bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

1 Is mRNA decapping activity of ApaH like phosphatases (ALPH’s) the reason for

2 the loss of cytoplasmic ALPH’s in all eukaryotes but Kinetoplastida?

3 Paula Andrea Castañeda Londoño1, Nicole Banholzer 1, Bridget Bannermann 2 and

4 Susanne Kramer #1

5

6

7 1 Zell- und Entwicklungsbiologie, Biozentrum, Universität Würzburg, Würzburg,

8 Germany

9 2 Department of Medicine, University of Cambridge, Cambridge, UK

10

11 # Corresponding author

12 Susanne Kramer

13 Tel.: +49 931 3186785

14 Email: [email protected]

15

16 Short title: Phylogeny of ALPHs

17

18 Keywords: ApaH like phosphatase, ApaH, ALPH, Trypanosoma brucei, mRNA

19 decapping, m7G cap, mRNA cap, ALPH1, Kinetoplastida

1 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

20 ABSTRACT

21 Background: ApaH like phosphatases (ALPHs) originate from the bacterial ApaH

22 protein and are present in eukaryotes of all eukaryotic super-groups; still, only two

23 proteins have been functionally characterised. One is ALPH1 from the Kinetoplastid

24 Trypanosoma brucei that we recently found to be the mRNA decapping enzyme of

25 the parasite. mRNA decapping by ALPHs is unprecedented in eukaryotes, which

26 usually use nudix hydrolases, but the bacterial ancestor protein ApaH was recently

27 found to decap non-conventional caps of bacterial mRNAs. These findings prompted

28 us to explore whether mRNA decapping by ALPHs is restricted to Kinetoplastida or

29 more widespread among eukaryotes.

30 Results: We screened 824 eukaryotic proteomes with a newly developed Python-

31 based algorithm for the presence of ALPHs and used the data to refine phylogenetic

32 distribution, conserved features, additional domains and predicted intracellular

33 localisation of ALPHs. We found that most eukaryotes have either no ALPH

34 (500/824) or very short ALPHs, consisting almost exclusively of the catalytic domain.

35 These ALPHs had mostly predicted non-cytoplasmic localisations, often supported by

36 the presence of transmembrane helices and signal peptides and in two cases (one in

37 this study) by experimental data. The only exceptions were ALPH1 homologues from

38 Kinetoplastida, that all have unique C-terminal and mostly unique N-terminal

39 extension, and at least the T. brucei enzyme localises to the cytoplasm. Surprisingly,

40 despite of these non-cytoplasmic localisations, ALPHs from all eukaryotic super-

41 groups had in vitro mRNA decapping activity.

42 Conclusions: ALPH was present in the last common ancestor of eukaryotes, but most

43 eukaryotes have either lost the enzyme since, or use it exclusively outside the

44 cytoplasm in organelles in a version consisting of the catalytic domain only. While

2 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

45 our data provide no evidence for the presence of further mRNA decapping enzymes

46 among eukaryotic ALPHs, the broad substrate range of ALPHs that includes mRNA

47 caps provides an explanation for the selection against the presence of a cytoplasmic

48 ALPH protein as a mean to protect mRNAs from unregulated degradation.

49 Kinetoplastida succeeded to exploit ALPH as their mRNA decapping enzyme, likely

50 using the Kinetoplastida-unique N- and C-terminal extensions for regulation.

3 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

51 BACKGROUND

52 Eukaryotic phosphatases play essential roles in regulating many cellular processes

53 and can be classified in several ways, based on catalytic mechanism, substrate

54 specificity, ion requirements and structure. One four-group classification

55 distinguishes phosphoprotein phosphatases (PPPs), metal-dependent protein

56 phosphatases (PPM, sometimes classified as a subgroup of PPP), protein tyrosine

57 phosphatases and aspartic acid-based phosphatases [1]. The eukaryotic PPP group

58 includes the Ser/Thr phosphatases PP1, PP2A, PP2B, PP4, PP5, PP6 and PP7 [1,2]

59 but has been extended to include three families of bacterial origin: Shewanella-like

60 SLP phosphatases, Rhizobiales-like (RLPH) phosphatases and ApaH-like (ALPH)

61 phosphatases [3-6].

62

63 ApaH like phosphatases (ALPHs) evolved from the bacterial ApaH protein and were

64 present in the last common ancestor of eukaryotes [3,5]. They are widespread

65 throughout the entire eukaryotic kingdom, albeit have been lost in certain sub-

66 branches such as land plants and chordates [6]. To the best of our knowledge, only

67 two ALPHs have been functionally characterised to date. One is the S. cerevisiae

68 ALPH protein (YNL217W), an Zn2+ dependent endopolyphosphatase of the vacuolar

69 lumen [7] that is also active with Co2+ and possibly Mg2+ [8]. The enzyme’s main

70 function is the cleavage of vacuolar poly(P). A function of Ppn2 in stress response is

71 suggested by the fact that the increase in short polyphosphate upon Ppn2

72 overexpression correlates with an increased resistance to peroxide and alkali [9]. The

73 second characterised ALPH protein is ALPH1 of the Kinetoplastida Trypanosoma

74 brucei: we recently found that ALPH1 is the only or major trypanosome mRNA

75 decapping enzyme [10], the enzyme that removes the m7 methylguanosine (m7G) cap

4 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

76 present at the 5´end of most eukaryotic mRNAs. This finding was surprising, as all

77 other known mRNA decapping enzymes belong to a different enzyme family, the

78 nudix hydrolases (the prototype is Dcp2). Trypanosomes lack orthologues to Dcp2

79 and all decapping enhancers, the likely reason why they use the ApaH like

80 phosphatase ALPH1 instead.

81

82 Interestingly, recent data indicate, that the bacterial precursor protein of ALPH,

83 ApaH, may have an analogous function to Trypanosome ALPH1 in decapping of

84 bacterial mRNAs. In vitro, ApaH cleaves diadenosine tetraphosphate (Ap4A) into two

85 molecules of ADP [11,12], but is also active towards other NpnN nucleotides (with

86 n≥3) [13-15]. Deletion of the ApaH gene in vivo causes a marked increase in Np4A

87 levels and a wide range of phenotypes [16-23]. Np4A was therefore suggested to act

88 as an alarmone (a signalling molecule involved in stress response), but no Np4A

89 receptor has yet been identified. Instead, recent data show that stress induced increase

90 in Np4A levels cause massive nucleoside-tetraphosphate capping of bacterial mRNAs

91 [24] mostly or entirely caused by usage of Np4A as non-canonical transcription

92 initiation nucleotide [25]. Many other dinucleoside polyphosphates can be used for

93 co-transcriptional capping even in the absence of stress, including methylated

94 versions [26]. ApaH is the major decapping enzyme for all dinucleoside

95 polyphosphate caps [24,26], suggesting that its main function is the regulation of gene

96 expression via regulating mRNA decapping. Intriguingly, the enzyme can both

97 enhance decapping (by decapping nucleoside-tetraphosphate capped RNA) and

98 inhibit decapping (by cleaving Np4A and preventing its incorporation to the mRNA).

99

5 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

100 Puzzled by these novel functions of both bacterial ApaH and trypanosome ALPH1 in

101 mRNA decapping, we here set out to investigate, whether mRNA decapping is a

102 major function of eukaryotic ALPHs, or whether this function is restricted to

103 trypanosomes. We developed a Python algorithm for the identification of ALPHs in

104 all available eukaryotic reference proteomes and identified 412 ALPHs in 324/824

105 proteomes. To our surprise, almost all ALPH proteins consisted exclusively of the

106 catalytic domain and almost all had predicted transmembrane regions or signal

107 peptides and predicted non-cytoplasmic localisation. Among the few exceptions were

108 all orthologues to trypanosome ALPH1 present in all organisms of the Kinetoplastida

109 group: these had C-terminal and, with only 2 exceptions N-terminal extensions and

110 cytoplasmic localisation. The absence of any (regulatory) domains and the non-

111 cytoplasmic localisation predictions argue against mRNA decapping being a

112 widespread function of eukaryotic ALPHs outside the Kinetoplastida. We tested three

113 randomly chosen ALPH proteins of three different eukaryotic super-groups for

114 mRNA decapping activity in vitro, and all were active, indicating a broad substrate

115 range of this enzyme class. The mRNA decapping activity of ALPHs provides a

116 possible explanation for the evolutionary selection against the presence of a

117 cytoplasmic ALPH in eukaryotes by either losing the gene, or targeting the protein to

118 organelles, as an effective mean to protect mRNAs from unregulated degradation.

119 Only trypanosomes have succeeded to exploit the mRNA decapping activity of ALPH

120 to their advantage, by adding regulatory domains, likely involved in regulating

121 enzyme activity.

122

123

124

125

6 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

126 RESULTS

127

128 Identification of ApaH like phosphatases in available eukaryotic proteomes

129 ALPHs belong to the PPP family of phosphatases and possess the four conserved

130 signature motifs (motif 1-4) of this family, GDxHG, GDxxDRG, GNHE, and HGG,

131 sometimes with conservative substitutions [2]. One distinctive feature of ALPHs are

132 two changes in the GDxxDRG motif: The second Asp is replaced by a neutral amino

133 acid and the Arg residue is replaced by Lys. In addition, ALPHs have two C-terminal

134 motifs [3,6] that we here call motif 5 and 6. We screened 824 complete eukaryotic

135 proteomes (Table S1a) for the presence of ApaH like phosphatases with a home-made

136 Python algorithm; these included all reference proteomes available on UniProt [27]

137 and all available Kinetoplastida proteomes available on TriTrypDB [28,29]. The

138 algorithm is based on recognising the six sequence-motifs characteristic for ALPHs.

