bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

1 Full title: Adaptive iron utilization compensates for the lack of an inducible uptake system in

2 Naegleria fowleri and represents a potential target for therapeutic intervention

3 Short title: Iron metabolism in Naegleria fowleri

4

5 Dominik Arbon1, Kateřina Ženíšková1, Jan Mach1, Maria Grechnikova1, Ronald Malych1, Pavel Talacko2,

6 Robert Sutak1*

7

8 1Department of Parasitology, Faculty of Science BIOCEV, Charles University, Vestec u Prahy, Czech

9 Republic

10 2BIOCEV proteomics core facility, Faculty of Science BIOCEV, Charles University, Vestec u Prahy,

11 Czech Republic

12

13 *Corresponding author

14 E-mail: [email protected] (RS)

1 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

15 Abstract

16 Warm fresh water is a natural habitat for many single-celled organisms, including protozoan parasites

17 such as the infamous brain-eating amoeba, Naegleria fowleri, which can become pathogenic for

18 mammals, including humans. The condition caused by N. fowleri is known as primary amoebic

19 meningoencephalitis, which is a generic and usually fatal infection of the brain with rapid onset. One of

20 the important factors influencing a wide spectrum of pathogens, including N. fowleri, is the bioavailability

21 of iron in the environment. The strategy of withholding iron evolved in mammalian hosts, and the

22 different strategies of pathogens to obtain it are an important part of host-parasite interactions.

23 In the present study, we employ different biochemical and analytical methods to explore the effect of

24 decreased iron availability on the cellular processes of N. fowleri. We show that under iron starvation,

25 nonessential, iron-dependent, mostly cytosolic pathways in N. fowleri are downregulated, while the metal

26 is utilized in the mitochondria to maintain vital respiratory processes. Surprisingly, N. fowleri fails to

27 respond to acute shortages of iron by the induction of a reductive iron uptake system that seems to be the

28 main iron-obtaining strategy of the parasite. Our work aims to demonstrate the importance of

29 mitochondrial iron in the biology of N. fowleri and to explore the plausibility of exploiting it as a potential

30 target for therapeutic interference.

2 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

31 Author Summary

32 Naegleria fowleri is undoubtedly one of the deadliest parasites of humans, hence the name “brain-eating

33 amoeba”. Being dangerous, but rare, it may be regarded as a highly understudied pathogen of humans.

34 Unfortunately, the symptoms of primary amoebic meningoencephalitis may be confused with much more

35 common bacterial meningoencephalitis; therefore, it is quite probable that many infections caused by

36 N. fowleri have been misdiagnosed as bacterial infections without further inquiry. In many cases, fast

37 diagnosis is vital for commencing the correct therapy, and even then, complete success of the treatment is

38 very rare. Our laboratory focuses on the uptake and intracellular metabolism of metals in unicellular

39 eukaryotes, so we decided to explore the biology of N. fowleri from this aspect. Changes in the proteome,

40 as a direct effect of iron-deficient conditions, were described, and these data were used to further explore

41 the ways in which N. fowleri responds to these conditions on a cellular level and how its biology changes.

42 Based on these findings, we propose that the struggle of N. fowleri to obtain iron from its host could be

43 exploited for therapeutic interference purposes in primary amoebic meningoencephalitis patients.

3 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

44 Introduction

45 There are several amoebae that, in certain conditions, are able to parasitize suitable hosts. One particular

46 member of this group, Naegleria fowleri, is widely known as the “brain-eating amoeba”. This rather vivid

47 nickname is attributed to the condition known as primary amoebic meningoencephalitis (PAM), which is

48 caused by N. fowleri. It is found in warm fresh waters across most continents and can alter between three

49 distinguishable forms: a durable cyst, trophozoite amoeba and mobile flagellar form [1]. Its presence is

50 dangerous for people partaking in recreational activities involving water bodies [2]. The disease is

51 acquired when N. fowleri trophozoites forcefully enter through the nasal cavity and intrude the olfactory

52 neuroepithelium [3]. The disease has a rapid onset, and the nonspecific symptoms resemble those of

53 bacterial meningoencephalitis – headache, fever, vomiting – with rapid progress, causing seizures, coma,

54 and death [4,5]. Since PAM occurs commonly in healthy individuals, N. fowleri is not regarded as an

55 opportunistic parasite but as a pathogen [3]. Several other nonpathogenic species of Naegleria exist, and

56 Naegleria gruberi and perhaps Naegleria lovaniensis receive particular interest because these strains are

57 generally accepted as suitable laboratory systems [6].

58 The fatality rate for PAM is reported to be above 97% [7]. Its treatment is difficult not only because of the

59 interchangeable symptoms and rapid onset but also because there is no specific medication for the disease.

60 For successful treatment, it is crucial to make a correct diagnosis in time, followed by immediate therapy.

61 The currently accepted treatment includes administering several medications simultaneously, such as

62 amphotericin B, fluconazole, rifampin, azithromycin or miltefosine, in combination with methods that

63 decrease brain swelling [4,8].

64 Iron is an essential constituent of many biochemical processes, including redox reactions, the

65 detoxification of oxygen, cell respiration or various enzymes. This is mainly due to its ability to alter

66 between different oxidation states, therefore partaking in redox reactions, often in the form of iron-sulfur

67 clusters [9]. Iron is essential for virtually all known forms of life, and its availability is shown to be the

4 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

68 limiting factor of growth in certain locations [10]. Iron acquisition becomes even more challenging for

69 pathogens that inhabit another living organism, as demonstrated in the case of Plasmodium, where iron-

70 deficient human and mouse models manifested unfavorable conditions for parasite development [11,12].

71 The importance of iron acquisition mechanisms is known for other parasites as well [13]. Similarly, it is

72 suggested that iron-deficient conditions have an adverse effect on the viral life cycle [14], and several

73 bacteria are known to produce iron-chelating compounds, siderophores, to take up iron from their

74 surroundings. As a defense mechanism, mammals minimize the presence of free iron in their body using

75 several proteins, such as lactoferrin, ferritin or transferrin, which bind the metal, decreasing its

76 bioavailability [15]. Parasitic organisms are known to have adapted various means of acquiring iron from

77 their environments, ranging from opportunistic xenosiderophore uptake to specific receptors for

78 mammalian iron-containing proteins [13]. This two-sided competition between pathogens and their hosts

79 indicates the importance of iron metabolism in disease and underlines the importance of further research

80 on this topic to search for new methods of therapeutic intervention.

81 This work demonstrates how N. fowleri struggles to successfully adapt to iron-deficient conditions. No

82 increase in iron uptake from its surroundings, either directly or through phagocytosis, was detected, and

83 no alternative metabolic pathway was shown to adjust for the condition. Proteomics and metabolomics

84 approaches showed that the response of N. fowleri to low iron levels is to maintain iron trafficking to

85 mitochondria, while nonvital cytosolic iron-requiring pathways decline. These findings reveal a possible

86 exploitable weakness in the survival strategy of the amoeba within its host. Although the host brain is

87 relatively rich in iron content [16], not all of the iron is readily available for the parasite, as we show in the

88 case of iron bound to transferrin, the main human extracellular iron-binding protein. Thus, disrupting the

89 process of iron uptake by N. fowleri in the human brain, such as using artificial chelators, could represent

90 a safe complementary antiparasitic strategy against this deadly pathogen.

5 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

91 Results

92 Iron chelators have the potential to significantly hinder N. fowleri growth. Iron plays a vital

93 role in many biochemical processes; therefore, it is rational to expect that decreasing the bioavailability of

94 iron in the surrounding environment hinders cell growth. To determine the effect of iron on the cell

95 propagation of N. fowleri, three different iron chelators were added to N. fowleri growth medium:

96 bathophenanthroline disulfonic acid (BPS), 2,2′-dipyridyl (DIP) and deferoxamine (DFO). The

97 compounds inhibited the growth of the cultures to different extents. The IC50 values were calculated for

98 each compound after cultivation for 48 hours and are summarized in Table 1. The most potent effect in

99 hindering the culture growth was observed with the siderophore DFO. Both BPS and DIP exhibited

100 notably higher IC50 values. For all further experiments presented in this work, unless stated otherwise, iron

101 deprivation was achieved using 25 µM of the common extracellular chelator BPS, a condition when cell

102 growth was significantly affected, but microscopy observation showed no encystation or flagellated form,

103 and the cells did not lose their ability to multiply or attach to the surface. This is an important prerequisite,

104 particularly for proteomic and transcriptomic analysis, where too complex changes are accompanied by

105 unfavorable cell processes that could bias the data. To achieve iron-rich conditions, the cultivation

106 medium was supplemented with 25 µM ferric nitrilotriacetate (Fe-NTA).

