Lipocalin 2 and the pathophysiology of Alzheimer's disease Dekens, Doortje W.
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Dekens, D. W. (2019). Lipocalin 2 and the pathophysiology of Alzheimer's disease. Rijksuniversiteit Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Lipocalin 2 and the pathophysiology
of Alzheimer’s disease
Neurobiology, Groningen Institute for Evolutionary Life Sciences (GELIFES), University of Groningen, Groningen, The Netherlands and at the Department of Neurology and Alzheimer Centrum Groningen (ACG), University Medical Center Groningen, Groningen, The Netherlands. The studies in this thesis were financially supported by the Research School of Behavioral and Cognitive Neurosciences (BCN), the Internationale Stichting Alzheimer Onderzoek (ISAO#06511), and Stichting Hadders-De Cock (#2017-30).
Printing of this thesis was financially supported by the University of Groningen, the Graduate School of Medical Sciences (GSMS), University Medical Center Groningen (UMCG), and the Research School of Behavioral and Cognitive Neurosciences (BCN).
ISBN: 978-94-034-1611-3 (printed version) ISBN: 978-94-034-1610-6 (electronic version) Layout and cover design: Doortje Dekens
Printed by: Ipskamp Printing (www.proefschriften.net)
Copyright © D.W.Dekens, 2019.
All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, without prior written permission of the author.
Lipocalin 2 and the pathophysiology
of Alzheimer’s disease
Proefschrift
ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen
op gezag van de
rector magnificus prof. dr. E. Sterken
en volgens besluit van het College voor Promoties. De openbare verdediging zal plaatsvinden op
woensdag 15 mei 2019 om 11:00 uur
door
Dorine Willemijn Dekens
geboren op 2 juli 1990Prof. dr. U.L.M. Eisel
Copromotor
Dr. P.J.W. Naudé
Beoordelingscommissie
Prof. dr. P.J. Lucassen Prof. dr. J.A. Palha Prof. dr. E.A. van der Zee
Chapter 1
7General introduction
Chapter 2
19Neutrophil Gelatinase-Associated Lipocalin and its Receptors in Alzheimer's Disease (AD) Brain Regions: Differential Findings in AD with and without Depression
Chapter 3
43Lipocalin 2 contributes to brain iron dysregulation but does not affect cognition, plaque load, and glial activation in the J20 Alzheimer mouse model
Chapter 4
71Iron chelators inhibit amyloid-β-induced production of Lipocalin 2 in cultured astrocytes
Chapter 5
83Dynamics of neutrophil gelatinase-associated lipocalin plasma and cerebrospinal fluid concentrations in older males
Chapter 6
97Lipocalin 2 as a link between aging, risk factor conditions and age-related brain diseases
Chapter 7
145 General discussionAppendix
165 Summary 167 Nederlandse samenvatting 171 Dankwoord 175 Curriculum vitae 179 List of publications 180Chapter 1
9
1
General introduction
Our abilities to think, perceive, feel, create and store and retrieve memories, make decisions, plan and carry out actions and display behavior depend on the functioning of our brains. This proper functioning of the brain relies on the health of the brain cells that inhabit our brain. The brain consists of different types of brain cells, including amongst others: neurons, astrocytes, microglia and oligodendrocytes.
Neurons are nerve cells, which communicate with each other via electrical signals. A neuron can send electrical signals along its axon, which is a long extension that protrudes from the neuron’s cell body, until the signals reach the end of the axon. Here, electrical signals are often converted into chemical signals known as neurotransmitters. These neurotransmitters travel through the space that connects neurons, called the synaptic cleft. Upon arrival at the target neuron, the chemical signals are converted back into electrical signals, after which the targeted neuron itself may send out electrical signals to other connected neurons, and so on. Each neuron has communication points (called synapses) with many other neurons, and healthy neuronal communication within and between different brain regions is essential for the functioning of the brain.
Besides neurons, the brain consists of cell types including microglia, astrocytes and oligodendrocytes, which are collectively called glia. Glia – Greek for glue – were once believed to simply be the glue that holds the brain together. However, it is clear now that glia execute many more crucial functions in the brain. Microglia are known as the immune cells of the brain. Microglia scan the brain for signs of danger and damage, and become activated when they encounter pathogens or damaged brain cells. In this activated immune state they will try to remove these pathogens and damaged brain cells (by ‘eating’ them, a process called phagocytosis), to protect the brain and to promote repair of damaged brain tissue. In addition, microglia can modulate connections between neurons, by removing synapses. Astrocytes exert immune functions as well, and can also modulate synapses. Moreover, astrocytes provide nutrients to neurons, and protect neurons against overstimulation by taking up excess neurotransmitters from the synaptic cleft. Furthermore, astrocytes are an important component of the blood-brain barrier. The blood-brain barrier protects the brain from potential damaging factors and cells present in the blood by preventing their entry into the brain, while allowing oxygen and nutrients to pass into the brain. Also, astrocytes can influence the constriction and dilation of blood vessels, and thereby affect blood flow in the brain. Oligodendrocytes wrap the axons of neurons with a fatty substance called myelin, thereby insulating axons and improving fast conduction of electrical signals along axons. Due to the white appearance of myelin, brain regions that mainly contain myelinated axons are called white matter (as opposed to grey matter, which contains neuronal cell bodies).
Healthy neurons and glia, and healthy communication between them, are crucial to maintain brain health. Unfortunately, the functioning of neurons and glia can become disturbed, as evidenced by different brain diseases, such as Alzheimer’s disease (AD).
10
Alzheimer’s disease
AD is the most common cause of dementia, resulting in memory loss and problems with performing normal activities of daily living, which become increasingly worse with time [1,2]. In addition, AD is often accompanied by behavioral changes and neuropsychiatric problems, such as agitation, anxiety and depression [3].
AD accounts for 60-75% of all cases of dementia, which worldwide is estimated to affect 5-11% of people over the age of 60, and 20-50% of people over the age of 85 years [4– 14]. Since the risk of dementia increases with age, and life expectancy continues to increase globally, the number of dementia patients will only grow further in the coming decades. Currently, worldwide around 50 million people have dementia, and this number is expected to triple to around 150 million in 2050 [14]. AD (and other types of dementia) comes with a very high social burden for family and caregivers of AD patients, as well as with great economic costs. Many past and present investigations have/are focused on finding possible treatments for AD, in the hope to slow down, halt or even prevent this awful disease. Unfortunately, besides a few drugs that temporarily improve AD symptoms, no effective treatments have been identified so far. The fact that no effective treatments for AD are present yet, may depend for an important part on the incomplete understanding of AD.
What goes wrong in the Alzheimer brain?
AD pathology causes brain damage in the brains of AD patients. AD pathology starts in brain regions important for learning and memory, such as the brain region named the hippocampus. With time, AD pathology worsens and spreads out further throughout the brain, resulting in worsening of cognitive and behavioral problems over time. How does AD pathology start, and how does AD pathology cause brain damage? Despite years of intensive research, the pathological processes that underlie the brain damage in AD, and the causes that initiate these pathological processes, are not fully understood yet.
It is known that the AD brain is characterized by different pathological hallmarks. First of all, as was already reported in the first description of AD by Dr. Alois Alzheimer in 1906, the brain of AD patients is marked by abnormal accumulations of certain proteins, present between brain cells and within neurons [15]. The protein accumulations between brain cells were later found to consist of aggregated amyloid-β (Aβ) protein, and were named plaques. The protein accumulations within neurons were identified as aggregates of hyperphosphorylated tau protein, and were named tangles. Secondly, it has become clear that chronic inflammation in the brain (neuroinflammation) is an important hallmark of AD [16,17]. This chronic neuroinflammation is mediated by chronically activated microglia and astrocytes, and is apparent from for example the altered shape of these cells when they are activated as well as from increased levels of inflammatory factors, which are released (secreted) by these cells. Next to aggregation of Aβ and tau and chronic neuroinflammation, brain regions affected by AD are marked by multiple other pathological changes, including accumulation of iron, disturbed energy metabolism, dysregulated neurotransmitter
11
1
metabolism, disturbed blood flow, disruption of the blood-brain barrier, white matter damage, synaptic damage and neuronal loss [18–26].