139 The matrices used to define these motifs were stepwise optimised on yeast and

140 Kinetoplastida proteomes to not miss any ALPH (controlled by BLAST) while on the

141 other hand not to recognise PPPs, and in particular not the closely related

142 phosphatases SLP, RLPH and ApaH (sequences taken from [6]). The final algorithm

143 also included restrictions on distances between the motifs 1 and 2, 2 and 3 and 5 and 6

144 that we found to be highly conserved. BLAST screens on selected proteomes without

145 ALPH revealed that ALPHs were either fully absent or present in a truncated version

146 missing at least one of the motifs, often the most N or C-terminal one. Such truncated

147 ALPHs were not included to the final list, even though a subset may have arisen from

148 sequencing or annotation errors. Only for the Kinetoplastida, ALPHs with wrongly

149 annotated start codons were manually included (based on comparison with related

150 Kinetoplastida).

7 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

151

152 Figure 1 summarises all organisms of this study in phylogenetic groups based on the

153 latest eukaryotic classification suggested by [30]. For each group, the fraction of

154 organisms with and without ALPH is indicated in orange and blue, respectively. 324

155 of all 824 organisms included in this study have at least one ALPH, and these

156 organisms are distributed in a patchy way throughout all eukaryotes. Most

157 Euglenozoa have ALPHs, many Stramenopiles and Dikarya, but also Rhodophyceae

158 and Chloroplastida and many Metazoa. ALPHs are absent from land plants,

159 Apicomplexa and Ciliata. They are largely absent from Chordata (with the exception

160 of Branchiostoma floridae) and from Ecdysozoa (with 4 exceptions) and fully absent

161 from the few available proteomes of Amoebozoa and Metamonada. A phylogenetic

162 tree built from the catalytic domains of ALPHs mostly reflects the eukaryotic tree

163 (Supplementary Figure S1 and Supplementary Table S1b). Taken together, the data

164 indicate that ALPH was present in the last common ancestor of all eukaryotes and

165 was then selectively lost in certain sub-branches. Our data extend and agree to the

166 data of [6].

167

168 The aim of this work was not to analyse horizontal gene-transfer between eukaryotes,

169 bacteria and archaea; however, we could confirm the presence of ALPHs in a

170 subgroup (11/25) of Myxococcales, as described in [6] (Supplementary Table S2A)

171 and we detected ALPH in 1/285 archaean proteomes (OX=1906665

172 GN=EON65_52185, UniProt: UP000292173) (Supplementary Table S2B). All

173 prokaryotic ALPHs consisted mostly of the catalytic domain with almost no N- or C

174 terminal extensions.

175

8 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

176 General features of ApaH-like phosphatases

177 The dataset was used to refine the characteristics of ALPH proteins (Figure 2). 39%

178 of all analysed organisms have at least one ALPH isoform. The highest percentage is

179 found in Discoba with 94%, the lowest in Diaphoretickes with 22% (Figure 2A). Of

180 the 324 APLH-positive organisms, 20% have more than one ALPH isoform: most

181 (77%) have two and with one exception no organism has more than four (Figure 2B).

182 Organisms with multiple ALPHs were with 47% mostly enriched among the Discoba

183 (Figure 2B). The vast majority of all ALPH proteins is very short and consist mostly

184 of the catalytic domain (Figure 2C). The median size of the C-termini is with 26

185 amino acids very short. Only 49 ALPHs have C-termini longer than 100 amino acids

186 and most of these (31) are ALPHs of Discoba. The size of the N-terminus is slightly

187 more variable and has a median of 88 amino acids. Only 58 N-termini are longer than

188 200 amino acids and many of these (28) belong to ALPHs of Discoba. The largest

189 variance in the size of ALPH N- and C-termini is found in Discoba, reflecting the

190 presence of two very different ALPH variants in the Kinetoplastida (discussed

191 below).

192 The amino acid distances between some of the ALPH motifs were highly conserved

193 (Figure 2D): 95.2 of all ALPHs have between 28 and 30 amino acids between motif 1

194 and 2. The distance between motif 2 and 3 is 26 amino acids for 83.1% of ALPHs and

195 between 25 and 33 for 98.1% of ALPHs. The distance between the two C-terminal

196 motifs 5 and 6 is 19 amino acids for 93.2% of ALPHs. The distances between motif 3

197 and 4 and between motif 5 and 6 are less well conserved within all eukaryotes or

198 within Amorphea, but well conserved within the groups of Diaphoretickes and

199 Discoba. Sequence motifs were created for all six motifs (Figure 2E) [31]. Mostly,

9 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

200 these are conserved with some group-specific preferences at certain positions

201 indicated with orange bars (Figure 2E).

202

203 ALPHs of Opisthokonts (Dikarya and Holozoa)

204 The majority of available ALPH sequences (286) are from Dikarya, because the

205 number of available proteomes is high (288) and 81% of these proteomes contain at

206 least one ALPH. In this paper, we include Mucoromycota and Zoopagomycota to the

207 Dikarya as suggested by [30] and indicated in Figure 1. We investigated these ALPH

208 sequences further by looking for predicted domains (Interpro [32]), signal peptides

209 and trans-membrane helices (Phobius [33]) and predicted localisation (WoLF PSORT

210 [34]) (Fig. 3A and Table S1c). ALPHs of Dikarya have very short C-termini

211 (median=26 amino acids) and only slightly larger N-termini (median=97 amino acids)

212 and most (95%) do not contain any predictable domain in addition to the catalytic

213 domain. Of the 13 ALPHs that have a further domain, three have domains with

214 functions in cytochrome c complex assembly (IPR021150, IPR018793) indicating

215 mitochondrial functions and five ALPHs have a THIF-type NAD/FAD binding fold,

216 usually found in the ubiquitin activating E1 family, indicating a possible function in

217 protein degradation. ALPH of Lentinula edodes has a Peroxin-3 domain, indicating a

218 peroxisomal function. Two ALPHs of Rachicladosporium have Pectate lyase

219 domains, indicating a possible function degradation of cell wall material. ALPH of

220 Rhizopogon vesiculosus has a second ALPH domain.

221 The most interesting finding was the presence of a predicted trans-membrane region

222 and/or signal peptide within the C-terminal region in 79% of all Dikaryan ALPH

223 proteins. 184 ALPH proteins have predicted membrane helices (mostly one), a further

224 44 ALPH proteins have predicted signal peptides and only 61 ALPH proteins have

10 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

225 neither predicted (Fig. 3A, Table S1c). In agreement with these data, only 63 ALPH

226 proteins have predicted cytoplasmic localisation (mostly the ones without predicted

227 membrane helices and signal peptides). The remaining proteins are predicted to be

228 extracellular (31.8%), in the nucleus (27.8%), at the plasma membrane (21.1%), in the

229 mitochondrion (16.1), in the golgi (2.7%) or in peroxisomes (0.4%). Prediction data

230 need to be considered with care in the absence of experimental evidence, but taken

231 together, the data provide strong evidence for most Dikaryan APLH proteins being

232 non-cytoplasmic. Experimental data confirm non-cytoplasmic localisation for the

233 ALPH protein of S. cerevisiae (YNL217w, Ppn2): two high-throughput studies

234 indicate vacuolar localisation [35,36] and recent data show that Ppn2 is delivered to

235 the vacuolar lumen via the multivesicular body pathway, where it functions as an

236 endopolyphosphatase [7].

237

238 ALPHs are underrepresented in Holozoans (Fig. 1). In particular, all 130 vertebrate

239 proteomes lack ALPHs and of the three available non-vertebrate proteomes of

240 Chordata, only the Lancelet Branchiostoma floridae is ALPH-positive. Only four of

241 140 available Ecdysozoan proteomes contain ALPH; all are Arachnida. ALPHs are

242 present in subgroups of Cnidarians (2/3), Echinodermatans (1/2), Lophotrochozoens

243 (8/11), Placozoans (1/2) and Ichthyosporeans (1/2). 7 of these 18 Holozoan ALPH

244 proteins have a predicted signal peptide or transmembrane helix at their C-termini and

245 a non-cytoplasmic localisation prediction (mitochondrion, extracellular or ER); an

246 additional 5 proteins have a predicted non-cytoplasmic localisation only (Fig. 3B,

247 supplementary table S1d). Two proteins have an additional domain: ALPH of

248 Pomacea canaliculata has a domain of the FAD/NAD(P)-binding superfamily N-

249 terminal of the catalytic domain and ALPH of Leptotrombidium deliense may be

11 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

250 interacting with actin, as it has a Kaptin domain C-terminal of its ALPH domain (Fig.

251 3B, and Supplementary Table S1d).

252

253 ALPHs of Diaphoretickes

254 We found no ALPHs in land plants (100 proteomes), Apicomplexa (20 proteomes) or

255 Ciliata (3 proteomes). ALPH is present in 3/4 species of red algae, 11/18 species of

256 green algae (Chlorophyta), in the filamentous green algae Klebsormidium nitens, in

257 the photosynthetic Alveolate Vitrella brassicaformi and in the cryptophyte

258 algae Guillardia theta. 23/28 Stramenopiles have ALPHs: These are mostly (non-

259 photosynthetic) Oomycetes, including all strains of Phytophthora parasitica; these

260 plant pathogens have with 3 ALPH isoforms an unusually large number. Predicted

261 signal peptides or transmembrane helices are present at the C-termini of many

262 Chloroplastida ALPHs, ALPHs of Vitrella brassicaformi and in the ALPH of the

263 diatom Phaeodactylum tricornutum; in many cases this correlated with a predicted

264 localisation to the chloroplast. In contrast, ALPHs of Oomycetes have very short N-

265 and C-termini and mostly a predicted cytoplasmic localisation. Of all 55 ALPHs of

266 Diaphoretickes, only two have additional domains: Chlamydomonas reinhardtii

267 ALPH has a predicted Transposase IS605 domain (a transposon of bacterial origin)

268 and ALPH of Helicosporidium has a coatomer delta subunit domain.

269

270 ALPHs of Euglenozoa

271 ALPHs are present in all Kinetoplastida and in their close relative Euglena gracilis

272 (Supplementary Table Sf). The only possible exception is the free-living, non-

273 parasitic Bodo Saltans, but this genome is not yet complete.

