2 echinococcosis: a h nmr-based metabolomics study · 49 echinococcosis or hydatid disease is a...

1

1 Deciphering the metabolic perturbation in hepatic alveolar

2 echinococcosis: a 1H NMR-based metabolomics study

3 Caigui Lin1, Lingqiang Zhang3,4, Zhiliang Wei1,2, Kian-Kai Cheng5, Guiping Shen1,

4 Jiyang Dong1,, Zhong Chen1, Haining Fan3,4,

5 1. Department of Electronic Science, Fujian Provincial Key Laboratory for Plasma and

6 Magnetic Resonance, Xiamen University, Xiamen, Fujian Province, China;

7 2. Department of Radiology, The Johns Hopkins University, Baltimore, Maryland, USA;

8 3. Department of Hepatopancreatobiliary Surgery, Affiliated Hospital of Qinghai University,

9 Xining, Qinghai Province, China;

10 4., X Qinghai Province Key Laboratory of Hydatid Disease Research ining, Qinghai

11 Province, China;

12 5. Innovation Centre in Agritechnology, Universiti Teknologi Malaysia, Muar, Johor,

13 Malaysia

14

15 Submit to PLoS Pathogens

Corresponding Authors: [email protected] (J. Dong) or [email protected] (H. Fan)

.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted May 14, 2018. . https://doi.org/10.1101/320333doi: bioRxiv preprint

https://doi.org/10.1101/320333

http://creativecommons.org/licenses/by/4.0/

2

16 Abstract

17 Hepatic alveolar echinococcosis (HAE) is a chronic and potentially lethal parasitic

18 disease. It is caused by growth of Echinococcus multilocularis larvae in liver. To date,

19 early-stage diagnosis for the disease is not mature due to its long asymptomatic

20 incubation period. In this study, a proton nuclear magnetic resonance (1H NMR) -based

21 metabolomics approach was applied in conjunction with multivariate statistical analysis

22 to investigate the altered metabolic profiles in blood serum and urine samples from

23 HAE patients and to identify characteristic metabolic markers associated with HAE.

24 The current results identified 21 distinctive metabolic difference between the HAE

25 patients and healthy individuals, which can be associated with perturbations in energy

26 metabolism, amino acid metabolism, oxidative stress, and neurotransmitter imbalance.

27 In addition, the Fischer ratio, which is the molar ratio of branched-chain amino acids to

28 aromatic amino acids was found significantly lower (p<0.001) in blood serum from

29 HAE patients. The ratio, together with changes in other metabolic pathways may

30 provide new insight into mechanistic understanding of HAE pathogenesis, and may be

31 useful for early-stage HAE diagnosis.


https://doi.org/10.1101/320333


3

32 Author Summary

33 Hepatic alveolar echinococcosis (HAE) is a life-threatening disease caused by

34 Echinococcus multilocularis infection. The disease has a long asymptomatic early stage

35 (5~15 years), which complicates effective diagnosis of early-stage HAE even with

36 advanced imaging techniques. Metabolomics is an emerging analytical platform that

37 comprises of analysis of all small molecule metabolites that are present within an

38 organism. The applications of metabolomics method on HAE may help to reveal the

39 molecular biology mechanisms of HAE. In the current study, we had used 1H NMR-

40 based metabolomics technique to investigate blood serum and urine samples from HAE

41 patients. Altered metabolic responses and characteristic differential metabolites for

42 HAE were identified. The metabolic profiling of human biofluids provided valuable

43 information for early-stage HAE diagnosis and for therapeutic interventions, without

44 having to extract HAE vesicles from patients. By featuring global and comprehensive

45 metabolic status, the metabolomics approach holds considerable promise as a

46 noninvasive, dynamic, and effective tool for probing the underlying mechanism of

47 HAE.


https://doi.org/10.1101/320333


4

48 Introduction

49 Echinococcosis or hydatid disease is a near-cosmopolitan zoonotic parasitic

50 infection caused by larval stage of pathogenic cestode parasite in the genus

51 Echinococcus [1]. Different species of Echinococcus cause different diseases. The two

52 main types of diseases are cystic echinococcosis (CE) and alveolar echinococcosis (AE),

53 which are caused by Echinococcus granulosus and Echinococcus multilocularis,

54 respectively [2]. Compared with CE, AE is associated with less occurrence but higher

55 lethality if timely and proper treatments are unavailable [3]. AE has been reported in

56 definitive hosts (canids) as well as in humans [4].

57 In humans, AE affects liver in approximately 95% of the diagnosed cases, which

58 is known as Hepatic Alveolar Echinococcosis (HAE) [5]. HAE leads to liver-tissue

59 injury or even hepatic failure primarily through an infiltrative behavior. It resembles

60 the infiltrative proliferation in tumor growth and thus is clinically known as “worm

61 cancer” or “parasite liver cancer”. After a long period of latent and asymptomatic stage,

62 HAE can progress into cirrhotic stage [6] or metastasize into other organs (e.g., lung

63 and brain [7-9]), and cause local organ-function impairment and metastatic infiltration

64 [10].

65 The HAE disease is endemic in the Northern hemisphere, including North

66 America, Central European and Central Asian countries [11] (e.g. France [12],

67 Germany [13], Austria [14], Poland [15], northern Iran [16], Mongolia [17], and

68 western China [18]). Its global occurrence is mainly attributed to chronic anthropogenic

69 influences, including increased globalisation of animals and animal products, and


https://doi.org/10.1101/320333


5

70 altered human-animal interfaces [4].

71 HAE patients often suffer from symptoms including fatigue, abdominal pain,

72 hepatomegaly, nausea and vomiting. Notably, its long asymptomatic period of the early

73 infection stage leads to difficulty in determining time-point or place of infection [19].

74 In general, HAE diagnosis is based on the integrated considerations of medical history,

75 contact history with suspicious livestock, clinical and pathologic findings, imaging

76 examination, nucleic acid detection and serologic tests [11]. Modern imaging

77 techniques, e.g. ultrasonography (US), magnetic resonance imaging (MRI),

78 fluorodeoxyglucose positron emission tomography (FDG-PET), and computed

79 tomography (CT), have been applied for clinical diagnosis of HAE disease and long-

80 term follow-up after therapeutic interventions [20]. However, these imaging methods

81 have several limitations: CT is unable to delineate disease-induced perihepatic

82 extension; MRI provides good definition of hydatid lesions while fails to assess hydatid

83 fertility (viability); FDG-PET provides noninvasive and accurate evaluation of

84 metabolic activity in HAE at the complexity as well as high expense of administering

85 radiotracer with on-site cyclotron. Besides, effective diagnosis of early-stage HAE

86 remains difficult even with these advanced imaging techniques. This leads to common

87 diagnosis at middle or advanced stage as well as failure to have timely curative

88 resection [21]. Successful early diagnosis and treatment for HAE may prevent

89 complications, reduce postoperative reoccurrence, improve recovery rate, and prove

90 helpful for prognosis.

91 Metabolomics is an “omics” technique at the downstream of genomics,


https://doi.org/10.1101/320333


6

92 transcriptomics and proteomics [22]. Among the postgenomic technologies, high-

93 throughput metabolomics has gained increasing attention as it provides accurate

94 determination and quantification of small molecule metabolites in living systems

95 resulting from physiological stimuli or genetic modification [23], which is of diagnostic

96 and prognostic values. Nuclear magnetic resonance (NMR) spectroscopy is one of the

97 leading analytical approaches in metabolomics. It offers unique qualitative and

98 quantitative insight into all of the more abundant compounds present in biological

99 samples [24]. Generally, proton nuclear magnetic resonance (1H NMR) -based

100 metabolomics is utilized along with multivariate statistical analysis to capture metabolic

101 states of organisms at the global or “-omics” level and determine characteristic global

102 disease-induced metabolic changes.

