ANTI-S100B AUTOANTIBODIES IN CEREBRAL SMALL VESSEL DISEASE AND BRAIN METASTASIS IN A LUNG CANCER POPULATION

CHANDA MULLEN

Bachelor of Science in Biochemistry

John Carroll University

August 2011

Submitted in partial fulfillment of requirements for the degree

DOCTOR OF PHILOSOPHY IN CLINICAL BIOANALYTICAL CHEMISTRY WITH CELLULAR AND MOLECULAR MEDICINE SPECIALIZATION

at the

CLEVELAND STATE UNIVERSITY

December 2015

We hereby approve this dissertation for Chanda Mullen Candidate for the Doctor of Philosophy in Clinical Bioanalytical Chemistry with Cellular and Molecular Medicine Specialization in the Department of Chemistry And the CLEVELAND STATE UNIVERSITY College of Graduate Studies

______Dissertation Chairperson : Michael Kalafatis, Ph.D. ______Department of Chemistry Cleveland State University Date

______Dissertation Committee Member : Aimin Zhou, Ph.D. ______Department of Chemistry Cleveland State University Date

______Dissertation Committee Member : Yan Xu, Ph.D. ______Department of Chemistry Cleveland State University Date

______Dissertation Committee Member : David Anderson, Ph.D. ______Department of Chemistry Cleveland State University Date

______Dissertation Committee Member : Andre Machado M.D. , Ph.D. ______Department of Neurosciences, Cleveland Clinic Date

Date of Defense: November 13, 2015

Dedication

I dedicate this dissertation to the most supportive family, without whom the completion

of this degree would not have been possible:

To my husband Dan, I am forever grateful for your unwavering love and encouragement.

To my mother Andrea, who has been there for me through it all, always motivating me to achieve my goals and to not let anything get in the way, I love you.

To my father Gary, my rock, thank you for being my voice of reason and for never allowing me to doubt my potential.

Finally, to my grandmother Louise, you have given me so much happiness, spending time with you every weekend since the day I was born has shaped me into the woman I am today.

Acknowledgement

First and foremost, I would like to thank my mentor, Damir Janigro, Ph.D, for his encouragement over the last three years. Dr. Janigro introduced me to cerebrovascular research for which I am forever grateful. The time spent performing research in the Janigro lab has provided me with the knowledge and skills to pursue a career in clinical research.

The opportunity to work with human subjects and to develop diagnostic tests that have the potential be translated directly to patient care was an honor. Thank you to my colleagues, the members of the Janigro lab: Chaitali Ghosh, Ph.D; Phillip Iffland, Ph.D; Mohammed

Hossain, MS; Michael Deblock, Ph.D; Kyle Lopin, Ph.D; Vikram Puvenna, MS; and

Nicola Marchi, Ph.D. I am extremely appreciative for your patience and willingness to teach. I would also like to thank the clinicians with whom I had the pleasure of working with; Peter Mazzone, MD, MPH, FCCP; Humberto Choi, MD; Jeffrey Bazarian, MD, and

Michael Phillips, MD. Your insight allowed for the bridging of scientific research and clinical relevance. And to my fellow student and closest friend, Jennifer Rodgers, without whose companionship through coursework, research, studying, practicing presentations, proofreading, and everything else when we aren't being scientists, I would not have made it through these years with the same outlook, we are destined for great things!

Additionally I would like to thank my dissertation committee. I have learned so much from this truly inspiring group of individuals. First, Michael Kalafatis, Ph.D, thank you for being a great advisor and for guiding me through my Ph.D. I will be forever grateful for your support during difficult times. David Anderson, Ph.D, you introduced me to

clinical chemistry and all it has to offer, I am thankful. Aimin Zhou, Ph.D, you taught me the basic biotechnology techniques needed to be a successful scientist, I appreciate the time

I spent in your research lab, and all of your encouragement throughout this program from academic advising to career advice. Yan Xu, Ph.D, thank you for giving me the best feedback during our meetings your knowledge of techniques pushed me to continually optimize my research plans. Andre Machado, MD, you have given me the clinical perspective I needed, your insight is much appreciated.

ANTI-S100B AUTOANTIBODIES IN CEREBRAL SMALL VESSEL DISEASE AND BRAIN METASTASIS IN A LUNG CANCER POPULATION

CHANDA MULLEN

ABSTRACT

The astrocytic protein S100B is a biomarker of blood brain barrier (BBB) permeability. Prolonged BBB permeability, due to the presence of brain metastases and cerebral small vessel disease (SVD), allows for systemic exposure of CNS protein and has the potential to trigger production of autoantibodies against the unmasked antigen S100B in human sera. The aim of the study was to determine the clinical utility of a blood test for the detection of brain metastasis and SVD. A total of 128 patients diagnosed with lung neoplasms underwent MRI at time of diagnosis were enrolled. The measurement of serum biomarkers S100B, and autoantibodies against both the whole protein S100B and its hypothesized immunogenic target sequence ELINNELSHF were measured in human sera by ELISA. Results show that both brain metastasis and SVD are associated with the production of S100B autoantibodies. An S100B autoantibody test can serve as a non- invasive and inexpensive diagnostic tool in the early detection of brain metastasis and

SVD.

vi Table of Contents

Abstract …………………………………………………………………………….. vi

List of Tables ……………………………………………………………………….. xi

List of Figures ………………………………………………………………………. xii

List of Abbreviations ………………………………………………………………... xiii

CHAPTER 1: BACKGROUND AND SIGNIFICANCE

1.1 The Blood-Brain Barrier ……………………………………………. 1

1.1.1 The BBB in CNS Disease …………………………………... 3

1.1.2 Imaging the BBB ………………………………………….... 5

1.1.3 Neuroimaging of SVD …………………………………….... 6

1.2 Cerebral Small Vessel Disease …………………………………….. 8

1.2.1 Cerebral Blood Flow and SVD …………………………...… 8

1.2.2 Endothelial Dysfunction: A Mechanism of SVD …………… 9

1.3 S100B Structure and Function …………………………………….... 12

1.3.1 S100B in Biological Fluids …………………………………. 14

1.3.2 S100B in CNS Disease ……………………………………... 17

1.3.3 S100B Autoimmunity ………………………………………. 18

1.4 The Immune Response ……………………………………………… 19

1.5 Overarching Goals of the Study ……………………………………. 22

1.5.1 AIM 1: S100B and Anti-S100B Autoantibodies as Biomarkers for Early Detection of Brain Metastases in Lung Cancer ……………………………………………...….. 22

1.5.2 AIM 2: Anti-S100B Autoantibodies in Cerebral Small Vessel Disease ……………………………………………….. 24

1.6 References …………………………………………………………… 28

vii CHAPTER II: MATERIALS AND METHODS

2.1 Enzyme-Linked Immunosorbent Assay (ELISA) …………………... 50

2.1.1 Indirect ELISA …………………………………....……….... 52

2.2 S100B Protein Measurement ……………………………....…….….. 53

2.3 Anti-S100B Monomer Autoantibody Measurement ……...... ……... 54

2.4 Anti-S100B Multimer Autoantibody Measurement ………...... ……… 57

2.5 Anti-S100B Peptide Autoantibody Measurement …………..………... 61

2.6 References …………………………………………………..……….... 70

CHAPTER III: S100B AND ANTI-S100B AUTOANTIBODIES AS BIOMARKERS FOR EARLY DETECTION OF BRAIN METASTASES IN LUNG CANCER

3.1 Abstract ………………………………………………………..……… 71

3.2 Introduction …………………………………………………..……….. 73

3.3 Methods ……………………………………………………………….. 75

3.3.1 Patient Enrollment…………………………………………...… 75

3.3.2 Serum Measurements of S100B and anti-S100B Autoantibodies by ELISA …………………………..………... 75

3.3.3 Neuroimaging …………………………………………………. 76

3.3.4 Statistical Analysis …………………………………………….. 77

3.4 Results …………………………………………………………………. 78

3.4.1 Demographics ………………………………………………….. 78

3.4.2 Demographic effects of serum S100B and anti-S100B Autoantibodies …………………………………….. 83

3.4.3 S100B, anti-S100B autoantibodies and neuroimaging ………… 84

3.4.4 S100B, anti-S100B autoantibodies and lung cancer staging ……………………………………….………………….. 88

3.5 Discussion ……………………………………………………………… 89

viii 3.6 References ………………………………………...……………………. 92

CHAPTER IV: ANTI-S100B AUTOANTIBODIES IN CEREBRAL SMALL VESSEL DISEASE

4.1 Abstract ……………………………………………………………….. 95

4.2 Introduction …………………………………………………………… 97

4.3 Methods …………………………………………………………….... 100

4.3.1 Human Subjects …………………………………………….... 100

4.3.2 SVD Diagnosis by MRI ……………………………………… 100

4.3.3 Serum S100B Protein Measurement …………………………. 101

4.3.4 Anti-S100B Autoantibody Measurement …………………...... 101

4.3.5 Peptide Synthesis ……………………………………………... 102

4.3.6 Anti-S100B Peptide Autoantibody Measurement ……………. 104

4.3.7 Total Serum IgG Measurement ……………………………….. 105

4.3.8 Statistical Analysis ……………………………………………. 106

` 4.4 Results ………………………………………………………………… 107

4.4.1 Patient Characteristics, Neuroradiology and Demographics …………………………………………...... 107

4.4.2 SVD Grading …………………………………………………. 111

4.4.3 Serum Markers ……………………………………………….. 114

4.4.4 Clinical Utility ………………………………………………... 116

4.4.5 Lung Cancer Staging …………………………………………. 118

4.5 Discussion ……………………………………………………………. 120

4.6 Conclusion …………………………………………………………..... 124

4.7 References ……………………………………………………………. 125

ix

CHAPTER V: FUTURE DIRECTIONS

5.1 Antibody-Antigen Interaction: Epitope Mapping and Kinetics ……… 132

5.2 Limitations of Enrolled Patient Population …………………………... 135

5.3 S100B Expression in Lung Cancer …………………………………… 137

5.4 Conclusion ……………………………………………………………. 138

5.5 References .……………………………………………………………. 139

x

List of Tables

Table 1: S100B in Biological Fluids of Disease States ……………………….. 15

Table 2: Negative controls anti-S100B peptide autoantibody ELISA ……...... 69

Table 3: Lung cancer patient demographics ………………………………….... 79

Table 4: Cancer staging, histology, and neuroimaging characteristics ……….... 81

Table 5: Brain metastases demographics ………………………………………. 82

Table 6: Multivariable analysis of demographic effects on serum S100B …..... 83

Table 7: Multivariable analysis of demographic effects on serum anti-S100B autoantibodies ……………………………………………..………. 84

Table 8: S100B and anti-S100B autoantibody levels in the presence of SVD and brain metastases ………………………………………………..…. 85

Table 9: ROC analyses of S100B and anti- S100B autoantibodies for the detection of brain metastasis ……………………………………………..…..... 87

Table 10: S100B and anti-S100B autoantibodies for the detection of brain metastasis in patients stage I or II before brain imaging ……………...…. 88

Table 11: Patient Characteristics ……………………………………………… 110

Table 12: ARWMC scale and serum marker levels …………………………… 113

Table 13: Serum Marker levels in patients …………………………………….. 115

Table 14: ROC analyses for the prediction of SVD ……………………………. 117

Table 15: Amino acid properties of S100B peptide sequence ELINNELSHF …………………………………….………………... 134

xi List of Figures

Figure 1: Decreased CBF in SVD …………………………………………….. 11

Figure 2: Proposed mechanism of anti-S100B autoimmunity in SVD ……….. 26

Figure 3: Western blot analysis of S100B using two anti-S100B monoclonal antibodies ………………………………………………………… 55

Figure 4: Anti-S100B Monomer Autoantibody ELISA Standard Curves …….. 56

Figure 5: Silver stain of S100B Multimer …………………………………….. 59

Figure 6: Serum Titrations for optimization of anti-S100B peptide autoantibody ELISA ………………………………………………………...... 63

Figure 7: Conjugated Anti-Human IgG Antibody Titration …………………... 65

Figure 8: Monoclonal anti-S100B antibody binding to ELINNELSHF ………. 67

Figure 9: Pooled Human IgG Titration ………………………………………... 68

Figure 10: Radiologic appearance of SVD …………………………………… 108

Figure 11: SVD severity and serum marker levels …………………………… 112

Figure 12: Lung cancer staging and serum marker levels ……………………. 119

xii List of Abbreviations

AD Alzheimer's disease

AJs Adherens junctions

ANOVA Analysis of variance

ARWMC Age related white matter changes

AU Absorbance Unit

AUC Area under curve

BBB Blood-brain barrier

BBBD Blood brain barrier disruption

BMI Body mass index

BSA Bovine serum albumin

CBF Cerebral blood flow

CNS Central nervous system

COPD Chronic obstructive pulmonary disease

CrCl Creatinine clearance

CSF Cerebral spinal fluid

xiii CT Computed tomography

DAMP Danger associated molecular pattern

DTI Diffusion tensor imaging

DWI Diffusion weighted imaging

ELISA Enzyme-linked immunosorbent assay

FLAIR Fluid attenuated inversion recovery

Gd Gadolinium

GLUT Glucose transporter

HIV Human immunodeficiency virus

HPLC High performance liquid chromatography

HRP Horseradish peroxidase

IgG Immunoglobulin G

IL-1β Interleukin 1 beta

MALDI-TOF Matrix-assisted laser desorption/ionization time-of-flight met Metastasis

MHC Major histcompatibility complex

MRI Magnetic resonance imaging

xiv mRNA Messenger ribonucleic acid

MS Multiple sclerosis

NPV Negative predictive value

NSCLC Non-small cell lung cancer

OPD o-phenylenediamine dihydrochloride

PBS Phosphate buffered saline

PPV Positive predictive value

RAGE Receptor for activated glycation end products

ROC Receiver operator characteristic

SRS Stereotactic radiosurgery

SVD Cerebral small vessel disease

TJs Tight junctions

TNFα Tumor necrosis factor alpha

VEGF Vascular endothelial growth factor

xv

CHAPTER I

BACKGROUND AND SIGNIFICANCE

1.1 The Blood-Brain Barrier

The blood-brain barrier (BBB) is essential for the functioning of the central nervous system (CNS). It is a diffusion barrier that impedes the entrance of compounds from the blood to enter the brain. The BBB also provides immunological privilege in the brain. The

BBB is made up of endothelial cells, astrocyte end feet, and pericytes. Tight junctions

(TJs) allow for the formation of a barrier. TJs are located between the cerebral endothelial cells and selectively allow the diffusion of blood compounds. Small lipophilic molecules can freely cross the BBB yet large and hydrophilic molecules cross the BBB via active transport or concentration gradients. The mechanism of active transport across the BBB is regulated based on the requirements of the cerebral environment. For example, glucose

1 transporters are upregulated on the luminal side of the membrane when cerebral nutrient supply is low (1).

TJs maintain ion homeostasis of the BBB. Tight junctions link endothelial cells together to create an impermeable barrier. TJs are made up of three integral membrane proteins occludin, claudin, and junction adhesion molecules. Below the TJs is a secondary barrier, the adherens junction, which is located in the paracellular space. Adherens junctions (AJs) are composed of a cadherin-catenin complex and further limit vascular permeability (2).

Claudins are localized exclusively at TJs and bind other claudins on adjacent endothelial cells to form the primary seal of the TJ (3). The carboxy terminal of claudins bind the cytoplasmic proteins zonula occludins (ZO); ZO-1, ZO-2, and ZO-3. ZO-1 and

ZO-2 bind to actin in order to provide structural support to brain endothelial cells in the

BBB (4).

Occluidns are also localized at TJs and are much larger phosphoproteins than claudins. Occludin expression is highest in brain endothelial cells in comparison to non neural tissues. It functions as a regulatory protein to alter permeability (5). It has been shown in the context of brain tumors that increased BBB permeability is associated with a loss of occludin expression (6). The combination of claudins and occludins form the extracellular component of TJs and are required in the formation of the BBB (7).

The structure of the BBB does not prevent the movement of all solutes from the circulation to extracellular fluid, allowing for lipophilic agents to penetrate the brain parenchyma. Other large, non lipid soluble agents must also move from the circulation to 2 the parenchyma and vice versa. The movement of these solutes is facilitated by transporters in the endothelial cell membrane, such as the GLUT-1 transmembrane glucose transporter

(8).

The BBB also contains astrocytes and pericytes. Astrocytic end feet cover a large portion of the capillary walls. Astrocytes play a supportive role in the BBB, without astrocytes the BBB endothelial cells do not form an optimal barrier (9). In addition they perform spatial buffering by which water and ions are shuttled to the area near the cerebral microvasculature (10). Pericytes are cells of the microvessels that surround endothelial cells. They provide structural stability to the vessel wall and vasodynamic activity to the microvessels. Studies have shown that pericytes regulate angiogenesis (11).

1.1.1 The BBB in CNS Disease

Under healthy physiologic conditions the BBB is intact and selectively permeable, yet in CNS diseases an increase BBB permeability is observed. In the context of stroke or ischemia, TJs are disrupted and mediated by cytokines and vascular endothelial growth factor (VEGF), among others, leading to BBBD. Proinflammatory cytokines including IL-

1β and TNF-α are also elevated after ischemia (12). These proinflammatory mediators were shown up regulate endothelial and neutrophil adhesion molecules leading to the transmigration of leukocytes across the BBB. Increased leukocytes activate signal transduction pathways leading to the breakdown of TJs and subsequently the BBB (13).

Hypoxia induces increased VEGF release causing increased transcytosis and decreased expression of ZO-1 (14).

3 Increased BBB permeability has also been observed in brain tumors. The loss of occludin expression in the microvessels was observed in brain metastasis leading to an opening of endothelial TJs (15, 6). The mechanism of loss of TJ molecules in brain tumors is incompletely known, but VEGF secreted by tumors may play a role in the down regulation of proteins leading to TJ opening (16).

A major role of the BBB is to maintain the immune privilege of the CNS. Studies have shown that the CNS is an anti-inflammatory environment characterized by low MHC-

1 molecules and cytokines (17-19). Inflammatory molecules are neurotoxic and can promote cell death, in addition the skull is unable to tolerate inflammatory actions that lead to swelling (20). Immune cells can be neurodegenerative in the brain, in multiple sclerosis

(MS) T cells are found within demyelinated lesions suggesting a causal role (21). TJs of the BBB have the ability to prevent the passage of leukocytes from blood to brain (22). All immune privilege in the brain is compromised once inflammation occurs. This is due to the breakdown of the BBB which lessens the immunosuppressive effect of the CNS microenvironment. Additionally, immune privilege of the brain is dependent upon age, both the very young and very old have altered BBB integrity and immune privilege.

Immune privilege is relative and is not solely dependent on the BBB. The current hypothesis is that immune privilege in the CNS is compartmentalized, and that immune privilege is confined to the parenchyma (23). The lack of communication between the brain parenchyma and the periphery may be due to afferent and efferent immune responses. The afferent arm is responsible for adaptive immune privilege in the brain. The afferent arm is

4 responsible for the primary immune response in which antigen is presented to T cells for priming and activation. This can occur by the drainage of soluble antigens in the lymph.

The brain lacks a lymphatic system, yet drainage has been shown to occur along the perivascular space of the cerebral small vessels into the subarachnoid space through channels of the cribriform plate into the lymphatics of the nasal submucosa and cervical lymph (24, 25). Antigen injected into the parenchyma was shown to drain into the cervical lymph where it elicits an antibody response (26). Cytotoxic T cell responses were not observed, suggesting that B cell and T helper 2 responses may be favored. Soluble antigen was also shown to elicit in an immune response before it enters the cervical lymph and was taken up by dendritic cells in the CSF and meninges (27).

The efferent arm of the immune response in the CNS is also relative. T cells are able to enter the CNS and there is evidence of CNS targeted T cells which are able to detect antigens across an intact BBB (28). Once in the CNS, T cells recognize antigen via MHC, yet in normal CNS MHC expression is low. Inflammation must be present for upregulation of MHC, which is seen in multiple neurological disease states.

1.1.2 Imaging the BBB

The BBB does not maintain a constant integrity throughout life. For example, nonspecific BBB changes due to aging can be assessed using the CSF to plasma albumin ratio. Albumin is not normally synthesized in the brain, and its size does not permit passage across the BBB, resulting in very low levels in the CSF. Increasing albumin levels in the

CSF are indicative of loss of BBB integrity. Another marker of BBB function is the

5 astrocytic protein S100B. S100B is primarily synthesized in the brain. Thus, when increasing levels of S100B are detected in blood, it is indicative of increased BBB permeability (29).

BBB permeability can be imaged by contrast-enhanced MRI or CT (30, 31).

Gadolinium contrast is used to enhance areas of BBB breakdown. In these areas, contrast is able to penetrate into the CNS, and tumors, infections, and demyelinating diseases are able to be imaged. Scans are obtained before and after contrast injection. Due to its size, gadolinium does not diffuse across an intact BBB. Thus, gadolinium extravasation in MRI after IV contrast injection is an effective marker of BBB permeability. Increased contrast uptake is seen in multiple disease states including lacunar stroke (31, 32) vascular dementia

(33), diabetes (34), Alzheimer’s disease (35), and cerebral small vessel disease (SVD) (30).

BBB imaging can also be performed by contrast CT. Clinically relevant CT contrast agents include iodine, which attenuates x-rays and allows for hyperdense signal on CT.

1.1.3 Neuroimaging of SVD

Brain imaging modalities allow for the estimation of the extent and severity of

SVD. The presence of SVD is accompanied by white matter changes seen on CT and MRI

(36). SVD white matter changes are generally symmetrical and are most commonly seen in the periventricular and deep white matter. Advanced MRI techniques allow for more in- depth analysis of the structural features within white matter lesions. It also allows for assessment of the normal-appearing white matter on conventional CT or MRI (37).

On CT, SVD-related white matter changes are seen by a decrease of x-ray

6 attenuation or hypodensity. Pathologically, these areas indicate a loss of the myelin and axons (38). On MRI, white matter lesions are hyperintense on T2-weighted imaging.

However, T2-signal hyperintensity alone is insufficient to diagnose SVD as these hyperintensities are associated with many different types of tissue damage. A combination of signal patterns and distributions is used to define SVD. On T1-weighted imaging, SVD- related white matter changes are faintly seen, differing greatly from the more extensive white matter changes seen in multiple sclerosis (39). Magnetization transfer imagining is an example of quantitative MRI, a technique that can determine the difference in demyelination severity between SVD and Multiple Sclerosis (MS) based on the different signal characteristics of white matter changes (40).

DWI can also be employed to study SVD. White matter lesions seen on MRI can be further investigated by DWI as it is highly sensitive in the detection of early brain infarcts. In SVD, there are multiple small infarcts of varying age, meaning that all infarcts happen at different time points. DWI provides insight into the differentiation of age and subsequent severity of white matter lesions in SVD (41). In summary, SVD is an extremely complex disease, requiring a myriad of imaging modalities for the characterization and advancement towards treatment.

7

1.2 Cerebral Small Vessel Disease

SVD is an etiological mechanism of lacunar stroke, which accounts for approximately 25% of ischemic stroke, and is a major risk factor of vascular dementia (42,

43). While the pathological significance of SVD is not entirely understood, evidence exists linking small vessel disease to cognitive impairment (44, 45). The etiological mechanisms involved in SVD are incompletely known, but risk factors for SVD include age, hypertension, smoking, and heart disease, which are also associated with stroke and dementia (46). SVD lesions are associated with an increased risk of ischemic stroke, dementia, gait disturbances, cerebral hemorrhage, vascular death, and mortality (47, 48).

