiency show moderate cognitive deficits.

Journal of inherited metabolic disease 37: 63-68.

Hebb AL, Robertson HA (2007) Role of phosphodiesterases in neurological and psychiatric

disease. Current opinion in pharmacology 7: 86-92.

Höfgen N, Stange H, Schindler R, Lankau -H J, Grunwald C, Langen B, Egerland U, Tremmel P,

Pangalos MN, Marquis KL (2010) Discovery of imidazo [1, 5-a] pyrido [3, 2-e] pyrazines

as a new class of phosphodiesterase 10A inhibitiors. Journal of medicinal chemistry 53:

4399-4411.

Jindal A, Mahesh R, Bhatt S (2013) Etazolate, a phosphodiesterase 4 inhibitor reverses chronic

unpredictable mild stress-induced depression-like behavior and brain oxidative damage.

Pharmacology Biochemistry and Behavior 105: 63-70.

Jindal A, Mahesh R, Gautam B, Bhatt S, Pandey D (2012) Antidepressant-like effect of

etazolate, a cyclic nucleotide phosphodiesterase 4 inhibitor—an approach using rodent

behavioral antidepressant tests battery. European journal of pharmacology 689: 125-

131.

Kelly MP, Adamowicz W, Bove S, Hartman AJ, Mariga A, Pathak G, Reinhart V, Romegialli A,

Kleiman RJ (2014) Select 3′, 5′-cyclic nucleotide phosphodiesterases exhibit altered

expression in the aged rodent brain. Cellular signalling 26: 383-397.

Kennedy SH (2013) A review of antidepressant therapy in primary care: current practices and

future directions. The Primary Care Companion for CNS Disorders 15.

Kim D, Rybalkin SD, Pi X, Wang Y, Zhang C, Munzel T, Beavo JA, Berk BC, Yan C (2001)

Upregulation of phosphodiesterase 1A1 expression is associated with the development

of nitrate tolerance. Circulation 104: 2338-2343.

Kleiman RJ, Kimmel LH, Bove SE, Lanz TA, Harms JF, Romegialli A, Miller KS, Willis A, des

Etages S, Kuhn M (2011) Chronic suppression of phosphodiesterase 10A alters striatal

expression of genes responsible for neurotransmitter synthesis, neurotransmission, and

signaling pathways implicated in Huntington's disease. Journal of Pharmacology and

Experimental Therapeutics 336: 64-76.

Lakics V, Karran EH, Boess FG (2010) Quantitative comparison of phosphodiesterase mRNA

distribution in human brain and peripheral tissues. Neuropharmacology 59: 367-374.

Li P, Zheng H, Zhao J, Zhang L, Yao W, Zhu H, Beard JD, Ida K, Lane W, Snell G (2016)

Discovery of potent and selective inhibitors of Phosphodiesterase 1 for the treatment of

cognitive impairment associated with neurodegenerative and neuropsychiatric diseases.

Journal of medicinal chemistry 59: 1149-1164.

Mason SS, Baker KB, Davis KW, Pogorelov VM, Malbari MM, Ritter R, Wray SP, Gerhardt B,

Lanthorn TH, Savelieva KV (2009) Differential sensitivity to SSRI and tricyclic

antidepressants in juvenile and adult mice of three strains. European journal of

pharmacology 602: 306-315.

Maurice DH, Ke H, Ahmad F, Wang Y, Chung J, Manganiello VC (2014) Advances in targeting

cyclic nucleotide phosphodiesterases. Nature Reviews Drug Discovery 13: 290-314.

Natesan S, Ashworth S, Nielsen J, Tang S, Salinas C, Kealey S, Lauridsen J, Stensbøl T, Gunn

R, Rabiner E (2014) Effect of chronic antipsychotic treatment on striatal

phosphodiesterase 10A levels: a [ 11C] MP-10 PET rodent imaging study

with ex vivo confirmation. Translational psychiatry 4: e376.

O'Donnell JM, Zhang H-T (2004) Antidepressant effects of inhibitors of cAMP

phosphodiesterase (PDE4). Trends in pharmacological sciences 25: 158-163.

Polli JW, Kincaid RL (1994) Expression of a calmodulin-dependent phosphodiesterase isoform

(PDE1B1) correlates with brain regions having extensive dopaminergic innervation. The

Journal of neuroscience 14: 1251-1261.

Porsolt RD, Bertin A, Blavet N, Deniel M, Jalfre M (1979) Immobility induced by forced

swimming in rats: effects of agents which modify central catecholamine and serotonin

activity. European journal of pharmacology 57: 201-210.

Reed TM (2000) Characterization of phosphodiesterase 1B-deficient mice: Role of PDE1B in

central nervous system function. ProQuest Disseratations & Theses.

Reed TM, Repaske DR, Snyder GL, Greengard P, Vorhees CV (2002) Phosphodiesterase 1B

knock-out mice exhibit exaggerated locomotor hyperactivity and DARPP-32

phosphorylation in response to dopamine agonists and display impaired spatial learning.

The Journal of neuroscience 22: 5188-5197.

Reierson GW, Guo S, Mastronardi C, Licinio J, Wong M-L (2011) cGMP signaling,

phosphodiesterases and major depressive disorder. Current neuropharmacology 9: 715.

Rybalkin SD, Rybalkina I, Beavo JA, Bornfeldt KE (2002) Cyclic nucleotide phosphodiesterase

1C promotes human arterial smooth muscle cell proliferation. Circulation research 90:

151-157.

Schaefer T, Braun A, Amos‐Kroohs R, Williams M, Ostertag E, Vorhees C (2012) A new model

of Pde4d deficiency: genetic knock‐down of PDE4D enzyme in rats produces an

antidepressant phenotype without spatial cognitive effects. Genes, Brain and Behavior

11: 614-622.

Schmidt CJ, Chapin DS, Cianfrogna J, Corman ML, Hajos M, Harms JF, Hoffman WE, Lebel

LA, McCarthy SA, Nelson FR (2008) Preclinical characterization of selective

phosphodiesterase 10A inhibitors: a new therapeutic approach to the treatment of

schizophrenia. Journal of Pharmacology and Experimental Therapeutics 325: 681-690.

Siuciak JA, Chapin DS, Harms JF, Lebel LA, McCarthy SA, Chambers L, Shrikhande A, Wong

S, Menniti FS, Schmidt CJ (2006a) Inhibition of the striatum-enriched phosphodiesterase

PDE10A: a novel approach to the treatment of psychosis. Neuropharmacology 51: 386-

396.

Siuciak JA, McCarthy SA, Chapin DS, Fujiwara RA, James LC, Williams RD, Stock JL, McNeish

JD, Strick CA, Menniti FS (2006b) Genetic deletion of the striatum-enriched

phosphodiesterase PDE10A: evidence for altered striatal function. Neuropharmacology

51: 374-385.

Siuciak JA, McCarthy SA, Chapin DS, Martin AN (2008) Behavioral and neurochemical

characterization of mice deficient in the phosphodiesterase-4B (PDE4B) enzyme.

Psychopharmacology 197: 115-126.

Siuciak JA, McCarthy SA, Chapin DS, Reed T, Vorhees C, Repaske D (2007) Behavioral and

neurochemical characterization of mice deficient in the phosphodiesterase-1B (PDE1B)

enzyme. Neuropharmacology 53: 113-124.

Snyder GL, Prickaerts J, Wadenberg M-L, Zhang L, Zheng H, Yao W, Akkerman S, Zhu H,

Hendrick JP, Vanover KE (2016) Preclinical profile of ITI-214, an inhibitor of

phosphodiesterase 1, for enhancement of memory performance in rats.

Psychopharmacology: 1-12.

Stappenbeck TS, Virgin HW (2016) Accounting for reciprocal host–microbiome interactions in

experimental science. Nature 534: 191-199.

Svenningsson P, Lindskog M, Ledent C, Parmentier M, Greengard P, Fredholm BB, Fisone G

(2000) Regulation of the phosphorylation of the dopamine-and cAMP-regulated

phosphoprotein of 32 kDa in vivo by dopamine D1, dopamine D2, and adenosine A2A

receptors. Proceedings of the National Academy of Sciences 97: 1856-1860.

Svenningsson P, Nishi A, Fisone G, Girault J-A, Nairn AC, Greengard P (2004) DARPP-32: an

integrator of neurotransmission. Annu Rev Pharmacol Toxicol 44: 269-296.

Svenningsson P, Tzavara ET, Carruthers R, Rachleff I, Wattler S, Nehls M, McKinzie DL,

Fienberg AA, Nomikos GG, Greengard P (2003) Diverse psychotomimetics act through

a common signaling pathway. Science 302: 1412-1415.

Svenningsson P, Tzavara ET, Liu F, Fienberg AA, Nomikos GG, Greengard P (2002) DARPP-

32 mediates serotonergic neurotransmission in the forebrain. Proceedings of the

National Academy of Sciences 99: 3188-3193.

Wang ZZ, Zhang Y, Liu YQ, Zhao N, Zhang YZ, Yuan L, An L, Li J, Wang XY, Qin JJ (2013)

RNA interference‐mediated phosphodiesterase 4D splice variants knock‐down in the

prefrontal cortex produces antidepressant‐like and cognition‐enhancing effects. British

journal of pharmacology 168: 1001-1014.

Xu Y, Pan J, Chen L, Zhang C, Sun J, Li J, Nguyen L, Nair N, Zhang H, O'Donnell JM (2013)

Phosphodiesterase-2 inhibitor reverses corticosterone-induced neurotoxicity and related

behavioural changes via cGMP/PKG dependent pathway. International Journal of

Neuropsychopharmacology 16: 835-847.

Xu Y, Zhang H-T, O’Donnell JM (2011) Phosphodiesterases in the central nervous system:

implications in mood and cognitive disorders Phosphodiesterases as Drug Targets.

Springer, pp 447-485

Zhang H-T, Huang Y, Jin SC, Frith SA, Suvarna N, Conti M, James M (2002) Antidepressant-

like profile and reduced sensitivity to rolipram in mice deficient in the PDE4D

phosphodiesterase enzyme. Neuropsychopharmacology 27: 587-595.

Zhang H-T, Huang Y, Masood A, Stolinski LR, Li Y, Zhang L, Dlaboga D, Jin SC, Conti M,

O'Donnell JM (2008) Anxiogenic-like behavioral phenotype of mice deficient in

phosphodiesterase 4B (PDE4B). Neuropsychopharmacology 33: 1611-1623.

Tables

Table 1. Description of mice used in experiments Experiment Mice Used Assays Figure Males RT-qPCR 1 1 WT: n=5 KO: n=4 Males Locomotor Activity, 2 A,B,D 1 WT: n=12 TST, and FST (2 Day) KO: n=12 Males FST (1 Day) 2 C 1 WT: n=8 WT: n=8 Males & Females FST (2 Day) 2 E WT: n=15 1 HET: n=25 KO: n=20 Males FST (2 Day) 3 A WT-SAL: n=8 1 WT-FLX: n=8 KO-SAL: n=8 KO-FLX: n=7 Males FST (2 Day) 3 B WT-SAL: n=13 1 WT-BUP: n=14 KO-SAL: n=14 KO-BUP: n=11 WT Males Plasma Corticosterone 4 2 Control: n=8 and Western Blots Acute Stress: n=8 WT Males Plasma Corticosterone 5 2 Control: n=12 and Western Blots Chronic Stress: n=12

Table 2. Primer sequences Gene Primer Sequence (5’->3’) PDE1A GAAGCAAGCGGGGAGCATAG AAAGGCAATTAGGCAAGAAACAGG PDE1B TTATCAATCTCACCAAGGATG GCTGTCTTCATAGTCTTCAC

PDE1C TTGGTTATTGAGATGGTAATGG ATGAGGGATAAGGCTTTCG PDE4A CCGTATCCAGGTCCTCAG

ATGCGATCAGTCCATTGT PDE4B CCAGCAGGGAGACAAAGAAC ACAATGTAGTCAATGAAACCAACC PDE4D GCTTCATAGACTATATCGTTCATC GTCCTCCAAAGTGTCCAAG PDE2 CACATTGCCATGCCTATCTAC CCTTGGTCCAGTGCTCAC PDE10A CACTTTGACATTGGTCCTTTCG TTCTTCACAGACATGATAAAACGG PSMB2 AAATGCGCAATGGATATGAATTG GAAGACAGTCAGCCAGGTT

Table 3. Chronic Variable Stress Paradigm

AM (8:00-12:00) PM (13:00-17:00) Restraint (2 h) Tilted Cage (24 h) Shaker (1 h) Flooded Cage (18 h) Predator-Restraint (30 min) Dirty Rat Cage (18 h) Cold Room (1 h) Grid Floor (24 h) Hypoxia (30 min) Dirty Rat Cage (18 h) Restraint (2 h) Flooded Cage (18 h) Shaker (1 h) Grid Floor (24 h) Hypoxia (30 min) Tilted Cage (24 h) No Test Restraint (2 h) Tilted Cage (24 h) Predator-Restraint (30 min) Dirty Rat Cage (18 h) Hypoxia (30 min) Cold Room (1 h) Predator-Restraint (30 min) Flooded Cage (18 h) Cold Room (1 h) Dirty Rat Cage (18 h) Shaker (1 h) Grid Floor (24 h) Predator-Restraint (30 min) Hypoxia (30 min) No Test Restraint (2 h) Flooded Cage (18 h) Cold Room (1 h) Grid Floor (24 h) Hypoxia (30 min) Tilted Cage (24 h) Cold Room (1 h) Shaker (1 h)

Figures

Figure 1. Pde expression in WT and KO mice striatum and cerebellum. Pde1a, Pde1b, Pde1c,

Pde2, Pde4a, Pde4b, Pde4d, and Pde10a mRNA expression levels were measured by RT- qPCR in the striatum A, and cerebellum B, in WT and KO Pde1b mice. Percent mRNA expression was normalized to Pde1b WT striatum, set at 100%. Data are represented as LS

Mean ± SEM (WT n=5, KO n=4). ***p ≤ 0.001.

Figure 2. Pde1b KO produced resistance to induced immobility in FST and TST compared with

WT littermates. A, KO mice have increased locomotor activity (WT n=11, KO n=12). B, TST

(WT n=12, KO n=12). C, 1-day 6 min FST method (WT n=8, cKO n=8). D, 2 day FST method with 5 min on day 2 (WT n=10, KO n=12). E, 2 day FST method with 5 min on day 2 (WT n=15,

Het n=25, KO n=20). KO mice differ from both the Het and WT littermates. *p ≤ 0.05, **p ≤

0.01, ***p ≤ 0.001.

Figure 3. Pde1b KO mice have a similar antidepressant-like phenotype as currently marketed antidepressants when compared with WT littermates. Note that antidepressant efficacy occurred independent of the genotype. A, FST 2-day method, day for day-2 (5 min) (WT-Saline n=8 WT-Fluoxetine n=8 KO-Saline n=8 KO-Fluoxetine n=7). B, FST 2 day method 5 min (WT-

Saline n=13; WT-Bupropion n=14; KO-Saline n=14; KO-Bupropion n=11). **p ≤ 0.01, ***p ≤

0.001.

Figure 4. PDE1B is elevated by acute stress. A, Plasma corticosterone was collected 48 h prior to stress, right after day-1 FST, and again at sacrifice 24 h later. B, Fluorescent intensity of PDE1B normalized to actin (ab182565, ab3280). C, Fluorescent intensity of PDE10A normalized to actin (ab177933, ab3280). Tissue was collected 24 h after completion of day-2 of the FST. *p ≤ 0.05, ***p ≤ 0.001 (Control n=8, Stress n=8).

Figure 5. PDE1B is elevated while PDE10A is reduced in chronically stressed mice. A, Plasma for corticosterone was collected 24 h prior to stress, after the 21st day of stress, and upon sacrifice. B, Fluorescent intensity of PDE1B normalized to actin (ab170441, ab3280). C,

Fluorescent intensity of PDE10A normalized to actin (ab177933, ab3280). Tissue and blood was collected 24 h after the 3 days of TST and FST in both the stressed and control mice. *p ≤

0.05, **p ≤ 0.01, ***p ≤ 0.001 (Control n=12, Stress n=12).

Acknowledgments

This work was supported by NIH T32 ES007051 and funds from Cincinnati Children’s Hospital

Research Foundation

CHAPTER 3

Phosphodiesterase-1b deletion confers depression-like behavioral resistance separate

from stress-related effects in mice

Jillian R. Hufgard, Michael T. Williams, and Charles V. Vorhees*

Division of Neurology, Dept. of Pediatrics, Cincinnati Children’s Research Foundation and University of Cincinnati College of Medicine, Cincinnati, OH, USA

*Published in Genes, Brain and Behavior (2017)

Abstract

Phosphodiesterase-1b (Pde1b) is highly expressed in striatum, dentate gyrus, CA3, and substantia nigra. In a new Floxed Pde1b x CreCMV global knock-out (KO) mouse model we show an immobility resistance phenotype that recapitulates that found in constitutive Pde1b KO mice. We use this new mouse model to show that the resistance to acute stress-induced depression-like phenotype is not the product of changes in locomotor activity or reactivity to other stressors (learned helplessness, novelty suppressed feeding, or dexamethasone suppression), and is not associated with anhedonia using the sucrose preference test. Using tamoxifen inducible Cre, we show that the immobility-resistant phenotype depends on the age of induction. The effect is present when Pde1b is deleted from conception, P0 or P32, but not if deleted as adults (P60). We also mapped regional brain expression of PDE1B protein and of the Cre driver. These data add to the suggestion that PDE1B may be a target for drug development with therapeutic potential in depression alone or in combination with existing antidepressants.

