Mass spectrometry imaging shows major derangements in neurogranin and in purine

metabolism in the triple-knockout 3×Tg Alzheimer mouse model.

Clara Esteve1, Emrys A. Jones1, Douglas B. Kell2,3, Hervé Boutin4,5, Liam A.

McDonnell*1,6

1Center for Proteomics and Metabolomics, Leiden University Medical Center, Leiden,

The Netherlands

2School of Chemistry, The University of Manchester, Manchester, Lancs M13 9PL, UK

3Manchester Institute of Biotechnology, The University of Manchester, 131 Princess St,

Manchester, Lancs, UK

4Faculty of Medicine and Human Sciences, The University of Manchester, Manchester,

UK

5Wolfson Molecular Imaging Center, The University of Manchester, Manchester, UK

6Fondazione Pisana per la Scienza ONLUS, Pisa, Italy.

* Corresponding author and reprint requests

Dr. Liam A. McDonnell, Center for Proteomics and Metabolomics, Leiden University

Medical Center, Einthovenweg 20, 2333 ZC Leiden, The Netherlands; E-mail:

l.a.mcdonnell@lumc.nl; Phone: +31 71 526 8744; Fax: +31 71 526 6907

1

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

Abstract

Matrix-assisted laser desorption/ionization (MALDI) mass spectrometry imaging (MSI)

can simultaneously measure hundreds of biomolecules directly from tissue. Using

different sample preparation strategies, proteins and metabolites have been profiled to

study the molecular changes in a 3×Tg mouse model of Alzheimer’s disease. In

comparison with wild-type (WT) control mice MALDI-MSI revealed Alzheimer’s

disease-specific protein profiles, highlighting dramatic reductions of a protein with m/z

7560, which was assigned to neurogranin and validated by immunohistochemistry. The

analysis also revealed substantial metabolite changes, especially in metabolites related

to the purine metabolic pathway, with a shift towards an increase in

hypoxanthine/xanthine/uric acid in the 3×Tg AD mice accompanied by a decrease in

AMP and adenine. Interestingly these changes were also associated with a decrease in

ascorbic acid, consistent with oxidative stress. Furthermore, the metabolite N-

arachidonyl taurine was increased in the diseased mouse brain sections, being highly

abundant in the hippocampus. Overall, we describe an interesting shift towards pro-

inflammatory molecules (uric acid) in the purinergic pathway associated with a decrease

in anti-oxidant level (ascorbic acid). Together, these observations fit well with the

increased oxidative stress and neuroinflammation commonly observed in AD.

Keywords: Alzheimer’s, 3×Tg mouse, neurogranin, purigenic pathway, mass

spectrometry imaging

2

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

Introduction

Alzheimer’s disease or Alzheimer-type dementia (AD) is a neurodegenerative disorder

and the most common form of dementia. Today AD is the largest unmet medical need in

neurology [1, 2]. As such, it represents a massive and growing health-care problem with

about 35 million patients in 2010, expected to reach 115 million by 2050, with 2/3 of

those patients living in low- to middle-income countries and the cost of long-term care

predicted to double over the next 50 years (from 1.2% to 2.5% of European countries’

GDP) [3-6]. To date, only symptomatic treatments are available and there is an urgent

need to investigate new approaches and concepts to develop new therapies, which

essentially depend on a much better understanding of the pathophysiology of AD. From

a physiopathological point of view, Alzheimer’s disease is characterised by several

abnormalities, such as elevated levels of amyloid- (Aβ) peptides leading to Aβ plaque

formation, neurofibrillary tangles (NFT) made of aggregates of abnormally

phosphorylated Tau proteins, alteration of the cholinergic system, brain atrophy,

decreased brain metabolism and neuroinflammation [7-9], not all of which contribute

[10] to the end point of synaptic and neuronal loss, and cognitive dysfunction. Iron

dysregulation leading to oxidative stress has also been strongly implicated in disease

progression [11-16]. Iron dysregulation may also interact with a potential microbial

component of inflammation [17-19] for which there is other and wide-ranging evidence

[20-23].

Traditionally, histopathological analysis of AD tissue relies on a priori hypotheses or

knowledge, and the use of known biomarkers and associated probes (e.g. antibodies,

radiolabelled tracers). Conversely, mass spectrometry imaging (MSI), commonly based

3

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

on matrix-assisted laser desorption/ionization (MALDI), can be used to image hundreds

of biomolecules directly from tissue [24, 25]. A number of mass spectrometric methods

have been developed for MSI, enabling high sensitivity analysis of a range of molecular

classes, including proteins, peptides and metabolites directly from tissue [26]. This is a

data-driven strategy [27], and thus requires no a priori knowledge about hypothetical

biomarkers. MSI can be applied to different molecular classes just by altering the tissue

preparation strategy [28, 29]. In many neurological diseases the pathophysiology is not

entirely known and a beneficial approach involves systematic investigations of the

biomolecular differences between diseased and healthy tissue. MSI has been used to

investigate spatiotemporal molecular changes in several neurological pathological

conditions in rodent models, including ischemic stroke [30-32], cortical spread

depression [29], Parkinson’s disease [33, 34] and Alzheimer’s disease [35, 36]. A key

advantage of MSI is that it can annotate tissues based on their MS profiles and thereby

distinguish biomolecularly distinct regions even if they are unexpected or cannot be

distinguished using established histological and histochemical methods [37, 38]. This is

especially true for disorders such Alzheimer’s disease, which is a complex and

multifactorial disease and for which the pathophysiology is not well understood.

The triple transgenic (3Tg) mouse model [39] expressed the human mutations for

Presenilin (PresenilinM146V), the amyloid protein precursor APPSwe essential to obtain

amyloid pathology in mice [40] as well as the tauP301L mutation that induces tau

pathology [41, 42], and is considered to present all the pathological hallmarks of AD.

Here we report the results of a MALDI MSI study of the biomolecular changes related

to Alzheimer-type disease in the brain of the 3Tg mouse, including both metabolites and

proteins.

4

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

Materials and Methods

Chemicals

9-aminoacridine (9-AA), sinapinic acid (SA), poly-L-lysine solution, isopropanol,

methanol (MeOH), ethanol and trifluoroacetic acid (TFA) were obtained from Sigma-

Aldrich (St. Louis, MO, USA). Indium tin oxide (ITO)-obtained slide glass were

purchased from Bruker Daltonics (Bremen, Germany).