Current treatment addresses only vascular risk factors by controlling blood pressure and lipid levels (49). The radiologic means to grade SVD is the Fazekas scale (50, 51).

The pathology of SVD is performed post mortem and is primarily characterized by concentric hyaline wall thickening and calcification of the vessels (52). This is due to atherosclerosis of the cerebral small vessels in the white matter lesions. Areas of white matter lesions also show demyelination and loss of axons.

1.2.1 Cerebral Blood Flow and SVD

The relevance of cerebral blood flow (CBF) and its regulation in the brain is seen in SVD, a progressive, chronic vascular dysfunction of the cerebral vessels. While the pathological significance of SVD is not entirely understood, evidence exists linking SVD

8 to cognitive decline (44, 45, 53). The pathophysiology of SVD involves arteriopathy of cerebral small vessels, which leads to a decrease in blood flow. The resulting hypoperfusion impairs autoregulation and causes ischemia (54, 55). In the brain, autoregulation maintains CBF. It has been shown that CBF decreases due to impaired autoregulation in patients with SVD risk factors including aging and hypertension (56). In prolonged SVD, cerebral small vessels undergo inward remodeling, reducing vasodilator responses and increasing susceptibility to chronic hypoperfusion. A chronically hypoperfused state results in ischemia and, ultimately white matter lesions (57). White matter lesions are associated with cognitive decline and dementia (58, 59). Previous studies have reported the presence of altered CBF in white matter lesions (54) and decreased auto- regulation (55) in SVD.

1.2.2 Endothelial Dysfunction: A Mechanism of SVD

A current hypothesis of the pathophysiology of SVD implicates endothelial dysfunction. The normal endothelium functions to regulate CBF, autoregulation, and maintains the BBB by active and passive transport mechanisms (60, 61). A proinflammatory response is induced when the endothelium is activated by leukocyte infiltration. Endothelial cell activation causes an increase in vascular permeability allowing entry of serum proteins into the vascular wall and perivascular neural parenchyma, producing potentially toxic effects (62).

In addition, endothelial dysfunction promotes atherosclerosis and premature arterial aging (59, 63). Endothelial dysfunction contributes to brain aging in several ways including impairing neurogenesis, altering neuronal transmission, inhibiting amyloid 9 precursor protein processing, and affecting survival of oligodendrocyte precursor cells

(64).

Increased BBB permeability measured by contrast MRI or serum S100B is seen in patients with SVD (30, 65). Another mechanism of white matter lesions in SVD is the breakdown of the endothelium of the BBB causing extravasation of toxic fluids into white matter. BBB breakdown was shown in white matter lesions by the measurement of increased CSF proteins in patients with SVD (66). Also, increased BBB permeability was shown to precede the development of white matter lesions, thereby underscoring the causal role of altered BBB permeability in SVD (30). Furthermore, decreased BBB integrity correlates with the severity of white matter lesions, and the damaged white matter may be the site of BBB leakage, see Figure 1 (67).

10

Figure 1: Decreased CBF in SVD. Arterial stiffness (atherosclerosis) and endothelial dysfunction contribute to accelerated arterial aging along with risk factors of SVD and increased BBB permeability. Arteriopathy of cerebral small vessels leads to a decrease in blood flow; this hypoperfusion impairs auto-regulation, causing ischemia, lesions, and dementia. Shown is a contrast enhanced MRI of SVD. Arrows indicate regions of white matter hyperintensity.

11 1.3 S100B Structure and Function

S100B was first discovered in 1965 by Dr Moore, it was a protein fraction isolated from brain that was soluble in 100% saturated ammonium sulfate hence its name S100.

S100B is part of a family of calcium binding proteins with more than 20 members. The

S100 group of proteins is the largest subgroup of the EF-hand superfamily. S100 proteins contain two distinct EF hands; a modified S100-specific EF hand at the N-terminus and a classical calcium binding EF-hand. The genes that encode S100 proteins are located on chromosome 1 at region 1q21 with the exception of S100B which is located on chromosome 21 at region 21q22.

Each monomer of S100B is approximately 12kDa and can form homodimers in solution (68). The S100B homodimer is the major component of the protein fraction isolated from brain extract. In extracellular space S100B forms disulfide crosslinked homodimers (69). The structure of S100B contains two EF-hand calcium binding regions connected by a loop. The C-terminal calcium binding motif is common to all EF-hand proteins with a signature 12 amino acid sequence. The N-terminal EF hand is S100 specific and called a “pseudo EF-hand” the binding motif is 14 amino acids. Upon binding the helix

III of the classical EF-hand undergoes a conformational change exposing a hydrophobic cleft formed by residues in the hinge region, helix III, and C-terminal loop region (70).

This hydrophobic surface is the interaction site of S100 proteins with their targets (71).

S100B is primarily found in astrocytes and other glial cell types including; oligodendrocytes, Schwann cells, and retinal Muller cells (72, 73). Extracranial sources of

S100B have been reported in melanocytes, dendritic cells in lymphoid organs, Leydig cells,

12 and adipose tissue (74-76). Yet our studies have shown that extracranial sources of S100B do not effect serum levels (77).

The exact biological function of S100B is unknown, but intracellular and extracellular functions have been proposed. Intracellular S100B acts as a "regulator" while extracellular S100B acts as a "signal". Intracellularly S100B acts as a calcium sensor protein in order to induce a conformational change for binding of target proteins. However not all intracellular S100B target proteins are calcium dependent (78). S100B also regulates cell proliferation as seen in the context of melanoma where S100B interacts with the tumor suppressor p53by inhibiting phosphorylation and therefore p53 expression (79).

Other intracellular functions of S100B have also been proposed; protein phosphorylation, enzyme activity, cell differentiation, and cytoskeleton interaction (80).

S100B is released by astrocytes and extracellularly acts as a cytokine where it binds

RAGE and activates signaling pathways (81, 82). Nanomolar concentrations of extracellular S100B can display neuroprotective properties while micromolar concentrations can have deleterious effects. At nanomolar concentrations S100B upregulates anti-apoptotic factor upon binding RAGE and activates signaling pathways

(83). S100B also promotes neurite extension, a process mediated by RAGE dependent activation (84). High concentrations of S100B cause astrocyte apoptosis via inducible nitric oxide synthase activity (85). In addition, increased S100B can activate astrocytes allowing for the participation in brain inflammatory responses in a RAGE dependent manner (86). Micromolar S100B levels have been shown to stimulate apoptosis in vitro

13 (87). High levels of S100B have been measured in many neurodegenerative disorders including Alzheimer's disease, Down's syndrome, and epilepsy among others (88, 89).

1.3.1 S100B in Biological Fluids

Serum S100B is a biomarker of BBB permeability as it enters the blood when the

BBB is breached (90, 91). The ratio of CSF to serum levels of S100B are 10:1 allowing

S100B to be an ideal peripheral marker of BBB integrity (92-94). S100B protein was first detected in the CSF multiple sclerosis (MS) patients (95). Its detection in biological fluids including; CSF, blood, urine, saliva, amniotic fluid, and breast milk give S100B the potential to be a biomarker of CNS integrity, see Table 1. S100B can be measured at high concentrations in CSF and serum. In brain trauma and in acute brain injury and increased

S100B levels correlated with the size of brain lesions. Serum S100B was shown to correlate with the severity of systemic inflammation, as measured by CRP in a population of acute ischemic stroke patients (96). The origin of S100B whether leaked form damaged cells or secreted is yet unknown.

14 Biological S100B Disease State Reference Fluid Level

CSF Acute brain injury Increase 98, 99

Brain Tumor Increase 100

MS Increase 95, 101

AD Increase 102, 103

Schizophrenia Increase 104

Depression/Bipolar Increase 105

Neonatal BLOOD intraventricular Increase 106 haemorrhage

Acute brain injury Increase 96, 107, 108, 109, 110, 111

Cardiac events Increase 112, 113

Brain metastases Increase 65, 114

Melanoma Increase 115, 116, 117

MS Increase 118

AD Decrease 119

Schizophrenia Increase 120, 121

Depression/Bipolar Increase 122, 123

Obesity Increase 124

15

Preterm newborns with Increase 125 Intraventricular haemorrage

Neonatal hypoxic- URINE ischemic Increase 126 encephalopathy

Early neonatal Increase 127, 128 death

Vigorous physical SALIVA Increase 129 activity

AMNIOTIC Intratrauterine Increase 130 FLUID death

CNS Increase 131 malformations

Trisomy 21 Increase 132, 133

Table 1: S100B in Biological Fluids of Disease States. (Adapted from (97) )

16 1.3.2 S100B in CNS Disease

Increased S100B in biological fluids are a hallmark of many acute and chronic neurological diseases. S100B has been measured in the CSF and blood of patients with brain tumors (65, 114). However, the most common diagnostic use of S100B is in the detection of melanoma. S100B was first discovered in human melanocytes and melanoma tissue in 1981 and has since become an immunohistochemical marker for pigmented skin lesions (74,134). Serum S100B levels are also used as a biomarker in melanoma patients to monitor the progression of disease and to evaluate the efficacy of therapy (115).

S100B can be regarded as an inflammatory danger signal protein consistent with the inflammatory processes of carcinogenesis (135). The extracellular binding of S100B to the RAGE receptor plays a role in tumor development and metastasis (136, 137). The high concentration of S100B released in neurodegenerative diseases and melanoma allow for the binding of RAGE in macrophages to induce an inflammatory response (138). S100B can act as a chemoattractant for RAGE+-CD4+ T cells as evidenced in a mouse model of

MS allowing for the entrance of T cells into the brain (139).

In the context of Alzheimer’s disease (AD) S100B has been studied as a biomarker and as an effector. S100B is released in high concentrations from activated astrocytes, this up regulation of S100B is seen in AD tissue (140, 141). S100B has been linked to cognitive behavior suggesting that extracellular S100B plays a role in memory processing yet the mechanism has yet to be discovered (142). There is conflicting evidence of serum S100B concentrations in AD. In one study serum S100B concentrations were lower in AD patients compared with elderly controls, and no correlation with MRI-based assessment of brain 17 atrophy was found (143). Other studies have reported the correlation of CSF S100B levels and AD (103).

1.3.3 S100B Autoimmunity

Some CNS diseases are linked to mechanisms of autoimmunity. A common hypothesis of the mechanism of autoimmunity is the process of antigen unmasking due to the loss of CNS immunoprivilege (17). The current hypothesis is that the production of autoantibodies against brain specific antigen is due to prolonged exposure to the periphery following BBB permeability. In particular, S100B autoantibodies have been found in the context of neurodegenerative diseases (144-146). In AD the development of S100B autoantibodies was shown to precede the development of autoantibodies against amyloid beta in the timeline of disease suggesting its clinical utility as a biomarker of early AD diagnosis (144). Our lab has shown that in sub concussive head hits the release of S100B in serum allows for an auto immune response against S100B at the end of a football season

(111). Additionally the autoantibody levels correlated with MRI-DTI changes suggesting that autoimmunity against CNS protein may play a role in the initial progression of post traumatic cognitive decline.

There are many studies linking S100B to the immune system, however little is known about the mechanism of brain derived S100B in peripheral autoimmunity (147).

Protein levels of S100B have been reported in extracranial tissue. However, our lab has shown that S100B protein is observed in cells of lymph nodes, spleen, skin, and heart muscle, yet the mRNA expression of S100B was limited to barrier organs (148). This suggests that S100B can be both produced in barrier organs and taken up by cells in 18 systemic tissues. The sequestration of S100B by dendritic cells allow for the consideration of S100B as an antigen (148). The specificity of S100B was demonstrated by comparing uptake of S100A1, which only differs in sequence by 4 amino acids. It was shown that

S100B accumulation in dendritic predominates S100A1, yet both were shown to accumulate (148). S100B was found in both CD4+ and CD8+ dendritic cells of the spleen.

The main finding of the study is that S100B is present in immune cells, which may constitute a step of self-antigen recognition, causing the initiation of an immune response.

S100B can be perceived as a danger associated molecular pattern (DAMP) which has the ability to initiate an immune response. S100B uptake by dendritic cells is observed, but may remain inactive and "immune tolerant" until a threshold of antigen by prolonged

BBB permeability is reached (148). The low basal level of circulating S100B does not appear to mount an immune response as detectable levels of autoantibodies are only see after repeated increases of S100B as in the case of neurological diseases in which there is a breach of the BBB. The production of anti-S100B antibodies may persist in blood until the integrity of the BBB is breached, when they may then enter the brain (149).

1.4 The Immune Response

The main function of the human immune system is to defend against infection by distinguishing foreign agents from self and non-self. An antigen is a molecule that can initiate an immune response. Lymphocytes (B cells) respond to antigens by producing antibodies. Antibodies are secreted into blood to provide humoral immunity. Other lymphocytes (T cells) do not secrete antibodies rather they attack cells that carry antigens which is part of cell-mediated immunity. 19 The immune response mechanism involves leukocytes. Leukocytes are white blood cells and include; B cells, T cells, neutrophils, eosinophils, basophils, and monocytes. T cells play a role in the cell mediated immune response while B cells play a role in the humoral response. Neutrophils and macrophages are phagocytes which engulf foreign antigen. Neutrophils are short lived cells that release granules to assist in pathogen elimination. Macrophages are long lived cells that play a role in antigen presentation to T cells. Dendritic cells are a type of antigen presenting cell that also phagocytose foreign antigen.

T cells originate in bone marrow from hematopoietic stem cells and then migrate to the thymus where they mature and develop the ability to recognize antigen. T cells express a unique antigen binding receptor on their membrane, the T cell receptor, which is specialized to recognize a particular antigen. There are two main types of T cells; T helper cells and cytotoxic T cells. Cytotoxic T cells are involved in the destruction of cells infected with foreign antigen. T helper cells maximize the immune response and do not play a cytotoxic role.

B cells mature in the bone marrow and then migrate to the periphery including blood and lymph. B cells express a unique antigen binding receptor on their membrane.

Unlike T cells, B cells can recognize free antigen. The main function of B cells is to produce antibodies against foreign antigen. B cells become activated when they encounter antigen and then divide and differentiate into plasma cells and memory cells. Memory cells are survivors of past infection that express antigen binding receptors and are called upon when re-exposed to antigen. Plasma cells produce antibodies against the target

20 antigen. Plasma cells are short lived and undergo apoptosis when the foreign antigen is eliminated.

Macrophages survey cell surfaces for the presence of glycoproteins. The major histocompatibility complex (MHC) are a group of genes that produce the glycoproteins present on cell surfaces. These glycoproteins are considered MHC proteins which are polymorphic and different combinations exist for each individual. MHC proteins act as cell surface markers for T cells in the recognition of cells as self versus non self. When a foreign antigen infects the body it is taken up by antigen presenting cells, partially degraded, and moved to the plasma membrane. At the membrane the antigen fragments are complexed with MHC proteins to allow for T cell recognition. There are two types of MHC proteins;

MHC I and MHC II. MHC I is present on all nucleated cells while MHC II is only present on macrophages, B cells, and CD4+ T cells. MHC I proteins present intracellular peptides while MHC II proteins present extracellular peptides.

T cells become activated when they encounter an antigen presenting cell that displays an antigen fragment bound to its MHC protein. Cytotoxic T cells only recognize antigens complexed with MHC I proteins. T helper cells only recognize antigens complexed with MHC II proteins. The restriction of the T cell recognition is the result of co receptor proteins associated with T cell receptors. The CD 8 co receptor is associated with cytotoxic T cells and can interact only with MHC I proteins. The the CD 4 co receptor is associated with T helper cells and only interacts with MHC II proteins. CD 4 is expressed on the surface of T helper cells, macrophages and dendritic cells.

21 The cell mediated immune response protects the body from infection of virus or cancer by killing infected cells. The cell mediated immune response is carried out by T cells. Cytotoxic T cells are activated when they recognize cells displaying foreign antigen complexed with MHC I. This activation induces apoptosis of the cells displaying the foreign antigen including virus infected cells, cells with intracellular bacteria, and cancer cells displaying tumor antigens. T helper cells can also be activated in the cell mediated immune response. There are two types of T helper cell responses that can be induced by an antigen presenting cell; T helper 1 cells (Th1) and T helper 2 cells (Th2). The Th1 response is characterized by the release of interferon gamma which activates other cytokines to induce B cells to produce antibodies (IgG). The Th1 response is stimulated by intracellular antigen accumulation. The Th2 response is stimulated by extracellular antigen to activate the humoral immune response.

The humoral immune response is carried out by B cells. The B cell receptor recognizes and binds to antigen. This attracts T helper cells which secrete cytokine (IL-2) to stimulate B cells to divide and differentiate into plasma cells and memory cells. Plasma cells produce antibodies that are released into circulation. The secreted antibodies bind antigen on the surface of pathogens marking them for destruction and elimination. There are five types of antibodies produced by B cells; IgA, IgD, IgE, IgG and IgM. Each antibody has a specific biological function to recognize and neutralize pathogen.

1.5 Overarching Goals of the study

1.5.1 Aim 1: S100B and Anti-S100B Autoantibodies as Biomarkers for Early

Detection of Brain Metastases in Lung Cancer

22 At the time of diagnosis 10-15% of patients with lung cancer have brain metastasis at diagnosis (150). Brain metastases are a common site of disease recurrence after definitive treatment. In non small cell lung cancer (NSCLC), 30% of patients fail in the brain as the first site of failure (151). The early detection of brain metastases in patients with cancer have important implications for treatment and prognosis. Brain MRI with contrast is the gold standard for the detection of brain metastases.

A biomarker of brain metastases in lung cancer has several potential roles. Blood based biomarkers have the ability improve the early detection, diagnosis and treatment of lung cancer. Given the cost of radioimaging a serum marker can be a useful, non invasive and inexpensive tool. Patients with a high likelihood of having brain metastases, as identified by a sensitive serum marker, like S100B, at the time of diagnosis, could proceed to further radiological investigation. Those with a negative blood test could avoid imaging, therefore the number of unnecessary scans would be minimized.

There is evidence that brain metastases may be detectable using serum markers of

BBB damage. The neovascularisation associated with the growth of metastatic lesions causes increased BBB permeability. Allowing for CNS protein to extravasate into the periphery. The astrocytic protein S100B one such marker of BBB permeability. Elevated

S100B levels in CSF and serum after head trauma, subarachnoid hemorrhage, and stroke were shown to correlate with the extent of brain damage (94). Studies have confirmed the association between serum S100B and contrast enhanced MRI in the context of metastatic brain disease (29, 65). Increased serum S100B has been shown to trigger the production of anti-S100B autoantibodies in BBB disrupted football players who experience mild TBI

23 (111).

The main goal of this study is to determine the clinical utility of an S100B biomarker panel including S100B protein and anti-S100B autoantibodies in the detection of brain metastasis. Serum levels of these markers are measured by ELISA in patients with newly diagnosed lung cancer and brain MRI.

1.5.2 Aim 2: Anti-S100B Autoantibodies in Cerebral Small Vessel Disease

The environment produced in the brain of cerebral small vessel disease is subject to chronic opening of the BBB. SVD is one of the cerebrovascular risk factors for accelerated cognitive and motor impairment, ultimately leading to dementia (46). The current means of diagnosing SVD include imaging modalities such as CT, MRI, and DTI, which are expensive and often impractical. A marker for SVD would have important implications for early interventions and prognosis of pre-senile dementia. On the one hand, a blood marker with a high negative predictive value may be used to rule out SVD to avoid unnecessary scans and thus to decrease health care delivery costs. On the other hand, a marker with a high positive predictive value may aid in SVD diagnosis and be more sensitive than CT scans. In the case that a treatment for pre-senile vascular changes becomes available, it would be easy to predict a high clinical advantage due to early diagnosis of disease.

Small vessel disease seen on MRI scans is thought to be due to or associated with a disrupted blood-brain barrier (BBB) (152). BBB “leakage” may provide systemic exposure of CNS protein (antigen unmasking), specifically the astrocytic protein S100B,

24 See Figure 2. Given the fact that SVD is persistent, a long term exposure to unmasked antigen may lead to the production of auto-antibodies as shown in post-concussive syndrome or in pre-senile dementia (111, 145). Serum levels of S100B alone do not have clinically useful predictive values because any acute or transient blood-brain barrier disruptions elevate S100B regardless of the presence of SVD. While it has been shown that

S100B protein is elevated in subjects with SVD, whether SVD also lead to a measurable levels of auto-antibodies against S100B or other CNS protein is presently unknown (65,

114). A preliminary bio-informatic analysis revealed that a specific sequence of 10 amino acids in human S100B has the potential to be highly immunogenic. Our results support the presence of autoimmune IgGs targeting this sequence.

25

Figure 2: Proposed mechanism of anti-S100B autoimmunity in SVD

The main hypothesis is that, in patients at risk, SVD can be diagnosed by measurements of autoantibodies against S100B. To address this general hypothesis the following research aims were investigated:

AIM 2.1: Develop a blood test to diagnose or detect the presence of SVD. Rationale: SVD has potentially devastating consequences and currently a blood test for SVD does not exist.

Hypothesis: Autoimmunity against brain protein predicts radiological findings consistent with SVD. Approach: Enroll patients with risk factors for SVD (hypertension, age, diabetes, etc.), and measure serum levels of S100B protein and anti-S100B IgGs by ELISA.

26 AIM 2.2: Measure auto-immune IgG levels against a specific 10 amino acid sequence of

S100B as a more specific predictor and better surrogate of radiologically evident SVD.

Rationale: Preliminary results revealed a highly immunogenic target sequence of human

S100B. Hypothesis: The ELINNELSHF sequence in S100B is an immune trigger for autoantibody production and S100B autoimmunity in SVD.

27 1.6 References

1. Fung C, Evans E, Shin D. Hypoxic-ischemic brain injury exacerbates

neuronal apoptosis and precipitates spontaneous seizures in glucose transporter

isoform 3 heterozygous null mice. J Neurosci Res. 2010;88(15):3386-98

2. Correale J, Villa A. Cellular elements of the blood-brain barrier.

Neurochem Res. 2009;34(12):2067-77

3. Furuse M, Sasaki H, Tsukita S. Manner of interaction of heterogeneous

claudin species within and between tight junction strands. J. Cell Biol. 1999; 147

(1999), pp. 891–903

4. Haskins J, Gu L, Wittchen ES, Hibbard J, Stevenson BR. ZO-3, a novel

member of the MAGUK protein family found at the tight junction, interacts with

ZO-1 and occludin. J. Cell Biol., 141 (1998), pp. 199–208

5. Hirase T, Staddon JM, Saitou M, Ando-Akatsuka Y, Itoh M, Furuse M,

Fujimoto K, Tsukita S, Rubin LL. Occludin as a possible determinant of tight

junction permeability in endothelial cells. J. Cell. Sci., 110 (1997), pp. 1603–1613

6. Papadopoulos MC, Saadoun S, Woodrow CJ, Davies DC, Costa-Martins P,

Moss RF, Krishna S, Bell BA. Occludin expression in microvessels of neoplastic

and non-neoplastic human brain. Neuropathol. Appl. Neurobiol., 27 (2001), pp.