12 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

274 ALPHs of Kinetoplastida fall into two groups: Each organism has exactly one ALPH

275 that is homologous to T. brucei ALPH1, the mRNA decapping enzyme [10]. These

276 ALPHs all have a C-terminal extension of between 220 and 278 nucleotides and, with

277 two exceptions (Leptomonas pyrrhocoris and T. grayi), they all have N-terminal

278 extensions of a similar size. The in vitro mRNA decapping activity of T. brucei

279 ALPH1 does not require the N-terminus [10], and the two ALPHs that lack the N-

280 terminus are therefore likely active in mRNA decapping too. The fact that no

281 Kinetoplastida strain has lost its ALPH1 homologue, and the absence of homologues

282 to the canonical mRNA decapping enzymes (DCP1/DCP2), indicates that all

283 Kinetoplastida rely on ALPH1 for mRNA decapping. In addition to the ALPH1

284 homologue, some Leishmania strains and all Trypanosoma and Paratrypanosoma

285 have between one and three additional ALPH proteins that consist exclusively of the

286 catalytic domain; four Leishmania strains have extensions within this domain. The

287 absence of non-ALPH1 homologues in some Leishmania strains as well as in

288 Leptomonas, Crithidia and Blechomonas indicates a less general and perhaps less

289 essential function of these ALPH proteins.

290 When a phylogenetic tree is constructed from the sequences of the catalytic domains,

291 the Kinetoplastida ALPH1 homologues form a separate group, distant to the group of

292 the non-ALPH1 homologues, indicating that the ALPH1 decapping enzyme has

293 evolved only ones in the last common ancestor of the Kinetoplastida (Figure 5A).

294 None of the ALPH sequences from Kinetoplastida contains any recognisable domain

295 or localisation signal. The decapping enzyme T. brucei ALPH1, fused to eYFP either

296 C- or N terminally, localises to the cytoplasm, P-bodies and to the posterior pole of

297 the cell and this cytoplasmic localisation is consistent with the essential function of

298 ALPH1 in mRNA decapping ([10] Figure 5A). To investigate the localisation of an

13 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

299 non-ALPH1 homologue, T. brucei ALPH2 was expressed as C-terminal eYFP fusion

300 in procyclic trypanosomes using an inducible expression system [37]. ALPH2 showed

301 a localisation pattern characteristic of mitochondrial proteins and co-localised with a

302 mitochondrial stain (MitoTrackerTM Orange) (Figure 5A).

303

304 Differences between the two different groups of Kinetoplastida ALPH proteins are

305 also obvious within the phosphatase and ALPH motifs: three positions show major

306 differences; most pronounced is the preference of GDVHG in decapping enzymes and

307 GDIHG in the non-decapping ALPH proteins.

308

309 Euglena is too distantly related to the Kinetoplastida to unequivocally design its

310 ALPH to the ALPH1 or non ALPH1 group, but its first and second sequence motif

311 (GDIHG and GDLVGKG) indicates a non-ALPH1, and this is also suggested by the

312 absence of N- and C-terminal sequences (Figure 5B). Interestingly, Euglena ALPH is

313 the only Kinetoplastida ALPH with a predicted trans-membrane domain. Euglena

314 ALPH was not enriched within purified mitochondria fractions [38] and also not

315 within chloroplasts (Martin Zoltner, Charles University in Prague, personal

316 communication). Like Kinetoplastida, Euglena has no recognisable homologue to the

317 canonical mRNA decapping enzyme DCP1/DCP2 in its genome the absence of an

318 ALPH1 homologue raises the question of how mRNA decapping is achieved in this

319 organism.

320

321 ALPHs have in vitro mRNAs decapping activity

322 The phylogeny data from us and others [6] show that ApaH like phosphatases were

323 present in the last common ancestor of eukaryotes. Since then, large percentages of

14 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

324 eukaryotes have lost the enzyme and of those ALPHs that are still present, large

325 percentages have predicted non-cytoplasmic localisations. The finding that the

326 trypanosome ALPH1 evolved to be the parasites mRNA decapping enzyme [10]

327 prompted us to test, whether ALPHs from other organisms have mRNA decapping

328 activity too. If true, this could provide an explanation for the evolutionary thrive

329 against the presence of cytoplasmic ALPHs in eukaryotes, concurrent with the

330 evolvement of the eukaryotic mRNA cap.

331 We produced recombinant ALPH proteins from randomly selected organisms of the

332 all three eukaryotic kingdoms that have ALPHs: we used ALPH of the Ichthyosporea

333 Sphaeroforma arctica, ALPH of the green algae Auxenochlorella protothecoides and

334 ALPH2 from Trypanosoma brucei (ALPH2 has mitochondrial localisation and is not

335 the mRNA decapping enzyme). As a control, ALPH of Sphaeroforma arctica was

336 also produced as an inactive mutant by mutating a conserved amino acid in the metal

337 ion binding motif (GDVIG:GNVIG [39]). All four proteins were purified from Arctic

338 express cells (Supplementary Figure S2A) and subsequently tested in in vitro

339 decapping assays using a 39 nucleotide long RNA oligo with a m7G cap structure as a

340 substrate. Capped and uncapped oligos can be distinguished by differences in gel

341 mobility on urea acrylamide gels complemented with acryloylaminophenyl boronic

342 acid [40,41]. To identify the optimal reaction conditions for each enzyme, the

343 decapping activity of the three active enzymes was first tested in the presence of

344 different bivalent ions (Mg2+, Mn2+, Co2+ and Zn2+) (Supplementary Figure S2B). All

345 enzymes had mRNA decapping activity, but the ion requirements for optimal activity

346 differed: ALPH of Sphaeroforma arctica showed highest decapping activity with

347 Mn2+ and some activity with Co2+, but none with Mg2+ or Zn2+. T. brucei ALPH2 had

348 best decapping activity with Co2+ and some with Mg2+ and possibly Mn2+, but none

15 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

349 with Zn2+. Auxenochlorella protothecoides ALPH had best activity with Mg2+ and

350 some with Co2+, but none with Zn2+ and Mn2+. Figure 6 shows the results of one

351 representative decapping assay for each of the enzymes, in the presence of the optimal

352 ion, and, as a control with one ion that promotes enzyme activity not or little.

353 Importantly, the catalytically inactive mutant ALPH of Sphaeroforma arctica had no

354 decapping activity. Together with the differences in ion requirements, this is strong

355 evidence that the activity of the wild type proteins is not caused by a contaminating

356 bacterial enzyme. The T. brucei decapping enzyme ALPH1 served as a positive

357 control: it has highest activity with Mg2+ and reduced activity with Mn2+. The data

358 show that all ALPH enzymes tested accept capped mRNA as a substrate in vitro.

359

360 DISCUSSION

361 mRNAs are essential molecules in both eukaryotes and prokaryotes and need to be

362 protected from unregulated exonucleolytic degradation. 5´-3´exoribonucleases require

363 monophosphorylated RNA 5´ends and the major protection strategy is therefore to

364 avoid monophosphates at the mRNAs 5´end. Most eukaryotic mRNAs are modified at

365 their 5´end by an m7 methylguanosine (m7G) cap, linked via an inverted 5´-5´bridge

366 [42]. A small fraction of eukaryotic RNAs carries non-canonical nucleotide

367 metabolite caps, for example an NAD+ cap [43,44]. Bacteria were long believed to

368 simply keep the triphosphate that is naturally present at their 5´end after transcription,

369 but more recently were found to have diphosphate 5´ends [45], and also non-

370 canonical metabolite caps, such as the NAD+ cap [46], the 3-dephospho-coenzyme A

371 cap (dpCoA) [47] or dinucleoside polyphosphate caps [24-26]. Archaea mRNAs are

372 not well studied by do carry triphosphates caps [48].

16 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

373 In common to the increasing number of different mRNA caps and mRNA 5´ end

374 modifications discovered throughout all kingdoms of life is the pyrophosphate bond,

375 and therefore the requirement of a pyrophosphatase to remove the cap or the

376 polyphosphate. Enzymes of four different families can cleave pyrophosphate bonds at

377 the 5´end of mRNAs [49]. DXO and histidine triad (HIT) proteins, act mainly on

378 faulty RNAs or on cap remnants or remove non-canonical NAD+ caps without

379 cleaving the pyrophosphate bond [49]. This leaves two pyrophosphatase families that

380 act on intact mRNAs. Nudix domain proteins include the prototypic eukaryotic

381 mRNA decapping enzyme Dcp2 [50] and several further eukaryotic nudix hydrolases

382 with activity to the m7G cap and non-canonical metabolite caps [51-53] as well as the

383 bacterial RppH [54] and NudC proteins [55] that act on di- or triphosphorylated

384 mRNAs and on NAD+ caps [49]. Enzymes of the fourth group, the ApaH family only

385 entered the field of mRNA decapping most recently. ApaH like phosphatase ALPH1

386 from Trypanosoma brucei was shown to be the only or major mRNA decapping

387 enzyme in these parasites [10]. Bacterial ApaH cleaves of non-canonical nucleoside-

388 tetraphosphate caps [24-26] and likely regulates gene expression during stress.

389

390 The major aim of this work was to find out, how widespread mRNA decapping by

391 ApaH like phosphatases is. To identify ALPH proteins in all available eukaryotic

392 proteomes, we have developed a Python based algorithm based on recognising the six

393 conserved motifs of ALPH. We have started with a well-defined training set [6] and

394 stepwise optimised the matrix of the motifs to recognise all ALPHs, but no related

395 phosphatases. Our comprehensive datasets on the 412 ALPH proteins are summarized

396 in detail in Table S1 and contribute to a more comprehensive knowledge of this

397 enzyme family.

17 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

398

399 Our data provide no evidence for the presence of additional mRNA decapping

400 enzymes among eukaryotic ApaH-like phosphatases. We expected putative mRNA

401 decapping enzymes to fulfil two criteria: they must be cytoplasmic and they must

402 possess additional sequences next to the catalytic domain that can regulate enzyme

403 activity. The Trypanosoma brucei decapping enzyme ALPH1 is cytoplasmic and has

404 C- and N- terminal extensions. Outside the Kinetoplastida, most ALPHs consist of the

405 catalytic domain only and have predicted non-cytoplasmic localisations, often further

406 supported by predicted transmembrane helices and signal peptides. In the few cases

407 when additional domains were present, these did not indicate any connection to

408 mRNA metabolism or showed any similarities to the Kinetoplastida C- and N-

409 terminal extensions.