107 Table 1. Iron chelator IC50 values for N. fowleri cultures after 48 hours.

Compound IC50 (µM)

DIP 30.39 (±3.49)

BPS 17.01 (±2.42)

DFO 6.32 (±0.85)

108 Iron uptake is limited by ferric reductase activity. Iron is generally available in the environment in

109 two distinct oxidation states. Due to the different solubilities of the two forms, many organisms use ferric

110 reductase to reduce ferric to ferrous iron, which is more soluble and therefore easier to acquire. To explore

6 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

111 the mechanism by which N. fowleri acquires iron from its surroundings, the cell cultures were

112 supplemented with different sources of iron: 55Fe-transferrin, 55Fe(III)-citrate and 55Fe(II)-ascorbate. The

113 results presented in Fig 1A show that ferrous iron is taken up and incorporated into intracellular protein

114 complexes more rapidly than its trivalent counterpart, suggesting the involvement of ferric reductase in the

115 effective uptake and utilization of iron. The insignificant uptake of transferrin shown in S1 Fig (A) shows

116 that this protein is not a viable source of iron for N. fowleri, corresponding to the fact that it is not found in

117 the usual habitat of this organism. Surprisingly, there was no observable difference in iron uptake between

118 cells precultivated in iron-rich or iron-deficient media, showing that the organism fails to stimulate its iron

119 uptake machinery.

120 Fig 1. Effect of iron deficiency on N. fowleri.

121 (A) Ferrous and ferric iron uptake by N. fowleri precultivated in iron-rich and

122 iron-deficient conditions. Autoradiography of blue native electrophoresis gels of whole cell extracts

123 from N. fowleri previously cultivated for 72 hours in iron-deficient conditions (25 µM BPS) or iron-rich

124 conditions (25 µM Fe-NTA) and further incubated with 55Fe(II) (ferrous ascorbate) and 55Fe(III) (ferric

125 citrate). Equal protein concentrations were loaded. (B) Growth restoration of N. fowleri culture by

126 hemin in iron-deficient conditions. Cells grown in regular growth medium (control) and with 50 µM

127 hemin (hemin) exhibited similar propagation, while cells in iron-deficient conditions achieved with 50 µM

128 chelator bathophenanthroline disulfonic acid (BPS) were stagnate in growth in the first 24 hours. The

129 addition of 50 µM hemin to the iron-deficient conditions (BPS+hemin) partially restored culture growth.

130 Boxed area signifies time, where growth restoration was calculated from. Data are presented as means ±

131 SD from three independent experiments. (C) Downregulation of N. fowleri hemerythrin in iron-

132 deficient conditions. Western blot analysis of cells cultivated in iron-rich (Fe) and iron-deficient (BPS)

133 conditions using an antibody against Naegleria hemerythrin. The amount of protein loaded was equalized.

134 (D) Cellular content of selected amino acids in N. fowleri cells cultivated in iron-rich and

7 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

135 iron-deficient conditions. Relative amounts of phenylalanine and tyrosine were significantly increased

136 in the iron-deficient condition, while tryptophan and lysine remained unchanged. T-test p-values <0.01 are

137 marked with a star. The total protein concentration was equal in all samples, and values are given relative

138 to the iron-rich conditions for each amino acid. Fe, cells cultivated in iron-rich conditions; BPS, cells

139 cultivated in iron-deficient conditions. Data are presented as means ± SD from three independent

140 experiments. (E) Respiration of N. fowleri grown in iron-rich and iron-deficient conditions.

141 Using selective inhibitors of complex IV and alternative oxidase, the contribution of alternative oxidase

142 and complex IV activity to the total respiration was assessed. AOX, alternative oxidase; Fe, cells

143 cultivated in iron-rich conditions; BPS, cells cultivated in iron-deficient conditions. The rates of

144 respiration are shown relative to the values of cells cultivated in iron-rich conditions. The T-test p-values

145 <0.01 are marked with a star. Data are presented as means ± SD from five independent experiments.

146

147

148 To confirm that N. fowleri employs a reductive iron uptake mechanism, cell cultures were presented with

149 a source of a ferric iron radionuclide in the presence and absence of the ferrous iron chelator BPS. The

150 results shown in S1 Fig (B) demonstrate that N. fowleri readily incorporates ferric iron into its protein

151 complexes, while BPS chelates initially reduced ferrous iron and prevents it from being further utilized,

152 confirming that ferric iron uptake requires the reduction step.

153 Ferric reductase activity is not affected by iron availability. We have shown that neither ferrous

154 nor ferric iron uptake is induced by iron starvation (Fig 1A). To further investigate this observation, the

155 activity of ferric reductase was assessed in cells preincubated in iron-rich and iron-deficient conditions.

156 Measuring ferric reductase activity showed that the conversion of ferric to ferrous iron was not

157 significantly changed in iron-deprived cells. This further confirms that N. fowleri cannot efficiently adjust

158 the rate of iron acquisition from its surroundings in dependence on iron availability.

8 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

159 Hemin partially restores N. fowleri growth in the iron-deficient medium. As shown in Fig 1B,

160 cells cultivated in regular growth medium have a similar propagation pattern in comparison to a culture

161 grown in medium supplemented with 50 µM hemin; therefore, at a given concentration, hemin is not toxic

162 for N. fowleri, nor does it significantly support its growth under standard cultivation conditions. However,

163 while 50 µM BPS arrested the culture growth at 31% (±12%) during the first 24 hours, the addition of 50

164 µM hemin partially restored culture growth to 42% (±10%) in 48 hours and up to 52% (±7%) by 72 hours.

165 The T-test P-values for cultures in BPS and BPS supplemented with hemin are <0.01 for 48 and 72 hours

166 of growth.

167 The proteomic analysis illuminates the iron-starvation response of N. fowleri. Since iron has

168 an irrefutable role as a cofactor for various enzymes and its metabolism is dependent on a great number of

169 proteins, we aimed to examine the overall effect of iron availability on the N. fowleri proteome. Therefore,

170 we compared the whole-cell proteomes of cells grown in iron-rich and iron-deficient conditions, and we

171 have additionally analyzed membrane-enriched fractions of cells to obtain a higher coverage of the

172 membrane proteins. The aim was to reveal the metabolic remodeling accompanying iron starvation and to

173 identify proteins responsible for iron homeostasis, such as membrane transporters, signaling and storage

174 proteins or proteins involved in iron metabolism, for example, the formation of iron-sulfur clusters.

175 S1 Table lists the proteins that were significantly upregulated or downregulated in iron-deficient

176 conditions based on the analysis of the N. fowleri whole-cell proteome, and S2 Table contains the

177 regulated proteins in the membrane-enriched fraction. The proteins most relevant for this work are

178 summarized in Table 2. The list of downregulated proteins based on whole-cell proteomics under iron-

179 deficient conditions contained 20% of predicted iron-containing proteins, most of which were nonheme

180 enzymes such as desaturases and oxygenases, or hydrogenase (NF0008540) and its maturases HydE

181 (NF0081220) and HydG (NF0081230). Importantly, most of the downregulated iron-containing proteins

182 are typically located outside mitochondria.

183 Table 2. Effect of iron deprivation on the abundance of selected N. fowleri proteins.

9 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

Whole cell (membrane fraction) Gene ID Protein proteomics fold-change in iron-deficiency NF0008540 Hydrogenase -3.7 NF0081220 HydE -2.8 NF0081230 HydG -2.4 NF0127030 Hemerythrin -7045.1 NF0119700 Hemerythrin -11.4 NF0117840 Protoglobin Infinite downregulation NF0106930 Phenylalanine hydroxylase -5.7 NF0084900 4-Hydroxyphenylpyruvate dioxygenase -1.5 NF0073220 Homogentisate 1,2-dioxygenase -1.8 NF0001440 Cysteine desulfurase 3.7 NF0060430 IscU * 3.0 NF0001420 Mitochondrial phosphate carrier protein * 8.0 (8.4) NF0079420 Mitoferrin * 1.8 (1.7) NF0071710 Deferrochelatase 2.3 Mitochondrial carnitine/acylcarnitine NF0014630 transferase * Not found (1.5) NF0020800 Fe superoxide dismutase 1.8 NF0044680 Iron III Superoxide Dismutase 1.1 NF0030360 Cu-Zn superoxide dismutase 5.8 NF0121200 Catalase -1.3 NF0015750 Manganese superoxide dismutase 1.6 NF0048730 Glutathione peroxidase 2.0 Iron-containing cholesterol desaturase daf-36- NF0036800 like -2.0 (-3.4) NF0001470 Tyrosine aminotransferase -2.0 NF0004720 Alternative oxidase 1.5 184 Proteins were manually annotated using HHPRED or by sequence alignment with homologous proteins

185 from other organisms (denoted with a star). The fold change based on proteomics analysis with the whole

186 cells and membrane (for selected proteins; in brackets) fractions is given for cells cultivated in iron-

187 deficient conditions. Significantly regulated proteins (with a fold change above 2.3) are highlighted in

188 green.