Animals

Wild type (WT) and triple transgenic (3×Tg) incorporating the transgenes PS1M146V,

APPSwe, and tauP301L [39] adult male mice (n = 3 per group) were kept under a 12 h

light–dark cycle (08:00/20:00 h) at 22 °C with free access to food and water. Mice were

killed by anaesthetic overdose with isoflurane and decapitated. Brains were rapidly (<60

s) removed and frozen in cooled (-40 °C) isopentane. All procedures were carried out in

accordance with the Animals (Scientific Procedures) Act 1986.

MSI Data Acquisition

For the MALDI MSI experiments sagittal tissue sections of 12 µm thickness were cut

with a cryostat (1720 Digital, Leica, Rijswijk, The Netherlands) at -20 °C. Next these

were thaw-mounted onto an ITO-coated glass slides (Delta Technologies, Stillwater,

MN) previously coated with poly-L-lysine (0.05 % in water), and stored at -80 °C.

Before use, the tissues were slowly brought to room temperature in a desiccator.

Energy-rich metabolites, such as ATP, can degrade in tissue. It has been shown that

post-mortem degradation is more rapid under normal physiological conditions

(Gündisch et al., 2012). Note: there are technologies available specifically designed to

5

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

limit metabolite degradation (Sugiura et al., 2014; Blatherwick et al., 2013) but they

were not available at the University of Manchester where the animal experiments were

performed. Instead, we prepared the tissues in a manner to try to limit metabolic

degradation, specifically

• All brains were excised and frozen in cooled (-40 C) isopentane within 60s of

sacrifice;

• All ITO slides were pre-cooled in the cryostat chamber prior to tissue mounting;

• All samples were mounted within the chamber and at the same time - tissues not

being cut were covered in foil to avoid drying;

• Sections were taken and placed onto the relevant slide with just enough heat

applied to the underside of the slide to adhere the tissue section, and then immediately

refrozen in the cryostat chamber;

• The slides were then placed in a slidebox within the chamber until all sectioning

was complete;

• At the end of sectioning the slidebox was transferred to the -80 C freezer;

• For preparation the tissue sections were lyophilized in a freeze drier immediately

from the -80 C freezer.

To ensure reducibility and minimize the impact of any systematic source of bias the

experiments were performed in biological triplicate and the MSI data acquisition from

the 3xTg and WT mice performed in a pairwise manner (3xTg-WT etc); each slide

contained one 3xTg and one WT tissue section, thus were simultaneously prepared and

measured during the same MSI run. After MSI data acquisition the matrix was washed

off with 70 % ethanol and the tissue samples stained with cresyl violet (Nissl stain).

6

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

Histological images were scanned with a MIRAX DESK digital slide scanner (3D

Histech, Hungary).

Protein MSI

The tissues were first washed in ice-cold isopropanol to fix the tissues and remove salts

and lipids. The washed tissue sections were then allowed to dry in air prior to matrix

deposition using the ImagePrep (Bruker Daltonics, Bremen, Germany) device. A

solution of 20 mg/mL SA in 70 % isopropanol with 0.1 % TFA was used as organic

matrix. MALDI MSI experiments were then performed using an Autoflex III MALDI-

linear-TOF (Bruker Daltonics), 100-200 µm pixel size, and 600 laser shots per pixel (50

laser shots per position of a random walk within each pixel). Positively charged ions

between m/z 2000 and 25000 were detected. Data acquisition, pre-processing (mass

spectral smoothing, baseline subtraction and total-ion-count normalization) and data

visualization were performed using the Flex software suite (FlexControl 3.0,

FlexAnalysis 3.0, FlexImaging 2.1).

Metabolite MSI

A uniform coating of 9-AA was added using the ImagePrep and a solution of 10 mg/mL

in 70 % MeOH. Metabolite MSI experiments were performed using a 9.4 T Apex Q

MALDI-FTICR (Bruker Daltonics) in negative ion mode, using a 50-150 µm pixel size,

in the range 50-1000 m/z by averaging signals from 600 laser shots per pixel (50 laser

shots per position of a random walk within each pixel; random walk enabled through

adapting the experiment pulse program). Each pixel’s mass spectrum was recalibrated

using the matrix peak of 9-AA (m/z 193,0771219) as an internal lock mass and

7

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

normalized using the root-mean-square method. Data acquisition, pre-processing and

data visualization were performed using the Flex software suite (Compass 1.3,

FlexControl 3.0, FlexImaging 2.1).

MSI datasets contain an individual mass spectrum for each pixel. The ultrahigh mass

resolution of FTICR mass spectrometry generates large files and, thus, very large MSI

datasets (often >20 GB for a single tissue). An automated feature identification and

extraction routine was applied to extract the images of the detected peaks, thereby

reducing the data load by approximately a thousand fold [43].

MSI data filtering and pathway analysis

To further focus the analysis on known metabolites and to introduce metabolic pathway

information, the reduced MALDI MSI datasets were filtered using the Human

Metabolome Database. For each entry in the database, the compound names, exact

masses (converted to [M-H]- ions), and pathway memberships were extracted. Only

metabolite ions that corresponded to [M-H]- ions, within a mass tolerance of 1 ppm, and

only those whose isotopic profile matched that of the theoretical distribution (Pearson

correlation > 0.95, performed in Matlab) were retained.

Validation of proteins by immunohistochemistry

For all the procedure described below Phosphate Buffered saline (PBS) at 100 mM was

used. Frozen mouse brain sections were post-fixed in paraformaldehyde (4 % in PBS)

for 30 min and washed (6×5 min) in PBS. Sections were incubated for 30 min in 0.1 %

Triton X-100 containing 2 % normal donkey serum in PBS to block non-specific

binding. Without further washing, sections were incubated overnight at 4 °C with Anti-

8

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

Neurogranin antibody (Abcam ab23570, 1:500) in 2 % normal donkey serum/0.1 %

Triton X-100 in PBS. Sections were then washed (3×10 min) in PBS and incubated for

2 h at room temperature with secondary antibody (AlexaFluor 594 nm donkey anti-

rabbit IgG, Molecular Probes, Invitrogen, 1:500) in 2 % normal donkey serum/0.1 %

Triton X-100 in PBS) and then washed again (3×10 min) in PBS. Sections were

mounted with a Prolong Antifade kit (Molecular Probes, Invitrogen) with DAPI.

Sections incubated without the primary antibodies served as negative controls.

Images were collected on a Olympus BX51 upright microscope using a 4×/0.13,

10×/0.30 or 40×/0.50 UPlanFLN objectives and captured using a Coolsnap ES camera

(Photometrics) through MetaVue Software (Molecular Devices). Images were then

processed and analysed using ImageJ (http://rsb.info.nih.gov/ij).

