<<

JOURNAL OF PROTEOMICS 109 (2014) 276– 289

Available online at www.sciencedirect.com

ScienceDirect

www.elsevier.com/locate/jprot

Resolution of and -oxidation pathways of cuprina Ar-4 via comparative proteomics

Cheng-Ying Jianga, Li-Jun Liua, Xu Guoa, Xiao-Yan Youa, Shuang-Jiang Liua,c,⁎, Ansgar Poetschb,⁎⁎ aState Key Laboratory of Microbial Resources, Institute of Microbiology, Chinese Academy of Sciences, Beijing, PR China bPlant Biochemistry, Ruhr University Bochum, Bochum, Germany cEnvrionmental Microbiology and Biotechnology Research Center, Institute of Microbiology, Chinese Academy of Sciences, Beijing, PR China

ARTICLE INFO ABSTRACT

Article history: Metallosphaera cuprina is able to grow either heterotrophically on organics or autotrophically

Received 16 March 2014 on CO2 with reduced sulfur compounds as electron donor. These traits endowed the Accepted 6 July 2014 desirable for application in . In order to obtain a global overview of physiological Available online 14 July 2014 adaptations on the proteome level, proteomes of cytoplasmic and membrane fractions

from cells grown autotrophically on CO2 plus sulfur or heterotrophically on yeast extract Keywords: were compared. 169 were found to change their abundance depending on growth Quantitative proteomics condition. The proteins with increased abundance under autotrophic growth displayed

Bioleaching candidate /proteins of M. cuprina for fixing CO2 through the previously identified Autotrophy 3-hydroxypropionate/4-hydroxybutyrate cycle and for oxidizing elemental sulfur as energy Heterotrophy source. The main enzymes/proteins involved in semi- and non-phosphorylating Entner– Industrial microbiology Doudoroff (ED) pathway and TCA cycle were less abundant under autotrophic growth. Also some transporter proteins and proteins of metabolism changed their abundances, suggesting pivotal roles for growth under the respective conditions.

Abbreviations: ED pathway, Entner–Doudoroff pathway; EMP, Embden–Meyerhof–Parnas pathway; GA, glyceraldehyde; GAP, glyceralde- hyde 3-phosphate; KDG, 2-keto-3-deoxygluconate; KD(P)G, 2-keto-3-deoxy-(6-phospho)-gluconate; PEP, phosphoenolpyruvate; GDH, dehydrogenase; GAD, gluconate ; GAOR, glyceraldehyde ; GADH, glyceraldehyde dehydratase; GAPDH, phosphorylating GAP dehydrogenase; GAPN, non-phosphorylating GAP, dehydrogenase; GAPOR, GAP oxidoreductase; PGK, phosphoglyc- erate ; PGM, phosphoglycerate mutase; FBPA/ase, fructose 1,6-bisphosphate aldolase/bisphosphatase; FBPA, fructose 1,6-bisphosphate aldolase; TIM, triosephosphate ; BPG, 2,3-bisphosphoglycerate; PGI/PMI, bifunctional phosphoglucose/phosphomannose isomer- ase; PGM/PMM, phosphoglucomutase/phosphomannomutase; 3-HP/4-HB, 3-hydroxypropionyl/4-hydroxybutyryl; ACC/PCC, acetyl-CoA carboxylase/propionyl-CoA carboxylase; MSR, malonate semialdehyde reductase; 4-HBCS, 4-hydroxybutyryl-CoA synthetase; 4-HBCD, 4-hydroxybutyryl-CoA dehydratase; SSR, succinic semialdehyde reductase; ACK, acetoacetyl-CoA β-ketothiolase; 3-HPCD, 3-hydroxypropionyl- CoA dehydratase; ACR, acryloyl-CoA reductase; MSR, malonate semialdehyde reductase; 3-HPCS, 3-hydroxypropionyl-CoA synthetase (AMP-forming); MCE, methylmalonyl-CoA epimerase; MCM, methylmalonyl-CoA mutase; CCH, crotonyl-CoA hydratase; SOR, sulfur oxidoreductase; SQR, quinone oxidoreductase; NSR, NADPH:sulfur oxidoreductase; SR, sulfite reductase; SAOR, sulfite:acceptor oxidoreductase; TQO, thiosulfate:quinone oxidoreductase; TetH, tetrathionate ; Hdr, heterodisulfide reductase; APS, adenosine-5′- phosphosulfate; APSR, adenosine-5′-phosphosulfate reductase; APAT, adenylylsulfate:phosphate adenylyltransferase; AK, adenylate kinase. ⁎ Correspondence to: S.-J. Liu, Institute of Microbiology, Chinese Academy of Sciences, Beijing 100101, PR China. Tel.: +86 10 64847423; fax: +86 10 64807421. ⁎⁎ Corresponding author. E-mail addresses: [email protected] (S.-J. Liu), [email protected] (A. Poetsch).

http://dx.doi.org/10.1016/j.jprot.2014.07.004 1874-3919/© 2014 Elsevier B.V. All rights reserved. JOURNAL OF PROTEOMICS 109 (2014) 276– 289 277

Biological significance The described work is of great significance: For general microbiology:

• How do extremophile use their unique metabolic capabilities in adapting to autotrophic and hetetrotrophic growth conditions? • Which are important enzymes involved in the metabolic adaptation and which candidateshouldbeinvestigatedinmoredetailwith microbiological/biochemical approaches?