All of these pathological processes may be important contributors to the brain damage and symptoms that arise in AD. For example, Aβ aggregates (especially smaller aggregates, named oligomers) are known to promote all other above-mentioned pathological processes [27,28]. In addition, it has become clear that also chronic neuroinflammation has widespread effects [16]. Namely, when microglia and astrocytes become chronically activated, they may lose some of their protective functions such as removing damaged or dead cells and Aβ aggregates. Instead, chronically activated microglia and astrocytes may promote brain damage. For example, chronically activated microglia and astrocytes can continuously secrete pro-inflammatory and other toxic factors, which may amongst others disturb iron and energy metabolism, and provoke synaptic damage and cell death [16,29–31]. Interestingly, many of the pathological processes in AD influence each other, and promote each other’s progression. For example, Aβ aggregation, iron accumulation and chronic neuroinflammation seem to stimulate each other’s continuation, thereby together driving the progression of AD pathology [16,24,32,33].
What are the causes of Alzheimer’s disease?
What initiates these pathological processes in AD, as described above? A small part (~1%) of AD cases is directly caused by genetic mutations [34]. In this familial form of AD, patients carry mutations in genes that are involved in the production of Aβ. This results in an overproduction of Aβ, and early onset of AD (before 65 years of age). However, for the vast majority (~99%) of AD cases, the underlying causes are far less clear [35,36]. This main type of AD is called sporadic AD, and usually starts later in life (after the age of 65). The major risk factor for sporadic AD is increasing age. In addition, certain genetic variants have been linked with an increased risk to develop sporadic AD [37]. Furthermore, risk factors for sporadic AD include for example: unhealthy lifestyle (e.g. smoking, physical inactivity and stress), various diseases (e.g. cardiovascular disease, diabetes, obesity and depression), certain injuries (e.g. surgery and traumatic brain injury), specific environmental factors (e.g. certain metals, air pollutants and pesticides), and infections (including certain bacterial, viral and fungal infections) [38–41]. Interestingly, many of the risk factors for sporadic AD, including aging, genetic variants, unhealthy lifestyle, injury and disease, have been linked with Aβ accumulation and chronic neuroinflammation, thereby supporting the importance of Aβ aggregation and chronic inflammation in the development and progression of AD [37,42–44]. Taken together, different risk factors for sporadic AD have been identified, and many pathological processes have been recognized to arise in the AD brain, which may all contribute to the brain damage in and symptoms of AD (also see Fig. 1). As such, it appears that sporadic AD is likely caused by a complex combination of aging and genetic, lifestyle and environmental factors, which may together provoke different AD-related pathological processes in the brain, such as Aβ aggregation and chronic neuroinflammation. However,
12
Fig. 1 Risk factors for sporadic AD, and pathological processes known to be involved in AD pathology, resulting
in AD neurodegeneration and dementia.
many questions regarding the causes and pathophysiology of sporadic AD remain, including the molecular mechanisms that are involved. For example, for many pathological processes in AD it is not fully understood yet via what exact molecular mechanisms they are initiated. Moreover, the mechanisms that underlie the toxic effects of these pathological processes are not completely clear yet. In order to develop effective treatments for AD, it is essential to gain more insight into the molecular mechanisms and factors that are involved in the development and progression of AD pathology. Since the importance of chronic (neuro)inflammation in AD and risk factors of AD is increasingly recognized, a better understanding of (neuro)inflammatory processes and involved inflammatory factors may be key to elucidate how AD starts and progresses, and how it may be stopped or prevented.
Recently, a protein called Lipocalin 2 (Lcn2) was identified as a potential important new player in AD. As described below, Lcn2 may significantly affect (neuro)inflammatory processes. Moreover, Lcn2 may affect several other processes, including iron and energy metabolism, and cell death/survival.
Lipocalin 2 (Lcn2)
Lcn2 is a member of the lipocalin family of proteins, which binds and transports small hydrophobic factors [45]. Lcn2 has been implicated to play a role in different physiological and pathophysiological processes. One of the most well-known functions of Lcn2 is its role as an acute-phase protein in the defense against certain bacteria [46,47]. Bacteria require iron for their growth, and secrete small iron-binding factors called siderophores to collect iron. Lcn2 however can interfere with this bacterial iron uptake, by binding to bacterial siderophores and thereby preventing their delivery to bacteria. As such, Lcn2 exerts important antibacterial effects by hijacking bacterial iron acquisition. Interestingly, more recently it has become clear that Lcn2 may also be involved in normal physiological iron metabolism, by binding mammalian siderophores [46,47]. Besides its functions in antibacterial defense and mammalian iron metabolism, Lcn2 is known to play a role in
13
1
various other processes, including inflammation, cell migration, energy metabolism and cell death/survival signaling (Fig. 2a) [46,47].
Under healthy conditions, the gene expression and protein levels of Lcn2 are low. Interestingly, Lcn2 levels in blood were found to gradually increase with advancing age, which may reflect the chronic low-grade inflammation that accompanies aging [48]. In addition, local and circulating protein levels of Lcn2 are greatly increased upon pathogenic threats and different kinds of injury, as well as in a wide range of diseases [47]. Lcn2 was found to significantly influence the severity, development and/or progression of several of these infections, injuries and diseases, such as chronic kidney disease and different types of cancer [45–47]. The role of Lcn2 in these conditions may rely for example on its involvement in inflammation, iron metabolism and cell death/survival signaling. Interestingly, contradicting effects have been reported for Lcn2, including e.g. anti- and pro-inflammatory, and anti- and pro-cell survival effects [46,47]. These contradictory effects of Lcn2 may depend on many factors, such as the precise disease condition and the cell types and tissues that are involved.
A role for Lcn2 in Alzheimer’s disease?
The role of Lcn2 in the healthy, injured and diseased brain has only more recently been explored. As in most tissues elsewhere in the body, the expression of Lcn2 in the brain is low under healthy conditions [47]. Also, Lcn2 expression in the brain was reported to gradually increase with advancing age in mice, which may correspond with the chronic low-grade (neuro)inflammatory state that arises in the body and brain with increasing age [49]. Besides in aging, elevated peripheral and brain levels of Lcn2 were also observed in other risk factor conditions of age-related brain diseases, such as obesity, cardiovascular disease and depression [50–53]. Moreover, brain Lcn2 levels were found to increase manifold upon inflammatory stimulation and brain injury, as well as in various brain conditions such as multiple sclerosis and stroke [47,54]. The effects of Lcn2 have been studied in several cell culture and animal models of these brain conditions. Although a few studies noted a beneficial effect of Lcn2, most studies found that increased Lcn2 levels significantly aggravated different neuropathological processes. For example, Lcn2 was reported to promote pro-inflammatory activity of microglia and astrocytes, to provoke iron accumulation in the brain, to aggravate disruption of the blood-brain barrier and white matter damage, and to stimulate neuronal cell death. Furthermore, Lcn2 has been suggested to be involved in disease-related behavioral and cognitive changes [47,54].