384–395

7. Schroeter ML, Mertsch K, Giese H, Muller S, Sporbert A, Hickel B, Blasig

IE. Astrocytes enhance radical defence in capillary endothelial cells constituting

the blood–brain barrier. FEBS Lett., 449 (1999), pp. 241–244

28 8. Seidner G, Alvarez MG, Yeh JI, O'Driscoll KR, Klepper J, Stump TS, Wang

D, Spinner NB, Birnbaum MJ, De Vivo DC. GLUT-1 deficiency syndrome caused

by haploinsufficiency of the blood-brain barrier hexose carrier. Nat

Genet. 1998;18:188–91

9. Rubin LL, Barbu K, Bard F, Cannon C, Hall DE, Horner H, Janatpour M,

Liaw C, Manning M, Morales J. Differentiation of brain endothelial cells in cell

culture. Ann. N.Y. Acad. Sci., 633 (1991), pp. 420–425

10. Macaulay N, Zeuthen T. Glial K(+) clearance and cell swelling: key roles

for cotransporters and pumps. Neurochem Res. 2012;37(11):2299-309

11. Balabanov R, Dore-Duffy P. Role of the CNS microvascular pericyte in the

blood–brain barrier. J. Neurosci. Res., 53 (1998), pp. 637–644

12. Feuerstein GZ, Liu T, Barone FC. Cytokines, inflammation, and brain

injury: role of tumor necrosis factor-alpha. Cerebrovasc. Brain Metab. Rev., 6

(1994), pp. 341–360

13. Bolton SJ, Anthony DC, Perry VH. Loss of the tight junction proteins

occludin and zonula occludens-1 from cerebral vascular endothelium during

neutrophil-induced blood–brain barrier breakdown in vivo. Neuroscience, 86

(1998), pp. 1245–1257

14. Fischer S, Wobben M, Marti HH, Renz D, Schaper W. Hypoxia-induced

hyperpermeability in brain microvessel endothelial cells involves VEGF-mediated

changes in the expression of zonula occludens-1. Microvasc. Res., 63 (2002), pp.

70–80

29 15. Long DM. Capillary ultrastructure in human metastatic brain tumors. J.

Neurosurg., 51 (1979), pp. 53–58

16. Lamszus K, Laterra J, Westphal M, Rosen EM. Scatter factor/hepatocyte

growth factor (SF/HGF) content and function in human gliomas. Int. J. Dev.

Neurosci., 17 (1999), pp. 517–530

17. Galea I, Bechmann I, Perry VH. What is immune privilege (not)? Trends

Immunol. 2007;28(1):12-8

18. Corriveau RA, Huh GS, Shatz CJ. Regulation of class I MHC gene

expression in the developing and mature CNS by neural activity. Neuron.

1998;21(3):505-20

19. Stoll G, Jander S, Schroeter M. Detrimental and beneficial effects of injury-

induced inflammation and cytokine expression in the nervous system. Adv Exp

Med Biol. 2002;513:87-113

20. Hendriks JJ, Teunissen CE, de Vries HE, Dijkstra CD. Macrophages and

neurodegeneration. Brain Res Brain Res Rev. 2005;48:185–195

21. Lucchinetti C, Bruck W, Parisi J, Scheithauer B, Rodriguez M, Lassmann

H. Heterogeneity of multiple sclerosis lesions: implications for the pathogenesis of

demyelination. Ann Neurol. 2000;47:707–717

22. Abbott NJ, Ronnback L, Hansson E. Astrocyte-endothelial interactions at

the blood-brain barrier. Nat Rev Neurosci.2006;7:41–53

23. Anderson PB, Perry VH, Gordon S. The acute inflammatory response to

lipopolysaccharide in CNS parenchyma differs from that in other body tissues.

Neuroscience. 1992;48(1):169-86

30 24. Yamada S, DePasquale M, Patlak CS, Cserr HF. Albumin outflow into deep

cervical lymph from different regions of rabbit brain. Am J Physiol. 1991 Oct;261(4

Pt 2):H1197-204

25. Kida S, Pantazis A, Weller RO. CSF drains directly from the subarachnoid

space into nasal lymphatics in the rat. Anatomy, histology and immunological

significance. Neuropathol Appl Neurobiol. 1993 Dec;19(6):480-8

26. Harling-Berg CJ, Park TJ, Knopf PM.Role of the cervical lymphatics in the

Th2-type hierarchy of CNS immune regulation. Neuroimmunol. 1999 Nov

15;101(2):111-27

27. Pashenkov M, Teleshova N, Kouwenhoven M, Smirnova T, Jin

YP, Kostulas V, Huang YM, Pinegin B, Boiko A, Link H. Recruitment of dendritic

cells to the cerebrospinal fluid in bacterial neuroinfections. J Neuroimmunol. 2002

Jan;122(1-2):106-16

28. Tsai CY, Chow NH, Ho TS, Lei HY. Intracerebral injection of myelin basic

protein (MBP) induces inflammation in brain and causes paraplegia in MBP-

sensitized B6 mice. Clin Exp Immunol. 1997;109(1):127-33

29. Kanner, A.A., Marchi, N., Fazio, V., Mayberg, M.R., Koltz, M.T., Siomin,

V., Stevens, G.H., Masaryk, T., Aumayr, B., Vogelbaum, M.A., Barnett, G.H., &

Janigro, D. 2003. Serum S100beta: a noninvasive marker of blood-brain barrier

function and brain lesions. Cancer, 97, (11) 2806-2813

30. Topakian, R., Barrick, T.R., Howe, F.A., & Markus, H.S. 2010. Blood-brain

barrier permeability is increased in normal-appearing white matter in patients with

lacunar stroke and leucoaraiosis. J.Neurol.Neurosurg.Psychiatry, 81, (2) 192-197

31 31. Wardlaw, J.M., Doubal, F., Armitage, P., Chappell, F., Carpenter, T.,

Munoz, M.S., Farrall, A., Sudlow, C., Dennis, M., & Dhillon, B. 2009. Lacunar

stroke is associated with diffuse blood-brain barrier dysfunction. Ann.Neurol., 65,

(2) 194-202

32. Logothetis, N.K. 2008. What we can do and what we cannot do with fMRI.

Nature, 453, (7197) 869-878

33. Taheri, S., Gasparovic, C., Huisa, B.N., Adair, J.C., Edmonds, E.,

Prestopnik, J., Grossetete, M., Shah, N.J., Wills, J., Qualls, C., & Rosenberg, G.A.

2011. Blood-brain barrier permeability abnormalities in vascular cognitive

impairment. Stroke, 42, (8) 2158-2163

34. Starr, J.M., Wardlaw, J., Ferguson, K., MacLullich, A., Deary, I.J., &

Marshall, I. 2003. Increased blood-brain barrier permeability in type II diabetes

demonstrated by gadolinium magnetic resonance imaging.

J.Neurol.Neurosurg.Psychiatry, 74, (1) 70-76

35. Starr, J.M., Farrall, A.J., Armitage, P., McGurn, B., & Wardlaw, J. 2009.

Blood-brain barrier permeability in Alzheimer's disease: a case-control MRI study.

Psychiatry Res., 171, (3) 232-241

36. Awad, I.A., Spetzler, R.F., Hodak, J.A., Awad, C.A., & Carey, R. 1986.

Incidental subcortical lesions identified on magnetic resonance imaging in the

elderly. I. Correlation with age and cerebrovascular risk factors. Stroke, 17, (6)

1084-1089

37. Ropele, S., Seewann, A., Gouw, A.A., van der Flier, W.M., Schmidt, R.,

Pantoni, L., Inzitari, D., Erkinjuntti, T., Scheltens, P., Wahlund, L.O., Waldemar,

32 G., Chabriat, H., Ferro, J., Hennerici, M., O'Brien, J., Wallin, A., Langhorne, P.,

Visser, M.C., Barkhof, F., & Fazekas, F. 2009. Quantitation of brain tissue changes

associated with white matter hyperintensities by diffusion-weighted and

magnetization transfer imaging: the LADIS (Leukoaraiosis and Disability in the

Elderly) study. J.Magn Reson.Imaging, 29, (2) 268-274

38. Janota, I., Mirsen, T.R., Hachinski, V.C., Lee, D.H., & Merskey, H.

1989.Neuropathologic correlates of leuko-araiosis. Arch.Neurol., 46, (10) 1124-

1128

39. Truyen, L., van Waesberghe, J.H., van Walderveen, M.A., van Oosten,

B.W., Polman, C.H., Hommes, O.R., Ader, H.J., & Barkhof, F. 1996.

Accumulation of hypointense lesions ("black holes") on T1 spin-echo MRI

correlates with disease progression in multiple sclerosis. Neurology, 47, (6) 1469-

1476

40. Fazekas, F., Ropele, S., Enzinger, C., Gorani, F., Seewann, A., Petrovic, K.,

& Schmidt, R. 2005. MTI of white matter hyperintensities. Brain, 128, (Pt 12)

2926-2932

41. O'Sullivan, M., Rich, P.M., Barrick, T.R., Clark, C.A., & Markus, H.S.

2003. Frequency of subclinical lacunar infarcts in ischemic leukoaraiosis and

cerebral autosomal dominant arteriopathy with subcortical infarcts and

leukoencephalopathy. AJNR Am.J.Neuroradiol., 24, (7) 1348-1354

42. Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. Classification and

natural history of clinically identifiable subtypes of cerebral infarction. Lancet 1991

Jun 22;337(8756):1521-6

33 43. Garde E, Mortensen EL, Krabbe K, Rostrup E, Larsson HB. Relation

between age-related decline in intelligence and cerebral white‐matter

hyperintensities in healthy octogenarians: a longitudinal study. Lancet 2000 Aug

19;356(9230):628-34

44. O'Sullivan M, Morris RG, Huckstep B, Jones DK, Williams SC, Markus

HS. Diffusion tensor MRI correlates with executive dysfunction in patients with

ischaemic leukoaraiosis. J.Neurol.Neurosurg.Psychiatry 2004 Mar;75(3):441‐7

45. Prins ND, van Dijk EJ, den HT, Vermeer SE, Jolles J, Koudstaal PJ,

Hofman A, Breteler MM. Cerebral small vessel disease and decline in information

processing speed, executive function and memory. Brain 2005 Sep;128(Pt

9):2034-41

46. Smith EE, Schneider JA, Wardlaw JM, Greenberg SM. Cerebral

microinfarcts: the invisible lesions. Lancet Neurology 2012 Mar;11(3):272-82

47. Baezner H, Oster M, Henning O, Cohen S, Hennerici MG. Amantadine

increases gait steadiness in frontal gait disorder due to subcortical vascular

encephalopathy: a double‐blind randomized placebo-controlled trial based on

quantitative gait analysis. Cerebrovasc.Dis. 2001;11(3):235-44

48. Debette S, Markus HS. The clinical importance of white matter

hyperintensities on brain magnetic resonance imaging: systematic review and meta-

analysis. BMJ 2010;341 :3666; Altmann-Schneider I, Trompet S, de Craen AJ, et

al. Cerebral microbleeds are predictive of mortality in the elderly. Stroke 2011

;42:638-644

34 49. Vuorinen M, Solomon A, Rovio S, et al. Changes in vascular risk factors

from midlife to late life and white matter lesions: a 20-year follow-up study.

Dement Geriatr Cogn Disord 2011 ;31 :119-125

50. Fazekas F, Chawluk JB, Alavi A, Hurtig HI, Zimmerman RA. MR signal

abnormalities at 1.5 T in Alzheimer's dementia and normal aging. AJR Am.J

Roentgenol. 1987 Aug;149(2):351-623

51. Wahlund LO, Barkhof F, Fazekas F, Bronge L, Augustin M, Sjogren M,

Wallin A, Ader H, Leys D, Pantoni L, et al. A new rating scale for age-related white

matter changes applicable to MRI and CT. Stroke 2001 Jun;32(6):1318-22

52. Lammie GA, Brannan F, Slattery J, Warlow C. Nonhypertensive cerebral

small-vessel disease. An autopsy study. Stroke 1997;28:2222-2229

53. O'Sullivan, M. 2008. Leukoaraiosis. Pract.Neurol., 8, (1) 26-38

54. Hassan, A., Hunt, B.J., O'Sullivan, M., Parmar, K., Bamford, J.M., Briley,

D., Brown, M.M., Thomas, D.J., & Markus, H.S. 2003. Markers of endothelial

dysfunction in lacunar infarction and ischaemic leukoaraiosis. Brain, 126, (Pt 2)

424-432

55. Bakker, S.L., de Leeuw, F.E., de Groot, J.C., Hofman, A., Koudstaal, P.J.,

& Breteler, M.M. 1999. Cerebral vasomotor reactivity and cerebral white matter

lesions in the elderly. Neurology, 52, (3) 578-583

56. Pantoni, L. & Garcia, J.H. 1997. Pathogenesis of leukoaraiosis: a review.

Stroke, 28, (3) 652-659

57. Joutel, A., Monet-Lepretre, M., Gosele, C., Baron-Menguy, C., Hammes,

A., Schmidt, S., Lemaire-Carrette, B., Domenga, V., Schedl, A., Lacombe, P., &

35 Hubner, N. 2010. Cerebrovascular dysfunction and microcirculation rarefaction

precede white matter lesions in a mouse genetic model of cerebral ischemic small

vessel disease. J.Clin.Invest, 120, (2) 433-445

58. Kearney-Schwartz, A., Rossignol, P., Bracard, S., Felblinger, J., Fay, R.,

Boivin, J.M., Lecompte, T., Lacolley, P., Benetos, A., & Zannad, F. 2009. Vascular

structure and function is correlated to cognitive performance and white matter

hyperintensities in older hypertensive patients with subjective memory complaints.

Stroke, 40, (4) 1229-1236

59. Scuteri, A., Nilsson, P.M., Tzourio, C., Redon, J., & Laurent, S. 2011.

Microvascular brain damage with aging and hypertension: pathophysiological

consideration and clinical implications. J.Hypertens., 29, (8) 1469-1477

60. Abbott, N.J. 2004. Evidence for bulk flow of brain interstitial fluid:

significance for physiology and pathology. Neurochem.Int., 45, (4) 545-552

61. Hunt, B.J. & Jurd, K.M. 1998. Endothelial cell activation. A central

pathophysiological process. BMJ, 316, (7141) 1328-1329

62. Tomimoto, H., Akiguchi, I., Wakita, H., Kinoshita, A., Ikemoto, A.,

Nakamura, S., & Kimura, J. 1996. Glial expression of cytokines in the brains of

cerebrovascular disease patients. Acta Neuropathol., 92, (3) 281-287

63. Lazarov, O., Mattson, M.P., Peterson, D.A., Pimplikar, S.W., & van, P.H.

2010. When neurogenesis encounters aging and disease. Trends Neurosci., 33, (12)

569-579

36 64. Austin, S.A., Santhanam, A.V., & Katusic, Z.S. 2010. Endothelial nitric

oxide modulates expression and processing of amyloid precursor protein. Circ.Res.,

107, (12) 1498-1502

65. Vogelbaum, M.A., Masaryk, T., Mazzone, P., Mekhail, T., Fazio, V.,

McCartney, S., Marchi, N., Kanner, A., & Janigro, D. 2005. S100beta as a predictor

of brain metastases: brain versus cerebrovascular damage. Cancer, 104, (4) 817-

824

66. Pantoni L, Inzitari 0, Pracucci G, et al. Cerebrospinal fluid proteins in

patients with leucoaraiosis: possible abnormalities in blood-brain barrier function.

J Neurol Sci 1993;115:125-131

67. Brennan C , Achey R , Whaten C , Janigro D. Imaging as a means to study

cerebrovascular pathophysiology. Brain mapping: An encyclopedic reference

2015; 139:1113-1121

68. Fritz G, Heizmann CW. 3D structures of the calcium and zinc binding S100

proteins. A. Messerschmidt, W. Bode, W. Cygler (Eds.), Handbook of

Metalloproteins, Wiley, Chichester (2004), pp. 529–540

69. Barger SW, Wolchok SR, Van Eldik LJ. Disulfide-linked S100β dimers and

signal transduction. Biochim. Biophys. Acta, 1160 (1992), pp. 105–112

70. Donato R. Functional roles of S100 proteins, calcium-binding proteins of

the EF-hand type. Biochim Biophys Acta. 1999 Jul 8;1450(3):191-231

71. Marenholz I, Heizmann CW, Fritz G. S100 proteins in mouse and man:

from evolution to function and pathology. Biochem Biophys Res Commun. 2004

Oct 1;322(4):1111-22

37 72. Brockes J. P., Fields K. L. and Raff M. C. (1979) Studies on cultured rat

Schwann cells. I. Establishment of purified populations from cultures of peripheral

nerve. Brain Res. 165, 105–118

73. Rickmann M. and Wolff J. R. (1995) S100 protein expression in

subpopulations of neurons of rat brain. Neuroscience 67, 977–991

74. Cocchia D., Michetti F. and Donato R. (1981) S100 antigen in normal

human skin. Nature 294, 85–87

75. Michetti F., Lauriola L., Rende M., Stolfi V., Battaglia F. and Cocchia

D. (1985) S-100 protein in the testis. An immunochemical and

immunocytochemical study. Cell Tissue Res. 240, 137–142

76. Michetti F., Dell’Anna E., Tiberio G. and Cocchia

D. (1983) Immunochemical and immunocytochemical study of S-100 protein in rat

adipocytes. Brain Res. 262, 352–356

77. Pham N, Fazio V, Cucullo L, Teng Q, Biberthaler P, Bazarian JJ, Janigro

D. Extracranial sources of S100B do not affect serum levels. PLoS One. 2010 Sep

10;5(9)

78. L. Santamaria-Kisiel, A.C. Rintala-Dempsey, G.S. Shaw. Calcium-

dependent and -independent interactions of the S100 protein family. Biochem. J.,

396 (2006), pp. 201–214

79. J. Lin, Q. Yang, Z. Yan, J. Markowitz, P.T. Wilder, F. Carrier, D.J. Weber.

Inhibiting S100B restores p53 levels in primary malignant melanoma cancer cells.

J. Biol. Chem., 279 (2004), pp. 34071–34077

38 80. Donato R, Sorci G, Ruzzi F, Arcuri C, Bianchi R, Brozzi F, Tubaro C,

Giambanco I. S100B's double life: intracellular regulator and extracellular signal.

Biochim Biophys Acta. 2009 Jun;1793(6):1008-22

81. Hsieh HL, Schäfer BW, Sasaki N, Heizmann CW. Expression analysis of

S100 proteins and RAGE in human tumors using tissue microarrays. Biochem.

Biophys. Res. Commun., 307 (2003), pp. 375–381

82. Arumugam T, Simeone DM, Schmidt AM, Logsdon CD. S100P stimulates

cell proliferation and survival via receptor for activated glycation end products

(RAGE). J. Biol. Chem., 279 (2004), pp. 5059–5065

83. Businaro R, Leone S, Sorci G, Donato R, Lauro GM, Fumagalli L. S100B

protects LAN-5 neuroblastoma cells against Aβ amyloid-induced neurotoxicity via

RAGE engagement at low doses but increases Aβ amyloid neurotoxicity at high

doses. J. Neurosci. Res., 83 (2006), pp. 897–906

84. Huttunen HJ, Kuja-Panula J, Sorci G, Agneletti AL, Donato R, Rauvala H.

Coregulation of neurite outgrowth and cell survival by amphoterin and S100

proteins through RAGE activation. J. Biol. Chem., 275 (2000), pp. 40096–40105

85. Hu J, Castets F, Guevara JL, Van Eldik LJ. S100β stimulates inducible nitric

oxide synthase activity and mRNA levels in rat cortical astrocytes. J. Biol. Chem.,

271 (1996), pp. 2543–2547

86. Ponath G, Schettler C, Kaestner F, Voigt B, Wentker D, Arolt V,

Rothermundt M. Autocrine S100B effects on astrocytes are mediated via RAGE. J.

Neuroimmunol. (2007), pp. 214–222

39 87. Scotto C, Deloulme JC, Rousseau D, Chambaz E, Baudier J. Calcium and

S100B regulation of p53-dependent cell growth and apoptosis. Mol Cell Biol. 1998

Jul;18: 4272-81

88. Griffin WST, Stanley LC, Ling C, White L, McLeod W, Perrot LJ, White

CL, Araoz C. Brain Interleukin 1 and S100 immunoreactivity are elevated in

Down’s syndrome and Alzheimer disease. Proc. Natl. Acad. Sci. USA, 86 (1989),

pp. 7611–7615

89. Griffin WT, Yeralan O, Sheng JG, Boop FA, Mrak RE, Rovnaghi CR,

Burnett BA, Feokistova A, Van Eldik LJ. Overexpression of the neurotrophic

cytokine S100β in human temporal lobe epilepsy. J. Neurochem., 65 (1995), pp.

228–233

90. Blyth BJ, Farahvar A, Gee CA, Hawthorn BA, Bazarian JJ (2008) Serum

S-100b concentrations are an accurate indicator of blood brain barrier integrity.

Annals of Emergency Medicine 51: 538–539

91. Marchi N, Cavaglia M, Bhudia S, Hallene K, Janigro D (2004) Peripheral

markers of blood-brain barrier damage. Clinica Chimica Acta 342: 1–12

92. Reiber H (1998) Cerebrospinal fluid—physiology, analysis and

interpretation of protein patterns for diagnosis of neurological diseases. Mult Scler

4: 99–107

93. Reiber H (2001) Dynamics of brain-derived proteins in cerebrospinal fluid.

Clin Chim Acta 310: 173–186

40 94. Marchi N, Rasmussen PA, Kapural M, Fazio V, Cavaglia M, et al. (2003)

Peripheral markers of brain damage and blood-brain barrier dysfunction.

Restorative Neurology and Neuroscience 21: 109–121

95. Michetti F., Massaro A. and Murazio M. (1979) The nervous system-

specific S100 antigen in cerebrospinal fluid of multiple sclerosis patients. Neurosci.

Lett. 11, 171–175

96. Beer C., Blacker D., Bynevelt M., Hankey G. J. and Puddey I.

B. (2010) Systemic markers of inflammation are independently associated with

S100B concentration: results of an observational study in subjects with acute

ischaemic stroke. J. Neuroinflammation 7, 71–75

97. Michetti F, Corvino V, Geloso MC, Lattanzi W, Bernardini C, Serpero L,

Gazzolo D. The S100B protein in biological fluids: more than a lifelong biomarker

of brain distress. J Neurochem. 2012 Mar;120(5):644-59

98. Aurell A., Rosengren L. E., Karlsson B., Olsson J. E., Zbornikova

V. and Haglid K. G. (1991) Determination of S-100 and glial fibrillary acidic

protein concentrations in cerebrospinal fluid after brain infarction. Stroke 22, 1254–

1258

99. Persson L., Hårdemark H. G., Gustafsson J., Rundström G., Mendel-

Hartvig I., Esscher T. and Påhlman S. (1987) S-100 protein and neuron-specific

enolase in cerebrospinal fluid and serum: markers of cell damage in human central

nervous system. Stroke 18,911–918

41 100. Michetti F., Massaro A., Russo G. and Rigon G. (1980) The S100 antigen

in cerebrospinal fluid as a possible index of cell injury in the nervous system. J.