103 In the last decade, metabolomics approaches have been broadly studied and

104 validated in various liver diseases to determine potential early biomarkers and perturbed

105 metabolic pathways [25]. However, reports on applications of 1H NMR-based

106 metabolomics method on echinococcosis are limited, except: a 1H NMR study to

107 determine metabolites in rodent cyst biomass by Novak et al. [26] and an ex vivo 1H

108 NMR study to explore metabolic profiles with parasite viability in cyst liquid samples

109 from a larger patient collective by Hosch et al. [27]. Therefore, a gap of evaluating

110 metabolic variations in biofluids (blood serum and urine) of HAE patients remains to

111 be filled to facilitate the early diagnosis of HAE patients. The metabolic profiling of

112 human biofluids may provide important diagnostic and prognostic information. Blood

113 serum and urine samples serve as two common and important biofluids to be studied in


https://doi.org/10.1101/320333


7

114 metabolomics due to easy sample collection, pretreatment, restoration, and their ability

115 of giving an overview of global metabolism. In this study, by adopting the high-

116 resolution 1H NMR based metabolomics and multivariate statistical analyses, we aim

117 at: firstly, determining biomarkers to differentiate HAE patients from healthy subjects

118 and establishing metabolic fingerprint of biofluids obtained from HAE patients;

119 secondly, consolidating research foundation for molecular biology mechanism of HAE

120 disease; thirdly, demonstrating the feasibility and effectiveness of using metabolomics

121 in echinococcosis studies; finally, providing valuable diagnostic reference for early-

122 stage HAE.

123 Results

124 1H NMR spectra of biological samples

125 Representative 1H NMR spectra of biofluids collected from the HAE patients and

126 healthy volunteers are shown in Fig 1. In general, several metabolite classes were

127 detected in serum and urine samples, including metabolites involved in energy

128 metabolism, neurotransmission, membrane metabolism, and osmoregulation. In

129 addition, serum metabolome comprised of a few glycoprotein, while a number of gut

130 microbiota-related metabolites were detected in urine samples.

131 Pattern recognition analysis

132 To uncover HAE-induced metabolic changes, we performed multivariate analyses

133 on the processed NMR data. First, the unsupervised principal components analysis

134 (PCA) was used (Fig 2). Each point in the scores plot in Fig 2 represented an individual


https://doi.org/10.1101/320333


8

135 serum or urine sample and inter-point distance reflected the scale of their metabolic

136 differences. Moderate groupings could be observed between the control and the HAE

137 groups in the PCA scores plots. By contrast, a distinctive group separation can be

138 achieved when the data was analysed using supervised orthogonal partial least squares

139 discrimination analysis (OPLS-DA). From Fig 3, control group primarily distributed in

140 the left hemisphere, while HAE group fell into the right hemisphere. Such a distinct

141 separation between the control and HAE groups indicated that HAE induced distinctive

142 metabolic changes in human serum and urine. The OPLS-DA models were found robust

143 with excellent predictive powers following permutation tests (200 permutations) (Fig

144 3).

145 Determination of distinctive metabolites for HAE

146 Next, an enhanced volcano plot was used to identify metabolic markers that

147 differentiate the HAE patients from the healthy controls [28]. In the volcano plot, the

148 importance and significance of the metabolic changes were determined using the

149 following criteria: variable importance projection (VIP) > top 30%, absolute correlation

150 coefficient values ( ) > 0.5, -log10 (p-value) > 2 (i.e., p < 0.01), and absolute log2 (fold |𝑟|

151 change) > 0.25. Generally, identified metabolites with significant changes (those have

152 significant p-value, high fold change, VIP, and ) by multivariate statistical analyses |𝑟|

153 tended to locate at the upper-left or upper-right zones of the enhanced volcano plot in

154 larger circle shapes and warmer colors, as marked with abbreviations in Fig. 4.

155 For blood serum samples, the enhanced volcano plot showed significant increases

156 of phenylalanine (Phe), glutamate (Glu), tyrosine (Tyr), formate (For), and lactate (Lac),


https://doi.org/10.1101/320333


9

157 together with decreases of valine (Val), leucine (Leu), isoleucine (Ile), lysine (Lys),

158 serine (Ser), glutamine (Gln), betaine (Bet), creatine (Cr), and 1-methylhistidine (1-MH)

159 in the HAE group (Fig 4a). Using similar approach, urine samples from the HAE group

160 were characterized by significantly higher levels of glutamate (Glu), malate (Mal),

161 glycolate (Gla), 3-hydroxybutyrate (3-HB), γ-aminobutyrate (GABA), 3-

162 methylhistidine (3-MH), and p-hydroxyphenylacetate (p-HPA), as well as lower

163 concentration of creatinine (Cn) (Fig 4b).

164 Statistical power analysis

165 The statistical power analysis of the selected metabolites with significant changes

166 are presented with Table 1. First, the absolute difference between two independent

167 means (the HAE and control groups) and the intra-group standard deviation were used

168 to calculate the effect size. The effect size was considered as large at a value of 0.80

169 according to the report by Cohen [29]. In addition, post hoc analysis of the achieved

170 statistical power calculation yielded an average value of 0.92 (SD = 0.09) under the

171 given condition of significance level α = 0.05, sample size n = 18, and specified

172 effective sizes. The reliability and effectiveness of the current results were validated

173 based on sufficiently high effect sizes and statistical power.

174

175

176

177

178


https://doi.org/10.1101/320333


10

179

180 Table 1. Post-hoc power analyses of characteristic metabolites with G*power.

Control group HAE groupMetabolites

CONCa CONC p-valueb ESc SPd 1-Methylhistidine 15.52 ± 2.37 12.6 ± 3.16 3.55×10-3 1.04 0.86

3-Hydroxybutyrate 11.51 ± 1.52 13.30 ± 2.21 7.56×10-3 0.95 0.79

3-Methylhistidine 48.17 ± 6.95 69.37 ± 30.36 6.70×10-3 0.96 0.80

Betaine 10.9 ± 2.61 6.68 ± 3.51 2.50×10-4 1.36 0.98

Creatine 52.02 ± 5.06 39.87 ± 9.02 1.78×10-5 1.66 1.00

Creatinine 1438.09 ± 191.86 1128.99 ± 261.67 2.87×10-4 1.35 0.98

Formate 1.31 ± 0.42 2.20 ± 1.03 1.69×10-3 1.14 0.91

Glutamate 11.39 ± 5.49 20.71 ± 5.82 2.05×10-5 1.65 1.00

Glutamine 2.96 ± 1.49 1.77 ± 1.01 1.48×10-3 0.93 0.78

Glycolate 36.33 ± 5.67 49.44 ± 18.25 6.32×10-3 0.97 0.81

Isoleucine 17.14 ± 2.66 13.08 ± 3.01 1.44×10-4 1.43 0.99

Lactate 102.96 ± 43.02 140.90 ± 40.49 1.01×10-2 0.91 0.75

Leucine 15.70 ± 2.39 11.79 ± 2.02 7.05×10-6 1.77 1.00

Lysine 4.33 ± 0.80 2.47 ± 1.50 4.80×10-5 1.55 0.99

Malate 18.06 ± 1.39 20.95 ± 3.10 9.94×10-4 1.20 0.94

Phenylalanine 6.67 ± 2.58 10.02 ± 2.43 1.48×10-5 1.14 0.91

p-Hydroxyphenylacetate 2.57 ± 0.36 3.91 ± 1.44 5.13×10-4 1.28 0.96

Serine 11.19 ± 3.18 7.52 ± 2.97 1.09×10-3 1.19 0.93

Tyrosine 8.11 ± 1.40 9.98 ± 2.21 4.59×10-3 1.01 0.84

Valine 28.35 ± 5.20 19.90 ± 3.30 3.66×10-4 1.94 1.00

γ-Aminobutyrate 29.29 ± 3.31 33.44 ± 4.59 3.80×10-3 1.04 0.68

0.92 ± 0.09

181 a CONC: the relative concentration (percentage of normalized integrals, mean±standard error).