It may be hypothesized that Lcn2 may exert similar effects in the AD brain, and as such could contribute significantly to the development and progression of AD (Fig. 2b). A recent study has shown that Lcn2 protein levels are increased in the hippocampus, in human post-mortem brain tissue of AD patients [55]. Moreover, cell culture studies showed that Aβ induces Lcn2 production in cultured astrocytes, and that Lcn2 sensitizes neurons and astrocytes to Aβ-induced cell death [55,56]. However, most of these findings have not yet been replicated. Moreover, the effects of Lcn2 have not been investigated yet in vivo in animal models of AD. Therefore, further research is essential to gain more insight into the
14
Fig. 2 (a) Physiological processes in which Lcn2 is known to play a role, and (b) AD-related (patho)physiological
processes to which Lcn2 might contribute. Also see chapter 6 of this thesis and reviews [47,54] for further explanation.
potentially important role of Lcn2 in AD. More research is also needed to clarify whether Lcn2 may be a promising therapeutic target or diagnostic marker for AD.
Aims and outline of the thesis
In this thesis, we aim to gain a better understanding of the role of Lcn2 in AD. First, in
chapter 2, we studied Lcn2 protein levels in blood, cerebrospinal fluid and post-mortem brain
tissue of human AD patients and healthy age-matched persons. This study confirms that Lcn2 levels are significantly increased in the AD brain as compared to the healthy aged brain, in multiple brain regions. Furthermore, we found that co-existing depression in AD was related to significantly altered Lcn2 protein levels in different brain regions. After confirming increased Lcn2 levels in the brains of human AD patients, we in chapter 3 studied the role of Lcn2 in the J20 mouse model of AD. J20 mice overexpress two mutated human genes that in humans are known to cause familial AD. Indeed, J20 mice develop AD-like pathological characteristics, including Aβ plaque pathology, neuroinflammation and cognitive impairment. As expected, we found that Lcn2 protein levels are increased in the brain of J20 AD mice, as compared to normal (wildtype) mice. We then compared J20 mice with J20 mice that are deficient in Lcn2. We found that J20 and Lcn2-deficient J20 mice show equally severe memory impairment, Aβ plaque load, and activation of microglia and astrocytes. Interestingly, Lcn2-deficient J20 mice showed less severe brain iron accumulation, as compared to J20 mice. In chapter 4 we aimed to gain a better understanding of the regulation of Lcn2 production, and to identify potential inhibitors of Lcn2 overproduction, in cultured mouse astrocytes. We confirmed that Aβ induces a strong increase in Lcn2 protein production in astrocytes. Moreover, we show that iron chelators can inhibit this Aβ-induced Lcn2 production, and that Aβ may directly disturb iron metabolism in cultured astrocytes. In
chapter 5, we explored the potential suitability of Lcn2 as a biomarker in age-related
diseases, by assessing whether Lcn2 protein levels in blood and cerebrospinal fluid remain stable throughout the day in healthy elderly people. Indeed, Lcn2 levels did not fluctuate significantly during the day, which is a favorable characteristic for a biomarker. In chapter 6 we review the current knowledge regarding the potential role of Lcn2 in age-related brain
15
1
diseases including AD, Parkinson’s disease (PD) and vascular dementia (VaD). In addition, we discuss the role of Lcn2 in risk factor conditions for these disorders, and explore the possibility that Lcn2 is an inflammatory link between risk factor conditions and age-related brain diseases. Finally, in chapter 7 we end with a general discussion of this thesis, overviewing the strengths and limitations of the thesis, the main findings and implications that arise from it, and suggestions for future research.
Of note: whereas Lcn2 describes the mouse form, the human form of this protein is usually referred to as neutrophil gelatinase-associated lipocalin (NGAL). For consistency, we will use the term Lcn2 throughout this thesis, except for the chapters in which specifically the human protein (NGAL) was studied (Chapters 2 & 5).
16
References
[1] Tarawneh R, Holtzman DM. The Clinical Problem of Symptomatic Alzheimer Disease and Mild Cognitive Impairment. Cold Spring Harb Perspect Med [Internet]. 2012 [cited 2019 Jan 4];2.
[2] Prizer LP, Zimmerman S. Progressive Support for Activities of Daily Living for Persons Living With Dementia. The Gerontologist. 2018;58:S74–87.
[3] Cerejeira J, Lagarto L, Mukaetova-Ladinska EB. Behavioral and Psychological Symptoms of Dementia. Front Neurol [Internet]. 2012 [cited 2019 Jan 4];3.
[4] Alzheimer’s Association. What Is Alzheimer’s? [Internet]. Alzheimers Assoc. [cited 2018 Aug 2]. Available from: https://alz.org/alzheimers-dementia/what-is-alzheimers
[5] Brookmeyer R, Evans DA, Hebert L, Langa KM, Heeringa SG, Plassman BL, et al. National estimates of the prevalence of Alzheimer’s disease in the United States*. Alzheimers Dement J Alzheimers Assoc. 2011;7:61–73.
[6] Ferri CP, Prince M, Brayne C, Brodaty H, Fratiglioni L, Ganguli M, et al. Global prevalence of dementia: a Delphi consensus study. Lancet Lond Engl. 2005;366:2112–7.
[7] Gardner RC, Valcour V, Yaffe K. Dementia in the oldest old: a multi-factorial and growing public health issue. Alzheimers Res Ther. 2013;5:27.
[8] Langa KM, Larson EB, Crimmins EM, Faul JD, Levine DA, Kabeto MU, et al. A Comparison of the Prevalence of Dementia in the United States in 2000 and 2012. JAMA Intern Med. 2017;177:51–8.
[9] Plassman BL, Langa KM, Fisher GG, Heeringa SG, Weir DR, Ofstedal MB, et al. Prevalence of Dementia in the United States: The Aging, Demographics, and Memory Study. Neuroepidemiology. 2007;29:125–32.
[10] Prince M, Knapp M, Guerchet M, McCrone P, Prina M, Comas-Herrera A, et al. Dementia UK: Update. Alzheimers Soc. :137.
[11] Prince M, Bryce R, Albanese E, Wimo A, Ribeiro W, Ferri CP. The global prevalence of dementia: A systematic review and metaanalysis. Alzheimers Dement J Alzheimers Assoc. 2013;9:63-75.e2.
[12] Prince MJ, Wimo A, Guerchet M, Ali G-C, Wu Y-T, Prina M, et al. World Alzheimer Report 2015: The Global Impact of Dementia [Internet]. 2015 [cited 2018 Aug 2]. Available from: https://www.alz.co.uk/research/world-report-2015 [13] Qiu C, Kivipelto M, von Strauss E. Epidemiology of Alzheimer’s disease: occurrence, determinants, and strategies
toward intervention. Dialogues Clin Neurosci. 2009;11:111–28.
[14] World Health Organization. Dementia [Internet]. World Health Organ. [cited 2018 Aug 2]. Available from: http://www.who.int/news-room/fact-sheets/detail/dementia
[15] Selkoe DJ. Alzheimer’s disease. Cold Spring Harb Perspect Biol. 2011;3.
[16] Heneka MT, Carson MJ, El Khoury J, Landreth GE, Brosseron F, Feinstein DL, et al. Neuroinflammation in Alzheimer’s Disease. Lancet Neurol. 2015;14:388–405.
[17] Millington C, Sonego S, Karunaweera N, Rangel A, Aldrich-Wright JR, Campbell IL, et al. Chronic Neuroinflammation in Alzheimer’s Disease: New Perspectives on Animal Models and Promising Candidate Drugs. BioMed Res Int [Internet]. 2014 [cited 2019 Jan 4];2014.
[18] Crews L, Masliah E. Molecular mechanisms of neurodegeneration in Alzheimer’s disease. Hum Mol Genet. 2010;19:R12–20.
[19] Sweeney MD, Sagare AP, Zlokovic BV. Blood–brain barrier breakdown in Alzheimer’s disease and other neurodegenerative disorders. Nat Rev Neurol. 2018;14:133–50.
[20] Nasrabady SE, Rizvi B, Goldman JE, Brickman AM. White matter changes in Alzheimer’s disease: a focus on myelin and oligodendrocytes. Acta Neuropathol Commun [Internet]. 2018 [cited 2019 Jan 4];6.