189

190

10 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

191 The dramatic downregulation of the protein hemerythrin was confirmed by Western blotting using an

192 antibody against N. gruberi hemerythrin. The result shown in Fig 1C shows that the corresponding band

193 was diminished under iron-deficient conditions, which corresponds to the proteomics data. Ponceau S

194 staining to confirm a uniform amount of loaded proteins is shown in S1 Fig (C).

195 In the phenylalanine catabolism pathway, all three iron-dependent enzymes, phenylalanine hydroxylase

196 (NF0106930), p-hydroxyphenylpyruvate dioxygenase (NF0084900) and homogentisate 1,2-dioxygenase

197 (NF0073220), were downregulated in iron-deprived cells, even though the decrease in the expression of

198 the last two enzymes was below our set threshold.

199 The proteins upregulated under iron-deficient conditions included two essential mitochondrial components

200 of iron-sulfur cluster assembly machinery, namely, cysteine desulfurase (NF0001440) and iron-sulfur

201 cluster assembly enzyme IscU (NF0060430); two mitochondrial transporters, phosphate carrier

202 (NF0001420) and iron-transporting mitoferrin (NF0079420); and a potential homologue of

203 deferrochelatase (NF0071710). The increase in the expression of mitoferrin was below our set threshold,

204 but the results from the comparative analysis of the membrane-enriched fraction of N. fowleri grown

205 under different iron conditions (S2 Table and Table 2) confirmed the upregulation of this transporter as it

206 did for the mitochondrial phosphate carrier. Additionally, in the membrane-enriched proteomics analysis,

207 a carnitine/acylcarnitine transferase (NF0014630) was identified and found to be slightly, although not

208 significantly, upregulated.

209 Iron metabolism is interconnected with the production and detoxification of reactive oxygen species.

210 Among the antioxidant defense proteins, one family of superoxide dismutases contains iron as a cofactor.

211 Our proteomic analysis showed that while iron-dependent SODs (NF0020800 and NF0044680) were not

212 significantly changed, Cu-Zn-dependent SOD (NF0030360) was upregulated in iron-deficient conditions,

213 suggesting a compensatory mechanism for the mismetallated iron-dependent enzyme. Other radical

214 oxygen species detoxification enzymes, such as catalase (NF0121200), manganese SOD (NF0015750), or

215 glutathione peroxidase (NF0048730), were not significantly regulated in iron-deficient conditions.

11 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

216 Transcriptomics does not reflect the proteomic changes of N. fowleri. To uncover a broader

217 spectrum of the affected proteins that we were unable to detect in proteomic analysis, we performed

218 transcriptomic analysis of N. fowleri grown under the same conditions of iron availability (S3 Table). To

219 our surprise, the obtained data did not correspond to the proteomics results. The only relevant genes

220 regulated in the same way as the proteins observed by the proteomics analysis were hemerythrin

221 (NF0127030, NF0119700), protoglobin (NF0117840) and iron-containing cholesterol desaturase daf-36-

222 like (NF0036800). This result could indicate that the N. fowleri response to iron starvation is mostly

223 posttranslational.

224 The amino acid assay shows downregulation of the phenylalanine degradation pathway as a

225 result of decreased iron availability. Considering the iron-dependent changes in the abundance of

226 proteins participating in the phenylalanine degradation pathway, the next aim of this work was to analyze

227 this effect by directly detecting metabolites, namely, quantifying the levels of corresponding amino acids

228 in the cells by metabolomics. Cells grown in iron-rich and iron-deficient conditions were lysed, the protein

229 concentrations were equalized, and the resulting materials were analyzed for amino acid content by liquid

230 chromatography coupled with mass spectrometry.

231 As a control, relative amounts of tryptophan and lysine, the levels of which were not expected to change,

232 were determined in the same samples. Relative amounts of both phenylalanine and tyrosine were

233 significantly increased in iron-deficient conditions, as predicted by the proteomic data. The T-test values

234 for tryptophan and lysine were 0.472 and 0.236, respectively; therefore, these values were not significant.

235 However, the values for phenylalanine and tyrosine were 0.007 and less than 0.001, respectively,

236 confirming the significant difference between the iron-deficient and iron-rich conditions. Data are shown

237 in Fig 1D.

238 Intracellular lactate production is decreased in iron-deficient cells. N. fowleri cells possess a

239 lactate dehydrogenase and are therefore potentially able to produce lactate from pyruvate while

12 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

240 replenishing the level of the cofactor NAD+. To analyze the effect of iron starvation on this metabolic

241 pathway, lactate production was analyzed by gas chromatography coupled with mass spectrometry

242 detection. The level of intracellular lactate in iron-deficient cells decreased by 32% (±13%, T-test p-value

243 <0.05).

244 Alternative oxidase activity shifts in iron-deficient conditions to preserve the total

245 respiration rate. Alternative oxidase (AOX), present in N. fowleri, is a part of the mitochondrial

246 respiratory chain. Its function lies in accepting electrons from ubiquinol to reduce the final electron

247 acceptor, oxygen. Therefore, function of AOX is similar to complex IV; however, branching electron flow

248 towards AOX bypasses some of the proton pumping complexes, decreasing the effectivity of the

249 respiration chain. Using selective inhibitors of respiration complex IV and AOX enables the study of their

250 participation in respiration. Here, the effect of decreased iron availability on the respiratory chain was

251 depicted. The results shown in Fig 1E demonstrate that the total respiration of cells grown under iron-

252 deficient conditions was decreased, although not significantly (T-test P-value >0.01), and this decrease

253 was based on the lower activity of complex IV (T-test P-value <0.01). At the same time, in the iron-

254 deficient cells, the activity of AOX significantly increased (T-test P-value <0.01).

255 Phagocytosis reflects N. fowleri viability in iron-rich and iron-deficient conditions.

256 Trophozoites of the amoeba feed on bacteria in their natural environment, presenting a potential source of

257 iron. To investigate the overall cell viability and to test whether N. fowleri cells induce phagocytosis as an

258 effort to stimulate the acquisition of a potential iron source, the ability of iron-starved cells to phagocytose

259 bacteria was investigated. Quantification of phagocytosis was determined by flow cytometry using

260 Escherichia coli that present increased fluorescence within the acidic endocytic compartments. Values

261 were calculated as a percentage of amoeba containing phagocytosed bacteria from the total population. In

262 iron-deficient conditions, the rate of phagocytosis was lower by 25% (±12%, T-test p-value <0.001) in

263 comparison to iron-rich conditions, showing that iron availability plays a role in this process. A typical

264 cell phagocytizing fluorescent bacteria is shown in S2 Fig.

13 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

265 Discussion

266 Due to the irreplaceable role of iron in cellular processes, it is unsurprising that iron chelators have the

267 ability to hinder the proliferation of N. fowleri [17]. The different chelators used in this work have distinct

268 properties that influence their impact on cells. DIP is a membrane-permeable compound with affinity to

269 ferrous iron [18] and a ratio of three chelator molecules binding one molecule of metal [19]. BPS is a

270 membrane-impermeable chelator that binds ferrous iron with a ratio of three chelator molecules to one

271 metal ion [20]. Finally, DFO is a membrane-impermeable, ferric iron-binding siderophore [21] with a

272 binding ratio of one iron per molecule [22]. Quite surprisingly, the only tested membrane-permeable iron

273 chelator, DIP, had the highest IC50 value (30.39 µM; SD = 3.49). In comparison with BPS, with which it

274 shares an affinity to ferrous iron and the same denticity, DIP was almost half as effective in inhibiting

275 culture growth. BPS and DFO differed not only by the oxidation state of bound iron but also by the ratio

276 of molecules bound to the chelated iron. Their IC50 values were 17.01 µM (SD = 2.42) and 6.32 µM (SD =

277 0.85), respectively, showing an apparent tendency of the approximately three-fold amount of BPS

278 required to have the same effect as DFO, probably because of the binding ratio. Therefore, surprisingly, it

279 appears that from the selected chelators, the membrane-impermeable chelators are more effective against

280 N. fowleri. It is important to consider that the growth medium used in this work was adjusted to the

281 amoeba requirements, while in its host, the parasite is expected to meet much harsher conditions. Thus,

282 the anticipated in vivo antiparasitic effect of suitable chelators may be more pronounced.