410

411 Surprisingly, all three ALPH proteins that we tested, had mRNA decapping activity in

412 vitro, even though none of these enzymes is likely to use mRNA as their

413 physiological substrate: T. brucei ALPH2 localises to the mitochondrion (Figure 5A)

414 and ALPHs of Sphaeroforma arctica and Auxenochlorella protothecoides both have

415 predicted signal peptides and predicted localisation to the mitochondrion and

416 chloroplast, respectively. This rather broad substrate range of ALPHs is not

417 unexpected: the related bacterial ApaH protein cleaves NpnN nucleotides [13-15] as

418 well was mRNA capped with NpnN [24,26] and even has phosphatase and ATPase

419 activity in addition to its pyrophosphatase activity [56]. The only other characterised

420 ALPH protein, the S. cerevisiae ALPH protein Ppn2 has, to our knowledge, not been

421 tested for mRNA decapping activity. Substrate unspecificity is not restricted to ApaH

422 and ApaH like phosphatases, but is also found in enzymes of the nudix domain family

18 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

423 [57]: the major bacteria decapping enzyme RppH was initially identified as

424 diadenosine polyphosphate hydrolase [58] and can also cleave NAD+ caps [53] and

425 recent in vitro studies of mammalian nudix proteins shows that many have mRNA

426 decapping activity towards a range of caps and some also towards free metabolites

427 [51,53,59,60]; how many of these nudix proteins have mRNAs as their natural

428 substrates is still debated.

429

430 The broad substrate range of ALPHs that includes mRNA caps provides an

431 explanation for its loss or non-cytoplasmic localisation in most eukaryotes. The last

432 common ancestor of all eukaryotes had ALPH. This assumption is supported by the

433 presence of ALPH proteins in all eukaryotic kingdoms found in this and earlier [6]

434 studies. With the evolvement of the eukaryotic mRNA cap, eukaryotes faced two

435 problems: first, the cap needed to be protected from unregulated degradation by

436 pyrophosphatases and second, an mRNA decapping enzyme was needed for regulated

437 degradation. The first challenge was likely addressed in multiple ways, including for

438 example protection by cap binding proteins like the eIF4G complex. However, one

439 important evolutionary strategy may have been to inactivate the bacterial-inherited

440 ALPH proteins. Many eukaryotes lost ALPH proteins, others have transferred them

441 into organelles away from mRNA substrates, where they fulfil essential functions, for

442 example in poly-phosphate metabolism [7]. Evolution has solved the second

443 challenge by transforming available pyrophosphatases with wide substrate ranges into

444 regulatable mRNA decapping enzymes. Mostly nudix hydrolases were exploited for

445 this task, but Kinetoplastida uniquely use one of their ALPH proteins.

446

19 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

447 One open question is, whether all organisms outside the Kinetoplastida use DCP2

448 orthologues as their major mRNA decapping enzyme. While DCP2 has been

449 functionally characterised in yeast [50], human [61] and plants [62], there are no

450 experimental data on Metamonada and Discoba, and DCP2 orthologues are difficult

451 to identify by blast searches alone. Metamonada possibly have a DCP2 homologue:

452 in Trichomonas this can be identified by Blast and in Giardia by more sophisticated

453 in silico methods [63]. Whether Discoba have DCP2 is unclear. A possible DCP2

454 homologue is identifiable by blast in the Discoba Naegleria, but not in Euglena. The

455 question whether DCP2 was present in the last common ancestor and selectively lost

456 in Kinetoplastida, or whether it evolved subsequently to the separation of the Discoba

457 cannot be answered without experimental data.

458

459 In the last few years, more and more mRNA cap variants have been discovered in

460 both prokaryotes and eukaryotes, and more and more enzymes that act in mRNA

461 decapping. Our findings that ApaH like phosphatases have mRNA decapping activity,

462 but, with the exception of the Kinetoplastida enzymes, are not involved in mRNA

463 decapping in eukaryotes is an important contribution towards a more comprehensive

464 picture of mRNA decapping.

465

466 CONCLUSIONS

467 Despite of being present in all eukaryotic kingdoms, ApaH like phosphatases are

468 poorly studied. Here, we present a new algorithm for unequivocal classification of a

469 protein as an ALPH and we provide a comprehensive dataset of ApaH like

470 phosphatases of all available reference proteomes (Table S1). 500/824 proteomes had

471 no ALPH and the remaining 324 proteomes contained, in total, 412 ALPH proteins.

20 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

472 Almost all ALPHs consisted of the catalytic domain only and had non-cytoplasmic

473 localisation predictions supported by the presence of transmembrane helices and

474 signal peptides. The marked exceptions are ALPHs of the Kinetoplastida that are

475 orthologues to the Trypanosoma brucei mRNA decapping enzyme ALPH1: all have

476 C- and most have N-terminal extensions. Despite of their non-cytoplasmic

477 localisations that excluded mRNAs as their natural substrates, we found that ALPH

478 proteins from all eukaryotic kingdoms had mRNA decapping activity in vitro,

479 suggesting that (i) the loss or non-cytoplasmic localisation of ALPHs in most

480 eukaryotes was caused by evolutionary pressure to protect mRNAs from uncontrolled

481 degradation and (ii) the N- and C- terminal extensions of the Kinetoplastida ALPH

482 proteins serve to regulate the mRNA decapping activity of ALPH. mRNA decapping

483 is one of only two known ALPH functions, but our data strongly suggest that this

484 function is restricted to ALPHs of Kinetoplastida. The physiological functions of

485 most other ALPHs remain to be discovered. In vivo experiments are essential, as the

486 wide substrate range of ALPH will impede the identification of physiological

487 substrates in vitro.

488

489 METHODS

490 Identification of ApaH like phosphatases

491 A novel algorithm was developed using Python 3 to identify ALPHs in a set of

492 proteome fasta files. The algorithm initially used narrow matrices to identify the 6

493 conserved motifs (4 common to all PPP and 5-6 unique to ALPHs) based on the data

494 of [3,6], these matrices were successively extended after training on a set of 46 yeast

495 proteomes, controlled by BLAST, and later used to screen 824 eukaryotic proteomes.

496 This screen gave information about conserved and non-conserved distances between

21 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

497 the 6 motifs. All proteins that had only one motif missing were manually inspected

498 further. If they possessed the missing motif in an only slightly, conservatively-altered

499 version and, at the same time, had the correct distances between the motifs (where

500 these distances are conserved), the motif matrix was extended to include this motif.

501 The final version of the algorithm uses one narrow matrix for the initial screen:

502 motive 1 ["G", "D", "VILT", "HQ", "G"]; motive 2 ["G", "DN", "LMIFVT",

503 "IVTLEAC", "NSATFGVEQIDHM", "KQN", "GASHT"]; motive 3 ["G", "N",

504 "HNWQD", "ED"]; motive 4= ["H", "AG", "G"]; motive 5 ["VIT", "IVYFLAMT",

505 "FY", "G", "H"]; motive 6 ["LVIMT", "DE", "STGL", "GASRN"]. If not all motifs

506 were found, the algorithm tolerates up to two motives from an extended matrix:

507 motive 1 ["GxX", "DPEKRxX", "QWERTIPASDFGHKLYCVNMxX", "HAQxX",

508 "GFSAxX"]; motive 2 ["GxX", "DNxX", "QWERTIPASDFGHKLYCVNMxX",

509 "ILCVTMEAxX", "NDSTAGCHMFEVQIxX", "KQNxX", "GASHTxX"]; motive 3

510 ["GxX", "NxX", "HNTWQDxX", "EDxX"]; motive 4 ["HxX", "AGLVxX", "GxX"];

511 motive 5 ["VILCMTxX", "QWERTIPASDFGHKLYCVNMxX", "FYCxX", "GxX",

512 "HxX"]; motive 6 ["LVIMTGxX", "DExX", "ATGSVLxX", "GRASNExX"].

513 Furthermore, the programme restricts the distances between the motifs to 14-90

514 (motive 1 to 2), 25-80 (motive 2 to 3) and 17-30 (motive 5 to 6). ALPHs with missing

515 motives due to wrong-annotated start codons or sequencing errors will not be

516 detected, with the exception of an unidentified amino acid included in the extended

517 matrix (x, X). Only for the Kinetoplastida, such ALPHs were manually included.

518 Initially, the software tolerated R at position 6 in motive 2. However, distinction

519 between ALPHs and ApaH / RLPH was not possible in all cases and therefore R at

520 this position was not tolerated by either matrix any longer. This affected 19 proteins

22 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

521 that had been previously recognised as an ALPH (14 discoba, 3 metazoa and 2

522 embryophyta). The Python programme is available upon request.

523

524 Cloning and expression of recombinant proteins

525 The mRNA decapping enzyme T. brucei ALPH1 was expressed as an N-terminal

526 truncation with a C-terminal 6xHis-Tag in RosettaBlue competent bacteria cells

527 (Novagen) as previously described [10].

528 All other proteins were expressed either full length (T. brucei ALPH2

529 (Tb927.4.4330)) or with minor N-terminal truncations to exclude the transmembrane

530 regions (Sphaeroforma arctica: amino acids 21-316; Auxenochlorella protothecoides:

531 amino acids 21-327) in Arctic Express DE competent cells (Agilent). At the N-

532 terminus, the proteins were fused to a tag consisting of a 6xHis-Tag, a Thrombin

533 cleavage site, a SUMO tag and a TEV cleavage site, resulting in the following

534 additional protein sequence:

535 MGSSHHHHHHSSGLVPRGSASMSDSEVNQEAKPEVKPEVKPETHINLKVSDG

536 SSEIFFKIKKTTPLRRLMEAFAKRQGKEMDSLRFLYDGIRIQADQTPEDLDME

537 DNDIIEAHREQIGGHMENLYFQGEASAT. At the C-terminus, all proteins had

538 additional GSGSGC, due to cloning reasons. For the inactive mutant of the

539 Sphaeroforma arctica ALPH, the highly conserved GDVIG motif involved in metal

540 ion binding was mutated to GNVIG [39]. The purification of the recombinant proteins

541 was done via Nickel-agarose beads using standard procedures as previously described

542 [10], except that expression in Arctic Express DE cells was done at 10°C, following

543 the instructions of the company. The concentration of all purified proteins was

544 estimated from Coomassie gels using a titration of a protein with known

545 concentration for calibration.