Neurol. Sci. 44, 259–263

101. Massaro A., Michetti F., Laudisio A. and Bergonzi P. (1985) Myelin basic

protein and S-100 antigen in cerebrospinal fluid of patients with multiple sclerosis

in the acute phase. Ital. J. Neurol. Sci. 6, 53–56

102. Peskind E. R., Griffin W. S., Akama K. T., Raskind M. A. and Van Eldik

L. J. (2001) Cerebrospinal fluid S100B is elevated in the earlier stages of

Alzheimer’s disease. Neurochem. Int. 39, 409–413

103. Petzold A., Jenkins R., Watt H. C., Green A. J., Thompson E. J., Keir

G., Fox N. C. and Rossor M. N. (2003) Cerebrospinal fluid S100B correlates with

brain atrophy in Alzheimer’s disease. Neurosci. Lett. 336, 167–170

104. Rothermundt M., Falkai P., Ponath G. et al. (2004a) Glial cell dysfunction

in schizophrenia indicated by increased S100B in the CSF.Mol. Psychiatry 9, 897–

899

105. Graabe H. J., Ahrens N., Rose H. J., Kessler C. and Freyberger H.

J. (2001) Neurotrophic factor S100 beta in major

depression.Neuropsychobiology 44, 88–90

106. Gazzolo D., Marinoni E., Di Iorio R. et al. (2006) High maternal blood

S100B concentrations in pregnancies complicated by intrauterine growth restriction

and intraventricular hemorrhage. Clin. Chem. 52, 819–826

42 107. Abraha H. D., Butterworth R. J., Bath P. M., Wassif W. S., Garthwaite

J. and Sherwood R. A. (1997) Serum S-100 protein, relationship to clinical outcome

in acute stroke. Ann. Clin. Biochem. 34, 366–370

108. Raabe A., Grolms C., Sorge O., Zimmermann M. and Seifert

V. (1999b) Serum S-100B protein in severe head injury. Neurosurgery 45,477–483

109. Elting J. W., De Jager A. E., Teelken A. W., Schaaf M. J., Maurits N.

M., Van der Naalt J., Sibinga C. T., Sulter G. A. and De Keyser

J.(2000) Comparison of serum S-100 protein levels following stroke and traumatic

brain injury. J. Neurol. Sci. 181, 104–110

110. Vos P. E., Jacobs B., Andriessen T. M., Lamers K. J., Borm G. F., Beems

T., Edwards M., Rosmalen C. F. and Vissers J. L. (2010)GFAP and S100B are

biomarkers of traumatic brain injury: an observational cohort

study. Neurology 75, 1786–1793

111. Marchi N, Bazarian JJ, Puvenna V, Janigro M, Ghosh C, Zhong J, Zhu

T, Blackman E, Stewart D, Ellis J, Butler R, Janigro D. Consequences of repeated

blood-brain barrier disruption in football players. PLoS One. 2013;8(3):e56805

112. Li J. P., Lu L., Wang L. J., Zhang F. R. and Shen W. F. (2011a) Increased

serum levels of S100B are related to the severity of cardiac dysfunction, renal

insufficiency and major cardiac events in patients with chronic heart failure. Clin.

Biochem. 44, 984–988

113. Mazzini G. S., Schaf D. V., Vinadé E. R. et al. (2007) Increased S100B

serum levels in dilated cardiomyopathy patients. J. Card. Fail.13, 850–854

43 114. Marchi N, Mazzone P, Fazio V, Mekhail T, Masaryk T, Janigro D.

ProApolipoprotein A1: a serum marker of brain metastases in lung cancer patients.

Cancer. 2008 Mar 15;112(6):1313-24

115. Bouwhuis M. G., Suciu S., Kruit W. et al. (2011) Prognostic value of serial

blood S100B determinations in stage IIB-III melanoma patients: a corollary study

to EORTC trial 18952. Eur. J. Cancer 47, 361–368

116. Guo H. B., Stoffel-Wagner B., Bierwirth T., Mezger J. and Klingmüller

D. (1995) Clinical significance of serum S100 in metastatic malignant

melanoma. Eur. J. Cancer 31, 1898–1902

117. Henze G., Dummer R., Joller-Jemelka H. I., Böni R. and Burg

G. (1997) Serum S100-a marker for disease monitoring in metastatic

melanoma. Dermatology 194, 208–212

118. Missler U., Wandinger K. P., Wiesmann M., Kaps M. and Wessel

K. (1997a) Acute exacerbation of multiple sclerosis increases plasma levels of S-

100 protein. Acta Neurol. Scand. 96, 142–144

119. Chaves M. L., Camozzato A. L., Ferreira E. D., Piazenski I., Kochhann

R., Dall’Igna O., Mazzini G. S., Souza D. O. and Portela L. V.(2010) Serum levels

of S100B and NSE proteins in Alzheimer’s disease patients. J.

Neuroinflammation 7, 6–12

120. Lara D. R., Gama C. S., Belmonte-de-Abreu P., Portela L. V., Gonçalves

C. A., Fonseca M., Hauck S. and Souza D. O. (2001)Increased serum S100B protein

in schizophrenia: a study in medication-free patients. J. Psychiatr. Res. 35, 11–14

44 121. Rothermundt M., Missler U., Arolt V., Peters M., Leadbeater J., Wiesmann

M., Rudolf S., Wandinger K. P. and Kirchner H. (2001b)Increased S100B blood

levels in unmedicated and treated schizophrenic patients are correlated with

negative symptomatology. Mol. Psychiatry 6, 445–449

122. Machado-Vieira R., Lara D. R., Portela L. V., Gonçalves C. A., Soares J.

C., Kapczinski F. and Souza D. O. (2002) Elevated serum S100B protein in drug-

free bipolar patients during first manic episode: a pilot study. Eur.

Neuropsychopharmacol. 12, 269–272

123. Andreazza A. C., Cassini C., Rosa A. R. et al. (2007) Serum S100B and

antioxidant enzymes in bipolar patients. J. Psychiatr. Res. 41,523–529

124. Steiner J., Schiltz K., Walter M. et al. (2010a) S100B serum levels are

closely correlated with body mass index: an important caveat in neuropsychiatric

research. Psychoneuroendocrinology 35, 321–324

125. Gazzolo D., Bruschettini M., Lituania M., Serra G., Bonacci

W. and Michetti F. (2001b) Increased urinary S100B protein as an early indicator

of intraventricular hemorrhage in preterm infants: correlation with the grade of

hemorrhage. Clin. Chem. 47, 1836–1838

126. Liu L., Zheng C. X., Peng S. F., Zhou H. Y., Su Z. Y., He L. and Ai

T. (2010) Evaluation of urinary S100B protein level and lactate/creatinine ratio for

early diagnosis and prognostic prediction of neonatal hypoxic-ischemic

encephalopathy. Neonatology 97,41–44

45 127. Gazzolo D., Florio P., Ciotti S., Marinoni E., Di Iorio R., Bruschettini

M., Sacchi R., Serra G., Lituania M. and Michetti F. (2005a) S100B protein in urine

of preterm newborns with ominous outcome. Ped. Res. 58, 1170–1174

128. Gazzolo D., Frigiola A., Bashir M. et al. (2009) Diagnostic accuracy of

S100B urinary testing at birth in full-term asphyxiated newborns to predict neonatal

death. PLoS ONE 4, e4298

129. Michetti F., Bruschettini M., Frigiola A. et al. (2011) Saliva S100B in

professional sportsmen: high levels at resting conditions and increased after

vigorous physical activity. Clin. Biochem. 44, 245–247

130. Florio P., Michetti F., Bruschettini M., Lituania M., Bruschettini P., Severi

F. M., Petraglia F. and Gazzolo D. (2004) Amniotic fluid S100B protein in mid-

gestation and intrauterine fetal death. Lancet 364, 270–272

131. Sindic C. J., Freund M., Van Regemorter N., Verellen-Dumoulin

C. and Masson P. L. (1984) S-100 protein in amniotic fluid of anencephalic

fetuses. Prenat. Diagn. 4, 297–302

132. Gazzolo D., Bruschettini M., Corvino V., Lituania M., Sarli R., Bruschettini

P. and Michetti F. (2003a) Amniotic fluid levels of S100B protein in normal and

trisomy-21 foetuses. Clin. Chim. Acta 330, 131–133

133. Tort A. B., Portela L. V., da Purificação Tavares M., Gonçalves C. A., Netto

C., Giugliani R. and Souza D. O. (2004) Specificity and sensitivity of S100B levels

in amniotic fluid for Down syndrome diagnosis. Life Sci. 76, 379–384

46 134. Cochran A. J., Lu H. F., Li P. X., Saxton R. and Wen D. R. (1993) S-100

protein remains a practical marker for melanocytic and other tumours. Melanoma

Res. 3, 325–330

135. Srikrishna G. and Freeze H. H. (2009) Endogenous damage-associated

molecular pattern molecules at the crossroads of inflammation and

cancer. Neoplasia 7, 615–628

136. Logsdon C. D., Fuentes M. K., Huang E. H. and Arumugam

T. (2007) RAGE and RAGE ligands in cancer. Curr. Mol. Med. 7, 777–789

137. Sparvero L. J., Asafu-Adjei D., Kang R. et al. (2009) RAGE (Receptor for

Advanced Glycation Endproducts), RAGE ligands, and their role in cancer and

inflammation. J. Transl. Med. 7, 17–38

138. Yoshimura, A. Signal transduction of inflammatory cytokines and tumor

development. Cancer Sci., 97 (2006), pp. 439–447

139. Yan SS, Wu ZY, Zhang HP, Furtado G, Chen X, Yan SF, Schmidt AM,

Brown C, Stern A, LaFaille J, Chess L, Stern DM, Jiang H. Suppression of

experimental autoimmune encephalomyelitis by selective blockade of

encephalitogenic T-cell infiltration of the central nervous system. Nat. Med., 9

(2003), pp. 287–293

140. Marshak DR, Pesce S A, Stanley LC, Griffin WS. (1992) Increased S100

beta neurotrophic activity in Alzheimer’s disease temporal lobe. Neurobiol.

Aging 13, 1–7

47 141. Van Eldik LJ, Griffin WS. (1994) S100 beta expression in Alzheimer’s

disease: relation to neuropathology in brain regions.Biochim. Biophys.

Acta 1223, 398–403

142. Gromov LA, Syrovatskaya LP, Ovinova GV. Functional role of the

neurospecific S100 protein in the processes of memory. Neurosci. Behav. Physiol.,

22 (1992), pp. 25–29

143. Chaves ML , Camozzato AL, Ferreira ED, Piazenski I, Kochhann

R, Dall’Igna O, Mazzini GS, Souza DO, Portela LV.(2010) Serum levels of S100B

and NSE proteins in Alzheimer’s disease patients. J. Neuroinflammation 7, 6–12

144. Gruden MA, Davidova TB, Malisauskas M, Sewell RDE, Voskresenskaya

NI, et al. (2007) Differential neuroimmune markers to the onset of Alzheimer's

disease neurodegeneration and dementia: Autoantibodies to A beta((25-35))

oligomers, S100b and neurotransmitters. J Neuroimmunol 186: 181–192

145. Poletaev AB, Morozov SG, Gnedenko BB, Zlunikin VM, Korzhenevskey

DA (2000) Serum anti-S100b, anti-GFAP and anti-NGF autoantibodies of IgG

class in healthy persons and patients with mental and neurological disorders.

Autoimmunity 32: 33–38

146. Sorokina EG, Semenova JB, Granstrem OK, Karaseva OV, Meshcheryakov

SV, et al. (2010) S100B protein and autoantibodies to S100B protein in diagnostics

of brain damage in craniocerebral trauma in children

147. Sorci G, Giovannini G, Riuzzi F, Bonifazi P, Zelante T, et al.. (2011) The

Danger Signal S100B Integrates Pathogen- and Danger-Sensing Pathways to

Restrain Inflammation. Plos Pathogens 7

48 148. Bargerstock E, Puvenna V, Iffland P, Falcone T, Hossain M, Vetter S, Man

S, Dickstein L, Marchi N, Ghosh C, Carvalho-Tavares J, Janigro D.

Is peripheral immunity regulated by blood-brain barrier permeability changes?

PLoS One. 2014 Jul 2;9(7):e101477

149. Iffland PH, Carvalho-Tavares J, Trigunaite A, Man SM, Rasmussen P, et

al. (2013) Intracellular and circulating neuronal antinuclear antibodies in human

epilepsy. Neurobiol Dis 59: 206–219

150. Mazzone P, Chao S, Suh J. Lung cancer. In: Biller J, ed. Neurology of

Internal Medicine. Philadelphia: Lippincott, Williams & Wilkins; 2007

151. Stuschke M, Eberhardt W, Pottgen C, et al. Prophylactic cranial irradiation

in locally advanced non-smallcell lung cancer after multimodality treatment: long-

term follow-up and investigations of late neuropsychologic effects. J Clin Oncol.

1999;17(9):2700-2709

152. Mazzone PJ, Marchi N, Fazio V, Taylor JM, Masaryk T, Bury L, Mekhail

T, Janigro D. Small vessel ischemic disease of the brain and brain metastases in

lung cancer patients. PLoS.One. 2009;4(9):e7242

49

CHAPTER II

MATERIALS AND METHODS

2.1 Enzyme-Linked Immunosorbent Assay (ELISA)

The enzyme-linked immnosorbent assay (ELISA) is a laboratory technique used to measure the concentration of an analyte in solution. ELISA was developed in 1971 by

Doctors Engall, Perlman, van Weeman and Schuurs (1,2). The general steps of an ELISA involve: 1) Immobilization of antigen or antibody to solid phase, 2) Washing, 3) Blocking,

4) Incubation, 5) Addition of enzyme linked antibody, 6) Addition of colored substrate detection system, 7) Stopping reaction, 8) Absorbance reading by spectrophotometer. The most commonly used solid phase in ELISA is the 96 well microtiter plate which is supplied by a commercial manufacturer in order to provide a standard and stable product. Antigens or antibodies can be immobilized onto the solid phase by passive adsorption. Proteins are able to adsorb to the plastic as a result of hydrophobic interaction which is independent of the protein's net charge. This interaction is exploited in order to increase protein-plastic

50 binding because exposed protein residues are generally hydrophilic while internal residues are hydrophobic. The optimal antigen coating concentration should be determined by titration in a buffered solution such as PBS. Washing is another important step in ELISA used to separate unbound reagents. Washing steps are done in triplicate using a solution such as PBS or PBS with detergent (tween). In order to reduce non specific adsorption of antibody or protein to the solid phase a blocking step is performed. Non specific protein, such as bovine serum albumin (BSA), is used to coat any remaining plastic sites not bound to coated antigen or antibody. Incubation of antibody and antigen in ELISA depends upon proximity, thus during incubation periods plates are typically placed on a shaking surface to ensure mixing and contact of antibody-antigen. An enzyme linked antibody is crucial to the ELISA method. The quantification of results are made possible by the production of colored product from the reaction of substrate and enzyme linked antibody. An example of this is the horseradish peroxidase (HRP) hydrogen peroxide substrate system. HRP is the enzyme that catalyzes a reaction with the substrate that causes a detectable colored product.

Ortho-phenylenediamine (OPD) is a water soluble substrate for HRP. The product produced by the HRP-OPD reaction is yellow-orange and detectable at 490nm by spectrophotometer. The final steps of an ELISA include stopping the enzymatic reaction and reading the absorbance of the colored product. Reactions are stopped using a specific time for an individual assay, usually within 30 minutes after the addition of substrate.

Strong acids or bases are used to stop the enzymatic activity, for the HRP-OPD system sulfuric acid is commonly used as a stopping solution. The reading of the ELISA plate is done using a spectrophotometer which transmits light at a specific wavelength through the

51 colored product measuring the amount of light absorption. Once absorbance values are obtained data interpretation can be performed.

There are numerous advantages to using ELISA including the adaptability of the technique to allow for the development of specific methods in the detection of different analytes of interest. ELISA tests are rapid and sensitive making them ideal for diagnostic purposes. ELISA is a common diagnostic tool in medicine as seen with the human immunodeficiency virus (HIV) test in which serum antibody concentrations against the virus are measured (3). Analytes commonly measured by ELISA include proteins, peptides and antibodies. ELISA is dependent upon the specificity of antibodies used in the test.

Commercially available antibodies are produced in animals against an injected antigenic stimuli. Antibodies can be used in ELISA for the detection of antigens, specifically proteins and peptides, demonstrated by the measurement of serum levels of S100B protein in this study. Conversely, antibodies against a specific antigen can be measured by indirect ELISA

(see section 2.1.1), used to measure autoantibodies in human sera against putative antigen

S100B.

2.1.1 Indirect ELISA

The indirect ELISA is used to measure antibodies against an antigen of interest in solution. In this type of ELISA the antigen is immobilized onto the solid phase and targeted by the addition of detection antibodies (not enzyme linked). The bound detection antibodies are then targeted by enzyme linked antibodies against immunoglobulins. The enzyme linked antibodies must be against the species of the detection antibody. For example if the detection antibodies are human IgG then the enzyme linked antibodies must be anti-human 52 IgG. Substrate is then added to the bound conjugate and color develops, which is then stopped and read by a spectrophotometer.

Indirect ELISA is used in diagnostic applications in order to examine antibody levels in human sera, however, there is no generally accepted to express ELISA results.

Endpoint titer is commonly used in which the highest dilution that results in an absorbance value greater than two standard deviations of the negative control is reported. Yet samples tested at a single dilution can be expressed as direct absorbance values of the test samples.

Using single dilutions permits testing of large numbers of samples at the same time and since each sample is only tested at a single dilution, not at multiple dilutions, less material is required compared with the end point titer method. Indirect ELISA systems at a single dilution are used to measure anti-S100B autoantibody levels in this study.

2.2 S100B Protein Measurement

The commercially available ELISA was used to measure S100B protein levels in human sera (Diasorin, Stillwater MN #364701). The S100B ELISA is a two-site, one-step, enzyme linked immunosorbent assay. In the assay calibrators, controls and unknown samples react simultaneously with two solid phase capture antibodies and a detector antibody conjugated with HRP during the incubation in the microtiter wells. After a washing step a TMB chromogen (Tetramethylbenzidine) is added and the reaction is allowed to proceed for 15 minutes. The enzyme reaction is stopped by adding a Stop

Solution and the absorbance is measured at 450 nm. The limit of detection is 0.03 ng/mL

53 2.3 Anti-S100B Monomer Autoantibody Measurement

The indirect ELISA used to measure anti-S100B autoantibodies was developed and published by our lab (4). First, 96 well plates were coated with a PBS solution containing

S100B protein (human brain, catalog number-559291, EMD Chemicals). In this study we refer to this S100B antigen coating as the S100B monomer (autoantibody levels reported in Chapter III and Chapter IV). This source of S100B is considered to be in the monomeric form. This was confirmed by western blot analysis performed by the manufacturer and western blot analysis performed by us, both of which yield a single protein band at 12kDa, corresponding to the molecular weight of a single S100B monomer. Western blot analysis was performed as native polyacrylamide gel electrophoresis (PAGE) in which a non reducing and non denaturing sample buffer (in the absence of beta mercaptoethanol) was used in order to maintain the protein secondary structure. Two monoclonal antibodies produced in mouse against S100B were tested for reactivity with this S100B source

(Meridian Life Science # Q86610M and Sigma #S2657), both revealed a signal band at

12kDa again suggesting the presence of S100B in the monomeric form, see Figure 3.

54

Figure 3: Western blot analysis of S100B using two anti-S100B monoclonal antibodies. Western blot was performed on varying concentrations of S100B protein

(human brain, catalog number-559291, EMD Chemicals). Two anti-S100B monoclonal antibodies made in mouse were analyzed for binding to S100B at a dilution of 1:500

(Meridian Life Science Inc # Q86610M and Sigma # S2657).

Optimization of this ELISA was achieved by testing two concentrations of S100B protein (1 or 5 µg per well in a total volume of 100uL). No significant differences in sample signal were observed between the two concentrations of S100B coating suggesting that 1ug is sufficient to bind antibodies in human sera and at 5ug unbound antigen is being washed away. An S100B monoclonal antibody (catalog number: Q86610M, Meridian Life Science

Inc.) was used as a standard. Standard curves were obtained with 100 µl of serially diluted

55 S100B monoclonal antibodies run in duplicate. Figure 4 shows four standard curves from this study.

Plate 1 Plate 2 Plate 3 0.30 Plate 4

0.25

0.20

0.15

0.10 Absorbance Units (490nm) 0.05

0.00 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Anti-S100B Monocloncal Antibody Concentration (ug/mL)

Figure 4: Anti-S100B Monomer Autoantibody ELISA Standard Curves. Plates were coated with 1ug per 100 uL per well of S100B protein (EMD Chemicals #559291) and serial dilutions of monoclonal antibody to S100B were added as primary antibodies

(Meridian Life Science # Q86610M). Antibody concentrations are reported as ug/mL and

ELISA signal as average absorbance units (AU) at 490nm. Standard curve results are reported for a total of four ELISA plates performed in this study.

56 Next, plates were coated overnight at 4oC with S100B protein (1 µg/ 100uL/ per well). Wells were then washed 3 times with PBS. Subsequently, 100 µL of a 1% BSA blocking solution was added in each well and incubated for 2 hours at room temperature.

Wells were then washed 3 times with 200 µL of PBS containing 0.05% Tween 20. Serum samples and standards were added and incubated for 1 hour at room temperature. Serum samples were added at a single dilution of 1:5000 in PBS. Samples were aspirated and wells washed 3 times using 200 µL of PBS containing 0.05% Tween-20. A secondary antibody solution of 200 µL of horseradish-peroxidase (HRP) conjugated goat anti-mouse

IgG at a dilution of 1:5000 was added to standard samples. A secondary antibody solution of 200 µL of HRP goat anti-human at a dilution of 1:5000 was added to serum samples.

After a 1 hour incubation at room temperature wells were washed 3 times with 200 µL of

PBS containing 0.05% Tween-20. Finally, 100 µL of OPD solution was added and the reaction incubated for 30 minutes at room temperature. The reaction was stopped by adding

100 µL of 2.5 M sulfuric acid. Samples were analyzed using an ELISA plate reader at 490 nm.

2.4 Anti-S100B Multimer Autoantibody Measurement

The anti-S100B multimer autoantibody measurement in human sera was performed by indirect ELISA. A second source of S100B was obtained from Sino Biological (10181-

H07E). This protein is different from the monomeric form in Section 2.3 as it is a recombinant protein expressed by E. Coli and not isolated from human brain tissue. The manufacturer reports that the protein was purified by Ni-His tag affinity chromatography and has a molecular weight of 12kDa confirmed by SDS PAGE with a purity of 98.1%. 57 Western blot analysis was performed as a native PAGE in which a non reducing and non denaturing sample buffer was used in order to maintain the protein secondary structure. No signal was observed with either monoclonal S100B autoantibody (Meridian Life Science

# Q86610M and Sigma # S2657). The anti-S100B monomer autoantibody ELISA system was used in order to test the binding of this new source of S100B to anti-S100B monoclonal antibodies. The only change in protocol included the use of the new S100B source as the antigen coating at a concentration of (1 µg per well in a total volume of 100uL). Serial dilutions of monoclonal antibody to S100B were added as primary antibodies as in Figure

2 yet no signal was observed for either antibody.

Silver stain analysis revealed multiple bands at 12, 24, and 36 kDa, see Figure 5.

These results suggest that S100B is present in the sample in multiple forms including the monomer, dimer, and trimer forms, hence we refer to this S100B antigen coating as the

S100B multimer.

58 trimer

dimer

monomer

Figure 5: Silver stain of S100B Multimer. Native PAGE was run. Bio-Rad silver stain plus (#161-0449).Silver stain used to detect protein in PAGE gels after electrophoresis.