283 To maintain an optimal level of cellular iron in a hostile environment, such as host tissues, pathogens

284 possess selective and effective mechanisms for iron uptake. These strategies of obtaining iron include the

285 utilization of various sources from the host, including transferrin, lactoferrin or heme, and some parasites

286 even exploit bacterial siderophores as sources of iron [13]. Transferrin, an abundant human blood protein

287 that transports iron to various tissues, including the brain [23], serves as a viable source of iron for

288 different parasites [13]. It was demonstrated that N. fowleri possesses proteases able to degrade human

289 holotransferrin, although the work did not identify the intracellular fate of the iron [24]. Our work shows

14 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

290 that N. fowleri does not have the means of efficiently utilizing iron from this host protein, perhaps because

291 it is a facultative pathogen with no advantage of such an iron uptake mechanism in its natural

292 environment. We further demonstrated the preference of N. fowleri for ferrous iron compared to ferric

293 iron, the inhibitory effect of ferrous iron chelator on ferric iron uptake and the presence of extracellular

294 ferric reductase activity. Based on these observations, we argue that the main strategy of iron acquisition

295 by this parasite is the reductive two-step iron uptake mechanism, as described for Saccharomyces

296 cerevisiae [13].

297 S. cerevisiae possesses the ability to upregulate the expression of ferric reductase, which is responsible for

298 the first step of the reductive iron uptake mechanism, up to 55 times in iron-deficient conditions [25],

299 therefore increasing the rate of iron uptake. Our work shows that in N. fowleri, the activity of ferric

300 reductase is not induced by iron starvation nor is the ferric and ferrous iron uptake and further

301 incorporation of iron into cellular proteins. Phagocytosis of bacteria as a proposed way to obtain iron is

302 not regulated by iron availability in the parasite; on the contrary, it is decreased. Considering this, it is

303 important to note that the relationship between decreased phagocytosis and decreased hemerythrin

304 expression was previously described [26].

305 The lack of an inducible iron uptake system has also been described in the parasite Tritrichomonas foetus

306 [27]. However, contrary to N. fowleri, the obligatory parasite T. foetus appears to be able to utilize a wide

307 range of iron sources, including host transferrin or bacterial siderophores. Another potential source of iron

308 for N. fowleri can be heme from heme-containing proteins. The ability to utilize exogenous

309 metal-containing porphyrins may be advantageous for protists feeding on bacteria. Moreover, Naegleria

310 species phagocyte erythrocytes [28,29], and their ability to degrade hemoglobin using proteases was also

311 described [24]. However, it appears that heme oxygenase is not present in the genome of N. fowleri;

312 therefore, it is unlikely to be able to employ this enzyme to obtain iron from hemoglobin, as is the case of

313 T. foetus [27]. A potential homologue of bacterial deferrochelatase, a protein able to directly acquire iron

314 from heme [30], was identified as significantly upregulated in iron-deficient conditions by proteomic

15 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

315 analysis; however, further investigation is required to clarify the function of this protein. Nevertheless,

316 hemin was shown to partially restore N. fowleri growth in very strong iron-deficient conditions. N. fowleri

317 requires exogenous porphyrins for growth [31], and the fact that the addition of hemin partly surpasses the

318 conditions of iron starvation can be attributed to a metabolic rearrangement towards heme-dependent

319 pathways.

320 Proteomic analysis proved to be a valuable resource in determining the cellular changes brought upon N.

321 fowleri by iron starvation and gave a foundation for further discoveries shown in this work. The most

322 fundamental finding of this experiment is that mainly cytosolic iron-containing proteins were

323 downregulated. These are components of probably nonessential pathways, such as the phenylalanine

324 degradation pathway, hydrogenase maturation and hydrogen production; the latter is surprisingly shown to

325 take place in the cytosol of N. gruberi [32]. On the other hand, mitochondrial iron-dependent proteins

326 were generally unchanged, while some components of the iron-sulfur cluster synthesis machinery were

327 upregulated in iron-deficient conditions, emphasizing the essential role of the iron-dependent respiration

328 process. Consistent with this observation, the mitochondrial iron transporter was slightly upregulated in

329 iron-deficient conditions. Moreover, the carnitine/acylcarnitine carrier was identified in the membrane-

330 enriched proteomic analysis, and its expression was slightly increased in iron-starved cells. This

331 mitochondrial membrane-bound protein is involved in lipid metabolism, which was recently shown to be

332 vital for N. gruberi [33]. Another mitochondrial transport protein, the phosphate carrier, was strongly

333 upregulated in iron-deficient conditions, probably as an effort to compensate for impaired respiration and

334 decreased ATP production in the mitochondria of iron-deficient cells.

335 N. fowleri, as well as N. gruberi, possesses an AOX in the mitochondria, and our work indicates one of the

336 possible advantages of this respiratory chain element for these organisms. Under iron-deficient conditions,

337 the activity of AOX was significantly increased, even though it required iron, suggesting that this unusual

338 branch of the respiratory chain takes over a portion of the activity in the iron-demanding conventional

339 pathway of respiratory complexes III, IV and cytochrome c, an observation noted in the nonpathogenic

16 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

340 amoeba N. gruberi previously [34]. This represents a favorable compensation pathway, even though the

341 overall generation of the proton gradient and therefore ATP synthesis is hindered. Considering the reduced

342 efficiency of respiration by iron-starved cells and the presence of lactate dehydrogenase in the genome of

343 N. fowleri, it would be reasonable to expect the employment of the lactate dehydrogenase pathway in the

344 regeneration of the cofactor NAD+. Such an effect was observed in Trichomonas vaginalis, where the cells

345 modulate this pathway as a way of compensating for the metronidazole-induced loss of hydrogenosomal

346 metabolism [35]. However, our metabolomic analysis showed the opposite change; the production of

347 lactate was decreased upon iron starvation, suggesting another, most likely nonessential, function of this

348 pathway that is attenuated due to the hindered rate of overall energy metabolism. Another possible

349 compensatory pathway, ethanol production, is improbable because of the absence of pyruvate

350 decarboxylase or bifunctional aldehyde/alcohol dehydrogenase in the N. fowleri genome. The function of

351 cytosolic hydrogenase, which was downregulated in the iron-limited conditions together with its

352 maturation factors, is unknown.

353 Consistent with our previous study of iron metabolism in N. gruberi [34], hemerythrin was dramatically

354 downregulated in iron-deficient conditions, while its expression appeared to be high, based on the

355 intensities obtained in proteomic data from iron-rich cells. The involvement of hemerythrin in the iron

356 metabolism of Naegleria is suggestive but unclear. The presence of unbound metals in the cell must be

357 strictly regulated since the imbalance in iron homeostasis can lead to the formation of ROS,

358 mismetallation or other anomalies leading to the incorrect function of proteins. The relationship between

359 hemerythrin and defense against oxidative stress was previously suggested in bacteria [36,37] and so was

360 the role of hemerythrin-related proteins in iron homeostasis [38] or oxygen sensing [39]. One of the basic

361 mechanisms of maintaining the proper intracellular level of metals is the regulation of their acquisition.

362 Since our study shows the lack of such regulation for iron, it is possible that the sequestration of toxic free

363 iron, as well as its storage for use in iron-deficient conditions, is ensured by hemerythrin functioning as a

364 cytosolic iron pool. This hypothesis is supported by the fact that hemerythrin is a nonheme, noniron-sulfur

17 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

365 metalloenzyme and is among the most strongly regulated proteins by iron availability, and unlike other

366 proteins, its regulation was detected even at the mRNA level. Another oxygen-binding protein with

367 unclear function, protoglobin, was detected only in iron-sufficient conditions. The role of protoglobin in

368 the metabolism of N. fowleri remains to be elucidated.

369 In conclusion, our work shows that N. fowleri possesses only limited capabilities of adaptation to an iron-

370 deficient environment and is surprisingly not able to utilize transferrin as an alternative source of the

371 metal; neither can it induce the rate of iron acquisition under iron starvation, reflecting the lifestyle of a

372 facultative parasite with limited ability of survival in a host. The main strategy of acquiring iron from

373 media appears to be reductive iron uptake. Proteomic analysis of the response to iron starvation

374 demonstrated that a large amount of downregulated proteins in the iron-deficient conditions were

375 nonmitochondrial and nonheme enzymes, except for the heme-containing protein protoglobin, whose

376 expression is regulated at the transcriptional level, unlike the expression of most other affected proteins.

377 Therefore, it can be hypothesized that the fundamental effect of iron deprivation is the degradation of

378 mismetallated cytosolic proteins with a simultaneous increase in iron delivery to mitochondria and

379 induction of iron-sulfur cluster synthesis machinery to ensure essential cell processes. The overall changes

380 in cellular processes in the iron-deficient conditions discussed in this paper are illustrated in Fig 2. These

381 findings are in agreement with our previous study focused on iron metabolism in the nonpathogenic model

382 organism N. gruberi [34], where the mitochondrion was shown to be the center of the iron economy.

