Effect of Imipramine and Classical Benzodiazepines on Stress-Induced

Effect of Imipramine and Classical Benzodiazepines on Stress-induced

Neuroimmune Dysregulation and Behavior

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy

in the Graduate School of The Ohio State University

Karol Gabriela Ramirez Chan, DDS, MS

Graduate Program in Oral Biology

The Ohio State University

2015

Dissertation Committee:

Dr. John F. Sheridan, advisor

Dr. Ning Quan

Dr. Michael Bailey

Dr. John Walters

 Copyright by

Karol Gabriela Ramírez Chan

2015

Abstract

Psychosocial stress promotes brain-to-immune and immune-to-brain communication that can impact neurobiology and behavior. For example, exposure to stress is capable of causing peripheral immune dysregulation and neuroinflammatory signaling by repeated activation of neuroendocrine and autonomic pathways; that may contribute to the development of diseases and mental health disturbances.

In order to relieve anxiety and depression accompanying stress, physicians resort to anxiolytics and antidepressants. Lorazepam and clonazepam are routinely prescribed anxiolytics that act by enhancing GABAergic activity in the brain. Moreover, imipramine, a tricyclic antidepressant effective in the treatment of mood disorders, has been reported to influence immune function in depressed patients.

The potent impact of social stress on human’s health led to the development of representative animal models that emulate the pathophysiology of mental health disorders. The stress model of repeated social defeat (RSD), recapitulates many of the stress-driven alterations in both the periphery and central nervous system seen in humans experiencing chronic or repeated forms of stress (Hanke et al., 2012). For example, RSD triggered egress of inflammatory myeloid progenitor cells that traffic to blood, spleen and brain. RSD promoted brain region-specific activation of microglia/macrophages that led

ii to prolonged anxiety. In parallel, RSD promoted long-lasting social avoidant behavior.

Thus, the overall aim of this dissertation was to determine if pharmacologic blockade of the GABAergic and monoaminergic circuits using lorazepam, clonazepam, and imipramine, respectively, would: 1) prevent stress-induced peripheral and central inflammatory responses, and 2) block anxiety and social avoidance behavior in mice subjected to RSD.

Treatment with lorazepam or clonazepam prior to stressor exposure affected central and peripheral responses. Treatment with either drug was effective in attenuating mRNA expression of corticotropin-releasing hormone (CRH) in the hypothalamus and corticosterone in plasma in mice subjected to RSD. Both drugs blocked stress-induced elevated levels of IL-6 in plasma. Lorazepam and clonazepam had different effects on stress-induced enhancement of myelopoiesis and inhibited the release of monocytes and granulocytes to circulation. Furthermore, lorazepam, but not clonazepam, inhibited splenomegaly and the production of pro-inflammatory cytokines in the spleen following

RSD. Additionally, lorazepam and clonazepam blocked stress-induced accumulation of macrophages in the brain. In a similar manner, both lorazepam and clonazepam prevented neuroinflammatory signaling and reversed anxiety-like and depressive-like behavior in mice exposed to RSD.

Imipramine inhibits the reuptake of serotonin, norepinephrine, and dopamine on presynaptic neurons, and thus increases synaptic levels of these neurotransmitters.

Antidepressants can stimulate adaptive changes in the central monoaminergic circuitry, which can modulate immune reactivity and the central actions of cytokines (Miller et al.,

iii

2009). In the present study, imipramine attenuated stress-induced corticosterone and IL-6 in plasma. Imipramine also decreased the percentage of monocytes and granulocytes in the bone marrow and circulation. However, imipramine did not prevent splenomegaly, stress-related increased percentage of granulocytes in this organ, and the production of pro-inflammatory cytokines in the spleen, following RSD. Imipramine abated the accumulation of macrophages in the brain in mice exposed to RSD. Imipramine treatment prior to stress exposure blocked neuroinflammatory signaling and stress-related anxiety- and depressive-like behavior.

Finally, administration of imipramine for 24 days after stress termination reversed social avoidance behavior. Moreover, 24 days of imipramine treatment decreased stress- induced mRNA levels for IL-6 in brain microglia. Following ex vivo LPS stimulation, microglia from mice exposed to RSD, had higher mRNA expression of IL-6, TNF-α, and

IL-1β, and this was reversed by imipramine treatment. These data suggest that the antidepressant imipramine may exert its effect, in part, by down-regulating microglial activation.

These data support the notion that pharmacomodulation of the GABAergic and monoaminergic systems, besides exerting anxiolytic and antidepressant effects, may have therapeutic potential as neuroimmunomodulators during stress.

Dedication

Dedicated to María Rosa and Claudia, my beautiful grandmothers.

The lessons you both taught me have shaped who I am today.

My brother, Gabriel.

You’re the greatest gift of life!

And my parents, heroes forever, Rodrigo and Dory.

Thank you for everything you have done for me!

Acknowledgments

First, I would like to extend my gratitude to Dr. John Sheridan for believing in my potential and opening the doors of his laboratory to obtain my Doctor in Philosophy

Degree. You have inspired me to pursue my goals with hard work and perseverance.

Thank you for your time and mentoring. I really look forward to the day I can do the same for someone else.

Second, I would like to thank fellow graduate students Dan McKim and Anzela

Niraula for their generous help and guidance. Special thanks to post-doctoral fellows Dr.

Brenda Reader, Dr. Brant Jarrett, and Dr. Seung Yung (David) for sharing their scientific knowledge.

Third, I would like to thank technicians Daniel Shea, David Hammond, Yufen

Wang, Juan Carlo Avalon, undergraduate students, Tom Zhang, January Kim, Ana

Sucaldito, Rahul Gupta, Jenna Patterson, and post-doctoral fellow Sabrina Lisboa for their assistance in the experiments outlined below.

I would like to acknowledge the committee members for their time and dedication in all the stages in the pursuit of my degree. Thanks to Dr. Quan, Dr. Bailey, and Dr.

Walters.

I will be forever grateful with the College of Dentistry of The Ohio State

University for supporting my doctoral training financially.

To conclude, I would like to thank the University of Costa Rica for providing the means in getting further training in research. Special thanks, to faculty members that have facilitated my professional development.

vii

Vita

October 29, 1977 ...... Born, San José, Costa Rica

1996...... Lincoln High School, San José, Costa Rica

2002...... Doctor of Dental Surgery, Universidad

Latina de Costa Rica, San José, Costa Rica

2006...... Certificate in Periodontology, Pontificia

Universidad Javeriana, Bogotá, Colombia

2007-Present ...... Professor, Facultad de Odontología,

Universidad de Costa Rica, San José, Costa Rica

2012...... Master of Science, Post-Doctoral Master in

Clinical and Translational Research, University of Puerto Rico, Medical Sciences

Campus, San Juan, Puerto Rico

2012-Present ...... Graduate Research Associate, Department of Biosciences, College of Dentistry, The Ohio State University, Columbus, Ohio, USA

viii

Publications

Elias Boneta AR , Ramirez K, Naboa J, Mateo LR, Stewart B, Panagokos F, De Vizio W.

J. Efficacy in reducing dentine hypersensitivity of a regimen using a toothpaste containing 8% arginine and calcium carbonate, a mouthwash containing 0.8% arginine, pyrophosphate and PVM/MA copolymer and a toothbrush compared to potassium and negative control regimens: an eight-week randomized clinical trial. Journal of Dentistry.

2013 Mar; 41 Suppl 1:S42-9.

Ramirez K, García-Rodríguez O, Murillo-Arocho M, Fernández-López O, Elías-Boneta

AR. Dentogingival complex: dimension based on biotypes. Puerto Rico Health Sciences

Journal. 2013 Dec;32 (4):182-6

Ramirez K, Shea DT, McKim DB, Reader BF, Sheridan JF. Imipramine attenuates neuroinflammatory signaling and reverses stress-induced social avoidance. Brain

Behavior and Immunity. 2015 Feb 18. pii: S0889-1591(15)00024-0. doi:10.1016/j.bbi.2015.01.016.

Ramirez K, Niraula A, Shea DT, Sheridan JF. GABAergic modulation with classical benzodiazepines prevent stress-induced neuro-immune dysregulation and behavioral alterations. Brain, Behavior, and Immunity (In Review).

Ramirez K, Sheridan JF. Antidepressant imipramine rescues stress-induced inflammation in the periphery and central nervous system and related anxiety-and depressive like behaviors (In preparation).

Fields of Study

Major Field: Oral Biology

Area of specialization: Immunology and Inflammation

Research Area: Stress

Table of Contents

Abstract ...... ii

Dedication ...... v

Acknowledgments...... vi

Vita ...... viii

List of Tables ...... xii

List of Figures ...... xiii

List of Abbreviations ...... xv

Chapter 1: Introduction ...... 1

Chapter 2: GABAergic modulation with classical benzodiazepines prevents stress- induced neuro-immune dysregulation and behavioral alterations ...... 19

Chapter 3: Imipramine attenuates neuroinflammatory signaling and reverses stress- induced anxiety- and depressive- like behaviors ...... 65

Chapter 4: Discussion ...... 108

References ...... 119

List of Tables

Table 1- Effect of BDZs on stress-induced increase of pro-inflammatory cytokines in total splenocytes ...... 62

Table 2- Effect of BDZs on neuroinflammatory signaling in microglia ...... 63

Table 3- Effect of BDZs on neuroinflammatory signaling in hypothalamus/hippocampus coronal section ...... 64

Table 4 – Effect of imipramine on stress-induced mRNA expression of pro-inflammatory cytokines in the spleen ...... 102

Table 5- Imipramine blocked stress-induced neuroinflammatory signaling ...... 102

Table 6- Effect of imipramine on total splenocytes gene expression 24 days after stress termination ...... 104

Table 7- Imipramine blocked long-lasting neuroinflammatory signaling provoked by stress exposure ...... 104

Table 8- Imipramine treatment blocked microglia activation 24 days after stress cessation

...... 105

Table 9- Imipramine given orally ameliorated stress-induced long-lasting neuroinflammatory responses ...... 107

xii

List of Figures

Figure 1- Effect of BDZs on stress-induced increase of CRH, corticosterone, and norepinephrine ...... 54

Figure 2- Effect of BDZs on stress-driven increase of plasma IL-6 ...... 55

Figure 3- Effect of BDZs on stress-induced shift of hematopoiesis ...... 56

Figure 4 - Effect of BDZs on stress-induced increased trafficking of monocytes and granulocytes in circulation ...... 57

Figure 5- Effect of BDZs on the spleen ...... 58

Figure 6 – Effect of BDZs on monocyte trafficking to the CNS ...... 59

Figure 7- BDZs blocked stress-associated anxiety-like behavior ...... 60

Figure 8- BDZs prevented stress-related depressive-like behavior ...... 61

Figure 9- Imipramine attenuated RSD-induced increase in corticosterone and prevented

IL-6 in plasma ...... 96

Figure 10- Effect of imipramine on stress-induced shift of hematopoiesis ...... 97

Figure 11- Effect of imipramine on stress-induced increase of monocytes and granulocytes in blood and spleen ...... 98

Figure 12- Imipramine prevented stress-induced accumulation of macrophages in the brain ...... 99

Figure 13- Imipramine prevented stress-associated anxiety-like behavior ...... 100

xiii

Figure 14- Imipramine rescued stress-related depressive-like behavior...... 101

Figure 15- Imipramine treatment reversed RSD-induced social avoidance 24 days after stress cessation ...... 103

Figure 16- Imipramine blocked stress-induced exaggerated levels of IL-1β in the supernatant of LPS-stimulated microglia...... 105

Figure 17- Imipramine in drinking water reversed stress-induced social avoidance behavior 24 days after stress cessation ...... 106

xiv

List of Abbreviations

6-FAM 6-carboxyfluorescein

α alpha

ACTH adrenocorticotropic hormone

AMYG amygdala

APC antigen presenting cell or allophycocyanin

ATP adenosine triphosphate

β beta

BDZs benzodiazepines

BM bone marrow

BNST bed nucleus of stria terminalis

BSA bovine serum albumin c° celsius

CA3 cornu ammonis region 3 cAMP cyclic adenosine monophosphate

CBR central benzodiazepine receptor

CCL chemokine ligand

CCON clonazepam control

CD cluster of differentiation cDNA complementary deoxyribonucleic acid cm centimeter

CNS central nervous system

CO2 carbon dioxide

CON control

CRH corticotropin-releasing hormone

CRSD clonazepam repeated social defeat

Ct comparative threshold cycle

CX3CL1 fractalkine ligand

CX3CR1 fractalkine receptor

DG dentate gyrus

DMSO dimethyl sulfoxide

DNA deoxyribonucleic acid

EDTA ethylenediaminetetraacetic acid

EIA enzyme immunoassay

ELISA enzyme-linked immunoabsorbant assay

EST eastern standard time

FACS fluorescence-activated cell sorting

FBS fetal bovine serum

FITC fluorescein isothiocyanate

FGF fibroblast growth factor

xvi

FKBP51 FK506 binding protein-51

γ gamma g gram

GABA gamma-aminobutyric acid

GAPDH glyceraldehyde 3-phosphate dehydrogenase

GC glucocorticoid

GCs glucocorticoids

GILZ glucocorticoid-induced leucine zipper

GM-CSF granulocyte macrophage colony-stimulating factor

h hour

H2O water

HBSS Hank’s balanced salt solution

HCC home cage control

HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

HPA hypothalamic-pituitary-adrenal

HPC hippocampus

HYPO hypothalamus

IкB inhibitor of кB

Iba-1 ionized binding adaptor molecule-1

ICAM intercellular adhesion molecule

ICON imipramine control

IRSD imipramine RSD

xvii

IFN interferon

IKK inhibitory кB kinase

IL interleukin

IL-1β interleukin-1beta

IL-1R1-/- interleukin-1 receptor type 1 deficient mice

IL-6 interleukin-6 i.p. intraperitoneal

κ kappa kg kilogram

LC locus coereleus

LCON lorazepam control

L liter(s)

LPS lipopolysaccharide

LS lateral septum

LRSD lorazepam repeated social defeat

MAPK mitogen-activated protein kinase

MCS-F macrophage colony- stimulating factor

MCP-1 monocyte chemoattractant protein-1

MCP-2 macrophage inflammatory protein-2

µg microgram

μl microliter

μm micrometre

xviii mg milligram ml milliliter mm millimeter min minute

M molar

MPC myeloid progenitor cell

MPCs myeloid progenitor cells mRNA messenger ribonucleic acid

NAc nucleus accumbens

NaCl sodium chloride

NF-кB nuclear factor-кB

NFQ non-fluorescent quencher ng nanogram nm nanometer

PBS phosphate buffered saline

PCR polymerase chain reaction

PE phycoerythrin

PerCP phycoerythrin peridinin-chlorophyll-protein-complex

PDE4 phosphodiesterase type 4

PFC pre-frontal cortex

PKA protein kinase

PVN paraventricular nucleus of the hypothalamus

xix rpm revolutions per minute

RSD repeated social defeat

SEM standard error of the mean

SNS sympathetic nervous system

SSRI selective serotonin reuptake inhibitor

TCA tricyclic antidepressant (s)

TLR toll-like receptor

TNF-α tumor necrosis factor alpha

TSPO translocator protein

VCON vehicle control

VTA ventral tegmental area

VRSD vehicle repeated social defeat

Chapter 1: Introduction

The Stress Response

A stressor is defined as a physiological, physical or psychological stimulus that disrupts homeostatic function. Following perception of stress, physiologic and behavioral changes are enacted to help reestablish homeostasis. The specific response of an organism to a stressor is dependent on the magnitude, duration, and individual perception. Walter Cannon, an American physiologist was a pioneer in understanding bodily changes related to pain and fear. He coined responses to threat in animals as a

“fight or flight response”, which is related with the activation of the autonomic nervous system (Cannon, 1929). Hans Selye performed further work in the biological effects of exposure to stressful stimuli. He described the physiological and behavioral outcomes of the “general adaptation syndrome” to describe stress as the “nonspecific response of the body to any demand for change” (Selye, 1946). Selye’s research focused on the role of secreted factors that affected organs such as the thymus, spleen, and liver. Years later, scientists interested in the field of psychoneuroimmunology, further studied the interactions among the nervous, endocrine, and immune system. A focus of research

nowadays is the impact of these interactions on health.

Neuro-endocrine-immune communication in the response to stress

Brain and immune system communication involves a bidirectional interaction between these two platforms by a shared and specialized pathway known as the neuroimmune axis. The neuro-immune cross talk cannot be discussed without the inclusion of the endocrine system as a major bridge that unifies the first two systems during host’s homeostasis. The brain, endocrine and immune systems communicate with one another at different molecular levels with synthesis and release of endogenous chemical messengers, primarily cytokines, chemokines, neurotransmitters, and hormones.

Synchronized neuroendocrine and neuroimmune responses to infectious or stressful stimuli are mandatory for adaptive physiological and behavioral responses. For example, endogenous and exogenous stressors stimulate neuroendocrine, sympathetic, and immune responses, ensuring the activation of the hypothalamic–pituitary–adrenal (HPA)-axis and the sympathetic nervous system (SNS). HPA activation leads to the release of glucocorticoids (GC)s and activation of the SNS leads to the release of cathecolamines, epinephrine and norepinephrine, in circulation and tissues. Jointly, the activation of the

HPA and SNS increase the breakdown of glucose; and increase heart rate and muscle tone during acute stress (Sternberg et al., 2006). This provides the organism increased energy to respond to aversive stimuli. Via the stimulation of the HPA and SNS, the stress response is relayed from the brain to peripheral organs and to the immune system. In 2

contrast, when homeostasis is disrupted repeatedly, like for example with repeated psychosocial stress, there is an increase in inflammation in both the periphery and central nervous system (CNS). Exposure to prolonged stress promotes brain-to-immune and immune-to-brain communication that can influence neurobiology and behavior by repeated activation of neuroendocrine and autonomic pathways (Wohleb et al., 2014).

Psychosocial stress-induced, peripheral immune dysregulation and neuroinflammation contribute to the development of mental health disturbances, including anxiety and depression (Wohleb et al., 2013).

The primary mechanisms mediating physiological and behavioral responses to stress have been elucidated. However, the biological mechanisms connecting stress to psychiatric symptoms and disorders are not well understood. A growing body of evidence indicates that stress-related inflammation and altered immune signaling contribute in the pathogenesis of mental health complications, specifically in the context of psychosocial stress, with anxiety and depression being two of the most common outcomes in humans.

Interestingly, anxiety and depression show considerable overlap and co-occurrence. The fact that both anxiety and depression respond to the same pharmacological treatments, such as with anxiolytics and antidepressants, suggests a common biological dysfunction.

Repeated Social Defeat

The potent impact of social stress on humans’ health led to the development of representative preclinical animal models. There are several stress models that recapitulate 3

the pathophysiology of mental health disorders. Some clinically relevant murine models include repeated social defeat (RSD), inescapable tail-shock, restraint stress, and chronic unpredictable stress or chronic mild stress. RSD, a model of psychosocial stress in mice, provides a valuable and predictable probe to study the mechanisms leading to stress- related alterations in inflammation in both the periphery and CNS. These inflammatory responses are frequently associated with anxiety- and depressive-like behaviors (Kinsey et al., 2007; Wohleb et al., 2011; Wohleb et al., 2013).

The natural tendency of male mice living together in a cage is to form social hierarchies. During RSD, the social hierarchy is disrupted by introducing an aggressive male intruder into the cage of an established cage of resident mice. RSD in done once a day for six consecutive nights, during a predetermined time (i.e. two hours), to mimic a recurring stressor. During RSD, the resident mice are defeated and they display submissive behaviors. RSD elicits unique central, endocrine, and immune responses and invokes a series of individual differences in inbred mice (Avitsur et al., 2009).

Peripheral Effects of Stress

The prolonged inflammatory state associated with social stress has the potential to contribute to the etiology of anxiety and depression. However, it is not clear whether cytokines derived from the CNS or the periphery, or both, are responsible for the effects of these molecules in mood dysfunctions.

Analysis of peripheral inflammation markers in patients with mood disorders reveals constant elevations in interleukin-6 (IL-6), interleukin-1 (IL-1), and tumor necrosis factor- (TNF-) (Russo and Nestler, 2013). RSD caused an increase in pro- inflammatory molecules released into circulation. Specifically, RSD caused an increase of IL-6, TNF-, keratinocyte chemoattractant, macrophage inflammatory protein-2

(MCP-2), and monocyte chemoattractant protein-1 (CCL2, also known as MCP-1)

(Reader et al., 2015). These pro-inflammatory factors enhance inflammation and the trafficking of immune cells to enter circulation and tissues, such as the spleen, lung, and brain. It has been postulated that the increase of these pro-inflammatory mediators is mediated by beta ()-adrenergic signaling since pretreatment with propranolol, a nonselective -adrenergic blocker, ablated increases of IL-6, TNF, and CCL2 in circulation

(Hanke et al., 2012). Also, RSD enhanced levels of adrenocorticotropic hormone

(ACTH) in circulation, with a peak after the second cycle, and then the levels of the hormone decreased. However, increased levels of corticosterone elevated following the second cycle of RSD, and peaked at the sixth cycle (Engler et al., 2005).

In the RSD model, mice exposed to stress display an increase in plasma and tissue cathecolamines. This is meaningful, since peripheral immune cells express receptors for norepinephrine, and when these receptors are stimulated, functional responses occur that influence the development and mobility of these cells, as well as their inflammatory phenotype (Powell et al., 2013). For example, repeated activation of the SNS caused an increase of norepinephrine in the bone marrow (BM) (Hanke et al., 2012) that promoted a

shift in myelopoiesis after RSD exposure. Myeloid cells produced in the BM as a result of RSD were less mature and more inflammatory (Engler et al., 2005). These inflammatory myeloid progenitor cells (MPC)s trafficked throughout the body and had an increased capacity of releasing pro-inflammatory cytokines. Also, these pro- inflammatory cells were found to be GC-resistant. CD11b+ cells from RSD mice showed an exaggerated production of inflammatory cytokines, IL-6, TNF-, and IL-1 in response to toll-like receptor (TLR) and antigen stimulation (Avitsur et al., 2001; Bailey et al., 2009).

A hallmark of RSD is splenomegaly, which corresponds with an increased accumulation of CD11b+ monocytes and granulocytes (Avitsur et al., 2002).

Additionally, splenic monocytes, but not peritoneal monocytes, from RSD mice are GC- resistant (Avitsur et al., 2002), which suggests a tissue-specific response. GC- insensitivity is prevalent among infectious disease, but also in depression and anxiety

(Reader et al., 2015). The inverse relationship between ACTH and corticosterone and the

GC-insensitivity tissue-specific response suggests that the negative GC feedback mechanism is not disrupted in the pituitary gland; therefore RSD does not provoke a central state of

GC-insensitivity (Engler et al., 2005).

Interpretation of stress in the central nervous system

Neurobiological and behavioral responses occur when physiological stress is interpreted within the brain leading to activation of fear and threat appraisal circuitry.

Among the brain regions that are activated in response to stress are the prefrontal cortex

(PFC), hypothalamus (HYPO), amygdala (AMYG), and the CA3 and dentate gyrus of the hippocampus (HPC). Thus, repeated activation of neurophysiological circuits by stress, may lead to dysfunction in these limbic regions. In fact, activation of this pathway caused anxiety and depression in humans (Sheehan et al., 2004). Other brain regions that are activated in response to social stressors are the bed nucleus of the stria terminalis

(BNST), lateral septum (LS), and nucleus accumbens (NAc). These limbic regions were implicated in regulating mood (Wohleb et al., 2014).

Studies suggest that anxiety and depressive-like behaviors related to stress are associated with major neurobiological dysfunctions within brain regions that regulate emotional and behavioral responses such as the PFC, AMYG, and HPC (Wohleb et al.,

2014). For example, rodent stress models that induce depression- and anxiety-like symptoms have demonstrated that stress caused neuronal atrophy and dendritic retraction of neurons in the PFC (Radley et al., 2004). Additionally, stress was related to dendritic atrophy and changes in spine density (Russo and Nestler, 2013). Social defeat stress was reported to reduce neurogenesis in the HPC (Bokhoven et al., 2011). In addition, it has been shown in animal studies, that stress promotes neuronal hypertrophy and increased dendritic complexity in the AMYG (Wohleb et al., 2014). Recent literature has identified 7

structural and functional impairments within the brain’s reward circuitry, specifically in the ventral tegmental (VTA) area to the NAc neurocircuitry, that are associated with anhedonia (decrease sucrose preference), a depressive-like symptom in rodents (Russo and Nestler, 2013).