23 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

546

547 Phylogeny

548 The Maximum likelihood tree of the Euglenozoa ALPHs (Figure 5) was done with all

549 ALPH catalytic domains (defined as starting 6 amino acids upstream of motif 1 and

550 ending 12 amino acids downstream of motif 6) using MEGA 7 [64] with 500

551 bootstrap cycles. The sequences of Crithidia fasciculata and Trypanosoma vivax were

552 gap-corrected: one gap was removed).

553

554 Proteomes

555 Proteomes were downloaded from Uniprot, using all available reference proteomes of

556 UniProt_release 2019_02 (For Archaea: 2019_11). [27]. Table S1a lists all proteomes

557 used in this work. For Kinetoplastida, we downloaded all available proteomes from

558 TriTrypDB (February 2019) [28,29].

559

560 Work with trypanosomes

561 Trypanosoma brucei Lister 427 procyclic cells expressing a TET repressor (pSRP2.1)

562 were used for all experiments [37]. Cells were cultured in SDM-79 [65] at 27 ̊C and

563 5% CO2. Transgenic trypanosomes were generated using standard procedures [66].

564 All experiments used logarithmically growing trypanosomes. The mitochondria were

565 stained by adding 1 µM MitoTrackerTM Orange CMTMRos (Invitrogen) to living

566 cells; imaging was done within 30 minutes.

567

568 Microscopy

569 Z-stack images (100 stacks at 100 nm distance) were taken with a custom build TILL

570 Photonics iMic microscope equipped with a sensicam camera (PCO), deconvolved

24 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

571 using Huygens Essential software (Scientific Volume Imaging B. V., Hilversum, The

572 Netherlands). All images are presented as Z-projections (method sum sliced)

573 produced by ImageJ software.

574

575 In vitro decapping assays

576 Each decapping reaction was done in 10 µl volume (with RNAse-free water)

577 containing 50 mM Tris-HCl (pH 7.9), 100 mM NaCl, 1 mM metal ions (variable), 40

578 units Ribolock RNAse inhibitor (ThermoFisher Scientific), 0.25 µM recombinantly-

579 produced ALPH enzyme and 3.8 µM capped RNA oligo m7G-

580 AACUAACGCUAUUAUUAGAACAGUUUCUGUACUAUAUUG (Bio-Synthesis,

581 Texas, USA). The sample was incubated for 1 hour at 37°C. Control reactions were

582 done without enzyme. The reaction was stopped by ethanol precipitation: 1 µl 3 M

583 RNAse-free sodium acetate (pH 5.5) (Ambion), 1 µl RNAse-free glycogen

584 (ThermoFisher Scientific) and 30 µl cold 100% ethanol was added, the sample

585 incubated at -20°C for 30 minutes, followed by centrifugation (15 minutes, 20,000 g).

586 The pellet (containing the RNA oligo) was washed once in 80% ethanol, air-dried,

587 resolved in 5 µl RNAse-free water (ThermoFisher Scientific) and either stored at -

588 80°C or directly prepared for loading to the gel.

589 RNA Samples were separated on urea acrylamide gels that contained

590 acryloylaminophenyl boronic acid [40,41]: boronyl groups form stable complexes

591 with the m7G cap structure this way increasing the difference in gel mobility between

592 capped and uncapped oligos [40,41]. For 12% gels, a 10 ml gel solution was freshly

593 prepared with distilled water, containing 12% Acrylamide/Bis-acrylamide (19:1)

594 (Sigma-Aldrich), 7M urea (Sigma-Aldrich), 0.1% ammonium persulfate, 4 µl

595 TEMED, 0.25% acryloylaminophenyl boronic acid (Sigma-Aldrich) and 100 mM

25 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

596 Tris-acetate (pH 9.0). Gels were run in 0.5x TBE buffer (45 mM Tris-HCl pH 8.3, 45

597 mM boric acid, 1 mM EDTA) using the Mini-PROTEAN Tetra Vertical

598 Electrophoresis cell (gel size 10x8 cm, BioRad). Each run was preceded by a 30 min

599 pre-run at 300 V to warm the gel. The purified RNA samples were prepared for gel

600 electrophoresis by adding 1 µl loading dye (gel loading buffer II, Invitrogen) followed

601 by incubation at 95°C for 5 minutes. Samples were loaded immediately to the pre-

602 rinsed wells of the pre-run gel and the gel was run at 200 V (after a 10 min run at

603 170V) for 30-40 minutes, stained for 20 min in SYBR™ Green II RNA Gel Stain

604 (1:10,000 in 0.5x TBE, ThermoFisher Scientific) and imaged on the iBrightTM

605 (Invitrogen) with nucleic acid settings.

606

607 DECLARATIONS

608 Ethics approval and consent to participate

609 Not applicable

610 Consent for publication

611 Not applicable

612 Availability of data

613 All data generated or analysed during this study are included in this published article

614 [and its supplementary information files].

615 Competing interests

616 The authors declare that they have no competing interests.

617 Funding

618 This work was funded by the Deutsche Forschungsgemeinschaft (Kr4017 4-1).

619 Authors' contributions

26 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

620 SK wrote the manuscript and did the in silico work of this study, with the exception of

621 the phylogenetic tree of Figure S1 that was done by BB. PACL and NB did all wet

622 work (protein expression, decapping assays).

623 Acknowledgement

624 We like to thank Martin Zoltner (Charles University in Prague, Czech Republic),

625 Mark Carrington (University of Cambridge, UK) and Mark Field (University of

626 Dundee, UK) for sharing unpublished proteomes and data on Euglena and Bodo

627 Saltans.

628

629 REFERENCES

630 1. Kerk D, Templeton G, Moorhead GBG. Evolutionary radiation pattern of novel 631 protein phosphatases revealed by analysis of protein data from the completely 632 sequenced genomes of humans, green algae, and higher plants. Plant Physiol. 633 American Society of Plant Biologists; 2008;146:351–67.

634 2. Shi Y. Serine/threonine phosphatases: mechanism through structure. Cell. 635 2009;139:468–84.

636 3. Andreeva AV, Kutuzov MA. Widespread presence of “bacterial-like” PPP 637 phosphatases in eukaryotes. BMC Evol Biol. 2004;4:47.

638 4. Uhrig RG, Moorhead GB. Two ancient bacterial-like PPP family phosphatases 639 from Arabidopsis are highly conserved plant proteins that possess unique properties. 640 Plant Physiol. American Society of Plant Biologists; 2011;157:1778–92.

641 5. Uhrig RG, Labandera A-M, Moorhead GB. Arabidopsis PPP family of 642 serine/threonine protein phosphatases: many targets but few engines. Trends Plant 643 Sci. 2013;18:505–13.

644 6. Uhrig RG, Kerk D, Moorhead GB. Evolution of bacterial-like phosphoprotein 645 phosphatases in photosynthetic eukaryotes features ancestral mitochondrial or 646 archaeal origin and possible lateral gene transfer. Plant Physiol. 2013;163:1829–43.

647 7. Gerasimaitė R, Mayer A. Ppn2, a novel Zn2+-dependent polyphosphatase in the 648 acidocalcisome-like yeast vacuole. J Cell Sci. 2017;130:1625–36.

649 8. Andreeva N, Ledova L, Ryazanova L, Tomashevsky A, Kulakovskaya T, Eldarov 650 M. Ppn2 endopolyphosphatase overexpressed in Saccharomyces cerevisiae: 651 Comparison with Ppn1, Ppx1, and Ddp1 polyphosphatases. Biochimie. Elsevier B.V; 652 2019;163:101–7.

27 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

653 9. Ryazanova LP, Ledova LA, Andreeva NA, Zvonarev AN, Eldarov MA, 654 Kulakovskaya TV. Inorganic Polyphosphate and Physiological Properties of 655 Saccharomyces cerevisiae Yeast Overexpressing Ppn2. Biochemistry Mosc. 656 2020;85:516–22.

657 10. Kramer S. The ApaH-like phosphatase TbALPH1 is the major mRNA decapping 658 enzyme of trypanosomes. PLoS Pathog. 2017;13:e1006456.

659 11. Barton GJ, Cohen PT, Barford D. Conservation analysis and structure prediction 660 of the protein serine/threonine phosphatases. Sequence similarity with diadenosine 661 tetraphosphatase from Escherichia coli suggests homology to the protein 662 phosphatases. Eur J Biochem. 1994;220:225–37.

663 12. Koonin EV. Bacterial and bacteriophage protein phosphatases. Mol Microbiol. 664 1993;8:785–6.

665 13. Plateau P, Fromant M, Brevet A, Gesquière A, Blanquet S. Catabolism of bis(5'- 666 nucleosidyl) oligophosphates in Escherichia coli: metal requirements and substrate 667 specificity of homogeneous diadenosine-5',5'‘’-P1,P4-tetraphosphate 668 pyrophosphohydrolase. Biochemistry. 1985;24:914–22.

669 14. Guranowski A, Jakubowski H, Holler E. Catabolism of diadenosine 5',5"'-P1,P4- 670 tetraphosphate in procaryotes. Purification and properties of diadenosine 5‘,5"’- 671 P1,P4-tetraphosphate (symmetrical) pyrophosphohydrolase from Escherichia coli 672 K12. J Biol Chem. 1983;258:14784–9.

673 15. Guranowski A. Specific and nonspecific enzymes involved in the catabolism of 674 mononucleoside and dinucleoside polyphosphates. Pharmacol Ther. 2000;87:117–39.

675 16. Kimura Y, Tanaka C, Sasaki K, Sasaki M. High concentrations of intracellular 676 Ap4A and/or Ap5A in developing Myxococcus xanthus cells inhibit sporulation. 677 Microbiology (Reading). 2017;163:86–93.

678 17. Ismail TM, Hart CA, McLennan AG. Regulation of dinucleoside polyphosphate 679 pools by the YgdP and ApaH hydrolases is essential for the ability of Salmonella 680 enterica serovar typhimurium to invade cultured mammalian cells. J Biol Chem. 681 American Society for Biochemistry and Molecular Biology; 2003;278:32602–7.