[21] Kisler K, Nelson AR, Montagne A, Zlokovic BV. Cerebral blood flow regulation and neurovascular dysfunction in Alzheimer’s disease. Nat Rev Neurosci. 2017;18:419–34.
[22] Kelly SC, He B, Perez SE, Ginsberg SD, Mufson EJ, Counts SE. Locus coeruleus cellular and molecular pathology during the progression of Alzheimer’s disease. Acta Neuropathol Commun [Internet]. 2017 [cited 2019 Jan 4];5.
[23] Francis PT. The interplay of neurotransmitters in Alzheimer’s disease. CNS Spectr. 2005;10:6–9.
[24] Belaidi AA, Bush AI. Iron neurochemistry in Alzheimer’s disease and Parkinson’s disease: targets for therapeutics. J Neurochem. 2016;139 Suppl 1:179–97.
[25] de la Monte SM. Brain Insulin Resistance and Deficiency as Therapeutic Targets in Alzheimer’s Disease. Curr Alzheimer Res. 2012;9:35–66.
[26] Berger AL. Insulin resistance and reduced brain glucose metabolism in the aetiology of Alzheimer’s disease. J Insul Resist. 2016;1:7.
[27] Sengupta U, Nilson AN, Kayed R. The Role of Amyloid-β Oligomers in Toxicity, Propagation, and Immunotherapy. EBioMedicine. 2016;6:42–9.
[28] Carrillo-Mora P, Luna R, Colín-Barenque L. Amyloid Beta: Multiple Mechanisms of Toxicity and Only Some Protective Effects? Oxid Med Cell Longev [Internet]. 2014 [cited 2019 Jan 5];2014.
[29] Yin F, Sancheti H, Patil I, Cadenas E. Energy Metabolism and Inflammation in Brain Aging and Alzheimer’s Disease. Free Radic Biol Med. 2016;100:108–22.
[30] Mottahedin A, Ardalan M, Chumak T, Riebe I, Ek J, Mallard C. Effect of Neuroinflammation on Synaptic Organization and Function in the Developing Brain: Implications for Neurodevelopmental and Neurodegenerative Disorders. Front Cell Neurosci [Internet]. 2017 [cited 2019 Jan 5];11.
17
1
[31] Urrutia P, Aguirre P, Esparza A, Tapia V, Mena NP, Arredondo M, et al. Inflammation alters the expression of DMT1, FPN1 and hepcidin, and it causes iron accumulation in central nervous system cells. J Neurochem. 2013;126:541–9. [32] Ong W-Y, Farooqui AA. Iron, neuroinflammation, and Alzheimer’s disease. J Alzheimers Dis JAD. 2005;8:183–200;
discussion 209-215.
[33] Domingues C, Cruz e Silva OAB, Henriques AG. Impact of Cytokines and Chemokines on Alzheimer’s Disease Neuro-pathological Hallmarks. Curr Alzheimer Res. 2017;14:870–82.
[34] Ryan NS, Nicholas JM, Weston PSJ, Liang Y, Lashley T, Guerreiro R, et al. Clinical phenotype and genetic associations in autosomal dominant familial Alzheimer’s disease: a case series. Lancet Neurol. 2016;15:1326–35.
[35] Iqbal K, Grundke-Iqbal I. Alzheimer Disease, a Multifactorial Disorder Seeking Multi-therapies. Alzheimers Dement J Alzheimers Assoc. 2010;6:420–4.
[36] Piaceri I, Nacmias B, Sorbi S. Genetics of familial and sporadic Alzheimer’s disease. Front Biosci Elite Ed. 2013;5:167–77. [37] Van Cauwenberghe C, Van Broeckhoven C, Sleegers K. The genetic landscape of Alzheimer disease: clinical implications
and perspectives. Genet Med. 2016;18:421–30.
[38] Livingston G, Sommerlad A, Orgeta V, Costafreda SG, Huntley J, Ames D, et al. Dementia prevention, intervention, and care. Lancet Lond Engl. 2017;390:2673–734.
[39] Hickman R, Faustin A, Wisniewski T. Alzheimer Disease and its Growing Epidemic: Risk Factors, Biomarkers and the Urgent Need for Therapeutics. Neurol Clin. 2016;34:941–53.
[40] Campdelacreu J. Parkinson’s disease and Alzheimer disease: environmental risk factors. Neurol Engl Ed. 2014;29:541–9. [41] Sochocka M, Zwolińska K, Leszek J. The Infectious Etiology of Alzheimer’s Disease. Curr Neuropharmacol. 2017;15:996–
1009.
[42] Eikelenboom P, Hoozemans JJ, Veerhuis R, van Exel E, Rozemuller AJ, van Gool WA. Whether, when and how chronic inflammation increases the risk of developing late-onset Alzheimer’s disease. Alzheimers Res Ther. 2012;4:15.
[43] McKenzie JA, Spielman LJ, Pointer CB, Lowry JR, Bajwa E, Lee CW, et al. Neuroinflammation as a Common Mechanism Associated with the Modifiable Risk Factors for Alzheimer’s and Parkinson’s Diseases. Curr Aging Sci. 2017;10:158–76. [44] Vijaya Kumar DK, Choi SH, Washicosky KJ, Eimer WA, Tucker S, Ghofrani J, et al. Amyloid-β Peptide Protects Against
Microbial Infection In Mouse and Worm Models of Alzheimer’s Disease. Sci Transl Med. 2016;8:340ra72.
[45] Chakraborty S, Kaur S, Guha S, Batra SK. The multifaceted roles of neutrophil gelatinase associated lipocalin (NGAL) in inflammation and cancer. Biochim Biophys Acta. 2012;1826:129–69.
[46] Xiao X, Yeoh BS, Vijay-Kumar M. Lipocalin 2: An Emerging Player in Iron Homeostasis and Inflammation. Annu Rev Nutr. 2017;37:103–30.
[47] Ferreira AC, Dá Mesquita S, Sousa JC, Correia-Neves M, Sousa N, Palha JA, et al. From the periphery to the brain: Lipocalin-2, a friend or foe? Prog Neurobiol. 2015;131:120–36.
[48] Naudé PJW, Eisel ULM, Comijs HC, Groenewold NA, De Deyn PP, Bosker FJ, et al. Neutrophil gelatinase-associated lipocalin: a novel inflammatory marker associated with late-life depression. J Psychosom Res. 2013;75:444–50. [49] Sharman EH, Bondy SC, Sharman KG, Lahiri D, Cotman CW, Perreau VM. Effects of melatonin and age on gene
expression in mouse CNS using microarray analysis. Neurochem Int. 2007;50:336–44.
[50] Wang Y, Lam KSL, Kraegen EW, Sweeney G, Zhang J, Tso AWK, et al. Lipocalin-2 is an inflammatory marker closely associated with obesity, insulin resistance, and hyperglycemia in humans. Clin Chem. 2007;53:34–41.
[51] Yan Q-W, Yang Q, Mody N, Graham TE, Hsu C-H, Xu Z, et al. The adipokine lipocalin 2 is regulated by obesity and promotes insulin resistance. Diabetes. 2007;56:2533–40.
[52] Marques FZ, Prestes PR, Byars SG, Ritchie SC, Würtz P, Patel SK, et al. Experimental and Human Evidence for Lipocalin-2 (Neutrophil Gelatinase-Associated Lipocalin [NGAL]) in the Development of Cardiac Hypertrophy and heart failure. J Am Heart Assoc. 2017;6.
[53] Gouweleeuw L, Naudé PJW, Rots M, DeJongste MJL, Eisel ULM, Schoemaker RG. The role of neutrophil gelatinase associated lipocalin (NGAL) as biological constituent linking depression and cardiovascular disease. Brain Behav Immun. 2015;46:23–32.