Introduction

Major depression is the third leading cause of disability (WHO, 2008) and is predicted to be first by 2030 (Mathers et al., 2008). Antidepressants fail to work for many patients creating the need to find new targets. Most antidepressants are presynaptic reuptake inhibitors and extend the action of serotonin (5-HT), dopamine (DA), and/or norepinephrine (NE). An alternative method of extending neurotransmitter action is prolonging postsynaptic signaling.

Postsynaptic receptors activate second messengers (cAMP and cGMP) that in turn active downstream targets such as protein kinase A (PKA), protein kinase G (PKG), exchange protein activated by cAMP, and cyclic nucleotide gated channels (Conti and Beavo, 2007). The duration of signaling is determined by the rate of cyclic nucleotide hydrolysis and is controlled by phosphodiesterases (PDEs). There are 11 PDE families composed of 21 genes. Each has a specific tissue distribution and selectivity (Maurice et al., 2014).

Inhibition of PDE2 reverses impaired cognition caused by corticosterone or chronic stress (Xu et al., 2013, Xu et al., 2015). PDE4 inhibition using drugs, RNAi, or gene deletion decreases immobility in tail suspension (TST) and forced swim tests (FST) (Duman et al., 1999,

Zhang et al., 2002, O'Donnell and Zhang, 2004, Fujita et al., 2012, Jindal et al., 2012, Schaefer et al., 2012, Jindal et al., 2013, Wang et al., 2013). Clinical trials show that PDE4 inhibitors are effective antidepressants but have unacceptable side-effects (Hansen and Zhang, 2015). PDE5 inhibition increases antidepressant efficacy (Katarzyna et al., 2012, Socała et al., 2012b, a), and

PDE10A has been linked to schizophrenia positive symptoms (Hebb and Robertson, 2007,

Dlaboga et al., 2008, Xu et al., 2011, Natesan et al., 2014), and haloperidol or clozapine treatment increases Pde10a expression (Xu et al., 2013). PDE10A inhibition decreases stimulant-induced locomotor activity, blocks apomorphine-induced climbing, inhibits conditioned avoidance, blocks NMDA antagonist-induced deficits in acoustic startle, improves sensorimotor gating, increases sociability and social odor recognition, reverses stereotypy, and improves novel object recognition (Siuciak et al., 2006a, Siuciak et al., 2006b, Schmidt et al., 2008,

Grauer et al., 2009, Höfgen et al., 2010). These data support the potential of PDEs to modulate

CNS function.

Pde1b is abundantly expressed in striatum, an area with involvement in stress and anxiety (Lakics et al., 2010, Kelly et al., 2014, Hufgard et al., 2017). We showed that constitutive Pde1b knock-out (KO) mice have reduced immobility in TST and FST (Hufgard et al., 2017), suggesting its potential as a drug target. These KO mice also show additive effects to two widely used antidepressants: fluoxetine, a selective serotonin reuptake inhibitor, and bupropion, a norepinephrine-dopamine reuptake inhibitor (Hufgard et al., 2017). Constitutive

KO mice have limitations on the types of hypotheses that may be tested because the genes are disrupted from conception and compensatory changes during development cannot be ruled-out.

Therefore, we created a floxed Pde1b mouse that we describe here for the first time. First, floxed mice were crossed to a ubiquitous Cre line to create a global Pde1b KO mouse and these mice were compared with the constitutive Pde1b KO mice (Hufgard et al., 2017) to ensure that the new model had the same phenotype. Next, the global Pde1b KO mice were used to test the role of PDE1B in response to stress. We then used a tamoxifen inducible Cre to determine the age of onset of the Pde1b KO mice immobility-resistant phenotype.

Methods

Animals and Husbandry

Mice were bred in-house in a pathogen free vivarium using Modular Animal Caging

System (Alternative Design, Siloam Spring, AR) with HEPA filtered air at 30 air changes/h.

Food and reverse-osmosis water were available ad libitum; cages had corncob bedding and cotton nest material. Mice were maintained on a 14:10 h light:dark cycle (lights on at 600 h).

Protocols were approved by the Institutional Animal Care and Use Committee. The vivarium is accredited by AAALAC International. At weaning mice were housed with littermates by sex with

2-4 mice per cage. Mice were tested at ~postnatal day (P) 60 with not more than one mouse

per genotype per sex per litter used. Mice were tested by personnel blind to genotype in the

Animal Behavioral Core.

Generation of Pde1b ubiquitous conditional KO mice

Floxed mice were created on a C57BL/6J (albino) background. Flanking loxP sites were inserted between exons 2 and 3 and 5 and 6 with a neo cassette with flanking flippase recognition sites (Fig. 1A). Homologous recombination created a targeted allele with inserted

LoxP sites and the neo cassette. Flippase recombinase was used to remove the neo cassette.

The ubiquitous expressed cytomegalovirus (CMV) Cre driver, B6.C-Tg(CMV-cre)1Cgn/J, was used to create global Pde1b KO mice (KOCMV). Pde1bflox/flox x Cre+/- mice were bred to create

Pde1b-/- x Cre+/- offspring. These were bred with wildtype (WT) C57BL/6J mice to remove Cre.

Thereafter, Pde1b+/- x Pde1b+/- breeding was used to generate WT, heterozygous (HET), and

KOCMV mice containing litters.

Western Blot Analysis

Western blots were used to confirm KO on brain samples of WT and KOCMV mice; actin was used as reference (Fig. 1B). Frozen tissue was homogenized in radioimmuno-precipitation assay buffer (25 mM Tris, 150 mM NaCl, 0.5% sodium deoxychlorate, and 1% Triton X-100 adjusted to 7.2 pH with protease inhibitor (Pierce Biotechnology, Rockford, IL). Protein was quantified using the BCATM Protein Assay Kit (Pierce Biotechnology, Rockford, IL) and diluted to

3 µg/µL. Western blots were performed using LI-COR Odyssey® (LI-COR Biosciences, Lincoln,

NE) procedures. Briefly, 25 µL of sample mixed with Laemmli buffer (Sigma, USA) were loaded in a 12% gel (Bio-Rad Laboratories, Hercules, CA) and run at 200 volts for 35 min in running buffer (25 mM Tris, 192 mM glycine, 0.1% SDS). The gel was then transferred to Immobilon-FL transfer membrane (Millipore, USA) in 1X rapid transfer buffer (AMRESCO, Solon, OH) at 40 V for 1.5 h. Membrane was soaked in Odyssey PBS blocking buffer for 1 h, primary antibody in blocking buffer with 0.2% Tween 20 incubated overnight at 4ºC, secondary antibody incubated in blocking buffer with, 0.2% Tween 20, and 0.01% SDS for 1 h at room temperature.

Antibodies were rabbit anti-PDE1B C-terminal (Ab170441 or Ab182565) at 1:500 or 1:5000, respectively, and mouse anti-actin (Ab3280) at 1:2000 as a loading control. Odyssey IRDye

680 and 800 secondary antibodies were used at a 1:15,000 dilution. Relative protein levels were quantified using the LI-COR Odyssey® scanner and Image Studio software that read the fluorescent intensity of the sample normalized to actin.

Fluorescent Immunohistochemistry

WT and KOCMV mouse brains were perfused with 4% ice cold paraformaldehyde and stored overnight. Following at least 24 h in 30% sucrose they were cut on a freezing microtome at 20 µm and stained with cresyl violet. Free floating slices of prefrontal cortex, striatum, hippocampus, and substantia nigra were stained for PDE1B. Briefly, slices were washed in 1X

PBS three times for 10 min a piece followed by blocking at 4ºC in 2.5% donkey serum, 0.3%

Triton X-100, 1% BSA in 1X PBS for 1 h. Primary antibodies were incubated in the blocking buffer overnight at 4ºC and secondary was incubated for 1 h at room temperature in the dark.

Rabbit anti-PDE1B (Ab182565) and Alexa Fluor 488 were used as primary (1:200) and secondary (1:500) antibodies with DAPI counterstain. Images were taken using a Nikon confocal microscope (Microscopy Imaging Core); all images are 4x magnification.

Open-Field

We used 40 x 40 cm activity chambers (PAS System, San Diego Instruments, San

Diego, CA) (Hautman et al., 2014). Mice were tested for 1 h and data recorded in 5 min intervals.

Tail Suspension Test

The method of Cryan et al. was used (Cryan et al., 2005a). The mouse’s tail was inserted through a hole in a transparent horizontal acrylic plate mounted on four legs and pulled snuggly against the underneath surface so there was no space between the base of the tail and the plate. The test was scored in 1 min intervals for 5 min. Immobility (no movement) and latency to the first immobile episode were scored.

Forced Swim Test

Mice were placed in a glass cylindrical vessel 10 cm in diameter (i.d.), 25 cm tall filled to a depth of 6 cm with 22 ± 1 ºC water and tested using a 2-day method (Porsolt et al., 1979,

Cryan et al., 2005b). On day-1, mice were tested for 15 min and minutes 1-6 scored. On day-2, mice were tested for 5 min. Mice were scored for immobility, latency, and active swimming.

Immobility was minimal movement sufficient to keep afloat.

Home Cage Activity

Mice were tested in standard cages, singly housed (Tang et al., 2002). The home-cage system has metal frames with infrared photodetectors that surround the cage. Data were collected for 48 h (PAS System, San Diego Instruments, San Diego, CA).

Open Swim

We showed that constitutive Pde1b KO mice are more active than WT mice (Hufgard et al., 2017), therefore, swimming activity was tested to determine if open-field activity generalized to swimming activity since the FST relies on swimming. Mice were placed in a 122 cm pool of water for 5 min and tracked using ANY-maze software (Stoelting, Wood Dale, IL) on two consecutive days.

Sucrose Preference

A sucrose preference test was used to assess anhedonia (Pothion et al., 2004). Mice were adapted to drinking from a sipper sack for 7 days. Next, sacks were filled with 2% sucrose for 2 days, then they were returned to water for 5 more days. Mice were then given two sipper sacks, one with water and one with 2% sucrose, balanced for position across mice. Sacks were weighed before and 24 h later. Preference was: (sucrose solution ÷ sucrose + water) x 100.

Hot Plate

A 1 L beaker was placed on a hotplate at 100 ºC. Mice were placed in the beaker and latency to paw licking or jumping was recorded (Eddy and Leimbach, 1953) up to a limit of 30 s.

Learned Helplessness

For learned helplessness (LH), mice of both genotypes were randomly assigned to shock or no-shock conditions (Ridder et al., 2005, Malkesman et al., 2012). Using a Gemini shuttle-box (SDI, San Diego, CA) helplessness was induced over two days by placing mice on one side and exposing them to 360 trials that consisted of a 0.3 mA scrambled shock through the grid floor that lasted 2 s with intertrial intervals (ITI) of variable lengths (1-10 s; averaging 5 s). Controls were placed in the apparatus for the same length of time but without shock. On day 3, mice were given 30 trials with 10 s ITIs. Trials consisted of the door opening and activation of foot shock for 24 s or until the mouse crossed into the non-shock side. Number of escapes and latency to escape were recorded.

Novelty Suppressed Feeding

Following LH, mice were food and water deprived for 24 h and presented with food for

10 min in a novel environment (Bessa et al., 2009) that consisted of a 40 cm x 40 cm acrylic box. Latency to pellet investigation and amount eaten after returning to their home cage for 5 min were recorded.

Dexamethasone Suppression Test

Three days later, mice had blood collected by submandibular sampling. The next day they received a 3 µg/100 g body weight intraperitoneal injection of dexamethasone. Six hours later they were decapitated and blood collected (Ridder et al., 2005). Blood was collected at the same time of day both times.

Corticosterone Analysis

Blood was collected in tubes containing 2% EDTA. Samples were kept on ice until spun at 610 RCF for 15 min at 4 °C. Plasma was removed and stored at -80 °C. Samples were assayed using one lot of Immunodiagnostic Systems ® Corticosterone EIA Kit in duplicate following the kit instructions.

Tamoxifen-induced PDE1B KO

Forebrain dominant calcium/calmodulin-dependent protein kinase II alpha (CaMKIIα), a known regulator of PDE1B, Cre driver B6.Cg- Tg (Camk2a-cre/ERT2)1Aibs/J, was used to create a tamoxifen inducible Pde1b mouse (Tsien et al., 1996, Madisen et al., 2010).

Pde1bflox/flox :: Cre-/- mice were crossed with Pde1bflox/flox :: Cre+/- to obtain Pde1bflox/flox :: Cre+/- offspring. Recombination was initiated via tamoxifen administration. A subset of mice were also bred to a Gt(ROSA)26Sortm14(CAG-tdTomato)Hze mouse line to confirm expression of CaMKIIα- cre using td-tomato fluorescence (red) and immunohistochemistry (IHC) staining of PDE1B

(green). Western blots were used to verify PDE1B deletion.

Tamoxifen Administration

Male mice were gavaged with 20 mg/kg tamoxifen in 10 mL/kg of corn oil or corn oil alone for 5 days (Madisen et al., 2010) beginning on P0, P32, or P60 (KO(P0, P32, or P60)). No more than one male per group per age was used per litter. Mice were tested in open-field, TST, and

FST as adults.

Experiments

Experiment-1 was to test the global KO mice for consistency with the constitutive KO mice and determine the effects in HET mice. Accordingly, we used male WT, HET, and KOCMV mice for open-field on day 1, TST on day 2, and FST on day 3 and 4 (N: WT=11, HET=8,

KOCMV=11 (Fig. 3a-c)). No differences were seen between WT and HET mice, therefore, HET mice were not tested further.

Experiment-2 was to determine if the phenotype was also present in female mice.

Accordingly, we compared female WT (n=14) and Pde1b KOCMV (n=14) mice as in Experiment-1

(Fig. 3d-f). The immobility phenotype was similar in males and females, therefore, males were used in Experiment-3.

Experiment-3 was designed to determine if the open-field increased activity in KO mice affected swimming activity. Accordingly, we tested both genotypes as follows: home-cage on

day 1 and 2, open-swim on the morning of day 3 and open-field on the afternoon, repeat open- swim on the morning of day 4 and open-field on the afternoon, TST on day 5, and FST on day 6 and 7 (Fig. 4) (N: WT=11 and KOCMV=12).

Experiment-4 was designed to determine if the KO mice phenotype in TST and FST was associated with stress responsiveness or anhedonia. Accordingly, WT and Pde1b KOCMV male mice were divided into subgroups for LH: WT/No Shock (n=10), WT/Shock (n=14), KOCMV/No

Shock (n=11), KOCMV/Shock (n=12). The sequence was: corticosterone assessment on day 1, sipper sack habituation on days 1-5, hot plate on day 6, sucrose preference on day 6-7, LH on days 7-9, repeat corticosterone assessment on day 8, 24 h food/water deprivation on day 9-10, novelty suppressed feeding on day 10, repeat sucrose preference on days 10-11, and dexamethasone suppression test on day 14 (Fig. 5).

Experiment-5 was designed to determine the developmental onset of the immobility- resistant phenotype in KO mice. Accordingly, the experiment consisted of KOP0, P32, P60 mice

(Fig. 6) tested as adults in open-field on day 1, TST on day 2, and FST on days 3-4 (Fig. 7).

Group sizes were: WTP0 (n=14), KOP0 (n=16), WTP32 (n=12), KOP32 (n=12), WTP60 (n=10), and

P60 KO (n=13).

Data Analysis

Data were analyzed using SAS (v9.3, SAS Institute, Cary, NC) with p ≤ 0.05 as the threshold for significance. To control for litter, only one mouse per genotype per sex per litter was used. T-tests were used when there were two groups. When there were three groups, mixed linear ANOVA was used and data presented as least square mean (LS Means) ± SEMs.

Repeated measure Mixed model ANOVAs were used for corticosterone, home-cage, open- swim, and open-field and used the autoregressive-1 covariance structure and Kenward-Roger first order degrees of freedom. Litter was a random factor in these models.

Results

Generation of Global Knockout Mice

Whole brain western blots and IHC confirmed the deletion of PDE1B in the brain of

Pde1b KOCMV mice (Fig. 1b, Fig. 2a’-d’). We previously showed that Pde1b mRNA is highly expressed in the striatum (Hufgard et al., 2017). In WT mice using IHC, we show little to no staining in the prefrontal cortex (Fig. 2a), confirmed the presence of PDE1B protein in striatum

(Fig. 2b), show protein in the hippocampus in the dentate gyrus and CA3 (Fig 2c,d), and also in the substantia nigra (Fig. 2d).

Pde1b KOCMV mice are healthy, well groomed, and not visibly different from WT mice.

Experiment 1: The male Pde1b KOCMV mice do not differ from HET or WT littermates in open- field activity [F(2,32.6=0.8, p>0.4)] (Fig. 3a,c). Experiment 2: KOCMV females were more active than WT littermates [F(1,44.7)=6.1, p<0.05] (Fig. 3b,d).