682 18. Hansen S, Lewis K, Vulić M. Role of global regulators and nucleotide metabolism 683 in antibiotic tolerance in Escherichia coli. Antimicrob Agents Chemother. 684 2008;52:2718–26.

685 19. Nishimura A, Moriya S, Ukai H, Nagai K, Wachi M, Yamada Y. Diadenosine 686 5',5'“-”P1,P4-tetraphosphate (Ap4A) controls the timing of cell division in 687 Escherichia coli. Genes Cells. 1997;2:401–13.

688 20. Johnstone DB, Farr SB. AppppA binds to several proteins in Escherichia coli, 689 including the heat shock and oxidative stress proteins DnaK, GroEL, E89, C45 and 690 C40. EMBO J. 1991;10:3897–904.

691 21. Monds RD, Newell PD, Wagner JC, Schwartzman JA, Lu W, Rabinowitz JD, et 692 al. Di-adenosine tetraphosphate (Ap4A) metabolism impacts biofilm formation by

28 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

693 Pseudomonas fluorescens via modulation of c-di-GMP-dependent pathways. J. 694 Bacteriol. 2010;192:3011–23.

695 22. Ji X, Zou J, Peng H, Stolle A-S, Xie R, Zhang H, et al. Alarmone Ap4A is 696 elevated by aminoglycoside antibiotics and enhances their bactericidal activity. Proc 697 Natl Acad Sci USA. 2019;116:9578–85.

698 23. Farr SB, Arnosti DN, Chamberlin MJ, Ames BN. An apaH mutation causes 699 AppppA to accumulate and affects motility and catabolite repression in Escherichia 700 coli. Proc Natl Acad Sci USA. 1989;86:5010–4.

701 24. Luciano DJ, Levenson-Palmer R, Belasco JG. Stresses that Raise Np4A Levels 702 Induce Protective Nucleoside Tetraphosphate Capping of Bacterial RNA. Mol Cell. 703 Elsevier Inc; 2019;:1–19.

704 25. Luciano DJ, Belasco JG. Np4A alarmones function in bacteria as precursors to 705 RNA caps. Proc Natl Acad Sci USA. 2020;117:3560–7.

706 26. Hudeček O, Benoni R, Reyes-Gutierrez PE, Culka M, Šanderová H, Hubálek M, 707 et al. Dinucleoside polyphosphates act as 5'-RNA caps in bacteria. Nat Commun. 708 Nature Publishing Group; 2020;11:1052–11.

709 27. UniProt Consortium. UniProt: a worldwide hub of protein knowledge. Nucleic 710 Acids Res. 2019;47:D506–15.

711 28. Aslett M, Aurrecoechea C, Berriman M, Brestelli J, Brunk BP, Carrington M, et 712 al. TriTrypDB: a functional genomic resource for the Trypanosomatidae. Nucleic 713 Acids Res. 2010;38:D457–62.

714 29. Warrenfeltz S, Basenko EY, Crouch K, Harb OS, Kissinger JC, Roos DS, et al. 715 EuPathDB: The Eukaryotic Pathogen Genomics Database Resource. Methods Mol 716 Biol. 2018;1757:69–113.

717 30. Adl SM, Bass D, Lane CE, Lukes J, Schoch CL, Smirnov A, et al. Revisions to 718 the Classification, Nomenclature, and Diversity of Eukaryotes. J Eukaryot Microbiol. 719 John Wiley & Sons, Ltd (10.1111); 2018;66:jeu.12691–116.

720 31. Crooks GE, Hon G, Chandonia J-M, Brenner SE. WebLogo: a sequence logo 721 generator. Genome Res. Cold Spring Harbor Lab; 2004;14:1188–90.

722 32. Mitchell AL, Attwood TK, Babbitt PC, Blum M, Bork P, Bridge A, et al. InterPro 723 in 2019: improving coverage, classification and access to protein sequence 724 annotations. Nucleic Acids Res. 2019;47:D351–60.

725 33. Käll L, Krogh A, Sonnhammer ELL. Advantages of combined transmembrane 726 topology and signal peptide prediction--the Phobius web server. Nucleic Acids Res. 727 2007;35:W429–32.

728 34. Horton P, Park K-J, Obayashi T, Fujita N, Harada H, Adams-Collier CJ, et al. 729 WoLF PSORT: protein localization predictor. Nucleic Acids Res. 2007;35:W585–7.

730 35. Huh W-K, Falvo JV, Gerke LC, Carroll AS, Howson RW, Weissman JS, et al.

29 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

731 Global analysis of protein localization in budding yeast. Nature. 2003;425:686–91.

732 36. Breitkreutz A, Choi H, Sharom JR, Boucher L, Neduva V, Larsen B, et al. A 733 global protein kinase and phosphatase interaction network in yeast. Science. 734 2010;328:1043–6.

735 37. Sunter J, Wickstead B, Gull K, Carrington M. A new generation of T7 RNA 736 polymerase-independent inducible expression plasmids for Trypanosoma brucei. 737 PLoS ONE. 2012;7:e35167.

738 38. Hammond MJ, Nenarokova A, Butenko A, Zoltner M, Dobáková EL, Field MC, 739 et al. A uniquely complex mitochondrial proteome from Euglena gracilis. Ãvila- 740 Arcos MC, editor. Mol. Biol. Evol. 2020;66:4.

741 39. Ogris E, Mudrak I, Mak E, Gibson D, Pallas DC. Catalytically inactive protein 742 phosphatase 2A can bind to polyomavirus middle tumor antigen and support complex 743 formation with pp60(c-src). J Virol. 1999;73:7390–8.

744 40. Nübel G, Sorgenfrei FA, Jäschke A. Boronate affinity electrophoresis for the 745 purification and analysis of cofactor-modified RNAs. Methods. 2017;117:14–20.

746 41. Igloi GL, Kössel H. Affinity electrophoresis for monitoring terminal 747 phosphorylation and the presence of queuosine in RNA. Application of 748 polyacrylamide containing a covalently bound boronic acid. Nucleic Acids Res. 749 1985;13:6881–98.

750 42. Furuichi Y. Discovery of m(7)G-cap in eukaryotic mRNAs. Proc Jpn Acad Ser B 751 Phys Biol Sci. 2015;91:394–409.

752 43. Kiledjian M. Eukaryotic RNA 5'-End NAD+ Capping and DeNADding. Trends 753 Cell Biol. 2018;28:454–64.

754 44. Wang J, Alvin Chew BL, Lai Y, Dong H, Xu L, Balamkundu S, et al. Quantifying 755 the RNA cap epitranscriptome reveals novel caps in cellular and viral RNA. Nucleic 756 Acids Res. 2019;47:e130.

757 45. Luciano DJ, Vasilyev N, Richards J, Serganov A, Belasco JG. A Novel RNA 758 Phosphorylation State Enables 5′ End-Dependent Degradation in Escherichia 759 coli. Mol Cell. Elsevier Inc; 2017;67:44–6.

760 46. Chen YG, Kowtoniuk WE, Agarwal I, Shen Y, Liu DR. LC/MS analysis of 761 cellular RNA reveals NAD-linked RNA. Nat Chem Biol. Nature Publishing Group; 762 2009;5:879–81.

763 47. Kowtoniuk WE, Shen Y, Heemstra JM, Agarwal I, Liu DR. A chemical screen for 764 biological small molecule-RNA conjugates reveals CoA-linked RNA. Proc Natl Acad 765 Sci USA. National Academy of Sciences; 2009;106:7768–73.

766 48. Clouet-d'Orval B, Batista M, Bouvier M, Quentin Y, Fichant G, Marchfelder A, et 767 al. Insights into RNA processing pathways and associated-RNA degrading enzymes 768 in Archaea. FEMS Microbiol Rev. 2018.

30 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

769 49. Kramer S, McLennan AG. The complex enzymology of mRNA decapping: 770 Enzymes of four classes cleave pyrophosphate bonds. WIREs RNA. 2019;10:e1511.

771 50. Dunckley T, Parker R. The DCP2 protein is required for mRNA decapping in 772 Saccharomyces cerevisiae and contains a functional MutT motif. EMBO J. 773 1999;18:5411–22.

774 51. Sharma S, Grudzien-Nogalska E, Hamilton K, Jiao X, Yang J, Tong L, et al. 775 Mammalian Nudix proteins cleave nucleotide metabolite caps on RNAs. Nucleic 776 Acids Res. 2020;48:6788–98.

777 52. Grudzien-Nogalska E, Kiledjian M. New insights into decapping enzymes and 778 selective mRNA decay. WIREs RNA. Wiley-Blackwell; 2017;8:e1379.

779 53. Grudzien-Nogalska E, Wu Y, Jiao X, Cui H, Mateyak MK, Hart RP, et al. 780 Structural and mechanistic basis of mammalian Nudt12 RNA deNADding. Nat Chem 781 Biol. 2019;15:575–82.

782 54. Deana A, Celesnik H, Belasco JG. The bacterial enzyme RppH triggers messenger 783 RNA degradation by 5' pyrophosphate removal. Nature. 2008;451:355–8.

784 55. Cahová H, Winz M-L, Höfer K, Nübel G, Jäschke A. NAD captureSeq indicates 785 NAD as a bacterial cap for a subset of regulatory RNAs. Nature. Nature Publishing 786 Group; 2015;519:374–7.

787 56. Sasaki M, Takegawa K, Kimura Y. Enzymatic characteristics of an ApaH-like 788 phosphatase, PrpA, and a diadenosine tetraphosphate hydrolase, ApaH, from 789 Myxococcus xanthus. FEBS Lett. Federation of European Biochemical Societies; 790 2014;588:3395–402.

791 57. McLennan AG. Substrate ambiguity among the nudix hydrolases: biologically 792 significant, evolutionary remnant, or both? Cell Mol Life Sci. 2012;70:373–85.

793 58. Bessman MJ, Walsh JD, Dunn CA, Swaminathan J, Weldon JE, Shen J. The gene 794 ygdP, associated with the invasiveness of Escherichia coli K1, designates a Nudix 795 hydrolase, Orf176, active on adenosine (5“)-pentaphospho-(5”)-adenosine (Ap5A). J 796 Biol Chem. American Society for Biochemistry and Molecular Biology; 797 2001;276:37834–8.