[54] Jha MK, Lee S, Park DH, Kook H, Park K-G, Lee I-K, et al. Diverse functional roles of lipocalin-2 in the central nervous system. Neurosci Biobehav Rev. 2015;49:135–56.
[55] Naudé PJW, Nyakas C, Eiden LE, Ait-Ali D, van der Heide R, Engelborghs S, et al. Lipocalin 2: novel component of proinflammatory signaling in Alzheimer’s disease. FASEB J Off Publ Fed Am Soc Exp Biol. 2012;26:2811–23.
[56] Mesquita SD, Ferreira AC, Falcao AM, Sousa JC, Oliveira TG, Correia-Neves M, et al. Lipocalin 2 modulates the cellular response to amyloid beta. Cell Death Differ. 2014;21:1588–99.
Chapter 2
Neutrophil gelatinase-associated lipocalin (NGAL) and its
receptors in Alzheimer (AD) brain regions: differential
findings in AD with and without depression
Doortje W. Dekensa,b, Petrus J.W. Naudéa,b, Sebastiaan Engelborghsd,e, Yannick Vermeirena,e, Debby Van Dama,e, Richard C. Oude Voshaarc, Ulrich L.M. Eiselb,c, Peter P. De Deyna,d,e
a
Department of Neurology and Alzheimer Research Center, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands.
bDepartment of Molecular Neurobiology, University of Groningen, Groningen, the Netherlands.
c
University Center of Psychiatry & Interdisciplinary Center of Psychopathology of Emotion Regulation, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands.
d
Department of Neurology and Memory Clinic, Hospital Network Antwerp (ZNA), Antwerp, Belgium.
e
Laboratory of Neurochemistry and Behavior, Biobank, Institute Born-Bunge, University of Antwerp, Antwerp, Belgium.
20
Abstract
Co-existing depression worsens Alzheimer’s disease (AD) pathology. Neutrophil gelatinase-associated lipocalin (NGAL) is a newly identified (neuro)inflammatory mediator in the pathophysiologies of both AD and depression. This study aimed to compare NGAL levels in healthy controls, AD without depression (AD–D) and AD with co-existing depression (AD+D) patients. Protein levels of NGAL and its receptors; 24p3R and megalin, were assessed in nine brain regions from healthy controls (n=19), AD–D (n=19) and AD+D (n=21) patients. NGAL levels in AD–D patients were significantly increased in brain regions commonly associated with AD. In the hippocampus, NGAL levels were even further increased in AD+D subjects. Unexpectedly, NGAL levels in the prefrontal cortex of AD+D patients were comparable to those in controls. Megalin levels were increased in BA11 and amygdala of AD+D patients, while no changes in 24p3R were detected. These findings indicate significant differences in neuroimmunological regulation between AD patients with and without co-existing depression. Considering its known effects, elevated NGAL levels might actively promote neuropathological processes in AD with and without depression.
21
2
Introduction
Alzheimer’s disease (AD) is a multifactorial neurodegenerative disorder [1]. Apart from the classical amyloid beta (Aβ) cascade hypothesis and tau hypothesis of AD, mounting evidence has led to the understanding that more mechanisms are involved in the etiology of AD [2–4]. For example, another important factor in the pathophysiology of AD is neuroinflammation, characterized by pro-inflammatory processes in the brain [5–7]. Increased pro-inflammatory processes have been proposed as a biological constituent linking AD and depression [8,9]. Depression is a neuropsychiatric symptom that is frequently present in AD [10] and associated with an increased risk for the progression of mild cognitive impairment (MCI) to AD [11–14]. Furthermore, AD patients with co-existing depression (AD+D) have a poorer prognosis than AD patients without depression (AD–D) [15–17]. Interestingly, AD+D patients present aggravated Aβ and tau pathology in the hippocampus, as compared to AD–D patients [17,18]. Several studies have demonstrated that a low grade inflammatory environment is present in the brain and blood in depressed [19–21] and AD patients [5–7]. However, immune regulation in AD patients with co-existing symptoms of depression is poorly understood and existing studies are limited.
Recently a pro-inflammatory mediator called neutrophil gelatinase-associated lipocalin (NGAL) was identified to be involved in the pathophysiology of both AD and depression [22,23]. NGAL is also known as lipocalin 2, 24p3, uterocalin and siderocalin and belongs to the lipocalin family of small hydrophobic proteins [24–27]. NGAL acts via its two known receptors, megalin (also known as Lrp2) and 24p3R (also known as SLC22A17 and BOCT1) [28,29]. NGAL functions as an acute phase protein in the innate immune response and plays an important role in the defense against certain bacteria [30]. It exerts various important functions in the central nervous system, depending on the cellular environment, as recently reviewed by Jha and colleagues [31]. NGAL mRNA and protein are expressed in very low levels in the brain under normal physiological conditions [32,33]. NGAL protein levels are robustly increased in human AD post-mortem brain regions affected by AD pathology, particularly the hippocampus [23]. Furthermore, increased plasma NGAL levels were found in people with MCI [34]. In the cerebrospinal fluid (CSF), NGAL levels were significantly decreased in AD patients as well as in MCI patients that later converted to AD, which mimicks the characteristically lowered CSF Aβ levels in AD [23,35]. We recently showed that plasma NGAL levels are increased in depressed older persons [22]. In addition, plasma NGAL levels are significantly increased in depressed elderly females with impaired recall memory [36].
Cell culture experiments showed that NGAL is increased in different brain cell types upon exposure to Aβ and stimulation of tumor necrosis factor-α receptor 1 (TNFR1), and sensitizes primary cortical neurons and astrocytes to Aβ- and oxidative stress-induced cell death [23,37,38]. These effects may in part be caused by silencing of neuroprotective tumor necrosis factor-α receptor 2 (TNFR2) signaling [23]. NGAL is significantly increased in the hippocampus of mice following psychological stress, where it reduces dendritic spine formation of hippocampal neurons, potentially by inhibiting Akt phosphorylation of the
22
protein kinase B (PKB)/Akt pathway [23,33]. Furthermore, increased NGAL aggravates neuroinflammation upon neuronal damage and promotes the pro-inflammatory phenotypes of astrocytes and microglia [39–41].
The above-mentioned studies indicate that NGAL functions as a neuroinflammatory constituent that is involved in the pathophysiology of both depression and AD, and might exacerbate AD pathology in the presence of co-existing depression. Therefore our objectives were (1) to investigate NGAL protein levels in serum, CSF and human post-mortem brain tissue (nine different brain regions) of control subjects, AD–D and AD+D patients, and (2) to assess whether changes in brain NGAL levels correlate with alterations in the protein expression of its receptors, megalin and 24p3R. We hypothesize that NGAL levels are increased in AD-related brain regions in AD–D patients as compared to healthy age-matched controls, and further increased in AD+D patients in certain brain regions that are involved in both depression and AD.
Materials and methods
Human samples
Human tissue samples (serum, CSF and brain tissue) from healthy controls, AD–D and AD+D patients were obtained from the Biobank of the Institute Born-Bunge (University of Antwerp, Belgium) and stored at -80 °C until analysis. Informed consent for behavioral assessments and tissue collection was provided by all participants. Ethical approval for human sample collection was granted by the Medical Ethical Committee of the Middelheim General Hospital (Antwerp, Belgium) (approval numbers 2805 and 2806). Participants who were known to suffer from diseases that could influence NGAL levels, such as renal failure and cancer, were excluded. In this respect, C-reactive protein (CRP) levels were measured in serum by ELISA analysis (Abnova, KA0238, performed according to the manufacturer’s protocol) (Table 1). Serum CRP levels in control patients on average were below 10 mg/L, which is within normal range for healthy adults, indicating that active inflammation was not present in these participants [42]. Analyses of variance showed that serum CRP levels in AD patients were not significantly different from those in control patients (F=2.30, df=2, p=0.11). Further details and inclusion criteria are described by Vermeiren et al. [43,44].