Stress induces increased neuroinflammatory signaling

Stress activates microglia, the resident immune cells of the brain. Microglia activation refers to changes in the morphology, for example after stress exposure there is an increased size in soma. Microglia play a pivotal role in immune monitoring in the

CNS, and when activated they propagate similar immune functions as peripheral macrophages (Wohleb et al., 2014), including the production of pro-inflammatory cytokines, chemokines, and prostaglandins. Microglia have been demonstrated to be the source of neuroinflammatory signaling following stress exposure. For example, minocycline, an antibiotic that prevents microglia activation, ameliorated restraint stress- induced expression of pro-inflammatory cytokines in the brain and attenuated cognitive deficits, depressive and anxiety-like behaviors (Hinwood et al., 2012). Thus, physiological microglia alterations are likely to contribute to the dysfunctional neurobiological stress interpretation in the CNS.

It has been reported that prolonged stress caused neuronal and microglia activation in stress-responsive brain regions. As stated previously, RSD evokes a “fight or flight” response that causes neuronal and microglia activation within brain regions 8

associated with fear, anxiety and threat appraisal, including the PFC, HYPO, AMYG, and the CA3 and dentate gyrus of the HPC (Wohleb et al., 2014). It has been suggested this neuronal activation precedes microglia activation since evidence from the RSD stress model revealed an increase in region-specific neuronal activation after just one cycle of stress (Wohleb et al., 2014). The increased mRNA expression of pro-inflammatory cytokines in microglia occurs generally after three cycles of RSD, indicating that neuronal activation occurs first and then it is followed by activation of microglia.

It is well established that neuronal activity can regulate local microglia activation.

Likewise, microglia survey neurons through chemokine signaling pathways (i.e. fractalkine ligand (CX3CL1), growth factors such as macrophage colony stimulating factor (M-CSF), release of adenosine triphosphate (ATP), and levels of neurotransmitters.

Moreover, there is evidence that activation of microglia potentiates HPA axis stimulation through the release of IL-1 within the HYPO (Goshen and Yirmiya, 2009), augmenting neuroendocrine outflow that may reinforce stress-associated behaviors.

Microglia isolated from socially defeated mice have high levels of IL1- mRNA expression and reduced levels of GC responsive genes (GC-induced leucine zipper

(GILZ) and FK506 binding protein-51 (FKBP51) (Wohleb et al., 2011). Furthermore, microglia isolated from these mice and cultured ex vivo produced increased levels of IL-

6, TNF-, and CCL-2 following mitogen-stimulation with lipopolysaccharide (LPS) compared to microglia from home cage controls (Wohleb et al, 2011).

Neuroinflammatory mediators such as IL1-, TNF-, and IL-6, are related in the

neurobiological changes that reinforce fear/anxiety and threat circuitry (Wohleb et al.,

2014), promoting the development and maintenance of anxiety-like behavior.

Increasing attention has focused on the relationship between neuroinflammation and mood disorders. As stated previously, bouts of repeated stress have been reported to induce immune-enhancement and cause a pro-inflammatory phenotype by promoting a state of GC-insensitivity to develop in innate immune cells. This GC-insensitivity prevents the suppression of inflammation through immune cell apoptosis and inhibition of transcription factor nuclear factor kappa B (NFK-B) (Reader et al., 2015). NFK-B is a downstream target of pro-inflammatory mediators IL-6, IL-1, and TNF-, both in peripheral tissues and in the brain. It has been shown that NFK-B has the potential of regulating the brain reward circuitry in depression models. Altered central levels of these pro-inflammatory factors act in the HPC, increasing depressive-like behavior in response to chronic stress (Russo and Nestler, 2013). For instance, NFK-B activation in the HPC is required for the stress-related impairment of neurogenesis and induction of anhedonia

(Koo et al., 2010). Chronic social defeat stress increased levels of inhibitor of kappa B kinase (IKK) in the NAc, which then increased downstream signaling of NFK-B by phosphorylation IKB and causing its dissociation from NFK-B (Russo and Nestler,

2013). New immature excitatory spine structures on NAc dendrites are formed by the activation of NFK-B as well (Christoffel et al., 2011). Moreover, pro-inflammatory cytokines such as IL-6, IL-1, and TNF- modulate neuronal activity (Ozaktay et al.,

2006). For example, stress-induced neuroinflammatory signaling increases neuroplasticity that can lead to modification in the connectivity between neurons and 10

neuronal circuits underlying behavioral disorders such as prolonged anxiety and depressive symptoms (Koo and Duman, 2008; Koo et al., 2010; Christoffel et al., 2011).

It has also been reported pro-inflammatory cytokines increase central monoamine metabolism; this system is thought to be disturbed in depression (Kenis and Maes, 2002).

RSD-induced neuronal and microglia activation and production of pro- inflammatory molecules, promote the development of a reactive endothelium (Wohleb et al., 2014). Vascular brain endothelial cells increase cell adhesion molecule expression, which facilitates the adherence and extravasation of peripherally derived monocytes, which differentiate into perivascular and parenchymal macrophages (Wohleb et al.,

2014). Furthermore, parenchymal infiltration of peripherally derived monocytes is region specific and co-occurs within the fear, anxiety, and threat appraisal circuitry (Wohleb et al., 2013). The increased accumulation of macrophages in the CNS elicited by RSD enhanced neuroinflammatory signaling. RSD also enhanced reactivity of microglia and macrophages in a brain-dependent manner. In a previous study we determined reactivity of microglia and macrophages through Iba-1 staining in the medial AMYG, PFC, and paraventricular nucleus of the HPC, after 6 days of RSD. These findings showed that social defeat enhanced the active microglia phenotype in several areas of the brain associated with fear and threat appraisal, after 6 days of RSD (Wohleb et al., 2011).

Interestingly, recent findings from our laboratory showed that RSD-induced anxiety-like behavior persisted for at least eight days after social defeat but was resolved by twenty-four days. Additionally, twenty-four days after RSD termination, markers of immune alterations associated with RSD, such as splenomegaly, plasma IL-6, and the 11

number of circulating CD11b+ cells returned to control levels. Neuroinflammatory signaling returned to baseline by 24 days, nonetheless IL-6 mRNA level was still elevated at 24 days (Wohleb et al., 2013; Ramirez et al., 2015). Iba-1 labeling of microglia and increased Iba-1 proportional area was detected in the PFC 24 days after

RSD. However, immunoreactivity was no longer detected in the AMYG, HPC-CA3, and dentate gyrus of the HPC by 24 days. Brain macrophage accumulation, was no longer detected at 24 days after stress cessation (Wohleb et al., 2013). It was apparent that microglia return to a surveying state after RSD in a time-dependent manner.

It should be noted, nevertheless, that social avoidance behavior developed after 1 cycle of social defeat (Wohleb et al., 2013), and long-lasting social avoidance to an aggressor was present 24 days after stress cessation when the majority of macrophages had left the CNS. Splenomegaly and IL-6 responses abated, and hematopoiesis had returned to a homeostatic pattern (Wohleb et al., 2013).

Classical benzodiazepines, lorazepam and clonazepam

As stated before, psychosocial stress plays a major role in the pathophysiology of many mental health disturbances, including anxiety and depression, which are two of the most frequent diseases (Kalueff and Nutt, 2007). Depression and anxiety share common symptoms and are associated with increased release of peripheral cytokines in both animals and humans (Maes et al., 2011; Hodes et al., 2014). In order to relieve anxiety and depression accompanying stress, physicians resort to anxiolytic drugs, such as 12

benzodiazepines (BDZs). What makes BDZs one of the most prescribed classes of drugs is that, besides having a marked anxiolytic effect, they possess anticonvulsant, muscle relaxant, and sedative-hypnotic effects (Arvat et al., 2002), depending on the dose used.

The action of BDZs depends on the activation of two receptors, the central benzodiazepine receptor (CBR) and the peripheral benzodiazepine receptor, now renamed translocator protein (TSPO). The CBR is present in several regions of the CNS and is a component of the γ-aminobutyric acid (GABA)-A receptor (Papadopoulos et al.,

2006). Binding to this site, enhances GABAergic neurotransmission (Ferrarese et al.,

1993), which plays an inhibitory effect in the CNS by counteracting stress-induced hyperactivation of the HPA axis. BDZs also stimulate neuroendocrine effects that seem to be mediated at the hypothalamic and/or supra-hypothalamic level by suppressing the production of CRH (Arvat et al., 2002). In this manner, peripheral injections of BDZs have been shown to inhibit stress-induced CRH release (Imaki et al., 1995). On the other hand, the TSPO is structurally and functionally different from the GABA-A receptor.

TSPO is mostly, but not only, located in peripheral tissues, (e.g., heart, kidney, liver, lung, and adrenal glands (Papadopoulos et al., 2006). Specifically, the TSPO is present in the mitochondria, coupled to an anion channel, in platelets, lymphocytes, mononuclear cells, endothelium, vascular smooth muscle, mast cells, and also in microglia and neurons.

Several in vitro studies have shown that BDZs that bind to the TSPO can affect cellular and immune functions by suppressing cytokine secretion (Taupin et al., 1991; Kim et al.,

2006; Joo et al., 2009; Yousefi et al., 2013), modifying cell proliferation, affecting immune cell migration (Ruff et al., 1985) and cellular phagocytic activity (Jin et al., 13

2013). BDZs appear to have anti-inflammatory effects due to their interaction with the innate immune system (Kim et al., 2006). Binding of BDZs to these receptors modulates immune function and regulation of inflammation (Casellas, 2002; Veenman and Gavish,

2006). Taken together, these clues prompted us to explore the use of lorazepam and clonazepam in RSD to evaluate the effect of these two drugs on the neuroimmune system during psychosocial stress. Lorazepam and clonazepam are BDZs that act as positive allosteric modulators of GABA action with similar pharmacological mechanisms and share anxiolytic properties (Verleye et al., 2008; Saari et al., 2011). These two drugs have been used extensively to distinguish between the CBR and the TSPO (Casellas, 2002).

Lorazepam acts directly on the GABAA-R complex, and has affinity for the TSPO. On the other hand, clonazepam is a selective agonist of the GABAA-R (CBR) but does not bind with high affinity to the TSPO (Anholt et al., 1986). Clonazepam was found to have a strong binding capacity to TSPO in rat aortic smooth muscles, and these binding sites were concentrated in the mitochondria (Cox et al., 1991). These two different positive allosteric modulators of the GABAA-R produce enhancement of GABAergic activity

(Uusi-Oukari and Korpi, 2010) by locking the GABA receptor complex into a conformational change that is able to bind GABA with high affinity (Nutt and Malizia,

2001; Saari et al., 2011). Since both drugs affect functions of the endocrine system, it would be logical to expect that these two drugs might modulate the neuroendocrine regulation of inflammation during psychosocial stress by decreasing stress-induced neuroinflammatory signaling, thus, modifying stress-associated behavioral changes such as anxiety- and depressive-like behaviors. Therefore, the aims in the present study 14

(Chapter 2) were to determine if lorazepam and clonazepam would: 1) prevent stress- induced peripheral and central inflammatory responses, and 2) block anxiety and social avoidance behavior in mice subjected to RSD.

Imipramine

In recent years a focus on the efficacy of antidepressants on the reversal of neurobiological deficits caused by stress has dominated biological psychiatry. For instance, chronic administration of fluoxetine, a selective serotonin reuptake inhibitor, increased hippocampal neurogenesis (Malberg et al., 2000). Moreover, antidepressant tianeptine restored neuronal proliferation and hippocampal volume after stress exposure

(Czeh et al., 2001). These pro-neurogenic effects were found to be essential for the attenuation of altered behavioral responses. For example, the antidepressant behavioral effects of fluoxetine and imipramine, the latter a tricyclic antidepressant, were prevented in mice that had ablated neurogenesis (Santarelli et al., 2003). Similarly, in another study, imipramine treatment restored stress-induced decrease in neurogenesis in the HPC

(Bokhoven et al., 2011).

Immune cells, including monocytes, store serotonin, and transport of serotonin into immune cells is promoted by pro-inflammatory cytokines, such as IL1- and TNF-

(Janssen et al., 2010). Pro-inflammatory cytokines, such as the ones just mentioned, are also responsible for the up-regulation of serotonin transport, thus increasing its depletion extracellularly (Maes et al., 2002). In vitro, serotonergic antidepressants increase release 15

of serotonin and inhibit its reuptake, leading to low intracellular and increased extracellular levels of serotonin. Low intracellular serotonin levels appear to decrease cytokine levels through the inhibition of mRNA expression (Maes, 2001), while increased extracellular serotonin appears to exert an immunosuppressive effect (Kubera et al., 2000).

Clinical and experimental approaches indicate that antidepressants can also attenuate brain expression of pro-inflammatory cytokines and evoke neuroprotective and immunomodulatory effects (Sluzewska et al., 1995; Xia et al., 1996; Yirmiya et al., 2001;

Castanon et al., 2002; Hashioka et al., 2007; Hwang et al., 2008). In clinical studies, the therapeutic effects of antidepressants seem to be related to the immune status of depressed patients when treatment is initiated. For example, when depressed patients had enhanced immune activation, antidepressants attenuated secretion of cytokines. Elevated plasma levels of IL-6 in patients suffering from acute depression were reduced when these patients were treated with fluoxetine (Sluzewska et al., 1995). Conversely, when immunity was not altered in depressed patients at the initiation of treatment, antidepressants had no effects on immune function.

In animal models, imipramine and fluoxetine produced immune suppression and anti-inflammatory effects by suppressing the production of cytokines such as TNF-, IL-

1, and IL-6 by glial cells (Ha E et al., 2006, Lim et al., 2009; Liu et al., 2011).

Imipramine inhibited interferon (IFN)- stimulated microglial production of IL-6 and nitric oxide (Hashioka et al., 2007), and TNF- production in microglia and astrocyte

cultures (Hwang et al., 2008). In addition to the effects on immune function, antidepressants can also modulate behavior. Specifically, imipramine treatment ameliorated LPS-induced depressive-like behavior in rats, decreased anhedonia, anorexia, weight loss, reduced social, locomotor, and exploratory behaviors (Yirmiya, 1996;

Yirmiya et al., 2001). In mice subjected to social stress, 28 days of chronic administration of fluoxetine or imipramine, but not acute administration (1 day), improved social interaction in the social avoidance behavioral test (Berton et al., 2006;

Tsankova et al., 2006).

Taken together, these clues prompted us to explore the effect of antidepressant imipramine on the neuroimmune system in the context of psychosocial stress.

Imipramine has been on the market for more than 50 years and is effective in treating mood disorders, specifically major depression, and treatment resistant depression.

Imipramine’s mechanism of action in relation with the monoaminergic system has been well established, however further research needs to be done in order to determine its effects on the cathecolaminergic system. Therefore, the aim of the study presented in

Chapter 3 was to evaluate if imipramine could affect stress-induced shift in myelopoiesis, and trafficking of these MPCs to blood, spleen, and brain, and associated anxiety-and depressive like behaviors. Moreover, recent findings from our laboratory showed that

RSD promoted long-lasting microglial activation associated with social avoidance behavior, which was maintained for at least 24 days after cessation of RSD (Wohleb et al.

2013). Thus, other objectives were to determine if imipramine treatment reversed RSD- induced long-lasting social avoidance behavior and if imipramine treatment could reduce 17

the long-lasting neuroinflammatory profile that was still present 24 days after stress termination.

Chapter 2: GABAergic modulation with classical benzodiazepines prevents stress-

induced neuro-immune dysregulation and behavioral alterations

2.1 Introduction

As stated in Chapter 1, stress activates the HPA axis and the SNS, triggering the release of catecholamines, GCs and pro-inflammatory cytokines. Converging translational evidence suggests that stress-induced release of sympathetic, neuroendocrine, and immune molecules signal back to the CNS contributing to the development of behavioral changes. It has been demonstrated that the activity of the HPA axis to stressful stimuli is regulated by CRH neurons. Recent studies have shown that

CRH neurons are to a great extent governed by robust GABAergic afferents. Despite the well-recognized function of GABAergic control of the HPA axis during stress-adaptive processes, less well established is how modulating GABAergic activity with classical

BDZs can regulate psychosocial stress reactivity. Specifically, the mechanism whereby commonly prescribed anxiolytics such as BDZs influences the SNS and consequently peripheral and central immunity during psychosocial stress, needs to be

further elucidated. Therefore the purpose of this project was to test the overarching hypothesis that enhancing GABAergic activity with classical BDZs, lorazepam and clonazepam, prevent stress-induced HPA and SNS activity thus inhibiting stress-induced peripheral and central immune dysregulation, and associated behavioral alterations.

Aim 1: To test the hypothesis that enhancing GABAergic activity with classical BDZs, clonazepam and lorazepam attenuates stress-induced peripheral immune dysregulation.

A) The first objective of this aim was to assess if enhancing GABAergic activity with lorazepam and clonazepam, decreased stress-induced levels of CRH, corticosterone, norepinephrine, and pro-inflammatory cytokines in blood and splenocytes.

B) The second objective of this aim was to determine the contribution of enhancing

GABAergic activity with classical BDZs, lorazepam and clonazepam, on stress-induced development and egress of MPCs cells from BM to blood, and spleen.

Aim 2: To test the hypothesis that stressed-induced accumulation of macrophages in the brain, neuroinflammatory signaling, and associated stress-related behaviors are prevented by increasing GABAergic activity with lorazepam and clonazepam treatment.

A) The first objective of this aim was to ascertain the degree to which enhancing

GABAergic activity with lorazepam and clonazepam could decrease stress-induced 20

increased accumulation of macrophages in the brain and attenuate the expression of pro- inflammatory molecules in microglia.

B) The second objective was to elucidate the effect of enhancing GABAergic activity with lorazepam and clonazepam on stress-induced anxiety- and depressive-like behavior.

2.2 Methods

2.2.1 Animals

Male C57BL/6 (6–8 weeks old) and CD-1 (12 months, retired breeders) mice were purchased from Charles River Breeding Laboratories (Wilmington, Massachusetts) and allowed to acclimate to their surroundings for 7–10 days prior to initiation of experiments. C57BL/6 mice were housed in cohorts of three and CD-1 mice were singly housed in 11.5 x 7.5 x 6 inch polypropylene cages. Mice were maintained at 21°C under a

12:12 h light: dark cycle with ad libitum access to water and rodent chow in the animal facility at The Ohio State University. All procedures were in accordance with the NIH

Guidelines for the Care and Use of Laboratory Animals and approved by The Ohio State

University Institutional Laboratory Animal Care and Use Committee.

2.2.2 RSD

RSD was performed as described previously (Avitsur et al., 2001; Wohleb et al.,

2011). In brief, an intruder male CD-1 mouse was introduced into home cages of male

C57BL/6 mice (three per cage) for 2 hours (hrs) on 6 consecutive nights. Behavior was observed to make certain that the intruder was aggressive. If the CD-1 mouse did not initiate an attack within 5-10 minutes or was attacked by resident mice, a new CD-1 mouse was introduced. At the end of the 2 hrs the CD-1 mouse was removed and the resident mice were left undisturbed until the next day when the same paradigm was repeated. During RSD, resident mice display submissive behaviors such as upright posture, fleeing, and crouching (Avitsur et al., 2001; Hanke et al., 2012). Home cage control (HCC) cohorts were left undisturbed in a separate room.

2.2.3 Pharmacological treatments and administration procedures

Figure 1A shows experimental schematic. C57BL/6 mice were randomly selected for inclusion into experimental treatment groups using two different drugs, lorazepam and clonazepam. For lorazepam experiments, adult male C57BL/6 mice were allocated into different experimental groups. The groups were: RSD treated with vehicle (VRSD),

RSD treated with lorazepam (LRSD), HCC treated with vehicle (VCON) and HCC treated with lorazepam (LCON). Lorazepam (100 µg/kg) was diluted in distilled water, and a vehicle control solution (0.9% NaCL) was administered. The dose of lorazepam 22

(100µg/kg/day) was chosen based on prior studies (Miller et al., 1987, de Angelis et al.,

1995). A pilot study to determine dose-response in the open field test was also performed at this concentration in our laboratory. In this pilot study, only HCC were injected for eight days with lorazepam (100 µg/kg) or vehicle (n=6 per group) and locomotor activity was assessed at day eight. No differences were detected in the open field test in total distance traveled, total time spent in the periphery, and total time spent in the center.

Spleens were weighed and no differences were detected between the groups (data not shown). These results demonstrated that lorazepam, at this concentration, did not affect locomotor activity in these mice or spleen weight. In experiments using vehicle administration, preliminary studies indicated that the response of vehicle-injected and non-treated mice, HCC, were indistinguishable (data not shown). Lorazepam, or vehicle, was administered by intraperitoneal (i.p.) injection in a volume of 0.01ml/g body weight

(i.e. for a mouse that weights 30 grams the volume injected was 0.3ml). Mice were pre- treated with lorazepam or vehicle, two days before starting RSD at 1900 EST. From day

1 to day 6 of RSD, LRSD and VRSD groups each received their respective treatment immediately after the RSD cycle. LCON and VCON groups were injected at this time as well. It was expected that mice would become sedated with lorazepam 2 hours post injection, but they could can still ambulate and remain responsive.

For clonazepam experiments, the groups were: RSD treated with vehicle (VRSD),

RSD treated with clonazepam (CRSD), HCC treated with vehicle (VCON) and HCC treated with clonazepam (CCON). Clonazepam was diluted in saline (0.9% NaCL) and a vehicle control solution was administered. The dose of clonazepam (50g/kg) was 23

chosen based on literature in a model of restraint stress (Verleye et al., 2008). A pilot study was conducted with HCC in order to obtain the same dose-response behavioral and spleen weight data obtained with lorazepam (data not shown). Clonazepam or vehicle was administered i.p. in a volume of 0.01 ml/g body weight (i.e. for a mouse that weighed 30 grams the volume injected was 0.3ml). Mice were pre-treated with clonazepam or vehicle, two days before starting RSD at 1900 EST. From day 1 to day 6 of RSD, CRSD and VRSD groups received their respective treatment immediately after the RSD cycle. VCON and CCON groups were injected at this time too.

The experiments described in the following paragraphs were performed in different replicates of mice. Each replicate of mice consisted of four different experimental groups. Each consisted of three mice receiving a different treatment. The experiments were done two or three times to confirm the findings in different replicates.

For example the social avoidance test, and the collection of bone marrow, spleen, brain

(and brain slice) and blood for flow cytometry was done in two replicates. In a second experiment anxiety-like behavior, plasma IL-6, and isolation of microglia and total splenocytes was done in another replicate, and confirmed in a second replicate. In a third experiment, plasma was collected for corticosterone and norepinephrine, and the HYPO was dissected in order to evaluate gene expression of CRH in two replicates. Spleen weight was assessed in each mouse to evaluate the RSD effect.

2.2.4 Plasma corticosterone determination

Blood was collected from the submandibular vein immediately following the last cycle of RSD (i.e. at 1900 EST). Blood samples were immediately placed on ice and plasma was collected after centrifuging samples at 1800 x g for 15 minutes at 4C. Plasma was stored at -80C until assayed in duplicate using a Corticosterone ELISA kit from Enzo

Life Sciences (Farmingdale, New York) following manufacturer’s instructions (n=6-7 per group).

2.2.5 Plasma norepinephrine determination

Blood was collected from the submandibular vein, immediately after the last cycle of RSD (i.e. at 1900 EST). Blood samples were placed in lithium-heparin tubes from

Fisher Scientific (Waltham, MA), and then centrifuged at 3500 rpm for 15 minutes at

22C. Plasma supernatant was pulled and immediately stored at -80C until assayed in duplicate using a Noradrenaline Research ELISA kit (BA E-5200) from Labor

Diagnostika Nord GmbH & Co. KG (Nordhorn, Germany) following manufacturer’s instructions (n=5-6 per group).

2.2.6 Anxiety-like behavior testing and social avoidance test

Behavior was determined the morning following the last cycle of RSD as previously described (Wohleb et al., 2011). Each behavioral test was assessed in different cohorts of mice. Anxiety-like behavior was measured in the open field test, consisting of a 40 x 40 x 25 cm square plexiglass box with a solid floor and a 6 x 6 grid drawn on the floor separating the open field into 36 identical squares. This test takes advantage of mice’s natural tendencies to explore the environment while avoiding open spaces. For this experiment, the dependent variables were: total time spent in the center of the open field, latency to entering the center, and the number of entries into the center of the open field in a 5 min period. The test apparatus was cleaned with water between experimental subjects. Behavior was taped, digitized and coded using the Observer program (Noldus

Information Technologies, Netherlands) (n=7-11 per group for lorazepam experiment; n=6 per group for clonazepam experiment).