Acute Stress

Experiment 1: Pde1b KOCMV mice showed decreased immobility in both TST

[F(2,28)=12.7, p<0.001] (Fig. 3e) and FST [F(2,27)=29.3, p<0.001] (Fig. 3g) compared with WT and HET mice, thereby verifying that the Pde1b KOCMV mice have the same phenotype as the constitutive KO mice (Hufgard et al., 2017). Heterozygous male mice did not differ from WT mice; for this reason HET females were not tested.

Experiment 2: Female KOCMV mice showed the same immobility as Pde1b KOCMV males, i.e., they showed reduced immobility in the TST [t(25)=3.0, p<0.01] (Fig. 3f); however, in FST

[t(24)=0.8, p>0.05] the effect was only a trend and not significant (Fig. 3h) compared with female WT mice.

Activity

Experiment 3: Since Pde1b KO mice are more active, we tested whether this had a diurnal component and whether open-field differences translate to swimming. For the former, home-cage activity was assessed and for the latter swimming activity. Pde1b KOCMV mice

showed greater increases in activity during the dark cycle compared with WT mice during the light cycle [Genotype x Interval: F(23,444)=2.2, p<0.01] (Fig. 4a). However, when Pde1b

KOCMV mice were placed in an open pool of water, they showed no difference in swim speed

(Fig. 4b and d) or distance traveled. A trend towards increased open-field activity was seen in

Pde1b KOCMV mice when compared with WT mice when tested for 1 h during the light cycle but the effect was not significant [Genotype x Interval: F(11,210)=1.6, p<0.10] (Fig. 4c and e).

Following these procedures, Pde1b KOCMV mice still had decreased immobility in TST and FST compared with WT mice [TST: t(19)=2.4, p<0.05; FST day 1: t(18)=1.9, p<0.05] (Fig. 4f and g).

Sucrose Preference

Experiment 4: Pde1b KOCMV and WT mice did not differ in the amount of water consumed when given water, but Pde1b KOCMV mice (13.9 ± 1.4 g, consumed in 7 days) consumed more sucrose when given alone than WT mice (10.6 ± 0.6 g, consumed in 2 days)

[t(21)=2.3, p<0.05]. When a choice was presented, Pde1b KOCMV mice did not differ in sucrose preference compared with WT mice (WT: 72.3 ± 7.1%, KO: 64.2 ± 7.8%) (data not shown).

Learned Helplessness

Before LH testing, mice were tested for pain thresholds on a hot plate. There were no differences in reactivity between groups (Fig. 5a). During the LH test, mice in the inescapable shock condition had longer escape latencies when escape was provided compared with no- shock groups [F(1,105)=317.0, p<0.001] (Fig. 5b). There were no differences in latency between WT and Pde1b KOCMV mice on this test.

Novelty Suppressed Feeding and Sucrose Preference

For novelty suppressed feeding, latency to eat food (Fig. 5c) and the amount of food consumed (not shown) were not different across LH conditions or genotypes. Mice were also given a second sucrose preference test after 24 h food and water deprivation. There were no differences between LH and no-shock mice of either genotype (Fig. 5d).

There was an increase in corticosterone after the second day of LH between the no- shock and shock conditions, but there were no differences between the WT and KOCMV groups

(Fig. 5e). Corticosterone was not different among groups 5 days after LH (Fig. 5f, left). There were no differences in corticosterone 6 h after dexamethasone in KOCMV vs. WT mice (Fig. 5f, right).

Critical Period

Experiment 5: Pde1b was selectively deleted in cells containing the CaMKIIα promoter after tamoxifen exposure. Tamoxifen was started on P0, P32, or P60. There were no changes in general appearance or body weight from tamoxifen exposure. Pde1b KO mice at P32, but not at P0 or P60, showed reduced striatal PDE1B [P0: t(3.0)=1.1, p>0.05; P32: t(2.0)=2.9, p<0.05; P60: t(1.3)=2.0, p<0.1] (Fig. 6e, f, g, left). In the hippocampus PDE1B was reduced in

KO mice at P0 and P32 but not at P60 [P0: t(2.3)=3.5, p<0.05; P32: t(1.9)=3.0, p<0.05; P60: t(2.4)=1.3, p>0.05] compared with WT mice (Fig. 6e, f, g, right). Figure 6a, b, c, d compares corn oil control and tamoxifen mice at P0 and shows reduction of PDE1B (green) and induction of ROSA (red) in striatum. Pde1b KOP0, KOP32, and KOP60 mice showed no differences from WT mice in open-field activity (not shown). Pde1b KOP0 mice had decreased immobility compared with WT littermates in FST [t(27.7)=1.8, p<0.05] (Fig. 7b), but not TST (Fig. 7a). Pde1b KOP32 mice had decreased immobility compared with WT in both TST and FST [TST: t(21.9)=2.5, p<0.05; FST: t(21.4)=1.8, p<0.05] (Fig. 7c, d). Pde1b KOP60 mice show no difference on either test (Fig. 7e, f).

Discussion

The antidepressant-like phenotype of constitutive Pde1b KO mice was described

(Hufgard et al., 2017). Constitutive Pde1b KO mice are resistant to the induction of immobility in both the TST and FST and these mice show no Pde1a or Pde1c compensatory changes. We also showed that FST and chronic variable stress increases PDE1B protein expression in

striatum and hippocampus in WT mice. Since constitutive models can cause compensatory changes or fail to induce neuronal activity similarly to pharmacological targets, we developed a floxed mouse targeting Pde1b exons 3-5. Pde1b floxed mice were bred to CMV Cre mice to create a global Pde1b KO mouse. These mice recapitulated the phenotype of constitutive KO mice in both TST and FST. Furthermore, we found that the Pde1b KO phenotype was more pronounced in males than in females. By fluorescent IHC, PDE1B protein expression is highest in caudate-putamen, nucleus accumbens, dentate gyrus, CA3, and substantia nigra in WT mice.

This is in agreement with prior protein and mRNA expression data in rats and humans and confirms that our global KO is specific to Pde1b (Polli and Kincaid, 1992, 1994, Yan et al., 1994,

Yu et al., 1997, Lakics et al., 2010, Kelly et al., 2014).

In the global Pde1b KO line, spontaneous locomotor activity was increased but only in females during the diurnal phase as seen previously (Reed et al., 2002). However, other data show increased activity before and after stimulant exposure in both sexes of constitutive Pde1b

KO mice (Reed et al., 2002, Ehrman et al., 2006, Siuciak et al., 2007). Here, for the first time, we determined activity over a period of 48 h and found that Pde1b KOCMV mice showed greater hyperactivity during the dark cycle. However, the nocturnal activity increase in Pde1b KOCMV mice returned to WT mice levels during most of the light cycle.

To determine if increased spontaneous activity contributed to the decreased TST and

FST immobility in Pde1b KO mice, we tested activity in open-swim and open-field tests given the same day. Pde1b KOCMV mice showed no differences in average or maximum swim speed, or distance traveled in an open pool of water and minor differences in the open-field. The same mice tested in TST and FST after open-swim and open-field still showed the immobility-resistant phenotype. This demonstrates that the Pde1b KOCMV depression-resistant phenotype is not confounded by activity differences.

We questioned if the antidepressant-like phenotype of Pde1b KOCMV mice was a byproduct of stress resistance since FST and TST both cause stress. We reasoned that if the

underlying effect was stress-related then other stressors would also show a resistance phenotype. For this we used LH. Induction of LH caused longer escape latencies than in non- shocked controls, but there were no differences in escape latencies between Pde1b KOCMV and

WT mice, indicating that the KO phenotype is not directly related to stress.

To further examine the depression-related phenotype of the Pde1b KO mice, we used sucrose preference as a test of anhedonia. There were no differences between Pde1b KOCMV and WT mice. We also used novelty suppressed feeding and dexamethasone suppression, however, the latter test did not work as well as it had in preliminary experiments which compromises its interpretation. However, novelty suppressed feeding did work and showed no differences between Pde1b KO mice compared with WT mice, reinforcing the view that the KO phenotype is not stress-related.

When we deleted Pde1b at different ages, the early deletions resulted in the antidepressant-like phenotype, i.e., when Pde1b was deleted by tamoxifen starting at P0 or

P32, the same antidepressant-like phenotype was seen as in global Pde1b KOCMV mice.

However, the behavior results of tamoxifen at P60 were inconclusive because at this age tamoxifen did not effectively reduce PDE1B either because the dose was insufficient or a longer onset time than we used was needed. If there are genetic variants of Pde1b in people (this is not yet known), these data suggest there may be protective variations against depression as a function of inter-individual differences in variant subtypes. Alternatively, early age phenotype induction could be the result of differences in Cre driver expression. Pyramidal cell neurogenesis is embryonic and enters post-mitotic phases by P0 and are differentiated by P7

(Angevine Jr, 1965, Stanfield and Cowan, 1979, Pokorný and Yamamoto, 1981a, b). PDE1B is located in hippocampal pyramidal cells, therefore, P0 tamoxifen induction would take place during neurogenesis (Kincaid et al., 1987). Further, expression of CaMKIIα recombination at

P29 is present in hippocampus, striatum, cortex, and Purkinje cells and differs from that prior to

P29 (Tsien et al., 1996). This suggests that different populations of cells are more sensitive to

recombination at different ages. Sometimes there are effects from tamoxifen inducible Cre, including leakage (non-tamoxifen induced expression) that may also vary by age. Alternatively, age-dependent phenotype differences may reflect the lag between tamoxifen efficacy at deleting the gene and its effects on the emergence of the immobility phenotype that may not be the same at P0, P32, and P60. The antidepressant efficacy of current drugs in patients can take 4-

6 weeks to reach full effect (Belzung, 2014). Whether this interacts with the age at which a protein’s adult expression emerges, such as for PDE1B, is unknown.

In sum, PDE1B is expressed in areas associated with neuropsychiatric disorders, including depression (Lakics et al., 2010). Our data support the view that PDE1B may be a target for pharmacological intervention for depression.

References

Angevine Jr, J.B. (1965) Time of neuron origin in the hippocampal region: An autoradiographic

study in the mouse. Experimental Neurology.

Belzung, C. (2014) Innovative Drugs to Treat Depression: Did Animal Models Fail to Be

Predictive or Did Clinical Trials Fail to Detect Effects&quest. Neuropsychopharmacology,

39, 1041-1051.

Bessa, J., Ferreira, D., Melo, I., Marques, F., Cerqueira, J., Palha, J., Almeida, O. & Sousa, N.

(2009) The mood-improving actions of antidepressants do not depend on neurogenesis

but are associated with neuronal remodeling. Molecular psychiatry, 14, 764-773.

Conti, M. & Beavo, J. (2007) Biochemistry and physiology of cyclic nucleotide

phosphodiesterases: essential components in cyclic nucleotide signaling. Annu. Rev.

Biochem., 76, 481-511.

Cryan, J.F., Mombereau, C. & Vassout, A. (2005a) The tail suspension test as a model for

assessing antidepressant activity: review of pharmacological and genetic studies in

mice. Neuroscience & Biobehavioral Reviews, 29, 571-625.

Cryan, J.F., Valentino, R.J. & Lucki, I. (2005b) Assessing substrates underlying the behavioral

effects of antidepressants using the modified rat forced swimming test. Neuroscience &

Biobehavioral Reviews, 29, 547-569.

Dlaboga, D., Hajjhussein, H. & O'Donnell, J.M. (2008) Chronic haloperidol and clozapine

produce different patterns of effects on phosphodiesterase-1B,-4B, and-10A expression

in rat striatum. Neuropharmacology, 54, 745-754.

Duman, R.S., Malberg, J. & Thome, J. (1999) Neural plasticity to stress and antidepressant

treatment. Biological psychiatry, 46, 1181-1191.

Eddy, N.B. & Leimbach, D. (1953) Synthetic analgesics. II. Dithienylbutenyl-and

dithienylbutylamines. Journal of Pharmacology and Experimental Therapeutics, 107,

385-393.

Ehrman, L., Williams, M., Schaefer, T., Gudelsky, G., Reed, T., Fienberg, A., Greengard, P. &

Vorhees, C. (2006) Phosphodiesterase 1B differentially modulates the effects of

methamphetamine on locomotor activity and spatial learning through DARPP32‐

dependent pathways: evidence from PDE1B‐DARPP32 double‐knockout mice. Genes,

Brain and Behavior, 5, 540-551.

Fujita, M., Hines, C.S., Zoghbi, S.S., Mallinger, A.G., Dickstein, L.P., Liow, J.-S., Zhang, Y.,

Pike, V.W., Drevets, W.C. & Innis, R.B. (2012) Downregulation of Brain

Phosphodiesterase Type IV Measured with 11 C-( R)-Rolipram Positron Emission

Tomography in Major Depressive Disorder. Biological psychiatry, 72, 548-554.

Grauer, S.M., Pulito, V.L., Navarra, R.L., Kelly, M.P., Kelley, C., Graf, R., Langen, B., Logue, S.,

Brennan, J. & Jiang, L. (2009) Phosphodiesterase 10A inhibitor activity in preclinical

models of the positive, cognitive, and negative symptoms of schizophrenia. Journal of

Pharmacology and Experimental Therapeutics, 331, 574-590.

Hansen, R. & Zhang, H.-T. (2015) Phosphodiesterase-4 modulation as a potential therapeutic

for cognitive loss in pathological and non-pathological aging: possibilities and pitfalls.

Current pharmaceutical design, 21, 291-302.

Hautman, E.R., Kokenge, A.N., Udobi, K.C., Williams, M.T., Vorhees, C.V. & Skelton, M.R.

(2014) Female mice heterozygous for creatine transporter deficiency show moderate

cognitive deficits. Journal of inherited metabolic disease, 37, 63-68.

Hebb, A.L. & Robertson, H.A. (2007) Role of phosphodiesterases in neurological and

psychiatric disease. Current opinion in pharmacology, 7, 86-92.

Höfgen, N., Stange, H., Schindler, R., Lankau, H.-J., Grunwald, C., Langen, B., Egerland, U.,

Tremmel, P., Pangalos, M.N. & Marquis, K.L. (2010) Discovery of imidazo [1, 5-a] pyrido

[3, 2-e] pyrazines as a new class of phosphodiesterase 10A inhibitiors. Journal of

medicinal chemistry, 53, 4399-4411.

Hufgard, J.R., Williams, M.T., Skelton, M.R., Grubisha, O., Ferreira, F.M., Sanger, H., Wright,

M.E., Reed-Kessler, T.M., Rasmussen, K., Duman, R.S. & Vorhees, C.V. (2017)

Phosphodiesterase-1b (Pde1b) knockout mice are resistant to forced swim and tail

suspension induced immobility and show upregulation of Pde10a. Psychopharmacology.

Jindal, A., Mahesh, R. & Bhatt, S. (2013) Etazolate, a phosphodiesterase 4 inhibitor reverses

chronic unpredictable mild stress-induced depression-like behavior and brain oxidative

damage. Pharmacology Biochemistry and Behavior, 105, 63-70.

Jindal, A., Mahesh, R., Gautam, B., Bhatt, S. & Pandey, D. (2012) Antidepressant-like effect of

etazolate, a cyclic nucleotide phosphodiesterase 4 inhibitor—an approach using rodent

behavioral antidepressant tests battery. European journal of pharmacology, 689, 125-

131.

Katarzyna, S., Dorota, N., Ewa, P. & Piotr, W. (2012) Influence of the phosphodiesterase type 5

inhibitor, sildenafil, on antidepressant-like activity of magnesium in the forced swim test

in mice. Pharmacological Reports, 64, 205-211.

Kelly, M.P., Adamowicz, W., Bove, S., Hartman, A.J., Mariga, A., Pathak, G., Reinhart, V.,

Romegialli, A. & Kleiman, R.J. (2014) Select 3′, 5′-cyclic nucleotide phosphodiesterases

exhibit altered expression in the aged rodent brain. Cellular signalling, 26, 383-397.

Kincaid, R.L., Balaban, C.D. & Billingsley, M.L. (1987) Differential localization of calmodulin-

dependent enzymes in rat brain: evidence for selective expression of cyclic nucleotide

phosphodiesterase in specific neurons. Proceedings of the National Academy of

Sciences, 84, 1118-1122.

Lakics, V., Karran, E.H. & Boess, F.G. (2010) Quantitative comparison of phosphodiesterase

mRNA distribution in human brain and peripheral tissues. Neuropharmacology, 59, 367-

374.

Madisen, L., Zwingman, T.A., Sunkin, S.M., Oh, S.W., Zariwala, H.A., Gu, H., Ng, L.L., Palmiter,

R.D., Hawrylycz, M.J. & Jones, A.R. (2010) A robust and high-throughput Cre reporting

and characterization system for the whole mouse brain. Nature neuroscience, 13, 133-

140.

Malkesman, O., Austin, D., Tragon, T., Henter, I., Reed, J., Pellecchia, M., Chen, G. & Manji, H.

(2012) Targeting the BH3-interacting domain death agonist to develop mechanistically

unique antidepressants. Molecular psychiatry, 17, 770-780.

Mathers, C., Fat, D.M. & Boerma, J.T. (2008) The global burden of disease: 2004 update, World

Health Organization.

Maurice, D.H., Ke, H., Ahmad, F., Wang, Y., Chung, J. & Manganiello, V.C. (2014) Advances in

targeting cyclic nucleotide phosphodiesterases. Nature Reviews Drug Discovery, 13,

290-314.