Results

The non-targeted nature of MSI led us to investigate whether it could be used to

investigate the chemical and spatial extent of the molecular disturbances occurring in an

Alzheimer’s disease rodent model. MSI is able to analyse different molecular classes by

changing sample preparation strategies and optimizing the mass spectrometer for the

mass range of the molecular class of interest. Furthermore, by combining MSI with

histology the differential MS profiles found in specific histopathological entities can be

used to identify candidate biomarkers [44].

Visualization of protein changes in Alzheimer’s disease mouse brain

MSI is a key technique in the visualization of intact proteins in tissue [45]. To detect

protein signals associated with Alzheimer’s disease, protein profiles from mouse brain

9

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

sagittal tissue sections from three WT and three 3×Tg mice were acquired by MALDI-

MSI using SA as a matrix. Figure 1a) shows a typical average mass spectrum obtained

from a single WT mouse brain sagittal tissue section (green line), a single 3×Tg mouse

brain tissue section (red line), and the average mass spectrum of the entire data set

(purple line). A comparison of the WT and 3xTg average mass spectra revealed that

while many peaks were identical, two signals with m/z values of 7560 and 7870 showed

a much higher intensity in the spectra from WT mouse brains. Neurogranin is a small

78aa [46] post-synaptic protein. The peak at m/z 7560 was tentatively assigned to (the

sodium adduct of) neurogranin from the literature [46]. The distribution of this

neurogranin peak in the brain is shown in Figure 1b) in green. The peak at m/z 6400

corresponding to PEP-19 (in blue) and the peak at m/z 14200 corresponding to myelin

basic protein (in red) are shown as morphological references. Neurogranin is mainly

detected in healthy brain sections, being localized in the isocortex (indicated as area “a”

in the brain scheme), in particular in the prelimbic, somatomotor and anterior cingulate

areas. In order to ensure reproducibility, different animal pairs were analysed, showing

in all cases a m/z 7560 intensity difference between WT and 3×Tg mouse brain sections

(Figure 1.c). A t-test of the average intensity of the neurogranin MSI signal in the

isocortex, and using the Benjamini-Hochberg multiple testing correction, confirmed the

statistical significance of the detected changes, p<0.01. For a final validation, the

presence and distribution of the protein neurogranin was confirmed by

immunohistochemistry (IHC), as shown in Figure 1.d. No detectable signal was

observed for brain sections incubated without the primary antibody.

Visualization of metabolites changes in Alzheimer’s disease mouse brain

10

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

For fundamental reasons, metabolomics is necessarily more sensitive than is proteomics

for detecting biochemical changes [47]. Metabolite MSI offers enormous clinical

potential by enabling molecule-specific imaging of a class of biomolecules for which

alternative histological stains typically lack molecular specificity, e.g. Oil Red stains

lipids and trialycerides. Furthermore, when combined with known metabolic pathways

[48], metabolite MSI provides a means to image the activities of pathways in tissues

directly. The matrix used for metabolite analysis was 9-AA, commonly used due to the

low intensity of matrix background ions [49]. The MS data acquisition was performed

on a 9.4T FTICR mass spectrometer, which provides routine ultrahigh mass resolution

and accurate mass in the low m/z range, enabling each mass spectral peak to be fully

resolved. The assignments were based on accurate mass. By using known pathway

information, it was observed that control and Alzheimer’s disease mouse brain sagittal

sections showed highly distinct metabolic signatures. Figure 2 shows the purine

metabolic pathway, indicating the distributions of detected ions within this pathway for

WT and 3×Tg mice. By using MALDI-MSI we were able to detect a number of the

constituent metabolites, although others were below the limit of detection. Figure 2

shows several clear differences in metabolite concentration between WT and 3×Tg

mouse brain sections. The concentration of adenine and AMP are higher in the WT

mouse, while other metabolites like inosine, hypoxanthine, xanthine, and uric acid are

increased in the 3×Tg mouse. Furthermore, ascorbic acid, albeit not in the purine

metabolic ‘pathway’, shows a much higher abundance in WT brain sections.

Apart from the purine metabolic pathway, and as shown for ascorbic acid, the

concentrations of some other metabolites appear to be altered in the case of 3×Tg mouse

brain. As shown in Figure 3, when mass spectra of both healthy and AD mouse brains

11

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

were compared, a peak at m/z 410.2373 appeared to be a lot more intense for AD brain.

For the identity assignment of this peak, the METLIN database was chosen. This a

freely accessible web-based data repository, that has been developed to assist in

metabolite research and to facilitate metabolite identification through mass analysis

[50]. The mass of the peak was assigned, with a mass tolerance of < 1 ppm, to N-

arachidonoyl taurine, a fatty acyl amide of the amino acid taurine. The tissue

distribution of N-arachidonyl taurine in brain sections imaged at 150 µm pixel center-

to-pixel center distance is also shown in Figure 3 (in green). The peak at m/z 885.6,

corresponding to phosphatidyl inositol (18:0) (in red), is shown as a morphological

reference. As shown in the 50 µm resolution image, N-arachidonyl taurine (in orange) is

mainly detected in the 3×Tg brain, and localized in the hippocampal region. However, it

is not homogeneously distributed, showing higher intensities in the stratum oriens and

dentate gyrus regions, and lower intensities in the pyramidal layer and stratum radiatum

regions.

Discussion

MALDI MSI has several distinct advantages for neuroscience and neuropathological

research. MALDI MSI provides simultaneous label free, multiplex imaging of

endogenous biomolecules. Using essentially the same technology but different sample

preparation methods MALDI MSI may be used to analyze peptides, proteins, lipids,

metabolites, neurotransmitters and even N-linked glycans; for several of which there

exists no generally applicable method, analogous to immunohistochemistry, that may be

used to record molecule-specific distributions. Furthermore, any enzymatic/metabolic

change that results in a change in mass can be resolved. For example MALDI MSI has

12

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

detected changes in post-translational modifications in neuropeptides and proteins and

even, via the introduction of 13C-labeled glucose, traced metabolic flux (Sugiura et al.,

2014).

MALDI MSI data is routinely registered to histology, and has been registered to both

fluorescence microscopy and magnetic resonance imaging, thus enabling cell specific

molecular profiles to be obtained, within their correct histopathological context, and to

investigate/utilize associations with in-vivo imaging methods. These capabilities are

particularly suited to discovery research, such as reported here, which benefit from the

broad molecular scope of the analysis for the discovery of previously unknown

molecular changes.