383 Together, these results show that iron deficiency is a highly unfavorable condition for N. fowleri, and

384 targeted interference with its uptake could be an effective method of controlling the propagation or

385 viability of this organism in the host.

386 Fig 2. Illustration of the main effects of iron-deficient conditions on the selected cellular

387 processes of N. fowleri.

388 The results of proteomic analysis for selected proteins are depicted in red, the results from measured

389 metabolite levels are in green and the assessed enzyme activities are in purple. Upwards pointing arrows

18 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

390 stand for increased in the iron-deficient condition, downwards pointing arrows represent decreased in the

391 iron-deficient conditions and dashes mean no significant change in the different iron conditions. Fe

392 represents iron-containing/involving protein/process. C, cytochrome C; MTF, mitoferrin; MPCP,

393 mitochondrial phosphate carrier protein; P, phosphate; Q, ubiquinol/ubiquinone.

19 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

394 Materials and Methods

395 Organisms

396 Naegleria fowleri, strain HB-1, kindly provided by Dr. Hana Pecková (Institute of Parasitology, Biology

397 Center CAS) was maintained in 2% Bacto-Casitone (Difco, USA) supplemented with 10% heat-

398 inactivated fetal bovine serum (Thermo Fisher Scientific, USA), penicillin (100 U/ml) and streptomycin

399 (100 µg/ml) in 25 cm2 aerobic cultivation flasks at 37°C. When required, the cells were cultivated for 72

400 hours with the addition of 25 µM BPS (Sigma-Aldrich, USA), simulating the iron-deficient environment,

401 or with 25 µM Fe-NTA (Sigma-Aldrich, USA) to ensure iron-rich conditions.

402 Chelators

403 The growth dependence of N. fowleri on iron availability was assessed using the iron chelators BPS

404 (Sigma-Aldrich, USA), DIP (Sigma-Aldrich, USA) and DFO (Sigma-Aldrich, USA). N. fowleri cells were

405 cultivated on 24-well plates in a total volume of 1 ml in a humid chamber. Each chelator and control were

406 tested in biological tetraplicates (final concentrations of 100, 50, 25, 12.5, 6.3, 3.1, 1.6 and 0.8 µM). Cells

407 were cultivated for 48 hours, the plates were placed on ice for 10 min, and the medium was gently pipetted

408 to detach the cells. The number of cells in each sample was counted using a Guava easyCyte 8HT flow

409 cytometer (Merck, Germany). To describe and compare the effect of each chelator, the value of half-

410 maximal inhibitory concentration (IC50) was calculated using the online calculator on the AAT Bioquest

411 webpage [40].

412 Iron uptake

413 N. fowleri cells grown for 72 hours in iron-rich and iron-deficient conditions were washed with phosphate

414 buffered saline (PBS) (1000 g for 15 min) and transferred to measuring buffer (50 mM glucose; 20 mM

415 HEPES; pH 7.2). Cells were counted using a Guava easyCyte 8HT flow cytometer, and 2.5×105 cells were

416 equally split onto a 24-well plate. To assess iron uptake, cells were supplemented with 2 µM 55Fe-citrate,

20 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

417 2 µM 55Fe-citrate with 1 mM ascorbate, or 6.3 µM 55Fe-transferrin. Samples were incubated at 37°C for 1

418 hour, and then EDTA was added to a final concentration of 1 mM to chelate extracellular iron. Cells were

419 washed three times by measuring buffer, and the protein concentration was assessed using a BCA kit

420 (Sigma-Aldrich, USA). Samples were diluted to equal concentrations and separated using the Novex

421 Native PAGE Bis–Tris Gel system (4–16%; Invitrogen, USA). The gel was vacuum-dried for 2 hours and

422 autoradiographed using a tritium storage phosphor screen.

423 Ferric reductase activity

424 To assess the activity of N. fowleri ferric reductase in different iron environments, a ferrozine assay was

425 used to compare the formation of ferrous iron, as described previously [41]. Cells were grown in iron-rich

426 and iron-deficient environments for 72 hours. Samples containing no cells and samples without Fe-EDTA

427 were used as controls. The cells were washed twice and resuspended in glucose buffer (50 mM glucose,

428 0.5 mM MgCl2, 0.3 mM CaCl2, 5.1 µM KH2PO4, 3 µM Na2HPO4, pH 7.4). The total amount of proteins

429 was assessed using a BCA kit, and samples were diluted to equal concentrations. All further work was

430 performed with minimum light exposure. Ferrozine was added to a total concentration of 1.3 mM and Fe-

431 EDTA to a concentration of 0.5 mM. Samples were incubated at 37°C for three hours, pelleted (1000 g for

432 10 min) and the supernatant was used to determine the formation of the colored Fe(II)-ferrozine complex,

433 accompanied by a change in absorbance at 562 nm using 1 cm cuvettes and a UV-2600 UV-VIS

434 spectrophotometer (Shimadzu, Japan).

435 Hemin utilization

436 To assess the ability of N. fowleri to utilize hemin, 2×104 cells were cultivated in each well of a 24-well

437 plate in fresh growth medium. Cultures were supplemented with 50 µM BPS, 50 µM hemin, or 50 µM

438 hemin and 50 µM BPS. Cells without chelator or hemin were used as a control. The cell concentration was

439 measured every 24 hours for four days using a Guava easyCyte 8HT flow cytometer as described above.

440 Comparative proteomic analysis

21 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

441 N. fowleri cells were cultivated in 75 cm2 aerobic cultivation flasks in iron-rich and iron-deficient

442 environments. For whole-cell proteomic analysis, cells were washed three times in PBS (1000 g, 15 min,

443 4°C) and pelleted.

444 In addition, a membrane-enriched fraction was prepared. Approximately 1.5×107 cells were harvested

445 (1000 g, 15 min, 4°C), washed in 15 ml of sucrose-MOPS buffer (250 mM sucrose, 20 mM

446 3-morpholinopropanesulfonic acid, pH 7.4, supplemented with complete EDTA-free protease inhibitors,

447 Roche, Switzerland) and resuspended in 4 ml of sucrose-MOPS buffer. The suspension was sonicated

448 using a Q125 sonicator (Qsonica, USA) (45% amplitude, total time of 4 min using 1 s pulse and 1 s

449 pause). The resulting suspension was centrifuged (5000 g, 5 min) to spin down the unlysed cells. The

450 obtained supernatant was further centrifuged (200 000 g, 20 min). The supernatant was discarded, and the

451 membrane-enriched pellet was washed with distilled water.

452 Label-free proteomic analysis of the samples was assessed using the method described in Mach et al. 2018

453 [34], utilizing liquid chromatography coupled with mass spectrometry. The resulting MS/MS spectra were

454 compared with the Naegleria fowleri database, downloaded from amoebaDB [42] on 25/7/2017. The set

455 thresholds to filter proteins were Q-value =0, unique peptides detected >2, and the protein had to be

456 identified at least twice in one condition from the six runs. For proteins not identified only in one of the

457 conditions, intensity of 23 was selected as a lowest value to be included, on basis of previous imputation

458 experience. To distinguish significantly downregulated or upregulated proteins in the iron-deficient

459 conditions in S1 and S2 Tables, the threshold of fold change >2.3 and <-2.3 was chosen.

460 The annotations and identifications of the selected proteins discussed in this work were confirmed using

461 the tool HHPRED [43]. In addition, alignments were constructed using the chosen proteins and their

462 homologues from different organisms: NF0014630 identified as mitochondrial carnitine/acylcarnitine

463 transferase [44], NF0001420 identified as mitochondrial phosphate carrier [45], NF0060430 identified as

464 IscU [46,47] and NF0079420 identified as mitoferrin [48,49]. Alignments with the indicated conserved

22 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

465 sequences are shown in S3 Fig and were constructed using Geneious version 11.1.5 software with the

466 muscle alignment tool.