The social avoidance test was performed as previously described (Wohleb et al.,

2013). Briefly, the social avoidance test consists of two trials. In the first trial, an experimental mouse was placed into the arena with an empty wire mesh cage and activity was recorded for 2.5 minutes. In the second trial, the experimental mouse was placed in the arena with an unfamiliar CD-1 mouse in the wire mesh cage and activity was recorded for the same amount of time. Time in the interaction zone and time spent in the corners was video-recorded and analyzed using Noldus EthoVision Software (Leesburg,

Virginia). Fourteen hours after the sixth cycle of RSD, mice were killed by CO2 26

asphyxiation after the completion of either anxiety-like behavior testing or the social avoidance test and tissues were collected for processing (n=9-11 for lorazepam experiment; n=8-9 per group for clonazepam experiment).

2.2.7 Plasma IL-6 determination

Blood was collected by cardiac puncture with EDTA lined 1ml syringes. Whole blood was centrifuged; plasma was collected and stored at -80C until assayed. IL-6 levels were determined using the BD OptEIA Mouse IL-6 ELISA as previously described

(Wohleb et al., 2011) according to the manufacturer’s instructions (BD Biosciences, San

Jose, California). Briefly, 96-well enzyme immunoassay plates were coated with anti- mouse IL-6 capture antibody and incubated overnight at 4°C. Samples and IL-6 standards

(0–1000 pg/ml) were added and incubated for 2 hrs at room temperature. Plates were washed and incubated with biotinylated anti-mouse IL-6 antibody. Then, plates were washed and incubated with streptavidin-horseradish peroxidase conjugate. After 1 hr incubation at room temperature, plates were washed and incubated with tetramethylbenzidine liquid substrate for 15 min. Reactions were ended and absorbance was read at 450 nm using a Spectramax Plus 384 plate reader (Molecular Devices,

Sunnyvale, California). The assay was sensitive to 10 ng/ml IL-6 and the interassay and intra-assay coefficients of variation were <10% (n=6 per group).

2.2.8 Hypothalamus and slice of hippocampus/hypothalamus RNA isolation and real-time

PCR

For RNA isolation, brains were removed and dissected for the different brain regions: hypothalamus and a 1 mm coronal brain section (slice) through the hippocampus/hypothalamus and immediately flash frozen in liquid nitrogen. RNA was isolated from homogenized brain regions using tri-reagent/isopropanol precipitation,

RNA was reverse transcribed to cDNA using an RT-RETROscript kit (Ambion, Life

Technologies, Grand Island, New York) and RNA concentration was determined by spectrophotometry (Eppendorf, Hauppauge, New York) as described previously (Wohleb et al., 2013). Quantitative PCR was performed using the Applied Biosystems Assay-on-

Demand Gene Expression protocol as previously described (Wohleb et al., 2011). In brief, experimental cDNA was amplified by real-time PCR where a target cDNA (e.g.,

IL-1β, IL-6, TNF-α for brain slice and CRH for hypothalamus) and a reference cDNA

[glyceraldehyde-3-phosphate dehydrogenase (GAPDH)] were amplified simultaneously using an oligonucleotide probe with a 5’ fluorescent reporter dye (6-FAM) and a 3’ quencher dye (nonfluorescent quencher). Fluorescence was determined on an ABI

PRISM 7300-sequence detection system (Applied Biosystems, Life Technologies,

Grand Island, New York. Data were analyzed using a comparative threshold cycle method and results are expressed as a fold difference from GAPDH (n=5-6 per group for hypothalamus; n=6 per group for brain slice).

2.2.9 Microglia isolation and flow cytometry

Microglia was isolated from whole-brain homogenates as previously described

(Wohleb et al., 2011). In brief, brains were homogenized in Hanks’ balanced salt solution

(HBSS), with a pH of 7.4, by mashing the brain through a 70 m nylon mesh cell strainer. The homogenates were then centrifuged at 500 X g for 6 min at 10C.

Supernatants were decanted and the obtained pellets were resuspended in 70% isotonic

Percoll (GE Healthcare, Pittsburgh, Pennsylvania) at room temperature. A Percoll density gradient was layered in this manner: 70%, 50%, 35%, and 0% isotonic Percoll. The gradient was centrifuged at 2000 X g for 20 minutes at 10°C and microglia was taken by aspirating the interphase between the 50% and 70% Percoll layers (Wohleb et al., 2011).

The retrieved cells were washed and resuspended in sterile HBSS and centrifuged at 600

X g for 6 minutes at 10C. The supernatant was decanted and viable cells were counted using 0.1% trypan blue staining in an automated cell counter (Luna-FLTM dual fluorescence cell counting, Logos Biosystems, Annandale, Virginia). Each brain yielded approximately 6.5 x 10 5 cells. Studies (Henry et al., 2009; Wynne et al., 2010) demonstrate that viable cells isolated from brain homogenates through this protocol yields >90% microglia. For flow cytometry, staining of microglia surface antigens was executed as previously described (Henry et al., 2009 and Wohleb et al., 2011). In brief,

Fc receptors were blocked with anti-CD16/CD32 antibody (eBioscience, San Diego,

California). The cells were then incubated with anti-CD11b-APC, anti-CD45-FITC, and anti-MHC-II-PE antibodies (eBioscience, San Diego, California). Expression of these 29

surface receptors was evaluated using a Becton-Dickinson FACSCalibur™ four-color cytometer. Ten thousand events were recorded and microglia were identified by the expression of CD11b+ and CD45low markers and macrophages by the expression of

CD11b+ and CD45high (Nair and Bonneau, 2006). For each antibody, gating was determined based on appropriate isotype-stained controls. Flow data were analyzed using

FlowJo software (Tree Star, San Carlos, California).

2.2.10 Microglia RNA isolation and real-time PCR

RNA from Percoll-isolated microglia was isolated using the RNeasy plus mini-kit

(Qiagen). RNA concentration was determined by a spectrophotometer (Eppendorf,

Hamburg, Germany). RNA was reverse transcribed to get cDNA with the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) and quantitative PCR was done using the Applied Biosystems by Life Technologies Assay-on-Demand Gene

Expression protocol as previously described (Wohleb et al., 2011). Briefly, amplification of cDNA was performed by real-time PCR. A target cDNA (IL-1 β, IL-6, TNF-α, TLR4,

CCL2, and CX3CR1) and a reference cDNA (glyceraldehyde-3-phosphate dehydrogenase) were amplified at the same time using an oligonucleotide probe with fluorescent reporter dye (FAM). An ABI PRISM 7300-sequence detection system

(Applied Biosystems by Life Technologies, Grand Island, New York) was used to assess fluorescence. Data were analyzed by the comparative threshold cycle and the

results are given as the fold difference detected (n=8-11 per group for lorazepam experiments; n=6 per group for clonazepam experiments).

2.2.11 Total splenocytes RNA isolation and real-time PCR

Spleens were collected, weighed, and flash frozen in liquid nitrogen. Samples were stored at -80C until assayed. Samples were homogenized using a tissue tearor and then centrifuged at 4C and 12,000g for 10 minutes. Supernatant was collected and each sample received 200ul of chloroform, followed by vortexing and centrifugation at 12,000 x g for 15 minutes. The upper aqueous phase was decanted into a fresh tube, to which isopropanol was added to precipitate the RNA. This solution was centrifuged at 12,000 x g for 10 minutes, supernatant was removed, and the resulting pellet was washed with 1ml

75% EtOH to remove residual protein. The pellet was resuspended in 20ul nuclease free water. RNA concentration was measured by spectrophotometry (Implen, Westlake

Village, CA) and RNA was reverse transcribed to cDNA using a High Capacity cDNA

Reverse Transcription Kit (Applied Biosystems, Foster City, CA). Quantitative PCR was performed using a Taqman Gene Expression Assay to observe for potential changes in expression of pro-inflammatory cytokines (i.e. IL-1, IL-6, TNF-). Expression was analyzed using an ABI Prism 7000 Sequence Detection System (Applied Biosystems;

Foster City, CA) using the 2-ΔΔ Ct method with normalization glyceraldehyde-3 phosphate dehydrogenase (GAPDH) (n= 8-11 per group for lorazepam experiment; n=6 per group for clonazepam experiment). 31

2.2.12 Flow cytometry of bone marrow, blood, and spleen

Blood was collected after cardiac puncture with an EDTA filled syringe and bone marrow was removed, placed in ice-cold HBSS and processed. Bone marrow was flushed from bone into a 70µm strainer with 1xPBS into a 50 ml conical, until it reached a volume of 7.5 ml. Samples were centrifuged at 1800rpm for 8 minutes. Supernatant was poured off and the pellets of cells were brought up in 1ml FACS buffer and vortexed.

Fifty µl of each sample was added into a flow tube. Fc receptors of bone marrow cells and blood leukocytes were blocked with anti-CD16/CD32 antibody (eBioscience, San

Diego, California) as previously described (Wohleb et al., 2014). Cells were washed and then incubated with the appropriate antibodies (CD45, CD11b, CD34; eBioscience, San

Diego, California), and Ly6C (BD Biosciences, San Jose, California) for 1 h at 4C. Cells were washed and re-suspended in FACS buffer for analysis. Non-specific binding was assessed by using isotype-matched antibodies. Antigen expression was determined using a Becton-Dickinson FACSCalibur four-color cytometer (BD Biosciences, San Jose,

California).

For flow cytometric assessment of splenocytes, spleens were collected, weighed, placed in ice-cold HBSS and disrupted mechanically to obtain single cell suspensions as previously described (Hanke et al., 2012). Red blood cells were lysed by adding 2 ml of room temperature lysis buffer (0.16 M NH4Cl, 10 mM KHCO3, and 0.13 mM EDTA) for

2 min, followed by one wash with HBSS/10% heat inactivated fetal bovine serum (FBS).

Each cell pellet was resuspended in HBSS, filtered and washed a final time in HBSS. 32

Cells were counted and samples were resuspended (2.5 x 106 cells/ml) in supplemented

RPMI medium (10% heat-inactivated FBS, 0.075% sodium bicarbonate, 10 mM HEPES buffer, 100 U/ml penicillin G, 100 mg/ml streptomycin sulfate, 1.5 mM L-glutamine, and

0.0035% 2-mercaptoethanol). Single cell suspensions derived from spleen tissue (1 x106 cells per sample) were incubated with 1 g each of fluorescently labeled monoclonal antibodies (or the appropriate isotype controls). Antibody labeling was performed at 4 C for 45 min. The cells were then washed twice in PBS containing 1% FBS and 0.09%

NaN3. All antibodies were obtained from BD PharMingen (San Jose, California), APC- labeled anti-CD11b, PerCP-labeled anti-Gr-1/Ly-6C. Leukocytes were gated based on forward versus side scatter and a total of 10,000–100,000 events were analyzed on a

FACSCalibur flow cytometer using Cell Quest and Cell Quest Pro analysis software

(Becton-Dickenson; San Jose, California).

2.3 Statistical Analysis

To ensure a normal distribution, gathered data were subjected to Shapiro-Wilk tests using SPSS Statistics version 21 (IBM, Armonk, New York). Data were expressed as means ± SEM. To determine significant main effects and interactions between variables being evaluated, data were subjected to one-way (trial, stress, lorazepam, clonazepam), two-way (trial x stress); (trial x lorazepam); (trial x clonazepam); (stress x lorazepam); (stress x clonazepam) or three-way (trial x stress x lorazepam); (trial x stress x clonazepam) analysis of variance (ANOVAs). When appropriate, differences between 33

treatment group means were analyzed by an F-protected t test using the least-significant difference method. In all cases, the level of significance was set at p  .05.

2.4 Results

2.4.1 Effects of lorazepam and clonazepam on RSD increased levels of CRH, corticosterone and norepinephrine

It has been suggested BDZs have an anti-stress effect caused primarily by the inhibition of CRH release, followed by a decrease in corticosterone in response to stress

(File, 1991). Therefore, the objective of this experiment was to assess mRNA expression of CRH in the hypothalamus of mice exposed to RSD and the effect of the BDZs lorazepam and clonazepam on this expression. Furthermore, other objectives were to determine the effect of BDZs on stress-induced increased levels of plasma corticosterone and norepinephrine. The day after the last cycle of RSD, brains were collected and the hypothalamus was dissected, immediately flash frozen, and processed for RNA isolation and RT-qPCR. Figures 1B and 1C show RSD mice displayed higher expression of CRH in this brain region (main effect of stress in lorazepam experiment; F (1,22) = 17.15, p <

0.01; main effect of stress in clonazepam experiment; F(1,23) = 50.12, p< 0.01) and this response was attenuated with lorazepam treatment (stress x lorazepam interaction; F(1,22)

= 4.09 , p =0.05). In the same manner, clonazepam diminished expression of CRH in the hypothalamus of RSD mice (main effect of clonazepam; F(1,23) =7.48, p=0.01; stress x 34

clonazepam interaction; F(1,23)= 2.88 , p =0.10). Post-hoc analyses revealed RSD mice treated with clonazepam had CRH levels comparable to HCC and RSD mice treated with vehicle (p>0.05), whereas RSD mice treated with vehicle had increased levels of CRH

(p<0.05) compared to HCC.

Previous studies demonstrated that mice had significantly elevated levels of serum corticosterone immediately after the last cycle of RSD compared to HCC mice (Hanke et al., 2012). Figures 1D and 1E show that immediately after the last cycle of RSD, increased levels of corticosterone were found in plasma of mice subjected to RSD (main effect of stress in lorazepam experiment; F(1,24) = 13.02, p < 0.01; main effect of stress in clonazepam experiment; F(1,23) = 16.59, p< 0.05). The increased plasma corticosterone provoked by stress exposure was attenuated with both lorazepam and clonazepam treatment (stress x lorazepam interaction; F(1,24) = 5.45 , p <0.05 ; stress x clonazepam interaction; F(1,23)= 4.59, p<0.05).

RSD increased levels of norepinephrine in plasma, immediately after six cycles of social defeat, and returned to control levels by three hours after termination of the stressor (Hanke et al., 2012). Thus, we collected blood via submandibular vein immediately after the last cycle of RSD from the different experimental groups to evaluate if BDZs lorazepam and clonazepam mitigated levels of norepinephrine in plasma. As shown in Figures 1F and 1G, there were amplified levels of norepinephrine in plasma of socially defeated mice (main effect of stress in lorazepam experiment;

F(1,23)=73.59, p<0.01). Lorazepam decreased norepinephrine levels in stressed mice

(stress x lorazepam interaction; F(1,23)=35.57, p<0.01) comparable to HCC (p>0.05). On 35

the other hand, clonazepam treatment attenuated norepinephrine levels in stressed mice

(main effect of stress; F(1,22)=41.42, p<0.01; stress x clonazepam interaction; F(1,22)=5.61, p<0.05), but they were still high compared to HCC (p<0.05). These data demonstrated both lorazepam and clonazepam were effective in attenuating stress-induced enhanced mRNA expression of CRH in the hypothalamus and increased corticosterone levels in plasma. Regarding norepinephrine in plasma, lorazepam reduced levels of this hormone in stressed mice comparable to HCC, however stress-induced increased levels of norepinephrine remained high with clonazepam treatment compared to HCC, levels were just attenuated with treatment.

2.4.2 Lorazepam and clonazepam prevented stress-induced increase of IL-6 in plasma

We had previously shown RSD consistently promoted increased plasma levels of

IL-6 (Stark et al., 2002). Thus, the IL-6 response was used as a biomarker to gauge the robustness of the response to RSD (Hanke et al., 2012). Blood was collected the day after the last cycle of RSD and plasma IL-6 levels were determined. As expected, social defeat increased IL-6 levels in plasma (Figure 2A and 2B), and this increase was blocked by lorazepam and clonazepam treatment (main effect of stress; F(1,23) = 8.54, p<0.01; stress x lorazepam interaction; F(1,23)= 9.01, p <0.01; main effect of stress; F(1,23)=47.61, p< 0.01; stress x clonazepam interaction; F(1,23)= 43.05, p < 0.01), respectively. Together, these data indicate that anxiolytics such as lorazepam and clonazepam effectively decreased high levels of circulating IL-6 in plasma of C57BL/6 mice subjected to social defeat. 36

2.4.3 Effect of lorazepam and clonazepam on stress-induced alteration of bone marrow hematopoiesis

RSD enhances myelopoiesis and promotes the development, priming, and egress of a MPC population from the BM (Engler et al., 2004). Hence, the aim of this experiment was to determine the effects of BDZs, lorazepam and clonazepam, on stress- induced alterations in hematopoiesis in the BM (Figures 3A to 3D). BM was collected the next day after the last cycle of RSD from the different experimental groups, and CD31 and Ly-6C+ expression were determined by flow cytometry. Figure 3C shows, RSD mice treated with vehicle, had a decreased percentage of lymphocyte progenitors in the BM

(main effect of stress; F (1,22)= 41.00, p<0.01) and this effect was attenuated with lorazepam treatment in RSD mice (main effect of lorazepam; F (1,22) = 9.00, p<0.01), nonetheless, lymphocyte progenitors in RSD mice treated with lorazepam was still low compared to HCC (p<0.05). Lorazepam blocked stress-induced decrease in erythrocyte progenitors (main effect of stress; F (1,22) = 25.50, p<0.01; stress x lorazepam interaction;

F(1,22)=16.52, p<0.01) in the BM. Moreover, lorazepam blocked stress-induced increase percentage of granulocyte (main effect of stress; F(1,22)= 40.02, p<0.01; stress x lorazepam interaction; F(1,22) =14.96, p<0.01) and attenuated stress-induced increase of monocyte progenitors in the BM (main effect of stress; F(1,22)=47.89, p<0.01; stress x lorazepam interaction; F(1,23) =16.67, p<0.01). Monocyte progenitors in mice subjected to

RSD and treated with lorazepam, were still high compared to HCC (p<0.05).

As shown in Figure 3D, clonazepam did not have an effect on the reduction of lymphocyte progenitors in the BM as a consequence of RSD (main effect of stress; F (1,20)

=5.88, p<0.01; stress x clonazepam interaction; F(1,20)=0.05, p>0.05). Stress-induced decrease of erythrocyte progenitors in the BM (main effect of stress; F (1,20) =3.57, p<0.05) was attenuated with clonazepam (stress x clonazepam interaction; F (1,20) = 8.02, p =0.01), but they were still low compared to HCC (p<0.05). Additionally, stress-induced increase of granulocytes and monocytes (main effect of stress; F (1,20)=4.97, p<0.05; main effect of stress; F (1,20)=3.17, p=0.05; respectively) was not affected by clonazepam treatment in RSD mice. There were still an increased percentage of these cells in the BM compared to HCC (p<0.05). In sum, the BDZs lorazepam and clonazepam have different effects on stress-driven hematopoiesis.

2.4.4 Both lorazepam and clonazepam inhibit stress-induced increased percentage of monocytes and granulocytes in circulation

Following behavioral testing, mice were euthanized and blood was collected by cardiac puncture, and Ly-6C/CD11b labeling was performed to evaluate the percentage of monocytes and granulocytes in circulation using flow cytometry. As reported previously (Wohleb et al., 2013) and shown in Figures 4A and 4B, mice exposed to RSD had increased numbers of monocytes (main effect of stress; F (1,22) = 12.30, p< 0.01) and granulocytes in circulation (main effect of stress; F (1,22)= 4.26, p=0.05). Figure 4C shows lorazepam decreased stress-induced increase in monocytes (stress x lorazepam 38

interaction; F (1,22)=4.60, p<0.05) and granulocytes release from bone marrow into circulation (stress x lorazepam interaction; F (1,22)=14.96, p<0.01) (Figure 4D). In the same manner, in the clonazepam experiment, mice subjected to RSD and treated with vehicle, had an increased pool of monocytes (main effect of stress; F (1,20) =4.46, p=0.05) and granulocytes in circulation (main effect of stress; F (1,20) =12.09, p<0.01).

Clonazepam prevented the stress-driven increase in monocytes in circulation (Figure 4E)

(main effect clonazepam; F (1,20) =12.09, p=0.01), such that this percent was equal to

HCC (p>0.05). Also, clonazepam prevented stress-induced increase in granulocytes

(Figure 4F) (stress x clonazepam interaction; F (1,20) =12.09, p<0.01) in circulation. These data suggest that lorazepam and clonazepam prevented the release of monocytes and granulocytes from the BM into circulation, in mice subjected to social defeat.

2.4.5 Lorazepam, but not clonazepam, inhibited stress-induced splenomegaly

It was reported previously that RSD caused an increase in the number and percentage of monocytes and granulocytes in the spleen, which was associated with the development of splenomegaly (Avitsur et al., 2003). In line with this idea, total splenocytes were stained with Ly-6c/CD11b antibodies and flow cytometry was used in order to determine if lorazepam and clonazepam would affect the pool of these cells in the spleens of mice subjected to RSD (Figures 5A and 5B). Flow cytometric analysis indicated that stress-induced increase in monocytes (Figure 5C) (main effect of stress; F

(1,22) =10.32, p<0.01) was completely blocked by lorazepam treatment (stress x lorazepam 39

interaction; F (1,23)= 4.65, p<0.05) and this drug also reversed the stress driven increase in splenic granulocytes (Figure 5D) (main effect of stress; F (1,22) = 8.06, p = 0.01; stress x lorazepam interaction, F (1,22)= 4.18, p=0.05). A greater percentage of monocytes (Figure

5E) and granulocytes (Figure 6F) was still observed in the spleen of RSD animals treated with vehicle and clonazepam (main effect of stress; F (1,20)=4.97, p<0.05; stress x clonazepam interaction; F(1,20)=1.48, p>0.05; main effect of stress; F(1,20)=4.34, p=0.05; stress x clonazepam interaction; F(1,20)=2.79, p>0.05; respectively).

As mentioned previously, RSD has been shown to increase spleen size and weight

(i.e., splenomegaly) that is associated with an increased number of monocytes/macrophages (Avitsur et al., 2003). Therefore, we next aimed to determine if treatment with lorazepam and clonazepam would attenuate RSD-induced splenomegaly.

Lorazepam treatment significantly inhibited social defeat induced splenomegaly (Figure

5G) (main effect of stress; F(1,46)= 44.98, p<0.01; stress x lorazepam interaction, F(1,46) =

31.77, p <0.01). Interestingly, clonazepam attenuated splenomegaly (main effect of stress; F(1,47) = 34.72, p<0.01; stress x clonazepam interaction; F(1,47)= 5.98, p <0.05), nonetheless, post-hoc analysis indicate spleen weight in RSD mice treated with this drug was still high compared to HCC (p<0.05) as shown in Figure 5H.

We also sought to determine if both drugs would block RSD-related pro- inflammatory cytokine production by total splenocytes, specifically, IL-1, IL-6, and

TNF- (Table 1A and Table 1B). Lorazepam blocked mRNA expression of IL- (main effect of stress; F (1,40)= 6.15, p <0.05; stress x lorazepam interaction; F(1,40)= 5.67, p<0.05), IL-6 (main effect of stress; F(1,40)=12.94, p<0.01; stress x lorazepam 40

interaction; F(1,40)=3.84, p=0.05), and TNF- in total splenocytes (main effect of stress;

F(1,40) = 4.5, p< 0.05; main effect of lorazepam; F(1,40) = 14.68, p<0.01). On the contrary, clonazepam did not have an effect on stress-induced increase of mRNA expression of IL-

1 (main effect of stress; F (1,23)=14.23; stress x clonazepam interaction; F (1,23)=0.82 , p>0.05) . Clonazepam did not decrease mRNA expression of IL-6 and TNF- in total splenocytes of mice subjected to RSD (main effect of stress; F (1,23) = 9.30 p<0.01; stress x clonazepam interaction; F(1,23) =0.22 , p =0.07; main effect of stress; F(1,23)=7.99, p=0.01; stress x clonazepam interaction, F(1,23)= 0.17, p=0.08; respectively). Post-hoc analyses revealed these latter cytokines in the RSD group treated with clonazepam were not significantly different from HCC and RSD mice treated with vehicle (p>0.05). In sum, RSD promoted a pro-inflammatory phenotype in the spleen, increased pool of granulocytes and monocytes in the spleen, and splenomegaly, and these three effects were completely suppressed by lorazepam but partially suppressed with clonazepam treatment.