182 b p-value: statistical significance based on Student’s t-test.

183 c ES: the calculated effect size with G*power.

184 d SP: the calculated statistical power calculated as a function of critical significance level (α = 0.05),

185 given sample size (n = 18), and obtained variable effect size in G*power.

186 Metabolic pathway analysis

187 A metabolic pathway analysis of differential metabolites between HAE and

188 control groups was performed using the MetaboAnalyst 3.0 software to investigate the


https://doi.org/10.1101/320333


11

189 most perturbed metabolic pathways in HAE. We found that characteristic metabolites

190 represent perturbations in multiple pathways including alanine, aspartate and glutamate

191 metabolism, arginine and proline metabolism, D-glutamine and D-glutamate

192 metabolism, glycine, serine and threonine metabolism, glyoxylate and dicarboxylate

193 metabolism, lysine metabolism, methane metabolism, phenylalanine metabolism,

194 tyrosine metabolism and valine, leucine and isoleucine metabolism (Fig 5). These

195 pathways were selected based on Impact value ≥ 0.02 and –log(p) ≥ 5 and considered

196 as most relevant pathways in HAE (Table 2).

197 Table 2. Pathway analysis result for metabolomics data with MetaboAnalyst.

Key Pathway name Totala Hitb Impactc -log(p)

1 Alanine, aspartate and glutamate metabolism 24 2 0.38 10.94

2 Aminoacyl-tRNA biosynthesis 75 9 0.17 21.18

3 Arginine and proline metabolism 77 4 0.07 8.17

4 D-glutamine and D-glutamate metabolism 11 2 0.14 10.94

5 Glycine, serine and threonine metabolism 48 3 0.16 12.44

6 Glyoxylate and dicarboxylate metabolism 50 2 0.15 5.08

7 Lysine biosynthesis 32 1 0.10 9.94

8 Lysine degradation 47 1 0.15 9.94

9 Methane metabolism 34 2 0.16 7.15

10 Phenylalanine metabolism 45 3 0.12 11.72

11 Tyrosine metabolism 76 2 0.11 7.54

12 Valine, leucine and isoleucine biosynthesis 27 3 0.04 17.37

13 Valine, leucine and isoleucine degradation 40 3 0.02 17.37

198 a Total: total number of compounds in the pathway.

199 b Hits: number of the significant different metabolites in the pathway.

200 c Impact: pathway impact value calculated from pathway topology analysis (Betweenness).

201 d -log (p): logarithm of the p value calculated by Fisher exact test.

202 Discussion

203 The current study revealed the metabolic perturbation in response to HAE disease


https://doi.org/10.1101/320333


12

204 by using 1H NMR based metabolomics technique. Firstly, 1H NMR spectra and pattern

205 recognition analyses of serum and urine showed substantial changes of human

206 metabolome due to HAE. Secondly, a total number of 22 metabolites were identified

207 with moderate ability in discriminating HAE patients from healthy individuals. These

208 metabolites were determined from the enhanced volcano plot, which is the visualization

209 and integration of multivariate and statistical analysis results. Third, metabolic network

210 analyses revealed statistically-significant HAE-induced modulations to a series of

211 metabolic pathways, i.e., amino acid metabolism, energy metabolism, glyoxylate and

212 dicarboxylate metabolism, and methane metabolism (Fig 5 and 6). These metabolic

213 profiles of the host serum and urine opened important perspectives towards

214 understanding biological changes that occur during HAE infection.

215 Alveoli, host metastasis, and massive immune response due to HAE are known to

216 lead to generation of large amount of unstable reactive oxygen species (ROS) [11, 30].

217 Over-generation of ROS not only affects normal mitochondrial function but also affects

218 cellular homeostasis in case of infection, sterile damage, and metabolic imbalance [31,

219 32]. This will produce a status of oxidative stress with enhanced energy metabolism

220 and increased protein degradation due to cell damage or necrosis.

221 This oxidative-stress response often upregulates tricarboxylic acid (TCA) cycle

222 and glycolysis to produce more ATP for cell maintenance and proliferation. The

223 increased levels of malate and lactate in the HAE group than the control group support

224 the notion of enhanced TCA cycle and glycolysis. The decreased creatine concentration

225 in the HAE group provides further support of perturbed energy metabolism in liver


https://doi.org/10.1101/320333


13

226 mitochondria during HAE infection. Creatine acts as intracellular high-energy

227 phosphate shuttle and plays a crucial role in maintaining cellular energy homeostasis.

228 It is known that a portion of creatine, taken up by cells from the blood through a specific

229 Na＋ and Cl－ dependent transporter. It can then be phosphorylated to form creatine

230 phosphate, which accepts the high energy phosphate group from ATP and subsequently

231 catabolizes into creatinine to provide an immediate adequate ATP levels in case of

232 energy deficit [33]. Thus, decreases of creatine in blood serum may reflect their

233 consumption to alleviate HAE-induced energy deficit. Apart from being an ergogenic

234 aid, creatine has cytoprotective effect as a direct radical scavenger against ROS [34].

235 Therefore, the decline of creatine can partly be attributed to the antioxidant reactions

236 during HAE infections. In view of the cellular energy production and antioxidant

237 activity of creatine in living cells, creatine supplementation has been shown to have

238 beneficial effect in prevention and treatment of a number of muscular, neurological,

239 and cardiovascular diseases, including amyotrophic, lateral sclerosis [35], Parkinson’s

240 disease [36], ischemic brain [37], and congestive heart disease [38]. However, the

241 potential effect of creatine supplementation in the HAE patients remained to be

242 explored.