The literature on many diseases is data-rich and tested-hypothesis-poor. Given the

essential lack of treatments available, the dementias largely fall into this class. The

issues are compounded by (i) the difficulty of assessing cognitive function accurately,

and (ii) that (in the case of genuine Alzheimer’s in humans) a definitive diagnosis is

possible solely post mortem. Thus data-driven strategies that seek to discriminate

patients from controls are appropriate, as they can give strong indications as to which

surrogate variables are different in disease vs control samples. One hypothesis is then

that restoring the variable to their ‘normal’ range might be of therapeutic benefit, but

discovering the chief differences is necessarily the first need. Mass spectrometry

imaging strategies represent an exceptionally useful strategy for discovering such

markers, and were applied herein to assess differences between brain slices taken from

3×Tg mice and wild-type controls. We analysed both the proteome and the metabolome.

13

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

In the present case, we discovered a major change in the AD proteome, in that a peak at

m/z 7560 assigned to the post-synaptic protein neurogranin (Ng), and confirmed by

immunohistochemistry, was significantly down-regulated in the 3×Tg mouse. This

finding is in agreement with previous reports, which have shown a loss of Ng with

normal ageing in mice [51], in the AD mouse model Tg2576 [52] and in the brains of

AD patients by western blot and immunohistochemistry [53, 54]. Conversely Ng has

been found to be increased in the CSF of Alzheimer’s patients [55-60]. The levels of

proteins in brain parenchyma can correlate positively or negatively with levels of

proteins in CSF: increased production of pTau in AD brain results in increased levels in

CSF, while increased aggregation of A40-42 in AD brains results in decreased levels of

A in CSF; we can therefore hypothesise that neuronal loss results in decreased Ng

immunohistochemical staining and release of neurogranin in to the interstitial fluid and

CSF [61].

Even more interesting and novel were the findings that purine metabolism was seriously

deregulated, with substantial decreases in the concentration of adenine and AMP but

with other metabolites like inosine, hypoxanthine, xanthine, and uric acid being

noticeably increased in the 3×Tg mouse. Furthermore, ascorbic acid was much

decreased in the 3×Tg mouse. These findings indicate that the care taken to limit post-

mortem degradation of metabolites during the excision of the mouse brains, as well as

their sectioning, enabled the detection of hydrophilic small molecules that differ

reproducibly and discernibly between samples. Furthermore, these differences are

consistent with known AD pathology.

14

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

There is considerable evidence for the roles of purinergic signalling [62], and especially

the role of uric acid in Alzheimer’s disease (Kaddurah-Daouk et al., 2013; McFarland et

al., 2013) and in a variety of kinds of inflammatory process, including peanut allergy

[63, 64], pro-inflammatory cytokine production [65, 66], the Plasmodium falciparum-

induced inflammatory response [67] and the mechanistic basis for the action of alum as

an adjuvant [68]. Adenosine and inosine mediate anti-inflammatory effects via A2 and

A3 receptors [69, 70] (and IL-1β feeds back thereupon [71]), while purine metabolism

contributes to the anti-inflammatory action of aspirin [72]. The general consonance

between the metabolic ‘signature’ (i.e. those metabolites noticeably up- and down-

regulated in both the 3×Tg mouse and the hyperallergic response [64] is especially

striking, implying a similar overall regulation. We have also shown experimentally that

uric acid is raised significantly following heart failure [73]

Although a raised uric acid in serum is associated with a lower likelihood of AD and

dementia [74, 75], its relation to brain levels of uric acid is unknown.

The loss of ascorbate in the 3×Tg mouse is consistent with the well-known oxidative

stress accompanying, and probably contributing to, AD [76]. What has not been seen

previously is the derangement of purine metabolism that we report here. It would seem

that purine metabolism forms an important link with cytokine signalling as part of

neuroinflammation. Given the role of neuroinflammation [77, 78] and diet [79] in

exacerbating AD and other neurodegenerative disorders, this warrants considerable

further investigation.

Acknowledgments

15

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

DBK thanks the Biotechnology and Biological Sciences Research Council (grant

BB/L025752/1) for support. This is also a contribution from the Manchester Centre for

Synthetic Biology of Fine and Speciality Chemicals (SYNBIOCHEM) (BBSRC grant

BB/M017702/1). CE and LAM thank the support of the ZonMW Zenith project

“Imaging Mass Spectrometry-Based Molecular Histology: Differentiation and

Characterization of Clinically Challenging Soft Tissue Sarcomas” (No. 93512002;

B.H.) and the ICT consortium COMMIT project “e-biobanking with Imaging”.

16

347

348

349

350

351

352

353

References

[1] K.N. Fargo, P. Aisen, M. Albert, R. Au, M.M. Corrada, S. DeKosky, D. Drachman, H. Fillit, L. Gitlin, M. Haas, K. Herrup, C. Kawas, A.S. Khachaturian, Z.S. Khachaturian, W. Klunk, D. Knopman, W.A. Kukull, B. Lamb, R.G. Logsdon, P. Maruff, M. Mesulam, W. Mobley, R. Mohs, D. Morgan, R.A. Nixon, S. Paul, R. Petersen, B. Plassman, W. Potter, E. Reiman, B. Reisberg, M. Sano, R. Schindler, L.S. Schneider, P.J. Snyder, R.A. Sperling, K. Yaffe, L.J. Bain, W.H. Thies, M.C. Carrillo, W. Alzheimer's Association National Plan Milestone, 2014 Report on the Milestones for the US National Plan to Address Alzheimer's Disease, Alzheimer's & dementia : the journal of the Alzheimer's Association, 10 (2014) S430-452.

[2] A. Alzheimers, 2015 Alzheimer's disease facts and figures, Alzheimer's & dementia : the journal of the Alzheimer's Association, 11 (2015) 332-384.

[3] C.P. Ferri, M. Prince, C. Brayne, H. Brodaty, L. Fratiglioni, M. Ganguli, K. Hall, K. Hasegawa, H. Hendrie, Y.Q. Huang, A. Jorm, C. Mathers, P.R. Menezes, E. Rimmer, M. Scazufca, I. Alzheimers Dis, Global prevalence of dementia: a Delphi consensus study, Lancet, 366 (2005) 2112-2117.

[4] L. Jonsson, A. Wimo, The Cost of Dementia in Europe A Review of the Evidence, and Methodological Considerations, Pharmacoeconomics, 27 (2009) 391-403.

[5] A.P. Wimo, M. , World Alzheimer Report 2010 Executive Summary: The Global Economic Impact of Dementia., Alzheimer's Disease International, (2010).

[6] M.P. Prince, M.; Guerchet, M., World Alzheimer Report 2013 Executive Summary:An analysis of long-term care for dementia., Alzheimer's Disease International, (2013).

[7] H.-C. Huang, Z.-F. Jiang, Accumulated Amyloid-beta Peptide and Hyperphosphorylated Tau Protein: Relationship and Links in Alzheimer's Disease, Journal of Alzheimers Disease, 16 (2009) 15-27.