Natesan, S., Ashworth, S., Nielsen, J., Tang, S., Salinas, C., Kealey, S., Lauridsen, J.,

Stensbøl, T., Gunn, R. & Rabiner, E. (2014) Effect of chronic antipsychotic treatment on

striatal phosphodiesterase 10A levels: a [ 11C] MP-10 PET rodent imaging

study with ex vivo confirmation. Translational psychiatry, 4, e376.

O'Donnell, J.M. & Zhang, H.T. (2004) Antidepressant effects of inhibitors of cAMP

phosphodiesterase (PDE4). Trends in Pharmacological Sciences, 25, 158-163.

Pokorný, J. & Yamamoto, T. (1981a) Postnatal ontogenesis of hippocampal CA1 area in rats. I.

Development of dendritic arborisation in pyramidal neurons. Brain research bulletin, 7,

113-120.

Pokorný, J. & Yamamoto, T. (1981b) Postnatal ontogenesis of hippocampal CA1 area in rats. II.

Development of ultrastructure in stratum lacunosum and moleculare. Brain research

bulletin, 7, 121-130.

Polli, J.W. & Kincaid, R.L. (1992) Molecular cloning of DNA encoding a calmodulin-dependent

phosphodiesterase enriched in striatum. Proceedings of the National Academy of

Sciences, 89, 11079-11083.

Polli, J.W. & Kincaid, R.L. (1994) Expression of a calmodulin-dependent phosphodiesterase

isoform (PDE1B1) correlates with brain regions having extensive dopaminergic

innervation. The Journal of neuroscience, 14, 1251-1261.

Porsolt, R.D., Bertin, A., Blavet, N., Deniel, M. & Jalfre, M. (1979) Immobility induced by forced

swimming in rats: effects of agents which modify central catecholamine and serotonin

activity. European journal of pharmacology, 57, 201-210.

Pothion, S., Bizot, J.-C., Trovero, F. & Belzung, C. (2004) Strain differences in sucrose

preference and in the consequences of unpredictable chronic mild stress. Behavioural

brain research, 155, 135-146.

Reed, T.M., Repaske, D.R., Snyder, G.L., Greengard, P. & Vorhees, C.V. (2002)

Phosphodiesterase 1B knock-out mice exhibit exaggerated locomotor hyperactivity and

DARPP-32 phosphorylation in response to dopamine agonists and display impaired

spatial learning. The Journal of neuroscience, 22, 5188-5197.

Ridder, S., Chourbaji, S., Hellweg, R., Urani, A., Zacher, C., Schmid, W., Zink, M., Hörtnagl, H.,

Flor, H. & Henn, F.A. (2005) Mice with genetically altered glucocorticoid receptor

expression show altered sensitivity for stress-induced depressive reactions. The Journal

of neuroscience, 25, 6243-6250.

Schaefer, T., Braun, A., Amos‐Kroohs, R., Williams, M., Ostertag, E. & Vorhees, C. (2012) A

new model of Pde4d deficiency: genetic knock‐down of PDE4D enzyme in rats produces

an antidepressant phenotype without spatial cognitive effects. Genes, Brain and

Behavior, 11, 614-622.

Schmidt, C.J., Chapin, D.S., Cianfrogna, J., Corman, M.L., Hajos, M., Harms, J.F., Hoffman,

W.E., Lebel, L.A., McCarthy, S.A. & Nelson, F.R. (2008) Preclinical characterization of

selective phosphodiesterase 10A inhibitors: a new therapeutic approach to the treatment

of schizophrenia. Journal of Pharmacology and Experimental Therapeutics, 325, 681-

690.

Siuciak, J.A., Chapin, D.S., Harms, J.F., Lebel, L.A., McCarthy, S.A., Chambers, L., Shrikhande,

A., Wong, S., Menniti, F.S. & Schmidt, C.J. (2006a) Inhibition of the striatum-enriched

phosphodiesterase PDE10A: a novel approach to the treatment of psychosis.

Neuropharmacology, 51, 386-396.

Siuciak, J.A., McCarthy, S.A., Chapin, D.S., Fujiwara, R.A., James, L.C., Williams, R.D., Stock,

J.L., McNeish, J.D., Strick, C.A. & Menniti, F.S. (2006b) Genetic deletion of the striatum-

enriched phosphodiesterase PDE10A: evidence for altered striatal function.

Neuropharmacology, 51, 374-385.

Siuciak, J.A., McCarthy, S.A., Chapin, D.S., Reed, T., Vorhees, C. & Repaske, D. (2007)

Behavioral and neurochemical characterization of mice deficient in the

phosphodiesterase-1B (PDE1B) enzyme. Neuropharmacology, 53, 113-124.

Socała, K., Nieoczym, D., Wyska, E., Poleszak, E. & Wlaź, P. (2012a) Sildenafil, a

phosphodiesterase type 5 inhibitor, enhances the activity of two atypical antidepressant

drugs, mianserin and tianeptine, in the forced swim test in mice. Progress in Neuro-

Psychopharmacology and Biological Psychiatry, 38, 121-126.

Socała, K., Nieoczym, D., Wyska, E., Poleszak, E. & Wlaź, P. (2012b) Sildenafil, a

phosphodiesterase type 5 inhibitor, enhances the antidepressant activity of amitriptyline

but not desipramine, in the forced swim test in mice. Journal of Neural Transmission,

119, 645-652.

Stanfield, B.B. & Cowan, W.M. (1979) The development of the hippocampus and dentate gyrus

in normal and reeler mice. Journal of Comparative Neurology, 185, 423-459.

Tang, X., Orchard, S.M. & Sanford, L.D. (2002) Home cage activity and behavioral performance

in inbred and hybrid mice. Behavioural brain research, 136, 555-569.

Tsien, J.Z., Chen, D.F., Gerber, D., Tom, C., Mercer, E.H., Anderson, D.J., Mayford, M., Kandel,

E.R. & Tonegawa, S. (1996) Subregion- and Cell Type–Restricted Gene Knockout in

Mouse Brain. Cell, 87, 1317-1326.

Wang, Z.Z., Zhang, Y., Liu, Y.Q., Zhao, N., Zhang, Y.Z., Yuan, L., An, L., Li, J., Wang, X.Y. &

Qin, J.J. (2013) RNA interference‐mediated phosphodiesterase 4D splice variants

knock‐down in the prefrontal cortex produces antidepressant‐like and cognition‐

enhancing effects. British journal of pharmacology, 168, 1001-1014.

Xu, Y., Pan, J., Chen, L., Zhang, C., Sun, J., Li, J., Nguyen, L., Nair, N., Zhang, H. & O'Donnell,

J.M. (2013) Phosphodiesterase-2 inhibitor reverses corticosterone-induced neurotoxicity

and related behavioural changes via cGMP/PKG dependent pathway. International

Journal of Neuropsychopharmacology, 16, 835-847.

Xu, Y., Pan, J., Sun, J., Ding, L., Ruan, L., Reed, M., Yu, X., Lin, D., Li, J. & Chen, L. (2015)

Inhibition of phosphodiesterase 2 reverses impaired cognition and neuronal remodeling

caused by chronic stress. Neurobiology of aging, 36, 955-970.

Xu, Y., Zhang, H.-T. & O’Donnell, J.M. (2011) Phosphodiesterases in the central nervous

system: implications in mood and cognitive disorders. Phosphodiesterases as Drug

Targets. Springer, pp. 447-485.

Yan, C., Bentley, J.K., Sonnenburg, W.K. & Beavo, J.A. (1994) Differential expression of the 61

kDa and 63 kDa calmodulin-dependent phosphodiesterases in the mouse brain. The

Journal of neuroscience, 14, 973-984.

Yu, J., Wolda, S.L., Frazier, A.L., Florio, V.A., Martins, T.J., Snyder, P.B., Harris, E.A., McCaw,

K.N., Farrell, C.A. & Steiner, B. (1997) Identification and characterisation of a human

calmodulin-stimulated phosphodiesterase PDE1B1. Cellular signalling, 9, 519-529.

Zhang, H.-T., Huang, Y., Jin, S.C., Frith, S.A., Suvarna, N., Conti, M. & James, M. (2002)

Antidepressant-like profile and reduced sensitivity to rolipram in mice deficient in the

PDE4D phosphodiesterase enzyme. Neuropsychopharmacology, 27, 587-595.

Figures

CMV a. b. KO WT

PDE1B (61KDa)

Actin (42 KDa)

Figure 1. Generation of floxed mice and confirmation of global knockout mice (KOCMV). a.

Schematic of the Floxed mouse model. b. Whole brain western blot confirmation of complete lack of PDE1B in KOCMV mice.

100

Figure 2. Fluorescent immunohistochemistry analysis of PDE1B in the brain localizes to regions related to stress and depression. a. WT prefrontal cortex. a’. KOCMV prefrontal cortex. b. WT striatum. b’. KOCMV striatum. c. WT hippocampus. c’. KOCMV hippocampus. d. WT substantia nigra. d’. KOCMV substantia nigra. DAPI=Blue and PDE1B=Green. Scale = 1000 µm.

101

102

Figure 3. KOCMV mice have decreased immobility time in depression-related tasks compared with control littermates. a. Locomotor activity males genotype*interval (WT n=11, HET n=8,

KOCMV n=11). b. Locomotor activity females genotype*interval (WT n=14, KOCMV n=14). c.

Locomotor activity males. d. Locomotor activity females. e. TST males (WT n=11, HET n=8,

KOCMV n=10). f. TST females (WT n=13, KOCMV n=14). g. FST day 2, 5 min for males (WT n=9,

HET n=8, KOCMV n=11). h. FST day 2, 5 min for females (WT n=13, KOCMV n=13). * p ≤ 0.05, ** p ≤ 0.01, and *** p ≤ 0.001.

103

104

Figure 4. Pde1b KOCMV mice are hyperactive at night and when introduced to the novel home cage, although land hyperactivity did not predict water hyperactivity. a. 48 h home cage (WT n=11, KOCMV n=12). b. Day 1, 5 min open swim (WT n=11, KOCMV n=12). c. 1 h locomotor activity day 1 (WT n=11, KOCMV n=12). d. Day 2, 5 min open swim (WT n=11, KOCMV n=12). e.

1 h locomotor activity day 2 (WT n=11, KOCMV n=12). f. 5 min TST (WT n=10, KOCMV n=11). g.

6 min FST (WT n=9, KOCMV n=11). † p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001.

105

106

Figure 5. Pde1b KOCMV mice do not differ from WT littermates when exposed to the learned helplessness procedure. a. Latency to remove paw from hot plate (WT n=22, KOCMV n=21). b.

Latency to escape shock (WT/No Shock n=10, WT/Shock n=11, KOCMV/No Shock n=10,

KOCMV/Shock n=8). c. Latency to eat in a novel environment (WT/No Shock n=8, WT/Shock n=12, KOCMV/No Shock n=10, KOCMV/Shock n=11). d. Sucrose preference following 24 h food and water deprivation (WT/No Shock n=10, WT/Shock n=14, KOCMV/No Shock n=11,

KOCMV/Shock n=12). e. Plasma corticosterone response before, during, and after LH exposure

(WT/No Shock n=9, WT/Shock n=9, KOCMV/No Shock n=9, KOCMV/Shock n=8). f. Plasma corticosterone following dexamethasone (WT/No Shock n=8, WT/Shock n=8, KOCMV/No Shock n=8, KOCMV/Shock n=7).

107

108

109

Figure 6. Analysis of localization and quantity of PDE1B in KOP0, P32, P60 mice. a. Expression pattern of CaMKIIα Cre driver and PDE1B in striatum of P0 corn oil dosed mice. a’. Expression pattern of CaMKIIα Cre driver and PDE1B in striatum of P0 tamoxifen dosed mice. b. DAPI corn oil. b’. DAPI tamoxifen. c. PDE1B corn oil. c’. PDE1B tamoxifen. d. ROSA corn oil. d’. ROSA tamoxifen. e. Western blot analysis of striatum and hippocampus in P0 dosed mice. f. Western blot analysis of striatum and hippocampus in P32 dosed mice. g. Western blot analysis of striatum and hippocampus in P60 dosed mice. Blue = DAPI, Green = PDE1B, Red = Td-

Tomato. 2000 µm. † p < 0.1 and * p < 0.05.

110

111

Figure 7. When PDE1B is removed prior to sexual maturity KO mice had a reduction in immobility time in the TST and the FST. a. KOP0 Tail suspension test (WT n=14, KOP0 n=16). b.

KOP0 Forced swim test (WT n=14, KOP0 n=16). c. KOP32 Tail suspension test (WT n=12, KOP32 n=12). d. KOP32 Forced swim test (WT n=12, KOP32 n=12). e. KOP60 Tail suspension test (WT n=10, KOP60 n=13). f. KOP60 Forced swim test (WT n=10, KOP60 n=13). * p < 0.05

112

Acknowledgements

This work was supported by NIH T32 ES007051 and funds from Cincinnati Children’s Hospital

Research Foundation

113

CHAPTER 4

PDE1B is implicated in dopamine pathways and its deletion in dopamine receptor

1a cells confers an immobility-resistant phenotype in forced swim and tail

suspension tests

Jillian R. Hufgard, Michael T. Williams, and Charles V. Vorhees

Division of Neurology, Dept. of Pediatrics, Cincinnati Children’s Research Foundation and University of Cincinnati College of Medicine, Cincinnati, OH, USA

114

Abstract

Depression is a leading psychiatric disorder, but despite this current antidepressants fail to provide adequate relief for many patients. Current antidepressants act primarily presynaptically as reuptake inhibitors, whereas phosphodiesterases (PDE) regulate postsynaptic cyclic nucleotide signal duration by hydrolyzing second messengers terminating downstream signal transduction. Knocking out PDE1B in mice results in a phenotype resistant to immobility in forced swim (FST) and tail suspension (TST) tests. We showed PDE1B is enriched in striatum, hippocampus, and substantia nigra. Here we found that Pde1b localizes to dopamine receptor 1 (Drd1), Drd2, dopamine transporter (DAT), and serotonin transporter

(SERT) expressing cells. Floxed Pde1b mice were crossed with Cre mice for DAT, SERT, or

Drd1a, but only Pde1bDrd1a knockout (KO) mice exhibited the immobility-resistant phenotype

CMV found in global Pde1b KO mice. The dopamine direct pathway (D1) was therefore further tested here by intrastriatal injections of a D1 antagonist (SCH23390) and agonist (SKF38393).

D1 antagonism exaggerated the Pde1b KO mouse phenotype on open-field, TST, and FST while

D1 agonism neutralized the KO mouse phenotype. The data support the view that Pde1b KO mouse immobility-resistant phenotype is primarily driven by dopamine pathways.

115

Introduction

Major depressive disorder (MDD) continues to be one of the leading causes of suicide and disability (Vijayakumar, 2005). The lifetime prevalence of MDD is 16% and is projected to be the second leading health-related economic burden by 2020 (Artigas et al., 1996, Greist et al., 2002, Kesler et al., 2003). There are >50 antidepressants available, yet their efficacy can be summarized as providing relief for about one-third, partial relief for another third, and no relief in the last third of patients (Fava and Davidson, 1996, Holtzheimer and Nemeroff, 2006,

Furukawa et al., 2009). The majority of antidepressants are presynaptic monoamine (serotonin, dopamine, or norepinephrine) reuptake inhibitors that act to increase synaptic availability of transmitters after release that has the effect of lengthening synaptic signaling. An alternate way of increasing signaling would be to extend postsynaptic signaling.

Selective serotonin reuptake inhibitors are the most prescribed antidepressants, however drugs that inhibit dopamine and dopamine reuptake are also effective in some patients.

Dopamine acts through the dopamine receptor 1 (D1) and D2 by increasing or decreasing cyclic adenosine monophosphate/protein kinase A (cAMP/PKA) activation. This, in turn, alters the balance between dopamine- and cAMP-regulated neuronal phosphoprotein (DARPP-32) and phosphatase-1 (Greengard et al., 1999) that then changes the state of ion-gated channels leading to depolarization or hyperpolarization of the postsynaptic membrane (Wang et al., 1991,

Montminy, 1997, Snyder et al., 1998). cAMP also affects cAMP response element binding

(CREB); increasing CREB can increase brain derived neurotrophic factor (BDNF) levels and low

BDNF is associated with depression (Konradi et al., 1994, Nair and Vaidya, 2006, Tanis and

Duman, 2007). This suggests that regulation of second messengers may have antidepressant properties through regulation of CREB and BDNF.

The duration of cyclic nucleotide action is determined by the rate at which phosphodiesterases (PDE) hydrolyze cyclic nucleotides. There are 11 PDE families containing

21 isoforms. Each isoform has specific binding affinities for cAMP, cyclic guanosine

116

monophosphate (cGMP), or both, and each has distinct tissue distributions and cellular localization. PDE1B, 2A, 4A, 4B, 4D, 7B, 9A, and 10A are all expressed in the striatum (Lakics et al., 2010). PDE1B, 4, and 10A are expressed in striatonigral neurons, dopamine terminals, and striatopallidal neurons, respectively, and are linked to neuropsychiatric disorders (Hebb and

Robertson, 2007, Nishi and Snyder, 2010, Heckman et al., 2016). PDE4D inhibitors have antidepressant efficacy in depressed patients, but side-effects were too severe for them to be clinically useful (Zhu et al., 2001, Jeon et al., 2005). PDE1B and 10A are found in D1 and D2 expressing cells with PDE1B more abundant in D1 and PDE10A more abundant in D2 positive cells (Nishi and Snyder, 2010, Heckman et al., 2016).