For applied microbiology:

• Which are the key enzymes and reaction pathways for sulfur oxidation and autotrophic growth? This knowledge should accelerate future design of improved bioleaching processes in biomining industries or bioremediation.

© 2014 Elsevier B.V. All rights reserved.

1. Introduction at an optimal temperature of 65 °C and pH of 3.5. Strain Ar-4 grows chemolithoautotrophically on CO2 by oxidation of Extremely thermoacidophilic from the order elemental sulfur, ferrous , tetrathionate and , it , including , ,and also grows chemoorganoheterotrophically on various organic Metallosphaera, contribute greatly to biogeochemical element compounds such as yeast extract, amino acids, and in con- cycling and biohydrometallurgical processes with their physio- trast to M. sedula, D-glucose and few other monosaccharides logical versatility. So far, the most extensively studied model [8]. The M. cuprina genome is 16% smaller than the M. sedula for Crenarchaeotes has been . The genome, however, it harbors 233 ORFs that do not occur in complete arsenal of systems biology (genomics, transcriptomics, M. sedula[10]. – proteomics, metabolomics) together with modeling tools had Since 2D E is limited in its proteome coverage (membrane been used to understand its physiology under various condi- & alkalic proteins), we selected quantitative shotgun proteo- tions [1].However,Sulfolobus spp. lack or easily lost many mics instead to obtain a comprehensive and comparative physiological features, such as the iron or sulfur-oxidizing view on the physiology of M. cuprina under autotrophic and ability, exploited in biomining industries under autotrophic organotrophic growth. This first large-scale proteomics study growth condition [2]. Metallosphaera spp. stand out from for the genus Metallosphaera enabled a detailed description of Sulfolobales by their ability to grow heterotrophically on differences in central carbon metabolism, enzymes of the 3-HP/4-HB cycle, and oxidation pathway of RISCs. organics, and autotrophically by fixing CO2 with hydrogen, ferrous or reduced inorganic sulfur compounds (RISCs) as electronic donor or directly mobilizing metals from metal . The availability of the genome 2. Materials and methods sequence has enabled -omics studies, which mainly aimed M. cuprina at further elucidation of the CO2 fixation pathway and the 2.1. Cultivation of Ar-4 . DNA microarray analysis of M. sedula grown on yeast extract supplemented with different sulfur M. cuprina Ar-4 was cultivated aerobically at 65 °C in base salt compounds was used to identify the enzymes involved in medium (BSM) [10], supplemented with 5 g/L elemental sulfur iron and sulfur oxidation [3]. In addition autotrophic, hetero- (for autotrophic growth) or 2 g/L yeast extract (for heterotrophic trophic and mixotrophic growths were characterized [4] to growth). Autotrophic cultures were shaked three times every day narrow-down enzyme candidates for the 3-hydroxypropionate/ to maintain micro amount of CO2 in medium. Heterotrophic 4-hydroxybutyrate (3-HP/4-HB) cycle. Cultivation under strict culture was carried out without agitation. For proteome analysis, – H2 CO2 autotrophy was realized by growing M. sedula in a four 1-L flasks with aseptic filtration membrane, each containing bioreactor and transcriptomics together with biochemical 300-mL of BSM medium, were used. Cells were continuously assays was used to conclude on the physiologically relevant cultivated either autotrophically or heterotrophically for at least 4-hydroxybutyrate-CoA (4-HB-CoA) synthetase [5].Theobserved three transfers/generations in 100 mL (250 mL flask) before ino- negative impact of hydrogen on bioleaching efficiency of M. culated into 1-L flask. Autotrophically or heterotrophically grown sedula was confirmed at molecular level with transcriptomics [6]. cells were harvested in the late exponential growth phase (with 7– 7 – Metabolic flux analyses revealed interesting facts about carbon cell densities of 4 × 10 6×10/mL (OD600 =0.06 0.08) for auto- 8– 8 – metabolism, notably succinyl-CoA and not acetyl-CoA as main trophic cells and 2 × 10 4×10/mL (OD600 =0.3 0.5)), Fig. S1. anabolic precursor [7]. So far only once, proteomics, a compar- ison of M. sedula strains with high and low copper tolerance 2.2. Preparation of proteome fractions (2D-electrophoresis), was employed [8]. Metallosphaera cuprina Ar-4 was isolated from the muddy Preparation of cytoplasmic and membrane proteomes was water samples of a sulfuric [9]. It grows aerobically performed according to Haussmann et al. [11].Briefly,M. cuprina 278 JOURNAL OF PROTEOMICS 109 (2014) 276– 289