Experiment performed as per manufacturer instruction. Shown is the lane containing 5ug of S100B (Sino Biological (10181-H07E). Arrows indicate bands at 12, 24, and 36kDa corresponding to monomer, dimer, and trimer forms of S100B.

Since neither monoclonal antibody showed a measureable signal by western blot, mass spectrometry (MS) analysis was performed to confirm the identity of the commercially available recombinant protein S100B. MS analysis was performed by the core facility at Cleveland Clinic. The sample was in-solution tryptic digested and desalted by C18 solid phase extraction (SPE). The goal of MS is to identify protein S100B in the sample and confirm the protein sequence. For the protein digestion, the sample was precipitated by cold acetone and resuspended in Tris-HCl buffer containing 6M urea. The

59 samples were then reduced with dithiothreitol (DTT) and alkylated with iodoacetamide prior to digestion. Trypsin was added at 1:20 (trypsin : protein) ratio, and the sample was incubated overnight at room temperature to achieve complete digestion. The protein digest was then desalted using C18 SPE method and reconstituted in 30µL 1% acetic acid for LC-

MS analysis. The LC-MS system was a Thermo Finnigan Linear Trap Quadropole linear ion trap mass spectrometer system. The HPLC column was a self-packed 9 cm x 75 µm id

Phenomenex Jupiter C18 reversed-phase capillary chromatography column. Ten µL volumes of the extract were injected and the peptides eluted from the column by an acetonitrile/0.05 M acetic acid gradient at a flow rate of 0.3 µL/min were introduced into the source of the mass spectrometer online. The microelectrospray ion source is operated at 2.5 kV. The samples were analyzed using The digest was analyzed using the data dependent multitask capability of the instrument acquiring full scan mass spectra to determine peptide molecular weights and product ion spectra to determine amino acid sequence in successive instrument scans. The data was analyzed by using all Collision

Induced Dissociation spectra collected in the experiment to search the human reference sequence databases with the search program Mascot and also searched against human

S100B protein sequence using Sequest program. A search of the full human reference sequence database identified the most abundant component of this sample to be the protein of interest, S100B, by the presence of 10 unique peptides covering 91% of the protein sequence. Sequest searching also resulted 10 unique peptides covering 91% of protein

S100B sequence.

Given that the purity of the S100B is 91% the lack of binding of the monoclonal antibody to S100B may be explained by the epitope placement. It is possible that the

60 commercial monoclonal antibodies are directed against an epitope not present in the 91% pure S100B protein sample. Ultimately the use of a standard curve was unable to be generated using monoclonal anti-S100B antibodies as in the anti-S100B monomer autoantibody ELISA (Section 2.3). All other parameters of the anti-S100B monomer autoantibody ELISA protocol in Section 2.3 remained unchanged in the anti-S100B multimer autoantibody ELISA in the measurement of serum levels (Chapter IV).

2.5 Anti-S100B Peptide Autoantibody Measurement

The anti-S100B peptide autoantibody measurement in human sera is performed by indirect ELISA. This ELISA was developed by optimizing the protocol published and described in 2.3 (4). Preliminary bioinformatic analysis revealed a highly immunogenic target sequence of human S100B, ELINNELSHF, which is not found in any other human protein. We hypothesize that this sequence in S100B is a potential immune trigger for autoantibody production in cerebral small vessel disease. The ELINNELSHF sequence was synthesized by solid phase peptide synthesis at the proteomics core at Cleveland Clinic

(See Chapter IV section 4.3.5). The synthesized peptide sequence was used as the antigen coating in the indirect ELISA for the measurement of anti-S100B peptide autoantibodies

(serum levels reported in Chapter IV). First, 96 well plates were coated with a PBS solution containing ELINNELSHF (1 µg per well in a total volume of 100uL). Wells were then washed 3 times with PBS. Subsequently, 100 µL of a 1% BSA blocking solution was added in each well and incubated for 2 hours at room temperature. Wells were then washed 3 times with 200 µL of PBS containing 0.05% Tween 20. Serum samples and standards were

61 added and incubated for 1 hour at room temperature. Optimal sample dilution was determined by titer experiments. Serial dilutions of serum samples were performed and titer was chosen at a dilution factor at which an absorbance output was still distinguishable between groups, see Figure 6. A dilution of 1:5000 was chosen for the patient population because it is between 1:4000 and 1:8000 serial dilutions, dilutions higher than 1:8000 begin to show an overlap in signal suggesting the sample is too dilute to measure the autoantibody levels of interest.

62 SVD- 0.7 SVD+ SVD - SVD+ 0.6 SVD- SVD-

0.5

0.4

0.3

Absorbance Units (490nm) 0.2

0.1

1/1000 1/2000 1/4000 1/8000 1/16000 1/32000 1/64000 Serum Dilution

Figure 6: Serum Titrations for optimization of anti-S100B peptide autoantibody

ELISA. Serum titration curves for six patients enrolled in the study are shown. The diagnosis of cerebral small vessel disease (SVD) is noted in the key. Serum samples were serially diluted in PBS. A protein coating of ELINNELSHF (1 µg per well in a total volume of 100uL) was used. A secondary conjugated antibody (goat anti-human IgG

HRP Calbiochem # 401445) at a concentration of 1:5000 was used.

63 Samples were aspirated and wells washed 3 times using 200 µL of PBS containing

0.05% Tween-20. The selection of secondary antibody was based on the commercially available antibody concentration that yielded the highest signal at a chosen concentration.

Two different goat anti-human IgG HRP conjugated secondary antibodies were tested

(Calbiochem # 401445 and Milipore # AP309P), see Figure 7. On average the Calbiochem conjugated antibody showed the highest signal over the antibody dilution range. Titer was used in order to chose a dilution factor, a dilution of 1:5000 was chosen as it is between the 1:4000 and 1:8000 dilutions and after 1:8000 the absorbance unit signals begin to get closer to one another although the titration curves never intersect. An antibody concentration of 1:5000 in a volume of 200 µL of goat anti-human HRP conjugated antibody (Calbiochem #401445) were added to sample wells.

64 Calbiochem Milipore 1.4

1.2

1.0

0.8

0.6

0.4 Absorbance Units (490nm)

0.2

0.0 1/1000 1/2000 1/4000 1/8000 1/16000 1/32000 1/64000 Antibody Dilution

Figure 7: Conjugated Anti-Human IgG Antibody Titration. Conjugated antibody titration curves for two different HRP conjugates goat anti-human IgG antibodies are shown (Calbiochem # 401445 and Milipore # AP309P). Antibody solutions were serially diluted in blocking buffer (1% BSA in PBS). A protein coating of ELINNELSHF (1 µg per well in a total volume of 100uL) and serum sample (diluted 1:5000 in PBS) were used.

65 After a 1 hour incubation at room temperature wells were washed 3 times with 200

µL of PBS containing 0.05% Tween-20. Finally, 100 µL of OPD solution was added and the reaction incubated for 30 minutes at room temperature. The reaction was stopped by adding 100 µL of 2.5 M Sulfuric acid. Samples were analyzed using an ELISA plate reader at 490 nm.

Unlike the anti-S100B monomer autoantibody ELISA (Section 2.3) the use of a standard curve was unable to be generated using monoclonal anti-S100B antibodies. The monoclonal anti-S100B antibodies that displayed binding in the anti-S100B monomer autoantibody ELISA did not display binding to the peptide. There are many reasons why binding of the monoclonal antibodies to the S100B ELINNELSHF sequence is not observed. There is a large amount of sequence homology between the ELINNELSHF sequence and the mouse proteome, thus the mouse is unable to produce antibodies against this sequence as it is not perceived as "non-self". This was demonstrated in two ways. First, two anti-S100B monoclonal antibodies made in mouse (Meridian Life Science Inc #

Q86610M and Sigma # S2657) were analyzed by the anti-S100B peptide autoantibody

ELISA and the anti-S100B monomer autoantibody ELISA, see Figure 8. Secondly, the synthesized peptide sequence was used to immunize mice, a blood sample was taken from these mice and analyzed by the anti-S100B peptide autoantibody ELISA to determine if antibodies against the peptide sequence were produced. Neither the commercially available nor the immunized mouse serum showed measurable signal in the indirect ELISA setting and therefore could not be used as a standard curve for the proposed test.

66 Protein Prot ein 0.22 Peptide Peptide 0.7 0.20 0.6 0.18 0.5 0.16

0.4 0.14

0.3 0.12

0.10 0.2 Absorbance (490 nm) (490 Absorbance nm) (490 Absorbance

0.08 0.1

0.06 0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Concentration (ug/ml) Concentration (ug/ml)

Sigma Monoclonal Antibody S100B Merdian Monoclonal Antibody S100B

Figure 8: Monoclonal anti-S100B antibody binding to ELINNELSHF. Wells were coated with either S100B protein (EMD Chemicals #559291 at a concentration of 1 µg per well in a total volume of 100uL), seen in red, or S100B peptide (concentration of 1 µg per well in a total volume of 100uL), seen in black. Antibody solutions were serially diluted in

PBS. A secondary HRP conjugated goat anti-mouse IgG antibody solution diluted 1:5000 in blocking buffer (1% BSA in PBS) was used.

Another approach that was used in order to generate a standard curve was the use of pooled human IgG. The use of pooled human sera from patients with positive and negative disease diagnosis of the disease of interest are commonly used in the validation of diagnostic assays in order to determine cutoff values for the test (5). Yet it is difficult to obtain a sufficient quantity of sera needed for this type of validation in a small research study. A commercially available source of pooled IgG from human sera (Sigma #I4506) was analyzed for binding to the proposed peptide. These pooled IgG are heterogeneous and

6 7 of unknown disease state. The IgG was serially diluted 1:50, 1:500, 1:1000, 1:5000,

1:10000 with PBS. These samples were used in the anti-S100B peptide autoantibody

ELISA to measure the binding of pooled IgG to ELINNELSHF, see Figure 9. A titration curve displayed binding of IgGs to the peptide in a dose dependent manner. The higher the concentration of IgG the higher the absorbance unit signal.

Pooled Human IgG Titration 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2

Absorbance Units (490nm)Absorbance 0 0 0.0005 0.001 0.0015 0.002 0.0025 Dilution factor (1/dilution)

Figure 9: Pooled Human IgG Titration. Wells were coated with S100B peptide

(concentration of 1 µg per well in a total volume of 100uL). Pooled human IgG solutions were serially diluted in PBS. A secondary HRP conjugated goat anti-human IgG antibody

(Calbiochem # 401445) solution diluted 1:5000 in blocking buffer (1% BSA in PBS) was used. The plot reports IgG concentrations as 1/dilution factor including; 1:500, 1:1000,

1:5000, 1:10000. A 1:50 dilution factor was also measured but absorbance units overflowed, and were too high to detect. The linear fit of the titration curve has an R2 value of 0.9948.

68 Since the development of a uniform standard was unable to be generated for the anti-S100B peptide autoantibody ELISA the use of positive and negative controls were investigated. Negative controls are used to detect possible background. A blank was also run to determine the absorbance of the plastic plate. Blocking solution (1% BSA in PBS) was also measured in to determine possible non specific binding of conjugated goat anti- human IgG antibody. Negative controls were measured in the absence of peptide coating but in the presence of conjugated goat anti-human IgG antibody. Also negative controls were measure in the presence of peptide coating but in the absence of conjugated goat anti- human IgG antibody. See Table 2. Note all absorbance values are low background levels in a range of 0.037-0.067 absorbance units. The use of patient serum samples also provided a means of positive and negative controls. Specific samples were considered "high" patient sera and "low" patient sera and were consistently measured as an internal standards.

Average Absorbance Units (490nm)

- peptide coating + conjugated goat anti-human IgG 0.051

+ peptide coating - conjugated goat anti-human IgG 0.061

Blank 0.037

BSA 0.039

Table 2: Negative Controls anti-S100B peptide autoantibody ELISA. Negative controls were evaluated in ELISA plates, average absorbance units are reported.

69 2.6 References

1. Engvall E, Perlman P. Enzyme-linked immunosorbent assay

(ELISA) Quantitative assay of immunoglobulin G

Immunochemistry. 1971;8(9):871–4

2. Van Weemen BK, Schuurs AH. Immunoassay using antigen-enzyme

conjugates. FEBS Lett. 1971;15(3):232–236

3. Fearon M. The laboratory diagnosis of HIV infections. The Canadian

Journal of Infectious Diseases & Medical Microbiology. 2005;16(1):26-30

4. Marchi N, Bazarian JJ, Puvenna V, Janigro M, Ghosh C, Zhong J, Zhu T,

Blackman E, Stewart D., Ellis J, Butler R, Janigro D (2013) Consequences of

repeated blood-brain barrier disruption in football players. Plos One 8: e56805

5. Jacobson, RH. Validation of serological assays for diagnosis of infectious

disease. Rev. Sci. Tech. 1998;17:569-526

70

CHAPTER III

S100B AND ANTI-S100B AUTOANTIBODIES AS BIOMARKERS FOR EARLY DETECTION OF BRAIN METASTASES IN LUNG CANCER

3.1 Abstract

Significance: The astrocytic protein S100B is a serum biomarker of blood-brain barrier permeability. Prolonged blood-brain barrier permeability, due to the presence of brain metastases, has the potential to allow for the production of autoantibodies against S100B in human sera. We studied the clinical utility of serum S100B protein and anti- S100B autoantibody levels for the detection of brain metastasis in patients with lung cancer.

Methods: A total of 128 untreated patients were enrolled. Eighteen patients (14%) had brain metastasis at the time of lung cancer diagnosis. Brain imaging was performed as part of routine evaluation. Serum S100B protein and anti-S100B autoantibody levels were measured by ELISA.

Results: The clinical utility of these potential biomarkers was calculated by ROC analysis.

S100B serum protein levels had a sensitivity of 89% and specificity of 43% for brain

71 metastasis at a threshold of 0.058 ng/ml. The combinatory panel of S100B and anti-S100B autoantibody levels (at a threshold of 2.0 AU) had a sensitivity of 89%, and an increased specificity of 58%.

Conclusion: The measurement of serum S100B and anti-S100B autoantibody levels can serve as a non invasive and inexpensive diagnostic tool for the detection of brain metastases in lung cancer.

72 3.2 Introduction

The identification of brain metastases have important implications for treatment and prognosis of lung cancer. At time of diagnosis 10-15% of patients with lung cancer have brain metastases. The brain is a frequent site for progression of lung cancer because the blood-brain barrier (BBB) shelters the central nervous system (CNS) from systemic treatment. In NSCLC, 30% of patients fail in the brain as the first site of failure (1).

Contrast enhanced magnetic resonance imaging (MRI) is the gold standard for the detection of brain metastases (2). Guidelines suggest obtaining brain imaging at presentation in asymptomatic lung cancer patients with; advanced stage non-small cell lung cancer, all patients with small cell carcinoma, and anyone with symptoms that could be related to the presence of brain metastases (3,4). Longitudinal follow-up MRI scans of patients who are at risk for developing brain metastases is not routinely performed, primarily due to the cost of imaging (5-7). There are several treatment options for patients diagnosed with brain metastases from lung cancer, including corticosteroid therapy, whole brain radiation therapy (WBRT), and stereotactic radiosurgery (SRS). The choice of treatment is dependent on multiple factors including the patient`s performance status, control of the primary site, and the number and size of brain metastases (8).

The prediction of patients with brain metastasis could reduce the amount of unnecessary brain imaging, decrease costs and improve patient care. A blood based biomarker for the identification of the presence of brain metastases could be used as a tool for staging, prompting further characterization with brain imaging when positive. In

73 addition, the test could be used in a longitudinal setting in lieu of MRI for the early detection of brain metastases.

Elevated serum S100B levels have been reported in lung cancer patients with metastases (9,10). Studies have confirmed an association between elevated serum S100B and positive contrast enhancement in MRI scans in patients affected by metastatic brain disease (10,11). S100B has also been identified as a prognostic factor of brain metastasis in melanoma (12). The neovascularisation associated with the growth of metastatic lesions compromises the integrity of the BBB allowing for the extravasation of brain specific proteins into peripheral circulation (13). S100B was first isolated from bovine brain in 1965 and is a member of the EF hand calcium binding protein subgroup. S100B is expressed in the astrocytes and in Schwann cells of the peripheral nervous system. Extra cranial sources of S100B exist in melanocytes, adipocytes, and chondrocytes.

Increased serum S100B levels may also be found in patients with SVD (10).The population of older patients who are at risk of developing cancer often have comorbidities for cerebrovascular disease. Therefore SVD is a frequent finding on brain imaging performed for lung cancer staging. It is important to distinguish the SVD from brain metastases. We hypothesize that anti-S100B autoantibody levels may help to distinguish the elevation of serum S100B caused by SVD from that caused by brain metastases since the production of antibodies may be affected by the duration of exposure to S100B. The aim of our study was to verify that levels of S100B protein are elevated in lung cancer patients with brain metastases and determine if the addition of anti-S100B autoantibody levels can improve the accuracy of this biomarker.

74 3.3 Methods

3.3.1 Patient Enrollment

This prospective cohort study enrolled patients seen at the Cleveland Clinic from

2010 and 2012. We included adult patients > 18 years of age at the time of diagnosis of lung cancer, with or without neurological symptoms. The study was approved by the

Institutional Review Board (IRB#10-521) and all patients signed an informed consent. A total of 145 patients with lung cancer were enrolled. Demographic information including smoking history, comorbidities, anthropometric measurements, serum creatinine and brain imaging from the electronic medical record were collected. Creatinine clearance was calculated using the Cockcroft-Gault equation. Tumor histology was assessed by pathologists and staging was classified based on the American Joint Committee on Cancer

7th edition of tumor, node, metastasis staging criteria (14).

3.3.2 Serum Measurements of S100B and Anti-S100B Autoantibodies by ELISA

Serum S100B protein and anti-S100B autoantibodies were measured in each patient. S100B protein was measured using a commercially available, monoclonal, 2-site immunoluminometric assay from Diasorin (Stillwater, MI) (10,15,16).

A reverse ELISA was developed to detect anti-S100B autoantibodies (17). For the reverse ELISA 96 well plates were coated with PBS solution containing S100B protein

(human brain, catalog number 559291, EMD Chemicals). Optimization of this ELISA was achieved by testing two concentrations of S100B protein (1 or 5µg per well). No significant differences were observed at these two concentrations of S100B coating. Plates were coated overnight at 4oC with S100B protein (1 µg/well). Wells were then washed 3 times 75 with PBS. Subsequently, 100 µL of a 1% BSA blocking solution was added in each well and incubated for 2 hours at room temperature. Wells were then washed 3 times using

200µL of PBS containing 0.05% Tween-20. Serum samples and standards were added and incubated for 1 hour at room temperature. An S100B monoclonal antibody (catalog number: Q86610M Meridian Life Science Inc.) was used as a standard to allow for the conversion of absorbance unit values into concentration. Standard curves were obtained with 100µL of serially diluted S100B monoclonal antibody. A secondary antibody solution of 200µL of horseradish-peroxidase (HRP) goat anti-mouse IgG and 200µL of HRP goat anti-human were added to the standards and serum samples respectively. After 1 hour incubation at room temperature wells were washed 3 times with 200 µL of PBS containing

0.05% Tween-20. Finally, 100 µL of OPD solution was added and the reaction incubated for 30 minutes at room temperature. The reaction was stopped by adding 100 µL of 2.5M sulfuric acid. Samples were analyzed using an ELISA plate reader at 490nm.

3.3.3 Neuroimaging

Magnetic resonance imaging (MRI) or computed tomography scan (CT), was performed as part of routine care. Images were interpreted independently by an expert neuroradiologist who was blinded to the S100B and S100B autoantibody levels, and clinical information. Brain scans were evaluated for the presence of brain metastases and for white matter changes related to SVD. The Age Related White Matter Changes

(ARWMC) Rating Scale was used to grade the white matter changes (18). White matter changes on MRI were defined as ill-defined hyperintensities ≥5 mm on both T2 and FLAIR images, and on CT as ill-defined and moderately hypodense areas of ≥5 mm. Lacunes were

76 defined as well-defined areas of >2 mm with attenuation (on CT) or signal characteristics

(on MRI) the same as cerebrospinal fluid. If lesions with these characteristics were ≤2 mm, they were considered perivascular spaces. Changes in the basal ganglia were rated in the same way (18).

3.3.4 Statistical Analysis

Statistical analysis was performed using SPSS (version 13, Chicago-IL), JMP, and

OriginLab. Parametric statistical tests including ANOVA were used to compare populations. A difference with p<0.05 was considered statistically significant. Pairwise multivariable analyses were performed in order to determine demographic effect correlations within a population. Sensitivity, specificity and accuracy of S100B and anti-

S100B autoantibodies both individually and as a combinatory test were calculated to predict brain metastasis in patients with lung cancer. Logistic regression and Receiver

Operating Characteristic (ROC) curve analysis were used to determine the optimal cutoff values.

77 3.4 Results

3.4.1 Demographics

Table 3 shows the demographics of the patients enrolled in this study. We enrolled

145 patients newly diagnosed with lung cancer. Of these, 128 (88.3%) were included in our analysis as 17 (11.7%) patients had incomplete clinical or radiographic information.

Of the 128 patients included in analysis, 109 patients had brain imaging; 85 (78%) patients had an MRI and 24 (22%) had a CT scan. The majority of patients were male (60.2%), of

Caucasian race (87.5%), and had a positive history of smoking (95.3%, either current or former). The mean age was 67.9 + 10.5 years.

78 % or Demographics n or mean SD Male 77 60.2 Age (years) 67.9 10.5 BMI 28 5.8 Creatinine Clearance (ml/min) 89 31.9

Race Caucasian 112 87.5 African-American 15 11.7 Asian 1 0.8

Smoking History never 6 4.7 current 37 28.9 former 85 66.4

Comorbidities Diabetes 18 14.1 Hypertension 82 64.1 Hyperlipidemia 56 43.8 Stroke 9 43.8 Coronary Artery Disease 19 14.8 COPD 63 49.2

Table 3: Lung cancer patient demographics. Characteristics of all lung cancer patients enrolled in this study (n=128). Results are given as mean ± standard deviation for continuous data, or numbers (%) for categorical data.

79 Characteristics of lung cancer staging, histology and neuroimaging are reported in

Table 4. Forty three (33.5%) patients were diagnosed at stage IIIA-B and 38 (29.7%) stage

IV. The most common histological subtype was adenocarcinoma (47.7%). Nineteen patients did not have brain imaging during initial staging. These patients either had negative brain imaging during follow up or completed a clinical follow-up of at least 2 years with no signs of recurrent or metastatic disease. Thus they were included in the group negative for brain metastases. Of these, 15 patients were diagnosed at stage I-II, and 4 patients at stage IIIA-B. These patients were not included in the analysis of SVD.

80 Characteristics n % Final Stage with brain imaging IA 22 17.2 IB 10 7.8 IIA 3 2.3 IIB 8 6.3 IIIA 30 23.4 IIIB 13 10.2 IV 38 29.7 Incomplete 4 3.1

Histology Adenocarcinoma 61 47.7 Squamous 40 31.3 Small Cell 13 10.2 Other 14 10.9

Neuromaging Brain Metastasis 18 14 SVD 69 63.3

Table 4: Cancer staging, histology, and neuroimaging characteristics. Results are given as numbers and % of total population (n=128).

81 Lung cancer patient demographics were then stratified by the presence of brain metastasis as shown in Table 5. There is no statistically significant contribution of demographics on the presence of brain metastasis.