467 Comparative transcriptomic analysis

468 To obtain the transcriptome data of N. fowleri, biological pentaplicates of approximately 1×106 cells each

469 were grown in iron-rich and iron-deficient conditions for 72 hours. A High Pure RNA Isolation Kit

470 (Roche, Switzerland) was used to isolate cell RNA, and an Illumina-compatible library was prepared

471 using QuantSeq 3′ mRNA-Seq Library Prep Kit FWD for Illumina (Lexogen, Austria). The RNA

472 concentration was determined using a Quantus fluorometer (Promega, USA), and the quality of RNA was

473 measured on a 2100 Bioanalyzer Instrument (Agilent technologies, USA). Equimolar samples were

474 pooled to 10 pM and sequenced with MiSeq Reagent Kit v3 (Illumina, USA) using 150-cycles on the

475 MiSeq platform. The obtained results were filtered using the P-value of >0.05 followed by analysis on the

476 BlueBee platform with the method DESeq [50].

477 Amino acid quantification assay

478 Approximately 3×106 cells cultivated in iron-rich and iron-deficient environments were harvested by

479 centrifugation (1000 g, 15 min, 4°C), washed with PBS supplemented with cOmplete EDTA-free protease

480 inhibitor, and the total concentration of proteins was measured. Cells were transferred to 1 ml of buffer

481 solution (20 mM Tris, 1 mM MgCl2, pH 8, cOmplete EDTA-free protease inhibitor) and sonicated with

482 Sonopuls mini20 (Bandelin, Germany) (90% amplitude, 4°C, total time 120 s, 1 s pulse and 1 s pause).

483 The resulting suspension was mixed at a ratio of 1:4 with ice-cold acetonitrile and maintained overnight at

484 -20°C. Samples were centrifuged (16000 g, 20 min, 4°C) and filtered using Ultrafree Centrifugal Filter

485 Units (Merck Millipore, USA).

486 Samples were analyzed using liquid chromatography on a Dionex Ultimate 3000 HPLC system with on-

487 line mass spectrometry detection (Thermo Scientific, USA). The separation was achieved using a HILIC

488 column iHILIC-Fusion (150 x 2.1 mm, 1.8 μm particles, 100 Å pore size, HILICON, Sweden). The entire

23 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

489 analysis flow rate was 0.3 ml/min, and the column was equilibrated with 100% of solution A (80%

490 acetonitrile in water, 25 mM ammonium formate, pH 4.8) for 3 min. Amino acids were eluted by

491 increasing the gradient of solution B (5% acetonitrile in water, 25 mM ammonium formate, pH 4.8),

492 where 50% of solution B was reached in 7 min. After separation, the column was washed with 80%

493 solution B for 3 min and then equilibrated with 100% solution A for 5 min.

494 Amino acids were detected by mass spectrometry using the triple quadrupole instrument TSQ Quantiva

495 (Thermo Scientific, USA) in Selected Reaction Monitoring mode. Analytes were ionized using

496 electrospray ionization on an H-ESI ion source and analyzed with positive charge mode with a spray

497 voltage of 3500 V, ion transfer tube temperature of 325°C, and vaporizer temperature of 350°C. All

498 transitions, collision energies, and RF voltages were optimized prior to analysis using appropriate amino

499 acid standards. Each analyte was detected using at least two transitions. Cycle time was set to 1.8 s and

500 both Q1 and Q3 resolutions were set to 0.7 s. To analyze the ion chromatograms and calculate peak areas,

501 Skyline daily version 4.2.1.19004 [51] was used.

502 Lactate production

503 To assess the difference in the intracellular production of lactate, 3×106 cells cultivated in iron-rich or

504 iron-deficient conditions were prepared for analysis in the same way as for quantifying the amino acid

505 content. After incubation with acetonitrile and filtration, the cell sample was dried and resuspended in 100

506 µl of anhydrous pyridine (Sigma-Aldrich, USA), and 25 µl of a silylation agent (N-tert-butyldimethylsilyl-

507 N-methyl-trifluoroacetamide, Sigma-Aldrich, USA) was added. The sample was incubated at 70°C for 30

508 min. After incubation, 300 µl of hexane (Sigma-Aldrich, USA) and 10 µl of an internal standard (102

509 µg/ml 1-bromononane solution in hexane) were added. Selected compounds were analyzed as tert-butyl

510 silyl derivatives.

511 Samples were analyzed using two-dimensional gas chromatography coupled with mass detection

512 (GCxGC-MS; Pegasus 4D, Leco Corporation, USA) with ChromaTOF 4.5 software. Mass detection was

24 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

513 equipped with an EI ion source and TOF analyzer with unite resolution. A combination of Rxi-5Sil (30 m

514 x 0.25 mm, Restek, Australia) and BPX-50 (0.96 m x 0.1 mm, SGE, Australia) columns were used. The

515 input temperature was set to 300°C, the injection volume was 1 µl in spitless mode, and constant helium

516 flow of 1 ml/min, modulation time 3 s (hot pulse 1 s) and modulation temperature offset with respect to

517 the secondary oven 15°C were used. The temperature program applied on the primary oven was 50°C

518 (hold 1 min), which was increased by the rate of 10°C/min to a final temperature of 320°C (hold 3 min).

519 The temperature offset applied on the secondary column was +5°C.

520 Cell respiration

521 Five biological replicates of 3×106 N. fowleri cells grown for 72 hours in iron-rich and iron-deficient

522 conditions were washed twice and resuspended in 1 ml of glucose buffer, and the protein concentration

523 was assessed using a BCA kit. Total cell respiration was measured as the decrease in oxygen

524 concentration using an Oxygen meter model 782 (Strathkelvin instruments, UK) with Mitocell Mt 200

525 cuvette of total volume of 700 µl at 37°C. Measurements were carried out with 5×105 cells for 5 min, after

526 which potassium cyanide was added to a final concentration of 4 mM to block complex IV, and after 5

527 min, salicyl hydroxamic acid was added to a final concentration of 0.2 mM to completely block AOX.

528 Values gained after the addition of potassium cyanide and salicyl hydroxamic acid were subtracted to

529 acquire canonical respiratory chain and AOX activity, respectively.

530 N. fowleri bacterial phagocytosis

531 Approximately 3×106 cells cultivated in iron-rich and iron-deficient conditions were washed in cultivation

532 flasks by replacing the growth medium with 10 ml of PBS warmed to 37°C and resuspending in 7 ml of

533 37°C PBS. To assess the ability of the cells to phagocytose, 150 µl of pHrodo™

534 Green E. coli BioParticles™ Conjugate for Phagocytosis (Thermo Fisher Scientific, USA) was added, and

535 the cells were incubated for 3 hours at 37°C. Following incubation, the cells were washed with PBS,

536 detached on ice for 15 min, and the fluorescence caused by the phagocytosed particles was consecutively

25 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

537 analyzed using a Guava easyCyte 8HT flow cytometer as described above using a 488 nm laser and a

538 Green-B 525/30 nm detector. A negative control (without the addition of BioParticles) was used to

539 determine an appropriate threshold for N. fowleri cells. BioParticles resuspended in PBS were measured in

540 the same way to determine the background noise and gave a negligible signal. The effect of different iron

541 availability on the efficiency of phagocytosis of N. fowleri was established as the percentage of cells in

542 culture that had increased fluorescence.

543 To visualize the ability of N. fowleri to phagocytize, live amoebae incubated with fluorescent E. coli were

544 imaged with a Leica TCS SP8 WLL SMD-FLIM microscope (Leica, Germany) equipped with an HC PL

545 APO CS2 63x/1.20 water objective with 509 nm excitation and 526 nm-655 nm for HyD SMD and a PMT

546 detector for brightfield imaging. Images were processed using LAS X 3.5.1.18803 (Leica, Germany).

547

26 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

548 Acknowledgements

549 Special thanks to Ivánek for fruitful biochemical discussions.

27 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

550 References

551 1. De Jonckheere JF. What do we know by now about the genus Naegleria? Exp Parasitol. 2014

552 Nov;145(S):S2–9.

553 2. Kelly RB, Francine M-C, Charles PG. Occurrence of Naegleria fowleri in Arizona drinking water

554 supply wells . J Am Water Works Assoc. 2009 Nov;101(11):43–50.

555 3. Visvesvara GS, Moura H, Schuster FL. Pathogenic and opportunistic free-living amoebae:

556 Acanthamoeba spp., Balamuthia mandrillaris, Naegleria fowleri, and Sappinia diploidea. FEMS

557 Immunol Med Microbiol. 2007 Jun;50(1):1–26.

558 4. Grace E, Asbill S, Virga K. Naegleria fowleri: Pathogenesis, diagnosis, and treatment options.

559 Antimicrob Agents Chemother. 2015 Nov 1;59(11):6677–81.

560 5. Ma P, Visvesvara GS, Martinez AJ, Theodore FH, Daggett PM, Sawyer TK. Naegleria and

561 acanthamoeba infections: Review. Clin Infect Dis. 1990 May 1;12(3):490–513.

562 6. Zysset-Burri DC, Müller N, Beuret C, Heller M, Schürch N, Gottstein B, et al. Genome-wide

563 identification of pathogenicity factors of the free-living amoeba Naegleria fowleri. BMC

564 Genomics. 2014 Jun 19;15(1):496.

565 7. Centers for Disease Control and Prevention, Naegleria fowleri General Information. Available at:

566 https://www.cdc.gov/parasites/naegleria/general.html, last modified July 17/7/2018, accessed

567 15/3/2019 [Internet].