Ar-4 cells were harvested by centrifugation for 15 min at 8000 ×g. loaded onto a triphasic microcapillary column as described Cells were washed twice with BSM broth (pH 3.0) and with [12]; MudPIT analysis was conducted by using an Accela phosphate-buffered saline (PBS, pH 7.4) [11]. The washed cells gradient HPLC pump system (Thermo Electron) and flow- were resuspended in disintegration buffer (PBS containing addi- splitter coupled to an LTQ Orbitrap mass spectrometer tional 20 mmol/L MgCl2, 10 mmol/L MnCl2, 200 U/mL DNase I, (Thermo Electron) as described in [11,12]. protease inhibitor mixture for bacterial cells (Sigma, St. Louis, MO, USA, 1 mL for 6 g wet cells)) at a concentration of 6 mL of buffer/g 2.6. identification and quantification of wet cells. The cells were disrupted with a French Pressure Cell (40 K cell with a volume of 35 mL, Thermo Spectronic, Rochester, MS/MS data interpretations were performed using SEQUEST USA) with 4 passages at 40,000 psi. Cell debris and unbroken cells algorithm, embedded in Bioworks™ (Rev. 3.3, Thermo Electron & were precipitated twice by centrifugation for 15 min at 12,000 ×g 1998–2006), with a M. cuprina Ar-4 genomic database obtained and 4 °C. Membranes were enriched by 30 min of ultracentrifu- from EBI proteomes containing 2029 protein entries [10]. The gation at 100,000 ×g and 4 °C. The resulting supernatants, which selected enzyme specificities were different for predigest and the contained cytoplasmic proteins, were stored at −20 °C. The SIMPLE samples. For predigest samples only fully tryptic pellets were resuspended gently with ice-cold PBS and ultra- peptides with up to two missed cleavages were allowed, while centrifuged again. Membrane fraction was stored in PBS with 10% for SIMPLE samples no enzyme specificity was considered. All glycerol and protease inhibitors at −70 °C. other search parameters were: alkylation for cytoplas- mic proteins, no fixed modifications for membrane proteins, and 2.3. Digestion of membranes oxidation of was permitted as variable modification. The mass tolerance for precursor ions was set to 10 ppm; the The membranes with 300 μgproteincontentofM. cuprina Ar-4 mass tolerance for fragment ions was set to 1 amu. The (grown on 5 g/L sulfur or 2 g/L yeast extract) were predigested at threshold for both protein and peptide probability was set to 37 °C with trypsin (1:100 w/w protein) overnight in a 25 mmol/L 0.001, and at least two different peptides per protein were bicarbonate buffer (pH 8.6), then pelleted at accepted for protein identification. The false discovery rate (FDR) 100,000 ×g and 4 °C. The supernatants were stored at −70 °C for was estimated with a reversed database generated in Bioworks™ later processing. The pellets were processed further according to as described [13]. The FDR for protein identification was below the specific integral membrane peptide level enrichment 1% for all predigest, SIMPLE and cytoplasmic protein measure- (SIMPLE) protocol [12]; they were washed twice with ice-cold ments. Counting of MS spectra was used for relative protein deionized water and resuspended in 60% methanol and 40% quantification and carried out as described [13].Thelog2ofthe 25 mmol/L ammonium bicarbonate buffer (pH 8.6) by sonication individual spectral counts was calculated and the regulation for 10 min twice. Subsequently, trypsin and chymotrypsin (each factor of different proteins for autotrophy to heterotrophy was 1:100, w/w protein) were added to the sample for overnight definedastheaverageofsubtractionofthelog2ofthetwo digestion at 60 °C (SIMPLE). The peptides of predigest with proteins. The results were also statistically analyzed by student's trypsin and SIMPLE digest with trypsin/chymotrypsin were each t-test. Proteins were regarded significantly regulated, if the log2 desalted using Spec PT C18 AR solid phase extraction pipette tips regulation factor was ≥2.0andp-valuewas≤0.05. The mass (Varian, Lake Forest, CA, USA). Products of both digestion steps spectrometry proteomics data have been deposited to the were analyzed by LC-MS later. ProteomeXchange Consortium (http://www.proteomexchange. org) via the PRIDE partner repository with the dataset identifier 2.4. Precipitation and digestion of cytoplasmic proteins PXD000676[14].

Cytoplasmic proteins (supernatants obtained after disruption of the cells) were precipitated according to the protocol of Niessen 3. Results and discussion et al. [13]. At first, 1/4 volume of trichloroacetic acid was added to cytoplasmic protein (in PBS disruption buffer), then placed on ice 3.1. General survey of proteome dataset for 30 min. Upon acetone-washing and air-drying the pelleted protein was dissolved in 500 μL of 8 mol/L urea (in 100 mmol/L According to our chosen criteria, a total of 848 of the 2029 Tris–HCl, pH = 8.0) and incubated at 30 °C overnight. Proteins predicted proteins for M. cuprina Ar-4 [10] were identified and were reduced with Tris (2-carboxyethyl)-phosphine hydrochlo- mass spectrometry results deposited at ProteomeXchange; 545 ride and alkylated with iodoacetamide. The samples proteins were identified under both chemolithoautotrophic were diluted to 2 mol/L urea by adding 100 mmol/L Tris–HCl (sulfur) and chemoorganoheterotrophic (yeast extract) growth, buffer (pH 8.0), then 1 mol/L CaCl2 was added to a final concen- while 103 and 200 proteins were detected exclusively under tration of 2 mmol/L. The proteins were digested overnight with chemolithoautotrophic or chemoorganoheterotrophic growth. trypsin and desalted using Spec PT C18 AR solid phase extraction In total, 74 (21.8% of all 338 predicted) integral membrane pipette tips. proteins and 12 (37.5% of 32 predicted) secreted proteins were identified. 2.5. Multi-dimensional protein identification technology Out of all the 848 identified proteins, 169 changed their (MudPIT) abundances significantly. In the clusters of orthologous groups of proteins' (COGs) assignment, the proteomes covered all COG The desalted samples of tryptic predigestion, SIMPLE diges- categories (see Fig. S2) except of RNA processing & modification. tion and cytoplasmic protein digestion, respectively, were For about half of all proteins, only general function or no JOURNAL OF PROTEOMICS 109 (2014) 276– 289 279

Table 1 – Proteins with significantly changed abundance in the different cellular fractions/digests for cells growing autotrophically on sulfur or heterotrophically on yeast extract.