[8] D.M. Walsh, D.J. Selkoe, A beta Oligomers - a decade of discovery, Journal of Neurochemistry, 101 (2007) 1172-1184.

[9] A. Kadir, A. Nordberg, Target-Specific PET Probes for Neurodegenerative Disorders Related to Dementia, Journal of Nuclear Medicine, 51 (2010) 1418-1430.

[10] K. Herrup, The case for rejecting the amyloid cascade hypothesis, Nature Neuroscience, 18 (2015) 794-799.

[11] D.B. Kell, Iron behaving badly: inappropriate iron chelation as a major contributor to the aetiology of vascular and other progressive inflammatory and degenerative diseases, Bmc Medical Genomics, 2 (2009).

17

354

355356357358359360361362

363364

365366367368

369370

371372

373374

375376377

378379

380381

382383

384385386

[12] D.B. Kell, Towards a unifying, systems biology understanding of large-scale cellular death and destruction caused by poorly liganded iron: Parkinson's, Huntington's, Alzheimer's, prions, bactericides, chemical toxicology and others as examples, Archives of Toxicology, 84 (2010) 825-889.

[13] D.J. Bonda, H.-g. Lee, J.A. Blair, X. Zhu, G. Perry, M.A. Smith, Role of metal dyshomeostasis in Alzheimer's disease, Metallomics, 3 (2011) 267-270.

[14] G. Casadesus, M.A. Smith, X.W. Zhu, G. Aliev, A.D. Cash, K. Honda, R.B. Petersen, G. Perry, Alzheimer disease: Evidence for a central pathogenic role of iron-mediated reactive oxygen species, Journal of Alzheimers Disease, 6 (2004) 165-169.

[15] M.A. Smith, X. Zhu, M. Tabaton, G. Liu, D.W. McKeel, Jr., M.L. Cohen, X. Wang, S.L. Siedlak, B.E. Dwyer, T. Hayashi, M. Nakamura, A. Nunomura, G. Perry, Increased Iron and Free Radical Generation in Preclinical Alzheimer Disease and Mild Cognitive Impairment, Journal of Alzheimers Disease, 19 (2010) 363-372.

[16] S. Ayton, N.G. Faux, A.I. Bush, I. Alzheimers Dis Neuroimaging, Ferritin levels in the cerebrospinal fluid predict Alzheimer's disease outcomes and are regulated by APOE, Nature Communications, 6 (2015).

[17] M. Potgieter, J. Bester, D.B. Kell, E. Pretorius, The dormant blood microbiome in chronic, inflammatory diseases, Fems Microbiology Reviews, 39 (2015) 567-591.

[18] D.P. Kell, Marnie; Pretorius, Etheresia, Individuality, phenotypic differentiation, dormancy and ‘persistence’ in culturable bacterial systems: commonalities shared by environmental, laboratory, and clinical microbiology, F1000Research, 4 (2015).

[19] D.B.P. Kell, E., On the translocation of bacteria and their lipopolysaccharides between blood and peripheral locations in chronic, inflammatory diseases: the central roles of LPS and LPS-induced cell death., Integrative Biology, (2015).

[20] B.J. Balin, C.S. Little, C.J. Hammond, D.M. Appelt, J.A. Whittum-Hudson, H.C. Gerard, A.P. Hudson, Chlamydophila pneumoniae and the etiology of late-onset Alzheimer's disease, Journal of Alzheimers Disease, 13 (2008) 371-380.

[21] R.F. Itzhaki, Herpes simplex virus type 1 and Alzheimer's disease: increasing evidence for a major role of the virus, Frontiers in Aging Neuroscience, 6 (2014).

[22] J. Miklossy, Alzheimer's disease - a neurospirochetosis. Analysis of the evidence following Koch's and Hill's criteria, Journal of Neuroinflammation, 8 (2011).

[23] J. Miklossy, Historic evidence to support a causal relationship between spirochetal infections and Alzheimer's disease, Frontiers in Aging Neuroscience, 7 (2015).

[24] J.L. Norris, R.M. Caprioli, Analysis of Tissue Specimens by Matrix-Assisted Laser Desorption/Ionization Imaging Mass Spectrometry in Biological and Clinical Research, Chemical Reviews, 113 (2013) 2309-2342.

18

387388389390

391392

393394395

396397398399

400401402

403404

405406407

408409410

411412413

414415

416417

418419

420421422

[25] B. Balluff, C. Schoene, H. Hoefler, A. Walch, MALDI imaging mass spectrometry for direct tissue analysis: technological advancements and recent applications, Histochemistry and Cell Biology, 136 (2011) 227-244.

[26] K.E. Burnum, S.L. Frappier, R.M. Caprioli, Matrix-Assisted Laser Desorption/Ionization Imaging Mass Spectrometry for the Investigation of Proteins and Peptides, Annual Review of Analytical Chemistry, Place Published, 2008, pp. 689-705.

[27] D.B. Kell, S.G. Oliver, Here is the evidence, now what is the hypothesis? The complementary roles of inductive and hypothesis-driven science in the post-genomic era, Bioessays, 26 (2004) 99-105.

[28] A. Bodzon-Kulakowska, P. Suder, Imaging Mass Spectrometry: Instrumentation, applications, and combincation with other visualization techniques, Mass Spectrom. Rev., 35 (2015) 147-169.

[29] E.A. Jones, R. Shyti, R.J.M. van Zeijl, S.H. van Heiningen, M.D. Ferrari, A.M. Deelder, E.A. Tolner, A. van den Maagdenberg, L.A. McDonnell, Imaging mass spectrometry to visualize biomolecule distributions in mouse brain tissue following hemispheric cortical spreading depression, Journal of Proteomics, 75 (2012) 5027-5035.

[30] D. Miura, Y. Fujimura, M. Yamato, F. Hyodo, H. Utsumi, H. Tachibana, H. Wariishi, Ultrahighly Sensitive in Situ Metabolomic Imaging for Visualizing Spatiotemporal Metabolic Behaviors, Analytical Chemistry, 82 (2010) 9789-9796.

[31] M. Irie, Y. Fujimura, M. Yamato, D. Miura, H. Wariishi, Integrated MALDI-MS imaging and LC-MS techniques for visualizing spatiotemporal metabolomic dynamics in a rat stroke model, Metabolomics, 10 (2014) 473-483.