798 59. Abdelraheim SR, Spiller DG, McLennan AG. Mammalian NADH diphosphatases 799 of the Nudix family: cloning and characterization of the human peroxisomal NUDT12 800 protein. Biochem J. 2003;374:329–35.

801 60. Abdelraheim SR, Spiller DG, McLennan AG. Mouse Nudt13 is a Mitochondrial 802 Nudix Hydrolase with NAD(P)H Pyrophosphohydrolase Activity. Protein J. Springer 803 US; 2017;36:425–32.

804 61. van Dijk E, Cougot N, Meyer S, Babajko S, Wahle E, Séraphin B. Human Dcp2: a 805 catalytically active mRNA decapping enzyme located in specific cytoplasmic 806 structures. EMBO J. 2002;21:6915–24.

807 62. Xu J, Yang JY, Niu QW, Chua NH. Arabidopsis DCP2, DCP1, and VARICOSE

31 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

808 Form a Decapping Complex Required for Postembryonic Development. Plant Cell. 809 2006;18:3386–98.

810 63. Williams CW, Elmendorf HG. Identification and analysis of the RNA degrading 811 complexes and machinery of Giardia lamblia using an in silico approach. BMC 812 Genomics. BioMed Central; 2011;12:586–15.

813 64. Kumar S, Stecher G, Tamura K. MEGA7: Molecular Evolutionary Genetics 814 Analysis Version 7.0 for Bigger Datasets. Mol. Biol. Evol. 2016;33:1870–4.

815 65. Brun R, Schönenberger. Cultivation and in vitro cloning or procyclic culture 816 forms of Trypanosoma brucei in a semi-defined medium. Short communication. Acta 817 Trop. 1979;36:289–92.

818 66. McCulloch R, Vassella E, Burton P, Boshart M, Barry JD. Transformation of 819 monomorphic and pleomorphic Trypanosoma brucei. Methods Mol Biol. 820 2004;262:53–86.

821

822 FIGURE LEGENDS

823 Figure 1: Presence and absence of ALPH in the different eukaryotic subgroups

824 824 eukaryotic proteomes were screened for the presence of ALPH. Absence or

825 presence of ALPH is shown in pie diagrams in blue and orange, respectively, for each

826 phylogenetic group as indicated. The diameter of a circle roughly reflects the number

827 of available proteomes. Information for the phylogenetic tree was taken from [30]. All

828 organisms can be found in Supplementary Table S1.

829

830 Figure 2: General Features of ALPHs

831 (A) Proteomes with (orange) and without (blue) ALPHs.

832 (B) Number of ALPH proteins per organism.

833 (C) Sizes of the different ALPH ‘domains’ (N-terminus, catalytic domain, C-

834 terminus) are presented as box plot (waist is median; box is IQR; whiskers are ±1.5

835 IQR; only the smallest and largest outliers are shown). The catalytic domain is

836 defined as the range between the first and last motif, with an additional six N-terminal

837 amino acids and an additional eight C-terminal amino acids.

32 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

838 (D) Distances between the six different motifs (in amino acids) are presented as box

839 plot.

840 (E) Sequence motifs were created with WebLogo [31]. The most obvious differences

841 between the three groups are marked with orange bars.

842

843 Figure 3: ALPHs of Opisthokonts

844 ALPHs of Dikarya (A) and ALPHs of Holozoa (B) are presented with their catalytic

845 domain (dark blue) as well as further predicted sequence features, domains and

846 localisation predictions. The organism names for the Dikarya ALPHs can be found in

847 Supplementary Table S1c (number). More details on both Dikarya and Holozoa

848 ALPHs are listed in Supplementary Table S1c and S1d, respectively.

849

850 Figure 4: ALPHs of Diaphoretickes

851 ALPHs of Diaphoretickes are presented with their catalytic domain (dark blue),

852 further predicted sequence features and domains and localisation predictions. More

853 details are listed in Supplementary Table S1e.

854

855 Figure 5: ALPHs of Euglenozoa

856 (A) ALPHs of Euglenozoa are presented with their catalytic domain (dark blue) along

857 a maximum likelihood phylogenetic tree based on an alignment of gap corrected

858 sequences of ALPH catalytic domains. Different methods (minimal evolution,

859 neighbour joining) gave slightly different trees, but with all methods, the ALPH1

860 isoforms group together in one clade that never contains non-ALPH1 isoforms. All

861 details on Euglenozoa ALPHs are listed in Table S1f. The localisation of T. brucei

862 ALPH1 and ALPH2 was determined by expressing eYFP fusion proteins. One

33 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Eukaryotic ALPHs

863 representative fluorescent image of each cell line is shown. At least three different

864 clonal cell lines showed identical localisation.

865 (B) Sequence motifs of the ALPH1 orthologues and the Kinetoplastida orthologues to

866 other ALPHs. Major differences are marked by asterisks.

867

868 Figure 6: ALPH proteins have mRNA decapping activity

869 In vitro mRNA decapping assays were done with recombinant ALPH enzymes from

870 T. brucei (the decapping enzyme ALPH1 and the mitochondrial localised enzyme

871 ALPH2), from Sphaeroforma arctica (both the wild type and a catalytically inactive

872 mutant) and from Auxenochlorella protothecoides, using an m7G-capped RNA oligo

873 as a substrate. Note that a small portion of this oligo is uncapped even in the absence

874 of enzyme due to production issues (there is no auto-decapping activity during the

875 assay, data not shown). mRNA decapping activity was observed, as expected, with T.

876 brucei ALPH1 [10] and was fully absent in the catalytically inactive mutant of ALPH

877 from Sphaeroforma arctica. All three previously untested ALPH proteins had in vitro

878 decapping activity, albeit the ions requirements differed between the enzymes.

879

34 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Figure 1

Basidio- 25 Nucleariidae mycota number of proteomes and Fonticula 25 without ALPH (Fonticula alba) 1 47 Ascomycota 2 169 2 Blasto- 15 cladiales Dikarya 2 Chytridio- number of proteomes 7 15 133 mycetes (Branchiostoma with ALPH Mucoromycota 1 floridae) Opisthosporidia Zoopagomycota small circle: 1-19 (Cryptomycota) medium circle: 20-99 1 2 4 (Arachnida) 3 Chordata large circle: ≥100 Ecdysozoa

1 Nucletmycea 1 2 Cnidaria 8 Apusomonadida 140 Evosea Metazoa 1 1 1 Echinodermata Opisthokonta Holozoa 3 (Strongylocentrotus purpuratus) Discosea 8 Ichthyo- Lophotrochozoa Obazoa 2 sporea Amoebozoa 1 1 Placozoa 1 1 (Trichoplax) Choano- (Sphaeroforma Breviatea flagellida arctica)

Heterolo- Euglenida 1 (Euglena Gracilis) Amorphea Tsukubamonas bosea 1 (Naegleria gruberi) Diplonemea Discoba Euglenozoa Symbiontida 1(Bodo saltans: Kinetoplastea genome not Jakobida complete!) Diaphoretickes Metamonada Fornicata 2 Diplomonadida 31 Parabasalia (Trichomonas Preaxostyla 1 vaginalis) Cryptista 100 Cryptophyceae 1 (Guillardia theta) Embryophyta (land plants) CHLOROPLASTIDA 1 Haptista (Viridiplantae) Streptophyta Charophyceae Phaeo- phyta 1 (Klebsormidium Alveolata nitens) Klebsormidiophyceae Stramenopiles Rhizaria 1 (Heterokonts) 1 RHODOPHYCEAE 6 5 (red algae) Chlorophyceae Chromerida Chlorophyta 5 23 (Vitrella Mamiellophyceae 1 brassicaformis) 1 5 Apicomplexa Bangiophyceae Trebouxiophyceae Ciliata Florideophyceae 3 2 Dinoflagellata 20 1 1 (Gracilariopsis chorda) Figure 2

all Eukaryotes Diaphoretickes Discoba Amorphea

A number of proteomes number of proteomes number of proteomes number of proteomes bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this ALPH preprint (which500 was not certified324 by peer review) is the138 author/funder. All40 rights reserved.2 No reuse allowed32 without permission. 357 252 no ALPH 17 B 600 500 160 138 18 400 357 120 300 400 12 10 213 260 80 200 200 6 organisms organisms 30 organisms 4 organisms 49 40 2 100 34 9 5 1 5 5 1 4 1 0 0 0 0 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 number of ALPH number of ALPH number of ALPH number of ALPH

C 1000 1000 1000 1000 900 900 900 900 800 800 800 800 700 700 700 700 600 600 600 600 500 500 500 500 400 400 400 400 300 300 300 300

number of amino acids 200 number of amino acids 200 number of amino acids 200 number of amino acids 200 100 100 100 100 0 0 0 0 catalytic catalytic catalytic catalytic N-terminus C-terminus N-terminus C-terminus N-terminus C-terminus N-terminus C-terminus domain domain domain domain

D 231 349 243 231 349 200 200 200 200

160 160 160 160

120 120 120 120

80 80 80 80

40 40 40 40 number of amino acids number of amino acids number of amino acids number of amino acids 0 0 0 0 1 - 2 2 - 3 3 - 4 4 - 5 5 - 6 1 - 2 2 - 3 3 - 4 4 - 5 5 - 6 1 - 2 2 - 3 3 - 4 4 - 5 5 - 6 1 - 2 2 - 3 3 - 4 4 - 5 5 - 6 distance between motives distance between motives distance between motives E distance between motives bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. Figure 3 IPR013024 Gamma-glutamyl cyclotrans- Ubiquinol-cytochrome c chaperone ferase-like AIG2-like (IPR009288) A /UPF0174 (IPR021150) Ascomycota Basidiomycota WoLF WoLF 0 200 400 600 800 1000 1200 pSORT pSORT 0 200 400 600 800 1000 * SET domain * * * Peroxin-3 * * Pectate_lyase_SF_prot (IPR024535) 10 10