Brain tissue samples were collected from nine different brain regions, including the hippocampus, amygdala, thalamus, cerebellum, Brodmann area (BA) 9 (BA9; dorsolateral prefrontal cortex), BA10 (frontopolar cortex), BA11 (orbitofrontal cortex), BA22 (superior temporal cortex) and BA24 (ventral/dorsal anterior cingulate cortex (ACC)). Autopsy and collection procedures were performed as described previously [43,44]. Briefly, brain autopsy was performed within 6 hours after death of consented patients, upon which the left hemisphere was frozen at −80 °C for neurochemical analyses, and the right hemisphere was formaldehyde-fixed for neuropathological examination. Neuropathological examination was performed as previously described [44] using the criteria of amongst others Braak and Braak (1991) and Montine et al. (2012) [2,45]. Absence of AD pathology was confirmed in all
23
2
control brains used in this study, and definite AD diagnosis was confirmed in all AD brains (see Table 1 for Braak stages). No significant vascular pathology was detected in the brains. The frozen left hemisphere was routinely dissected into 21 brain regions, following a standard procedure [44]. The inclusion of the nine brain regions selected for this study was based on their involvement in both depression and AD.
Paired AD CSF and serum were not from the same AD patients as those from whom brain tissue was obtained. CSF and serum were from patients that had a diagnosis ranging from probable to definite AD. Numbers and demographic details of participants are listed in
Table 1.
Table 1 Patient demographics, behavioral and cognitive symptoms, and use of anti-depressant medication
(administered no more than one day before death).
Parameter
Serum Control (n=40) AD – D (n=40) AD + D (n=40)
Gender, female, n (%) 20 (50) 20 (50) 20 (50)
Age (years), mean (SD) 78.33 (7.19) 79.50 (9.20) 79.35 (8.69)
CSDD scale, mean (SD) 2.83 (2.48) 3.80 (2.15) 11.05 (2.57) MMSE, mean (SD) 28.00 (1.62) 14.97 (6.54) 12.54 (6.86) Anti-depressant medication, n (%) 5 (12.5) 8 (20) 16 (40) CRP (mg/L), mean (SD) 9.43 (19.24) 20.91 (32.38) 22.78 (36.02) CSF Control (n=26) AD – D (n=40) AD + D (n=40) Gender, female, n (%) 13 (50) 20 (50) 20 (50)
Age (years), mean (SD) 78.90 (5.74) 79.50 (9.20) 79.35 (8.69)
CSDD, mean (SD) N.D. 3.80 (2.15) 11.05 (2.57)
MMSE, mean (SD) N.D. 14,97 (6.54) 12.54 (6.86)
Anti-depressant medication, n (%) N.D. 8 (20) 16 (40)
Brain tissue Control (n=19) AD – D (n=19) AD + D (n=21)
Gender, female, n (%) 7 (36.84) 7 (36.84) 8 (38.10)
Age (years), mean (SD) 73.04 (9.24) 83.25 (9.09) 77.74 (10.18)
Braak stage, mean (SD) 0.06 (0.25) 4.15 (1.02) 4.58 (1.12)
CSDD, mean (SD) N.D. 4.74 (1.88) 13.33 (4.52)
Anti-depressant medication, n (%) 4 (21.05) 3 (15.79) 10 (47.62)
AD–D: Alzheimer’s disease without co-existing depression, AD+D: Alzheimer’s disease with co-existing depression, n: number of patients, SD: standard deviation, N.D.: not determined, CSDD: Cornell Scale for Depression in Dementia, MMSE: mini mental state examination.
Neuropsychiatric evaluation
For all patients, the Cornell Scale for Depression in Dementia (CSDD) was used to distinguish between AD–D and AD+D (see also Table 1). The CSDD scale examines signs of major depression in demented patients [46]. The presence of significant depressive symptoms is suggested by a total score of 8 or more, while scores above 18 are indicative of major depression [14]. For the current study AD patients with a CSDD score of ≥ 8 were classified as AD+D patients, whereas patients with a CSDD score of < 8 were defined as non-depressed, AD–D patients.
24
Mini-Mental State Examination (MMSE) scores were determined in all subjects from whom serum was obtained, and in AD patients from whom CSF was collected. In addition, the Middelheim Frontality Score (MFS) scale was included as a measure of frontal lobe features [47].
For the control patients from whom serum was obtained, the time between behavioral scoring and sampling of serum was on average 0.7 days (SD; 1.83 days), and the time between MMSE scoring and sampling was 4.25 days (SD; 11.65 days). From AD patients, serum and CSF were collected on average 0.36 days (SD; 1.41 days) apart from behavioral rating. For AD patients from whom brain tissue was obtained, the average time from baseline or follow-up (n=9) behavioral scoring until time of death was 6.23 days (SD; 11.49). No behavioral data was obtained from the control subjects from whom brain tissue was received.
Measurement of NGAL with ELISA
NGAL in serum and CSF samples was quantified via a sandwich ELISA (R&D Systems) according to the manufacturer’s protocol, as described previously [22,23]. Serum was diluted 1:100, and CSF samples were not diluted. The intra- and inter-assay coefficients of variation were 3% and 5%, respectively.
Western blot analyses of NGAL, megalin and 24p3R levels in brain tissue
Brain NGAL, megalin and 24p3R levels were determined by Western blot as previously described [23]. An NGAL antibody specifically directed against human NGAL was used at a concentration of 1:600 (rat anti-human LCN2, MAB17571, R&D systems). Both megalin (rabbit anti-megalin, Ab129198, Abcam) and 24p3R (rabbit anti-24p3R, Ab124506, Abcam) antibodies were used at 1:1000 dilution. Actin served as internal standard to control for loading variations and was used at 1:1000 000 dilution (mouse anti-actin, 691002, ImmunO, MP Biomedicals Inc.). Appropriate HRP-conjugated secondary antibodies were used at a 1:5000 dilution (donkey anti-rabbit, NA934, GE Healthcare; goat anti-mouse, Sc-2005, SantaCruz Biotechnology; goat anti-rat, 112-035-003, Jackson ImmunoResearch Laboratories Inc.). Examples of full Western blots for NGAL, 24p3R and megalin are shown in Suppl. Fig. 3.Immunohistochemistry for fluorescent double stainings
Paraformaldehyde-fixed hippocampal brain tissue from AD patients was used for immunofluorescent double stainings. For double stainings of NGAL with ionized calcium-binding adapter molecule 1 (Iba1) and glial fibrillary acidic protein (GFAP), brain sections were cut from paraffin embedded tissue (10 µm thick). These sections were deparaffinized before start of the staining protocol. For double staining of NGAL with neuronal nuclei (NeuN), brain sections were obtained from frozen, non-paraffin embedded tissue (16 µm thick). Sections were subjected to heat-induced antigen retrieval with 10 mM sodium citrate buffer (pH 6.0, no Tween) in the microwave. Tissue was then blocked with 10% normal serum
25
2
in PBS+0.3% Triton X-100 for 1h at room temperature, and incubated with the primary antibody overnight at 4 °C. Primary antibodies were diluted in PBS+0.3% Triton X-100, with 1% normal serum. The following primary antibodies were used: anti-NGAL (R&D systems, MAB1757, 1:50 dilution), anti-Iba1 (Wako, 019-19741, 1:1000 dilution), anti-GFAP (Sigma g3839, 1:1500 dilution), and anti-NeuN (Abcam, Ab177487, 1:300 dilution). Subsequently, sections were incubated with secondary antibodies for 1h at RT. Secondary antibodies were labelled with either Alexa Fluor 488 (Abcam Ab150153 or Molecular Probes A-21202, used at 1:400 dilution) or Cy3 (Jackson ImmunoResearch 712-165-150 or 711-165-152, used at 1:700 dilution). Stainings were mounted with Mowiol, and imaged with the Leica TCS SP8 microscope.