ID Description Log2 ratio t-Test TM/SP COG auto/hetero

cyto pre mem

Cytoplasm Mcup_0018 Adenylosuccinate −4.45 0.028 F Mcup_0032 DNA-directed RNA polymerase subunit H 4.73 0.013 K Mcup_0070 -independent phosphoglycerate mutase −5.04 0.010 G Mcup_0137 Tyrosyl-tRNA synthetase −5.33 0.016 J Mcup_0142 30S ribosomal protein S4P −4.15 0.003 J Mcup_0157 Glutamyl-tRNA synthetase −3.89 0.002 J Mcup_0163 Metal-dependent hydrolase −4.45 0.002 R Mcup_0209 NADPH:sulfur oxidoreductase −4.72 0.018 TMH R Mcup_0244 Conserved hypothetical protein Msed_2042 5.21 0.000 L Mcup_0253 Proteasome endopeptidase complex −5.79 0.006 O Mcup_0280 Peptidase U62, modulator of DNA gyrase −4.79 0.021 R Mcup_0294 Haloacid dehalogenase superfamily protein −3.80 0.001 J Mcup_0360 amidotransferase subunit PdxT −4.05 0.007 H Mcup_0368 V-type ATP synthase subunit A 2.39 0.011 C Mcup_0455 4-Hydroxyphenylacetate 3-hydroxylase −4.17 0.000 Q Mcup_0469 6-Phosphogluconate dehydrogenase, NAD-binding −4.87 0.010 I Mcup_0506 C/D box methylation guide ribonucleoprotein complex aNOP56 subunit −4.52 0.000 J Mcup_0524 Ornithine carbamoyltransferase −4.43 0.014 E Mcup_0533 Phosphoenolpyruvate synthase −1.77 0.001 G Mcup_0541 Anthranilate synthase component I −4.24 0.001 E Mcup_0565 cleavage system aminomethyltransferase T −4.17 0.000 E Mcup_0593 Acidic ribosomal protein P0 1.76 0.031 J Mcup_0594 50S ribosomal protein L1P −6.16 0.017 J Mcup_0596 Transcription antitermination protein NusG −4.09 0.000 K Mcup_0604 Conserved hypothetical protein Msed_1625 6.58 0.001 R Mcup_0605 30S ribosomal protein S19e −4.90 0.006 J Mcup_0607 SpoVT/AbrB -containing protein −4.77 0.012 P Mcup_0617 L-Glutamine synthetase 2.04 0.038 E Mcup_0650 Succinyl-CoA synthetase (ADP-forming) beta subunit −6.80 0.003 C Mcup_0676 Conserved hypothetical protein Msed_1555 −7.63 0.002 S Mcup_0752 Aconitate hydratase −6.06 0.000 TMH C Mcup_0817 Alpha-amylase −4.86 0.000 G Mcup_0818 Starch synthase −4.86 0.000 G Mcup_0823 Pyruvate kinase −5.04 0.006 G Mcup_0836 CoA-binding domain-containing protein −4.28 0.018 C Mcup_0848 Xanthine and Co dehydrogenase maturation factor −6.95 0.000 O Mcup_0891 Formate dehydrogenase, alpha subunit −4.25 0.000 R Mcup_0900 Gluconate dehydratase/galactonate dehydratase −4.25 0.000! M Mcup_0902 1-dehydrogenase/glucose 1-dehydrogenase −5.24 0.013 E Mcup_0904 Peptidase S9 prolyl oligopeptidase −6.30 0.014 E Mcup_0949 2-Keto-3-deoxygalactonate aldolase/2-keto-3-deoxy-phosphogalactonate aldolase/ −5.20 0.001 E 2-keto-3-deoxy-phosphogluconate aldolase/2-keto-3-deoxygluconate aldolase Mcup_0957 Extracellular solute-binding protein −5.28 0.000 EXC G Mcup_1023 Signal-transduction protein −5.36 0.000 T Mcup_1025 FAD dependent oxidoreductase −5.04 0.000 C Mcup_1026 Electron transfer flavoprotein, alpha subunit 6.29 0.000 C Mcup_1136 Conserved hypothetical protein Msed_1166 −5.28 0.001 L Mcup_1137 Conserved hypothetical protein Msed_1165 −5.75 0.000 L Mcup_1139 Conserved hypothetical protein Msed_1163 −5.58 0.000 TMH R Mcup_1143 CRISPR-associated Csx7 family protein −5.28 0.004 L Mcup_1152 CRISPR-associated Cas5 family protein −6.56 0.009 L Mcup_1171 Linocin_M18 bacteriocin protein −4.58 0.000 S Mcup_1191 Conserved hypothetical protein Msed_0991 −5.28 0.000 O Mcup_1210 TreZ −4.47 0.007 G Mcup_1211 Glycogen debranching enzyme −5.98 0.016 G Mcup_1216 Alcohol dehydrogenase −4.28 0.007 R Mcup_1244 Carbohydrate kinase, YjeF related protein −6.11 0.008 G Mcup_1257 Xanthine dehydrogenase, molybdenum binding subunit apoprotein −6.79 0.031 C Mcup_1260 Conserved hypothetical protein Msed_0785 −4.47 0.007 S