2.4.6 Lorazepam and clonazepam, prevented stress-induced trafficking of bone marrow- derived CD11b+ myeloid cells to the brain

RSD has been shown to alter the phenotype of myeloid-derived CD11b+ cells in the periphery (Bailey et al., 2009) and increase their propensity to traffic from the bone marrow to the brain (Wohleb et al., 2011). Therefore, we next sought to determine the effect of lorazepam and clonazepam on the trafficking of circulating MPCs to the CNS in 41

RSD mice. Figures 6A and 6B show representative bivariate dot plots of CD11b and

CD45 staining for macrophages (CD11b+/CD45high) (Nair et al., 2007) for the various experimental groups. The percentage of macrophages after RSD was increased, however, this was ablated by both lorazepam (main effect of stress; F (1,22)= 5.72, p<0.05; stress x lorazepam interaction; F (1,22) = 4.52, p <0.05) and clonazepam treatment (main effect of stress; F (1,20) = 3.13, p=0.05; p<0.05; stress x clonazepam interaction; F(1,20) = 6.19 , p

<0.05) as seen in Figures 6C and 6D. These data suggest that BDZ, lorazepam and clonazepam were effective in preventing RSD-induced accumulation of macrophages in the brain.

2.4.7 Lorazepam and clonazepam ameliorated neuroinflammatory signaling

To determine if the RSD-induced neuroinflammatory profile was attenuated by lorazepam and clonazepam treatment, gene expression of pro-inflammatory cytokines and activation markers was examined in microglia (Tables 2A and 2B). For lorazepam experiments, RSD mice treated with vehicle had increased relative gene expression of IL-

6 (main effect stress; F (1,37)= 7.89, p< 0.01), IL1- (main effect stress; F(1,37)= 4.17, p<

0.05), TNF- (main effect stress; F(1,37)= 8.02, p< 0.01), and activation markers TLR-4,

CCL2, and CX3CR1 (main effect stress; F(1,36)= 4.36, p< 0.05; main effect stress; F(1,23)=

11.24, p< 0.01; main effect stress; F(1,23)= 6.77, p< 0.01; respectively) and this effect was prevented with lorazepam (main effect of lorazepam; F(1,37)= 4.30, p< 0.05; F(1,37)= 8.94, p< 0.01; F(1,37)= 6.17, p< 0.01; F(1,36)= 11.20, p< 0.01; F(1,23)= 9.56, p< 0.01; F(1,23)= 4.70, 42

p< 0.05; respectively) (stress x lorazepam interaction; p<0.05 for all molecules evaluated). Similarly, clonazepam experiments showed a blockade of stress-induced increase of mRNA expression of IL-6 (main effect of stress; F (1,23) = 8.32, p<0.01; main effect of clonazepam; F(1,23)= 5.49, p< 0.05), IL1- (main effect of stress; F(1,23) = 6.98, p=0.01; main effect of clonazepam; F(1,23)= 9.95, p< 0.01), TNF- (main effect of stress;

F(1,23) = 6.60, p=0.01; main effect of clonazepam; F(1,23)= 6.45, p< 0.05), TLR-4 (main effect of stress; F(1,23) = 12.23, p<0.01; main effect of clonazepam; F(1,23)= 7.22, p= 0.01),

CCL2 (main effect of stress; F(1,23) = 4.53, p=0.05; main effect of clonazepam; F(1,23)=

4.54, p< 0.05), and CX3CR1 (main effect of stress; F(1,23) = 5.58, p<0.05; main effect of clonazepam; F(1,23)= 5.14, p<0.05) in microglia. Post-hoc analyses indicated there was no difference in mRNA expression of these pro-inflammatory molecules evaluated between RSD mice treated with clonazepam and HCC (p<0.05 for all molecules assessed).

To further characterize neuroinflammatory signaling, mRNA expression of IL-1, IL-

6, and TNF- was determined in a 1mm coronal brain slice of the hippocampus/hypothalamus (Tables 3A and 3B). These pro-inflammatory cytokines were increased in RSD mice treated with vehicle and blocked with lorazepam treatment

(main effect of stress; F (1,23)=8.76, p<0.01; stress x lorazepam interaction; F(1,23) = 4.74, p<0.05 ; main effect of stress; F(1,23)=5.77, p<0.05; stress x lorazepam interaction;

F(1,23)=4.52, p<0.05 ; main effect of stress; F(1,23) = 5.38, p<0.05; stress x lorazepam interaction; F(1,23) = 4.58, p<0.05; respectively) . Clonazepam was also effective in

blocking mRNA expression of these pro-inflammatory cytokines (main effect of stress; F

(1,23) =13.75, p<0.01; stress x clonazepam interaction; F(1,23) =4.99, p<0.05 ; main effect of stress; F(1,23) = 4.99, p<0.05; stress x clonazepam interaction; F(1,23)=8.45, p<0.01; main effect of stress; F(1,23) =4.93, p<0.05; stress x clonazepam interaction; F(1,23)=4.73, p<0.05; respectively). These data suggest that lorazepam and clonazepam may reduce RSD- induced neuroinflammatory signaling by decreasing gene expression of pro-inflammatory molecules in the CNS.

2.4.8 Lorazepam and clonazepam reversed stress-induced anxiety-like behavior

Consistent with previous studies, RSD promoted anxiety-like behavior in the open field test (Wohleb et al., 2011). The day after the last cycle of RSD, individual mice were placed in the corner of an open field box and allowed to freely explore the apparatus.

Figures 7B and 7E show mice subjected to RSD took longer to first enter the center of the open field compared to HCC mice, and this was reversed by both lorazepam (main effect of stress; F (1,41)= 38.00, p<0.01; stress x lorazepam interaction; F (1,23)=46.53, p<0.01) and clonazepam treatment (main effect of stress; F(1,23)=7.72, p<0.05; stress x clonazepam interaction; F(1,23)=7.27, p<0.05). Figures 7C and 7F show that mice subjected to RSD and treated with vehicle spent less time in the center of the open field compared to HCC, and this was reversed with lorazepam (main effect of stress; F (1,41)=

4.05, p=0.05; stress x lorazepam interaction, F(1,41)=5.73, p<0.05) and clonazepam treatment (main effect of stress; F(1,23) = 9.91, p<0.01; stress x clonazepam interaction; 44

F(1,23)=10.61, p<0.01). Likewise, Figures 7A and 7D show that both drugs were effective in increasing center entries in mice treated with lorazepam and subjected to RSD (main effect of stress; F (1,41)=15.10, p<0.01; stress x lorazepam interaction; F (1,41)= 7.84, p<0.01) and clonazepam treatment (main effect of stress; F(1,23)=9.91, p< 0.01; stress x clonazepam interaction; F(1,23)=10.61, p<0.01). Neither RSD nor benzodiazepine treatment of RSD mice, affected the time spent moving in the open field (main effect of stress; p>0.05; stress x BDZ interaction; p >0.05; for both drugs) (data not shown), indicating that differences were not due to changes in locomotion or activity level.

2.4.9 Lorazepam and clonazepam reversed social avoidance behavior

Along with promoting anxiety-like behavior, RSD triggers depressive-like behavior that can be measured as social avoidance (Wohleb et al., 2013). Therefore, in this experiment the aim was to determine if stress-induced social avoidance behavior could be reversed in mice exposed to RSD with anxiolytics such as lorazepam and clonazepam. The day after the 6th cycle of RSD, social avoidance was determined using a two-trial interaction paradigm with an empty social target trial (Trial 1) followed by a social target trial (Trial 2). Figures 8A and 8B show that mice subjected to RSD spent less time in the interaction zone when an intruder was introduced in both lorazepam and clonazepam experiments (main effect stress; F (1,41)=9.27, p<0.01; main effect stress;

F(1,34)=5.01, p<0.05; respectively). Lorazepam was effective in reversing stress-induced social avoidance behavior in mice subjected to RSD (main effect trial; F (1,38) = 14.12, 45

p<0.01; F (1,38) = 14.12; trial x stress interaction; F (1,38) = 13.82, p<0.01; trial x lorazepam;

F(1,38) = 16.81, p<0.01; trial x stress x lorazepam interaction; F (1,38) = 9.49, p<0.01). In the same manner, clonazepam blocked social avoidance behavior in mice subjected to RSD

(main effect trial; F (1,31)=15.50, p<0.01; trial x stress interaction; F(1,31) =10.02, p<0.01; trial x clonazepam; F(1,31)=8.17, p<0.01; trial x stress x clonazepam interaction; F(1,38) =

9.32, p<0.01). Together, these data indicate that BDZ, lorazepam and clonazepam effectively reversed stress-induced depressive-like behavior in the test of social avoidance.

2.5 Discussion

BDZs are the most widely used drugs in anxiety disorders and treatment of stress symptoms, and their biochemical effects have been used as a tool to elucidate mechanisms and behavioral outcomes in response to stress (Ferrarese et al., 1992).

Therefore, the aim of the present study was to evaluate the effects of BDZs on the neuroimmune system and behavior during psychosocial stress. These data show that lorazepam and clonazepam potently ameliorated neuroimmune responses in mice subjected to RSD, and blocked stress-related anxiety and depressive-like behaviors.

It has been suggested that clinical actions of BDZs are mediated by their inhibitory influence on the HPA axis, mediated by CRH suppression, reduction of ACTH, and corticosterone release (Arvat et al., 2002). Hyperactivity of the HPA axis has been accepted as a biological mechanism underlying mood disorders and depression (Marques 46

et al., 2009). HPA axis hyperactivity may be a consequence of a decreased feedback inhibition of GC signaling, thus increased CRH secretion from the paraventricular nucleus (PVN) in the HYPO and extra-hypothalamic neurons (Marques et al., 2009).

Elevated numbers of neurons expressing CRH and enhanced mRNA levels of CRH receptors in the PVN in depressed patients at time of autopsy reveal an increase in CRH cell activity premortem (Gadek-Michalska et al., 2013). We found lorazepam and clonazepam attenuated stress-induced mRNA expression of CRH in the hypothalamus, without affecting basal levels in HCC. Classic literature reported direct inhibition of BDZ on CRH secretion from rat hypothalamic cultures (Calogeero et al., 1988; De Souza,

1990). Prenatal administration of diazepam in rats attenuated the reduction in hypothalamic norepinephrine induced by restraint stress, and at the same time, inhibited corticosterone response to stress (Keim and Sigg, 1977) and attenuated stress-induced changes in plasma corticosterone in a dose-related manner (Simmons et al., 1984). In agreement with these studies, we found both BDZs to attenuate stress-induced levels of corticosterone in plasma of mice subjected to RSD.

Studies in humans (Kiecolt-Glaser and Glaser, 2002) and animals (Hanke et al.,

2012) confirm that psychosocial stress increases peripheral cytokine production, and patients with major depressive disorder and post-traumatic stress-disorder have higher levels of cytokine IL-6 (Cole et al., 2010; Baune et al., 2012). In the present study, we found higher levels of IL-6 in plasma of animals subjected to RSD and treated with vehicle, nonetheless this increase was blunted in mice subjected to RSD and treated with either lorazepam or clonazepam. This decrease of IL-6 in plasma was accompanied with 47

dampened neuroinflammatory signaling in the latter experimental group. This indicates that central neural pathways may trigger secretion of IL-6 into the circulation in response to psychosocial stress. However, we cannot rule out the source of IL-6 may be derived from other peripheral tissues (Jankord et al., 2010) in mice subjected to RSD and treated with vehicle. In agreement with another study using repeated social defeat stress (Hodes et al., 2014), this decrease in plasma IL-6 was paralleled by a reversal in social avoidance behavior in mice subjected to chronic social stress, which confirms the idea that IL-6 levels contribute to stress-induced behavioral alterations.

Microglia are also implicated in producing pro-inflammatory mediators and have a causal role in behavioral responses to stress (Wohleb et al, 2011; 2013). As stated in

Chapter 1, it is thought that activation of microglia triggers release of IL-1 within the

HYPO (Goshen and Yirmiya, 2009). There is strong evidence that noradrenaline and IL-

1 in the brain play pivotal roles in the regulation of stress-induced IL-6 in plasma

(Ishikawa et al., 2001). Consistent with this idea, we previously showed that -adrenergic receptor antagonism with propranolol (a non-selective -adrenergic receptor antagonist) prevented, IL-6 in serum, anxiety-like behavior and microglia neuroinflammatory signaling in mice subjected to RSD. Additionally, IL-1r1_/_ receptor knockout mice did not show anxiety-like behavior or microglia activation after being subjected to RSD

(Wohleb et al., 2011). Similarly, we prevented anxiety-like behavior and microglia neuroinflammatory signaling in mice subjected to RSD and treated with lorazepam and clonazepam. TSPO has been detected in microglia (Park et al., 1996) and macrophages

(Zavala et al., 1984), thus in addition to the anxiolytic activity of BDZs; these drugs have been reported to be anti-neuroinflammatory agents (Wilms, 2003) agent.

RSD is associated with increased myelopoiesis in the BM and an accumulation of monocytes and granulocytes in circulation and in the spleen (Engler et al., 2004). It is hypothesized that enhanced myelopoiesis observed in the BM is due, at least in part, to an increase in the production of GM-CSF (Engler et al., 2005). We previously reported that neutralizing GM-CSF with an antibody, or pre-treatment with propranolol before each stress cycle, blocked the RSD-induced increase in monocytic and granulocytic MPCs

(Powell et al., 2013). In the same manner, lorazepam inhibited stress-induced increase in monocytes and granulocytes in the BM, and prevented stress-induced decrease in erythrocytes, while attenuating stress-related decrease in lymphocytes in the BM.

Clonazepam only attenuated stress-induced decrease in erythrocytes and had no effect on lymphocytic, granulocytic or monocytic progenitors. We cannot rule out that a higher dose of clonazepam could have had the effects we saw on the BM with lorazepam treatment. Since the BM is a variable and dynamic microenvironment we cannot conclude exactly how these drugs affected myelopoiesis. However, it is plausible that lorazepam may be acting on myelopoiesis enhancement provoked by stress, via the action at the TSPO. Clear labeling of the BM using whole-body autoradiographic activity, showed high density of this receptor in this organ (Anholt et al., 1985). Evidence has accumulated that BDZs that target the TSPO can interact with phagocytic cells, and affect monocyte-macrophage functions (Taupin et al., 1991).

Like previously reported by our group and others (Powell et al., 2013; Heidt et al.,

2014) repeated stress resulted in a significant accumulation of monocytes and granulocytes in circulation. It is hypothesized that enhanced production and release of monocytes and granulocytes in the BM and subsequently their substantial accumulation in circulation was dependent upon SNS activity. In accordance with this idea, we showed that propranolol treatment in mice subjected to RSD prevented enhanced myelopoiesis and transient increase of these cells in the bloodstream (Hanke et al., 2012; Powell et al.,

2013). Likewise, BDZs, lorazepam and clonazepam inhibited stress-induced increase in monocytes and granulocytes in circulation. BDZs have been shown to be potent stimulators of human monocytes in vitro, and it is thought this effect is mediated through the TSPO (Ruff et al., 1985). However, it has also been shown benzodiazepine-induced chemotaxis is impaired in patients with generalized anxiety disorder (Sacerdote et al.,

1999).

A hallmark characteristic of the RSD response is splenomegaly corresponding with an increased accumulation of CD11b+ monocytes and granulocytes (Avitsur et al.,

2001). Our hypothesis is that repeated activation of the neuroendocrine systems causes the mobilization of monocytes and granulocytes from the BM to the spleen (Reader et al.,

2015). However, as already stated in the present study, we show a decrease in the percentage of monocytes and granulocytes in mice subjected to RSD and treated with

BDZs in circulation. Nonetheless, we found that lorazepam attenuated the stress-induced increase percentage of monocytes and granulocytes in the spleen, and at the same time prevented splenomegaly. It has been reported that diazepam, a benzodiazepine that 50

targets both CBR and TSPO like lorazepam, prevents changes in the subpopulation composition of the spleen, by inhibiting social defeat induced increased percentage proliferation index of splenocytes, and decreasing the percentage of S and G2-M cells in this organ (Kudryavtseva et al., 2011). In this study, clonazepam had no effect on stress- induced increase percentage of monocytes and granulocytes, while only attenuating splenomegaly. This suggests that the increase of monocytes and granulocytes from the latter experimental group might be cells produced in the spleen per se, since myelopoietic activity in the adult spleen can be stimulated by various challenges (Ohno et al., 1993).

Previous studies reported that murine splenocytes exhibit abundant TSPO, particularly the macrophage and T-cell populations (Zavala et al., 1984). Also, it was shown that BDZs that target the TSPO but not the CBR, prevented splenocytes from producing IL-1, TNF-, and IL-6 in vivo, after an immune challenge (Zavala et al.,

1990). Similarly, we found lorazepam but not clonazepam inhibited stress-induced increased mRNA expression of the mentioned pro-inflammatory cytokines in stressed mice. It is noteworthy to mention that lorazepam decreased baseline levels of IL-6 and

TNF- in HCC. Diazepam has been found to be a potent inhibitor of phosphodiesterase type 4 (PDE4) (Cherry et al., 2001). Inhibitors of PDE4 increase intracellular levels of cyclic AMP, which can reduce the activation of a wide range of pro-inflammatory responses, including the production of pro-inflammatory cytokines (Spina, 2008). Again, we cannot rule out that a higher dose of clonazepam could have inhibited splenomegaly and RSD-induced pro-inflammatory phenotype in the spleen.

It is widely accepted that the TSPO is upregulated in activated microglia and macrophages, and is thus a biomarker of neuroinflammation (Fujimura et al., 2009; Choi et al., 2011; Liu et al., 2014). Via interaction with the TSPO, BDZs, such as clonazepam, midazolam, and diazepam have been shown to inhibit microglial activation and to reduce inflammation in vitro (Wilms, 2003). We found that lorazepam and clonazepam prevented stress-induced accumulation of macrophages in the CNS. Equivalently, we found both drugs were effective in decreasing stress-induced elevated mRNA expression of pro-inflammatory cytokines and activation markers in isolated microglia. Furthermore, we found dampened neuroinflammatory signaling in brain slices dissected from the hippocampus/hypothalamus of socially defeated mice treated with both BDZs, these regions are stress-responsive areas of the brain (Wohleb et al., 2014).

As stated in Chapter 1, the interpretation of stress involves neuronal activation within brain areas associated with fear and threat appraisal, and these signals are relayed to the immune system via the HPA axis and SNS neuroendocrine pathways (Wohleb et al., 2014). It is thought neuronal activation precedes microglial activation, since cytokine expression was seen after 3 cycles of RSD (Wohleb et al., 2014). Moreover, we had demonstrated, pre-treatment with propranolol blocked stress-induced neuronal activation as well as microglia morphology (Wohleb et al., 2011). Other studies indicate ablation of noradrenergic locus coereleus (LC) projections diminished stress-induced IL-1 production in the HPC (Johnson et al., 2005). In concert, BDZs have been reported to inhibit cell firing and noradrenergic activation in the LC, which provides to CRH neurons central noradrenergic inputs, while CBRs have been found on presynaptic axons and 52

nerve terminals of noradrenergic neurons localized in the cerebral cortex and HPC (Arvat et al., 2002). Since GABA is the main inhibitory neurotransmitter in the CNS

(Trincavelli et al., 2012), it is possible that by enhancing GABAergic neurotransmission with BDZS, stress-induced pathways were alleviated, down regulating the activity of the

HPA axis and SNS, thus the neuroimmunomodulatory effects we found in the present study.

Depression and anxiety are multifactorial and complex disorders with social, psychological, and biological components. The RSD model recapitulates stress-induced neuroendocrine abnormalities with associated mood alterations. It is clear that the HPA axis and immune system are profoundly altered with stress exposure, and all of these systems are interconnected affecting one another. These interconnections play an important role in the pathogenesis of these behavioral alterations. Chronic activation of these systems leads to neurotoxic alterations in the CNS that may lead to depression. In the present study, we identified that BDZs could prevent these neuroendocrine alterations after RSD. We provided evidence that BDZs act on different interconnected systems by exerting anti-inflammatory actions in the CNS, apparently attenuating HPA axis and SNS products triggered by stress. Possibly, these systems influence each other in a synergistic manner. Further research will help clarify the complex mechanisms underlying depression and anxiety and the actions of BDZs upon all neuroendocrine systems.

Figure 1- Effect of BDZs on stress-induced increase of CRH, corticosterone, and norepinephrine

(A) Experimental schematic. (B,C) RSD mice had increased CRH in the HYPO, and this was attenuated with lorazepam and clonazepam. RSD animals treated with (D) lorazepam or (E) clonazepam had increased levels of corticosterone than HCC treated with vehicle, lorazepam or clonazepam, but lower than RSD animals treated with vehicle. (F) Lorazepam decreased levels of norepinephrine in plasma, while (G) clonazepam attenuated levels of norepinephrine in plasma at this time point. Bars represent the meanSEM. Means with different letters (a,b,c) are significantly different from each other (p<0.05). 54

Figure 2- Effect of BDZs on stress-driven increase of plasma IL-6

Levels of IL-6 were assessed in plasma of mice in the different experimental groups the next day after the last cycle of RSD. Mice subjected to RSD and treated with vehicle had increased levels of IL-6 in plasma. Levels of IL-6 in plasma of mice subjected to RSD and treated with BDZs (A) lorazepam and (B) clonazepam had lower levels of this cytokine, comparable to HCC. Bars represent the meanSEM. Means with different letters (a,b) are significantly different from each other (p<0.05). Control=Home cage controls; RSD=Repeated Social Defeat

Figure 3- Effect of BDZs on stress-induced shift of hematopoiesis

(A and B) Representative bivariate dot plots of BM staining. (A and C) Lorazepam reversed stress- induced decrease of erythrocyte progenitors and attenuated stress-induced decrease of lymphocytic progenitors. Stress-driven increase of granulocytic progenitors was reduced with lorazepam. Monocyte progenitors in both stress groups were higher than HCC. (B and D) Clonazepam attenuated erythrocytic progenitors. This drug had no effect on stress-driven decrease in lymphocytic progenitors or stress-induced increase in granulocytic and monocytic progenitors. Bars represent the meanSEM. Means with different letters (a,b,c) are significantly different from each other (p<0.05). 56

Figure 4 - Effect of BDZs on stress-induced increased trafficking of monocytes and granulocytes in circulation

Blood was collected by cardiac puncture A and B are representative flow bivariate dot plots of Ly- 6C/CD11b labeling in the blood showing that RSD mice treated with vehicle had elevated percentage of blood monocytes (M) (CD11b+/Ly6Chigh) and blood granulocytes (G) (CD11b+/Ly6Cintermediate) in circulation. (C and D) Lorazepam and (E and F) clonazepam decreased stress-induced increase in circulating monocytes and granulocytes in mice subjected to RSD. Bars represent the meanSEM. Means with different letters (a,b) are significantly different from each other (p<0.05). 57

Figure 5- Effect of BDZs on the spleen

(A and B) RSD mice treated with vehicle had elevated percentage of monocytes (M) and (G) granulocytes in the spleen. (C and D) Lorazepam blocked stress-induced increase of monocytes and granulocytes. Mice subjected to RSD and treated with (E and F) clonazepam still displayed a high percentage of these cells when compared to HCC. (G) Lorazepam inhibited splenomegaly. (H) RSD mice treated with clonazepam still displayed a high spleen weight than HCC. Bars represent the meanSEM. Means with different letters (a,b,c) are significantly different from each other (p<0.05). 58

Figure 6 – Effect of BDZs on monocyte trafficking to the CNS

(A and B) Representative bivariate dot plots of the different experimental groups with brain CD11b/CD45 staining, showing that (C) lorazepam blocked stress-induced increased percentage of macrophages in the brain. (D) Likewise, clonazepam reversed stress-induced increased percentage of macrophages in the CNS. RSD mice treated with vehicle had increased percentage of (CD11b+/CD45high) macrophages in the CNS. Bars represent the meanSEM. Means with different letters (a,b) are significantly different from each other (p<0.05).