243 HAE-induced perturbation in amino acids metabolism is indicated by decreased

244 levels of lysine, serine, glutamine, and branch-chain amino acids (BCAA; i.e., valine,

245 leucine, and isoleucine), together with increased levels of glutamate, and aromatic

246 amino acids (AAA; e.g. tyrosine, and phenylalanine) (Figs 4 and 6). Tyrosine, which is

247 the first product of phenylalanine catabolism, accumulates from the hydroxylation of


https://doi.org/10.1101/320333


14

248 phenylalanine with the aid of a rate-determining enzyme (i.e., phenylalanine

249 hydroxylase) in liver and kidney. Previously, it was reported that directionalities of

250 plasma-concentration shift of these two metabolites are the same and conversion of

251 phenylalanine to tyrosine is an exclusive function of the liver [39, 40]. Thus, the parallel

252 accumulations of AAA (i.e. phenylalanine and tyrosine) in blood serum suggested the

253 possible loss of liver function during HAE infection. As essential amino acids, BCAA

254 account for about 20% of our dietary protein and are key regulators of protein syntheses

255 in animals and humans. BCAA disposal is tightly mediated by phosphorylation and

256 dephosphorylation of the branched-chain α-keto acid dehydrogenase complex

257 (BCKDC) [41]. The BCKDC is the rate-limiting enzyme in BCAA oxidative

258 catabolism, the possible activation of BCKDC by HAE increases the BCAA catabolism

259 to promote protein synthesis in case of increased protein degradation due to cell damage

260 or necrosis during HAE infection. Moreover, the role of BCAA in immune function has

261 gained growing research interest in the last decade [42]. BCAA are oxidized by immune

262 cells as fuel sources and then incorporated as precursors for the synthesis of new

263 immune cells, effector molecules, and protective molecules [43]. The host immune

264 response induced by HAE may lead to higher consumption of BCAA to maintain

265 immune function.

266 In addition, HAE was found to down-regulate muscle function during HAE

267 infection. Down-regulated muscle function can be indicated by increased

268 concentrations of GABA and 3-methylhistidine, as GABA has relaxant effects on

269 muscle tone while 3-methylhistidine is an indicator of muscle catabolism [44]. In


https://doi.org/10.1101/320333


15

270 general, 3-methylhistidine is released during muscle-protein degradation and is not

271 reutilized for muscle-protein synthesis. This suggested that immune response to HAE

272 requires not only redirection of abundant energy but also metabolic resources from

273 other tissues, especially from skeletal muscle. Hence, the consumption of BCAA in

274 promoting protein synthesis may focus on the effect of muscle-protein synthesis. On

275 the other hand, ninety-eight percent of body creatine is found in skeletal muscle and a

276 constant fraction of the body creatine pool is converted each day to creatinine. Hence,

277 it has long been recognized that urinary creatinine excretion constitutes a good

278 reflection of skeletal muscle mass [45]. The decline of urine creatinine reflects a

279 corresponding muscle mass loss in the HAE patients. The symptom of fatigue for HAE

280 patients may attribute to muscle mass loss and down-regulated muscle function.

281 As large neutral amino acids, BCAA and AAA travel across blood-brain barrier

282 into the brain through a process competing for a same kind of neutral amino acid

283 transporter LAT-1. [46]. Therefore, the raised AAA (tyrosine, and phenylalanine) and

284 reduced BCAA (valine, leucine, and isoleucine) levels in blood serum of the HAE

285 patients may contribute to a subsequent increased influx of tyrosine and phenylalanine

286 in brain during HAE infection. However, the production of important bioamine

287 neurotransmitters (i.e., dopamine, epinephrine and norepinephrine) is relatively

288 insensitive to cerebral levels of their precursor (i.e., tyrosine and phenylalanine). By

289 contrast, the production of false neurotransmitters (e.g., octopamine or

290 phenylethanolamine) will be easier stimulated by the cerebral elevated levels of

291 tyrosine and phenylalanine, leading to increased amount of false neurotransmitters.


https://doi.org/10.1101/320333


16

292 These false neurotransmitters are structurally similar to “authentic” neurotransmitters,

293 but without ability in propagating neural excitation. This may result in imbalances of

294 neurotransmitter synthesis. Homeostasis between excitatory (e.g., glutamate) and

295 inhibitory (e.g., GABA) neurotransmitters is essential for maintaining the normal

296 functioning of central nervous system (CNS) [47]. In general, glutamate production

297 originates from two pathways: synthesized from α-ketoglutarate (one of intermediates

298 of TCA cycle) by the catalysis of glutamate synthase or glutamine-glutamate

299 conversion, and GABA can be produced from glutamate by the glutamate

300 decarboxylase. In our study, elevations of excitatory glutamate and inhibitory GABA

301 along with glutamine reduction indicated that neurotransmitter recycling disorder

302 occurs in the CNS during HAE infection due to the imbalance between excitation and

303 inhibition. Moreover, alterations of GABA and glutamate in the same direction have

304 been reported to contribute to pathophysiology of depression based on behavioral

305 observations [48, 49].

306 Previously, Fischer et al. had reported decreased BCAA and increased AAA levels

307 in blood during hepatic failures in patients with liver disease. The ratio between the

308 three BCAA (valine, leucine, and isoleucine) and two AAA (tyrosine, and

309 phenylalanine) has been termed as Fischer ratio for a quick assessment of liver disease

310 [50]. With a simultaneous consideration over variations of BCAA and AAA, the amino

311 acid imbalance will lead to a lower Fischer ratio in hepatic failures [51-53]. In this study,

312 a decreased Fischer ratio of serum samples for HAE group (p＜0.001) indicated that

313 HAE may have triggered hepatic failure in patients. Therefore, Fischer ratio may


https://doi.org/10.1101/320333


17

314 emerge as a reliable index for diagnosing HAE. Note that a precise Fischer ratio for

315 HAE patients was not available in our experiments, but this does not affect the

316 effectiveness of Fischer ratio in discriminating HAE patients from healthy individuals.

317 Moreover, the inadequate development of HAE among patients and high altitude

318 subjects live in our study may partly attribute to the difference in Fischer ratio as

319 compare with results of other researchers.

320 Based on these analyses, the imbalanced BCAA and AAA in serum play a crucial

321 pathophysiologic role in HAE. A raise in serum levels of BCAA is anticipated to restore

322 a normal Fischer ratio and reverse the imbalance, aiding the HAE patients in promoting

323 protein synthesis, improving muscle function, protecting against stress, maintaining

324 neurotransmitter equilibrium and strengthening immune response.

325 It should be noted that the present study has several limitations. First, a larger

326 number of subjects will further increase the reliability and accuracy of current results.

327 Then, an integrated application of NMR with LC-MS or GC-MS will extend the

328 metabolite coverage, and multiple -omic techniques (e.g., genomics and proteomics)

329 can cross-validate and better interpret the experimental results. Finally, a distinctive

330 method that can distinguish HAE from other types of hepatic disease is not available

331 with the current study. Such a method will prove helpful in avoiding clinic

332 misdiagnoses since HAE exhibits certain similarities (e.g., lower Fischer ratio) with

333 other hepatic disease. The possibility of using Fischer ratio to categorize different types

334 of hepatic diseases and unique metabolic marker for HAE apart from other hepatic

335 diseases will be explored in our future study, where groups with non-HAE hepatic


https://doi.org/10.1101/320333


18

336 diseases will be included.

337 The attractive properties of NMR based metabolomics approach, i.e., excellent

338 repeatability, short detection duration, and multiple metabolite coverage in a single

339 measurement, facilitate the HAE studies. A systematic exploration over multiple

340 metabolites in small molecules helps to reveal the comprehensive metabolic variations

341 induced by HAE infection. The early-stage HAE diagnose may be possible with a

342 combination of both metabolomics and imaging techniques. For instance, chemical-

343 exchange weighted magnetic resonance imaging techniques can be utilized to yield

344 maps weighted by metabolite of interest [54-56]. Such maps are obtained through an

345 exchange effect with water and thus are characterized by significant signal

346 enhancement to observe small metabolic variations unnoticeable with other imaging

347 modalities. A general issue in such techniques is the lack of specificity, and modulations

348 from adjacent resonances are inevitably and unfavorably introduced. The NMR based

349 metabolomics with high specificity can thus be combined for analyses and better

350 interpret the result by mutual reference.