Statistical analysis
Analysis of variance (ANOVA) with Tukey post hoc test for pair-wise comparisons was used to determine differences in NGAL protein levels between healthy controls, AD–D and AD+D patients. This was followed by analyses of covariance (ANCOVA) with Bonferroni post hoc test with CSF or brain NGAL levels as dependent variable to analyze NGAL levels between the studied groups, adjusted for age and gender as confounding factors for CSF NGAL levels, and age, gender and use of anti-depressants as confounding factors for brain NGAL levels. Due to the limited number of brain tissue samples, bootstrapping was used with 1000 bootstraps per sample for analyses of brain tissues. Results were considered statistically significant when p-values were <0.05. Pearson correlation was used to examine the association between hippocampal NGAL levels and the severity of depression, and associations between NGAL, megalin and 24p3R in the studied brain regions. A corrected p-value of less than 0.006 was considered significant for multiple comparisons between NGAL, megalin and 24p3R in the nine studied brain regions. As the correlations between the Middelheim frontality score with NGAL in the nine brain regions were exploratory, p-values were not corrected and p-values less than 0.05 were considered significant. All analyses were conducted with SPSS version 22.0.
Results
NGAL levels in different brain regions of control, AD–D and AD+D patients
Mean NGAL/actin ratios in all nine investigated brain regions are displayed in Fig. 1. The data shown is not yet adjusted for age, sex and use of anti-depressants, which are potential confounding factors. NGAL was ubiquitously expressed throughout the healthy human brain. NGAL levels in the hippocampus were higher in AD–D than in healthy controls, and even higher in AD+D (Fig. 1A). After adjustment for confounders (age, sex and use of anti-depressants), the increased NGAL levels as compared to controls remained significant for AD–D (p=0.017) and AD+D (p=0.003), but the difference between the AD–D and AD+D group was no longer significant (p=0.112).26
Fig. 1 NGAL levels in nine different brain regions of control, AD–D and AD+D patients, as assessed by Western
blot. Brain regions included (A) hippocampus, (B) amygdala, (C) thalamus, (D) BA9, (E) BA10, (F) BA11, (G) BA24, (H) BA22 and (I) cerebellum. Bars indicate the mean ratio between NGAL and the internal control protein actin, ± SEM. AD–D: Alzheimer’s disease without co-existing depression, AD+D: Alzheimer’s disease with co-existing depression, BA: Brodmann area.
27
2
As depicted in Fig. 2, hippocampal NGAL levels were found to be positively correlated with the severity of depression (as measured by the CSDD depression score) in all of the AD patients (r=0.359, p=0.027). The amygdala contains higher NGAL levels in both AD–D and AD+D patients as compared to controls (Fig. 1B); these differences remained significant (AD– D, p=0.027; AD+D, p=0.04 compared to control) after adjusting for confounding factors. NGAL in the thalamus showed a similar expression pattern to that found in the amygdala (Fig. 1C). After adjustments for confounding factors, NGAL remained significantly higher in AD–D (p=0.002) compared to controls. Furthermore, a distinct pattern of NGAL levels was observed for BA9, BA10, BA11 and BA24: NGAL levels were significantly increased in AD–D patients compared to controls, but were not different between controls and AD+D patients and even significantly lower in AD+D as compared to AD–D (Fig. 1D, E, F, G). For BA11, these differences were, however, not significant, as shown in Fig. 1F. Moreover, after adjusting for confounding factors, the differences in NGAL levels between control and AD–D were as follows: BA9 (p=0.029), BA10 (p=0.073), BA11 (p=0.098) and BA24 (p=0.05). Similarly, when comparing AD–D and AD+D participants, adjustment for confounding factors gave the following results: BA9 (p=0.029), BA10 (p=0.040), BA11 (p=0.112) and BA24 (p=0.069). In BA22 and the cerebellum, no differences between the three groups were found (Fig. 1H and
I).
Fig. 2 Pearson correlation between hippocampal NGAL levels and CSDD depression score, in all AD patients.
r=.359 and p=.027.
To assess to which brain cell types NGAL is localized in the AD hippocampus, double stainings between NGAL and markers for microglia (Iba1), astrocytes (GFAP) and neurons (NeuN) were performed (Fig. 3). Immunofluorescent stainings showed that there is little co-localization for NGAL with microglia (Fig. 3A-C) and neurons (Fig. 3G-I). NGAL did co-localize with GFAP-positive astrocytes (Fig. 3D-F).
28
Fig. 3 Double stainings for NGAL with Iba1 (A-C), GFAP (D-F) and NeuN (G-I), in AD hippocampal human brain
sections. Scale bar: 25 µm.
NGAL levels in serum and CSF
Serum NGAL levels were not significantly different between the study groups (ANOVA, F=0.26, df=2, p=0.77) (Fig. 4A). There were, however, differences between the studied groups in CSF NGAL levels (ANOVA, F=9.52, df=2, p<.001). NGAL levels were significantly lower in AD–D (p<0.001) and AD+D (p=0.001) CSF compared to control CSF (Fig. 4B). Further, analyses with ANCOVA (F=9.30, df=2, p<0.001) and Bonferroni post hoc test showed that CSF NGAL levels remained significantly lower in AD–D (p=0.001) and AD+D (p=0.001) after including age and gender as covariates.
Megalin and 24p3R levels in different brain regions
Megalin levels were significantly higher in the amygdala of AD+D patients as compared to controls (Fig. 5A); this remained significant (p=0.022) after adjusting for confounding factors (age, gender, and use of antidepressant medication). A similar pattern was found in BA11
29
2
(Fig. 5B), but this difference lost significance after adjusting for covariates (p=0.059). No differences in megalin levels between the three groups were found in the other brain regions (Suppl. Fig. 1). Moreover, for 24p3R, no differences between control, AD–D and AD+D were found for any of the studied brain regions (Suppl. Fig. 2).
Fig. 4 NGAL levels in human (A) serum and (B) CSF of control, AD–D and AD+D patients, as measured by ELISA.
Bars indicate mean protein concentration in ng/ml (for serum) and pg/ml (for CSF). Error bars represent SEM. **p=0.001, ***p<0.001. AD–D: Alzheimer’s disease without co-existing depression, AD+D: Alzheimer’s disease with co-existing depression.
Fig. 5 Megalin levels in (A) BA11 and the (B) amygdala of control, AD–D and AD+D patients. Bars indicate the
mean megalin/actin ratio, ± SEM. AD–D: Alzheimer’s disease without co-existing depression, AD+D: Alzheimer’s disease with co-existing depression, BA: Brodmann area.
30
Correlation between brain NGAL, megalin and 24p3R in different brain regions
No significant correlations were found between NGAL and 24p3R or megalin levels. Megalin and 24p3R levels were positively correlated in the hippocampus (r=.510, p<.001) (Table 2).Table 2 Correlations between brain levels of NGAL, megalin and 24p3R.
BA 9 BA 10 BA 11 BA 22 BA 24 Cerebellum Amygdala Hippocampus Thalamus
1 r=-.153 p=.259 r=.016 p=.905 r=.117 p=.406 r=.327 p=.026 r=-.096 p=.490 r=.065 p=.624 r=-.063 p=.659 r=.159 p=.266 r=.369 p=.009 2 r=-.155 p=.254 r=.005 p=.968 r=-.264 p=.054 r=.001 p=.995 r=-.193 p=.161 r=.037 p=.783 r=-.035 p=.814 r=-.057 p=.685 r=.049 p=.733 3 r=.363 p=.006 r=.238 p=.077 r=.148 p=.296 r=.206 p=.179 r=.338 p=.012 r=.359 p=.006 r=.000 p=.999 r=.510 p<.001 r=.369 p=.008 Pearson correlations in all nine brain regions between: 1) NGAL with megalin (whole population), 2) NGAL with 24p3R (whole population) and 3) Megalin with 24p3R (whole population). Significant correlations are written in bold. BA: Brodmann area. For correction of multiple comparisons, a p-value of less than .006 was considered significant.