Figure 7- BDZs blocked stress-associated anxiety-like behavior

Anxiety-like behavior was evaluated the next day after the last cycle of RSD in the open field test. Mice subjected to RSD and treated with vehicle, had significantly less number of center entries than HCC and RSD mice treated with (A) lorazepam and (D) clonazepam. Mice treated with vehicle and subjected to RSD had significant increased time to enter the center of the open field and this was reversed with (B) lorazepam and (E) clonazepam treatment. A decrease of total time spent in the open field was assessed in RSD mice treated with vehicle and this was prevented with (C) lorazepam and (F) clonazepam treatment in mice subjected to RSD. Bars represent the meanSEM. Means with asterisk (*) are significantly different from each other (p<0.05). #= number; (s)=seconds 60

Figure 8- BDZs prevented stress-related depressive-like behavior

The morning after the last cycle of RSD, depressive-like behavior was evaluated using the social avoidance test. Social withdrawal was exacerbated when an aggressor was introduced in the social target trial (Trial 2) in RSD mice treated with vehicle, with a decreased time spent in the interaction zone, and this was reversed with (A) lorazepam and (B) clonazepam treatment. Bars represent the mean  SEM. Means with asterisk (*) are significantly different from each other (p<0.05). Trial 1: time without intruder; Trial 2= time with intruder 61

Table 1- Effect of BDZs on stress-induced increase of pro-inflammatory cytokines in total splenocytes

Total splenocytes of RSD animals treated with vehicle had increased mRNA expression of IL1-, IL-6, and TNF-. (A) Lorazepam was effective in decreasing mRNA expression of these pro-inflammatory cytokines in mice subjected to RSD, while (B) clonazepam treated mice subjected to RSD had increased mRNA expression of IL-1 when compared to HCC. Relative gene expression of IL-6 and TNF- in RSD mice treated with clonazepam was not significantly different to HCC and RSD mice treated with vehicle. Means with different letters (a,b,c) are significantly different from each other (p<0.05). 62

Table 2- Effect of BDZs on neuroinflammatory signaling in microglia

Microglia were collected by Percoll gradient separation. Increased neuroinflammatory signaling was observed in RSD mice treated with vehicle and ameliorated in RSD mice treated with lorazepam and clonazepam. RSD markedly increased relative gene expression of pro-inflammatory cytokines IL-1, IL-6, TNF-, and activation markers TLR-4, CCL2 and CX3CR1. (A) Lorazepam significantly blocked mRNA expression of these pro-inflammatory molecules in microglia of mice subjected to RSD. In the same manner, (B) clonazepam dampened mRNA expression of these pro-inflammatory molecules in isolated microglia. Means with different letters (a,b) are significantly different from each other (p<0.05). 63

Table 3- Effect of BDZs on neuroinflammatory signaling in hypothalamus/hippocampus coronal section

The next day after the last cycle of RSD, brains were collected and a 1mm coronal section of the hippocampus/hypothalamus was dissected to assess mRNA expression of pro-inflammatory cytokines IL- 1, IL-6, and TNF-. (A) Lorazepam and (B) clonazepam ameliorated neuroinflammatory signaling. Both benzodiazepines blunted stress-induced expression of these pro-inflammatory cytokines in the brain slice evaluated. Means with different letters (a,b,c) are significantly different from each other (p<0.05). 64

Chapter 3: Imipramine attenuates neuroinflammatory signaling and reverses stress-

induced anxiety- and depressive- like behaviors

3.1 Introduction

It is apparent that cytokines play a major role in the development of anxiety and depression. Thus, if pro-inflammatory molecules play a causative role in anxiety and depression, then antidepressants should down-regulate these molecules or in part, interfere with their actions. Indeed, there is accumulating evidence that indicates that antidepressants have the capacity to modulate cytokine production. For example, it has been shown antidepressants can modify the expression of central GC receptors; resulting in re-establishment of the GC-mediated negative feedback (Castanon et al., 2002) that may be disrupted by stress. As stated previously in Chapter 1, GCs alter cytokine expression. Antidepressants can also stimulate adaptive changes in the central monoaminergic circuitry, which can modulate immune reactivity and thus, the central actions of cytokines. Chronic treatment with tricyclic antidepressant clomipramine abolished LPS-stimulated increase of serotonin concentration in the HYPO

(Yirmiya et al., 2001). Similarly, the HPA axis activation provoked by systemic LPS administration, and concomitant depressive-like behaviors, such as reduced social exploration, locomotor activity, and food consumption were attenuated with treatment of tricyclic antidepressant, tianeptine (Castanon et al., 2002).

More research is needed to establish the exact influence of antidepressants on stress-related cathecolaminergic mechanisms, specifically in the context of HPA and SNS activation, and cytokine production in both the periphery and brain. Therefore, the overarching objective of this study was to determine the effect of imipramine, one of the most commonly prescribed tricyclic antidepressants, on neuroimmune alterations caused by stress in both the periphery and CNS, and associated stress-related behaviors.

Aim 1: To test the hypothesis that imipramine attenuates stress-induced inflammation in the periphery.

The first objective was to assess if imipramine ameliorated stress-induced increase in corticosterone, IL-6, and splenomegaly, three hallmarks of RSD, after six cycles of social stress.

The second objective was to determine if imipramine prevented RSD-induced shift in myelopoiesis, and decreased the trafficking of inflammatory MPCs into circulation and spleen, after six cycles of RSD.

The third objective was to evaluate if imipramine decreased the stress-induced increased expression of pro-inflammatory cytokines in the spleen, after six cycles of RSD and at 24 days after stress cessation.

Aim 2: To test the hypothesis that imipramine prevented stress-induced neuroinflammatory signaling and reversed stress-associated behaviors.

The first objective was to determine if imipramine prevented the trafficking of monocytes from the periphery to the brain, after six cycles of RSD.

The second objective was to determine if imipramine ameliorated neuroinflammatory signaling in microglia induced by social defeat, after six days of RSD and at 24 days after stress termination.

The third objective was to evaluate if imipramine rescued anxiety-like behavior, after six cycles of RSD and prevented stress-associated depressive-like behavior at this time point, and at 24 days after stress termination.

3.2 Methods

3.2.1 Animals

Male C57BL/6 (6–8 weeks old) and CD-1 (12 months, retired breeders) mice were obtained from Charles River Breeding Laboratories (Wilmington, Massachusetts) and allowed to acclimate to their surroundings for 7–10 days before initiation of experimental procedures. C57BL/6 mice were housed in cohorts of three and CD-1 mice were singly housed and maintained at 21°C under a 12:12 h light: dark cycle with ad libitum access to water and rodent chow in the animal facility at The Ohio State

University. All procedures were in accordance with the NIH Guidelines for the Care and

Use of Laboratory Animals and were approved by the Ohio State University Institutional

Laboratory Animal Care and Use Committee.

3.2.2 RSD

RSD was performed as described previously, see Chapter 2, section 2.2.2 and see

Avitsur et al., 2001; Stark et al., 2001; Wohleb et al., 2011; Hanke et al., 2012.

3.2.3 Pharmacological treatments and administration procedures

When imipramine was used to evaluate the effects after the last cycle of RSD,

C57BL/6 mice were randomly selected for inclusion into different experimental treatment groups. The groups were: RSD/imipramine, RSD/vehicle, HCC/imipramine, and

HCC/vehicle. Figure 9A shows the experimental schematic. Mice in the RSD/imipramine received daily imipramine (15mg/kg) treatment in their drinking water, starting two days before the initiation of social defeat. HCC/imipramine received daily imipramine treatment in their drinking water at the same dose while RSD/vehicle and HCC/vehicle groups drank untreated water. The dose of imipramine was based on a previous study with C57BL/6 mice, in which chronic administration at 15mg/kg in drinking water, effectively increased sucrose and water intake, as well as enhanced home-cage and novelty exploration activities in naïve animals (Strekalova et al., 2013). The amount of water consumed for each cage was registered daily throughout the experiment. The calculation of the concentration of imipramine in drinking water was based on the evaluated mean volume of daily water consumption, assessed by weighing the bottles daily (from Day 0 to Day 8). Bottles of water were changed daily before and during the experimental protocol. An average of 9.0 ml per day/per cage intake was calculated.

Water consumption in the four groups of mice was in fact similar. Based on this average, the desirable dosage of treatment (15 mg/kg/day) was established.

The experiments were performed in different replicates of mice. Each replicate of mice consisted of four different experimental groups. Each consisted of three mice 69

receiving a different treatment. The experiments were done two times to confirm the findings in different replicates. For example the social avoidance test, and the collection of bone marrow, spleen, brain (and brain slice) and blood for flow cytometry was done in two replicates. In a second experiment collection of blood to assess corticosterone, anxiety-like behavior, plasma IL-6, and isolation of microglia and total splenocytes was done in another replicate, and confirmed in a second replicate.

In the study where imipramine was administered to evaluate the effects 24 days after stress cessation, C57BL/6 mice were randomly selected for inclusion into different experimental treatment groups. The groups were: RSD/imipramine, RSD/vehicle,

HCC/imipramine, and HCC/vehicle. Mice in the RSD/imipramine received daily intraperitoneal (i.p.) injections of imipramine (20 mg/kg) for 24 days after the 6 cycles of

RSD. HCC/imipramine received daily i.p. imipramine at the same dose while

RSD/vehicle and HCC/vehicle groups received i.p. injections of vehicle (sodium chloride, 0.9%) for 24 days at the same time point (Figure 15A). This dose and timing was chosen since previous studies had shown that chronic imipramine treatment after social defeat, at this concentration, reversed social avoidance behavior in C57BL/6 mice

(Berton et al., 2006; Tsankova et al., 2006). Additionally, we had previously shown social avoidance behavior was still present 24 days after RSD (Wohleb et al., 2013).

Therefore, based on these studies, we decided to treat the animals for 24 days after the last cycle of RSD to assess if there was a reversal of social avoidance. The day after the last injection of imipramine the interaction and avoidance toward an unfamiliar CD-1 mouse was measured. 70

The experiments were performed in different replicates of mice. Each replicate of mice consisted of four different experimental groups. Each consisted of three mice receiving a different treatment. The experiments were done two times to confirm the findings in different replicates. In a first experiment, the social avoidance test was performed, brains were collected to isolate microglia, and total splenocytes for PCR. This was done in two replicates. In a second experiment, social avoidance was determined and isolation of microglia was performed to stimulate these cells with LPS.

The experiment at 24 days after stress termination was repeated for social avoidance behavior and detection of pro-inflammatory cytokines in microglia, and modified by giving experimental mice imipramine (15mg/kg) in their drinking water.

Experimental schematic is shown in Figure 17A. C57BL/6 mice were randomly selected for inclusion into the four different experimental treatment groups as described in the paragraph above. Mice in the RSD/imipramine group were treated with imipramine in the drinking water for 24 days after the 6 cycles of RSD. HCC/imipramine were also given the same treatment at the same dose at this time point. Animals in the RSD/vehicle and

HCC/vehicle groups drank untreated water throughout the experiment. The amount of water consumed for each cage was registered daily throughout the experiment (from Day

0 to Day 35) (n=8-9 per group).

The experiments were performed in two different replicates of mice. Each replicate of mice consisted of four different experimental groups. Each consisted of three mice receiving a different treatment. The social avoidance test was performed and brains were collected to isolate microglia. 71

3.2.4 Anxiety-like behavior testing and social avoidance test

For the study assessing the effect of imipramine after six cycles of RSD, anxiety- like behavior was determined the morning following the last cycle of RSD as previously described in Chapter 2, and see Wohleb et al., 2011 (n=6 per group). Additionally, the social avoidance test was performed as previously described in Chapter 2 and also see

Wohleb et al., 2014 (n=6 per group).

Anxiety-like behavior was not evaluated 24 days after stress cessation, since previous studies had already shown that it was not present at this time point (Wohleb et al., 2014), however social avoidance behavior was determined at this time point as previously described in Chapter 2 and see Wohleb et al., 2013 (n=12-15 per group).

3.2.5 Plasma corticosterone determination

Blood was collected immediately after the last cycle of RSD and processed to get plasma as described in Chapter 2 assayed in duplicate using a Corticosterone ELISA kit from Enzo Life Sciences (Farmingdale, New York) following manufacturer’s instructions

(n=6 per group).

3.2.6 Plasma IL-6 determination

Blood was collected the day after RSD and processed as previously described in

Chapter 2 (n=5-6 per group).

3.2.7 Microglia isolation and flow cytometry

Microglia was isolated from whole-brain homogenates as previously described in

Chapter 2 and also refer to Henry et al., 2009; Wynne et al., 2010; Wohleb et al., 2011, the next day after the last cycle of RSD. Flow cytometry was performed in the study evaluating the effects of imipramine after six cycles of RSD as described previously in

Chapter 2 and also refer to Henry et al., 2009 and Wohleb et al., 2011. Flow cytometry was not performed at 24 days after stress cessation since previous data generated from our laboratory found that the increased accumulation of macrophages in the brain triggered by RSD returned to baseline at this time point (Wohleb et al., 2013) (n=6 per group).

3.2.8 Microglia RNA isolation and real-time PCR

RNA from Percoll-isolated microglia was isolated as described in Chapter 2 using the RNeasy plus mini-kit (Qiagen) and quantitative PCR was done using the Applied

Biosystems by Life Technologies Assay-on-Demand Gene Expression protocol as 73

previously described in Chapter 2 and refer to Godbout et al., 2005; Wohleb et al., 2011.

Amplification of cDNA was performed by real-time PCR. A target cDNA (IL-1 β, IL-6,

TNF-α, TLR4, CCL2, and CX3CR1 for the study evaluating the effect of imipramine after six cycles of RSD; IL-1 β, IL-6, TNF-α for the study evaluating the effect of imipramine 24 days after stress cessation) and a reference cDNA (glyceraldehyde-3- phosphate dehydrogenase) were amplified at the same time using an oligonucleotide probe with fluorescent reporter dye (FAM). An ABI PRISM 7300-sequence detection system (Applied Biosystems by Life Technologies, Grand Island, New York) was used to assess fluorescence. Data were analyzed by the comparative threshold cycle and the results are given as the fold difference detected (n=5-6 per group).

3.2.9 Total splenocytes RNA isolation and real-time PCR

Spleens were collected, weighed, and flash frozen in liquid nitrogen. Samples were processed as previously described see Chapter 2 (n=5-6 per group).

3.2.10 Flow cytometry of bone marrow, blood, and spleen

poured off and the pellet of cells was brought up in 1ml FACS buffer and vortexed. Fifty

µl of each sample was added into a flow tube. Fc receptors of bone marrow cells and blood leukocytes were blocked with anti-CD16/CD32 antibody (eBioscience, San Diego,

California) as previously described (Wohleb et al., 2013). Cells were washed and then incubated with the appropriate antibodies (CD45, CD11b, CD34; eBioscience, San

California).

2 min, followed by one wash with HBSS/10% heat inactivated fetal bovine serum (FBS).

Each cell pellet was resuspended in HBSS, filtered and washed a final time in HBSS.

Cells were counted and samples were resuspended (2.5 x 106 cells/ml) in supplemented

RPMI medium (10% heat-inactivated FBS, 0.075% sodium bicarbonate, 10 mM HEPES buffer, 100 U/ml penicillin G, 100 mg/ml streptomycin sulfate, 1.5 mM L-glutamine, and

for 45 min. The cells were then washed twice in PBS containing 1% FBS and 0.09%

FACSCalibur flow cytometer using Cell Quest and Cell Quest Pro analysis software

(Becton-Dickenson; San Jose, California) (n=6 per group).

3.2.11 Ex-vivo microglia and cytokine measurement by real-time PCR and ELISA

Microglia were isolated 24 days after stress termination, and identified as such cells using flow cytometry, and then stimulated ex vivo with LPS as described previously

(Wohleb et al., 2011). In brief, microglia isolated by Percoll gradient separation were counted and plated on poly-L-lysine-coated 96-well plates. Cells were placed in complete

RPMI containing 10% heat inactivated fetal bovine serum, 0.075% sodium bicarbonate,

10 mM HEPES buffer, 100U/ml penicillin G, 100g/ml streptomycin sulfate, 1.5 nM L- glutamine, and 0.0035% 2-mercaptoethanol. Cells were then stimulated with 400 nanograms/ml LPS (Sigma-Aldrich, St. Louis, MO), for 18 h and incubated at 37°C in

5% CO2 and controls were left unstimulated at same time point and temperature. After cell harvesting, the gene expression of pro-inflammatory cytokines (IL-6, IL-1, and

TNF-) was measured using real time-PCR as described above. Supernatants were

collected and the concentration of IL-1 was detected using Quantikine ELISA Mouse

IL-1/IL-1F2 (R&D SystemsTM, Minneapolis, MN) (n=4-6 per group).

3.3 Statistical Analysis

The data were subjected to Shapiro-Wilk tests to ensure a normal distribution using SPSS Statistics version 21. Data are expressed as means ± SEM. In order to determine significant main effects and interactions between variables, data were subjected to one-way (trial, stress, imipramine), two-way (trial x stress); (trial x imipramine); (stress x imipramine) or three-way (trial x stress x imipramine); (stress x imipramine x LPS) analysis of variances (ANOVAs). When appropriate, differences between treatment group means were analyzed by an F-protected t test using the least- significant difference method. In all cases, the level of significance was set at p  .05.

3.4 Results

3.4.1 Imipramine attenuated RSD-induced corticosterone in plasma and blocked IL-6 in plasma after six cycles of social defeat

Previous studies showed that mice had enhanced levels of serum corticosterone immediately after the last cycle of RSD compared to HCC mice (Avitsur et al., 2001;

Hanke et al., 2012). Additionally, RSD consistently increased plasma levels of IL-6 77

(Stark et al., 2002). Hence, corticosterone and IL-6 responses were used as biomarkers to gauge the robustness of the response to social defeat. Therefore, our first objective was to determine the effect of imipramine on stress-induced increased levels of plasma corticosterone, and the second objective the effect of this drug in levels of IL-6 in plasma.

Figure 9B shows that immediately after the last cycle of RSD, increased levels of corticosterone were found in plasma of mice subjected to RSD (main effect of stress; F

(1,23) = 32.15, p < 0.01; main effect of imipramine; F (1,23) = 6.76, p < 0.05; stress x imipramine interaction; F (1,23) = 1.30, p = 0.26). Post-hoc analyses revealed imipramine attenuated plasma corticosterone in socially defeated animals (p<0.05) compared to socially defeated mice treated with vehicle. However, plasma corticosterone was high in

RSD mice treated with imipramine compared to HCC (p<0.05).

Blood was collected the day after the last cycle of RSD and plasma IL-6 levels were determined. As reported previously, social defeat increased IL-6 levels in plasma

(Figure 9C), and this increase was blocked by imipramine treatment (main effect of stress; F (1,21) = 25.95, p<0.01; stress x imipramine interaction; F (1,21)= 4.46, p <0.05).

Together, these data indicate that imipramine was effective in attenuating stress-related increased corticosterone levels in plasma and blocked levels of circulating IL-6 in plasma of C57BL/6 mice subjected to social defeat.

3.4.2 Effect of Imipramine on stress-driven alteration of bone marrow hematopoiesis

As stated previously in this manuscript, RSD enhanced myelopoiesis and promoted the development, priming, and egress of a MPC population from the BM

(Engler et al., 2004). Thus, the aim of this experiment was to determine the effects of imipramine on stress-induced alterations in hematopoiesis in the BM (Figures 10A and

10B). RSD mice treated with vehicle and imipramine, had a decreased percentage of lymphocyte progenitors in the BM (main effect of stress; F(1,23)= 33.73, p<0.01) compared to HCC. Imipramine prevented stress-induced decrease in erythrocyte progenitors (main effect of stress; F(1,23) = 6.15, p<0.05; stress x imipramine interaction;

F(1,23)=9.61, p<0.01) in the BM. Moreover, imipramine blocked stress-induced increase percentage of granulocytes (main effect of stress; F (1,23)= 19.28, p<0.01; stress x imipramine interaction; F(1,23) =11.32, p<0.01) and blocked stress-induced increase of monocyte progenitors in the BM (main effect of stress; F(1,23)=10.76, p<0.01; stress x imipramine interaction; F(1,23) =9.77, p<0.01). In sum, imipramine partially blocked RSD- induced alteration in hematopoiesis.

3.4.3 Effect of imipramine in stress-induced increased percentage of monocytes and granulocytes in circulation and in the spleen

Previous findings reported increased numbers of monocytes and granulocytes in socially defeated mice, after six cycles of RSD (Hanke et al., 2012; Powell et al., 2013; 79

Wohleb et al., 2013). Figure 11A, shows mice exposed to RSD had increased numbers of

(Figure 11B) monocytes (main effect of stress; F(1,23)= 4.53, p=0.05) and (Figure 11C) granulocytes in circulation (main effect of stress; F(1,23)= 114.73, p<0.01). Post-hoc analyses revealed imipramine decreased stress-induced increase in monocytes (p<0.05) in circulation. Additionally, imipramine also ameliorated granulocyte trafficking in circulation (stress x imipramine interaction; F (1,23)=100.89, p<0.01).

RSD caused an increase in the number and percentage of monocytes and granulocytes in the spleen, which was associated with the development of splenomegaly, after six cycles of social defeat (Avitsur et al., 2003). Thus, total splenocytes were analyzed using flow cytometry to determine if imipramine would affect the percentage of these cells in the spleens of socially defeated mice (Figures 11D). Flow cytometric analysis indicated that stress-driven increase in (11E) monocytes (main effect of stress; F

(1,23) =4.77, p<0.05) was completely blocked by imipramine treatment (stress x imipramine interaction; F(1,23)= 4.38, p<0.05). However this drug did not reverse the stress driven increase in splenic (11F) granulocytes (main effect of stress; F (1,23) = 4.74, p

<0.05; stress x imipramine interaction, F (1,23)= 1.24, p>0.05). Mice subjected to RSD and treated with imipramine had a greater percentage of these cells in the spleen compared to

HCC (p<0.05).

We next aimed to determine if treatment with imipramine would attenuate splenomegaly in socially defeated mice. Imipramine did not attenuate splenomegaly

(main effect of stress; F (1,47) = 8.04, p<0.01; stress x imipramine interaction; F (1,47)=

1.78, p>0.05). Post-hoc analyses indicate spleen weight in RSD mice treated with this drug was still high compared to HCC (p<0.05) as shown in Figure 11G.

The data collected after the last cycle of social defeat suggests that imipramine prevented the release of monocytes and granulocytes from the bone marrow into circulation, in socially defeated mice. Also, imipramine decreased the pool of monocytes in the spleens of socially defeated animals, but was not effective in decreasing the pool of granulocytes in the spleen or splenomegaly associated with RSD.

3.4.4 Imipramine did not decrease stress-induced production of pro-inflammatory cytokines in total splenocytes after six cycles of social defeat

We sought to determine if imipramine would decrease RSD increased pro- inflammatory cytokine production by total splenocytes, specifically, IL-1, IL-6, and

TNF- (Table 4A), after six cycles of social defeat. Imipramine did not block mRNA expression of IL-1 (main effect of stress; F(1,20)= 5.12, p <0.05; stress x imipramine interaction; F(1,20)= 3.09, p=0.09), IL-6 (main effect of stress; F(1,21)=7.76, p=0.01; stress x imipramine interaction; F(1,21)=3.54, p=0.07), nor TNF- in total splenocytes (main effect of stress; F(1,21) = 7.36, p=0.01; stress x imipramine interaction; F(1,21) = 0.69, p>0.05). In sum, the production of pro-inflammatory cytokines in total splenocytes was enhanced after six cycles of RSD and imipramine treatment was not effective in preventing increased mRNA levels of IL1-, IL-6, and TNF- in socially defeated mice.

3.4.5 Imipramine prevented stress-induced trafficking of bone marrow-derived CD11b+ myeloid cells to the brain

As previously stated in Chapter 1, RSD altered the phenotype of myeloid-derived

CD11b+ cells in the periphery (Bailey et al., 2009) and increased their propensity to traffic from the bone marrow to the brain (Wohleb et al., 2011). In line with this idea, we next sought to determine the effect of imipramine on the trafficking of circulating MPCs to the CNS in socially defeated mice the next day after the last cycle of RSD. Figure 12A shows representative bivariate dot plots of CD11b and CD45 staining for macrophages

(CD11b+/CD45high) (Nair et al., 2007) in the brain for the various experimental groups.

The percentage of macrophages after RSD was increased, however, this was ablated by imipramine (main effect of stress; F(1,23)= 10.13, p<0.01; stress x imipramine interaction;

F(1,23)= 6.70, p=0.01) as seen in Figure 12B. These data suggest that imipramine was effective in preventing RSD-induced accumulation of macrophages in the CNS.

3.4.6 Imipramine prevented stress-induced neuroinflammatory signaling

To determine if RSD-driven neuroinflammatory signaling was attenuated after the last cycle of RSD with imipramine treatment, gene expression of pro-inflammatory cytokines and activation markers were examined in microglia (Table 5A). RSD mice treated with vehicle had increased relative gene expression of IL1- (main effect stress; F

(1,21)= 31.15, p< 0.01), IL-6 (main effect stress; F(1,21)= 9.57, p< 0.01), TNF- (main 82

effect stress; F(1,21)= 13.54, p< 0.01), and activation markers TLR-4, CCL2, and CX3CR1

(main effect stress; F(1,20)= 7.76, p=0.01; main effect stress; F(1,21)= 11.68, p< 0.01; main effect stress; F(1,20)= 9.99, p< 0.01; respectively) and this effect was prevented with imipramine treatment (stress x imipramine interaction; F(1,21) = 8.89, p<0.01; stress x imipramine interaction; F(1,21) = 6.28, p<0.05; stress x imipramine interaction; F(1,21) =

5.22, p<0.05; stress x imipramine interaction; F(1,20) = 4.05, p=0.06; stress x imipramine interaction; F(1,21) = 4.97, p<0.05; stress x imipramine interaction; F(1,20) = 4.17, p=0.05; respectively). Post-hoc analyses indicated there were no differences in mRNA expression of these pro-inflammatory molecules evaluated between RSD mice treated with imipramine and HCC (p<0.05 for all molecules assessed). These data suggest that imipramine abated RSD-induced neuroinflammatory signaling by decreasing gene expression of pro-inflammatory molecules in microglia.