568 8. Linam WM, Ahmed M, Cope JR, Chu C, Visvesvara GS, Da Silva AJ, et al. Successful treatment

569 of an adolescent with Naegleria fowleri primary amebic meningoencephalitis. Pediatrics. 2015 Mar

570 1;135(3):e744–8.

571 9. Lill R, Broderick JB, Dean DR. Special issue on iron-sulfur proteins: Structure, function,

572 biogenesis and diseases. Biochim Biophys Acta - Mol Cell Res. 2015 Jun 1;1853(6):1251–2.

28 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

573 10. Boyd PW, Jickells T, Law CS, Blain S, Boyle EA, Buesseler KO, et al. Mesoscale iron enrichment

574 experiments 1993-2005: Synthesis and future directions. Science (80- ). 2007 Feb

575 2;315(5812):612–7.

576 11. Koka S, Föller M, Lamprecht G, Boini KM, Lang C, Huber SM, et al. Iron deficiency influences

577 the course of malaria in Plasmodium berghei infected mice. Biochem Biophys Res Commun. 2007

578 Jun;357(3):608–14.

579 12. Kabyemela ER, Fried M, Kurtis JD, Mutabingwa TK, Duffy PE. Decreased Susceptibility to

580 Plasmodium falciparum Infection in Pregnant Women with Iron Deficiency . J Infect Dis. 2008 Jul

581 15;198(2):163–6.

582 13. Sutak R, Lesuisse E, Tachezy J, Richardson DR. Crusade for iron: iron uptake in unicellular

583 eukaryotes and its significance for virulence. Trends Microbiol. 2008 Jun 1;16(6):261–8.

584 14. Drakesmith H, Prentice A. Viral infection and iron metabolism. Nat Rev Microbiol. 2008

585 Jul;6(7):541–52.

586 15. Ganz T. Iron in innate immunity: starve the invaders. Curr Opin Immunol. 2009;21(1):63–7.

587 16. Singh N, Haldar S, Tripathi AK, Horback K, Wong J, Sharma D, et al. Brain Iron Homeostasis:

588 From molecular mechanisms to clinical significance and therapeutic opportunities. Antioxidants

589 Redox Signal. 2014 Mar 10;20(8):1324–63.

590 17. Newsome AL, Wilhelm WE. Inhibition of Naegleria fowleri by microbial iron-chelating agents:

591 Ecological implications. Appl Environ Microbiol. 1983 Feb; 45(2): 665–668.

592 18. Romeo AM, Christen L, Niles EG, Kosman DJ. Intracellular chelation of iron by bipyridyl inhibits

593 DNA virus replication: Ribonucleotide reductase maturation as a probe of intracellular iron pools. J

594 Biol Chem. 2001 Jun 29;276(26):24301–8.

595 19. Megger N, Welte L, Zamora F, Müller J. Metal-mediated aggregation of DNA comprising 2,2′-

29 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

596 bipyridine nucleoside, an asymmetrically substituted chiral bidentate ligand. Dalt Trans. 2011 Feb

597 8;40(8):1802–7.

598 20. Nyayapati S, Afshan G, Lornitzo F, Byrnes RW, Petering DH. Depletion of cellular iron by BPS

599 and ascorbate: Effect on toxicity of adriamycin. Free Radic Biol Med. 1996 Jan 1;20(3):319–29.

600 21. Ihnat PM, Vennerstrom JL, Robinson DH. Synthesis and solution properties of deferoxamine

601 amides. J Pharm Sci. 2000 Dec;89(12):1525–36.

602 22. Kontoghiorghe CN, Kontoghiorghes GJ. Efficacy and safety of iron-chelation therapy with

603 deferoxamine, deferiprone, and deferasirox for the treatment of iron-loaded patients with non-

604 transfusion-dependent thalassemia syndromes. Drug Des Devel Ther. 2016;10:465–81.

605 23. Moos T. Brain iron homeostasis. Dan Med Bull. 2002 Nov;49(4):279–301.

606 24. Martínez-Castillo M, Ramírez-Rico G, Serrano-Luna J, Shibayama M. Iron-binding protein

607 degradation by cysteine proteases of Naegleria fowleri. Biomed Res Int. 2015 May 18;2015:1–8.

608 25. Dancis A, Roman DG, Anderson GJ, Hinnebusch AG, Klausner RD. Ferric reductase of

609 Saccharomyces cerevisiae: Molecular characterization, role in iron uptake, and transcriptional

610 control by iron. Proc Natl Acad Sci U S A. 1992 May 1;89(9):3869–73.

611 26. Jung SY, Kim JH, Song KJ, Lee YJ, Kwon MH, Kim K, et al. Gene silencing of nfa1 affects the in

612 vitro cytotoxicity of Naegleria fowleri in murine macrophages. Mol Biochem Parasitol. 2009 May

613 1;165(1):87–93.

614 27. Sutak R, Chamot C, Tachezy J, Camadro JM, Lesuisse E. Siderophore and haem iron use by

615 Tritrichomonas foetus. Microbiology. 2004 Dec 1;150(12):3979–87.

616 28. Alonso P, Zubiaur E. Phagocytic Activity of Three Naegleria Strains in the Presence of

617 Erythrocytes of Various Types. J Protozool. 1985 Nov;32(4):661–4.

618 29. Scaglia M, Gatti S, Brustia R, Chichino G, Rondanelli EG. Phagocytosis of human erythrocytes by

30 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

619 Naegleria is not related to species pathogenicity. A phase-contrast cinemicrographic study.

620 Microbiologica. 1991 Jan;14(1):45–53.

621 30. Létoffé S, Heuck G, Delepelaire P, Lange N, Wandersman C. Bacteria capture iron from heme by

622 keeping tetrapyrrol skeleton intact. Proc Natl Acad Sci U S A. 2009 Jul 14;106(28):11719–24.

623 31. Bradley SG, Toney DM, Zhang Y, Marciano-Cabral F. Dependence of growth, metabolic

624 expression, and pathogenicity of Naegleria fowleri on exogenous porphyrins. J Parasitol. 1996

625 Oct;82(5):763–8.

626 32. Tsaousis AD, Nývltová E, Sutak R, Hrdý I, Tachezy J. A Nonmitochondrial hydrogen production

627 in Naegleria gruberi. Genome Biol Evol. 2014 Apr;6(4):792–9.

628 33. Bexkens ML, Zimorski V, Sarink MJ, Wienk H, Brouwers JF, De Jonckheere JF, et al. Lipids Are

629 the Preferred Substrate of the Protist Naegleria gruberi, Relative of a Human Brain Pathogen. Cell

630 Rep. 2018 Oct 16;25(3):537-543.e3.

631 34. Mach J, Bíla J, Ženíšková K, Arbon D, Malych R, Glavanakovová M, et al. Iron economy in

632 Naegleria gruberi reflects its metabolic flexibility. Int J Parasitol. 2018 May 5;48(9–10):719–27.

633 35. Kulda J, Tachezy J, Čerkasovova A. In Vitro Induced Anaerobic Resistance to Metronidazole In

634 Trichomonas Vaginalis. J Eukaryot Microbiol. 1993 May;40(3):262–9.

635 36. Li X, Li J, Hu X, Huang L, Xiao J, Chan J, et al. Differential roles of the hemerythrin-like proteins

636 of Mycobacterium smegmatis in hydrogen peroxide and erythromycin susceptibility. Sci Rep. 2015

637 Nov 26;5:16130.

638 37. Ma Z, Strickland KT, Cherne MD, Sehanobish E, Rohde KH, Self WT, et al. The Rv2633c protein

639 of Mycobacterium tuberculosis is a non-heme di-iron catalase with a possible role in defenses

640 against oxidative stress. J Biol Chem. 2018;293(5):1590–5.

641 38. Selote D, Samira R, Matthiadis A, Gillikin JW, Long TA. Iron-binding e3 ligase mediates iron

31 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

642 response in plants by targeting basic helix-loop-helix transcription factors1. Plant Physiol. 2015

643 Jan;167(1):273–86.

644 39. Zeng WB, Chen WB, Yan QP, Lin GF, Qin YX. Hemerythrin is required for Aeromonas

645 hydraphlia to survive in the macrophages of Anguilla japonica. Genet Mol Res. 2016;15(2).

646 40. IC50 Calculator, available at: https://www.aatbio.com/tools/ic50-calculator, accessed 07/03/2019

647 [Internet].