351 In conclusion, 1H NMR-based metabolomics approach was used to investigate

352 endogenous metabolic characterizations of HAE disease, thus helping to reveal the

353 molecular biology mechanisms of HAE. Particularly, multivariate statistical analyses

354 have highlighted several characteristic metabolic makers for HAE, e.g., decreased

355 BCAA (valine, leucine, and isoleucine) group, increased AAA (tyrosine, and

356 phenylalanine) group, increased lactate, and decreased creatine to list a few. The altered

357 Fischer ratio holds high diagnostic potential as a predictor for evaluating the status or


https://doi.org/10.1101/320333


19

358 future development of HAE. These findings may provide valuable reference

359 information for early-stage HAE diagnoses and for therapeutic interventions. By

360 featuring global and comprehensive metabolic status. The current metabolomics

361 approach holds considerable promise as a noninvasive, dynamic, and effective tool for

362 probing the underling mechanism of HAE.

363

364 Materials and Methods

365 Ethical statement

366 All procedures involving human subjects were approved by the ethical committee

367 of Affiliated Hospital of Qinghai University in Xining, Qinghai, China. Written

368 informed consents were obtained from all subjects involved in the study.

369 Participant recruitment

370 The recruited HAE patients were diagnosed at the Hepatopancreatobiliary Surgery

371 Department in the Affiliated Hospital of Qinghai University from July to September

372 2016. Diagnoses of HAE relied on integrated usage of imaging examination, serologic

373 test, nucleic acid detection and pathologic observation. In addition, patients’ medical

374 history and contact with potential animal host of the disease were also taken into

375 consideration. Healthy volunteers from patients’ families were recruited as the control

376 group. To prevent confounding metabolic effect from other diseases, patients with the

377 following diseases were excluded from the study: diabetes, nephrosis, autoimmunity

378 disease, malignant hepatic tumor, fever, severe hepatorenal dysfunction, and post-


https://doi.org/10.1101/320333


20

379 transplantation immunoreaction.

380 Sample size used in this study was estimated based on the prior power test with

381 some knowledge from our previous preliminary study. The power analyses was done

382 using MetaboAnalyst 3.0 software (http://www.metaboanalyst.ca) [57, 58], and the

383 result indicated that sample size of 18 is adequate to provide sufficiently high statistical

384 power (Fig 7).

385 Sample collections and preparations

386 All blood and urine samples were collected in the morning, 12 hours after the last

387 meal of the previous day (fasting conditions). Blood samples were collected into tubes

388 without anticoagulant and left to clot at room temperature for 20 minutes. Then blood

389 samples were centrifuged at 11000×g at 4 oC for 15 minutes to obtain blood serum.

390 Then, the blood serum and urine samples were aliquoted, snap-frozen in liquid nitrogen

391 and stored at -80 oC until further treatments.

392 Prior to analysis, samples were thawed at room temperature. An aliquot of 400 μL

393 blood serum was added with 200 μL phosphate buffer solution (90 mM

394 K2HPO4/NaH2PO4, pH 7.4, 0.9% NaCl, 99.9% D2O). On the other hand, 300 μL of

395 urine samples were mixed with a different phosphate buffer solution (300 μL, 1.5 M

396 K2HPO4/NaH2PO4, pH 7.4, 99.9% D2O with 0.3 mM 3-trimethylsilyl- propionic-

397 2,2,3,3-d4 acid (TSP) as a chemical-shift reference for 0 ppm). D2O was used to provide

398 the NMR spectrometer with a field frequency for locking. Buffered serum and urine

399 samples were then centrifuged (11000×g, 4 oC, 10 min) to remove debris, and 500 μL

400 supernatant from each 600 μL mixture was transferred into 5-mm NMR tube. In total,


https://doi.org/10.1101/320333


21

401 36 serum and urine samples in NMR tubes were prepared and stored at 4 oC before

402 NMR analysis.

403 1H NMR experiment

404 1H NMR experiments were performed using a Bruker NMR system operating at

405 the proton frequency of 600 MHz. The operating temperature was set at 298K. The

406 Carr-Purcell-Meiboom-Gill (CPMG) sequence (awaiting time ~ π/2 ~ [τ ~ π ~τ]n ~

407 acquisition) was used to acquire spectra of blood serum samples with an echo time (τ)

408 of 250 μs and a free relaxation duration (2nτ) of 100 ms. For urinary samples, Nuclear

409 Overhauser Effect Spectroscopy (NOESY, awaiting time ~ π/2 ~ t1 ~ π/2 ~ tm ~ π/2 ~

410 acquisition) was implemented with a 2 s water suppression and mixing time (tm) of 120

411 ms. Free induction decays were recorded into 32 k data points with 64 averages at a

412 spectral width of 10 kHz prior to Fourier transformation.

413 Data processing of 1H NMR spectra

414 Data pre-processing for the acquired 1H NMR spectra (including Fourier

415 transformation, baseline correction, and phase correction) was performed using the

416 MestReNova v.8.1.2 software (Mestrelab Research S.L.). TSP signal was set as δ 0.00

417 for urine sample, and left split of the doublet of lactate signals was set as δ 1.336 for

418 serum samples. Residual water signals (serum: δ 4.65-5.15, urine: δ 4.75-5.15), urea

419 resonances (δ 5.70-6.40), and peak-free regions were selectively excluded from further

420 analyses. The remaining spectra over ranges of δ 0.8-8.5 for blood serum and δ 0.8-9.5

421 for urine were segmented into bucketed data using self-adaptive integration [59], and


https://doi.org/10.1101/320333


22

422 the results were saved in Microsoft Excel files. For data normalization, the probabilistic

423 quotient normalization (PQN) [23] was utilized to compensate for overall concentration

424 variations. The peaks in the acquired 1H NMR spectra were assigned based on previous

425 published literatures [60, 61], the KEGG database [62], and the HMDB database [63].

426 The relative concentrations for all assigned metabolites were evaluated based on their

427 normalized peak integral area. For metabolites that gave rise to multiple peaks, peaks

428 in the least overlapping spectral region were chosen for quantification.

429 Multivariate and univariate statistical data analyses

430 Multivariate analyses of the pre-processed data were performed using the SIMCA

431 software (version 14.1, Umetrics, Umeå Sweden). The data were examined by the non-

432 supervised principal components analysis (PCA, unit variance scaling) and orthogonal

433 partial least squares-discrimination analysis (OPLS-DA, unit variance scaling). A 7-

434 fold cross-validation and permutation test (200 permutations) were performed and the

435 obtained values of R2 (total explained variation) and Q2 (model predictability) were

436 applied to validate the constructed models [64]. In general, model with Q2 ≥ 0.4 is

437 considered as a reliable model [65]. The variable importance projection (VIP) and

438 absolute correlation coefficient ( ) constructed from the OPLS-DA analysis were used |𝑟|

439 as parameters to select differential metabolites. VIP, which was denoted as a unitless

440 number, delineates the contribution of each predictor variable to the model and presents

441 the influence of each predictor on response variables. A larger VIP corresponds to a

442 greater discriminatory power for the metabolites. Student’s t-test constitutes a simple

443 statistical analyses to determine statistically significant metabolic variations in


https://doi.org/10.1101/320333


23

444 univariate analysis with transformed p-value [66]. For a metabolite, its fold change was

445 calculated based on the ratio of average concentrations between the HAE group and the

446 control group.