-Met +Met Characteristics (n=110) (n=18) p Male 67 10 0.67 Age (years) 68 67 0.79 BMI 27.7 26.2 0.28 CrCl (ml/min) 87.7 90.6 0.71

Race Caucasian 88% 83% 0.62 Afican-American 11% 17% 0.55 Asian 1% 0 n/a

Smoking History never 3.60% 11% 0.09 current 30% 22.20% 0.5 former 66.40% 66.80% 0.98

Comorbidities Diabetes 15.40% 5.50% 0.26 Hypertension 66.30% 50% 0.18 Hyperlipidemia 45.40% 33% 0.34 Stroke 5.40% 16.70% 0.07 Coronary Artery Disease 15.40% 11.10% 0.63 COPD 48.20% 55.50% 0.56

Table 5: Brain metastases demographics. Results are given as percent of subpopulation.

Comparisons are made between met+ patients and met- patients. ANOVA was performed and statistical significance defined as p<0.05.

82 3.4.2 Demographic Effects of Serum S100B and Anti-S100B Autoantibodies

A pairwise multivariable analysis was performed for each biomarker to reveal the potential effects of demographic data on serum levels. For S100B statistically significant correlations were observed for race, creatinine clearance, and age (Table 6). Anti-S100B autoantibody levels correlated only with smoking history (Table 7).

Variable Correlation P race (caucasian, afican american, asian) -0.24 0.0019 creatinine clearance -0.24 0.0037 Age 0.17 0.03 gender -0.12 0.13 hyperlipidemia 0.10 0.24 hypertension 0.08 0.33 diabetes 0.07 0.41 smoking (current, former, never) -0.07 0.44 Stroke 0.06 0.48 COPD -0.05 0.55 BMI -0.04 0.66 Coronary Artery Disease -0.02 0.80

Table 6: Multivariable analysis of demographic effects on serum S100B. Results of pairwise multivariable analysis of demographic data variables and S100B serum levels.

Analysis was performed using JMP 9.0 software. Statistical significance is defined as p<0.05 for correlations between variables.

83 Variable Correlation P smoking (current, former, never) -0.18 0.04 race (caucasian, afican american, asian) -0.14 0.07 creatinine clearance 0.08 0.32 hypertension 0.05 0.56 COPD -0.05 0.59 hyperlipidemia -0.05 0.59 BMI -0.04 0.61 Coronary Artery Disease 0.04 0.62 Age 0.01 0.87 diabetes -0.01 0.94 Stroke 0.00 0.96 gender 0.00 0.98

Table 7: Multivariable analysis of demographic effects on serum anti-S100B autoantibodies. Results of pairwise multivariable analysis of demographic data variables and anti-S100B autoantibody serum levels. Analysis was performed using JMP 9.0 software. Statistical significance is defined as p<0.05 for correlations between variables.

3.4.3 S100B, anti-S100B autoantibodies and neuroimaging

Serum levels of S100B and anti-S100B autoantibodies were compared between groups based on neuroimaging diagnoses of brain metastases and SVD (Table 8). Only patients who had brain imaging were included in this analysis (n=106). Patients were stratified by the presence of brain metastasis and SVD into four groups; -met -SVD, -met

+SVD, +met +SVD, and +met -SVD. In top table we see how neuroimaging diagnosis can confound biomarker measurement. The middle table shows results stratified by the presence of metastasis regardless of SVD diagnosis. In the bottom table results are stratified by SVD in the absence of metastasis.

84

Table 8: S100B and anti-S100B autoantibody levels in the presence of SVD and brain metastases. Results are reported as means of serum levels for each neuroimaging diagnosis group with standard deviations. Red asterisks indicate p<0.05 by ANOVA in the comparison of anti-S100B autoantibody levels in SVD+ vs SVD- in the absence of brain metastatsis.

85 A total of 18 patients had brain metastases (Table 4). An ROC for the predictive value of S100B for brain metastases was generated and yielded an AUC of 0.63. An S100B serum level threshold was chosen in order to maximize sensitivity (Table 5). At a serum

S100B level of 0.058 ng/mL the sensitivity was 89% and the specificity was 43%. The positive predictive value (PPV) was 21.1% and the negative predictive value (NPV) was

96.2%.

An ROC for the predictive value of anti-S100B autoantibodies for brain metastases was generated and a threshold that would improve the specificity while maintaining a high sensitivity of S100B to detect brain metastases was chosen (Table 9). An anti-S100B autoantibody serum level of 2.0 AU had a sensitivity of 5.6% and a specificity of 84.1%.

The combinatory serum test for the detection of brain metastases was developed using an anti-S100B autoantibody threshold of <2.0 AU and an S100B threshold of >0.058 ng/mL.

The combinatory test sensitivity was 89% with an improved specificity of 58.2% compared to an S100B serum test. The overall accuracy was 51% with S100B alone and it improved to 62.5% when combined with the autoantibody threshold.

86 Sensitivity Specificity Serum Marker (%) (%) S100B (ng/mL) 0.023 100 6 0.058 89 43 0.07 55 59 0.1 33 80 0.155 11 95

Anti-S100B Autoantibodies (AU) 1 94.4 2.8 1.5 66.7 33.6 2 5.6 84.1 2.5 0 98.9

Combinatory test S100B > 0.058 ng/mL & Anti-S100B Autoantibodies (AU) < 1.0 5.6 97.3 < 1.5 33.3 83.6 < 2.0 89 58.2 < 2.5 89 45.5

Table 9: ROC analyses of S100B and anti- S100B autoantibodies for the detection of brain metastasis.

87 3.4.4 S100B, anti-S100B autoantibodies and lung cancer staging

There were a total of 46 patients that would be diagnosed as stage I or II if they had not had brain imaging at diagnosis. Three of these patients had brain metastases bringing their stage to IV. This subgroup was used in order to predict the presence of brain metastases in the lowest stage cancers without brain imaging. An ROC analysis revealed that a serum S100B >0.058 ng/mL had a sensitivity of 100%, and NPV of 100% (Table

10). When combined with anti-S100B IgG <2.0 AU the sensitivity and NPV remained the same but specificity improved from 39.5% to 55.8% and PPV had a modest increased from

10.3% to 13.6%.

Sensitivity Specificity Serum Marker (%) (%) PPV NPV

S100B > 0.058 ng/mL 100 39.5 10.3 100

S100B > 0.058 ng/mL & anti-S100B autoantibody < 2.0 AU 100 55.8 13.6 100

Table 10: S100B and anti-S100B autoantibodies for the detection of brain metastasis in patients stage I or II before brain imaging (n=46).

88

3.5 Discussion

This is a prospective study investigating the accuracy of the S100B protein and anti-S100B autoantibody serum levels for the detection of brain metastasis in patients with lung cancer. Our results confirm that elevated serum levels of S100B protein have a high sensitivity to detect brain metastases, and the addition of S100B autoantibody levels improves the specificity.

Staging is an evaluation of the extent to which lung cancer has spread. Further investigation is necessary when a lesion suspicious for metastatic disease is identified.

Dedicated imaging and/or a biopsy are performed until a pathological diagnosis or a very high probability of metastatic disease is obtained. The diagnosis of a brain metastasis in particular relies on the detailed radiographic characterization of a suspicious lesion as the potential morbidity limits the frequency with which a brain biopsy is performed. A biomarker could be used to identify those at highest and lowest risk of having brain metastases which could translate into more targeted use of brain imaging.

The properties of the astrocytic protein S100B have a great potential for clinical utility as a biomarker to detect brain metastases. The transendothelial migration of metastatic cells causes BBB permeability (19). When the BBB is disrupted brain derived proteins such as S100B are allowed to enter into the periphery. In our study, we used elevated serum levels of S100B as an indirect sign of metastatic growth in the brain as it is a marker of BBB permeability. Thus the predictive quality of S100B alone was not clinically useful for the detection of brain metastases as it can be observed in a multitude of other CNS diseases. Anti-S100B autoantibodies are hypothesized to develop as a result

89 of peripheral exposure of the brain specific protein S100B. Multiple BBB disruptions measured by surges in serum S100B have been shown to trigger an autoimmune response characterized by elevated titers of anti-S100B autoantibodies (17). The measurement of both S100B and anti-S100B autoantibodies have the potential to improve specificity of brain metastases as there is a prolonged BBB disruption with the presence of brain metastases (Table 9).

We found the combinatory biomarker test of S100B and anti-S100B autoantibody serum levels resulted in improved specificity (Table 9). In addition, this relationship was maintained for those presenting with early stage disease (e.g.:stage I or II before CNS staging, Table 10). In this subgroup the sensitivity and negative predictive value were both

100%. The use of S100B and anti-S100B autoantibodies as a biomarker test to select patients who would need brain imaging could impact the cost of initial staging. The United

States is estimated to have 224,000 new cases of lung cancer in 2014 and approximately

33,600 (15%) patients will be diagnosed with localized disease. Of the later, 52.1% (true negatives) would not require further brain imaging based on our data (20). Considering that the cost of an MRI of the brain is about US $3,000 and the cost of measuring S100B is US

$50, the potential savings are more than 50 million dollars per year.

We have previously demonstrated that SVD is a common cause of BBB disruption and elevation of S100B levels in patients with lung cancer (10). SVD is typically silent and may cause a long-term (months to years) and sustained BBB disruption. On the other hand, brain metastases that are detectable on brain imaging are likely present for a short time

(weeks to months). We have shown that autoantibody levels are higher in SVD compared to brain metastases (Table 8).

90 Our study had limitations. Not all patients had brain imaging at initial staging. MRI and CT scans were obtained as part of routine care. Clinical practice may be heterogeneous in this aspect. Providers may choose not to obtain brain imaging when patients are presumed to have early stage lung cancer or when obvious extrathoracic metastatic disease is present and patients have no neurological symptoms. We only included patients who had brain imaging, and those who did not have brain imaging who were clinically metastases free after 2 years of follow-up in the analysis.

The ELISA for the measurement of anti-S100B autoantibodies is not available commercially and it was developed in our laboratory (17). We did not observe a significant difference in the absorbance levels compared to previous study but further validation of the assay is necessary (9,10). Further understanding of S100B physiology (e.g.: half-life, clearance), immune response, and the different factors that might affect the serum levels will also contribute to the efforts to improve its specificity. S100B should also be studied in the longitudinal follow-up of patients with lung cancer after treatment in order to determine if it can accurately detect newly developed asymptomatic brain lesions.

In conclusion, the use of biomarkers to help identify CNS metastatic disease is promising. Serum S100B combined with S100B antibody levels are able to detect brain metastases from lung cancer with high sensitivity and reasonable specificity.

91 3.6 References

1. Stuschke M, Eberhardt W, Pottgen C, et al. Prophylactic cranial irradiation

in locally advanced non-small cell lung cancer after multimodality treatment: long-

term follow-up and investigations of late neuropsychologic effects. J Clin Oncol.

1999;17(9):2700-2709

2. Higer HP, Pedrosa P, Schuth M. MR imaging of cerebral tumors: state of

the art and work in progress. Neurosurg Rev. 1989;12(2):91-106

3. Silvestri GA, Gonzalez AV, Jantz MA, et al. Methods for staging non-small

cell lung cancer: Diagnosis and management of lung cancer, 3rd ed: American

College of Chest Physicians evidence-based clinical practice guidelines. Chest.

May 2013;143(5 Suppl):e211S-250S

4. Ricciardi S, de Marinis F. Multimodality management of non-small cell

lung cancer patients with brain metastases. Curr Opin Oncol. Mar 2010;22(2):86-

93

5. Butler AR, Leo JS, Lin JP, Boyd AD, Kricheff, II. The value of routine

cranial computed tomography in neurologically intact patients with primary

carcinoma of the lung. Radiology. 1979;131(2):399-401

6. Cole FH, Jr., Thomas JE, Wilcox AB, Halford HH, 3rd. Cerebral imaging

in the asymptomatic preoperative bronchogenic carcinoma patient: is it

worthwhile? Ann Thorac Surg. 1994;57(4):838-840

7. Colice GL, Birkmeyer JD, Black WC, Littenberg B, Silvestri G. Cost-

effectiveness of head CT in patients with lung cancer without clinical evidence of

metastases. Chest. 1995;108(5):1264-1271 92 8. Ricciardi S, de Marinis F. Multimodality management of non-small cell

lung cancer patients with brain metastases. Curr Opin Oncol. 2010;22(2):86-93

9. Marchi N, Mazzone P, Fazio V, Mekhail T, Masaryk T, Janigro D.

ProApolipoprotein A1: a serum marker of brain metastases in lung cancer patients.

Cancer. Mar 15 2008;112(6):1313-1324

10. Vogelbaum MA, Masaryk T, Mazzone P, et al. S100beta as a predictor of

brain metastases: brain versus cerebrovascular damage. Cancer. Aug 15

2005;104(4):817-824

11. Kanner AA, Marchi N, Fazio V, et al. Serum S100beta: a noninvasive

marker of blood-brain barrier function and brain lesions. Cancer.

2003;97(11):2806-2813

12. Bouwhuis MG1, Suciu S, Kruit W, Salès F, Stoitchkov K, Patel P, Cocquyt

V, Thomas J, Liénard D, Eggermont AM, Ghanem G. Prognostic value of serial

blood S100B determinations in stage IIB-III melanoma patients: a corollary study

to EORTC trial 18952. Eur J Cancer. 2011 ;47(3):361-8

13. Marchi N, Cavaglia M, Fazio V, Bhudia S, Hallene K, Janigro D. Peripheral

markers of blood-brain barrier damage. Clin Chim Acta. 2004;342(1-2):1-12

14. Detterbeck FC, Postmus PE, Tanoue LT. The stage classification of lung

cancer: Diagnosis and management of lung cancer, 3rd ed: American College of

Chest Physicians evidence-based clinical practice guidelines. Chest. May

2013;143(5 Suppl):e191S-210S

15. Pham N, Fazio V, Cucullo L, et al. Extracranial sources of S100B do not

affect serum levels. PLoS One. 2010;5(9)

93 16. Georgiadis D, Berger A, Kowatschev E, et al. Predictive value of S-100beta

and neuron-specific enolase serum levels for adverse neurologic outcome after

cardiac surgery. J Thorac Cardiovasc Surg. Jan 2000;119(1):138-147

17. Marchi N, Bazarian JJ, Puvenna V, et al. Consequences of repeated blood-

brain barrier disruption in football players. PLoS One. 2013;8(3):e56805

18. Wahlund LO, Barkhof F, Fazekas F, et al. A new rating scale for age-related

white matter changes applicable to MRI and CT. Stroke. Jun 2001;32(6):1318-1322

19. Wilhelm I, Molnár J, Fazakas C, Haskó J, Krizbai I. Role of the Blood-

Brain Barrier in the Formation of Brain Metastases. nt J Mol Sci. 2013 Jan; 14(1):

1383–1411

20. Siegel R, Ma J, Zou Z, Jemal A. Cancer statistics, 2014. CA Cancer J Clin.

Jan-Feb 2014;64(1):9-29

94

CHAPTER IV

ANTI- S100B AUTOANTIBODIES IN CEREBRAL SMALL VESSEL DISEASE

4.1 Abstract

Objective: We aimed to investigate whether cerebrovascular changes in SVD trigger production of autoantibodies against the unmasked antigen S100B.

Background: Cerebral small vessel disease (SVD) is a cerebrovascular risk factor for accelerated cognitive and motor impairment, ultimately leading to vascular dementia. SVD severity is graded using the age-related white matter changes (ARWMC) in the periventricular region seen on MRI. The pathophysiology of SVD involves vascular endothelial dysfunction and increased blood-brain barrier (BBB) permeability. BBB permeability allows for systemic exposure of CNS protein, and provides access to the brain of otherwise excluded molecules, such as immunoglobulins. Extravasation of the astrocytic protein S100B across a disrupted BBB has been reported in individuals with SVD. Given

95 that SVD is persistent, long term exposure to unmasked antigen may lead to the production of autoantibodies.

Methods: We enrolled patients diagnosed with lung neoplasms who underwent MRI scans to rule out metastatic brain disease. The presence or absence of SVD was diagnosed radiologically (62% SVD+); patients with brain metastases (16%) were excluded. An additional cohort of age matched and comorbidity matched cancer free controls were enrolled. Serum biomarkers were measured by ELISA. S100B protein levels and anti-

S100B autoantibody levels in three forms (monomer, multimer, and peptide) were measured.

Results: Patients with SVD had increased levels of anti-S100B autoantibodies when compared to SVD negative patients and controls. Only anti-S100B monomer autoantibodies and anti-S100B peptide autoantibodies displayed predictive value in the diagnosis of SVD based on ROC analysis. Anti-S100B monomer autoantibodies have the greatest possibility for clinical with a specificity of 91% and positive predictive value of

88%.

Conclusion: Our results show that chronic cerebrovascular disease and BBB disruption are associated with production of anti-S100B autoantibodies. These may be used as a diagnostic tool for SVD and premature cognitive decline while also providing an insight into therapeutic prevention of vascular dementias by immunomodulation.

96 4.2 Introduction

Neurological diseases have long been treated as manifestations of changes in neuronal function. While neuronal dysfunction or death are downstream effects of neurological diseases, recent findings have revealed that the blood-brain barrier (BBB) and systemic inflammation are involved in neurophysiology (1-3). The examples are numerous and span from epilepsy (1) to stroke (4) and dementia (5). In particular, the role of cerebrovascular changes in predicting or causing neurodegenerative diseases has been underscored in recent clinical studies (1, 6-8).

Brain imaging modalities allow for the estimation of the extent and severity of

SVD. The presence of SVD is accompanied by white matter changes seen on contrast enhanced CT and MRI (9,10). SVD related white matter changes are generally symmetrical and are most commonly seen in the periventricular and deep white matter. SVD includes white matter lesions and lacunar infarcts (11). SVD is one of the vascular risk factors for accelerated cognitive and motor impairment, ultimately leading to vascular dementia (12).

A broad more operational definition of SVD is age-related white matter changes

(ARWMC); both terms refer to often subclinical radiological findings. While the significance of SVD is not entirely understood, evidence linking small vessel disease to subtle or severe cognitive impairment have been reported (11,13-15).

Known risk factors of SVD include age, hypertension, smoking, diabetes, and heart disease, which are also associated with stroke and dementia (16). Current treatment addresses only vascular risk factors by lowering blood pressure. Yet this type of treatment

97 has the potential to worsen white matter ischemia due to the impairment of auto-regulation

(11).

A current hypothesis of the pathophysiology of SVD involves endothelial dysfunction. The endothelium functions to regulate CBF, auto-regulation, and maintains the BBB (17,18). In addition, endothelial dysfunction promotes atherosclerosis and premature arterial ageing (19-21). Circulating soluble markers of endothelial cells have been reported in patients with SVD (21). Endothelial dysfunction results in increased BBB permeability which allows for the extravasation of blood constituents into vessel walls and white matter (11). An increase in BBB permeability was shown to precede the development of white matter lesions thereby underscoring the causal role of altered BBB permeability in SVD (22). In addition, BBB permeability correlates with the severity of white matter lesions, and damaged white matter may be the site of BBB leakage (11).

We have shown that SVD is associated with elevated serum levels of S100B (23).

Furthermore, multiple or prolonged episodes of blood-brain barrier disruption (BBBD) characterized by surges in serum S100B trigger an autoimmune response characterized by elevated titers of anti-S100B autoantibodies in blood (2). These autoantibodies may be pathogenic upon penetration into the CNS via a leaky BBB. In a population of Alzheimer's patients anti-S100B autoantibodies were shown to be elevated in early onset of the disease

(24). In healthy football players increased serum levels of anti-S100B autoantibodies correlated with white matter changes on diffusion tensor imaging (DTI) scans (2, 25).

These responses also correlated with indices of white matter changes measured by MRI-

DTI (25).

98 Additionally, we have identified and synthesized a 10 amino acid peptide sequence that has the potential to be the immune trigger for autoantibody production and S100B autoimmunity in SVD. This epitope on human S100B was identified by a bioinformatic analysis using the immune epitope database. The full protein sequence of S100B (92 amino acid length) was inserted into the database and the search resulted in a single study where an MHC ligand assay was performed (26). The aim of this study was to characterize MHC- bound peptides derived from brains of patients with MS. Peptides were separated by reversed- phase HPLC and analyzed by mass spectrometry. Using mass spectrometric analysis 309 peptides including ELINNELSHF derived from S100B. This 10 amino acid sequence is not found in any other human protein and has the potential to be the target of

S100B autoimmunity.

The measurement of serum autoantibodies against both the whole protein S100B and ELINNELSHF are measured in human sera. In summary, SVD is important in the progressive deterioration of cerebrovascular function. We wished to explore the possibility that SVD is characterized by increased levels of anti-S100B autoantibodies in serum.

99 4.3 Methods

4.3.1 Human Subjects

All patients signed an informed consent according to institutional review protocols at The Cleveland Clinic Foundation. Human research was conducted as per Institutional

Review Board (IRB) guidelines (approved protocol at Cleveland Clinic IRB# 10-521). This prospective cohort study enrolled patients seen at the Cleveland Clinic from 2010 and

2012. We included adult patients > 18 years of age at the time of diagnosis of lung cancer, with or without neurological symptoms. Patients who were previously treated for lung cancer, had synchronous malignancies, or HIV infection were excluded. An additional cohort of propensity matched control patient samples were collected.

4.3.2 SVD Diagnosis by MRI

We have relied on radiology to diagnose SVD. Neuroradiologists were blinded to

S100B results and patients clinical information. Patient recruitment was made possible by a collaboration with Drs. Peter Mazzone and Humberto Choi who are pulmonologists at the Cleveland Clinic. Clinical info was obtained from electronic medical record including height and weight at the time of diagnosis. MRI studies included T1-weighted sequences with and without gadolinium, T2-weighted images, and fluid-attenuated inversion recovery

(FLAIR). MRI scans were evaluated for the presence of brain metastases and SVD related white matter changes. The age-related white matter changes (ARWMC) rating scale was used to grade SVD (27). The definitions of rating scores (0-3) are shown in the table below.

The following brain areas were used for grading; frontal, parieto-occipital, temporal,

100 infratenorial/cerebellum, and basal ganglia (striatum, globus pallidus, thalamus, internal/external capsule, and insula).

ARWMC Scale

White Matter Lesions Basal Ganglia Lesions

0 No lesions 0 No lesions

1 Focal lesions 1 One focal lesion (>5mm)

2 Beginning confluence of lesions 2 > 1 focal lesion

3 Diffuse involvement of entire region 3 Confluent lesions

4.3.3 Serum S100B Protein Measurement

The commercially available S100B ELISA was used for the measurement of serum

S100B protein (Diasorin Stillwater, MN).

4.3.4 Anti-S100B Autoantibody Measurement

This ELISA is described in (2). First, 96 well plates were coated with a PBS solution containing S100B protein (human brain, catalog number-559291, EMD

Chemicals or Sino Biological Inc #10181-H07E). Optimization of the ELISA was achieved by testing two concentrations of S100B protein (1 or 5 µg per well). No significant differences were observed at these two concentrations of S100B coating. An S100B monoclonal antibody (catalog number: Q86610M, Meridian Life Science Inc.) was used as a standard. Standard curves were obtained with 100 µl of serially diluted S100B monoclonal antibodies. Plates were coated overnight at 4oC with S100B protein (1

µg/well). Wells were then washed 3 times with PBS. Subsequently, 100 µL of a 1% BSA 101 blocking solution was added in each well and incubated for 2 hours at room temperature.