Previously, we showed that constitutive Pde1b and conditional global Pde1b knockout

(KO) mice exhibit an immobility resistant phenotype in tail suspension (TST) and forced swim

(FST) tests, and these effects can be dissociated from changes in open-field activity, open swim activity, stress reactivity, and anhedonia (sucrose preference test) (Hufgard et al., 2017a,

Hufgard et al., 2017b). Using tamoxifen-inducible Cre, we further showed that the immobility- resistant phenotype occurs when Pde1b is deleted from conception, from postnatal day 0 (P0), or from P32.

To further understand factors influencing the Pde1b immobility-resistant phenotype, herein we determined mRNA and protein expression of Pde1b in serotoninergic, dopaminergic,

(Siuciak et al., 2007), and glutamatergic positive cells. With this information, we used

Pde1bflox/flox mice bred to specific mouse Cre lines for the serotonin transporter (Sert), dopamine transporter (Dat), and dopamine receptor 1a (Drd1a) and created specific KO mice to investigate the relationship between these factors and the immobility-resistant phenotype; we were unable to include a Cre Drd2 experiment in this study. Of the three Cre-specific lines tested, only the Pde1b-Drd1a specific KO mice showed the immobility-resistant phenotype.

Given this, we tested the effects of intra-striatal infusion of a D1 antagonist compared with a D1

117

agonist on the Pde1b phenotype, and showed that inhibiting the D1 receptor in Pde1b KO mice is sufficient to induce an immobility-resistant phenotype.

Methods

Animals and Husbandry

Mice were bred in-house. Litters were housed in polysulfone cages in a pathogen free vivarium using Modular Animal Caging System (Alternative Design, Siloam Spring, AR) with

HEPA filtered air (Alternative Design, Siloam Spring, AR) at 30 air changes/h. The NIH 5001 diet and water were provided ad libitum with automated reverse-osmosis UV purified water delivered through a lixit (SE Lab Group, Napa, CA). Cages had corncob bedding and cotton nest material as partial enrichment. Mice were maintained on a 14:10 h light:dark cycle (lights on at 600 h). Protocols were approved by the Institutional Animal Care and Use Committee.

The vivarium is accredited by AAALAC International. At weaning (P28) mice were housed with littermates by sex with 2-4 mice per cage. Mice began testing at ~P60. Only one male was tested per genotype per litter. Behavioral testing was done blind with respect to genotype or treatment.

Generation of Pde1b conditional KO mice

Floxed mice were created on a C57BL/6J (albino) background as described (Hufgard et al., 2017b). Four distinct KO lines were used. B6.SJL-Slc6a3tm1.1(cre)Bkmn/J, B6.Cg-Tg(Fev- cre)1Esd/J, and B6;129-Tg(Drd1a-cre)120Mxu/Mmjax mice were bred to Pde1bflox/flox mice to create Pde1bflox/+ x cre+/- breeding pairs. These combinations produced offspring with the following genotypes: Pde1b+/+::cre-/- (wildtype (WT)), Pde1bflox/flox::cre-/-(Flox), Pde1b+/+::cre+/- for

CreSert, CreDat, or CreDrd1a, and Pde1bflox/flox::cre+/- (KOSert, KODat, or KODrd1a). Global Pde1b

KOCMV mice were generated and described using the cytomegalovirus (CMV) cre driver, B6.C-

Tg(CMV-cre)1Cgn/J (Hufgard et al., 2017b), and these mice were used to create

Pde1bflox/flox::cre+/- mice that were Pde1b-/-::cre+/-. These mice were bred with WT C57BL/6J

118

mice to remove cre. Thereafter, Pde1b+/- x Pde1b+/- breeding was used to generate WT and global Pde1b KOCMV mice for the intrastriatal injection experiments. A subset of mice were bred to the Gt(ROSA)26Sortm14(CAG-tdTomato)Hze mice to visualize expression of the SERT, DAT, and

DRD1A-cres using td-tomato fluorescence.

Fluorescent Immunohistochemistry

Brains were perfused with ice cold 1X phosphate buffered saline followed by 4% paraformaldehyde and then cut on a freezing microtome at 20 µm. Free floating sections were stained with cresyl violet to assist in identifying anatomical landmarks in striatum, hippocampus, and substantia nigra. Antibodies were: PDE1B:ab182565 (1º = 1:200, 2º = 1:500), DAT -

Slc6a3:MAB369 (1º = 1:200, 2º = 1:278), DRD1 - DRD1A:ab78021 (1º = 1:1000, 2º = 1:500),

SERT - Slc6a4:MAB352 (1º = 1:200, 2º = 1:100), and glutamate receptor (NMDA) – GRIN1:

Ab77264 (1º = 1:200, 2º = 1:200). Images were taken using a Nikon confocal microscope in the

Confocal Imaging Core at Cincinnati Children’s Hospital Medical Center.

RNA Scope In Situ Hybridization

Mice were rapidly decapitated and brains were flash frozen from WT mice. Brains were sectioned on a cryostat at 20 µm and cresyl violet stained as above. Sections were probed with the RNA Scope Kit from Advanced Cell Diagnostics (Hayward, CA). The following probes were used: Pde1b, Dat - Slc6a3, Drd1a, Drd2, Sert - Slc6a4, gamma-aminobutyric acid (Gaba) transporter - Slc6a12, glutamate (Glut) transporter – Slc1a3, Nmda – Grin1, RNA polymerase II

- Polr2a (Fig. 2E positive control), dihydrodipicolinate reductase – Dapb (Fig 2F. negative control), and glial fibrillary acidic protein (Gfap). The Pde1b probe targets a region between exons 6-14 and was tested for non-specific staining using constitutive Pde1b KO mice that have exons 6-9 deleted (Fig. 2E) (Hufgard et al., 2017a). Images were taken using a Nikon confocal microscope with 100x magnification. All images are maximum intensity projections made from z-stacks using NIS Elements Viewer.

Behavioral Tests

119

Separate groups of mice were used for each experiment. Within each experiment the same mice were tested sequentially in open-field, TST, and FST. Sample sizes are shown in figure captions. Only male mice were used since we showed that Pde1b KO males and females show the same immobility-resistant phenotype (Hufgard et al., 2017b).

Open-Field Locomotor Activity

Spontaneous locomotor activity was assessed in 40 x 40 cm automated activity monitors

(PAS System, San Diego Instruments, San Diego, CA) as described (Hautman et al., 2014).

Mice were placed in the monitors for 1 h and data were collected in 5 min intervals.

Tail Suspension Test

Immobility time and latency to first immobile event were recorded. Immobility was defined as the absence of movement.

Forced Swim Test

Mice were placed in a transparent glass cylindrical vessel 10 cm in diameter (i.d.) and 25 cm tall, filled with water to a depth of 6 cm at 22 ± 1 ºC. Mice were tested over 2 days (Porsolt et al., 1979, Cryan et al., 2005b) with scoring on both days. On day 1, mice were tested for 15 min with minutes 1-6 scored. On day 2, mice were tested for 5 min. Mice were scored for immobility, latency to immobility, and active swimming. Immobility was defined as movement only sufficient to stay afloat.

Western Blot Analysis

Frozen brain tissue was homogenized in radioimmuno-precipitation assay buffer with protease inhibitors. Protein was quantified using the BCATM Protein Assay Kit and diluted to 3

µg/µL. Western blots were performed using LI-COR Odyssey® procedures. Primary antibodies

120

from Abcam and their dilutions were: rabbit anti-PDE1B C-terminal (Ab182565) at 1:5000 and mouse anti-actin (Ab3280) at 1:2000 as a loading control. Odyssey IRDye 680 and 800 secondary antibodies were used at 1:15,000 dilutions. Relative protein levels are quantified using the LI-COR Odyssey® scanner and Image Studio analysis software that reads fluorescent intensity of the sample normalized to actin.

Cannulation Surgery

Adult male mice either WT or KOCMV were anesthetized with continuous isoflurane inhalation throughout surgery. Mice were placed in a Stoelting (Wooddale, IL) stereotaxic apparatus where a midline incision of the scalp was made followed by retraction of the skin and muscles to expose the skull. In both hemispheres, a 3.5 mm guide cannula was inserted and cemented directly to the skull of the mouse with a staple used for stability. Cannula placement in the striatum followed the coordinates from the Franklin and Paxinos atlas (2008) and are relative to bregma (AP: +0.65 mm, ML: +/- 2.00 mm, DV: -2.75 mm) (Zhang et al., 2003). After the procedure, a cement cap was created and the skin sutured around the cap. Mice were given 0.04 mL buprenorphine hydrochloride subcutaneously to minimize pain and a 0.25 mL saline bolus after surgery. The mice were then single housed for a minimum of 14 days of recovery and the duration of testing to minimize cannula disturbance.

Antagonist/Agonist Treatment

Mice were dosed 24 h prior to testing using an infusion pump and a 26 gauge 3.75 mm internal cannula, making the final injection depth 3.00 mm. Injection rate was of 0.3 μL/min, and the internal cannula was left in place for 1 min following injection (Mohammadi et al., 2015).

The volume given was 0.6 μL, i.e., 0.3 μL given per side (Mohammadi et al., 2015). The D1 antagonist SCH23390 (SCH) [C17H18ClNO.HCl ] (MW 324.24) was given at 0.3 μg/mouse. The

D1 agonist SKF38393 (SKF) [C16H17NO2.HBr] (MW 336.23) was given at 0.1 μg/mouse

(Zarrindast et al., 2010, Mohammadi et al., 2015). Drugs were purchased from Tocris

121

Bioscience (Avonmouth, Bristol, United Kingdom). The vehicle was 0.9% saline or saline alone in controls.

Data Analysis

Data were analyzed using SAS (v9.3, SAS Institute, Cary, NC), where p ≤ 0.05 was the threshold for significance. To control for litter effects only one mouse per genotype per litter was used. General linear model ANOVA was used when genotype was the independent variable with three or four groups (WT, Flox, Cre, and/or global KO) and data presented as least square mean (LS Means) ± SEM. Where there were more factors, mixed linear model ANOVAs were used. Two-way ANOVAs were used in experiments where the factors were genotype (KO vs WT) and drug (SAL, SCH, or SKF). Repeated measure ANOVA was used for open-field with interval as the within factor. Mixed models used the autoregressive-1 covariance matrix and

Kenward-Roger first order adjusted degrees of freedom. Litter was a random factor in these models.

Results

Protein and mRNA fluorescent imaging of PDE1B

PDE1B is in highest abundance in striatum (Fig. 1A), dentate gyrus (Fig. 1G), and substantia nigra (not shown) (Hufgard et al., 2017b). At higher magnification, PDE1B is found in cytoplasm with some clustering around nuclei where DAPI appears. There is colocalization of

PDE1B with DRD1A (Fig 1B) in the striatum and of PDE1B with NMDA (Fig. 1H) in the dentate gyrus. RNAScope was used to visualize the cellular localization of Pde1b message in association with Gaba and Glut (Fig. 2A, 3A, 4A), with Sert and Nmda (Fig. 2B, 3B, 4B), with

Dat and Drd1a (Fig. 2C, 3C, 4C), and with Drd2 and Gfap (Fig. 2D, 3D, 4D). The proximity of

Pde1b surrounding the nucleus was closest to Nmda, Drd1a, and Drd2+ cells in the caudate- putamen, dentate gyrus, and substantia nigra. Nucleus accumbens, amygdala, and ventral tegmental area showed similar results (not shown).

122

Presynaptic serotonin and dopamine specific PDE1B knockout mice

Pde1b is associated with serotonergic and dopaminergic pathways (Siuciak et al., 2007).

Accordingly, we tested the behavioral phenotype following KO of Pde1b in serotonergic and dopaminergic presynaptic cells using KOSert and KODat mice. WT, Flox, CreSert, and Pde1b-

KOSert (Fig. 5A-C) male littermates and WT, CreDat, and Pde1b-KODat (Fig. 5E-G) male littermates did not differ significantly in open-field (Fig. 5A, E), TST (Fig. 5B, F), or FST (Fig.

5C, G). The expression pattern of CreSert and CreDat were confirmed with td-tomato in WT and

Pde1b-KODat and Pde1b-KOSert mice that expressed ROSA (Fig. 5D, H).

DRD1-specific PDE1B KO mice

We tested for effects operating through the Drd1a receptor. Pde1b-KODrd1a mice showed no differential phenotype for grooming, weight, or health compared with WT littermates (not shown). The expression of the CreDrd1a driver was confirmed with td-tomato in WT and Pde1b-

KODrd1a mice expressing ROSA (Fig. 6E); Western blots confirmed the knockdown of PDE1B in the striatum of Pde1b-KODrd1a mice compared with controls [F(2,10)=4.2, p<0.05] (Fig. 6A). WT,

CreDrd1a, and Pde1b-KODrd1a male littermates did not differ in open-field (Fig. 6B) or FST (Fig.

6D); although note that the insertion of Cre did cause a trend towards increased immobility in both the CreDrd1a and Pde1b-KODrd1a mice. However, Pde1b-KODrd1a male mice showed a significant decrease in immobility in the TST [F(2,25)=5.4, p<0.05] (Fig. 6C) compared with WT and CreDrd1a mice.

Intrastriatal dopamine receptor 1 antagonist and agonist infusion

The mRNA, protein, and genetic experiments showed that Pde1b is present in postsynaptic dopamine-positive cells; therefore, to further examine the role of D1 in the immobility-resistant phenotype, we infused either SKF38393 (agonist) or SCH23390

(antagonist) bilaterally into the dorsal striatum and tested the mice 24 h later. No genotype or drug main effect on open-field activity was seen, but there was a genotype x drug interaction

[F(2,31.8)=5.4, p<0.01]. The antagonist SCH23390 caused a small reduction in activity in WT

123

mice but increased activity in Pde1b KOCMV mice [F(1, 25.3)=9.4, p<0.01] (Fig. 7A). By contrast, the agonist SKF38393 reversed the pattern of activity of Pde1b KOCMV mice relative to

WT mice (Fig. 7A). In TST with SAL infusion, the typical reduction in immobility time was seen in Pde1b KOCMV mice compared with WT mice (Hufgard et al., 2017b). There was a main effect of drug [F(1,31.5)=24.9, p<0.001] and a genotype x drug interaction [F(2,21.2)=5.1, p<0.05]

(Fig. 7B). The antagonist SCH23390 slightly increased immobility in WT mice and reduced it in

Pde1b KOCMV mice further compared with the SAL control. The agonist SKF38393 had the opposite effect, reversing the immobility of Pde1b KOCMV versus WT mice [SAL: F(1,21.1)=6.9, p<0.05; SCH: F(1,17.4)=35.4, p<0.001]. In FST with SAL infusion, the typical Pde1b KOCMV reduced immobility was again seen compared with WT mice (Hufgard et al., 2017b). Antagonist

SCH23390 infusion reduced immobility in WT mice and further reduced it in Pde1b KOCMV mice; the genotype difference was still significant [F(1,39.7)=5.1, p<0.05] (Fig. 7C, middle). When agonist SKF38393 was infused, group differences disappeared and both had nearly identical immobility times (Fig. 7C, right).

Discussion

Pde1b is expressed in areas high in dopaminergic innervation (Lakics et al., 2010,

Hufgard et al., 2017b). Hufgard et al. (2017b) showed that PDE1B expression is highest in caudate-putamen, nucleus accumbens, dentate gyrus, CA3, and substantia nigra. Here we show co-localization of PDE1B in cells expressing DRD1A and NMDA. PDE1B is cytosolic, although some aggregation around nuclei was noted (R Fusco and Giampa, 2015). Pde1b mRNA is also cytosolic in cells expressing Drd1a and Nmda. In addition, Pde1b mRNA is present in cells expressing Drd2 and has proximal localization with presynaptic Gaba and Glut mRNA expressing cells. These localizations were consistent in striatum, hippocampus, and substantia nigra. Although Pde1b mRNA was found in proximity with Gaba and Glut mRNA, they may be in adjacent cells rather than the same cells since cell boundaries in these

124

preparations are not definitive. The expression of Pde1b in both Drd1a and Drd2 positive cells supports the view that Pde1b has a role in postsynaptic dopamine transmission. The mRNA and protein expression patterns suggest that PDE1B may be involved in both the indirect, D2, and direct, D1, pathways, although we did not test the D2 association here. This would be consistent with published data describing cAMP/PKA control via PDEs in these pathways (Nishi and Snyder, 2010, Heckman et al., 2016). Nishi et al. suggested that PDE1B plays a larger role in the direct pathway while PDE10A plays a larger role in the indirect pathway (Nishi and

Snyder, 2010, Nishi et al., 2011).

Aside from the mRNA and protein expression of PDE1B, we also determined its functional role. The lack of the immobility phenotype in SERT and DAT specific Pde1b KO mice suggests that PDE1B does not interact with these presynaptic elements as evidenced by the lack of PDE1B::ROSADAT co-fluorescing cells and the lack of reduction in PDE1B::ROSASERT.