3.4.7 Imipramine prevented stress-induced anxiety-like behavior

RSD promoted anxiety-like behavior in the open field test after six cycles of RSD

(Wohleb et al., 2011). Therefore, the aim of this experiment was to evaluate if imipramine treatment of socially defeated mice could reverse anxiety-like behavior. The day after the last cycle of RSD, individual mice were placed in the corner of an open field box and allowed to freely explore the apparatus. Figures 13A show mice subjected to

RSD had a decreased number of center entries compared to HCC mice, and this was prevented with imipramine treatment (main effect of stress; F(1,23)= 4.27, p=0.05; stress x 83

imipramine interaction; F(1,23)=4.28, p<0.05). Figures 13B shows that mice subjected to

RSD and treated with vehicle took more time to first enter the center of the open field compared to HCC treated with vehicle and this was reversed with imipramine (main effect of stress; F (1,23)= 4.54, p<0.05; stress x imipramine interaction, F(1,23)=3.85, p=0.06). Post-hoc analyses revealed mice subjected to RSD and treated with imipramine took less time to enter the center compared to RSD mice treated with vehicle (p<0.05), the time was not significantly different from HCC mice (p>0.05) Likewise, Figure 13C shows that imipramine was effective in increasing center duration in mice treated with imipramine and subjected to RSD (main effect of stress; F(1,23)=4.38, p<0.05; stress x imipramine interaction; F(1,23)= 0.86, p>0.05). Post-hoc analyses indicated this increase in center duration was significantly different from RSD mice treated with vehicle (p<0.05), but not from HCC mice (p>0.05). These data indicated imipramine acted as an anxiolytic in socially defeated mice.

3.4.8 Imipramine reversed social avoidance behavior

In this experiment the aim was to determine if eight-day imipramine treatment during RSD could prevent stress-induced social avoidance behavior. The day after the 6th cycle of RSD, social avoidance was determined using a two-trial interaction paradigm with an empty social target trial (Trial 1) followed by a social target trial (Trial 2). Figure

14A show that mice subjected to RSD spent less time in the interaction zone when an intruder was introduced in (main effect stress; F (1,20)=7.38, p=0.01). Imipramine was 84

effective in reversing stress-induced social avoidance behavior in mice subjected to RSD

(main effect trial; F(1,20) = 12.86; trial x stress interaction; F (1,20) = 31.17, p<0.01; trial x imipramine; F(1,20) = 17.25, p<0.01; trial x stress x imipramine interaction; F (1,20) = 29.28, p<0.01). Together, these data indicate that eight days of imipramine treatment during

RSD effectively reversed stress-induced depressive-like behavior in the test of social avoidance.

3.4.9 Imipramine reversed long-lasting RSD-induced social avoidance behavior, 24 days after stress cessation

Since RSD promoted long-lasting social avoidance behavior maintained for at least 24 days after RSD cessation (Wohleb et al., 2013), our aim was to determine in this experiment if stress-induced social avoidance behavior could be reversed in mice exposed to RSD with imipramine treatment. Mice were subjected to 6 cycles of RSD or left undisturbed (HCC) and subsequently treated chronically with daily i.p. injections of imipramine or vehicle for 24 days. At 24 days after RSD, social avoidance was determined using a two-trial interaction paradigm with an empty social target trial (Trial

1) followed by a social target trial (Trial 2). Figure 15B shows that mice in the control group treated with either vehicle or imipramine increased time spent in the interaction zone (p<0.05 for both) when a social target was introduced in the second trial (main effect trial; F (1, 28) = 14.53, p<0.01; main effect imipramine; F (1, 28) = 0.7, p>0.05; trial x imipramine interaction; F(1, 28) = 0.0006 , p= 0.98). Time spent in the interaction zone was 85

significantly decreased in RSD mice treated with vehicle (p<0.01), when an aggressor was introduced in the social trial (main effect stress; F(1,55) = 9.62, p<0.01; main effect imipramine; F (1,55) = 4.41, p< 0.05; stress x imipramine; F (1,55) = 5.22, p <0.05) and this was reversed with imipramine treatment (trial x stress x imipramine interaction; F (1,55) =

4.90, p=0.03). Figure 15C shows that there were no differences between the control treatment groups in the time spent in corners (p= 0.39), however, during the social trial

RSD mice treated with vehicle spent more time in the corners and less time in the interaction zone (p<0.01) (main effect stress; F (1,55) = 9.62, p<0.01; main effect imipramine; F (1,55) = 4.41, p<0.05; stress x imipramine; F (1,55) = 5.22, p<0.05). Together, these data indicate that imipramine reversed stress-induced social avoidance behavior.

3.4.10 RSD enhanced TNF-, IL-6 and IL1- expression in total splenocytes 24 days after stress cessation

In this experiment we sought to determine if RSD increased pro-inflammatory cytokine production by total spleen cells 24 days after cessation of stress (Table 6). TNF-

 relative gene expression was significantly enhanced in total splenocytes of RSD mice treated with vehicle in comparison to HCC and RSD mice treated with imipramine

(p<0.05); (main effect stress; F(1,23)= 11.81, p<0.01; main effect imipramine; F(1,23) =

8.53, p<0.01). Stress induced the relative gene expression of IL-6 (main effect stress;

F(1,23) = 14.08, p<0.01) and a trending decrease of the expression of this cytokine was observed in RSD mice treated with imipramine (main effect imipramine; F(1,23) = 3.74, p= 86

0.065). Likely, RSD mice had increased levels of IL1- in total spleen cells (effect of stress; F(1,23) = 15.08, p<0.01), and a trending attenuation of expression was detected with imipramine treatment in RSD mice (effect of imipramine; F(1,23) = 3.16, p= 0.088). In sum, RSD promoted the mRNA expression of pro-inflammatory cytokines 24 days after stress cessation in total splenocytes, and imipramine partially attenuated the production of

IL-6 and IL-1 in these cells.

3.4.11 Imipramine ameliorates RSD-induced long lasting over expression of IL-6 in microglia

To elucidate if the RSD-induced neuroinflammatory profile at 24 days after RSD was attenuated by imipramine treatment, gene expression of three major pro- inflammatory cytokines, IL-6, IL1-, and TNF- was examined in microglia (Table 7).

Increased relative gene expression of IL-6 was maintained at 24 days after cessation of

RSD in mice treated with vehicle (main effect stress; F (1,23)= 25.20, p< 0.01 and main effect imipramine; F (1,23)= 18.34, p<0.01). Imipramine attenuated relative gene expression of IL-6 (stress x imipramine interaction; F (1,23)= 4.91, p=0.04). There were no differences in TNF-α and IL-1β relative gene expression at 24 days between the RSD and control groups (p>0.05 for both). These data suggest that imipramine may reduce long-lasting RSD-induced neuroinflammation by decreasing gene expression of IL-6.

3.4.12 Imipramine attenuates RSD-induced reactivity of microglia to LPS stimulation

We had previously reported that microglia isolated from RSD mice and cultured ex vivo produced markedly higher levels of pro-inflammatory cytokines (IL-6 and TNF-

) compared to HCC (Wohleb et al., 2011), after 6 cycles of RSD. Therefore, the objective of the next experiment was to determine if microglia from RSD mice had increased reactivity after LPS stimulation 24 days after RSD cessation, and determine if imipramine treatment reduced microglia response to LPS stimulation. To address this objective, brains were collected at 24 days after RSD. Microglia were collected by

Percoll gradient separation and cultured ex vivo. Microglia were plated in growth media, in a 96-well tissue culture plate at 1 X 105 cells per well. Cells were incubated with LPS and supernatants and cells were collected 18 hours later. Relative gene expression of IL-

, IL-6, and TNF-α was determined from these cells. Stimulation with LPS induced higher gene expression of IL-1, IL-6, and TNF- (Table 8) in microglia from RSD mice treated with vehicle, 24 days after RSD cessation (main effect stress; F (1,41)= 9.2, p<0.01,

F (1,42) = 6.2, p=0.02, F (1,40) = 5.6, p=0.02, respectively), and this over expression was reduced with imipramine treatment (stress x imipramine x LPS interaction; F(1,41)= 4.4,

p=0.04 for IL-1, F (1,42) = 5.58, p=0.02 for IL-6, and F(1,40) = 5.25, p=0.03 for TNF-).

Figure 16 shows increased protein levels of IL1- in supernatants of RSD mice treated with vehicle after LPS stimulation compared to all the other mitogen-stimulated groups

(p<0.01 between each group), meaning that imipramine reversed the exaggerated

response to mitogen-stimulation (stress x imipramine x LPS interaction; F (1,41) = 17.00, p<0.01). Collectively, these data suggests that RSD microglia are more reactive and have an increased response to subsequent inflammatory stimuli, even 24 days after the RSD.

3.4.13 Treatment with imipramine in the drinking water reversed stress-induced social avoidance behavior 24 days after stress cessation

In order to confirm the reversal of RSD-induced social avoidance behavior with imipramine treatment, the drug was diluted in the animal’s drinking water. Animals from the RSD/imipramine group were given imipramine (15mg/kg) daily via drinking water, after 6 cycles of RSD for 24 days. HCC/imipramine group started the same treatment at this time point and dose, while RSD/vehicle and HCC/vehicle groups drank plain water throughout the whole experiment. Social avoidance was determined using the two-trial interaction paradigm with an empty social target trial followed by a social trial. Mice in the control group treated with either vehicle or imipramine showed no differences between the first and second trial in the interaction zone when a social target was introduced in the second trial (main effect trial, F (1,28) =0.25, p= 0.62; trial x imipramine interaction, F (1,28) = 0.13, p= 0.72). Time spent in the interaction zone was significantly decreased in RSD mice treated with vehicle (p<0.01), when an unfamiliar CD-1 mouse was introduced in the social trial (main effect stress; F (1,28) =5.76, p<0.05; stress x imipramine interaction; F (1,28) = 4.69, p<0.05; trial x stress interaction; F (1, 28) = 15.7 p<0.01) and was reversed with imipramine treatment in the drinking water (trial x stress x 89

imipramine interaction; F (1,28) = 10.97, p<0.01) (Figure 17B). There were no differences between the groups in time spent in the corners between trial 1 and trial 2 (main effect stress; F (1,28) = 0.02, p>0.05; main effect imipramine; F (1,28)= 0.27, p>0.05; stress x imipramine interaction; F (1,28) = 0.28, p>0.05 (Figure 17C).

3.4.14 Imipramine in the drinking water reduced RSD-induced IL-6 relative gene expression in microglia

Relative gene expression was determined in animals treated with imipramine (15 mg/kg) in the drinking water (Table 9). RSD markedly increased relative gene expression of IL-6 in microglia 24 days after stress cessation (main effect stress; F (1,32)=10.11, p<0.01), and this was decreased with imipramine treatment diluted in the drinking water

(stress x imipramine interaction; F (1,32) = 12.44, p<0.01). TNF-α relative gene expression was increased in the RSD/vehicle group compared to HCC of either groups (main effect stress; F(1,32) =10.06, p<0.01), nonetheless there was no difference when compared to

RSD/imipramine group (main effect imipramine; F(1,32) = 2.34, p=0.14). An increased relative gene expression level of IL-1 was present in the RSD groups treated with vehicle and imipramine (main effect stress; F (1,32)=11.46, p<0.01, main effect imipramine; F (1,32) = 0.83, p=0.37) compared to HCC.

3.5 Discussion

The present study confirms imipramine treatment in the drinking water (15mg/kg) prevented stress-induced neuroimmune dysregulation and associated behavioral alterations, after six cycles of social defeat. At this time point, imipramine attenuated corticosterone in plasma, a product of HPA activation. Also, imipramine prevented stress-induced increase in plasma IL-6 and the development and trafficking of MPCs from the BM to circulation and brain. Finally, treatment with imipramine, 6 days after social defeat, decreased neuroinflammatory signaling, paralleled by a blockade of co- morbid anxiety- and depressive-like behaviors in mice exposed to RSD. Moreover, the present study provides evidence that chronic treatment with imipramine by i.p. injection

(20 mg/kg) or by oral administration (in drinking water) (15 mg/kg), caused reversal of stress-induced social avoidance behavior after the last cycle of RSD and even 24 days after stress termination.

Imipramine inhibits the reuptake of serotonin, a neurotransmitter that plays a pivotal role in aggressive and socioaffective appraisal cues (Challis et al., 2014). An increase in serotonin, by pharmacological manipulation in several species, promotes socioaffective stimuli, whereas depletion of serotonin negatively shifts social responses to avoidance and aggressive behaviors (Canli and Lesch, 2007; Passamonti et al., 2012).

We had previously shown that 24 days after RSD cessation markers of immune alterations associated with RSD, such as splenomegaly, plasma IL-6, and the number of circulating CD11b+ cells returned to control levels. Moreover, brain macrophage 91

population returned to control levels by 24 days after RSD. IL-1 and TNF- mRNA levels returned to baseline by 24 days, nonetheless IL-6 mRNA level was still elevated at

24 days (Wohleb et al., 2013). In the present follow-up series of experiments, in parallel to the behavior, RT-qPCR analysis showed that imipramine either administered by i.p. injection or orally, significantly reduced stress-induced expression of IL-6. One of the ways imipramine may be attenuating stress-induced neuroinflammatory signaling is the fact that microglia have serotonin receptors. Physiological concentrations of serotonin are required to have a balance; low intracellular levels of serotonin are required for a reduced production of pro-inflammatory cytokines. Moreover, a number of in vivo studies have suggested that antidepressants increase intracellular levels of cAMP in immune cells, through activation of monoamine receptors (Xia et al., 1996; Edgar et al., 1999). As a result, pharmacological enhancement of cAMP dampens the expression of pro- inflammatory cytokines (Benbernou et al., 1997; Brideau et al., 1999; Eigler et al., 1998;

Platzer et al., 1999) by inhibiting the protein kinase A (PKA) pathway (Xia et al., 1996;

Hashioka et al., 2007).

In agreement with another study (Yirmiya et al., 2001), imipramine blunts inflammatory activation of microglia but may not fully suppress pro-inflammatory cytokine production in total splenocytes. At 24 days after stress cessation, microglia from

RSD mice treated with imipramine failed to produce exaggerated levels of pro- inflammatory cytokines IL-6, IL-1, TNF-, ex vivo, after LPS stimulation, compared to

RSD mice treated with vehicle. As a matter of interest, there was still a pro-inflammatory phenotype in total splenocytes, 24 days after stress cessation. Imipramine failed to fully 92

attenuate mRNA expression of pro-inflammatory cytokines in total splenocytes after six cycles of RSD and even 24 days after stress termination. It is plausible, that microglia are more sensitive to the anti-inflammatory effect of imipramine than splenocytes. There is evidence that shows a different regulation and responsiveness to LPS administration in central and peripheral cytokine production (Yirmiya et al., 2001).

Microglia from RSD mice, have long-lasting enhanced sensitivity to inflammatory challenges and an amplified response to external stimuli. Several lines of evidence indicate that antidepressants can negatively regulate LPS-induced nuclear translocation of NF-ĸ p65 subunit (Obuchowicz et al., 2014) in microglia. It has been reported that imipramine can shift the balance in the production of pro-inflammatory cytokines such as TNF- and IL1- toward IL-10, an anti-inflammatory cytokine, and can suppress LPS-stimulated cytokine release even at very low concentrations in rat glial cultures (Obuchowicz et al., 2014). It has also been found that imipramine reduced the production of IL1-, TNF- in LPS-activated murine BV2 microglia cells (Hwang et al.,

2008). Imipramine added to culture medium of LPS-stimulated mixed glial cells prevented morphological alterations induced by LPS and transformed microglia cells into cells with a neuron-like morphology (Obuchowicz et al., 2014).

Microglia and brain macrophages directly influence neurophysiology and behavior via cytokines and other secondary mediators. Pro-inflammatory cytokines enhance anorexia, weight loss, and sleep alterations in animal models, as well as decrease interest in social exploration (Castanon et al., 2002). Hence, the present results confirm the beneficial anti-inflammatory properties of antidepressants and the effect on 93

preventing glial activation and decreasing expression of pro-inflammatory molecules. In concert, the anti-inflammatory effect of antidepressants is associated with neuroprotection against cell death in microglia/neuron co-cultures (Hwang et al., 2008).

Nowadays it has been postulated that neuroinflammatory signaling plays a crucial role in the development of anxiety and depression. Pro-inflammatory cytokines can modify neurogenesis in the HPC (Koo et al., 2009) and affect mood. In contrast, as stated in Chapter 1, treatment with antidepressants stimulates HPC neurogenesis and could potentially block atrophy and damage caused by repeated stress (Bokhoven et al., 2011).

Circulating levels of IL-6 are associated with behavioral deficits and depressive-like behavior. Likewise, NF-ĸ signaling is essential to show deficits in social aversion

(Christoffel et al., 2011). Taken together, our data suggest that the long-lasting stress- induced microglial reactivity and pro-inflammatory phenotype at 24 days after RSD cessation may be important factors in prolonged social avoidance behavior. Specifically, social avoidance was associated with sensitized microglia. Imipramine treatment of RSD mice conferred protection against the production of pro-inflammatory molecules in microglia, impeding the social interaction deficit. Therefore, data support the notion that social avoidance behavior is related to neuronal activation and microglia stimulation.

Antidepressant pharmacotherapy is the most often used treatment to date for depression; however the exact mechanism of action is still unclear. Many hypotheses have been taken into consideration for antidepressant activity. It has been thought depressed patients have impairment in neuroimmune endocrine functions. Thus, high levels of circulating corticosteroids and a hyperactivation of the neuroimmune system, 94

causes high levels of pro-inflammatory cytokines in the periphery and CNS as a result of psychosocial stress. Imipramine treatment in the RSD model seemed to have corrected or at least interfered with these alterations.

Figure 9- Imipramine attenuated RSD-induced increase in corticosterone and prevented IL-6 in plasma

(A) Experimental schematic is shown. C57BL/6 mice were randomly allocated in different experimental groups. Mice were exposed to 6 cycles of Repeated Social Defeat (RSD) or left undisturbed and treated with imipramine in their drinking water or vehicle (untreated water) starting two days before social defeat. (B) RSD promoted higher levels of corticosterone in plasma, immediately after the last cycle of RSD. Animals subjected to RSD and treated with imipramine had increased levels of corticosterone than HCC, but lower than RSD animals treated with vehicle. Additionally, (C) mice subjected to RSD and treated with vehicle had increased levels of IL-6 in plasma. Levels of IL-6 in plasma of mice subjected to RSD and treated with imipramine had lower levels of this cytokine, comparable to HCC. Bars represent the meanSEM. Means with different letters (a,b,c) are significantly different from each other (p<0.05).

Figure 10- Effect of imipramine on stress-induced shift of hematopoiesis

(A) Representative bivariate dot plot of CD31/Ly-6c staining of BM. (B) Imipramine reversed stress- induced decrease of erythrocytes but had no effect on stress-induced decrease of lymphocytes. Stress- induced increase of granulocytes and monocytes was reduced with imipramine treatment. Bars represent the meanSEM. Means with different letters (a,b) are significantly different from each other (p<0.05).

Figure 11- Effect of imipramine on stress-induced increase of monocytes and granulocytes in blood and spleen

(A, D) Representative flow bivariate dot plots of Ly-6C/CD11b labeling in blood and spleen respectively. Imipramine decreased stress-induced increase in circulating (B) monocytes and (C) granulocytes. (E) Imipramine blocked stress-induced increase of monocytes in the spleens. RSD mice and with (F) imipramine had a high percentage of granulocytes when compared to HCC. RSD mice treated with vehicle had increased spleen weight compared to HCC. (G) Imipramine didn’t decrease splenomegaly in socially defeated animals. Bars represent the meanSEM. Means with different letters (a,b) are significantly different from each other (p<0.05). 98

Figure 12- Imipramine prevented stress-induced accumulation of macrophages in the brain

Brains were collected the day after the last cycle of RSD, CD11b+ cells were enriched by Percoll gradient separation, and CD45, CD11b expression levels were determined using flow cytometry. (A) Representative bivariate dot plots of the different experimental groups with CD11b/CD45 staining, showing that (B) imipramine blocked stress-induced increased percentage of macrophages in the brain. Bars represent the meanSEM. Means with different letters (a,b) are significantly different from each other (p<0.05). 99

Figure 13- Imipramine prevented stress-associated anxiety-like behavior

(A) Mice subjected to RSD and treated with vehicle, had significantly less number of center entries than HCC and RSD mice treated with imipramine. (B) Mice treated with vehicle and subjected to RSD took more time to enter the center of the open field, and this was reversed with imipramine treatment. (C) A decreased time spent in the open field was assessed in RSD mice treated with vehicle and this was prevented with imipramine treatment. Bars represent the meanSEM. Means with asterisk (*) are significantly different from each other (p<0.05).

100

Figure 14- Imipramine rescued stress-related depressive-like behavior

Depressive-like behavior was evaluated using the social avoidance test. Social withdrawal was exacerbated when an aggressor was introduced in the social target trial (Trial 2) in RSD mice treated with vehicle, with a decreased time spent in the interaction zone, and this was reversed with (A) imipramine treatment. Bars represent the mean  SEM. Means with asterisk (*) are significantly different from each other (p<0.05). Trial 1: time without intruder; Trial 2= time with intruder; RSD=Repeated Social Defeat; (s)=seconds

101

Table 4 – Effect of imipramine on stress-induced mRNA expression of pro-inflammatory cytokines in the spleen

Total splenocytes of socially defeated animals had increased mRNA expression of pro-inflammatory cytokines. (A) Imipramine was not effective in decreasing mRNA expression of these pro-inflammatory cytokines in mice subjected to RSD. Means with different letters (a,b) are significantly different from each other (p<0.05).

Table 5- Imipramine blocked stress-induced neuroinflammatory signaling

Microglia was collected by Percoll gradient separation. (A) Increased neuroinflammatory signaling was observed in RSD mice treated with vehicle and ameliorated in RSD mice treated with imipramine. RSD markedly increased relative gene expression of pro-inflammatory mediators. Means with different letters (a,b) are significantly different from each other (p<0.05).