Wells were then washed 3 times with 200 µL of PBS containing 0.05% Tween 20. Serum samples and standards were added and incubated for 1 hour at room temperature. Samples were aspirated and wells washed 3 times using 200 µL of PBS containing 0.05% Tween-

20. A secondary antibody solution of 200 µL of horseradish-peroxidase (HRP) goat anti- mouse IgG and 200 µL of HRP goat anti-human were added to the standards and serum samples respectively. After a 1 hour incubation at room temperature wells were washed 3 times with 200 µL of PBS containing 0.05% Tween-20. Finally, 100 µL of OPD solution was added and the reaction incubated for 30 minutes at room temperature. The reaction was stopped by adding 100 µL of 2.5 M Sulfuric acid. Samples were analyzed using an

ELISA plate reader at 490 nm.

4.3.5 Peptide Synthesis

The synthesis of ELINNELSHF was performed by the proteomics core at

Cleveland Clinic. Solid phase peptide synthesis was used to generate the resin-bound peptide sequence. The chemical peptide synthesis starts at the C-terminal end of the peptide and ends at the N-terminus. The amino acids used contained a 9 fluoromethoxycarbonyl

(Fmoc) group for protection and quantification of coupling. The first Fmoc-amino acid wass coupled to Wang resin. The basic steps for the synthesis of the peptide on a resin includes; 1) Deprotection: the Fmoc group is removed from the N-terminus of the growing peptide chain allowing the next amino acid to be coupled; 2) Wash: the resin containing the peptide is washed thoroughly with solvent before beginning the coupling reaction; 3)

Coupling: an amino acid is activated and added to the growing peptide chain; 4) Wash: the

102 resin containing the peptide is washed thoroughly before beginning the deprotection reaction of the next cycle. Steps 1 through 4 are repeated until the desired peptide sequence is assembled on the resin. 5) Cleavage : the last step is to cleave the peptide from the resin and collect it.

All the Fmoc amino acids required to synthesize the peptide ELINNELSHF were obtained from Chem-Impex International, Inc. (Wood Dale, IL) and contained the following side chain protecting groups: Asn(Trt), Asp(OtBu), His(Trt), Ser(tBu). The peptide was synthesized using CEM Liberty Automated Microwave Peptide Synthesizer

(CEM Corporation, Matthews, NC). This system can sequentially synthesize up to 12 different peptides, using the single-mode microwave reactor, Discover. The reaction vessel

(30 ml) is connected to a spray head for delivery of all the reagents and a fiber-optic temperature probe for controlling the microwave power delivery. The system utilizes up to

25 stock solutions for amino acids and seven reagent ports that perform the following functions: main wash, secondary wash, deprotection, capping, activator, activator base, and cleavage. The system uses nitrogen pressure for transfer of all reagents and to provide an inert environment during synthesis. The system uses metered sample loops for precise delivery of all amino acids, activator and activator base. The cleavage reaction is routinely performed manually. Following completion of the synthesis, both cleavage and deprotection were performed according to the manufacturer’s protocol. Next, the peptide was precipitated out and washed using ice-cold anhydrous tert-butyl ether. The precipitated and washed peptide was then dissolved in 25% acetonitrile and lyophilized to complete dryness.

103 Prior to precipitation with acetonitrile, a small volume of the peptide (10-25 microliter) was diluted 10-fold with 0.1% TFA for analysis. Analytical HPLC (Beckman

System Gold) was performed using a C18 column (Higgins Analytical Inc.). The purity of the preparation was checked by gradient elution with a single peak at 214 nm (solvent

A=98% acetonitrile in water; solvent B=0.085% TFA in water). The appropriate fractions were collected and subjected to Mass spectrometry. The analysis was performed using a

MALDI-TOF/TOF mass spectrometer (Model No. 4800; AB SCIEX, Framingham, MA).

4.3.6 Anti-S100B Peptide Autoantibody Measurement

Optimization of the ELISA described in (2) was performed using the peptide sequence as the antigen of the reverse ELISA. First, 96 well plates were coated with a PBS solution containing ELINNELSHF. Optimization of this ELISA was achieved by testing concentrations of ELINNELSHF (1, 2 or 5 µg per well). No significant differences in signal were observed at these concentrations of ELINNELSHF and a coating of 1 µg/well was chosen. Plates were coated overnight at 4oC with ELINNELSHF (1 µg/well). Wells were then washed 3 times with PBS. Subsequently, 100 µL of a 1% BSA blocking solution was added in each well and incubated for 2 hours at room temperature. Wells were then washed

3 times with 200 µL of PBS containing 0.05% Tween 20.

Serum samples and standards were added and incubated for 1 hour at room temperature. Optimal sample dilution was determined by titer experiments and a dilution of 1:5000 was chosen for the patient population. Samples were aspirated and wells washed

3 times using 200 µL of PBS containing 0.05% Tween-20. The selection of secondary antibody was based on the commercially available antibody concentration that yielded the

104 most reproducible results. Two goat anti-human HRP conjugated secondary antibodies were tested; (Calbiochem # 401445 and Milipore # AP309P). An antibody concentration of 1:5000 of 200 µL goat anti-human HRP conjugated antibody (Calbiochem #401445) were added to sample wells. After a 1 hour incubation at room temperature wells were washed 3 times with 200 µL of PBS containing 0.05% Tween-20. Finally, 100 µL of OPD solution was added and the reaction incubated for 30 minutes at room temperature. The reaction was stopped by adding 100 µL of 2.5 M Sulfuric acid. Samples were analyzed using an ELISA plate reader at 490 nm.

Positive and negative controls including pooled human IgG, highest sera sample, lowest sera sample were included. All experiments were run on the same day in duplicates.

Reproducibility experiments were performed between plates, between days, and over time to evaluate variability.

4.3.7 Total Serum IgG Measurement

The commercially available human IgG ELISA was used for the measurement of total serum IgG (abcam #ab100547) as per manufacturer's instructions.

105 4.3.8 Statistical Analysis

Origin 9.0 (Origin Lab, Northampton, MA, USA) and Jump 9.0 software were used for statistical analysis. The data was assessed for assumptions of normality and appeared to display a normal distribution. Parametric statistical tests were performed. ANOVA was used to compare populations. Data points were fitted by a least square fitting routine. R is the square root of the coefficient of determination R2. In all figures, symbols with error bars indicate mean ± standard deviation (SD); *p<0.05 and was considered statistically significant and p values are shown in the figures or accompanying legends. n.s. represents not significant differences. Pairwise multivariable analyses were performed in order to determine demographic effect correlations within a population. Univariate logistic regressions for each serum marker were performed. The Youden Index was used to chose a cutoff level for each serum marker that maximizes the relationship between sensitivity and specificity using a single statistic. In addition, the associated sensitivity, specificity, negative predictive value (NPV) and positive predictive value (PPV) were investigated and reported. The total area under the curve (AUC) for each respective ROC curve was considered statistically significant for those exhibiting AUC’s significantly different from

0.5.

106 4.4 Results

4.4.1 Patient Characteristics, Neuroradiology and Demographics

We report findings from 111 patients who underwent a routine MRI or CT scan to detect brain metastatic growth in the context of staging for lung cancer. MRI scans were interpreted by a neuroradiologist, CT scans were excluded. Patients with a diagnosis of metastatic disease were also excluded from this study. Examples of MRI scans are shown in Figure 10. Note that in addition to an unequivocal diagnosis of brain metastases (A) or

SVD (B), scans with uncertain interpretation were also encountered. When uncertainty as to the pathological significance of changes was present, the patient was excluded from subsequent analysis. Thus, of the 111 subjects initially enrolled, 18 (16%) were excluded because of presence of metastasis visible on MRI, and 10 because of lack of MRI or low sample volume.

107 Contrast Pre-Contrast A

B

C

Figure 10: Radiologic appearance of SVD. MRI images pre- and post-contrast injection of Gadolinium in three patients enrolled in the study. A) Normal MRI, negative for SVD, negative for brain metastasis, exemplary of non-SVD patients. Note the lack of contrast enhancement and normal appearing white-gray matter margins. B) SVD positive MRI, negative for brain metastasis, as typically seen in SVD patients in this study. Note the patchy appearance of contrast enhancement. C) SVD positive MRI in a patient with a large brain metastasis. Patients with an isolated brain metastasis, or with brain metastasis and a positive SVD were excluded from this study.

108 The demographic properties of the patients enrolled in this study are shown in Table

11. Note that positively diagnosed SVD patients were older than those without SVD

(p=0.05). Other demographic values including gender, BMI, and kidney function

(measured by creatinine clearance, CrCl) were consistent among patient populations. The presence of comorbidities including diabetes, hypertension, and hyperlipidemia were not statistically different between groups. However in the lung cancer population there was a higher incidence of COPD than in controls (p<0.01). This can be attributed to the fact that the control population did not have a history of lung disease.

109 Characteristics +SVD (n=51) -SVD (n=32) p Age (mean) 70 64 0.05 Gender (male, %) 55 62 0.5 BMI (mean) 27.3 28.9 0.23 CrCl (mean) 82 95 0.09 Diabetes (%) 17 12 0.51 Hypertension (%) 64 63 0.84 Hyperlipidemia (%) 41 50 0.43 COPD (%) 43 56 0.24

Control Lung Cancer Characteristics (n=85) (n=83) p Age (mean) 66 67 0.34 Gender (male, %) 49 58 0.27 Diabetes (%) 12 15 0.44 Hypertension (%) 63 64 0.96 Hyperlipidemia (%) 45 46 0.96 COPD (%) 16 50 p<0.01

Table 11: Patient Characteristics. Characteristics of lung cancer patients and control patients enrolled in this study are reported. The table is stratified into four different groups; lung cancer patients with SVD, lung cancer patients negative for SVD, control patients, and all lung cancer patients (both SVD+ and SVD-). Results are given as mean for continuous data, or percent for categorical data. Clinical comparisons are made between

SVD patients and non SVD patients in the absence of metastasis and between all lung cancer patients and controls. Statistical significance is defined as p<0.05.

110 4.4.2 SVD Grading

Fifty one of the patients who had brain imaging had evidence of SVD. We analyzed

S100B, anti-S100B peptide autoantibodies, anti-S100B monomer autoantibodies, and anti-

S100B multimer autoantibodies serum levels against the grading of white matter and basal ganglia lesions as a sum SVD severity score (Figure 11). Neither S100B protein levels nor any anti-S100B autoantibody levels significantly varied according to the severity of SVD.

Additionally, Table 12 reports serum biomarker levels with ARWMC scale for white matter lesions, basal ganglia lesions, and the sum SVD severity score. Note that the majority of lesions are found in white matter (n=47) as compared to basal ganglia (n=4).

Also sum SVD severity scores were mild (scores 1 and 2) with only 2 patients having a severity score of 3.

111

Figure 11: SVD severity and serum marker levels. Overall SVD severity is reported as the sum of basal ganglia and white matter scores to obtain SVD grade. Serum levels of

S100B, anti-S100B peptide autoantibodies, anti-S100B monomer autoantibodies, and

S100B peptide autoantibodies are plotted against SVD grade in A, B, C, and D respectively. The number of patients per SVD grade group are as follows; SVD 0: 32 patients, SVD 1: 41 Patients, SVD 2: 8 patients, and SVD 3: 2 patients. Statistical significance was analyzed by ANOVA. No significant differences between groups were observed. The SVD 3 group was excluded from ANOVA analyses due to insufficient sample size.

112 anti-S100B anti-S100B anti-S100B peptide monomer multimer S100B autoantibodies autoantibodies autoantibodies SVD Grade (ng/mL) (AU) (AU) (AU) White Matter Lesions 0 (n=32) 0.06 1.32 1.48 0.36 1 (n=39) 0.07 1.44 1.7 0.42 2 (n=8) 0.06 1.51 1.63 0.41 3 (n=0) 0 0 0 0

Basal Ganglia Lesions 0 (n=79) 0.065 1.39 1.6 0.39 1 (n=3) 0.071 1.62 1.66 0.49 2 (n=1) 0.13 1.57 1.81 0.3 3 (n=0) 0 0 0 0

Sum SVD Severity 0 (n=32) 0.06 1.32 1.48 0.36 1 (n=41) 0.07 1.42 1.69 0.42 2 (n=8) 0.052 1.57 1.7 0.41 3 (n=2) 0.116 1.46 1.57 0.42

Table 12: ARWMC scale and serum marker levels. Serum levels are reported as means.

Statistical significance was analyzed by ANOVA, no significant differences between groups were observed.

113 4.4.3 Serum Markers

Blood samples were collected at time of imaging or within 24 hrs. A total of five serum markers were measured in this study by ELISA; S100B protein, anti-S100B monomer autoantibodies, anti-S100B multimer autoantibodies, anti-S100B peptide autoantibodies, and total serum IgG. Serum levels are reported in Table 13. In the upper portion of Table 13 comparisons are within the lung cancer population between patients diagnosed with SVD. S100B protein serum levels are slightly elevated in the SVD positive group compared to the SVD negative group; however, this difference was not statistically significant (p=0.07). The difference in anti-S100B peptide autoantibodies and monomer autoantibodies between SVD+ and SVD- were statistically significant (p=0.05 and p<0.01 respectively). There was no difference in serum levels of anti-S100B multimer autoantibodies and total IgG serum levels between the SVD+ and SVD- groups. In the lower portion of Table 13 comparisons are made between the lung cancer population

(SVD- and SVD+ combined) and the control population. Three of the five biomarkers measured in the lung cancer population were also measured in controls; anti-S100B peptide autoantibodies, anti-S100B multimer autoantibodies, and total IgG. The difference in anti-

S100B peptide autoantibodies and multimer autoantibodies between lung cancer and controls were statistically significant, however total IgG levels remained unchanged.

114 Serum Marker +SVD (n=51) -SVD (n=32) p S100B Protein (ng/mL) 0.071 0.057 0.07 anti-S100B peptide autoantibodies (AU) 1.46 1.27 0.05 anti-S100B monomer autoantibodies (AU) 1.7 1.4 p<0.01 anti-S100B multimer autoantibodies (AU) 0.42 0.35 0.15

Total IgG (ng/mL) 2.5 2.1 0.3

Control Lung Cancer Serum Marker (n=85) (n=85) p anti-S100B peptide autoantibodies (AU) 0.33 1.3 p<0.01 anti-S100B multimer autoantibodies (AU) 0.23 0.39 p<0.01

Total IgG (ng/mL) 2.73 2.3 0.14

Table 13: Serum Marker levels in patients. Results are given as mean serum levels for each serum marker. Protein levels of S100B and total IgG are reported in ng/mL and autoantibody levels are reported as absorbance units measured at 490nm. Clinical comparisons are made between SVD patients and non SVD patients in the absence of metastasis and between all lung cancer patients (negative for metastasis) and controls.

Statistical significance is defined as p<0.05.

115 4.4.4 Clinical Utility

In order to demonstrate the clinical utility of serum measurements receiver operator curves (ROC) were generated. The clinical utility of each serum marker measured was calculated for the prediction of SVD (Table 14). The area under the curve (AUC) of the ROC for S100B protein in the determination of SVD was found to be 0.60. The AUCs improved with the measurements of autoantibodies. The AUCs for anti-S100B monomer, anti-S100B multimer, and anti-S100B peptide autoantibodies were 0.71, 0.61, and 0.63 respectively. Only anti-S100B peptide autoantibody and anti-S100B monomer autoantibody measurements were significantly different from an AUC of 0.5. Anti-S100B monomer autoantibodies have the greatest clinical utility in the prediction of SVD. At a cutoff value of 1.76 AU sensitivity and specificity were maximized to 43% and 91%, with a PPV of 88% and a NPV of 50%.

116 Serum Sensitivit Specificit Marker AUC Cutoff y y PPV NPV S100B Protein 0.60 0.06 61 63 72 50 anti-S100B monomer autoantibodie s 0.71 1.76 43 91 88 50 anti-S100B multimer autoantibodie s 0.61 0.34 67 56 71 51 anti-S100B peptide autoantibodie s 0.63 1.48 51 74 76 48

Table 14: ROC analyses for the prediction of SVD. Univariate logistic regressions for each serum marker were performed. The Youden Index was used to chose a cutoff level for each serum marker that maximizes the relationship between sensitivity and specificity using a single statistic. In addition, the associated sensitivity, specificity, negative predictive value (NPV) and positive predictive value (PPV) were investigated and reported.

The total area under the curve (AUC) for each respective ROC curve was considered statistically significant for those exhibiting AUC’s significantly different from 0.5.

117 4.4.5 Lung Cancer Staging

The effect of cancer staging on serum biomarker levels were investigated. The at risk population for SVD analyzed with brain imaging in this study all had a lung cancer diagnosis. We analyzed S100B, anti-S100B peptide autoantibodies, anti-S100B monomer autoantibodies, and anti-S100B multimer autoantibodies serum levels against lung cancer grading in order to determine the effect of cancer severity on serum biomarker levels

(Figure 12). Neither S100B protein levels nor any anti-S100B autoantibody levels significantly varied according to the cancer staging.

118

Figure 12: Lung cancer staging and serum marker levels. Lung cancer staging is represented as four groups; 1: Stages IA and IB, 2: Stages IIA and IIB, 3: Staged IIA and

IIB, and 4: Stage IV. Serum levels of S100B, anti-S100B peptide autoantibodies, anti-

S100B monomer autoantibodies, and S100B peptide autoantibodies are plotted against lung cancer staging in A, B, C, and D respectively. The number of patients per staging group are as follows; 1: 17 patients, 2: 9 Patients, 3: 39 patients, and 4: 18 patients.

Statistical significance was analyzed by ANOVA. No significant differences between groups were observed.

119 4.5 Discussion

The main finding from our work is that patients with SVD have autoantibodies against the astrocytic protein S100B. In particular, the presence of autoantibodies correlated with MRI white matter changes in the periventricular region. A lesser finding of this work is the discovery of autoantibodies as clinical biomarkers of radiologically detectable SVD. In an era where clinical practice is dominated by economic forces, any alternative to traditional, costly radiologic investigation is worth pursuing.

Serum autoantibodies against the antigen, S100B, were measured in three forms; monomeric, multimeric, and as a peptide (ELINNELSHF). Although autoantibodies against all three antigenic forms were elevated in patients with SVD compared to those without SVD, only elevations in the monomeric and peptide forms were statistically significant (Table 3). In addition, neither autoantibody levels nor S100B protein levels correlated with SVD severity as measured by the ARWMC scale for white matter lesions, basal ganglia lesions, and the sum SVD severity score (Figure 11 and Table 12). This could be due in part to the small number of severe SVD cases (SVD grade 2 n=8, grade 3 n=2).

Serum biomarker level comparisons were also made between the lung cancer population (SVD- and SVD+) and the control population (Table 12). Only antibodies against S100B multimer and S100B peptide forms were measured. The differences in anti-

S100B peptide autoantibodies and anti-S100B multimer autoantibodies between lung cancer and controls were statistically significant. In addition, both anti-S100B autoantibody levels were significantly higher in SVD- and SVD+ groups individually when compared to controls. The low levels of autoantibodies in the control group suggest that

120 age and risk factors of SVD alone are not enough to cause increased serum levels of autoantibodies (Table 11 and Table 13). Yet, the fact that autoantibody levels are increased in the SVD- group compared to controls suggests that not only can a positive SVD diagnosis cause an increase, but a diagnosis of cancer itself may also play a role. Just as the presence of brain metastases is seen as a confounder in the diagnosis of SVD using serum S100B, based on our previous work (23, 28-30), the presence of lung cancer may be a confounder in the detection of SVD using anti-S100B autoantibodies. Lung cancer staging however was not found to correlate with S100B or anti-S100B autoantibodies

(Figure 13).

The clinical utility of serum biomarkers in the detection of SVD were investigated by ROC. Only anti-S100B monomer autoantibodies and anti-S100B peptide autoantibodies had significant predictive qualities based on AUC (Table 14). Anti-S100B monomer autoantibodies had the greatest AUC of 0.71, and at a cutoff value of 1.76 AU sensitivity and specificity were maximized to 43% and 91%, with a PPV of 88% and a NPV of 50%.

While Anti-S100B peptide autoantibodies had an AUC of 0.63, and at a cutoff value of

1.76 AU sensitivity and specificity were maximized to 51% and 74%, with a PPV of 76% and a NPV of 48%. While the specificity and positive predictive value of anti-S100B monomer autoantibodies are high, the marginal values of specificity and negative predictive value limit its clinical utility.

There were limitations in this study, the lung cancer population consisted of patients undergoing a contrast MRI scan to rule out metastatic brain disease. Thus, this population may not be ideal to study markers of normal or pathological ageing in the context of SVD.

However, we did not include in the analysis patients’ data where a tumor was diagnosed

121 (e.g., Figure 10C) nor did we include scans where an unequivocal response was obtained.

In addition, the population enrolled was intentionally chosen because it encompasses individuals at high risk for SVD including hypertension and age. It is important to underscore that in a separate study on young athletes with no known risk for SVD but with a history of repeated BBBD a scenario encompassing autoantibodies against S100B and

MRI changes in white matter were documented (2, 25). Other studies report serum surges in UCH-L1 and phospho-tau in young impact-sport players (31,32). Autoimmunity against the astrocytic protein GFAP and S100B have been reported in post-TBI subjects and in

Alzheimer’s disease (25, 33-36). Taken in the context of current understanding of early and late events in brain ageing, our results show a surprising similarity in serum profiles of subjects with advanced SVD and their young counterpart with white matter changes.

Whether these conditions are a continuum of abnormal functional decline of the CNS remains to be proven with studies where the radiologic endpoints are comparable (e.g., DTI vs. DTI, fMRI vs. fMRI).

Since the patients enrolled in the study were affected by lung cancer, the observed autoimmune IgGs may be a consequence of the disease process, as in paraneoplastic syndromes. Approximately 3-5% of patients with small cell lung cancer develop paraneoplastic syndrome (37). Paraneoplastic autoimmune CNS disorders have also been recognized; in addition, autoimmune brain disorders such as multiple sclerosis with specificity for brain white matter occur in absence of neoplastic disease (37). Several mechanisms have been proposed to explain paraneoplastic syndromes but it is likely that immune mimicry triggered by tumor-secreted potential autoantigens is an important mechanism (38). 122 The presence of autoantibodies against S100B, or other endogenous, non-tumor derived brain protein, likely depends on a different mechanism. A recent paper has proposed a scenario where GFAP, microtubule associated Tau, and synapsin, not only

S100B show autoimmunity in instances of BBB permeability (2, 33, 39). Since during

BBBD an array of protein enter into periphery, the possibility exists that these proteins may trigger an immune response similar to what observed in our study. In the presence of BBB permeability, S100B is allowed to enter peripheral blood and reach antigen presenting cells

(3). According to this model, the presence of a systemic tumor is not necessary nor sufficient to trigger the autoimmune response. If this hypothesis proves correct, the presence of autoantibodies against S100B is a phenomenon distinct from paraneoplastic mechanisms of autoimmunity. This may have profound treatment implications, since paraneoplastic syndromes are typically resistant to immunosuppressive therapy while other autoimmune brain diseases respond to some extent to immunomodulators (40).