648 41. Stookey LL. Ferrozine-A New Spectrophotometric Reagent for Iron. Anal Chem. 1970

649 Jun;42(7):779–81.

650 42. Aurrecoechea C, Barreto A, Brestelli J, Brunk BP, Caler E V., Fischer S, et al. AmoebaDB and

651 MicrosporidiaDB: Functional genomic resources for Amoebozoa and Microsporidia species.

652 Nucleic Acids Res. 2011 Jan 1;39(SUPPL. 1):D612-9.

653 43. Zimmermann L, Stephens A, Nam SZ, Rau D, Kübler J, Lozajic M, et al. A Completely

654 Reimplemented MPI Bioinformatics Toolkit with a New HHpred Server at its Core. J Mol Biol.

655 2018 Jul 20;430(15):2237–43.

656 44. Indiveri C, Iacobazzi V, Tonazzi A, Giangregorio N, Infantino V, Convertini P, et al. The

657 mitochondrial carnitine/acylcarnitine carrier: Function, structure and physiopathology. Mol

658 Aspects Med. 2011 Aug 1;32(4–6):223–33.

659 45. Hamel P, Saint-Georges Y, De Pinto B, Lachacinski N, Altamura N, Dujardin G. Redundancy in

660 the function of mitochondrial phosphate transport in Saccharomyces cerevisiae and Arabidopsis

661 thaliana. Mol Microbiol. 2004 Feb 23;51(2):307–17.

662 46. Šmíd O, Horáková E, Vilimova V, Hrdý I, Cammack R, Horvath A, et al. Knock-downs of Iron-

663 Sulfur Cluster Assembly Proteins IscS and IscU Down-regulate the Active Mitochondrion of

664 Procyclic Trypanosoma brucei. J Biol Chem. 2006 Sep 29;281(39):28679–86.

32 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

665 47. Hoff KG, Cupp-Vickery JR, Vickery LE. Contributions of the LPPVK motif of the iron-sulfur

666 template protein IscU to interactions with the Hsc66-Hsc20 chaperone system. J Biol Chem. 2003

667 Sep 26;278(39):37582–9.

668 48. Mittra B, Laranjeira-Silva MF, Perrone Bezerra de Menezes J, Jensen J, Michailowsky V, Andrews

669 NW. A Trypanosomatid Iron Transporter that Regulates Mitochondrial Function Is Required for

670 Leishmania amazonensis Virulence. Horn D, editor. PLoS Pathog. 2016 Jan 7;12(1):e1005340.

671 49. Brazzolotto X, Pierrel F, Pelosi L. Three conserved histidine residues contribute to mitochondrial

672 iron transport through mitoferrins. Biochem J. 2014 May 15;460(1):79–89.

673 50. Anders S, Huber W. Differential expression analysis for sequence count data. Genome Biol. 2010

674 Oct 27;11(10):R106.

675 51. MacLean B, Tomazela DM, Shulman N, Chambers M, Finney GL, Frewen B, et al. Skyline: An

676 open source document editor for creating and analyzing targeted proteomics experiments.

677 Bioinformatics. 2010 Apr 1;26(7):966–8.

678

679

33 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

680 Supporting information

681 S1 Fig. (A). Lack of 55Fe-transferrin uptake in N. fowleri, cultivated in iron-rich and iron-

682 deficient conditions. The uptake of transferrin-bound iron was assessed by incubation of N. fowleri

683 with 55Fe-transferrin. Tf, pure 55Fe-transferrin; Fe N. fowleri cultivated in iron-rich conditions for 72

684 hours, consecutively incubated with 55Fe-transferrin for 1 hour; BPS, N. fowleri cultivated in iron-deficient

685 conditions for 72 hours, consecutively incubated with 55Fe-transferrin for 1 hour; Ctr, iron uptake control

686 of N. fowleri culture cultivated in iron-deficiency incubated with 55Fe(III)-citrate for 1 hour. The

687 utilization of iron was analyzed by blue native electrophoresis as described in the Methods section. (B)

688 Mechanism of ferric iron uptake involves the reductive step. N. fowleri culture was incubated for

689 1 hour with 55Fe(III)-citrate with and without the addition of 0.2 mM BPS. Incorporation of 55Fe(III)-

690 citrate to cellular proteins was higher in the sample without the presence of BPS, indicating that a

691 reductive iron uptake mechanism takes place. Several distinct signals on the lower part of +BPS probably

692 correspond to residues of BPS complexed with ferrous iron radionuclides. The utilization of iron was

693 analyzed by blue native electrophoresis as described in the Methods section. -BPS, cell sample without

694 addition of BPS chelator; +BPS, cell sample with the addition of BPS chelator. (C) Loading control of

695 N. fowleri cell lysate for Western blot analysis of hemerythrin expression. Ponceau S loading

696 stain shows equal protein concentrations of loaded samples of N. fowleri cultivated in iron-deficient (BPS)

697 and iron-rich (Fe) conditions.

698 S2 Fig. N. fowleri cell phagocytosis. N. fowleri cells phagocytizing several fluorescently modified

699 pHrodo™ Green E. coli BioParticles™ (blue). Images were acquired with a Leica TCS SP8 WLL SMD-

700 FLIM microscope (Leica, Germany) equipped with an HC PL APO CS2 63x/1.20 water objective with

701 509 nm excitation and 526 nm-655 nm for HyD SMD and a PMT detector for brightfield imaging. Images

702 were processed using LAS X 3.5.1.18803 (Leica, Germany).

34 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

703 S3 Fig. Alignments of N. fowleri proteins with homologues from other organisms (A)

704 Alignment of N. fowleri NF0060430 with the IscU proteins from Homo sapiens, Saccharomyces

705 cerevisiae and T. brucei. Red arrows point to the conserved cysteine required for iron-sulfur cluster

706 assembly, based on a previous study [46]. The red rectangle denotes the conserved LPPVK motif of the

707 IscU proteins [47]. (B) Alignment of N. fowleri NF0079420 with the mitoferrin proteins of Trypanosoma

708 brucei, Leishmania mexicana, Saccharomyces cerevisiae and Homo sapiens. Red arrows point to the

709 sequence motif Px(D/E)xx(K/R)x(K/R), and yellow circles mark residues in contact with substrate,

710 according to a previous study [48]. Conserved histidine residues responsible for iron transport are marked

711 with blue stars [49]. (C) Alignment of N. fowleri NF0001420 with the mitochondrial phosphate carriers of

712 Saccharomyces cerevisiae, Homo sapiens and Arabidopsis thaliana. Red arrows point to residues

713 important for the phosphate transport activity, according to a previous study [45]. (D) Alignment of N.

714 fowleri NF0014630 with mitochondrial carnitine/acylcarnitine transferases of Saccharomyces cerevisiae,

715 Arabidopsis thaliana, Homo sapiens and Caenorhabditis elegans. The red rectangle denotes the signature

716 motifs Px(D/E)xx(R/K)x(R/K), and the arrows point to conserved residues, according to a previous study

717 [44].

718 S1 Table. Comparison of whole-cell proteomes of N. fowleri in iron-rich and iron-deficient

719 environments. List of whole-cell proteomes of N. fowleri compared in iron-rich and iron-deficient

720 conditions sorted into four sheets: raw data, all detected proteins, significantly upregulated proteins and

721 significantly downregulated proteins in iron-deficient conditions. Proteins with >2.3 (denoting upregulated

722 in iron-deficient conditions) or <-2.3-fold change (denoting downregulated in iron-deficient conditions)

723 are regarded as significantly regulated. Proteins were annotated from amoebaDB [42] on 25/7/2017, and

724 manual annotation was performed for selected proteins, as described in the Methods section. Probable

725 iron-containing proteins of the significantly downregulated and upregulated proteins are highlighted in

726 yellow.

35 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

727 S2 Table. Comparison of membrane proteomes of N. fowleri in iron-rich and iron-deficient

728 environments. List of membrane-enriched proteomes of N. fowleri compared in iron-rich and iron-

729 deficient conditions sorted into four sheets: raw data, all detected proteins, significantly upregulated

730 proteins and significantly downregulated proteins in iron-deficient conditions. Proteins with >2.3

731 (denoting upregulated in iron-deficient conditions) or <-2.3-fold change (denoting downregulated in iron-

732 deficient conditions) are regarded as significantly regulated. Proteins were annotated from amoebaDB [42]

733 on 25/7/2017, and manual annotation was performed for selected proteins, as described in the Methods

734 section.

735 S3 Table. Comparison of transcriptomes of N. fowleri in iron-rich and iron-deficient

736 environments. List of genes significantly downregulated and upregulated in iron-deficient conditions.

737 Raw data are included, and manual annotation was performed for selected proteins, as described in the

738 Methods section.

36 bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/763011; this version posted September 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.