[32] K. Hattori, M. Kajimura, T. Hishiki, T. Nakanishi, A. Kubo, Y. Nagahata, M. Ohmura, A. Yachie-Kinoshita, T. Matsuura, T. Morikawa, T. Nakamura, M. Setou, M. Suematsu, Paradoxical ATP Elevation in Ischemic Penumbra Revealed by Quantitative Imaging Mass Spectrometry, Antioxidants & Redox Signaling, 13 (2010) 1157-1167.

[33] M. Shariatgorji, A. Nilsson, R.J.A. Goodwin, P. Kallback, N. Schintu, X. Zhang, A.R. Crossman, E. Bezard, P. Svenningsson, P.E. Andren, Direct Targeted Quantitative Molecular Imaging of Neurotransmitters in Brain Tissue Sections, Neuron, 84 (2014) 697-707.

[34] A. Ljungdahl, J. Hanrieder, M. Faelth, J. Bergquist, M. Andersson, Imaging Mass Spectrometry Reveals Elevated Nigral Levels of Dynorphin Neuropeptides in L-DOPA-Induced Dyskinesia in Rat Model of Parkinson's Disease, Plos One, 6 (2011).

[35] R.W. Hutchinson, A.G. Cox, C.W. McLeod, P.S. Marshall, A. Harper, E.L. Dawson, D.R. Howlett, Imaging and spatial distribution of beta-amyloid peptide and metal ions in Alzheimer's plaques by laser ablation-inductively coupled plasma-mass spectrometry, Analytical Biochemistry, 346 (2005) 225-233.

[36] L. Carlred, A. Gunnarsson, S. Sole-Domenech, B. Johansson, V. Vukojevic, L. Terenius, A. Codita, B. Winblad, M. Schalling, F. Hook, P. Sjovall, Simultaneous Imaging of Amyloid-beta and

19

423424425

426427428

429430431

432433434

435436437438

439440441

442443444

445446447448

449450451

452453454

455456457458

459460

Lipids in Brain Tissue Using Antibody-Coupled Liposomes and Time-of-Flight Secondary Ion Mass Spectrometry, Journal of the American Chemical Society, 136 (2014) 9973-9981.

[37] B.N. Mathur, R.M. Caprioli, A.Y. Deutch, Proteomic Analysis Illuminates a Novel Structural Definition of the Claustrum and Insula, Cerebral Cortex, 19 (2009) 2372-2379.

[38] A. Roempp, S. Guenther, Z. Takats, B. Spengler, Mass spectrometry imaging with high resolution in mass and space (HR2 MSI) for reliable investigation of drug compound distributions on the cellular level, Analytical and Bioanalytical Chemistry, 401 (2011) 65-73.

[39] S. Oddo, A. Caccamo, J.D. Shepherd, M.P. Murphy, T.E. Golde, R. Kayed, R. Metherate, M.P. Mattson, Y. Akbari, F.M. LaFerla, Triple-transgenic model of Alzheimer's disease with plaques and tangles: Intracellular A beta and synaptic dysfunction, Neuron, 39 (2003) 409-421.

[40] A. Willuweit, J. Velden, R. Godemann, A. Manook, F. Jetzek, H. Tintrup, G. Kauselmann, B. Zevnik, G. Henriksen, A. Drzezga, J. Pohlner, M. Schoor, J.A. Kemp, H. von der Kammer, Early-Onset and Robust Amyloid Pathology in a New Homozygous Mouse Model of Alzheimer's Disease, Plos One, 4 (2009).

[41] D.C. Lee, J. Rizer, M.-L.B. Selenica, P. Reid, C. Kraft, A. Johnson, L. Blair, M.N. Gordon, C.A. Dickey, D. Morgan, LPS- induced inflammation exacerbates phosphotau pathology in rTg4510 mice, Journal of Neuroinflammation, 7 (2010).

[42] P.J. Khandelwal, S.B. Dumanis, A.M. Herman, G.W. Rebeck, C.E.H. Moussa, Wild type and P301L mutant Tau promote neuro-inflammation and alpha-Synuclein accumulation in lentiviral gene delivery models, Molecular and Cellular Neuroscience, 49 (2012) 44-53.

[43] L.A. McDonnell, A. van Remoortere, N. de Velde, R.J.M. van Zeijl, A.M. Deelder, Imaging Mass Spectrometry Data Reduction: Automated Feature Identification and Extraction, Journal of the American Society for Mass Spectrometry, 21 (2010) 1969-1978.

[44] L.H. Cazares, D. Troyer, S. Mendrinos, R.A. Lance, J.O. Nyalwidhe, H.A. Beydoun, M.A. Clements, R.R. Drake, O.J. Semmes, Imaging Mass Spectrometry of a Specific Fragment of Mitogen-Activated Protein Kinase/Extracellular Signal-Regulated Kinase Kinase Kinase 2 Discriminates Cancer from Uninvolved Prostate Tissue, Clinical Cancer Research, 15 (2009) 5541-5551.

[45] L. MacAleese, J. Stauber, R.M.A. Heeren, Perspectives for imaging mass spectrometry in the proteomics landscape, Proteomics, 9 (2009) 819-834.

[46] C.M. deArrieta, L.P. Jurado, J. Bernal, A. Coloma, Structure, organization, and chromosomal mapping of the human neurogranin gene (NRGN), Genomics, 41 (1997) 243-249.

[47] R. Goodacre, S. Vaidyanathan, W.B. Dunn, G.G. Harrigan, D.B. Kell, Metabolomics by numbers: acquiring and understanding global metabolite data, Trends in Biotechnology, 22 (2004) 245-252.

[48] I. Thiele, N. Swainston, R.M.T. Fleming, A. Hoppe, S. Sahoo, M.K. Aurich, H. Haraldsdottir, M.L. Mo, O. Rolfsson, M.D. Stobbe, S.G. Thorleifsson, R. Agren, C. Boelling, S. Bordel, A.K.

20

461462

463464

465466467

468469470

471472473474

475476477

478479480

481482483

484485486487488

489490

491492

493494495

496497

Chavali, P. Dobson, W.B. Dunn, L. Endler, D. Hala, M. Hucka, D. Hull, D. Jameson, N. Jamshidi, J.J. Jonsson, N. Juty, S. Keating, I. Nookaew, N. Le Novere, N. Malys, A. Mazein, J.A. Papin, N.D. Price, E. Selkov, Sr., M.I. Sigurdsson, E. Simeonidis, N. Sonnenschein, K. Smallbone, A. Sorokin, J.H.G.M. van Beek, D. Weichart, I. Goryanin, J. Nielsen, H.V. Westerhoff, D.B. Kell, P. Mendes, B.O. Palsson, A community-driven global reconstruction of human metabolism, Nature Biotechnology, 31 (2013) 419-+.