447 In this study, results of multivariate statistical analysis were visualized with

448 volcano plot to identify metabolites with significant difference between the HAE and

449 control groups [67, 68]. Particularly, the interactive volcano plot with vertical axis

450 denoting -log10 (p-value) and horizontal axis denoting log2 (fold change) presents circles

451 with different size and color for displaying VIP and values, respectively.|r|

452 Additionally, a post hoc power analysis with specified significance level α, sample

453 size, and effect size was executed with the online statistics software G*power 3.1

454 (http://www.gpower.hhu.de/) [69, 70] to demonstrate the statistical power of reported

455 results.

456 Acknowledgments

457 We thank all the members for helpful discussions.

458 Figure captions

459 Fig 1. Resonance assignments of representative 1H NMR spectra at 600 MHz. (a)

460 serum, (b) urine, HAE (red line) and control (blue line) groups. Spectral regions of

461 5.2-9.5 ppm for serum and urine samples are displayed with different scales for better

462 visualization of weak and narrow spectral peaks. Assignments: 1-MH, 1-

463 methylhistidine; 3-HB, 3-hydroxybutyrate; 3-MH, 3-methylhistidine; Ace, acetate; AD,

464 acetamide; AH, aminohippurate; Ala, alanine; All, allantoin; Alt, allantoate; Bet,

465 betaine; Cho, choline; Ci, citrate; Cn, creatinine; Cr, creatine; DMA, dimethylamine;


https://doi.org/10.1101/320333


24

466 DMG, N,N-dimethylglycine; For, formate; G, glycerol; GA, ; guanidoacetate; GABA,

467 γ-aminobutyrate; Gla, glycolate; Gln, glutamine; Glu, glutamate; Gly, glycine; GPC,

468 glycerophosphocholine; HIB, 2-hydroxyisobutyrate; Hip, hippurate; IB, isobutyrate;

469 Ile, isoleucine; Lac, lactate; Leu, leucine; Lys, lysine; M, malonate; MA, methylamine;

470 Mal, malate; Met, methionine; MG, methylguanidine; m-I, myo-inositol; MM,

471 methylmalonate; Mol, methanol; NAG, N-acetylglutamate; OA, oxaloacetate; OAG,

472 O-acetylglycoprotein; PAG, phenylacetylglycine; PC, phosphocholine; Phe,

473 phenylalanine; p-HPA, para-hydroxyphenylacetate; Ser, serine; Tau, taurine; Thr,

474 threonine; TMAO, trimethylamine N-oxide; Tri, trigonelline; Tyr, tyrosine; Ura, uracil;

475 Val, valine; α-Glc, α-glucose; α-KG, α-ketoglutarate; β-Glc, β-glucose.

476 Fig 2. PCA score plots of control and HAE groups. (a) serum and (b) urine. Each

477 data point represents one subject. Two components accounted for 95.5% of total

478 variances in the serum samples and 61.1% of total variances in the urine samples. The

479 outer ellipse represents the 95% confidence interval (Hotelling’s T2 distribution).

480 Fig 3. OPLS-DA scores plot (left panel) and the corresponding validation plot

481 (right panel). (a) serum and (b) urine samples. The R2 and Q2 values reflecting the

482 fraction of variance and the model predictability, respectively, are given to evaluate the

483 model quality.

484 Fig 4. Enhanced volcano plots showing significantly-changed metabolites. (a)

485 serum and (b) urine samples. Volcano plot shows -log10 (p-value) on y-axis versus

486 log2 (fold change) on x-axes. Each point represents a different metabolite. The circles

487 size and color are determined based on the variable importance projection (VIP) and


https://doi.org/10.1101/320333


25

488 absolute correlation coefficient values ( ), respectively. For each comparison, VIP |𝑟|

489 values are categorized into two categories: top 30% and rest 70 % with each represented

490 by large and small circles, respectively. Warmer color corresponds to higher .|𝑟|

491 Fig 5. Bubble plots of altered metabolic pathways in bio-samples of HAE

492 compared with the control. Bubble area is proportional to the impact of each pathway

493 with color denoting the significance from highest (in red) to lowest (in white). 1, alanine,

494 aspartate and glutamate metabolism; 2, aminoacyl-tRNA biosynthesis; 3, arginine and

495 proline metabolism; 4, D-glutamine and D-glutamate metabolism; 5, glycine, serine

496 and threonine metabolism; 6, glyoxylate and dicarboxylate metabolism; 7, lysine

497 biosynthesis; 8, lysine degradation; 9, methane metabolism; 10, phenylalanine

498 metabolism; 11, tyrosine metabolism; 12, valine, leucine and isoleucine biosynthesis;

499 13, valine, leucine and isoleucine degradation.

500 Fig 6. Mapping of significantly-altered metabolites induced by HAE onto plant

501 biosynthetic pathways and potential cross-talks. Each bar graph represents one

502 metabolite in relative concentration (mean±standard error) of control (blue) and HAE

503 (red) groups. * indicates p < 0.05 statistical significance relative to control group; **

504 indicates p < 0.01 statistical significance relative to control group.

505 Fig 7. Sample size profile with false discovery rate (FDR)=0.005.

506 Reference507 1. Taxy JB, Gibson WE, Kaufman MW. Echinococcosis unexpected occurrence and 508 the diagnostic contribution of routine histopathology. Am J Surg Pathol. 509 2017;41(1):94-100.510 2. McManus DP, Zhang WB, Li J, Bartley PB. Echinococcosis. Lancet. 511 2003;362(9392):1295-1304.