* second catalytic domain 20 * 20 Dothideomycetes

30 30 THIF-type NAD/FAD binding fold (IPR000594) * Agaricomycotina

THIF-type NAD/FAD binding fold (IPR000594) 40 * * 40 THIF-type NAD/FAD binding fold (IPR000594)

MLDKYNKYTKYPGPSPTYWQQPSPRPTKPVPKPPVDKCSFWMENIKHQGVAAFNPSPSTYQVFRNVKDFGAQGDGVTDDTAAINNAISSGGRCGPQTCGPGTTTTSPAVVYFPSGTYMVNASIIDYYYTQIIGNPNCLPIIRAFATFTGGLGVIDGDQYGAHGLGWGSTNKFWTQIRNFVIDMTLVPASSAITGIHWPTAQATSLQNIVFNMSDANGTQHQGFFIEEGSGGFMNDLTFYGGLNAITFGNQQFTVRNLTFYNAVTAINQIWDWGMTYYGISINNCSVGLNMAAGGPTAQSVGSMTFFDSSITDTPIGIITAHDATSRPLTAGSLIVENVVLNNVPIAIQGPGATTALAGTTGQTTITGWGEGHSYTPNGPVNFEGPIVPNNRPANLLAGNRYYARSKPQYQNYPVSSFVSARTSGATGNGHTDDTAALQHAIYAARAQNKVLYVDHGDYLVTSTIYVPSGSKIVGESYSVILSSGSFFNNVNKPQPVVQIGKNGESGSIEWSDMIVSTQGQQRGAVLFEYNLRSSSSSPSGIWDVHSRIGGFAGSNLGLAECPKTPTVNVTSANLNQNCIAAFMHFHVTKQSSGLYLENVWLWTADHDVEDPALTQITVYTGRGMLVESSGPVWLVGTAVEHNVLYEYQFVKASNVFGGQMQIESAYYQPNPPAPIPFPYVASLNDPQFPVATITADGYVIPASEGWGLRIVDSSNLLFYGVGLYSFFNNYSTACSNQGAGETCQNHIFSLEGRNSAISLYNLNTVGTHYQITIDGKEVAYYADNANGFISTRSSRDEKRRDDERRGLLIDRISNAWQDEKPAYNGNEDGELGCCDLENETSVTNILRSSRGKKMILYALLFVAALVYAWRWHVRPSLIEDCEFVEGFRLRQNGTYDVAKGGHWKHGGAKIQFLPKQLLPGSEGDAYGKRRLVFVGDIHGCRKELLSLLEAVHFDAANDHLIATGDVISKGPDSVGVLDELIRLHATSVRGNHEDRIIASAKSSDIASDDDAVYVSRSKKAKEDR* 50 50* Pucciniomycotina 60 Mucoromycota WoLF 0 200 400 600 800 Eurotiomycetes pSORT

70 Glomeromycotina Mortierellomycotina 10

80

THIF-type NAD/FAD binding fold (IPR000594) 20 * Spore coat protein CotH (IPR014867) * 90

Pezizomycotina THIF-type NAD/FAD binding fold (IPR000594) * Mucoromycotina

100 Zoopagomycota WoLF 0 200 400 600 pSORT

3 Entomophthoro- 110 6 mycotina Leotiomycetes Pezizomycetes *9 Kickxellomycotina

B WoLF 0 100 200 300 400 500 600 700 120 pSORT Sphaeroforma arctica JP610 Trichoplax sp. H2 Ichthyosporea Placozoa Helobdella robusta Lingula unguis 130 * Biomphalaria glabrata Mizuhopecten yessoensis Crassostrea gigas FAD/NAD(P)-binding Lophotrochozoa Elysia chlorotica domain superfamily Lottia gigantea (IPR036188) 140 Pomacea canaliculata Strongylocentrotus purpuratus * Echinodermata Pocillopora damicornis * Metazoa Cnidaria Cytochrome c oxidase Stylophora pistillata Tetranychus urticae Kaptin (IPR029982) * assembly protein PET191 Leptotrombidium deliense 150 (IPR018793) Stegodyphus mimosarum * Ecdysozoa

Sordariomycetes Dinothrombium tinctorium Branchiostoma floridae Chordata

Cytochrome c oxidase 160* assembly protein PET191 (IPR018793)

170 Domains/SP/TM WoLF pSORT Xylonomycetes (A=fungi, B=animals) signal peptide predicted (Phobius) cytoplasm 180 TM helix predicted (Phobius) extracellular catalytic domain (ALPH) plasma membrane disordered regions mitochondrion 190 S. cerevisiae Interpro domain /family nucleus * protein was tested by Interpro Scan golgi

Saccharomycetes peroxisome

Saccharomycotina 200 ER Taphrinomycotina bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Figure 4

WoLF amino acids 0 200 400 600 800 1000 1200 pSORT Alveolata Vitrella brassicaformis (strain CCMP3155) Gracilariopsis chorda Rhodophyta Galdieria sulphuraria Cyanidioschyzon merolae (strain 10D) * Cryptophyta Guillardia theta (strain CCMP2712) * Klebsormidium nitens Klebsormidiophyceae Streptophyta Tetradesmus obliquus_A0A383VP00 Tetradesmus obliquus_A0A383VP43 Transposase IS605 Raphidocelis subcapitata OrfB, C-terminal Chlamydomonas eustigma (IPR010095) Chlamydomonas reinhardtii * Chloro- Gonium pectorale phyceae Volvox carteri f. nagariensis Chlorella sorokiniana * Chlorophyta Auxenochlorella protothecoides_A0A087SP29 * Auxenochlorella protothecoides_A0A087SQ73 Trebouxio- Chlorella variabilis_E1ZSP5 * phyceae Chlorella variabilis_E1ZSM2 Chloroplastida Coccomyxa subellipsoidea (strain C-169)_I0YK19 Coccomyxa subellipsoidea (strain C-169)_I0YZ98 Helicosporidium sp. ATCC 50920 * Phytophthora parasitica P1569_V9F465 Coatomer delta subunit Phytophthora parasitica P1569_V9F0V2 (IPR027059) Phytophthora parasitica P10297_W2Z9Y7 Phytophthora parasitica P10297_W2Z6X4 Phytophthora parasitica CJ01A1_W2WWN6 Phytophthora parasitica CJ01A1_W2WX98 Phytophthora parasitica (strain INRA-310)_W2R9X1 Phytophthora parasitica (strain INRA-310)_W2R8D4 Phytophthora parasitica P1976_A0A081A4F4 Phytophthora parasitica P1976_A0A081A4F2 Phytophthora parasitica (strain INRA-310)_W2RAD0 Phytophthora parasitica CJ01A1_W2WWR3 Phytophthora parasitica P10297_W2Z6M6 Phytophthora parasitica P1569_V9F302 Phytophthora parasitica P1976_A0A081A4F3 Phytophthora cactorum Oomycetes Phytophthora megakarya Phytophthora ramorum ( ) Phytophthora sojae (strain P6497) Phytophthora kernoviae Peronospora effusa Bremia lactucae ( ) Aphanomyces astaci Aphanomyces invadans Stramenopiles Thraustotheca clavata Achlya hypogyna ( ) Saprolegnia parasitica ( ) Saprolegnia diclina (strain VS20) ( ) Pythium insidiosum Albugo candida Phaeodactylum tricornutum (strain CCAP 1055/1)_B7G322 Thalassiosira pseudonana Bacillariophyta Phaeodactylum tricornutum (strain CCAP 1055/1)_B7FSD5 * *Ectocarpus siliculosus * Phaeophyceae

Domains/SP/TM WoLF pSORT (plants) signal peptide predicted (Phobius) cytoplasm TM helix predicted (Phobius) mitochondrion catalytic domain (ALPH) nucleus disordered regions peroxisome Interpro domain /family chloroplast * protein was tested by Interpro Scan ( ) organism has no chloroplast bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. Figure 5

0 200 400 600 800 ALPH1-eYFP DAPI merged A L. aethiopica C 42 L. tropica 28 L. gerbilli 93 L. turanica L. arabica 41 L. major 80 L. mexicana 82 L. amazonensis P-bodies 35 L. infantum 95 L. donovani L. enriettii 73 30 L. tarentolae 100 L. panamensis 98 L. braziliensis 33 E. monterogeii posterior 5 µm 49 L. MAR LEM2494 poleole 99 C. fasciculata 85 L. pyrrhocoris 59 61 L. seymouri ALPH1 B. ayalai P. confusum (decapping) T. brucei brucei 98 100 T. evansi 83 T. brucei gambiense 80 T. congolense T. vivax 83 T. theileri T. grayi 88 T. rangeli 39 81 57 T. cruzi Dm28c (2017) T. cruzi marinkellei other 83 ALPHs 99 L. infantum 100 L. donovani ALPH2-eYFP mitotracker merged 45 L. panamensis 100 L. braziliensis 32 P. confusum T. grayi 100 T. cruzi Dm28c (2017) 91 T. cruzi marinkellei 91 58 T. rangeli T. theileri 24 T. vivax TvY486_0807310 T. vivax TvY486_0807440 79 45 T. congolense TcIL3000_8_8080 100 T. congolense TcIL3000_8_8210 42 T. congolense TcIL3000_4_3670 90 T. brucei_gambiense Tbg972.4.4420 72 100 T. evansi TevSTIB805.4.4510 T. brucei brucei ----- = ALPH2 88 T. brucei brucei Tb927.8.8040 = ALPH3 T. brucei gambiense Tbg972.8.8370 100 5 µm T. evansi TevSTIB805.8.8520 E. gracilis predicted transmembrane region N-terminally tagged ALPH2 localises to the glycosome (tryptag.org). 0.20

B phosphatase motifs ALPH motifs ALPH1

other ALPHs ** * bioRxiv preprint doi: https://doi.org/10.1101/2020.12.17.423368; this version posted December 18, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. Figure 6

no Trypanosoma Sphaeroforma Trypanosoma Auxenochlorella enzyme brucei (ALPH1) arctica brucei (ALPH2) protothecoides inactive mutant

Mg2+ Mg2+ Mn2+ Mn2+ Mn2+ Mg2+ Co2+ Zn2+ Mg2+ Mn2+

capped uncapped

decapping * (*) ***(*) activity