[49] F. Benabdellah, D. Touboul, A. Brunelle, O. Laprevote, In Situ Primary Metabolites Localization on a Rat Brain Section by Chemical Mass Spectrometry Imaging, Analytical Chemistry, 81 (2009) 5557-5560.

[50] C.A. Smith, G. O'Maille, E.J. Want, C. Qin, S.A. Trauger, T.R. Brandon, D.E. Custodio, R. Abagyan, G. Siuzdak, METLIN - A metabolite mass spectral database, Therapeutic Drug Monitoring, 27 (2005) 747-751.

[51] N. Mons, V. Enderlin, R. Jaffard, P. Higueret, Selective age-related changes in the PKC-sensitive, calmodulin-binding protein, neurogranin, in the mouse brain, J. Neurochem., 79 (2001) 859-867.

[52] A.J. George, L. Gordon, T. Beissbarth, I. Koukoulas, R.M.D. Holsinger, V. Perreau, R. Cappai, S.-S. Tan, C.L. Masters, H.S. Scott, Q.-X. Li, A Serial Analysis of Gene Expression Profile of the Alzheimer’s Disease Tg2576 Mouse Model, Neurotox. Res., 17 (2010) 360-379.

[53] E. Bereczki, P.T. Francis, D. Howlett, J.B. Pereira, K. Höglund, A. Bogstedt, A. Cedazo-Minguez, J.-H. Baek, T. Hortobágyi, J. Attems, C. Ballard, D. Aarsland, Synaptic proteins predict cognitive decline in Alzheimer's disease and Lewy body dementia, Alzheimers Dement., 12 (2016) 1149-1158.

[54] P.H. Reddy, G. Mani, B.S. Park, J. Jacques, G. Murdoch, W. Whetsell Jr., J. Kaye, M. Manczak, Differential loss of synaptic proteins in Alzheimer's disease: Implications for synaptic dysfunction, J. Alzheimers Dis., 7 (2005) 103-117.

[55] A. De Vos, D. Jacobs, H. Struyfs, E. Fransen, K. Andersson, E. Portelius, U. Andreasson, D. De Surgeloose, D. Hernalsteen, K. Sleegers, C. Robberecht, C. Van Broeckhoven, H. Zetterberg, K. Blennow, S. Engelborghs, E. Vanmechelen, C-terminal neurogranin is increased in cerebrospinal fluid but unchanged in plasma in Alzheimer's disease, Alzheimer's & Dementia, (2015).

[56] K. Höglund, S. Kern, A. Zettergren, A. Börjesson-Hansson, H. Zetterberg, I. Skoog, K. Blennow, Preclinical amyloid pathology biomarker positivity: effects on tau pathology and neurodegeneration, Transl. Psychiatry, 2017, pp. e995.

[57] H. Kvartsberg, F.H. Duits, M. Ingelsson, N. Andreasen, A. Öhrfelt, K. Andersson, G. Brinkmalm, L. Lannfelt, L. Minthon, O. Hansson, U. Andreasson, C.E. Teunissen, P. Scheltens, W.M. Van der Flier, H. Zetterberg, E. Portelius, K. Blennow, Cerebrospinal fluid levels of the synaptic protein neurogranin correlates with cognitive decline in prodromal Alzheimer's disease, Alzheimer's & Dementia, (2015).

21

498499500501502503

504505506

507508509

510511512

513514515

516517518519

520521522

523524525526527

528529530

531532533534535

[58] J. Remnestål, D. Just, N. Mitsios, C. Fredolini, J. Mulder, J.M. Schwenk, M. Uhlén, K. Kultima, M. Ingelsson, L. Kilander, L. Lannfelt, P. Svenningsson, B. Nellgård, H. Zetterberg, K. Blennow, P. Nilsson, A. Häggmark-Månberg, CSF profiling of the human brain enriched proteome reveals associations of neuromodulin and neurogranin to Alzheimer's disease, Proteomics Clin. Appl., 10 (2016) 1242-1253.

[59] R. Tarawneh, G. D’Angelo, D. Crimmins, E. Herries, T. Griest, A.M. Fagan, G.J. Zipfel, J.H. Ladenson, J.C. Morris, D.M. Holtzman, Diagnostic and Prognostic Utility of the Synaptic Marker Neurogranin in Alzheimer Disease, JAMA Neurol., 73 (2016) 561-571.

[60] H. Wellington, R.W. Paterson, E. Portelius, U. Törnqvist, N. Magdalinou, N.C. Fox, K. Blennow, J.M. Schott, H. Zetterberg, Increased CSF neurogranin concentration is specific to Alzheimer disease, Neurology, 86 (2016) 829-835.

[61] P.H. Reddy, G. Mani, B.S. Park, J. Jacques, G. Murdoch, W. Whetsell, J. Kaye, M. Manczak, Differential loss of synaptic proteins in Alzheimer's disease: Implications for synaptic dysfunction, Journal of Alzheimers Disease, 7 (2005) 103-117.

[62] H.K. Eltzschig, M.V. Sitkovsky, S.C. Robson, MECHANISMS OF DISEASE Purinergic Signaling during Inflammation, New England Journal of Medicine, 367 (2012) 2322-2333.

[63] K.R. Chalcraft, J. Kong, S. Waserman, M. Jordana, B.E. McCarry, Comprehensive metabolomic analysis of peanut-induced anaphylaxis in a murine model, Metabolomics, 10 (2014) 452-460.

[64] J. Kong, K. Chalcraft, T.S. Mandur, R. Jimenez-Saiz, T.D. Walker, S. Goncharova, M.E. Gordon, L. Naji, K. Flader, M. Larche, D.K. Chu, S. Waserman, B. McCarry, M. Jordana, Comprehensive metabolomics identifies the alarmin uric acid as a critical signal for the induction of peanut allergy, Allergy, 70 (2015) 495-505.

[65] T. Lyngdoh, P. Vuistiner, P. Marques-Vidal, V. Rousson, G. Waeber, P. Vollenweider, M. Bochud, Serum Uric Acid and Adiposity: Deciphering Causality Using a Bidirectional Mendelian Randomization Approach, Plos One, 7 (2012).

[66] F. Martinon, Mechanisms of uric acid crystal-mediated autoinflammation, Immunological Reviews, 233 (2010) 218-232.

[67] J.M. Orengo, A. Leliwa-Sytek, J.E. Evans, B. Evans, D. van de Hoef, M. Nyako, K. Day, A. Rodriguez, Uric Acid Is a Mediator of the Plasmodium falciparum-Induced Inflammatory Response, Plos One, 4 (2009).