102

Figure 15- Imipramine treatment reversed RSD-induced social avoidance 24 days after stress cessation

(A) Experimental schematic. (B) Social withdrawal was exacerbated when an aggressor was introduced in the social target trial (Trial 2) in RSD mice treated with vehicle, with a decreased time spent in the interaction zone and this was reversed with imipramine i.p. treatment (20mg/kg). (C) RSD mice treated with vehicle spent more time in the corners during the social target trial (Trial 2). Bars represent the mean  SEM. Means with asterisk (*) are significantly different from each other (p<0.05). Trial 1: time without intruder; Trial 2= time with intruder 103

Table 6- Effect of imipramine on total splenocytes gene expression 24 days after stress termination

Total splenocytes of RSD animals treated with vehicle had increased mRNA expression of TNF- at 24 days after cessation of stress, and this was attenuated with imipramine i.p. treatment (20 mg/kg). Enhanced levels of IL-6 and IL-1 mRNA were detected in both stressed groups. Imipramine did not fully block the expression of either IL-6 or IL1-. Means with different letters (a,b,c) are significantly different from each other (p<0.05). CON= Home caged controls; RSD= Repeated Social Defeat; VCON= Vehicle Controls; ICON= Imipramine Controls; VRSD= Vehicle Repeated Social Defeat; IRSD= Imipramine Repeated Social Defeat

Table 7- Imipramine blocked long-lasting neuroinflammatory signaling provoked by stress exposure

Following behavioral testing, microglia was collected by Percoll gradient separation. RSD markedly increased relative gene expression of IL-6 even at 24 days after cessation of stress, and this was decreased with imipramine i.p. treatment. No differences in IL1-β and TNF-α mRNA expression at 24 days between the RSD and control groups were seen. Means with different letters (a,b,c) are significantly different from each other (p<0.05). CON= Home caged controls; RSD= Repeated Social Defeat; VCON= Vehicle Controls; ICON= Imipramine Controls; VRSD= Vehicle Repeated Social Defeat; IRSD= Imipramine Repeated Social Defeat 104

Table 8- Imipramine treatment blocked microglia activation 24 days after stress cessation

Brains were collected at 24 days after RSD. Enriched microglia were collected and cultured ex vivo. Cells were incubated with LPS (0.1µM) and IL-1, IL-6, and TNF-α relative gene expression was determined from these cells collected 18 h later. Means represent the relative fold change increase over vehicle controls in media ± SEM. Means with different letters (a, b) are significantly different from each other (p<0.05). CON= Home caged controls; RSD= Repeated Social Defeat; VCON= Vehicle Controls; ICON= Imipramine Controls; VRSD= Vehicle Repeated Social Defeat; IRSD= Imipramine Repeated Social Defeat

Figure 16- Imipramine blocked stress-induced exaggerated levels of IL-1β in the supernatant of LPS- stimulated microglia

IL-1 protein levels were determined in supernatant collected after LPS stimulation. Bars represent the relative fold change increase over vehicle controls in media ± SEM determined by RT-PCR, and bars represent protein levels (pg/ml) determined by ELISA. Means with different letters (a,b) are significantly different from each other (p<0.05). 105

Figure 17- Imipramine in drinking water reversed stress-induced social avoidance behavior 24 days after stress cessation

Social Avoidance test was repeated, but this time imipramine was given orally in the animal’s drinking water (15mg/kg). (A) Experimental schematic is shown. (B) Time spent in the interaction zone was significantly decreased in RSD mice treated with vehicle, when an aggressor was introduced in the social target trial (Trial 2), and was reversed with Imipramine treatment. (C) No differences were detected in the time spent in the corners between the groups. Bars represent the mean  SEM. Means with asterisk (*) are significantly different from each other (p<0.05). Trial 1: time without intruder; Trial 2= time with intruder 106

Table 9- Imipramine given orally ameliorated stress-induced long-lasting neuroinflammatory responses

RSD markedly increased relative gene expression of IL-6 in microglia after 24 days of stress cessation, and this was decreased with imipramine treatment diluted in the drinking water (15 mg/kg). TNF-α relative gene expression and IL1- relative gene expression are shown. Means represent the relative fold change increase over vehicle controls in media ± SEM. Means with different letters (a, b) are significantly different from each other (p<0.05). CON= Home caged controls; RSD= Repeated Social Defeat; VCON= Vehicle Controls; ICON= Imipramine Controls; VRSD= Vehicle Repeated Social Defeat; IRSD= Imipramine Repeated Social Defeat

107

Chapter 4: Discussion

Psychosocial stressors are well-established precipitants of anxiety and depression.

Research confirms that chronic psychosocial stress can activate inflammatory responses in both the periphery and the CNS, representing a major breakthrough in the association between stress and mood disorders. In terms of the mechanisms by which stress activates the inflammatory response, attention has been focused on the impact of cytokines on the

HPA axis, specifically related to major depression. Administered cytokines have been shown to stimulate the expression and release of CRH, ACTH, and cortisol, all of these products of HPA axis activation, have been found to be altered in depressed patients

(Miller et al., 2009). One of the suggested pathways in which cytokines may impact HPA axis function is through effects on negative feedback regulation. Stress generates an impaired negative feedback regulation by a decreased responsiveness to GCs. For example, in the RSD model of stress, splenocytes did not have the ability to suppress

NFK-B activity due to a failure of the nuclear translocation of the GC receptor (GR)

(Quan et al., 2003). Increased GC resistance may be related to high levels of cortisol in stress-related depression (Gadek-Michalska et al., 2013). Moreover, it has been shown, that in depressed individuals, CRH is hypersecreted from the PVN in the HYPO and from extrahypothalamic neurons, which leads as well to HPA axis hyperactivity in the

108

presence of GC insensitivity (Gadek-Michalska et al., 2013), suggesting that this enhancement of CRH neuronal activity may be responsible for the behavioral symptoms of mood disorders. Prolonged stress is thought to over-stimulate the HPA axis causing hypersecretion of CRH, which ultimately, if left untreated, leads to impaired HPA axis activity, potentially promoting the onset of psychiatric alterations.

The data presented in the present manuscript is consistent with previous findings that antidepressants, such as imipramine, and BDZs, may produce their effects partly through regulation of the HPA-CRH system. On a related point, anxiety and depression have long been associated with alterations in the secretion of hormones including ACTH, cortisol/corticosterone, CRH, and adrenal catecholamines (Kalueff and Nutt 2007).

Imipramine may have an intrinsic effect on central components of the HPA axis that can explain potentially its therapeutic effects. Several studies have indicated that the GR is upregulated with antidepressants, for example, rats treated with imipramine showed an increase in hippocampal and hypothalamic GR mRNA levels, whereas in vitro treatment for 24 hours increased expression of GR, promoted GR nuclear translocation, and enhanced GR function in rodent fibroblasts (Antonioli et al., 2012). There is evidence mirtazapine, a tricyclic antidepressant like imipramine, decreased cortisol levels in depressed patients (Antonioli et al., 2012). Likely, it has been clearly demonstrated that

BDZs can have neuronal control of the HPA axis through GABAergic neurotransmission modulation. Benzodiazepine administration has been shown to result in reduction of CRH concentration in the locus coeruleus, AMYG and the pyriform cortex, all of which are associated with stress behavior (Gilligan et al., 2000). Moreover, several BDZs decrease 109

ACTH/corticosterone responses to stressors, and have inhibited CRH liberation in the portal vein (Lopez et al., 1999).

Pro-inflammatory cytokines may exert their effect in the CNS by activating the

HPA axis. Cytokines induce gene expression and synthesis of CRH (Janssen et al., 2010).

HPA axis activation may lead to a further rise in pro-inflammatory cytokines, through the positive feedback loop. Stress enhances cytokine levels and induction of catecholamines via an activated HPA axis, which may further increase pro-inflammatory cytokines in the periphery and CNS. Here, we showed imipramine and BDZs were able to silence neuroinflammatory signaling in microglia, thus it is plausibile that stress-induced activation of the HPA axis might have been mitigated by these drugs initially at the CNS level.

It is therefore possible that GABAergic and monoaminergic drugs may correct neuroendocrine dysregulation, through normalization of HPA axis activity and reduced cortisol/corticosterone secretion. In clinical studies, the restoration of HPA axis hyperactivity in depressed patients was associated with remission, and a recent study found that resistance to pharmacological treatment was associated with abnormalities in the HPA axis negative feedback response (Antonioli et al., 2012). Taken all together, the data reported above indicate that imipramine and classical BDZs, lorazepam and clonazepam had an effect on HPA hormones. Imipramine and BDZs attenuated RSD- induced increase of corticosterone. BDZs attenuated CRH in the HYPO of mice exposed to RSD, and these drugs also modulated norepinephrine levels. Therefore, the effects of antidepressant imipramine and BDZs on HPA products should be taken into 110

consideration when endocrinological assessment is performed in patients with suspected

HPA axis dysregulation.

There are many studies showing a direct effect of antidepressants on inflammatory cytokines, and some evidence showing BDZs have this same effect. For example, two studies conducted in animal glial cells indicated a decrease of pro- inflammatory cytokines after exposing these cells to imipramine and stimulation with

LPS and interferon gamma (Antonioli et al., 2012). This goes in hand with the data presented here indicating that imipramine reduced stress-induced priming of microglia, and their exaggerated response to LPS-stimulation. A study from another laboratory showed in the same manner, midazolam, a BDZ that acts on the CBR and TSPO like lorazepam, inhibited pro-inflammatory mediators in LPS-activated macrophages and glial cells (Kim et al., 2006). Therefore it is plausible that lorazepam and clonazepam could also block stress-induced priming of microglia and be considered neuroprotective.

However, it should be noted that this type of assays do not simulate true biological processes and even though these methods have established themselves as standards in immunological research in neuropsychiatric disorders, some care needs to be taken in interpreting in vitro and ex vivo data generated to the human condition.

Over the last few years, evidence from in vivo studies show a clear trend to increased IL-6 plasma levels in both acute and chronic depression, and a tendency to normalize with antidepressants (Janssen et al., 2010). Interestingly, immune cells from patients that have undergone psychological stress do not show enhancement of cytokine production in vitro, even though they did show increased circulating levels of IL-6 111

(Jankord et al., 2010), suggesting that the production of IL-6 might be from another source other than peripheral immune cells. Central injections to block IL-1 or adrenergic signaling attenuated stress-induced increase of circulating IL-6 (Jankord et al.,

2010). There is evidence that psychosocial stress triggers activation of IL-6 neurons in the hypothalamo-neurohyphophyseal system, and this activation is paralleled by an increase in plasma IL-6 (Jankord et al., 2010). Studies indicate magnocellular neurons produce IL-6, and then it is transported to the neurohypophysis, where it is released into the peripheral circulation (Ghorbel et al., 2003; Jankord et al., 2007). In the same manner, pretreatment with propranolol in mice subjected to RSD, and the use of IL-1 receptor type-1 knockout mice in the RSD model, inhibited stress-associated increase in serum IL-

6 comparable to HCC (Wohleb et al., 2014). In the present study antidepressant imipramine and BDZs blocked IL-6 in plasma. Both types of drugs act at the CNS level and support the hypothesis that the stress-induced increase in plasma IL-6 plasma after stress exposure, might come from the brain.

We cannot rule out that the stress-induced increase of IL-6 may come from another source in the periphery. The decrease of peripheral IL-6 in plasma in mice subjected to RSD and treated with imipramine and BDZs, probably regulated microglia activation. Transduction of peripheral cytokines across the blood brain barrier and through circumventricular organs leads to microglia activation that secretes secondary mediators to engage neuronal circuits. In fact, the data presented here indicates IL-6 is an important modulator of RSD-induced social avoidance behavior. After 6 cycles of RSD, imipramine and BDZs blocked increases of IL-6 peripherally and centrally and restored 112

social interaction. At 24 days, social avoidant behavior was prevented with imipramine treatment, paralleled by a blockade of mRNA expression of IL-6 in microglia.

A proposed mechanism for the antidepressant effect on cytokine function is that in vitro, they inhibit pro-inflammatory cytokine activity through cAMP levels, via - adrenoreceptor and/ or dopamine receptor activation (in the case of imipramine) (Janssen et al., 2012). Increased adenylate cyclase, leads to intracellular cAMP levels to block the cAMP-dependent protein kinase (PKA) pathway and ultimately leads to decreased NFKB activity. Also, the PI-3 kinase/AKT pathway is inhibited by elevated cAMP levels

(Janssen et al., 2012). It has been reported antidepressants prevent the down regulation of adenylate cyclase. A similar mechanism is proposed on how BDZs suppress the production of pro-inflammatory cytokines. For example, midazolam, a benzodiazepine that targets both the CBR and TSPO like lorazepam, was found to block the p38MAPK pathway in a LPS-stimulated monocyte-macrophage lineage, while inhibiting NF-KB activation via suppression of IKB- degradation and thus, NF-KB translocation (Kim et al., 2006). Miller et al., 2013 reviewed the literature on how inflammatory cytokines have been shown to increase the reuptake pumps (transporters) for serotonin, norepinephrine, and dopamine. Pharmacological inhibition of p38 MAPK reversed TNF and IL-1 induced increased expression and activity of the serotonin transporter in vitro in rat cell lines and rat mid-brain and striatal synaptosomes. MAPK pathways have been related with functioning of dopamine transporter, where activation of the MAPK kinase increased dopamine reuptake in rat striatal synaptosomes, and the reuptake was rescued with MEK inhibitors. Furthermore, it has been suggested inflammatory cytokines can affect 113

neurotransmitter systems in the brain, and this way play a role in depression and anxiety, including the GABA circuitry. For example, studies in rodents have demonstrated that

GABA can decrease the release of inflammatory cytokines by inhibiting NF-KB and p38

MAPK signaling pathways (Bhat et al., 2010). It should be emphasized that imipramine as well clonazepam were able to suppress stress-induced neuroinflammatory signaling in microglia, but were not fully effective in blocking the expression of pro-inflammatory cytokines in total splenocytes. This suggests that these drugs may only act centrally, while lorazepam had the capacity to act in the periphery possibly through TSPO in the spleen.

In accordance with other studies, imipramine decreased the percentage of circulating monocytes and granulocytes enhanced by stress exposure. It has been shown antidepressants normalized increased counts of monocytes, leukocytes and neutrophils in depressed patients (Seidel et al., 1996; Maes et. al, 1997). This makes sense since imipramine as well, blocked the development of monocytes and granulocytes in the BM.

This manuscript provides novel data demonstrating BDZs, lorazepam and clonazepam were able to suppress the stress-induced increase of circulating monocytes and granulocytes. This supports the notion that these drugs impede the release of pro- inflammatory cells from the BM. The mechanism leading to this reduction in circulating monocytes and granulocytes in mice subjected to RSD and treated with imipramine and

BDZs needs to be further elucidated.

As it has been consistently shown, psychosocial stress promotes activation of peripheral and central myeloid cells (microglia/macrophages) that enhance 114

neuroinflammatory signaling and contribute to the development of anxiety-like behavior, and maintenance of long-lasting social withdrawal. In the follow-up experiments described in this manuscript, we provide evidence that by treating animals exposed to stress with an antidepressant and anxiolytics, silencing of primed resident microglia occurred, and BM-derived monocytes did not accumulate in the CNS. This confirms one more time, that these cells are culprits underlying stress-induced anxiety. However, neuronal activation plays a major part in subsequent microglia activation. Therefore, it is hypothesized one of the major mechanisms of actions of these drugs is inhibiting neuronal activation. Authors have suggested that dysfunctional neuron-microglia crosstalk may be the root of neuroinflammatory signaling after stress exposure (Jurgens and Johnson, 2010). By blocking stress-induced neuronal activation with imipramine and

BDZS, behavioral alterations are prevented.

One of imipramine’s mechanism of medicinal action is antagonizing adreno- receptors, hence its anti-anxiety properties. Additionally, adrenergic receptor antagonists have been found to block increased peripheral blood IL-6 related to stress in humans

(Miller et al., 2009). Previous work has demonstrated that RSD triggered anxiety-like behavior and enhanced the inflammatory state in the periphery and in the CNS in a - adrenergic dependent manner (Hanke et al., 2012; Wohleb et al., 2011). Pre-treatment of

RSD mice with propranolol, a nonselective adrenergic receptor antagonist blocked RSD- induced microglia/macrophage activation and social defeat-induced anxiety, as well as expression of c-Fos in brain regions associated with fear and threat appraisal (Wohleb et al., 2011). It is plausible that imipramine may be antagonizing directly the activation of 115

these adrenergic receptors and decreasing pro-inflammatory cytokine production, and preventing associated stress-related anxiety-like behavior.

There is evidence imipramine blocked reuptake of norepinephrine into presynaptic neurons (Michelson et al., 1997), in fact this is one of its medicinal actions.

The locus coeruleus, contains inhibitory-2 auto receptors, thus, accumulation of norepinephrine at the synapse would decrease the firing of noradrenergic neurons. Since norepinephrine stimulates CRH release, imipramine may contribute by decreasing noradrenergic activation in the brainstem. Likely, several studies showed that BDZs suppressed stress-induced norepinephrine and norepinephrine metabolites turnover

(Arvat et al., 2002). As stated in the discussion of Chapter 2, BDZs also inhibit cell firing and noradrenergic activity in the locus coeruleus. Ablation of noradrenergic locus coeruleus projections is accompanied by a reduction in stress-induced IL-1 production

(Johnson et al., 2005). In other words, central noradrenergic responses exert a major role in stress-induced microglia activation. This suggests these drugs act on noradrenergic function and may influence HPA axis and SNS activity, thus rescuing RSD-induced changes in hematopoiesis. Repeated activation of the SNS due to stress increases norepinephrine in the BM promoting the production of myeloid cells, including monocytes and granulocytes (Hanke et al., 2012). It is plausible that imipramine and lorazepam may also be inhibiting the activation of the SNS in mice subjected to RSD. In humans, the administration of alprazolam, a benzodiazepine that targets both the CBR and TSPO like lorazepam, decreased stress-induced norepinephrine levels (Santagostino et al., 1996). Furthermore, alprazolam has been found to antagonize neuroendocrine and 116

behavioral effects of yohimbine (which increases norepinephrine release by blocking alpha-2 noradrenergic receptors), and vice versa (Charney et al., 1986).

Conclusions

The experiments presented in Chapter 2 confirms that RSD triggered increased gene expression of CRH in the hypothalamus and increased plasma levels of corticosterone, which were attenuated with BDZs. The increased production of these hormones is related to the response of the HPA axis during stress. The neuroendocrine effects of BDZs have been reported previously in the literature (Arvat et al., 2002), however, we provide evidence that BDZs are potent neuroimmunomodulators during psychosocial stress. Furthermore, we confirm BDZs can decrease stress-induced IL-6 in plasma levels, in response to 6 cycles of RSD, and block stress-associated behavior alterations such as prolonged anxiety- and depressive-like behaviors. According to literature (Verleye et al., 2008; Saari et al., 2011), the anxiolytic effect of BDZs is an outcome of targeting the CBR related to the GABA-A receptor. The GABA system may be part of the cytokine signaling response within the CNS, which attempts to maintain an anti-inflammatory CNS environment, essential for optimal neuronal transmission. It is therefore apparent that GABA neurotransmission is an integral part of the stress circuitry and may be of great clinical significance in the development of drugs that modulate functional parts of the GABA receptors in the clinical setting. In recent years, it has been postulated that GABAA-R modulators may serve as novel antidepressants (Möhler, 117

2012). We provide evidence that BDZs improved stress-induced deficits in social avoidance associated with depressive-like behavior. Additionally, the anti-inflammatory actions produced by targeting the TSPO during psychosocial stress in the periphery and

CNS, offers new clinical therapeutic opportunities. By exploiting this knowledge, new tools for diagnosis and therapeutic interventions can be developed in order to treat stress- related disorders.

The experiments in Chapter 3 demonstrate that antidepressants, such as imipramine, may exert their effects, in part, by down regulating neuroinflammation. It is not well known whether immunomodulation is a side effect or part of the clinical activity of imipramine. Since, hyperactivation of the neuroimmune system is believed to be involved in pathogenesis/development of mood disorders and anxiety, the present study suggests that the anti-inflammatory effect of antidepressants may have protective effects by silencing RSD-induced priming and activation of microglia, thus down regulating the biosynthesis of high levels of pro-inflammatory cytokines. By elucidating the underlying mechanism of antidepressant action and neuroimmune responses clinicians may be able to optimize therapeutic interventions of mood disorders and anxiety.

118

References

Anholt RR, De Souza EB, Oster-Granite ML, Snyder SH. 1985. Peripheral-type benzodiazepine receptors: Autoradiographic localization in whole-body sections of neonatal rats. The Journal of Pharmacology and Experimental Therapeutics 233 (2): 517- 26.

Anholt RR, Pedersen PL, De Souza EB, Snyder SH. 1986. The peripheral-type benzodiazepine receptor. Localization to the mitochondrial outer membrane. The Journal of Biological Chemistry 261 (2): 576-83.

Antonioli M, Rybka J, Carvalho LA. 2012. Neuroimmune endocrine effects of antidepressants. Neuropsychiatric Disease and Treatment 8: 65-83.

Arvat,E, Giordano,R, Grottoli S, Ghigo, E. 2002. Review Article- Benzodiazepines and anterior pituitary function. Journal of Endocrinological Investigation 25 (8): 735.

Avitsur R, Padgett DA, Dhabhar FS, Stark JL, Kramer KA, Engler H, Sheridan JF. 2003. Expression of glucocorticoid resistance following social stress requires a second signal. Journal of Leukocyte Biology 74 (4): 507-13.

Avitsur R, Stark JL, Dhabhar FS, Sheridan JF. 2002. Social stress alters splenocyte phenotype and function. Journal of Neuroimmunology 132 (1-2):1-2.

Avitsur R, Powell N, Padgett DA, Sheridan JF. 2009. Social interactions, stress, and immunity. Immunology and Allergy Clinics of North America 29 (2): 285-93.

Avitsur R, Stark JL, Sheridan JF. 2001. Social stress induces glucocorticoid resistance in subordinate animals. Hormones and Behavior 39 (4): 247-57.

Ayelli EV, Sterin-Borda, L, Cremaschi GA, Genaro AM. 1999. Role of protein kinase C and cAMP in fluoxetine effects on human T-cell proliferation. European Journal of Pharmacology 372 (1): 65-73.

119

Bailey MT, Kierstein S, Sharma S, Spaits M, Kinsey SG, Tliba O, Amrani Y, Sheridan JF, Panettieri RA, Haczku A. 2009. Social stress enhances allergen-induced airway inflammation in mice and inhibits corticosteroid responsiveness of cytokine production. Journal of Immunology 182 (12): 7888-96.

Baune BT, Smith E, Reppermund S, Air T, Samaras K, Lux O, Brodaty H, Sachdev P, Trollor JN. 2012. Inflammatory biomarkers predict depressive, but not anxiety symptoms during aging: The prospective sydney memory and aging study. Psychoneuroendocrinology 37 (9): 1521-30.

Berton O, McClung CA, Dileone RJ, Krishnan V, Renthal W, Russo SJ, Graham D, Tsankova NM, Bolanos CA, Rios M, Monteggia LM, Self DW, Nestler EJ. 2006. Essential role of BDNF in the mesolimbic dopamine pathway in social defeat stress. Science 311 (5762): 864-8.

Bhat R, Axtell R, Mitra A, Miranda M, Lock C, Tsien RW, Steinman L. 2010. Inhibitory role for GABA in autoimmune inflammation. Proceedings of the National Academy of Sciences of the United States of America 107 (6): 2580-5.

Calogero AE, Gallucci WT, Chrousos GP, Gold PW. 1988. Interaction between GABAergic neurotransmission and rat hypothalamic corticotropin-releasing hormone secretion in vitro. Brain Research 463 (1): 28-36.

Canli T, Lesch KP. 2007. Long story short: The serotonin transporter in emotion regulation and social cognition. Nature Neuroscience 10 (9): 1103-9.

Cannon WB. 1929. Pharmacological injections and physiological inferences. Science 70 (1821): 500-1.

Casellas P. 2002. Peripheral benzodiazepine receptors and mitochondrial function. Neurochemistry International 40 (6): 475-86.

Castanon N, Leonard B, Neveu P, Yirmiya R. 2002. Effects of antidepressants on cytokine production and actions. Brain, Behavior, and Immunity 16 (5): 569-74.

Challis C, Beck SG, Berton O. 2014. Optogenetic modulation of descending prefrontocortical inputs to the dorsal raphe bidirectionally bias socioaffective choices after social defeat. Frontiers in Behavioral Neuroscience 8 :43.

Charney DS, Breier AJ, Peter I, Heninger GR. 1986. Behavioral, biochemical, and blood pressure responses to alprazolam in healthy subjects: Interactions with yohimbine. Psychopharmacology 88 (2): 133-40.

120

Cherry JA, Thompson BE, Pho V. 2001. Diazepam and rolipram differentially inhibit cyclic AMP-specific phosphodiesterases PDE4A1 and PDE4B3 in the mouse. Biochimica Et Biophysica Acta. Gene Structure and Expression. 1518 (1): 27.

Choi J, Ifuku M, Noda M, Guilarte TR. 2011. Translocator protein (18 kDa)/peripheral benzodiazepine receptor specific ligands induce microglia functions consistent with an activated state. Glia 59 (2): 219-30.

Christoffel DJ, Golden SA, Dumitriu D, Robison AJ, Janssen WG, Ahn HF, Krishnan V, Reyes CM, Han MH, Ables JL, Eisch AJ, Dietz DM, Ferguson D, Neve RL, Greengard P, Kim Y, Morrison JH, Russo SJ. 2011. IκB kinase regulates social defeat stress-induced synaptic and behavioral plasticity. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience 31 (1): 314-21.

Cole SW, Arevalo JM, Takahashi R, Sloan EK, Lutgendorf SK, Sood AK, Sheridan JF, Seeman TE. 2010. Computational identification of gene-social environment interaction at the human IL6 locus. Proceedings of the National Academy of Sciences of the United States of America 107 (12): 5681-6.

Cox DA, Ellinor PT, Kirley TL, Matlib MA. 1991. Identification of a 17-kDa protein associated with the peripheral-type benzodiazepine receptor in vascular and other smooth muscle types. The Journal of Pharmacology and Experimental Therapeutics 258 (2): 702- 9.

Czéh B, Michaelis T, Watanabe T, Frahm J, de Biurrun G, van Kampen M, Bartolomucci A, Fuchs E. 2001. Stress-induced changes in cerebral metabolites, hippocampal volume, and cell proliferation are prevented by antidepressant treatment with tianeptine. Proceedings of the National Academy of Sciences of the United States of America 98 (22): 12796-801. de Angelis, Luisa. 1995. Effects of valproate and lorazepam on experimental anxiety: Tolerance, withdrawal, and role of clonidine. Pharmacology Biochemistry and Behavior 52 (2): 329-33.