The only treatment available for SVD is directed towards blood pressure management. Our results, if confirmed as part of the pathophysiology of SVD, suggest that a novel therapeutic approach to this “BBB disease” is immunomodulation of an autoimmune response. There is growing evidence that autoimmunity, or the presence of autoantibodies, is a hallmark of many neurological diseases. For example, S100B, GFAP, amyloid beta, and tau have been shown to be elevated in TBI, epilepsy, stroke, and cancer

(2, 33, 41-46). It is thus important to understand whether SVD is a consequence of misguided immunity attacking the brain.

123 4.6 Conclusion

Our results are the first evidence of non-paraneoplastic anti-CNS immunity in subjects with SVD. The relevance for this misguided immune response for ageing during the lifespan needs to be further systematically evaluated. Our study underscores the importance of brain-derived blood biomarkers in the diagnosis and understanding of complex processes responsible for cognitive and functional decline of the human brain.

124 4.7 References

1. Marchi N, Granata T, Janigro D (2014) Inflammatory pathways of seizure

disorders. Trends Neurosci 37: 55-65

2. Marchi N, Bazarian JJ, Puvenna V, Janigro M, Ghosh C, Zhong J, Zhu T,

Blackman E, Stewart D., Ellis J, Butler R, Janigro D (2013) Consequences of

repeated blood-brain barrier disruption in football players. Plos One 8: e56805

3. Bargestock E, Puveena V, Iffland P, Falcone T, Hossain M, Vetter S, Man

S, Dickstein L, Marchi N, Ghosh C, Carvalho-Tavares J, Janigro D (2014) Is

peripheral immunity regulated by blood-brain barrier permeability changes? PLoS

One. 2014 Jul 2;9(7):e101477

4. Kim SY, Buckwalter M, Soreq H, Vezzani A, Kaufer D (2012) Blood-brain

barrier dysfunction-induced inflammatory signaling in brain pathology and

epileptogenesis. Epilepsia 53: 37-44

5. Carmeliet P, De Strooper B (2012) Alzheimer's disease: A breach in the

blood-brain barrier. Nature 485: 451-452

6. Janigro D, Iffland PH, Marchi N, Granata T (2013) A role for inflammation

in status epilepticus is revealed by a review of current therapeutic approaches.

Epilepsia 54: 30-32

125 7. Marchi N, Granata T, Ghosh C, Janigro D (2012) Blood-brain barrier

dysfunction and epilepsy: pathophysiologic role and therapeutic approaches.

Epilepsia 53: 1877-1886

8. Guo K, Janigro D (2013) New immunological approaches in treating and

diagnosing CNS diseases. Pharm Pat Anal 2: 361-371

9. Bradley WG, Jr., Waluch V, Yadley RA, Wycoff RR (1984) Comparison

of CT and MR in 400 patients with suspected disease of the brain and cervical spinal

cord. Radiology 152: 695-702

10. Awad IA, Spetzler RF, Hodak JA, Awad CA, Carey R (1986) Incidental

subcortical lesions identified on magnetic resonance imaging in the elderly. I.

Correlation with age and cerebrovascular risk factors. Stroke 17: 1084-1089

11. O'Sullivan M (2008) Leukoaraiosis. Pract Neurol 8: 26-38

12. Smith EE, Schneider JA, Wardlaw JM, Greenberg SM (2012) Cerebral

microinfarcts: the invisible lesions. Lancet Neurology 11: 272-282

13. O'Sullivan M, Morris RG, Huckstep B, Jones DK, Williams SCR, Markus

HS (2004) Diffusion tensor MRI correlates with executive dysfunction in patients

with ischaemic leukoaraiosis. Journal of Neurology Neurosurgery and Psychiatry

75: 441-447

14. Prins ND, van Dijk EJ, den Heijer T, Vermeer SE, Jolles J, Koudstaal PJ,

Hofman A, Breteler MM (2005) Cerebral small-vessel disease and decline in

126 information processing speed, executive function and memory. Brain 128: 2034-

2041

15. den Heijer T, Launer LJ, Prins ND, van Dijk EJ, Vermeer SE, Hofman A,

Koudstaal PJ, Breteler MM (2005) Association between blood pressure, white

matter lesions, and atrophy of the medial temporal lobe. Neurology 64: 263-267

16. Smith EE, Schneider JA, Wardlaw JM, Greenberg SM (2012) Cerebral

microinfarcts: the invisible lesions. Lancet Neurol 11: 272-282

17. Hunt BJ, Jurd KM (1998) Endothelial cell activation. A central

pathophysiological process. BMJ 316: 1328-1329

18. Abbott NJ (2004) Evidence for bulk flow of brain interstitial fluid:

significance for physiology and pathology. Neurochem Int 45: 545-552

19. Hassan A, Hunt BJ, O'Sullivan M, Parmar K, Bamford JM, Briley D, Brown

MM, Thomas DJ, Markus HS (2003) Markers of endothelial dysfunction in lacunar

infarction and ischaemic leukoaraiosis. Brain 126: 424-432

20. Scuteri A, Nilsson PM, Tzourio C, Redon J, Laurent S (2011)

Microvascular brain damage with aging and hypertension: pathophysiological

consideration and clinical implications. J Hypertens 29: 1469-1477

21. Lazarov O, Mattson MP, Peterson DA, Pimplikar SW, van PH (2010) When

neurogenesis encounters aging and disease. Trends Neurosci 33: 569-579

127 22. Topakian R, Barrick TR, Howe FA, Markus HS (2010) Blood-brain barrier

permeability is increased in normal-appearing white matter in patients with lacunar

stroke and leucoaraiosis. J Neurol Neurosurg Psychiatry 81: 192-197

23. Vogelbaum MA, Masaryk T, Mazzone P, Mekhail T, Fazio V, McCartney

S, Marchi N, Kanner A, Janigro D (2005) S100beta as a predictor of brain

metastases: brain versus cerebrovascular damage. Cancer 104: 817-824

24. Gruden MA, Davidova TB, Malisauskas M, Sewell RDE, Voskresenskaya

NI, et al. (2007) Differential neuroimmune markers to the onset of Alzheimer's

disease neurodegeneration and dementia: Autoantibodies to A beta((25-35))

oligomers, S100b and neurotransmitters. J Neuroimmunol 186: 181–192

25. Bazarian JJ, Zhu T, Zhong J, Janigro D, Rozen E, Roberts A, Javien H,

Merchant-Borna K, Abar B, Blackman EG. (2014) Persistent, Long-term Cerebral

White Matter Changes after Sports-Related Repetitive Head Impacts. Plos One 9:

e94734

26. Fissolo N, Haag S, de Graaf KL, Drews O, Stevanovic S, Rammensee

HG, Weissert R. Naturally presented peptides on major histocompatibility complex

I and II molecules eluted from central nervous system of multiple sclerosis patients.

Mol Cell Proteomics. 2009 Sep;8(9):2090-101

27. Wahlund LO, Barkhof F, Fazekas F, Bronge L, Augustin M, Sjogren M,

Wallin A, Ader H, Leys D, Pantoni L, Pasquier F, Erkinjuntti T, Scheltens P (2001)

A new rating scale for age-related white matter changes applicable to MRI and CT.

Stroke 32: 1318-1322

128 28. Kanner, A.A., Marchi, N., Fazio, V., Mayberg, M.R., Koltz, M.T., Siomin,

V., Stevens, G.H., Masaryk, T., Aumayr, B., Vogelbaum, M.A., Barnett, G.H., &

Janigro, D. 2003. Serum S100beta: a noninvasive marker of blood-brain barrier

function and brain lesions. Cancer, 97, (11) 2806-2813

29. Marchi N, Mazzone P, Fazio V, Mekhail T, Masaryk T, Janigro D.

ProApolipoprotein A1: a serum marker of brain metastases in lung cancer patients.

Cancer. 2008 Mar 15;112(6):1313-24

30. Mazzone PJ, Marchi N, Fazio V, Taylor JM, Masaryk T, Bury L, Mekhail

T, Janigro D. Small vessel ischemic disease of the brain and brain metastases in

lung cancer patients. PLoS.One. 2009;4(9):e7242

31. Puvenna V, Brennan C, Shaw G, Yang C, Marchi N, Bazarian JJ, Merchant-

Borna K, Janigro D (2014) Significance of ubiquitin carboxy-terminal hydrolase l1

elevations in athletes after sub-concussive head hits. Plos One 9: e96296

32. Shahim P, Tegner Y, Wilson DH, Randall J, Skillback T, Pazooki D,

Kallberg B, Blennow K, Zetterberg H (2014) Blood Biomarkers for Brain Injury in

Concussed Professional Ice Hockey Players. JAMA Neurol. 2014;(6):684-92

33. Zhang Z, Zoltewicz JS, Mondello S, Newsom KJ, Yang Z, Yang B,

Kobeissy F, Guingab J, Glushakova O, Robicsek S, Heaton S, Buki A, Hannay J,

Gold MS, Rubenstein R, Lu XC, Dave JR, Schmid K, Tortella F, Robertson CS,

Wang KK (2014) Human traumatic brain injury induces autoantibody response

against glial fibrillary acidic protein and its breakdown products. Plos One 9:

e92698

129 34. Maetzler W, Berg D, Synofzik M, Brockmann K, Godau J, Melms A,

Gasser T, Hornig S, Langkamp M (2011) Autoantibodies Against Amyloid and

Glial-Derived Antigens are Increased in Serum and Cerebrospinal Fluid of Lewy

Body-Associated Dementias. Journal of Alzheimers Disease 26: 171-179.

35. Nagele RG, Clifford PM, Siu G, Levin EC, Acharya NK, Han M, Kosciuk

MC, Venkataraman V, Zavareh S, Zarrabi S, Kinsler K, Thaker NG, Nagele EP,

Dash J, Wang HY, Levitas A (2011) Brain-reactive autoantibodies prevalent in

human sera increase intraneuronal amyloid-beta(1-42) deposition. J Alzheimers

Dis 25: 605-622

36. Levin EC, Acharya NK, Han M, Zavareh SB, Sedeyn JC, Venkataraman V,

Nagele RG (2010) Brain-reactive autoantibodies are nearly ubiquitous in human

sera and may be linked to pathology in the context of blood-brain barrier

breakdown. Brain Res 1345: 221-232

37. Dalmau J, Rosenfeld MR (2008) Paraneoplastic syndromes of the CNS.

Lancet Neurol 7: 327-340

38. Ankeny DP, Popovich PG (2010) B cells and autoantibodies: complex roles

in CNS injury. Trends in Immunology 31:332-338

39. Poletaev AB, Morozov SG, Gnedenko BB, Zlunikin VM, Korzhenevskey

DA (2000) Serum anti-S100b, anti-GFAP and anti-NGF autoantibodies of IgG

class in healthy persons and patients with mental and neurological disorders.

Autoimmunity 32: 33–38

40. Maverakis E, Goodarzi H, Wehrli LN, Ono Y, Garcia MS (2012) The

etiology of paraneoplastic autoimmunity. Clin Rev Allergy Immunol 42: 135-144

130 41. Donato R, Cannon BR, Sorci G, Riuzzi F, Hsu K, Weber DJ, Geczy CL

(2013) Functions of S100 Proteins. Current Molecular Medicine 13: 24-57

42. Lumpkins KM, Bochicchio GV, Keledjian K, Simard JM, McCunn M,

Scalea T (2008) Glial Fibrillary Acidic Protein is Highly Correlated With Brain

Injury. Journal of Trauma-Injury Infection and Critical Care 65: 778-784

43. Hauschild A, Engel G, Brenner W, Glaser R, Monig H, Henze E,

Christophers E (1999) S100B protein detection in serum is a significant prognostic

factor in metastatic melanoma. Oncology 56: 338-344

44. Bazarian JJ, Zemlan FP, Mookerjee S, Stigbrand T (2006) Serum S-100B

and cleaved-tau are poor predictors of long-term outcome after mild traumatic brain

injury. Brain Inj 20: 759-765

45. Emmerling MR, Morganti-Kossmann MC, Kossmann T, Stahel PF, Watson

MD, Evans LM, Mehta PD, Spiegel K, Kuo YM, Roher AE, Raby CA (2000)

Traumatic brain injury elevates the Alzheimer's amyloid peptide A beta 42 in

human CSF. A possible role for nerve cell injury. Ann N Y Acad Sci 903: 118-122

46. Liliang PC, Liang CL, Weng HC, Lu K, Wang KW, Chen HJ, Chuang JH

(2010) tau Proteins in Serum Predict Outcome After Severe Traumatic Brain Injury.

Journal of Surgical Research 160: 302-307

131

CHAPTER V

FUTURE DIRECTIONS

5.1 Antibody-Antigen Interaction: Epitope Mapping and Kinetics

We have identified and synthesized a 10 amino acid peptide sequence that may be the immune trigger for anti-S100B autoantibody production. This potential epitope on human S100B was identified by a bioinformatic analysis using the immune epitope database. The specific peptide sequence, ELINNELSHF, is not found in any other human protein, but is common in pathogens such as Listeria or Salmonella. This S100B peptide has also been identified in another study where it was found to bind to MHC (1). In the study presented herein human IgG in sera were found to bind to ELINNELSHF by a reverse

ELISA method.

In order to further investigate whether the exact binding site of IgG to S100B is part of the ELINNELSHF sequence a method such as Biacore may be used. Biacore measures protein protein interactions, antibody and antigen in thus case, and binding 132 affinity by measuring the binging kinetics of the interaction (2). Biacore uses surface plasmon resonance to study protein protein interactions which require no labels. The response is measured by a change in mass concentration on the surface. Measurements can be made on body fluids, including human sera, and the specificity depends on the ligand attached to the surface. Binding sites on S100B can be mapped using Biacore to analyze antibody specificity. There are two approaches to epitope mapping. First, if antibodies (in human sera) bind simultaneously to the same antigen (S100B) this indicates distinct epitopes. If binding is interfered epitopes may be the same or closely situated or a conformational change in the antigen (S100B) induced by binding of the first antibody may hide other possible epitopes. Second, peptide inhibition techniques can be used to inhibit binding of the antibody. Specific S100B regions can be tested and if that peptide blocks antibody binding it is considered to represent a portion of the epitope for the antibody (IgG in human sera).

Another important aspect that can be studied using Biacore is the binding affinity and kinetics of the antigen antibody binding reaction. Kinetics are analyzed for the reaction over a range of antigen concentrations as a function of time. There are two methods for kinetic analysis using Biacore; multi-cycle and single-cycle. Multi-cycle analysis runs each analyte concentration in a separate cycle, single-cycle runs a range of concentrations of analyte in one cycle. Affinity analysis can be performed using mathematical calculations from kinetic constants for equilibrium association and dissociation. Additionally, steady state affinity can be analyzed only if the binding is 1:1, but may be more complex of there are two different binding sites.

133 To study the specificity and cross reactivity of the peptide in the reverse ELISA setting, a scrambled peptide could be synthesized. The sequence can be changed in ways to investigate the effect of charge and arrangement in the antigen antibody binding. The hypothesized epitope presented herein, ELINNELSHF, contains two negatively charged residues, four hydrophobic residues, and three polar residues, see Table 15. Changing the negatively charged glutamic acid residues to positively charged residues such as Lysine or

Arginine, could potentially block antibody antigen binding. Whereas keeping the charge the same and changing some the polar residues to hydrophobic residues could result in a lower binding specificity.

Amino Acid Name Side Chain Properties

E Glutamic Acid Negatively Charged

L Leucine Hydrophobic

I Isoleucine Hydrophobic

N Asparagine Polar

N Asparagine Polar

E Glutamic Acid Negatively Charged

L Leucine Hydrophobic

H Histidine Polar

F Phenylalanine Hydrophobic

Table 15: Amino acid properties of S100B peptide sequence ELINNELSHF.

134 5.2 Limitations of Enrolled Patient Population

A major limitation to studies involving human samples is the unpredictability of patient variables. In this study, to obtain a population of serum samples matched with MRI scans a lung cancer population was enrolled. For the first phase of the study in which serum biomarkers were analyzed for clinical utility in the detection of brain metastases in lung cancer, this population contained many metastasis negative patients. In order to further investigate the possibility for clinical use of the proposed blood test, more patients, especially metastasis positive patients, must be studied as there were only a total of 18 confirmed metastasis cases in the current cohort.

The second phase of the study involved the use of biomarkers in the detection of cerebral small vessel disease (SVD). The lung cancer population enrolled was enriched with SVD positive cases (61%), however the presence of cancer may be a confounder in the measurement of the proposed biomarkers, see chapter III. In addition, the severity of

SVD was unable to be completely investigated by our blood test due to the low incidence of cases with a SVD severity score greater than one. The age matched and comorbidity matched cancer free controls also pose a study limitation. These patients have risk factors associated with SVD but do not have corresponding neuroimaging data as it is not part of routine care. Thus these patients are at high risk for SVD, yet there is no diagnosis due to the lack of MRI. In addition, the low levels of anti-S100B autoantibodies measured in the high risk control samples compared to the cancer population further suggest that cancer may play a role in serum elevations.

135 Another aspect of the study that could improve both phases, is the collection of longitudinal follow-up serum samples and MRI scans. Longitudinal follow-up MRI scans of patients who are at risk for developing brain metastases are not routinely performed, primarily due to the cost of imaging (3-5). This can affect patient care as early detection is key for treatment modalities. Thus the potential of a blood based biomarker test for the early detection of brain metastases would be an inexpensive and non invasive alternative.

As mentioned above, the incidence of metastasis in this cohort is low, only 18 cases, due to the fact that serum samples and MRI are taken at time of diagnosis. If these patients were followed longitudinally there is a potential to capture more brain metastases. In the same regard, for the second phase of the study the additional serum samples and MRI scans over time could reveal increased SVD severity. Finally, if budget were to allow, the scans of cancer free controls at risk for SVD would provide insight into the potential effects of cancer diagnosis in the detection of SVD.

136 5.3 S100B Expression in Lung Cancer

Altered expression of S100 proteins have been observed in many human cancers

(6). The majority of S100s play a role in tumor progression, metastasis, and angiogenesis, yet few act as tumor suppressors. The measurement of S100 protein expression is thus useful in the diagnosis, prognosis, treatment, and monitoring of patient disease.

Specifically S100B has been a clinical marker in the diagnosis and therapeutic monitoring of melanoma (7-9). In lung cancer S100A2, S100A10, and S100A11 among others, have been shown to participate in cancer progression (10-13). The expression of S100B in lung tumor tissue has yet to be studied. S100B is an astrocytic protein primarily expressed in the brain, yet extracranial sources have been reported. The source of S100B as an antigen in the production of the observed autoimmune response in patients' sera, presented herein, needs to be further elucidated. Resected tumor tissue from patients in the cohort can be evaluated for S100B protein expression. S100B protein expression can be demonstrated by western blot using the protein fraction from homogenized tumor tissue. Gene expression of S100B RNA can also be evaluated in tumor tissue, see method (14). In addition,

Immunohistochemical analysis may be performed on tumor tissue. If S100B were to be expressed in lung tumor tissue it could serve as a potential target for treatment. For example, S100B inhibitors have been shown to delay the progression of Alzheimer's disease (15).

137 5.4 Conclusion

The results presented herein underscore the importance of brain-derived blood biomarkers in the diagnosis and understanding of complex processes responsible for cognitive and functional decline of the human brain. S100B, a protein normally shielded from the systemic circulation by the blood-brain barrier can act as a diffusible signal after disruption of the endothelial tight junctions that separate the brain from the blood. We have demonstrated the presence of anti-CNS immunity in subjects with lung cancer. The implications of elevated serum anti-S100B autoantibody levels are two-fold. In early stage lung cancer increased anti-S100B autoantibodies in combination with increased S100B protein levels are useful in the detection of CNS metastasis. In addition, chronic cerebrovascular disease and subsequent prolonged BBB permeability allows for the systemic exposure of S100B needed to illicit an immune response. Increased anti-S100B autoantibody levels were observed in patients with a positive diagnosis of SVD compared to patients without SVD. Interestingly, cancer free, age matched and comorbidity matched controls had lower anti-S100B autoantibody levels compared to both the SVD positive and

SVD negative cancer patients. Thus the possibility that lung cancer affects anti-S100B autoantibody serum levels needs to be further investigated. The overall relevance of misguided immune response needs to be studied, but the data suggest that S100B may specifically connect BBB permeability to immune response.

138

5.5 References

1. Fissolo S, Haag S, de Graaf KL, Drews O, Stevanovic S, Rammensee

HG, Weissert R. Naturally presented peptides on major histocompatibility complex

I and II molecules eluted from central nervous system of multiple sclerosis patients.

Mol Cell Proteomics. 2009 Sep;8(9):2090-101

2. Murphy, M., Jason-Moller, L. and Bruno, J. 2006. Using Biacore to

Measure the Binding Kinetics of an Antibody-Antigen Interaction. Current

Protocols in Protein Science. 45:19.14:19.14.1–19.14.17

3. Butler AR, Leo JS, Lin JP, Boyd AD, Kricheff, II. The value of routine

cranial computed tomography in neurologically intact patients with primary

carcinoma of the lung. Radiology. 1979;131(2):399-401

4. Cole FH, Jr., Thomas JE, Wilcox AB, Halford HH, 3rd. Cerebral imaging

in the asymptomatic preoperative bronchogenic carcinoma patient: is it

worthwhile? Ann Thorac Surg. 1994;57(4):838-840

5. Colice GL, Birkmeyer JD, Black WC, Littenberg B, Silvestri G. Cost-

effectiveness of head CT in patients with lung cancer without clinical evidence of

metastases. Chest. 1995;108(5):1264-1271

6. Bresnick AR, Weber DJ, Zimmer DB. S100 proteins in cancer. Nat

Rev Cancer. 2015 Feb;15(2):96-109

7. Cocchia D., Michetti F. and Donato R. (1981) S100 antigen in normal

human skin. Nature 294, 85–87

139 8. Bouwhuis M. G., Suciu S., Kruit W. et al. (2011) Prognostic value of serial

blood S100B determinations in stage IIB-III melanoma patients: a corollary study

to EORTC trial 18952. Eur. J. Cancer 47, 361–368

9. Cochran A. J., Lu H. F., Li P. X., Saxton R. and Wen D. R. (1993) S-100

protein remains a practical marker for melanocytic and other tumours. Melanoma

Res. 3, 325–330

10. Hao, J. Selective expression of S100A11 in lung cancer and its role in

regulating proliferation of adenocarcinomas cells. Mol. Cell Biochem

2012;359, 323–332

11. Yang X, Popescu NC and Zimonjic DB. DLC1 interaction with S100A10

mediates inhibition of in vitro cell invasion and tumorigenicity of lung cancer cells

through a RhoGAP-independent mechanism. Cancer Res 2011; 71: 2916- 2925

12. Hao J, Wang K, Yue Y, Tian T, Xu A, Xiao X and He D. Selective

expression of S100A11 in lung cancer and its role in regulating proliferation of

adenocarcinomas cells. Mol Cell Biochem 2012; 359: 323-332

13. Bulk, E. et al. S100A2 induces metastasis in non-small cell lung

cancer. Clin. Cancer Res 2009; 15, 22–29

14. Bargerstock E, Puvenna V, Iffland P, Falcone T, Hossain M, Vetter S, Man

S, Dickstein L, Marchi N, Ghosh C, Carvalho-Tavares J, Janigro D.

Is peripheral immunity regulated by blood-brain barrier permeability changes?

PLoS One. 2014 Jul 2;9(7):e101477

15. Heil, A. et al. S100P is a novel interaction partner and regulator of

IQGAP1. J. Biol. Chem.2011; 286, 7227–7238

140