https://doi.org/10.1101/320333


26

512 3. Budke CM, Casulli A, Kern P, Vuitton DA. Cystic and alveolar echinococcosis: 513 successes and continuing challenges. PloS Neglect Trop Dis. 2017;11(4):4.514 4. Davidson RK, Romig T, Jenkins E, Tryland M, Robertson LJ. The impact of 515 globalisation on the distribution of echinococcus multilocularis. Trends in 516 Parasitology. 2012;28(6):239-247.517 5. Farrokh D, Zandi B, Rad MP, Tavakoli M. Hepatic alveolar echinococcosis. Arch 518 Iran Med. 2015;18(3):199-202.519 6. Sekiya C. Parasitic cirrhosis of the liver. Japanese Journal of Clinical Medicine. 520 1994;52(1):234-239.521 7. Tappe D, Weise D, Ziegler U, Muller A, Mullges W, Stich A. Brain and lung 522 metastasis of alveolar echinococcosis in a refugee from a hyperendemic area. J 523 Med Microbiol. 2008;57(11):1420-1423.524 8. Takci E, Sengul G, Akar A, Uslu H, Alper F, Erdogan F, et al. Alveolar echinococcosis 525 of the brain in five patients. J Clin Neurosci. 2008;15(10):1105-1109.526 9. Ohsaki Y, Sasaki T, Shibukawa K, Takahashi T, Osanai S. Radiological findings of 527 alveolar hydatid disease of the lung caused by echinococcus multilocularis. 528 Respirology. 2007;12(3):458-461.529 10. Wang H, Lu C, Liu X, Zhang W. Metastatic and prognostic factors in patients with 530 alveolar echinococcosis. Int J Clin Exp Pathol. 2015;8(9):11192-11198.531 11. Nunnari G, Pinzone MR, Gruttadauria S, Celesia BM, Madeddu G, Malaguarnera 532 G, et al. Hepatic echinococcosis: clinical and therapeutic aspects. World J 533 Gastroenterol. 2012;18(13):1448-1458.534 12. Piarroux M, Piarroux R, Giorgi R, Knapp J, Bardonnet K, Sudre B, et al. Clinical 535 features and evolution of alveolar echinococcosis in France from 1982 to 2007: 536 results of a survey in 387 patients. J Hepatol. 2011;55(5):1025-1033.537 13. Romig T, Kratzer W, Kimmig P, Frosch M, Gaus W, Flegel WA, et al. An 538 epidemiologic survey of human alveolar echinococcosis in southwestern 539 Germany. American Journal of Tropical Medicine and Hygiene. 1999;61(4):566-540 573.541 14. Kreidl P, Allerberger F, Judmaier G, Auer H, Aspock H, Hall AJ. Domestic pets as 542 risk factors for alveolar hydatid disease in Austria. Am J Epidemiol. 543 1998;147(10):978-981.544 15. Nahorski WL, Knap JP, Pawlowski ZS, Krawczyk M, Polanski J, Stefaniak J, et al. 545 Human alveolar echinococcosis in Poland: 1990-2011. PloS Neglect Trop Dis. 546 2013;7(1):8.547 16. Geramizadeh B, Baghernezhad M. Hepatic alveolar hydatid cyst: a brief review of 548 published cases from Iran in the last 20 years. Hepat Mon. 2016;16(10):4.549 17. Ito A, Budke CM. The present situation of echinococcoses in Mongolia. J 550 Helminthol. 2015;89(6):680-688.551 18. Zhang WB, Zhang ZZ, Wu WP, Shi BX, Li J, Zhou XN, et al. Epidemiology and control 552 of echinococcosis in central Asia, with particular reference to the People's 553 Republic of China. Acta Trop. 2015;141:235-243.554 19. Romig T, Dinkel A, Mackenstedt U. The present situation of echinococcosis in 555 Europe. Parasitol Int. 2006;55:S187-S191.


https://doi.org/10.1101/320333


27

556 20. Liu WY, Delabrousse E, Blagosklonov O, Wang J, Zeng HC, Jiang Y, et al. Innovation 557 in hepatic alveolar echinococcosis imaging: best use of old tools, and necessary 558 evaluation of new ones. Parasite. 2014;21:14.559 21. Qu B, Guo L, Sheng G, Yu F, Chen G, Wang YP, et al. Management of advanced 560 hepatic alveolar echinococcosis: report of 42 cases. American Journal of Tropical 561 Medicine and Hygiene. 2017;96(3):680-685.562 22. Lindon JC, Nicholson JK. Spectroscopic and statistical techniques for information 563 recovery in metabonomics and metabolomics. Annu Rev Anal Chem 1. Palo Alto: 564 Annual Reviews; 2008. p. 45-69.565 23. Dieterle F, Ross A, Schlotterbeck G, Senn H. Probabilistic quotient normalization 566 as robust method to account for dilution of complex biological mixtures. 567 Application in 1H NMR metabonomics. Anal Chem. 2006;78(13):4281-4290.568 24. Markley JL, Bruschweiler R, Edison AS, Eghbalnia HR, Powers R, Raftery D, et al. 569 The future of NMR-based metabolomics. Curr Opin Biotechnol. 2017;43:34-40.570 25. Amathieu R, Triba MN, Goossens C, Bouchemal N, Nahon P, Savarin P, et al. 571 Nuclear magnetic resonance based metabolomics and liver diseases: recent 572 advances and future clinical applications. World J Gastroenterol. 2016;22(1):417-573 426.574 26. Novak M, Hameed N, Buist R, Blackburn BJ. Metabolites of alveolar echinococcus 575 as determined by [31P]- and [1H]-nuclear magnetic resonance spectroscopy. 576 Parasitol Res. 1992;78(8):665-670.577 27. Hosch W, Junghanss T, Stojkovic M, Brunetti E, Heye T, Kauffmann GW, et al. 578 Metabolic viability assessment of cystic echinococcosis using high-field 1H MRS 579 of cyst contents. NMR Biomed. 2008;21(7):734-754.580 28. Lin CG, Wei ZL, Cheng KK, Xu JJ, Shen GP, She C, et al. 1H NMR-based investigation 581 of metabolic response to electro-acupuncture stimulation. Sci Rep. 2017;7:13.582 29. Cohen J. A powr primer. Psychol Bull. 1992;112(1):155-159.583 30. Cabrera G, Cabrejos ME, Morassutti AL, Cabezon C, Orellana J, Hellman U, et al. 584 DNA damage, RAD9 and fertility/infertility of echinococcus granulosus hydatid 585 cysts. J Cell Physiol. 2008;216(2):498-506.586 31. Apel K, Hirt H. Reactive oxygen species: metabolism, oxidative stress, and signal 587 transduction. Annu Rev Plant Biol. 2004;55:373-399.588 32. Dunn JD, Alvarez LAJ, Zhang XZ, Soldati T. Reactive oxygen species and 589 mitochondria: a nexus of cellular homeostasis. Redox Biol. 2015;6:472-485.590 33. Snow RJ, Murphy RM. Creatine and the creatine transporter: a review. Mol Cell 591 Biochem. 2001;224(1-2):169-181.592 34. Sestili P, Martinelli C, Bravi G, Piccoli G, Curci R, Battistelli M, et al. Creatine 593 supplementation affords cytoprotection in oxidatively injured cultured 594 mammalian cells via direct antioxidant activity. Free Radic Biol Med. 595 2006;40(5):837-849.596 35. Mazzini L, Balzarini C, Colombo R, Mora G, Pastore I, De Ambrogio R, et al. Effects 597 of creatine supplementation on exercise performance and muscular strength in 598 amyotrophic lateral sclerosis: preliminary results. J Neurol Sci. 2001;191(1-599 2):139-144.