De Souza EB. 1990. Neuroendocrine effects of benzodiazepines. Journal of Psychiatric Research 24 : 111-9.

Engler H, Bailey MT, Engler A, Sheridan JF. 2004. Effects of repeated social stress on leukocyte distribution in bone marrow, peripheral blood and spleen. Journal of Neuroimmunology 148 (1-2): 106-15.

121

Engler H, Engler A, Bailey MT, Sheridan JF. 2005. Tissue-specific alterations in the glucocorticoid sensitivity of immune cells following repeated social defeat in mice. Journal of Neuroimmunology 163 (1-2): 110-9.

Ferrarese C, Appollonio I, Bianchi G, Frigo M, Marzorati C, Pecora N, Perego M, Pierpaoli C, Frattola L. 1993. Benzodiazepine receptors and diazepam binding inhibitor: A possible link between stress, anxiety and the immune system. Psychoneuroendocrinology 18 (1): 3-22.

File SE. 1991. Chapter 2 – Interactions of anxyolitic and antidepressants drugs with hormones of the hypothalamic-pituitary-adrenal axis. In: File SE, Psychopharmacology of Anxiolytics and Antidepressants Amsterdam: Pergamon; 29-55.

Fujimura Y, Zoghbi SS, Simeon FG, Taku A, Pike VW, Innis RB, Fujita M. 2009. Quantification of translocator protein (18 kDa) in the human brain with PET and a novel radioligand, 18F-TSPO06. Journal of Nuclear Medicine 50 (7): 1047-53.

Gądek-Michalska A, Spyrka J, Rachwalska P, Tadeusz J, Bugajski J. 2013. Influence of chronic stress on brain corticosteroid receptors and HPA axis activity. Pharmacological Reports : PR 65 (5): 1163-75.

Gądek-Michalska A, Tadeusz J, Rachwalska P, Bugajski J. 2013. Cytokines, prostaglandins and nitric oxide in the regulation of stress-response systems. Pharmacological Reports: PR 65 (6): 1655-62.

Ghorbel MT, Sharman G, Leroux M, Barrett T, Donovan DM, Becker KG, Murphy D. 2003. Microarray analysis reveals interleukin-6 as a novel secretory product of the hypothalamo-neurohypophyseal system. The Journal of Biological Chemistry 278 (21): 19280-5.

Gilligan PJ, Robertson DW, Zaczek R. 2000. Corticotropin releasing factor (CRF) receptor modulators: Progress and opportunities for new therapeutic agents. Journal of Medicinal Chemistry 43 (9): 1641-60.

Goshen I, Yirmiya R. 2009. Interleukin-1 (IL-1): A central regulator of stress responses. Frontiers in Neuroendocrinology 30 (1): 30-45.

Ha E, Jung KH, Choe BK, Bae JH, Shin DH, Yim SV, Baik HH. 2006. Fluoxetine increases the nitric oxide production via nuclear factor kappa B-mediated pathway in BV- 2 murine microglial cells. Neuroscience Letters 397 (3): 185-9.

122

Hanke ML, Powell ND, Stiner LM, Bailey MT, Sheridan JF. 2012. Beta adrenergic blockade decreases the immunomodulatory effects of social disruption stress. Brain Behavior and Immunity 26 (7): 1150-9.

Hashioka S, Klegeris A, Monji A, Kato T, Sawada M, McGeer PL, Kanba S. 2007. Antidepressants inhibit interferon-induced microglial production of IL-6 and nitric oxide. Experimental Neurology 206 (1): 33-42.

Heidt T, Sager HB, Courties G, Dutta P, Iwamoto Y, Zaltsman A, Lin CP, Vinegoni C, Swirski FK, Weissleder R, Nahrendorf M, Von Zur MC, Bode C, Fricchione GL, Denninger J, Fricchione GL, Denninger J, Libby P, Weissleder R. 2014. Chronic variable stress activates hematopoietic stem cells. Nature Medicine 20 (7): 754-8.

Henry CJ, Huang, Y, Wynne AM, Godbout JP. 2009. Peripheral lipopolysaccharide (LPS) challenge promotes microglial hyperactivity in aged mice that is associated with exaggerated induction of both pro-inflammatory IL-1β and anti-inflammatory IL-10 cytokines. Brain, Behavior, and Immunity 23 (3): 309-17.

Hinwood M, Morandini J, Day TA, Walker FR. 2012. Evidence that microglia mediate the neurobiological effects of chronic psychological stress on the medial prefrontal cortex. Cerebral Cortex 22 (6): 1442-54.

Hodes GE, Pfau ML, Leboeuf M, Golden SA, Christoffel DJ, Bregman D, Rebusi N, Heshmati M, Aleyasin H, Warren BL, Labont B, Horn S, Lapidus KA, Stelzhammer V, Wong, Erik HF, Bahn S, Krishnan V, Bolanos-Guzman CA, Murrough, JW, Merad M, Russo, SJ. 2014. Individual differences in the peripheral immune system promote resilience versus susceptibility to social stress. Proceedings of the National Academy of Sciences 111 (45): 16136-41.

Hwang J, Zheng LT, Ock J, Lee MG, Kim SH, Lee HW, Lee WH, Park HC, Suk K. 2008. Inhibition of glial inflammatory activation and neurotoxicity by tricyclic antidepressants. Neuropharmacology 55 (5): 826-34.

Imaki T, Wang XQ, Shibasaki T, Harada S, Chikada N, Takahashi C, Naruse M, Demura H. 1995. Chlordiazepoxide attenuates stress-induced activation of neurons, corticotropin- releasing factor (CRF) gene transcription and CRF biosynthesis in the paraventricular nucleus (PVN). Molecular Brain Research 32 (2): 261-70.

Ishikawa I, Kitamura H, Kimura K, Saito M. 2001. Brain interleukin-1 is involved in blood interleukin-6 response to immobilization stress in rats. The Japanese Journal of Veterinary Research 49 (1): 19-25.

123

Jankord R, Turk JR, Schadt JC, Casati J, Ganjam VK, Price EM, Keisler DH, Laughlin MH. 2007. Sex difference in link between interleukin-6 and stress. Endocrinology 148 (8): 3758-64.

Jankord R, Zhang R, Flak JN, Solomon MB, Albertz J, Herman JP. 2010. Stress activation of IL-6 neurons in the hypothalamus. American Journal of Physiology. Regulatory, Integrative and Comparative Physiology 299 (1): 343-51.

Janssen DGA, Caniato RN, Verster, JC, Baune BT. 2010. A psychoneuroimmunological review on cytokines involved in antidepressant treatment response. Human Psychopharmacology: Clinical and Experimental 25 (3): 201-15.

Jin Z, Mendu SK, Birnir B. 2013. GABA is an effective immunomodulatory molecule. Amino Acids 45 (1): 87-94.

Johnson JD, Campisi J, Sharkey CM, Kennedy SL, Nickerson M, Greenwood BN, Fleshner M. 2005. Catecholamines mediate stress-induced increases in peripheral and central inflammatory cytokines. Neuroscience 135 (4): 1295-307.

Joo HK, Oh SC, Cho EJ, Park KS, Lee JY, Lee EJ, Lee SK, Kim HS, Park JB, Jeon BH. 2009. Midazolam inhibits tumor necrosis factor-alpha-induced endothelial activation: Involvement of the peripheral benzodiazepine receptor. Anesthesiology 110 (1): 106-12.

Jurgens HA, Johnson RW. 2012. Dysregulated neuronal-microglial cross-talk during aging, stress and inflammation. Experimental Neurology 233 (1): 40-8.

Kalueff AV, Nutt DJ. 2007. Role of GABA in anxiety and depression. Depression and Anxiety 24 (7): 495-517.

Keim KL, Sigg EB. 1977. Plasma corticosterone and brain catecholamines in stress: Effect of psychotropic drugs. Pharmacology Biochemistry and Behavior 6 (1): 79-85.

Kenis G, Maes M. 2002. Effects of antidepressants on the production of cytokines. The International Journal of Neuropsychopharmacology 5 (4): 401-12.

Kiecolt-Glaser JK, Glaser R. 2002. Depression and immune function: Central pathways to morbidity and mortality. Journal of Psychosomatic Research 53 (4):873-6.

Kim SN, Son SC, Lee SM, Kim CS, Yoo DG, Lee SK, Hur GM, Park JB, Jeon BH. 2006. Midazolam inhibits proinflammatory mediators in the lipopolysaccharide-activated macrophage. Anesthesiology 105 (1): 105-10.

124

Kinsey SG, Bailey MT, Sheridan JF, Padgett DA, Avitsur R. 2007. Repeated social defeat causes increased anxiety-like behavior and alters splenocyte function in C57BL/6 and CD-1 mice. Brain Behavior and Immunity 21 (4): 458-66.

Koo JW, Russo SJ, Ferguson D, Nestler EJ, Duman RS. 2010. Nuclear factor-kappaB is a critical mediator of stress-impaired neurogenesis and depressive behavior. Proceedings of the National Academy of Sciences of the United States of America 107 (6): 2669-74.

Koo JW, Duman RS. 2009. Interleukin-1 receptor null mutant mice show decreased anxiety-like behavior and enhanced fear memory. Neuroscience Letters 456 (1): 39-43.

Kubera M, Holan V, Mathison R, Maes M. 2000. The effect of repeated amitriptyline and desipramine administration on cytokine release in C57BL/6 mice. Psychoneuroendocrinology 25 (8): 785-97.

Kudryavtseva NN, Shurlygina AV, Melnikova EV, Tenditnik MV, Bondar NP, Panteleeva NG, Smagin DA, Kolesnikov NN, Trufakin, VA. 2011. Cell cycle arrest in the thymus and spleen in male mice under conditions of chronic social defeat stress: Effects of diazepam. Bulletin of Experimental Biology and Medicine 151 (4): 411-4.

Lim CM, Kim SW, Park JY, Kim C, Yoon SH, Lee JK. 2009. Fluoxetine affords robust neuroprotection in the postischemic brain via its anti-inflammatory effect. Journal of Neuroscience Research 87 (4): 1037-45.

Liu GJ, Middleton RJ, Hatty CR, Kam WW, Chan R, Pham T, Harrison-Brown M, Dodson E, Veale K, Banati RB. 2014. The 18 kDa translocator protein, microglia and neuroinflammation. Brain Pathology 24 (6): 631-53.

Liu R, Jolas T, Aghajanian G. 2000. Serotonin 5-HT2 receptors activate local GABA inhibitory inputs to serotonergic neurons of the dorsal raphe nucleus. Brain Research 873 (1): 34-45.

López JF, Akil H, Watson SJ. 1999. Neural circuits mediating stress. Biological Psychiatry 46 (11): 1461-71.

Maes M, Kubera M, Obuchowiczwa E, Goehler L, Brzeszcz J. 2011. Depression's multiple comorbidities explained by neuroinflammatory and oxidative & nitrosative stress pathways. Neuro Endocrinology Letters 32 (1): 7-24.

Maes M, Vandoolaeghe E, Van Hunsel F, Bril T, Demedts P, Wauters A, Neels H. 1997. Immune disturbances in treatment-resistant depression: Modulation by antidepressive treatments. Human Psychopharmacology: Clinical & Experimental 12 (2):153-162.

125

Maes M, Verkerk R, Bonaccorso S, Ombelet W, Bosmans E, Scharpé S. 2002. Depressive and anxiety symptoms in the early puerperium are related to increased degradation of tryptophan into kynurenine, a phenomenon which is related to immune activation. Life Sciences 71 (16): 1837-48.

Maes M. 2001. The immunoregulatory effects of antidepressants. HUP Human Psychopharmacology: Clinical and Experimental 16 (1): 95-103.

Malberg JE, Eisch AJ, Nestler EJ, Duman RS. 2000. Chronic antidepressant treatment increases neurogenesis in adult rat hippocampus. The Journal of Neuroscience : The Official Journal of the Society for Neuroscience 20 (24): 9104-10.

Marques AH, Silverman MN, Sternberg EM. 2009. Glucocorticoid dysregulations and their clinical correlates from receptors to therapeutics. Annals of the New York Academy of Sciences 1179 (01): 1-18.

Michelson D, Galliven E, Hill L, Demitrack M, Chrousos G, Gold P. 1997. Chronic imipramine is associated with diminished hypothalamic-pituitary-adrenal axis responsivity in healthy humans. The Journal of Clinical Endocrinology and Metabolism 82 (8): 2601-6.

Miller AH, Haroon E, Raison CL, Felger JC. 2013. Cytokine targets in the brain: Impact on neurotransmitters and neurocircuits. Depression and Anxiety 30 (4): 297-306.

Miller AH, Maletic V, Raison CL. 2009. Inflammation and its discontents: The role of cytokines in the pathophysiology of major depression. Biological Psychiatry 65 (9): 732- 41.

Miller LG, Thompson ML, Greenblatt DJ, Deutsch SI, Shader RI, Paul SM. 1987. Rapid increase in brain benzodiazepine receptor binding following defeat stress in mice. Brain Research 414 (2): 395-400.

Möhler H. 2012. The GABA system in anxiety and depression and its therapeutic potential. Neuropharmacology 62 (1): 42-53.

Nair A, Bonneau RH. 2006. Stress-induced elevation of glucocorticoids increases microglia proliferation through NMDA receptor activation. Journal of Neuroimmunology 171 (1-2): 72-85.

Nutt DJ, Malizia AL. 2001. New insights into the role of the GABA(A)-benzodiazepine receptor in psychiatric disorder. The British Journal of Psychiatry: The Journal of Mental Science 179: 390-6.

126

Obuchowicz E, Bielecka AM, Paul-Samojedny M, Pudełko A, Kowalski J. 2014. Imipramine and fluoxetine inhibit LPS-induced activation and affect morphology of microglial cells in the rat glial culture. Pharmacological Reports 66 (1): 34-43.

Ohno H, Ogawa M, Nishikawa S, Hayashi S, Kunisada T, Nishikawa S. 1993. Conditions required for myelopoiesis in murine spleen. Immunology Letters 35 (2): 197-204.

Ozaktay AC, Kallakuri S, Takebayashi T, Cavanaugh JM, Asik Ibrahim, DeLeo JA, Weinstein JN. 2006. Effects of interleukin-1 beta, interleukin-6, and tumor necrosis factor on sensitivity of dorsal root ganglion and peripheral receptive fields in rats. European Spine Journal 15 (10): 1529-1537.

Papadopoulos V, Baraldi M, Guilarte TR, Knudsen TB, Lacapère JJ, Lindemann P, Norenberg MD, Nutt D, Weizman A, Zhang MR, Gavish M. 2006. Translocator protein (18 kDa): New nomenclature for the peripheral-type benzodiazepine receptor based on its structure and molecular function. Trends in Pharmacological Sciences 27 (8): 402-9.

Park CH, Carboni E, Wood PL, Gee KW. 1996. Characterization of peripheral benzodiazepine type sites in a cultured murine BV-2 microglial cell line. Glia 16 (1): 65- 70.

Passamonti L, Crockett MJ, Apergis-Schoute AM, Clark L, Rowe JB, Calder AJ, Robbins TW. 2012. Effects of acute tryptophan depletion on prefrontal-amygdala connectivity while viewing facial signals of aggression. Biological Psychiatry 71 (1): 36- 43.

Powell ND, Sloan EK, Bailey MT, Arevalo JM, Miller GE, Chen E, Kobor MS, Reader BF, Sheridan JF, Cole SW. 2013. Social stress up-regulates inflammatory gene expression in the leukocyte transcriptome via β-adrenergic induction of myelopoiesis. Proceedings of the National Academy of Sciences of the United States of America 110 (41): 16574-9.

Quan N, Avitsur R, Stark JL, He L, Lai W, Dhabhar F, Sheridan JF. 2003. Molecular mechanisms of glucocorticoid resistance in splenocytes of socially stressed male mice. Journal of Neuroimmunology 137 (1-2): 51-8.

Radley JJ, Sisti HM, Hao J, Rocher AB, McCall T, Hof PR, McEwen BS, Morrison JH. 2004. Chronic behavioral stress induces apical dendritic reorganization in pyramidal neurons of the medial prefrontal cortex. Neuroscience 125(1):1-6.

Ramirez KG, Shea DT, McKim DB, BF Reader, Sheridan JF. Imipramine attenuates neuroinflammatory signaling and reverses stress-induced social avoidance. Brain Behavior and Immunity 2015. doi:10.1016/j.bbi.2015.01.016. 127

Reader BF, Jarrett BL, McKim DB, Wohleb ES, Godbout JP, Sheridan JF. 2015. Peripheral and central effects of repeated social defeat stress: Monocyte trafficking, microglial activation, and anxiety. Neuroscience 289 (1): 429-42.

Ruff MR, Pert CB, Weber RJ, Wahl LM, Wahl SM, Paul SM. 1985. Benzodiazepine receptor-mediated chemotaxis of human monocytes. Science 229 (4719): 1281-3.

Russo SJ, Nestler EJ. 2013. The brain reward circuitry in mood disorders. Nature Reviews. Neuroscience 14 (9): 609-25.

Saari TI, Uusi-Oukari M, Ahonen J, Olkkola KT. 2011. Enhancement of GABAergic activity: Neuropharmacological effects of benzodiazepines and therapeutic use in anesthesiology. Pharmacological Reviews 63 (1): 243-67.

Sacerdote P, Panerai AE, Frattola L, Ferrarese C. 1999. Benzodiazepine-induced chemotaxis is impaired in monocytes from patients with generalized anxiety disorder. Psychoneuroendocrinology 24 (2): 243-9.

Santagostino G, Amoretti G, Frattini P, Zerbi F, Cucchi ML, Preda S, Corona GL. 1996. Catecholaminergic, neuroendocrine and anxiety responses to acute psychological stress in healthy subjects: Influence of alprazolam administration. Neuropsychobiology 34 (1): 36- 43.

Santarelli L, Saxe M, Gross C, Surget A, Battaglia F, Dulawa S, Weisstaub N, Lee J, Duman R, Arancio O, Belzung C, Hen R. 2003. Requirement of hippocampal neurogenesis for the behavioral effects of antidepressants. Science 301 (5634): 805-9.

Seidel A, Arolt V, Hunstiger M, Rink L, Behnisch A, Kirchner H. 1996. Major depressive disorder is associated with elevated monocyte counts. Acta Psychiatrica Scandinavica 94 (3): 198-204.

Selye H. 1946. The general adaptation syndrome and the diseases of adaptation. The Journal of Allergy 17 (6): 117-230.

Sheehan TP, Chambers RA, Russell DS. Regulation of affect by the lateral septum: implications for neuropsychiatry. 2004. Brain Research Reviews 46(1):71-117.

Simmons RD, Miller RK, Kellogg CK. 1984. Prenatal exposure to diazepam alters central and peripheral responses to stress in adult rat offspring. Brain Research 307 (1-2): 39-46.

128

Sluzewska A, Rybakowski JK, Laciak M, Mackiewicz A, Sobieska M, Wiktorowicv K. 1995. Interleukin-6 serum levels in depressed patients before and after treatment with fluoxetine. Annals of the New York Academy of Sciences 762 (1): 474-6.

Spina, D. 2008. PDE4 inhibitors: Current status. BPH British Journal of Pharmacology 155 (3): 308-15.

Stark JL, Avitsur R, Hunzeker J, Padgett DA, Sheridan JF. 2002. Interleukin-6 and the development of social disruption-induced glucocorticoid resistance. Journal of Neuroimmunology 124 (1-2): 9-15.

Sternberg EM. 2006. Neural regulation of innate immunity: A coordinated nonspecific host response to pathogens. Nature Reviews. Immunology 6 (4): 318-28.

Strekalova T, Anthony DC, Dolgov O, Anokhin K, Kubatiev A, Steinbusch HM, Schroeter C. 2013. The differential effects of chronic imipramine or citalopram administration on physiological and behavioral outcomes in naïve mice. Behavioural Brain Research 245: 101-6.

Taupin V, Jayais P, Descamps-Latscha B, Cazalaa JB, Barrier G, Bach JF, Zavala F. 1991. Benzodiazepine anesthesia in humans modulates the interleukin-1 beta, tumor necrosis factor-alpha and interleukin-6 responses of blood monocytes. Journal of Neuroimmunology 35 (1-3):13-9.

Trincavelli ML, Da Pozzo E, Daniele S, Martini C. 2012. The GABAA-BZR complex as target for the development of anxiolytic drugs. Current Topics in Medicinal Chemistry 12 (4): 254-69.

Tsankova NM, Berton O, Renthal W, Kumar A, Neve RL, Nestler EJ. 2006. Sustained hippocampal chromatin regulation in a mouse model of depression and antidepressant action. Nature Neuroscience 9 (4): 519-25.

Uusi-Oukari M, Korpi ER. 2010. Regulation of GABA(A) receptor subunit expression by pharmacological agents. Pharmacological Reviews 62 (1): 97-135.

Van Bokhoven P, Oomen CA, Hoogendijk WJ, Smit AB, Lucassen PJ, Spijker S. 2011. Reduction in hippocampal neurogenesis after social defeat is long-lasting and responsive to late antidepressant treatment. The European Journal of Neuroscience 33 (10): 1833-40.

Veenman L, Gavish M. New Developments in Receptor Pharmacology. 2006. The peripheral-type benzodiazepine receptor and the cardiovascular system. Implications for drug development. Pharmacology and Therapeutics 110 (3): 503-24.

129

Verleye M, Heulard I, Gillardin JM. 2008. Investigation of the anticonvulsive effect of acute immobilization stress in anxious Balb/cByJ mice using GABA A-related mechanistic probes. Psychopharmacology 197 (4): 523-34.

Wilms, H. 2003. Involvement of benzodiazepine receptors in neuroinflammatory and neurodegenerative diseases: Evidence from activated microglial cells in vitro. Neurobiology of Disease 14 (3): 417-24.

Wohleb ES, Hanke ML, Corona AW, Powell ND, Stiner LM, Bailey MT, Nelson RJ, Godbout JP, Sheridan JF. 2011. β-Adrenergic receptor antagonism prevents anxiety-like behavior and microglial reactivity induced by repeated social defeat. The Journal of Neuroscience : The Official Journal of the Society for Neuroscience 31 (17): 6277-88.

Wohleb ES, McKim DB, Sheridan JF, Godbout JP. 2014. Monocyte trafficking to the brain with stress and inflammation: A novel axis of immune-to-brain communication that influences mood and behavior. Frontiers in Neuroscience 8:447.

Wohleb, ES, McKim DB, Shea DT, Powell ND, Tarr AJ, Sheridan JF, Godbout JP. 2013. Re-establishment of anxiety in stress-sensitized mice is caused by monocyte trafficking from the spleen to the brain. Biological Psychiatry 75(12):970-81.

Wynne AM, Henry CJ, Huang Y, Cleland A, Godbout JP. 2010. Protracted downregulation of CX3CR1 on microglia of aged mice after lipopolysaccharide challenge. Brain Behavior and Immunity 24 (7): 1190-201.

Xia Z, DePierre JW, Nassberger L. 1996. Tricyclic antidepressants inhibit IL-6, IL-1b and TNF-a release in human blood monocytes and IL-2 and interferon-g in T cells. Immunopharmacology 34 (2-3): 27.

Yirmiya R, Pollak Y, Barak O, Avitsur R, Ovadia H, Bette M, Weihe E, Weidenfeld J. 2001. Effects of antidepressant drugs on the behavioral and physiological responses to lipopolysaccharide (LPS) in rodents. Neuropsychopharmacology : Official Publication of the American College of Neuropsychopharmacology 24 (5): 531-44.

Yirmiya R. 1996. Endotoxin produces a depressive-like episode in rats. Brain Research 711 (1): 163-74.

Yousefi OS, Wilhelm T, Maschke-Neuß K, Kuhny M, Martin C, Molderings GJ, Kratz F, Hildenbrand B, Huber M. 2013. The 1,4-benzodiazepine Ro5-4864 (4-chlorodiazepam) suppresses multiple pro-inflammatory mast cell effector functions. Cell Communication and Signaling 11 (1): 13.

130

Zavala F, Taupin V, Descamps-Latscha B. 1990. In vivo treatment with benzodiazepines inhibits murine phagocyte oxidative metabolism and production of interleukin 1, tumor necrosis factor and interleukin-6. The Journal of Pharmacology and Experimental Therapeutics 255 (2): 442-50.

Zavala F, Haumont J, Lenfant M. 1984. Interaction of benzodiazepines with mouse macrophages. European Journal of Pharmacology 106 (3): 561-6.

131