Pde1b KODrd1a mice showed a phenotype in TST similar to that of global Pde1b KOCMV mice verifying out previous findings. Additionally, Western blots showed a reduction of PDE1B but not a complete deletion in striatum. The PDE1B:ROSADRD1A expression showed dual fluorescing cells, suggesting that although PDE1B and DRD1A are expressed in some of the same cells there was incomplete Cre efficiency, and this may explain why TST was affected but not FST. These data implicate the role of PDE1B in D1 pathways, but we do not yet know about possible D2 involvement and therefore cannot rule out a possible effect on the indirect pathway.

The Pde1b KODrd1a mice immobility phenotype was only seen in TST, not FST. This difference, may be explained by the fact that the PDE1B reduction was not complete but the exact reason is unknown. Both dopamine and 5-HT are altered in Pde1b KO mice (Siuciak et al., 2007) suggesting that future experiments might investigate 5-HT receptor involvement. Nevertheless, the current data suggest that postsynaptic factors may offer an alternate way to modulate signaling for the treatment of depression (Kulkarni and Dhir, 2009).

125

To further explore the role of PDE1B in the direct dopamine pathway we used intrastriatal injections of a D1 agonist and antagonist. We confirmed that in the saline condition the Pde1b KOCMV mouse phenotype was present, i.e., decreased immobility in TST and FST

(Hufgard et al., 2017a). The antagonist SCH23390 caused a differential response in Pde1b

KOCMV mice versus WT mice in the open-field, i.e., it increased activity in Pde1b KOCMV mice, similar to what is seen after stimulant exposure (Reed et al., 2002, Ehrman et al., 2006, Siuciak et al., 2007). For example, Pde1b KO mice treated with methamphetamine have exaggerated hyperactivity compared with WT mice (Reed et al., 2002). Methamphetamine stimulates Ca2+ signaling in D1 receptor-positive cells (Sotty et al., 2009). When PDE1B is present these two

2+ pathways, Ca and D1 signaling, act to regulate one another, whereas when PDE1B is absent an exaggerated response to a D1 stimulus from methamphetamine was seen. When mice in the present experiment were infused with the agonist SKF38393, the opposite response was seen, i.e., a trend toward hypoactivity.

There was a genotype by drug interaction in TST with saline showing the unaltered phenotype and the antagonist SCH23390 causing an exaggerated decrease in immobility in both TST and FST, suggesting that PDE1B inhibition in combination with D1 antagonism augments the effects, even though D1 antagonism does not induce immobility by itself.

Increased cGMP can cause a depression-like state, and antagonism of an increase can reverse the effect (Kulkarni and Dhir, 2009), suggesting that a balance of cGMP is important for mood stability. Knocking out Pde1b increases cAMP/cGMP so that in conjunction with the antagonism of D1, the combination decreases cAMP/cGMP that leads to a more balanced level of cAMP and cGMP that enhances the phenotype. A similar additive affect was seen when Pde1b KO mice were exposed to SSRIs or NDRIs (Hufgard et al., 2017a). Conversely, D1 agonists mitigated the genotype specific decrease in immobility causing Pde1b KO mice in both TST and FST to be similar to that of WT littermates, suggesting increased production of cGMP. Although

126

PDE1B is present in D1 expressing cells, dopamine/Drd1 interactions can be independent of

Ca2+/calmodulin-stimulated PDE1B expression allowing for a synergistic phenotype (Fig. 8).

Interestingly, neither SCH23390 nor SKF38393 produced locomotor, TST, or FST changes as reflected by a lack of overall drug effects. The behavioral changes were all drug by genotype interactions. Why this occurred is not clear but could be related to the time interval between drug infusion and testing (24 h). A time-course experiment will be needed to resolve this.

Protein and mRNA expression show that Pde1b localizes to postsynaptic cells, i.e., to direct and indirect dopamine pathways and to glutamatergic cells. Postsynaptic specificity was established by the lack of phenotype seen in presynaptic 5-HT and dopamine specific Pde1b combination KO mice and the TST immobility-resistant phenotype in Pde1b KODrd1a mice. The intrastriatal injections of D1 antagonist and agonist findings are consistent with these data and implicate the direct dopamine pathway; however, further investigations into PDE1B involvement in the indirect pathway are needed. The data suggest that PDE1B inhibition acts as an on- demand modulator of second messengers and a potential target for modulating depression- related symptoms (Fig. 8).

127

References

Artigas F, Romero L, de Montigny C, Blier P (1996) Acceleration of the effect of selected

antidepressant drugs in major depression by 5-HT 1A antagonists. Trends in

neurosciences 19:378-383.

Cryan JF, Mombereau C, Vassout A (2005a) The tail suspension test as a model for assessing

antidepressant activity: review of pharmacological and genetic studies in mice.

Neuroscience & Biobehavioral Reviews 29:571-625.

Cryan JF, Valentino RJ, Lucki I (2005b) Assessing substrates underlying the behavioral effects

of antidepressants using the modified rat forced swimming test. Neuroscience &

Biobehavioral Reviews 29:547-569.

Ehrman L, Williams M, Schaefer T, Gudelsky G, Reed T, Fienberg A, Greengard P, Vorhees C

(2006) Phosphodiesterase 1B differentially modulates the effects of methamphetamine

on locomotor activity and spatial learning through DARPP32‐dependent pathways:

evidence from PDE1B‐DARPP32 double‐knockout mice. Genes, Brain and Behavior

5:540-551.

Fava M, Davidson KG (1996) Definition and epidemiology of treatment-resistant depression.

Psychiatric Clinics of North America 19:179-200.

Furukawa T, Yoshimura R, Harai H, Imaizumi T, Takeuchi H, Kitamura T, Takahashi K (2009)

How many well vs. unwell days can you expect over 10 years, once you become

depressed? Acta Psychiatrica Scandinavica 119:290-297.

Greengard P, Allen PB, Nairn AC (1999) Beyond the dopamine receptor: the DARPP-32/protein

phosphatase-1 cascade. Neuron 23:435-447.

Greist JH, Mundt JC, Kobak K (2002) Factors contributing to failed trials of new agents: Can

technology prevent some problems? Journal of Clinical Psychiatry 63:8-13.

128

Hautman ER, Kokenge AN, Udobi KC, Williams MT, Vorhees CV, Skelton MR (2014) Female

mice heterozygous for creatine transporter deficiency show moderate cognitive deficits.

Journal of inherited metabolic disease 37:63-68.

Hebb AL, Robertson HA (2007) Role of phosphodiesterases in neurological and psychiatric

disease. Current opinion in pharmacology 7:86-92.

Heckman PR, van Duinen MA, Bollen EP, Nishi A, Wennogle L, Blokland A, Prickaerts J (2016)

Phosphodiesterase inhibition and regulation of dopaminergic frontal and striatal

functioning: clinical implications. International Journal of Neuropsychopharmacology

pyw030.

Holtzheimer PE, Nemeroff CB (2006) Advances in the treatment of depression. NeuroRx 3:42-

56.

Hufgard JR, Williams MT, Skelton MR, Grubisha O, Ferreira FM, Sanger H, Wright ME, Reed-

Kessler TM, Rasmussen K, Duman RS, Vorhees CV (2017a) Phosphodiesterase-1b

(Pde1b) knockout mice are resistant to forced swim and tail suspension induced

immobility and show upregulation of Pde10a. Psychopharmacology.

Hufgard JR, Williams MT, Vorhees CV (2017b) Phosphodiesterase-1b deletion confers

depression-like behavioral resistance separate from stress-related effects in mice.

Jeon Y, Heo Y-S, Kim C, Hyun Y-L, Lee T, Ro S, Cho J (2005) Phosphodiesterase: overview of

protein structures, potential therapeutic applications and recent progress in drug

development. Cellular and Molecular Life Sciences 62:1198-1220.

Kesler R, Berglund P, Dernler O, Jin R, Koretz D, Merikangas K, Rush A, Walters E, Wang P

(2003) National Comorbidity Survey Replication: The epidemiology of major depressive

disorder: results from the National Comorbidity Survey Replication (NCS-R). JAMA

289:3095-3105.

Konradi C, Cole RL, Heckers S, Hyman SE (1994) Amphetamine regulates gene expression in

rat striatum via transcription factor CREB. Journal of Neuroscience 14:5623-5634.

129

Kulkarni SK, Dhir A (2009) Current investigational drugs for major depression. Expert opinion on

investigational drugs 18:767-788.

Lakics V, Karran EH, Boess FG (2010) Quantitative comparison of phosphodiesterase mRNA

distribution in human brain and peripheral tissues. Neuropharmacology 59:367-374.

Mohammadi M, Nasehi M, Zarrindast M-R (2015) Modulation of the effects of the cannabinoid

agonist, ACPA, on spatial and non-spatial novelty detection in mice by dopamine D1

receptor drugs infused into the basolateral amygdala. Behavioural brain research

280:36-44.

Montminy M (1997) Transcriptional regulation by cyclic AMP. Annual review of biochemistry

66:807-822.

Nair A, Vaidya V (2006) Cyclic AMP response element binding protein and brain-derived

neurotrophic factor: molecules that modulate our mood? Journal of biosciences 31:423-

434.

Nishi A, Kuroiwa M, Shuto T (2011) Mechanisms for the modulation of dopamine D1 receptor

signaling in striatal neurons. Frontiers in neuroanatomy 5.

Nishi A, Snyder GL (2010) Advanced research on dopamine signaling to develop drugs for the

treatment of mental disorders: biochemical and behavioral profiles of phosphodiesterase

inhibition in dopaminergic neurotransmission. Journal of pharmacological sciences

114:6-16.

Porsolt RD, Bertin A, Blavet N, Deniel M, Jalfre M (1979) Immobility induced by forced

swimming in rats: effects of agents which modify central catecholamine and serotonin

activity. European journal of pharmacology 57:201-210.

R Fusco F, Giampa C (2015) Phosphodiesterases as therapeutic targets for Huntington’s

disease. Current pharmaceutical design 21:365-377.

Reed TM, Repaske DR, Snyder GL, Greengard P, Vorhees CV (2002) Phosphodiesterase 1B

knock-out mice exhibit exaggerated locomotor hyperactivity and DARPP-32

130

phosphorylation in response to dopamine agonists and display impaired spatial learning.

The Journal of neuroscience 22:5188-5197.

Siuciak JA, McCarthy SA, Chapin DS, Reed T, Vorhees C, Repaske D (2007) Behavioral and

neurochemical characterization of mice deficient in the phosphodiesterase-1B (PDE1B)

enzyme. Neuropharmacology 53:113-124.

Snyder GL, Fienberg AA, Huganir RL, Greengard P (1998) A dopamine/D1 receptor/protein

kinase A/dopamine-and cAMP-regulated phosphoprotein (M r 32 kDa)/protein

phosphatase-1 pathway regulates dephosphorylation of the NMDA receptor. Journal of

Neuroscience 18:10297-10303.

Sotty F, Montezinho LP, Steiniger‐Brach B, Nielsen J (2009) Phosphodiesterase 10A inhibition

modulates the sensitivity of the mesolimbic dopaminergic system to d‐amphetamine:

involvement of the D1‐regulated feedback control of midbrain dopamine neurons.

Journal of neurochemistry 109:766-775.

Tanis KQ, Duman RS (2007) Intracellular signaling pathways pave roads to recovery for mood

disorders. Annals of medicine 39:531-544.

Vijayakumar L (2005) Suicide and mental disorders in Asia. International Review of Psychiatry

17:109-114.

Wang L-Y, Salter M, MacDonald J (1991) Regulation of kainate receptors by cAMP-dependent

protein kinase and phosphatases. Science 253:1132.

Zarrindast MR, Dorrani M, Lachinani R, Rezayof A (2010) Blockade of dorsal hippocampal

dopamine receptors inhibits state-dependent learning induced by cannabinoid receptor

agonist in mice. Neuroscience research 67:25-32.

Zhang Y, Schlussman SD, Ho A, Kreek MJ (2003) Effect of chronic “binge cocaine” on basal

levels and cocaine‐induced increases of dopamine in the caudate putamen and nucleus

accumbens of c57bl/6j and 129/j mice. Synapse 50:191-199.

131

Zhu J, Mix E, Winblad B (2001) The antidepressant and antiinflammatory effects of rolipram in

the central nervous system. CNS drug reviews 7:387-398.

132

Figures

Figure 1. PDE1B co-localizes with DRD1A and NMDA but not DAT or SERT. A. Striatum staining for PDE1B (green), DAT (red), and DRD1A (purple) scale = 1000 µm. B. Striatum staining for PDE1B (green), DAT (red), and DRD1A (purple) scale = 50 µm. Arrows point to cells showing colocalization of DRD1A and PDE1B. C. Striatum DAPI. D. Striatum PDE1B. E.

Striatum DAT. F. Striatum DRD1A. G. Hippocampus staining for PDE1B (green), SERT (red), and NMDA (purple) scale = 1000 µm. H. Hippocampus staining for PDE1B (green), SERT

(red), and NMDA (purple) scale = 50 µm. Arrows point to cells showing colocalization of NMDA and PDE1B. I. Hippocampus DAPI. J. Hippocampus PDE1B. K. Hippocampus SERT. L.

Hippocampus NMDA.

133

Figure 2. Pde1b localizes in the striatum to the cytosolic space of cells also expressing Drd1a,

Drd2, and Nmda. Arrows point to mRNA fluorescence. A. Nuclei = Blue, Pde1b = green, Gaba

= red, Glut = purple. B. Nuclei = Blue, Pde1b = green, Nmda = red, Sert = purple. C. Nuclei =

Blue, Pde1b = green, Dat = red, Drd1a = purple. D. Nuclei = Blue, Pde1b = green, Drd2 = red,

Gfap = purple. E. Pde1b KO mouse Nuclei = Blue, Pde1b = green, Polr2a = red (Positive

Control). F. Nuclei = Blue, Pde1b = green, DapB = red (negative control). Scale = 10 µm.

134

Figure 3. Pde1b localizes in the hippocampus to the cytosolic space of cells also expressing

Drd1a, Drd2, and Nmda. Arrows point to mRNA fluorescence. A. Nuclei = Blue, Pde1b = green, Gaba = red, Glut = purple. B. Nuclei = Blue, Pde1b = green, Nmda = red, Sert = purple.

C. Nuclei = Blue, Pde1b = green, Dat = red, Drd1a = purple. D. Nuclei = Blue, Pde1b = green,

Drd2 = red, Gfap = purple. Scale = 10 µm.

135

Figure 4. Pde1b localizes in the substantia nigra to the cytosolic space of cells also expressing

Drd1a, Drd2, and Nmda. Arrows point to mRNA fluorescence. A. Nuclei = Blue, Pde1b = green, Gaba = red, Glut = purple. B. Nuclei = Blue, Pde1b = green, Nmda = red, Sert = purple.

C. Nuclei = Blue, Pde1b = green, Dat = red, Drd1a = purple. D. Nuclei = Blue, Pde1b = green,

Drd2 = red, Gfap = purple. Scale = 10 µm.

136

137

Figure 5. Neither Sert nor Dat specific KO of Pde1b induced an antidepression-like phenotype in comparison with WT littermates. A. Locomotor activity (WT n=20, Flox n=13, CRESERT n=12,

KOSERT n=9). B. TST (WT n=20, Flox n=13, CRESERT n=12, KOSERT n=9). C. FST (WT n=19,

Flox n=12, CRESERT n=13, KOSERT n=9). D. Hippocampus of KOSERT mouse Blue = DAPI, Green

= PDE1B, Red = ROSA. Arrows point to cells showing colocalization of SERT or DAT and

PDE1B. D’. PDE1B D’’. ROSA. E. Locomotor activity (WT n=18, CREDAT n=9, KODAT n=14).

F. TST (WT n=12, CREDAT n=8, KODAT n=16). G. FST (WT n=14, CREDAT n=9, KODAT n=18).

H. Striatum of KODat mouse Blue = DAPI, Green = PDE1B, Red = ROSA H’. PDE1B. H’’.

ROSA. Scale = 50 µm.

138

139

Figure 6. Drd1a targeted KO of Pde1b produced a decrease in immobility in TST when compared with WT littermates. A. Western blot analysis of PDE1B in the striatum and hippocampus, respectively (STR: WT n=3, CREDAT n=4, KODAT n=4; HIP: WT n=4, CREDAT n=4,

KODAT n=6). B. Locomotor activity (WT n=10, CREDAT n=9, KODAT n=9). C. TST (WT n=9,

CREDAT n=9, KODAT n=8). D. FST (WT n=10, CREDAT n=9, KODAT n=9). E. Striatum of KODrd1a mouse Blue = DAPI, Green = PDE1B, Red = ROSA. Arrows point to cells showing colocalization of DRD1A and PDE1B. E’. PDE1B E’’. ROSA. Scale = 50 µm. * p < 0.05.