(continued on next page) 280 JOURNAL OF PROTEOMICS 109 (2014) 276– 289

Table 1 (continued)

ID Description Log2 ratio t-Test TM/SP COG auto/hetero

cyto pre mem

Cytoplasm Mcup_1298 Band 7 protein −3.77 0.250 O Mcup_1386 ATPase −4.23 0.015 TMH R Mcup_1450 Leucyl-tRNA synthetase −5.44 0.008 J Mcup_1453 Conserved hypothetical protein Msed_0579 −5.63 0.004 J Mcup_1492 2Fe–2S iron–sulfur cluster binding domain-containing protein −4.99 0.044 C Mcup_1515 Phosphoribosyltransferase 4.50 0.024 R Mcup_1518 Heat shock protein Hsp20 −5.36 0.023 O Mcup_1563 Signal-transduction protein −4.28 0.007 R Mcup_1569 Pyruvate flavodoxin/ferredoxin oxidoreductase domain-containing protein −4.84 0.015 C Mcup_1599 Amidohydrolase 2.08 0.005 R Mcup_1654 Acetolactate synthase catalytic subunit −6.20 0.025 E Mcup_1676 Peptidase U62, modulator of DNA gyrase −5.14 0.003 R Mcup_1679 Acyl-CoA dehydrogenase domain-containing protein −5.35 0.017 I Mcup_1680 3-Hydroxyacyl-CoA dehydrogenase, NAD-binding −4.08 0.002 I Mcup_1685 AMP-dependent synthetase and −7.85 0.007 I Mcup_1687 −6.69 0.010 P Mcup_1688 3-Hydroxyacyl-CoA dehydrogenase, NAD-binding −4.53 0.032 I Mcup_1689 Phenylacetate-CoA ligase 7.00 0.006 TMH H Mcup_1692 Enoyl-CoA hydratase −4.93 0.003 I Mcup_1693 Short chain enoyl-CoA hydratase −4.75 0.001 I Mcup_1708 Delta-1-pyrroline-5-carboxylate dehydrogenase −5.14 0.029 C Mcup_1725 Conserved hypothetical protein Msed_0351 −4.72 0.005 S Mcup_1736 FAD dependent oxidoreductase −1.02 0.044 C Mcup_1789 2,3-Dimethylmalate lyase −6.03 0.001 G Mcup_1792 Acyl-CoA dehydrogenase domain-containing protein −4.73 0.029 I Mcup_1793 4-Aminobutyrate aminotransferase −5.83 0.000 E Mcup_1797 Acyl-CoA dehydrogenase domain-containing protein −6.17 0.010 I Mcup_1798 Acetyl-CoA acetyltransferase −6.63 0.005 I Mcup_1799 Acetyl-CoA acetyltransferase −4.00 0.001 I Mcup_1857 Delta-aminolevulinic acid dehydratase −5.02 0.003 H Mcup_1877 Alcohol dehydrogenase −4.84 0.008 E Mcup_1921 Conserved hypothetical protein Msed_0152 −4.84 0.008 T Mcup_1925 Biotin/lipoyl attachment domain-containing protein −4.32 0.000 C Mcup_1929 Glycosyl , group 1 6.06 0.016 M Mcup_1947 30S ribosomal protein S5P 7.87 0.000 J Mcup_1954 50S ribosomal protein L5P −3.91 #DIV/0! J Mcup_1955 30S ribosomal protein S4e −4.00 0.001 J Mcup_1965 50S ribosomal protein L23P −5.01 0.045 J Mcup_1979 Exosome complex RNA-binding protein Rrp42 −5.15 0.015 J Mcup_1998 Spermine synthase/spermidine synthase 6.32 0.003 E Mcup_2030 Ribose-5-phosphate isomerase 2.33 0.028 G

Pre-digest Mcup_0010 triad (HIT) protein 3.86 0.001 F Mcup_0017 5-Formaminoimidazole-4-carboxamide-1-(beta)-D-ribofuranosyl 4.97 0.011 R 5′-monophosphate synthetase Mcup_0018 Adenylosuccinate lyase 3.99 0.004 F Mcup_0031 Malate dehydrogenase 4.76 0.001 C Mcup_0040 30S ribosomal protein S7P −4.20 0.005 J Mcup_0070 Cofactor-independent phosphoglycerate mutase −1.33 0.011 G Mcup_0090 Conserved hypothetical protein Msed_2216 5.30 0.005 P Mcup_0141 30S ribosomal protein S13P 4.74 0.012 J Mcup_0214 Glu/Leu/Phe/Val dehydrogenase, C terminal −3.95 0.041 E Mcup_0232 Alpha/beta fold family hydrolase/acetyltransferase-like protein 4.06 0.009 R Mcup_0284 Homoserine dehydrogenase 4.33 0.007 E Mcup_0293 3-Hydroxyacyl-CoA dehydrogenase, NAD-binding 2.89 0.011 I Mcup_0297 Carbamoyl phosphate synthase small subunit 4.20 0.042 E Mcup_0298 Argininosuccinate lyase 6.22 0.001 E Mcup_0299 Argininosuccinate synthase 5.02 0.002 E Mcup_0308 Phosphoribosylaminoimidazole-succinocarboxamide synthase 5.87 0.000 F JOURNAL OF PROTEOMICS 109 (2014) 276– 289 281

Table 1 (continued)