https://doi.org/10.1101/320333


28

600 36. Matthews RT, Ferrante RJ, Klivenyi P, Yang LC, Klein AM, Mueller G, et al. Creatine 601 and cyclocreatine attenuate MPTP neurotoxicity. Exp Neurol. 1999;157(1):142-602 149.603 37. Holtzman D, Khait I, Mulkern R, Allred E, Rand T, Jensen F, et al. In vivo 604 development of brain phosphocreatine in normal and creatine-treated rabbit 605 pups. J Neurochem. 1999;73(6):2477-2484.606 38. Gordon A, Hultman E, Kaijser L, Kristjansson S, Rolf CJ, Nyquist O, et al. Creatine 607 supplementation in chronic heart failure increases skeletal muscle creatine 608 phosphate and muscle performance. Cardiovasc Res. 1995;30(3):413-418.609 39. Dejong CHC, van de Poll MCG, Soeters PB, Jalan R, Damink S. Aromatic amino acid 610 metabolism during liver failure. J Nutr. 2007;137(6):1579S-1585S.611 40. Abu-Omar MM, Loaiza A, Hontzeas N. Reaction mechanisms of mononuclear 612 non-heme iron oxygenases. Chem Rev. 2005;105(6):2227-2252.613 41. Harris RA, Joshi M, Jeoung NH. Mechanisms responsible for regulation of 614 branched-chain amino acid catabolism. Biochem Biophys Res Commun. 615 2004;313(2):391-396.616 42. Zhang SH, Zeng XF, Ren M, Mao XB, Qiao SY. Novel metabolic and physiological 617 functions of branched chain amino acids: a review. J Anim Sci Biotechnol. 618 2017;8:12.619 43. Negro M, Giardina S, Marzani B, Marzatico F. Branched-chain amino acid 620 supplementation does not enhance athletic performance but affects muscle 621 recovery and the immune system. Journal of Sports Medicine and Physical Fitness. 622 2008;48(3):347-351.623 44. Allen PJ, Wise D, Greenway T, Khoo L, Griffin MJ, Jablonsky M. Using 1-D 1H and 624 2-D 1H J-resolved NMR metabolomics to understand the effects of anemia in 625 channel catfish (Ictalurus punctatus). Metabolomics. 2015;11(5):1131-1143.626 45. Davies KM, Heaney RP, Rafferty K. Decline in muscle mass with age in women: a 627 longitudinal study using an indirect measure. Metab-Clin Exp. 2002;51(7):935-628 939.629 46. Fernstrom JD. Large neutral amino acids: dietary effects on brain neurochemistry 630 and function. Amino Acids. 2013;45(3):419-430.631 47. Benitez-Diaz P, Miranda-Contreras L, Pena-Contreras Z, Davila-Vera D, Mendoza-632 Briceno RV, Palacios-Pru E. Histotypic mouse parietal cortex cultures: 633 excitation/inhibition ratio and ultrastructural analysis. J Neurosci Methods. 634 2006;156(1-2):64-70.635 48. Cryan JF, Kaupmann K. Don't worry 'B' happy!: a role for GABAB receptors in 636 anxiety and depression. Trends Pharmacol Sci. 2005;26(1):36-43.637 49. Merali Z, Du LS, Hrdina P, Palkovits M, Faludi G, Poulter MO, et al. Dysregulation 638 in the suicide brain: mRNA expression of corticotropin-releasing hormone 639 receptors and GABAA receptorsubunits in frontal cortical brain region. J Neurosci. 640 2004;24(6):1478-1485.641 50. Fischer JE, Yoshimura N, Aguirre A, James JH, Cummings MG, Abel RM, et al. 642 Plasma amino acids in patients with hepatic encephalopathy. Effects of amino 643 acid infusions. Am J Surg. 1974;127(1):40-47.


https://doi.org/10.1101/320333


29

644 51. Hagenfeldt L, Eriksson LS, Wahren J. Amino acids in liver disease. Proc Nutr Soc. 645 1983;42(3):497-506.646 52. Benjamin IS, Engelbrecht GHC, Saunders SJ, Vanhoornhickman R. Amino acid 647 imbalance following portal diversion in the rat. The relevance of nutrition and of 648 hepatic function. J Hepatol. 1988;7(2):208-214.649 53. Blonde-Cynober F, Aussel C, Cynober L. Abnormalities in branched-chain amino 650 acid metabolism in cirrhosis: influence of hormonal and nutritional factors and 651 directions for future research. Clin Nutr. 1999;18(1):5-13.652 54. Cai KJ, Haris M, Singh A, Kogan F, Greenberg JH, Hariharan H, et al. Magnetic 653 resonance imaging of glutamate. Nat Med. 2012;18(2):302-306.654 55. Zhou JY, Tryggestad E, Wen ZB, Lal B, Zhou TT, Grossman R, et al. Differentiation 655 between glioma and radiation necrosis using molecular magnetic resonance 656 imaging of endogenous proteins and peptides. Nat Med. 2011;17(1):130-U308.657 56. van Zijl PCM, Lam WW, Xu J, Knutsson L, Stanisz GJ. Magnetization Transfer 658 Contrast and Chemical Exchange Saturation Transfer MRI. Features and analysis 659 of the field-dependent saturation spectrum. NeuroImage. 2017.660 57. Xia JG, Psychogios N, Young N, Wishart DS. MetaboAnalyst: a web server for 661 metabolomic data analysis and interpretation. Nucleic Acids Res. 2009;37:W652-662 W660.663 58. Xia JG, Sinelnikov IV, Han B, Wishart DS. MetaboAnalyst 3.0-making 664 metabolomics more meaningful. Nucleic Acids Res. 2015;43(W1):W251-W257.665 59. Anderson PE, Mahle DA, Doom TE, Reo NV, DelRaso NJ, Raymer ML. Dynamic 666 adaptive binning: an improved quantification technique for NMR spectroscopic 667 data. Metabolomics. 2011;7(2):179-190.668 60. Psychogios N, Hau DD, Peng J, Guo AC, Mandal R, Bouatra S, et al. The human 669 serum metabolome. PloS One. 2011;6(2):23.670 61. Bouatra S, Aziat F, Mandal R, Guo AC, Wilson MR, Knox C, et al. The human urine 671 metabolome. PloS One. 2013;8(9).672 62. Ogata H, Goto S, Sato K, Fujibuchi W, Bono H, Kanehisa M. KEGG: kyoto 673 encyclopedia of genes and genomes. Nucleic Acids Res. 1999;27(1):29-34.674 63. Wishart DS, Knox C, Guo AC, Eisner R, Young N, Gautam B, et al. HMDB: a 675 knowledgebase for the human metabolome. Nucleic Acids Res. 2009;37:D603-676 D610.677 64. Triba MN, Le Moyec L, Amathieu R, Goossens C, Bouchemal N, Nahon P, et al. 678 PLS/OPLS models in metabolomics: the impact of permutation of dataset rows 679 on the K-fold cross-validation quality parameters. Mol Biosyst. 2015;11(1):13-19.680 65. McLoughlin GA, Ma D, Tsang TM, Jones DNC, Cilia J, Hill MD, et al. Analyzing the 681 effects of psychotropic drugs on metabolite profiles in rat brain using 1H NMR 682 spectroscopy. J Proteome Res. 2009;8(4):1943-1952.683 66. Cui XQ, Churchill GA. Statistical tests for differential expression in cDNA 684 microarray experiments. Genome Biol. 2003;4(4):210.685 67. De Hertogh B, De Meulder B, Berger F, Pierre M, Bareke E, Gaigneaux A, et al. A 686 benchmark for statistical microarray data analysis that preserves actual biological 687 and technical variance. Bmc Bioinformatics. 2010;11:17.


https://doi.org/10.1101/320333


30

688 68. Hur M, Campbell AA, Almeida-de-Macedo M, Li L, Ransom N, Jose A, et al. A 689 global approach to analysis and interpretation of metabolic data for plant natural 690 product discovery. Nat Prod Rep. 2013;30(4):565-583.691 69. Faul F, Erdfelder E, Buchner A, Lang AG. Statistical power analyses using G*Power 692 3.1: tests for correlation and regression analyses. Behav Res Methods. 693 2009;41(4):1149-1160.694 70. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power 695 analysis program for the social, behavioral, and biomedical sciences. Behav Res 696 Methods. 2007;39(2):175-191.697698

699 Supporting Information Legends

700 File 1. (Serum.xlsx): Binned 1H NMR spectra of serum samples.

701 File 2. (Urine.xlsx): Binned 1H NMR spectra of urine samples.


https://doi.org/10.1101/320333



https://doi.org/10.1101/320333


2 echinococcosis: a h nmr-based metabolomics study · 49 echinococcosis or hydatid disease is a...

Documents