140

141

Figure 7. Pde1b KOCMV mice exhibit a decrease in immobility time in TST and FST that is exacerbated after D1 antagonism but mitigated after D1 agonism. A. Locomotor activity (WT:

Saline n=8, SCH n=8, SKF n=8; KO: Saline n=9, SCH n=8, SKF n=8). B. TST (WT: Saline n=8,

SCH n=8, SKF n=8; KO: Saline n=9, SCH n=8, SKF n=8). C. FST (WT: Saline n=8, SCH n=8,

SKF n=8; KO: Saline n=9, SCH n=8, SKF n=8). D. D1 cell specific activation and regulation of cAMP. Dopamine signaling increases intracellular cAMP levels through adenylyl cyclase stimulation. Increases in cAMP and PKA signaling can lead to upregulation of CREB and

DARPP-32. PDE1B is dependent on Ca2+/calmodulin signaling and upon stimulation it regulates cAMP and therefore the downstream effectors. * p < 0.05, ** p < 0.01, and *** p <

0.001.

142

Figure 8. D1 cell specific activation and regulation of cAMP. Dopamine signaling increases intracellular cAMP levels through adenylyl cyclase stimulation. Increases in cAMP and PKA signaling can lead to upregulation of CREB and DARPP-32. PDE1B is dependent on

Ca2+/calmodulin signaling and upon stimulation it regulates cAMP and therefore the downstream effectors.

143

Acknowledgements

This work was supported by NIH T32 ES007051 and funds from Cincinnati Children’s Hospital

Research Foundation

144

CHAPTER 5

Discussion

145

Secondary messenger regulation in depression

Secondary messengers and their regulation are linked to depression, predominantly through their control over neuroplasticity signaling pathways. Cyclic adenosine monophosphate

(cAMP) regulates protein kinase (PK) type A and G along with exchange proteins activated by cAMP (Epac). Epacs are known to drive many cellular functions such as: apoptosis, cellular proliferation/differentiation/migration, immune responses, muscle contraction, and some neuronal functions (Halene and Siegel, 2007). In the brain, Epacs regulate the Ras/MAPK/ERK pathway that facilitates neuroplasticity. cAMP regulation of PKA and PKG leads to increased neuroplasticity via the enhancement of cAMP response element binding (CREB) and brain derived neurotrophic factor (BDNF) that are transcription factors involved in neurogenesis.

Diminished PKA and CREB activity were found in the prefrontal cortex of suicide patients

(Marazziti et al., 2009). PKA is believed to provoke neurotransmitter increases at central synapses along with potentiating norepinephrine neuroprotection in dopaminergic neurons

(Marazziti et al., 2009). Cyclic guanosine monophosphate (cGMP) has been implicated in neuroplasticity regulation through its regulation in PKG/CREB/BDNF and cyclic nucleotide gated channels (Halene and Siegel, 2007). Taken together these findings demonstrate the importance of second messengers in depression phenotypes.

The role of cAMP and cGMP in depression has provided the impetus to target second messenger regulation as a treatment of depression. Phosphodiesterases (PDE) are the primary phosphatases controlling intracellular levels of cAMP and cGMP. PDEs are a superfamily (1-

11) of over 100 distinct proteins with varying cellular and tissue distribution and specificity toward cAMP and/or cGMP. The significant number of distinct PDEs within the human body allows for diverse control over many systems. Of the many PDEs a few are implicated in depression and its treatment. In a Mexican American population, polymorphisms and single nucleotide polymorphisms (SNPs) in PDEs were related to major depressive disorder (MDD).

146

The diagnosis of MDD was associated with polymorphisms in Pde9a and Pde11a (Wong et al.,

2006). Unfortunately, an independent study was not able to replicate these results (Cabanero et al., 2009). Wong et al. related SNPs in Pde1a and Pde11a to positive remission rates following antidepressant treatment (Wong et al., 2006). This further supports there is a relationship between PDE regulation and depression.

PDE1B regulation in activity and depression-like behaviors

PDE4 and in particular PDE4D has been studied as a potential target for antidepressant treatment. These investigations show a connection between PDE4D inhibition and decreases in depression-like behaviors. In humans, PDE4D inhibitors have antidepressant activity but also emetic side effects. The successes of PDE4 inhibitors suggest that other PDEs may have antidepressant activity. PDE1B and PDE10A are highly expressed in dopaminergic neurons of the striatum, a brain region linked to depression. PDE1B and PDE10A are found in medium spiny neurons, and both are associated with dopamine- and cAMP-regulated phosphoprotein of

Mr 32 kDa (DARPP-32). This overlap in localization has led to studies attempting to differentiate the distinct roles of these PDEs. Nishi and Snyder suggest that although both proteins are present in the same cell populations, PDE1B is primarily regulating striatonigral neurons and

PDE10A striatopallidal neurons (Nishi and Snyder, 2010). Pde1b KO mice have increased activity following AMPH and MK-801 stimulation implicating increased dopamine receptor dopamine 1 (D1) motor function while responses to PDE10A inhibitors mimic antipsychotics that act though D2 antagonism (Reed et al., 2002, Ehrman et al., 2006, Siuciak et al., 2007, Bateup et al., 2008).

We explored the relationship of PDE1B and PDE10A in wild-type (WT) mice after stress exposure. PDE1B was increased after acute and chronic stress, while PDE10A was decreased in response to only chronic stress. The increase in PDE1B in the striatum is indicative of a relationship between PDE1B and stress. If PDE1B functions like a D1 antagonist then the

147

increase in PDE1B following stress is indicative of decreased striatal dopaminergic function.

Dopamine dysfunction is implicated in the pathophysiology of depression, and rodent models show dopamine depletion following learned helplessness stress (Dunlop and Nemeroff, 2007).

Nestler and Carlezon hypothesized that manipulations of proteins-like CREB and BDNF may be the etiology of dopamine reward-related depression in animal models (Nestler and Carlezon,

2006).

The decrease of PDE10A after 21 days of chronic stress but not after acute stress suggests that PDE10A may be responding to changes in PDE1B as opposed to a direct response to the chronic stress. Since PDE1B is upregulated and antagonizing dopaminergic functions, changes in PDE10A may be compensatory by upregulating D2 signaling. Mouse studies of Pde10a knockouts (KO) suggest that this compensation may be partial but not sufficient to fully prevent depressive-like behaviors. Pde10a KO mice show no differences in anxiety and depression-like behaviors when compared with WT mice (Siuciak et al., 2006b).

Although no functional compensation is seen in response to decreases in PDE10A, this does not rule out a relationship between PDE1B and PDE10A. When Pde1b is deleted there is upregulation of Pde10a mRNA. Pde10a KO mice show decreased baseline activity levels and reduced sensitivity to N-methyl-D-aspartate (NMDA) antagonists, an indication that PDE10A plays a role in the inhibition of D2 motor function. We hypothesize that the hyperactivity phenotype previously reported (Ehrman et al., 2006) and seen here may be caused by increases in PDE10A expression that result from PDE1B decreases.

The mild hyperactivity and reduced immobility were shown to be distinct and non- confounding in these mice. Pde1b KO mice did not show increased swimming activity despite their mild diurnal and larger nocturnal hyperactivity. The repeated exposure to water in the open swim test also did not alter the decreased immobility in either tail suspension (TST) or forced swim tests (FST), suggesting serial testing within this mouse model does not extinguish

148

the phenotype. These data suggest that Pde1b KO mice have an activity and immobility phenotype. The activity phenotype might be further explored by using a PDE10A inhibitor or creating a Pde1b;10a double KO mouse.

Aside from the activity phenotype in Pde1b KO mice, they also exhibit reduced immobility in the TST and FST as noted. This decrease in immobility is indicative of an antidepressant-like phenotype and strengthens the evidence of PDE1B involvement in depression. Both TST and FST were used despite their apparent redundancies so as not to misappropriate these mice as antidepressant-like. FST is a test of behavioral despair that measures swimming/immobility behavior and attributes an immobile posture as a sign of depression-like behaviors (Cryan et al., 2005a, Cryan et al., 2005b). Aside from just

‘persistence in escape-directed behavior’, FST has also been thought to represent entrapment behaviors seen in depressed patients (Cryan et al., 2005a, Cryan et al., 2005b). FST has been shown to be efficacious in testing serotonin targeting reuptake inhibitors because increased serotonin neurotransmission has been associated with increased swimming behavior (Cryan et al., 2005b). TST, although also measuring immobility time, utilizes hemodynamic stress and is thought to predominantly measure the mouse’s reluctance to continue attempting to escape

(Cryan et al., 2005a). Both TST and FST have been validated as a mode of testing for antidepressant activity along with depression-like behaviors in genetically depressed mice

(Cryan et al., 2005a). Therefore, TST and/or FST are capable of assessing a depressed or antidepressant phenotype in mice, and the utilization of both tests minimizes a false positive antidepressant-like phenotype assessment.

Within our studies, depressive-like mice have increased levels of PDE1B and mice resistant to the induction of depressive-like behaviors have Pde1b deleted. When PDE1B is removed, D1 signaling is enhanced, increasing the amount of intracellular cAMP leading to amplified neuroplasticity. Although the present studies only explore the functional relationship

149

between PDE1B and depression, future studies may be able to relate PDE1B inhibition directly to increased CREB and BDNF levels.

PDE1B phenotype differences in male and female mice

There are significant differences in depression diagnosis rates of men and women, between 1.5 and 2.5 times the prevalence in women (Fava and Kendler, 2000). For this reason both male and female constitutive Pde1b KO and Pde1b KOCMV mice were tested for changes in activity, TST, and FST. Reed et al. (2002) showed female KO mice had increased activity in comparison with WT females and all male mice in locomotor activity prior to AMPH administration. We saw a similar increase in activity only in the female Pde1b KOCMV mice.

Studies in rodents show that females are more sensitive to acute and novel environment test procedures by displaying increased activity during these tests and this may explain why we consistently see increased activity in females but not males (Andersen, 2003). The increased activity phenotype is evident in males following stimulant administration or during the dark cycles of 48 h home cage activity testing. These data suggest that the hyperactivity phenotype is present in both male and female Pde1b KO mice, but future studies regarding PDE1B and activity will need to take into account these sex differences.

When female mice were tested in TST and FST they also showed a similar but not identical phenotype as the male mice. Constitutive Pde1b KO female mice had decreased immobility time in comparison with WT females in FST while the Pde1b KOCMV female mice did not. Although this difference in FST immobility may be due to the difference in exon targets for the two models it is more likely that these differences are related to estrogen cycling. In neither experiment, constitutive nor conditional, were estrogen levels monitored or controlled for in these females this in combination with number of subjects that was 12 or below, it is likely that female mice had varying estrogen levels. Research has shown that female rodents have heightened activity corresponding to low estrogen levels (Andersen, 2003). This may also be

150

evidenced by the apparently decreased immobility of WT females compared with males. These experiments were run independently and therefore cannot be directly compared, but reduced immobility time in WT female mice may be a likely cause for the differences of these two models. Although, female Pde1b KOCMV mice did show a reduction in immobility time in TST similarly to that seen in male mice, suggesting the resistance to acute-stress induced depression-like behaviors phenotype is still present. Because of this sex difference only male mice were further examined within these experiments. Future studies including female mice will need to take estrogen cycling into account as PDE1B inhibition may have differing effects depending on estrogen levels.

PDE1B reduction at developmental time points

The embryonic KO of Pde1b elicits an antidepressant-like phenotype as evidenced by a reduction in immobility time in TST and FST. We examined the relationship of Pde1b knockdown (KD) at various developmental time points to assess if there was a critical period for efficacy. The forebrain dominant calcium/calmodulin-dependent protein kinase II alpha

(CaMKIIα), a known regulator of PDE1B, Cre driver, B6.Cg- Tg(Camk2a-cre/ERT2)1Aibs/J, was used to create a tamoxifen inducible Pde1b mouse (Tsien et al., 1996, Madisen et al., 2010).

Although CaMKIIα is primarily restricted to the CA1, CaMKIIα expression is also seen in the

CA3, dentate gyrus, cortex, and striatum (Tsien et al., 1996, Dragatsis and Zeitlin, 2000,

Casanova et al., 2001). Specifically inhibition of CaMKIIα in medium spiny neurons of the striatum has been linked to the reduction in functional excitatory synapses and enhancements in intrinsic excitability (Klug et al., 2012). The localization to the hippocampus and striatum, more specifically medium spiny neurons, implicated CaMKIIα as a promising cre driver for KD of

Pde1b.

Aside from regional localization, CaMKIIα is detected as early as embryonic day 18.5 with significant recombination at postnatal day (P) 5 (Dragatsis and Zeitlin, 2000). Another

151

study showed gradual increases of CaMKIIα from P3 to P24, reaffirming that CaMKIIα is present at developmental ages (Casanova et al., 2001). Since CaMKIIα peaks around three weeks postnatally, expressing cre driven by CaMKIIα during adulthood should also result in a reduction of Pde1b (Klug et al., 2012). Therefore, the CaMKIIα cre driver would allow for directed KD of Pde1b at various developmental time points following tamoxifen exposure.

When mice were dosed with tamoxifen from P0 to P4 this corresponds to early detection of CaMKIIα in the hippocampus and explains why reductions in PDE1B were only evident in the hippocampus (Casanova et al., 2001). Early hippocampal development is dependent on serotonin signaling that promotes neurogenesis, dendritic development, cellular migration, and dopamine terminal growth (Whitaker-Azmitia, 2001). The reduction of hippocampal PDE1B during early postnatal development may play a role in serotonin signaling and cause long term alterations to neurogenesis. When PDE1B was KD during this perinatal period we saw a reduction in immobility time specific to FST, this suggests that these early life reductions may be altering serotonin signaling and promoting an antidepressant-like phenotype.

Tamoxifen exposure from P32-P36 resulted in a reduction of PDE1B in both the striatum and hippocampus in adulthood. This developmental time point coincides with significant amounts of synaptic pruning taking place within the hippocampus and striatum (Andersen,

2003, Khalaf-Nazzal and Francis, 2013). The periadolescent decrease in PDE1B may be altering the amount or specificity of pruning taking place during this developmental stage and this could be tested by measuring the volume of the striatum in mice exposed to tamoxifen or corn oil or by looking at synaptic densities in these brain regions. Interestingly, D1 and D2 receptor densities in the caudate putamen also experience their peak expression during this periadolescent period suggesting that PDE1B downregulation may be altering dopamine receptor availability. Striatal and hippocampal PDE1B were decreased in mice exposed to tamoxifen during this periadolescent time point that resulted in decreased immobility in both

152

TST and FST, suggesting that periadolescent decreases in PDE1B result in antidepressant-like behaviors in adulthood.

Dosing with tamoxifen beginning at P60 resulted in a decrease of PDE1B within both the striatum and hippocampus. Our data show a trend toward a reduction of PDE1B in the hippocampus but not in the striatum, but our fluorescent ROSA expression suggests that the cre driver is being expressed in these regions of interest. We propose that the seven day wait between the completion of tamoxifen dosing and the beginning of testing was not sufficient to allow for a reduction in PDE1B. As a result we cannot claim that a reduction of PDE1B during adulthood changes the amount of time spent immobile in TST or FST.

Reduction in PDE1B in the striatum and hippocampus affects adult depressive-like behaviors when knocked out during development. These periods of susceptibility may be indicative of PDE1B inhibition being a useful therapeutic in childhood and adolescent depression. The failure to reduce PDE1B in adult mice may also suggest that PDE1B clearance takes longer than 7 days and targeting PDE1B as an adult may not result in immediate reduction of MDD side effects. Current antidepressants can take up to 6 weeks to produce effects suggesting that further testing of adult PDE1B KD is needed to explore the efficacy of treatment in adulthood and lag time between inhibition and side effect reduction.

PDE1B regulation in dopaminergic pathways

The relationship between PDE1B and dopaminergic signaling is reinforced by the expression patterns of PDE1B, the recapitulation of the reduced immobility phenotype when

Pde1b is only knocked out in D1 specific cells, and also by the lack of phenotype following learned helplessness, a severe stressor. We used both fluorescent mRNA and protein imaging to show localization of Pde1b to dopaminergic neurons. The fluorescent imaging recapitulated data that Pde1b expression is highest in the striatum (Bender and Beavo, 2006) with significant

153

presence in the substantia nigra. Pde1b mRNA and protein are also co-localized with D1 neurons. This expression pattern is consistent with the role of PDE1B in the striatonigral pathway (Nishi and Snyder, 2010).

Following confirmation of PDE1B in D1 expressing neurons, we showed that Pde1b KO in D1 cells was sufficient to reproduce the immobility phenotype in TST but not in FST. The phenotype was absent in dopamine transporter specific PDE1B KO mice, and there was no colocalization of PDE1B and the dopamine transporter, thus showing PDE1B is localized to postsynaptic neurons. Interestingly, the reduced immobility phenotype was only present in TST and not FST in the D1 specific Pde1b KO mice. This may reflect limitations to the sensitivity of the FST. The FST was developed as a test of “behavioral despair” and continues to be the leading assay for antidepressant screening, but has limitations for some applications, particularly dopamine-related antidepressants. Because of these limitations the TST was developed, although both tests examine immobility as a readout they may have somewhat different neurochemical basis for a similar functional response (Kulkarni and Dhir, 2009).

The lack of phenotype in Pde1b KO mice following severe stress induced by learned helplessness may be explained by PDE1B having a primary role in dopaminergic and not in serotonergic pathways. Learned helplessness, although regarded as a method of inducing depressive-like behaviors in rodent models, has failed to be reversed or blocked by dopamine targeting antidepressants in C56BL/6J mice (Shanks and Anisman, 1989). This suggests the non-differential response in global Pde1b KO mice may be caused by a dominant dopaminergic mechanism induced by learned helplessness, although it may also be indicative of limitations to the types or amount of stress PDE1B can buffer. This hypothesis could be tested by exploring the response of Pde1b KO mice in different stress paradigms: corticosterone or lipopolysaccharide exposure, thermodynamic dysregulation by heat or cold exposure, chronic

154

variable stress, or olfactory bulbectomy. These models have been used to test other PDE inhibitors and their ability to reduce depressive-like behaviors with some success.