ID Description Log2 ratio t-Test TM/SP COG auto/hetero

cyto pre mem

Pre-digest Mcup_0360 Glutamine amidotransferase subunit PdxT −5.05 0.000 H Mcup_0367 V-type ATP synthase subunit B 1.07 0.021 C Mcup_0382 NADH dehydrogenase (quinone) 4.06 0.000 TMH C Mcup_0429 Conserved hypothetical protein Msed_1806 −3.24 0.004 EXC O Mcup_0456 3,4-Dihydroxyphenylacetate 2,3-dioxygenase −7.26 0.000 R Mcup_0469 6-Phosphogluconate dehydrogenase, NAD-binding 4.54 0.015 I Mcup_0483 Conserved hypothetical protein Msed_1746 5.20 0.012 S Mcup_0489 Aspartyl/glutamyl-tRNA amidotransferase subunit A 4.20 0.042 J Mcup_0506 C/D box methylation guide ribonucleoprotein complex aNOP56 subunit −1.66 0.029 J Mcup_0537 synthase subunit beta −5.33 0.001 R Mcup_0539 Anthranilate phosphoribosyltransferase 4.19 0.014 E Mcup_0544 Aspartate aminotransferase 4.00 0.029 E Mcup_0554 Aminotransferase, class V 1.15 0.015 E Mcup_0582 Acetyl-CoA acetyltransferase-like protein −5.01 0.000 I Mcup_0593 Acidic ribosomal protein P0 1.25 0.024 J Mcup_0680 Dihydrolipoamide dehydrogenase 3.87 0.021 C Mcup_0686 FAD-dependent pyridine -disulfide oxidoreductase 2.02 0.050 C Mcup_0744 AMP-dependent synthetase and ligase 6.37 0.022 I Mcup_0762 DSBA oxidoreductase 4.33 0.001 Q Mcup_0781 Na+- solute symporter 5.09 0.001 TMH E Mcup_0811 Alcohol dehydrogenase 4.66 0.008 C Mcup_0813 Coenzyme F390 synthetase-like protein 2.91 0.033 H Mcup_0835 synthase −4.57 0.011 E Mcup_0858 Propionyl-CoA carboxylase carboxyltransferase subunit alpha/propionyl-CoA 8.40 0.000 I carboxylase carboxyltransferase subunit beta Mcup_0888 Vinylacetyl-CoA Delta-isomerase 6.35 0.002 Q Mcup_0891 Formate dehydrogenase, alpha subunit −5.81 0.004 R Mcup_0892 Conserved hypothetical protein Msed_1317 6.08 0.003 V Mcup_0958 Binding-protein-dependent transport systems inner membrane component 5.18 0.009 G Mcup_0959 Binding-protein-dependent transport systems inner membrane component 4.67 0.002 TMH G Mcup_0979 Aldehyde oxidase and xanthine dehydrogenase, molybdopterin binding −1.98 0.024 TMH C Mcup_1006 proton symporter 5.48 0.003 TMH G Mcup_1025 FAD dependent oxidoreductase −6.18 0.000 C Mcup_1136 Conserved hypothetical protein Msed_1166 4.29 0.020 L Mcup_1139 Conserved hypothetical protein Msed_1163 5.17 0.002 TMH R Mcup_1151 CRISPR-associated autoregulator DevR family protein −6.08 0.002 L Mcup_1171 Linocin_M18 bacteriocin protein −5.21 0.024 S Mcup_1211 Glycogen debranching enzyme −6.16 0.002 G Mcup_1254 Phosphoribosyltransferase 4.33 0.007 R Mcup_1316 Radical SAM domain-containing protein 3.86 0.001 H Mcup_1370 FAD linked oxidase domain-containing protein 1.57 0.043 C Mcup_1371 Histidine triad (HIT) protein 4.56 0.001 F Mcup_1437 Acetyl-CoA acetyltransferase 5.14 0.010 I Mcup_1529 Type II secretion system protein 5.47 0.015 TMH N Mcup_1572 Conserved hypothetical protein Msed_0522 4.53 0.000 TMH S Mcup_1623 Sugar transporter 4.99 0.012 TMH G Mcup_1644 Binding-protein-dependent transport systems inner membrane component 3.72 0.012 TMH E Mcup_1654 Acetolactate synthase catalytic subunit −5.70 0.002 E Mcup_1679 Acyl-CoA dehydrogenase domain-containing protein −2.45 0.023 I Mcup_1680 3-Hydroxyacyl-CoA dehydrogenase, NAD-binding −1.98 0.017 I Mcup_1683 Acetyl-CoA acetyltransferase −1.92 0.022 I Mcup_1687 Carbonic anhydrase 5.09 0.001 P Mcup_1691 Acetyl-CoA acetyltransferase −5.80 0.000 I Mcup_1712 TQO small subunit DoxA domain-containing protein 5.00 0.023 EXC O Mcup_1713 TQO small subunit DoxD 6.14 0.000 TMH S Mcup_1717 Alkyl hydroperoxide reductase/thiol specific antioxidant/Mal allergen −3.89 0.042 O Mcup_1772 Rubrerythrin −4.85 0.015 C Mcup_1785 Conserved hypothetical protein Msed_0283 5.36 0.003 TMH S Mcup_1790 2-Methylcitrate dehydratase −4.85 0.005 R Mcup_1801 Conserved hypothetical protein Msed_0268 −5.95 0.018 R