[68] M. Kool, T. Soullie, M. van Nimwegen, M.A.M. Willart, F. Muskens, S. Jung, H.C. Hoogsteden, H. Hammad, B.N. Lambrecht, Alum adjuvant boosts adaptive immunity by inducing uric acid and activating inflammatory dendritic cells, Journal of Experimental Medicine, 205 (2008) 869-882.

[69] A. Ohta, M. Sitkovsky, Role of G-protein-coupled adenosine receptors in downregulation of inflammation and protection from tissue damage, Nature, 414 (2001) 916-920.

22

536537538539540

541542543

544545546

547548549

550551

552553554

555556557558

559560561

562563

564565566

567568569570

571572

[70] K. Varani, M. Padovan, F. Vincenzi, M. Targa, F. Trotta, M. Govoni, P.A. Borea, A(2A) and A(3) adenosine receptor expression in rheumatoid arthritis: upregulation, inverse correlation with disease activity score and suppression of inflammatory cytokine and metalloproteinase release, Arthritis Research & Therapy, 13 (2011).

[71] A.P. Simoes, J.A. Duarte, F. Agasse, P.M. Canas, A.R. Tome, P. Agostinho, R.A. Cunha, Blockade of adenosine A(2A) receptors prevents interleukin-1 beta-induced exacerbation of neuronal toxicity through a p38 mitogen-activated protein kinase pathway, Journal of Neuroinflammation, 9 (2012).

[72] L.M. Yerges-Armstrong, S. Ellero-Simatos, A. Georgiades, H. Zhu, J.P. Lewis, R.B. Horenstein, A.L. Beitelshees, A. Dane, T. Reijmers, T. Hankemeier, O. Fiehn, A.R. Shuldiner, R. Kaddurah-Daouk, N. Pharmacometabol Res, Purine Pathway Implicated in Mechanism of Resistance to Aspirin Therapy: Pharmacometabolomics-Informed Pharmacogenomics, Clinical Pharmacology & Therapeutics, 94 (2013) 525-532.

[73] W.B. Dunn, D.I. Broadhurst, S.M. Deepak, M.H. Buch, G. McDowell, I. Spasic, D.I. Ellis, N. Brooks, D.B. Kell, L. Neyses, Serum metabolomics reveals many novel metabolic markers of heart failure, including pseudouridine and 2-oxoglutarate, Metabolomics, 3 (2007) 413-426.

[74] S.M. Euser, A. Hofman, R.G.J. Westendorp, M.M.B. Breteler, Serum uric acid and cognitive function and dementia, Brain, 132 (2009) 377-382.

[75] N. Lu, M. Dubreuil, Y. Zhang, T. Neogi, S.K. Rai, A. Ascherio, M.A. Hernán, H.K. Choi, Gout and the risk of Alzheimer's disease: a population-based, BMI-matched cohort study, Annals of the Rheumatic Diseases, (2015).

[76] J.S. Bester, Prashilla; Kell, Doublas B; Pretorius, E., Viscoelastic and ultrastructural characteristics of whole blood and plasma in Alzheimer-type dementia, and the possible role of bacterial lipopolysaccharides (LPS), Oncotarget, (2015).

[77] C. Drake, H. Boutin, M.S. Jones, A. Denes, B.W. McColl, J.R. Selvarajah, S. Hulme, R.F. Georgiou, R. Hinz, A. Gerhard, A. Vail, C. Prenant, P. Julyan, R. Maroy, G. Brown, A. Smigova, K. Herholz, M. Kassiou, D. Crossman, S. Francis, S.D. Proctor, J.C. Russell, S.J. Hopkins, P.J. Tyrrell, N.J. Rothwell, S.M. Allan, Brain inflammation is induced by co-morbidities and risk factors for stroke, Brain Behavior and Immunity, 25 (2011) 1113-1122.

[78] N.J. Rothwell, G.N. Luheshi, Interleukin I in the brain: biology, pathology and therapeutic target, Trends in Neurosciences, 23 (2000) 618-625.

[79] E.M. Knight, I.V.A. Martins, S. Gumusgoz, S.M. Allan, C.B. Lawrence, High-fat diet-induced memory impairment in triple-transgenic Alzheimer's disease (3xTgAD) mice is independent of changes in amyloid and tau pathology, Neurobiology of Aging, 35 (2014) 1821-1832.

23

573574575576

577578579580

581582583584585

586587588

589590

591592593

594595596

597598599600601

602603

604605606

607

Figure captions

Figure 1. a) Average MALDI-TOF-MSI spectra from a single WT mouse (green), a

single 3×Tg mouse (red) and the average spectrum of all animals (purple). Differentially

expressed m/z species are marked by arrows. b) MALDI-TOF-MSI overview image

visualizing 7.5 kDa Neurogranin (green), PEP-19 (blue) and 14.2 kDa myelin basic

protein (red); scale bar = 2mm. A scheme of a sagittal brain section is also included to

highlight the a: isocortex, b: fibre tracts, c: thalamus, d: hippocampus, and e: midbrain,

pons, medulla and fibre tracts. c) Visualization of m/z 7560 in a MALDI-TOF-MSI

reproducibility study based on the analysis of four animal pairs. d)

Immunohistochemical validation of Neurogranin, scale bar 400 μm.

Figure 2. MALDI-FTICR-MSI visualization of accurate mass (<1 ppm) and isotope

profile filtered (Pearson correlation > 0.95) metabolites involved the purine metabolic

pathway for WT and 3×Tg mice. All tissue sections are sagittal with the cerebellum

located at the top. Scale bar = 2 mm.

Figure 3. MALDI-FTICR-MS spectra from WT mouse (blue) and 3×Tg mouse (red) for

the m/z range 405-412. MSI visualization of N-arachidonoyl taurine (green) in WT and

3×Tg sagittal mouse brain tissue sections at 150 µm resolution. Phosphatidylinositol

(18:0, 20:2) (red) is also shown to outline the structure and orientation of the sections.

The bottom right MSI image shows a higher mass resolution, 50 µm pixel size, MALDI

MSI analysis of the area indicated by the white square.

24

608

609

610

611

612

613

614

615

616

617

618

619

620

621

622

623

624

625

626

627

628

Figure 1.

25

629

630

631

632

633

Figure 2.

26

634

635

636

Figure 3.

27

637

638

639

640

641

642

643

644

645

646

Graphic abstract

28

647

648

649

650

651

652

653

654

Highlights

Protein and metabolite MALDI MSI comparison of AD transgenic mice with

wild type.

Independently validated differences in protein expression in AD transgenic

mice.

Metabolic differences in AD transgenic mice consistent with known AD

biology.

29

655

656

657

658

659

660

661