Reduction in PDE1B in the striatum and hippocampus affects adult depressive-like behaviors when knocked out during development but the adult tamoxifen experiment was inconclusive. As a result, our data cannot address the question of whether there is a critical period in development for when PDE1B reduction leads to the immobility resistant adult phenotype. The findings may hint that the downstream mechanisms, most likely CREB and

BDNF, underlying the reduced depressive-like behaviors, require PDE1B reduction for longer amounts of time than we tested. Current antidepressants can take up to 6 weeks to produce effects while PDE1B deletion in adult mice was evaluated after only 7 days.

PDE1B in serotonergic and glutamatergic signaling

Since the constitutive and global Pde1b KO mice have reduced immobility in both TST and FST it suggests that PDE1B may be acting in several neurotransmitter pathways. Suiciak et al. hypothesized multiple neurotransmitters are targeted since Pde1b KO mice had changes in serotonin and dopamine levels (Siuciak et al., 2007). The presence of PDE1B in the dentate gyrus and CA3 of the hippocampus suggests a possible role of PDE1B in serotonin signaling.

However, Pde1b KO in serotonin transporter (SERT) cells was insufficient to reproduce the reduced immobility phenotype. The protein expression of PDE1B and SERT suggest that

PDE1B and SERT are not expressed in the same cells. Along with the fluorescent ROSASERT expression showing that PDE1B was not knocked out in co-expressing cells, implicates that

PDE1B was not reduced in the KOSERT mice resulting in the lack of reduced immobility phenotype seen in the global KO mice. Postsynaptic serotonin was not explored in the present studies. If Pde1b KO in postsynaptic serotoninergic cells is sufficient to recapitulate the reduced immobility phenotype it would lead to the proposition that PDE1B inhibition is a dual pathway antidepressant. The insufficiency of current antidepressants in 60-67% of patients has led to

155

combination therapies. If PDE1B is able to act on several mechanisms involved in depression, it might increase the efficacy of current antidepressants. Hence, PDE1B is a possible target for combination therapies based on its additive effect with the selective serotonin reuptake inhibitor fluoxetine, the norepinephrine dopamine reuptake inhibitor bupropion, or the D1 antagonist,

SCH23390. In each of these cases Pde1b KO mice had decreased immobility in TST and/or

FST with additive effects in combination. Most combination therapies target the reuptake of multiple monoamines. Two studies explored the potential of PDE inhibition in combination with reuptake inhibitor therapies. PDE4 and serotonin/norepinephrine reuptake inhibitors or SSRIs selective serotonin reuptake inhibitors (SSRI) enhanced the beneficial effects on depressive-like behaviors in animal models (Cashman and Ghirmai, 2009, Cashman et al., 2009). These studies speak to the feasibility of targeting both pre- and postsynaptic pathways to gain additional antidepressant benefits.

The co-localization studies showed a relationship between Pde1b and glutamatergic cells. Specifically, Pde1b mRNA and protein were present in the same cells as NMDA receptors. Subcellular cross-talk between glutamate, serotonin, and dopamine has been reported in relation to MDD (Drago et al., 2011). Specifically, Drago et al. postulated that unbalanced striatal dopamine regulation leads to improper NMDA signaling and ultimately to

MDD (Drago et al., 2011). Interestingly, Pde1 is linked to the control of several glutamatergic receptors. α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors in D1 cells are regulated by the cross-talk between Pde1 and Pde2 (Song et al., 2016). Pharmacological inhibition of Pde1 elevated intracellular cGMP levels and stimulated Pde2; Pde2 activation, in turn, results in a decrease of surface AMPA receptors (Song et al., 2016).

PDE1B regulation by Ca2+/calmodulin-dependent protein kinase II (CaMKII) also suggests a possible glutamatergic relationship (Figure 1). CaMKII phosphorylates AMPA receptors (Drago et al., 2011) while blocking PDE1B or inhibition of PDE1B leads to an increase

156

of cGMP stimulation of PDE2 then decreases AMPA. This regulation is controlled by dopamine signaling since dopamine has an inhibitory effect on CaMKII (Drago et al., 2011).

Antidepressant treatment increases AMPA receptor levels suggesting an interplay between surface receptors and depressive-like behaviors (Bleakman et al., 2007). The interplay between

PDE1B, PDE2, dopamine, and AMPA receptors may partially explain why the D1 agonist

SKF38393 in combination with Pde1b KO induced the immobility phenotype in TST and FST.

We hypothesize that the removal of PDE1B causes a down regulation of outward-facing AMPA receptors that is further decreased by SKF38393 through inhibition of CaMKII. This hypothesis could be tested by adding a PDE2 inhibitor-treated group and testing if PDE2 inhibition blocks the reversal of PDE1B reduced immobility caused by intra-striatal SKF38393.

Conclusion

The localization of PDE1B to D1 and D2 medium spiny neurons in the striatum lends itself to therapeutic targeting. The reduction of immobility in depressive-like behavioral assays supports this since when Pde1b is knocked out only in D1 cells, dopamine changes appear to be the primary mechanism behind the antidepressant-like phenotype. The presence of PDE2 and

PDE10A in the same cell types leads to the idea of compensatory or functional relationships between PDEs. We showed that Pde1b reduction causes an upregulation of Pde10a mRNA, while Song et al. showed Pde2 to be increased by Pde1 inhibition (Song et al., 2016). Although each of these PDEs most likely plays a role, they appear to be inter-related, suggesting that pharmacological inhibition of PDE1B may alter each of the other PDEs. Further investigation of

PDE1 and /2 and PDE1 and /10A may shed further light on how PDE1B modulation may affect patients.

Intra-cellular Therapies Inc. recently reported a potent, specific, and brain penetrant

PDE1 inhibitor (Li et al., 2016). They showed that the inhibitor is capable of enhancing memory:

157

acquisition, consolidation, and retrieval in rats without altering exploratory behavior in novel object recognition (Li et al., 2016, Snyder et al., 2016). These studies indicate that PDE1B inhibition may have indications other than depression and it also suggests that inhibitors may be viable for multiple neuropsychiatric conditions. Snyder et al. reported no differences in exploratory behavior after drug-induced inhibition of PDE1B and suggested a primary involvement in cognitive and mood disorders (Snyder et al., 2016).

Although PDE4 inhibitors have undergone clinical trials for depression and failed because of side-effects, PDE1B continues to look promising. PDE1B is highly expressed in striatum and in the striatonigral pathway. KO of Pde1b only in D1 neurons was sufficient to reduce immobility, suggesting that PDE1B is a promising target. The negative side effects of

PDE4 may be due to off target effects caused by total PDE4 inhibition. Recently PDE4 inhibitors have been improved to reduce emetic effects but also PDE4D has been individually targeted and has shown efficacy (Rutter et al., 2014, Wang et al., 2015a). These data suggest that specifically targeting PDE1B may have reduced off target effects as compared with a nonspecific PDE1 inhibitor since Pde1b is primarily expressed in the brain (Lakics et al., 2010).

PDE1B is highly expressed in the striatum and in the striatonigral pathway. KO of Pde1b only in

D1 neurons was sufficient to reduce immobility suggesting the regulation of dopaminergic function. The co-localization of PDE1B with NMDA+ cells and changes in serotonin regulation when PDE1B is knocked out implies that PDE1B may be involved in these pathways as well.

Inhibition of PDE1B alone is promising and in combination SSRI, norepinephrine/dopamine reuptake inhibitor, and D1 antagonists and combinations may be the most promising. The development of a brain penetrant PDE1 inhibitor, now in Phase I clinical trials, suggests the feasibility of PDE1 inhibitors and perhaps the future development of an inhibitor even more specific to PDE1B. In conclusion, reduction of PDE1B attenuates the induction of depressive-

158

like behaviors in mouse models of “behavioral despair” at least partly through D1 neurons in the striatum, suggesting PDE1B has therapeutic potential in depression treatment.

159

References

Andersen SL (2003) Trajectories of brain development: point of vulnerability or window of

opportunity? Neuroscience & Biobehavioral Reviews 27:3-18.

Bateup HS, Svenningsson P, Kuroiwa M, Gong S, Nishi A, Heintz N, Greengard P (2008) Cell

type–specific regulation of DARPP-32 phosphorylation by psychostimulant and

antipsychotic drugs. Nature neuroscience 11:932-939.

Bender AT, Beavo JA (2006) Cyclic nucleotide phosphodiesterases: molecular regulation to

clinical use. Pharmacological reviews 58:488-520.

Bleakman D, Alt A, Witkin J (2007) AMPA receptors in the therapeutic management of

depression. CNS & Neurological Disorders-Drug Targets (Formerly Current Drug

Targets-CNS & Neurological Disorders) 6:117-126.

Cabanero M, Laje G, Detera-Wadleigh S, McMahon FJ (2009) Association Study of

Phosphodiesterase Genes in the Sequenced Alternatives to Relieve Depression (STAR*

D) sample. Pharmacogenetics and genomics 19:235.

Casanova E, Fehsenfeld S, Mantamadiotis T, Lemberger T, Greiner E, Stewart A, Schütz G

(2001) A CamKIIα iCre BAC allows brain‐specific gene inactivation. Genesis 31:37-42.

Cashman JR, Ghirmai S (2009) Inhibition of serotonin and norepinephrine reuptake and

inhibition of phosphodiesterase by multi-target inhibitors as potential agents for

depression. Bioorganic & medicinal chemistry 17:6890-6897.

Cashman JR, Voelker T, Johnson R, Janowsky A (2009) Stereoselective inhibition of serotonin

re-uptake and phosphodiesterase by dual inhibitors as potential agents for depression.

Bioorganic & medicinal chemistry 17:337-343.

Cryan JF, Mombereau C, Vassout A (2005a) The tail suspension test as a model for assessing

antidepressant activity: review of pharmacological and genetic studies in mice.

Neuroscience & Biobehavioral Reviews 29:571-625.

160

Cryan JF, Valentino RJ, Lucki I (2005b) Assessing substrates underlying the behavioral effects

of antidepressants using the modified rat forced swimming test. Neuroscience &

Biobehavioral Reviews 29:547-569.

Dragatsis I, Zeitlin S (2000) CaMKIIα‐cre transgene expression and recombination patterns in

the mouse brain. Genesis 26:133-135.

Drago A, Crisafulli C, Sidoti A, Serretti A (2011) The molecular interaction between the

glutamatergic, noradrenergic, dopaminergic and serotoninergic systems informs a

detailed genetic perspective on depressive phenotypes. Progress in neurobiology

94:418-460.

Dunlop BW, Nemeroff CB (2007) The role of dopamine in the pathophysiology of depression.

Archives of general psychiatry 64:327-337.

Ehrman L, Williams M, Schaefer T, Gudelsky G, Reed T, Fienberg A, Greengard P, Vorhees C

(2006) Phosphodiesterase 1B differentially modulates the effects of methamphetamine

on locomotor activity and spatial learning through DARPP32‐dependent pathways:

evidence from PDE1B‐DARPP32 double‐knockout mice. Genes, Brain and Behavior

5:540-551.

Fava M, Kendler KS (2000) Major depressive disorder. Neuron 28:335-341.

Halene TB, Siegel SJ (2007) PDE inhibitors in psychiatry–future options for dementia,

depression and schizophrenia? Drug discovery today 12:870-878.

Khalaf-Nazzal R, Francis F (2013) Hippocampal development–Old and new findings.

Neuroscience 248:225-242.

Klug JR, Mathur BN, Kash TL, Wang H-D, Matthews RT, Robison A, Anderson ME, Deutch AY,

Lovinger DM, Colbran RJ (2012) Genetic inhibition of CaMKII in dorsal striatal medium

spiny neurons reduces functional excitatory synapses and enhances intrinsic excitability.

PloS one 7:e45323.

161

Kulkarni SK, Dhir A (2009) Current investigational drugs for major depression. Expert opinion on

investigational drugs 18:767-788.

Lakics V, Karran EH, Boess FG (2010) Quantitative comparison of phosphodiesterase mRNA

distribution in human brain and peripheral tissues. Neuropharmacology 59:367-374.

Li P, Zheng H, Zhao J, Zhang L, Yao W, Zhu H, Beard JD, Ida K, Lane W, Snell G (2016)

Discovery of potent and selective inhibitors of Phosphodiesterase 1 for the treatment of

cognitive impairment associated with neurodegenerative and neuropsychiatric diseases.

Journal of medicinal chemistry 59:1149-1164.

Madisen L, Zwingman TA, Sunkin SM, Oh SW, Zariwala HA, Gu H, Ng LL, Palmiter RD,

Hawrylycz MJ, Jones AR (2010) A robust and high-throughput Cre reporting and

characterization system for the whole mouse brain. Nature neuroscience 13:133-140.

Marazziti D, Dell'Osso MC, Consoli G, Baroni S (2009) Second messenger modulation: a novel

target of future antidepressants? Current medicinal chemistry 16:4679-4690.

Nestler EJ, Carlezon WA (2006) The mesolimbic dopamine reward circuit in depression.

Biological psychiatry 59:1151-1159.

Nishi A, Snyder GL (2010) Advanced research on dopamine signaling to develop drugs for the

treatment of mental disorders: biochemical and behavioral profiles of phosphodiesterase

inhibition in dopaminergic neurotransmission. Journal of pharmacological sciences

114:6-16.

Reed TM, Repaske DR, Snyder GL, Greengard P, Vorhees CV (2002) Phosphodiesterase 1B

knock-out mice exhibit exaggerated locomotor hyperactivity and DARPP-32

phosphorylation in response to dopamine agonists and display impaired spatial learning.

The Journal of neuroscience 22:5188-5197.

Rutter AR, Poffe A, Cavallini P, Davis TG, Schneck J, Negri M, Vicentini E, Montanari D, Arban

R, Gray FA (2014) GSK356278, a potent, selective, brain-penetrant phosphodiesterase

4 inhibitor that demonstrates anxiolytic and cognition-enhancing effects without inducing

162

side effects in preclinical species. Journal of Pharmacology and Experimental

Therapeutics 350:153-163.

Shanks N, Anisman H (1989) Strain-specific effects of antidepressants on escape deficits

induced by inescapable shock. Psychopharmacology 99:122-128.

Siuciak JA, McCarthy SA, Chapin DS, Fujiwara RA, James LC, Williams RD, Stock JL, McNeish

JD, Strick CA, Menniti FS (2006) Genetic deletion of the striatum-enriched

phosphodiesterase PDE10A: evidence for altered striatal function. Neuropharmacology

51:374-385.

Siuciak JA, McCarthy SA, Chapin DS, Reed T, Vorhees C, Repaske D (2007) Behavioral and

neurochemical characterization of mice deficient in the phosphodiesterase-1B (PDE1B)

enzyme. Neuropharmacology 53:113-124.

Snyder GL, Prickaerts J, Wadenberg M-L, Zhang L, Zheng H, Yao W, Akkerman S, Zhu H,

Hendrick JP, Vanover KE (2016) Preclinical profile of ITI-214, an inhibitor of

phosphodiesterase 1, for enhancement of memory performance in rats.

Psychopharmacology 1-12.

Song RS, Tolentino R, Sobie EA, Neves-Zaph SR (2016) Cross-regulation of

Phosphodiesterase 1 and Phosphodiesterase 2 Activities Controls Dopamine-mediated

Striatal α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic Acid (AMPA) Receptor

Trafficking. Journal of Biological Chemistry 291:23257-23267.

Tsien JZ, Chen DF, Gerber D, Tom C, Mercer EH, Anderson DJ, Mayford M, Kandel ER,

Tonegawa S (1996) Subregion- and Cell Type–Restricted Gene Knockout in Mouse

Brain. Cell 87:1317-1326.

Wang Z-Z, Yang W-X, Zhang Y, Zhao N, Zhang Y-Z, Liu Y-Q, Xu Y, Wilson SP, O'Donnell JM,

Zhang H-T (2015) Phosphodiesterase-4D knock-down in the prefrontal cortex alleviates

chronic unpredictable stress-induced depressive-like behaviors and memory deficits in

mice. Scientific reports 5:11332.

163

Whitaker-Azmitia PM (2001) Serotonin and brain development: role in human developmental

diseases. Brain research bulletin 56:479-485.

Wong M-L, Whelan F, Deloukas P, Whittaker P, Delgado M, Cantor RM, McCann SM, Licinio J

(2006) Phosphodiesterase genes are associated with susceptibility to major depression

and antidepressant treatment response. Proceedings of the National Academy of

Sciences 103:15124-15129.

164

Figures

Figure 1. The role of PDE1B in AMPA receptor phosphorylation and cellular surface regulation.

PDE1B KO mice have increased cAMP and cGMP leading to elevated levels of CaMKII and cGKII that promotes increased cellular membrane levels of AMPA receptors. Increases in cGMP by PDE1B inhibition also increases PDE2 that blocks AMPA receptor insertion into the cellular membrane.

165