(continued on next page) 282 JOURNAL OF PROTEOMICS 109 (2014) 276– 289

Table 1 (continued)

ID Description Log2 ratio t-Test TM/SP COG auto/hetero

cyto pre mem

Pre-digest Mcup_2017 Conserved hypothetical protein Msed_2271 −4.44 0.000 P Mcup_2027 Exonuclease-like protein 3.87 0.021 L

Main digest Mcup_0202 Acyl-CoA dehydrogenase domain-containing protein 8.18 0.005 I Mcup_0293 3-Hydroxyacyl-CoA dehydrogenase, NAD-binding 5.11 0.000 I Mcup_0366 V-type ATP synthase subunit D 4.44 0.027 C Mcup_0858 Propionyl-CoA carboxylase carboxyltransferase subunit alpha/propionyl-CoA 6.88 0.020 carboxylase carboxyltransferase subunit beta Mcup_0959 Binding-protein-dependent transport systems inner membrane component 4.45 0.046 TMH G Mcup_1006 Sugar proton symporter 6.32 0.006 TMH G Mcup_1185 Conserved hypothetical protein Msed_0990 −4.03 0.005 TMH S Mcup_1216 Alcohol dehydrogenase −8.81 0.006 R Mcup_1305 Aminotransferase, class V −3.88 0.005 E Mcup_1322 Peptidase S8/S53 subtilisin kexin sedolisin −3.88 0.001 O Mcup_1563 Signal-transduction protein 4.76 0.020 R Mcup_1569 Pyruvate flavodoxin/ferredoxin oxidoreductase domain-containing protein −4.46 0.010 C Mcup_1609 Quinol oxidase polypeptide I 6.96 0.000 TMH C Mcup_1627 Extracellular solute-binding protein −5.96 0.029 TMH G Mcup_1713 TQO small subunit DoxD 4.71 0.003 TMH S Mcup_1717 Alkyl hydroperoxide reductase/thiol specific antioxidant/Mal allergen −4.63 0.018 O Mcup_1725 Conserved hypothetical protein Msed_0351 7.40 0.001 S Mcup_1729 Conserved hypothetical protein Msed_0347 −4.74 0.001 S Mcup_1792 Acyl-CoA dehydrogenase domain-containing protein −4.26 0.014 I Mcup_1926 Carbamoyl-phosphate synthase L chain, ATP-binding 7.03 0.018 I

function at all could be predicted, which is not surprising in view described for the genus Metallosphaera so far. According to of the limited knowledge about Archaea protein functions from our shotgun proteomics data, most of the enzymes related genomes. The three best-covered COGs with function prediction to ED pathway were expressed at least 2 folds lower under were 1) energy production & conversion, 2) amino acid transport autotrophic versus heterotrophic conditions (Fig. 1), in- & metabolism, and 3) translation, ribosomal structure & cluding glucose dehydrogenase (GDH, Mcup_0902), gluco- biogenesis. nate dehydratase (GAD, Mcup_0900), and glycerate kinase Proteome comparison was achieved by relative quantifi- (Mcup_0274). Function of these enzymes has been con- cation using the spectral counting technique. Proteins that firmed for other hyperthermophilic Archaea, i.e. GDH and display significantly changed abundance are listed in Table 1. GAD in tenax [21–23],glyceratekinasein In the following, detailed results of the proteome comparison , S. solfataricus and T. tenax [24]. Similar to will be presented with a focus on carbon metabolism, sulfur the members of Sulfolobus, Metallosphaera did not contain oxidation and metabolite transport. homologs to glyceraldehyde dehydratase (GADH) in Picrophilus torridus (pto0332)andThermoplasma acidophilum 3.2. Central carbon metabolism (ta0809) [25], indicating that Sulfolobales harbor different enzymes for ED pathway. In S. acidocaldarius and S. (i) Entner–Doudoroff (ED) pathway: solfataricus, GA (GAOR: Saci_2269, 2270, Glycolysis and gluconeogenesis pathways are widely 2271 and SSO2636, 2637, 2639) were thought to oxidize GA distributed in Bacteria, Eukarya as well as Archaea and to glycerate in the npED branch [18,20]. In the genome of more conserved in and . Archaea Ar-4, one gene cluster mcup_1768 to 1770 encoded proteins developed particular variations of these pathways, such homologous to Saci_2269, 2270, 2271 and SSO2636, 2637, as inventing some unusual glycolytic enzymes [15,16].In 2639, yet abundances of Mcup_1768, 1769, 1770 were un- thegenomeofM. cuprina the phosphofructokinase gene is affected by growth condition. On the contrary, two ORFs absent, suggesting that glucose cannot be degraded (Mcup_0979 and Mcup_1567), annotated as molybdopterin through the classical EMP pathway. Instead, M. cuprina binding aldehyde oxidase and xanthine dehydrogenase, should degrade glucose like many other hyperthermo- decreased more than 3fold under autotrophic growth philic Archaea through the ED-like pathway [10,17]. condition, although with p > 0.05 and should be verified Although the use of the branched ED pathway has been for their functions. reported in the close relatives Sulfolobus solfataricus[18,19] The ED pathway can branch into non-phosphorylating and Sulfolobus acidocaldarius [20], such feature has not been and semi-phosphorylating branches (Fig. 1). It had been JOURNAL OF PROTEOMICS 109 (2014) 276– 289 283