Correlations between prefrontal NGAL levels and frontal lobe features
NGAL levels in BA9, BA10, BA11 and BA24 were found to be inversely correlated with the MFS in AD patients (Table 3). There were no significant correlations between MFS scores and megalin or 24p3R, in any of the studied brain regions.
Table 3 Correlations between the Middelheim Frontality score and AD brain levels of NGAL, megalin, and
24p3R.
BA 9 BA 10 BA 11 BA 22 BA 24 Cerebellum Amygdala Hippocampus Thalamus
1 r=-.373 p=.023 r=-.452 p=.005 r=-.375 p=.022 r=-.057 p=.736 r=-.399 p=.016 r=-.018 p=.913 r=-.057 p=.754 r=.078 p=.648 r=.152 p=.376 2 r=.028 p=.872 r=.050 p=.772 r=.174 p=.309 r=.079 p=.651 r=.001 p=.996 r=-.025 p=.882 r=.210 p=.233 r=.047 p=.788 r=-.044 p=.803 3 r=.044 p=.797 r=.009 p=.818 r=.292 p=.084 r=.251 p=.141 r=.131 p=.447 r=.035 p=.833 r=.-155 p=.413 r=-.160 p=.487 r=-.069 p=.693 Pearson correlations in all nine brain regions between: 1) NGAL with Middelheim Frontality score (only in AD patients), 2) Megalin with Middelheim Frontality score (only in AD patients) and 3) 24p3R with Middelheim Frontality score (only in AD patients). Significant correlations are written in bold. BA: Brodmann area.
Discussion
This study shows the following three main findings; (1) NGAL levels are significantly higher in regions throughout the brain, which are commonly affected by AD pathology. (2) NGAL levels were significantly increased in the hippocampus in AD, and even more so in AD with co-existing depression. Hippocampal NGAL levels were positively correlated with the severity of depression. (3) Surprisingly, as compared to control patients, AD patients with co-existing depression did not show higher NGAL levels in the studied prefrontal cortical regions and BA24.
31
2
NGAL in AD–D brains
We have previously shown that NGAL levels were significantly higher in the entorhinal cortex and especially the hippocampus of AD patients as compared to age-matched controls, while cerebellar NGAL levels remained unchanged [23]. Our current observations confirm these previous findings. Moreover, the previous findings are complemented by the novel findings of increased NGAL levels in several other brain regions, including the prefrontal cortex, which are known to be affected by AD [2,48,49]. Whether NGAL expression in the brain merely follows AD pathology or also precedes it remains to be elucidated.
NGAL in AD+D brains
Consistent with our hypothesis, a comparison between controls, AD–D and AD+D patients showed that hippocampal NGAL levels were higher in AD–D, and even higher in AD+D patients, as compared to controls. This may reflect aggravated hippocampal Aβ and tau pathology in AD+D vs. AD–D as described in literature [17,18], which might contribute to higher NGAL levels since Aβ directly stimulates the production of NGAL [38].
The prefrontal cortical regions BA9, BA10 and BA11 and the anterior cingulate cortical area BA24 unexpectedly had lower NGAL levels in AD+D compared to AD–D patients. Evidence exists of dissimilarities between prefrontal cortex and hippocampus in major depressive disorder that could explain our findings. Circuitries involved in mood regulation may be dysfunctional in late-life depression, due to disruptions in underlying white matter tracts. Such mood-related connections also exist between the hippocampus and prefrontal cortical regions [50–55]. Immunohistochemical staining of human AD post-mortem brain tissue illustrated a robust punctate and diffuse distribution of NGAL in the cell bodies, axons and dendrites of hippocampal pyramidal neurons of the CA1 region [23]. Hypothetically, hippocampal NGAL may normally be transported via neuronal projections to prefrontal cortical regions. Disruptions in such projections, which may be present in co-existing symptoms of depression in AD, might therefore impair transport of NGAL to prefrontal regions, and lead to reduced NGAL levels as shown in this study. This hypothesis is further supported by the finding that prefrontal NGAL levels negatively correlate with the MFS score. The MFS provides a measure of frontal lobe features, with a higher score indicating greater abnormalities in frontal lobe functioning [47]. Interestingly, it has been previously reported that MFS scores are positively correlated with severity of depressive symptoms in AD [56], which we also found in patients in this study (data not shown). Taken together, our current results support the association between higher MFS scores and depression, and suggest that frontal lobe changes (which may be involved in the pathology of depression) may be linked with lower NGAL levels in the prefrontal cortex. However, due to the exploratory nature of the correlations between the MFS with NGAL, these findings should be interpreted with caution.
32
Of interest, there may be other factors that, like NGAL, show lower levels in AD+D patients than in AD–D patients, in the prefrontal cortex. In a recent study by Vermeiren et al. [44], the levels of monoamines and their metabolites were determined in the same human brain samples as investigated in the current study. Interestingly, the levels of 3-methoxy-4-hydroxyphenylglycol (MHPG), the metabolite of norepinephrine (NE), were found to be significantly decreased in BA9 and BA10 in AD+D patients as compared to AD–D patients. It was suggested by Vermeiren et al. that these lower prefrontal MHPG levels in AD+D patients may be due to altered norepinephrinergic neurotransmission in the prefrontal cortex, caused by a more severe neurodegeneration of the locus coeruleus in AD+D patients. The locus coeruleus is the primary center of NE synthesis in the brain, and has strong connections with the prefrontal cortex. Although the lower prefrontal NGAL and MHPG levels in AD+D patients as compared to AD–D patients probably have different underlying mechanisms, it is interesting to see that they share this feature. Furthermore, it might be that NGAL expression may in part be induced via NE. Although not shown in brain cells, it was reported that NE treatment of adipocytes causes a significant increase in intracellular NGAL levels [57].
NGAL levels in serum and CSF
Corresponding to our previous findings [23] we did not find significant changes in serum NGAL levels between the studied groups. As shown by [34] it may be that increased serum NGAL levels appear especially in early stages of AD. Moreover, although it was previously shown that plasma NGAL levels were elevated in elderly depressed patients [22], this effect may be lost in AD+D patients.
Lower CSF NGAL levels in AD are in accordance with our previous findings [23]. The decreased NGAL levels in CSF of AD patients might in part be explained by lower megalin expression in the choroid plexus of AD patients [58]. Since megalin in the choroid plexus may transport NGAL from the brain into the CSF, lower megalin levels therefore can result in less NGAL clearance from the brain, causing increased brain NGAL levels and decreased CSF NGAL levels in AD. This potential explanation is supported by the findings that megalin can also transport Aβ from the brain to CSF, and that significantly reduced CSF Aβ levels are characteristic for AD [23,35,58–63]. Our findings further show that CSF NGAL levels are not affected by the presence of co-existing depression in AD.
Megalin and 24p3R
The neurobiological functions of increased megalin levels in BA11 and the amygdala, shown in this study, are unclear. Megalin is expressed throughout the brain, by different cell types like neurons, microglia, astrocytes and endothelial cells. It was reported that neuronal and astrocytic megalin may play an important role in modulating Aβ-mediated neurotoxicity, by indirectly reducing Aβ secretion into the extracellular space [64]. Previous studies showed that megalin expression is decreased in the choroid plexus of AD patients [58]. Our findings indicate that this is not the case for the brain regions in this study.