Fig. 1 – Regulated proteins of central carbon metabolism (including 3-HP/4-HB cycle, branched ED, gluconeogenesis and TCA pathways). The abbreviations of the substrates: KDG, 2-keto-3-deoxygluconate; KD(P)G, 2-keto-3-deoxy-(6-phospho)-gluconate; GA, glyceraldehyde; GAP, glyceraldehyde 3-phosphate; PG, phosphoglycerate; PEP, phosphoenolpyruvate; BPG, 2,3-bisphosphoglycerate; G-1-P, glucose 1-phosphate; G-6-P; glucose 6-phosphate; FP, fructose 1-phosphate; F1,6BP, Fructose 1,6-bisphosphate; DHAP, dihydroxyacetone phosphate. The key enzymes of ED and gluconeogenesis pathways: glucokinase/hexokinase (Mcup_0845); GAD (Mcup_0900); KD(P)G aldolase (Mcup_0949); GAOR (Mcup_0979 & Mcup_1567); glycerate kinase (Mcup_0274); KDG kinase (Mcup_0948); GAPN (Mcup_0943); GAPDH (Mcup_0577); PGK (Mcup_0576); PGM (Mcup_1229 & Mcup_0070); (Mcup_0560); pyruvate kinase (Mcup_0823); putative pyruvate: ferredoxin oxidoreductase alpha subunit (Mcup_1587); pyruvate, water dikinase (Mcup_1655); phosphoenolpyruvate carboxykinase (Mcup_0741). The key enzymes of 3-HB/4-HP cycle (based on the M. sedula model proposed in reference [37]): ACC/PCC (Mcup_0858, 1925 & 1926); MCR/SCR (Mcup_1427); MSR (Mcup_0293); 3-HPCS (Mcup_0744); 3-HPCD (Mcup_0286); ACR (Mcup_0809); MCE (Mcup_1517); MCM (Mcup_0235 & 1516); SSR (Mcup_0811); 4-HBCS (Mcup_0813?); 4-HBCD (Mcup_0888); CCH (Mcup_1680); 3-HBCD (Mcup_1680); ACK (Mcup_1437). The key enzymes of TCA: citrate synthase (Mcup_0716, 1787); aconitate hydratase (Mcup_0752); isocitrate dehydrogenase (Mcup_1421); α-oxoglutarate ferredoxin oxidoreductase α & β subunits (Mcup_1569 & 1570); succinyl-CoA synthetase α& β subunits (Mcup_0651 & 0650); succinate dehydrogenase ABCD subunits (Mcup_1935,1934, 1933 & 1932); fumarate hydratase (Mcup_0749); malate dehydrogenase (Mcup_1632, 0031); phosphoenolpyruvate carboxylase (Mcup_0741, 1246). The codes of enzymes were shown in different colors: up-regulated proteins for autotrophy with p value <0.05 were given in red, and the up-regulated proteins with p value >0.05 were given in pink. The down-regulated proteins with p value <0.05 were in green, with p value >0.05 were in blue, and the proteins with no significant change to growth mode were given in black; the proteins not identified with proteomics were given in purple.

found that semi-phosphorylating ED branch exists in (GAPN) dehydrogenase. KDG kinase (Mcup_0948) level Sulfolobales, and is operated mainly by 2-keto-3- did not depend significantly on the growth condition, deoxygalactonate/2-keto-3-deoxygluconate (KDG) kinase, even though its orthologous protein SSO3195 had been 2-keto-3-deoxy-(6-phospho)-gluconate (KD(P)G) aldolase, reported to be the sole KDG kinase in S. solfataricus [19]. and nonphosphorylating glyceraldehyde-3-phosphate KD(P)G aldolase (Mcup_0949), another important enzyme 284 JOURNAL OF PROTEOMICS 109 (2014) 276– 289

of the semi-phosphorylating ED branch, decreased more fold (p > 0.05) down-regulated for autotrophic cells, might than 5fold under the autotrophic condition. Its homolog be like the Msed_0431 which had the function for SSO3197 of S. solfataricus had been described as bi- catalyzing pyruvate phosphorylation to PEP [7]. This functional enzyme for KDG or KDPG cleavage or for- will be a complement of PEPS for enhancing gluconeogen- mation with activities on both phosphorylated and esis function. For phosphoenolpyruvate carboxykinase non-phosphorylated substrates (GAP and KD(P)G or GA (PEPCK, Mcup_0741), also expressed in proteomics, the and KDG) [21]. Comparative proteomics of Ar-4 showed activity of catalyzing the reversible decarboxylation of that the protein of GAPN (Mcup_0943) decreased 3fold in oxaloacetate to form PEP had been detected for a homol- autotrophic cells (p > 0.05), which displays 63% amino ogous protein (Msed_ 1452) in M. sedula[7]. The GAPDH acid identity to SSO3194 (S. solfataricus). Some works had (Mcup_0577) decreased more than 4 folds under autotro- confirmed the function of SSO3194 for oxidizing GAP to phic condition, while the amounts of PGK did not change 3-phosphoglycerate [21,26].Theresearchersalsofound according to the growth condition. Classical GAPDH that GAD, KD(P)G aldolase, KDG kinase and GAPN of S. coupled with PDK, commonly present in (hyper)thermo- solfataricus and T. tenax or part of them from S. tokodaii, S. philic Archaea, had been verified to function in anabolism acidocaldarius and Halobacterium formed a gene cluster and (gluconeogenic direction) rather than in catabolism (glyco- suggested that GAPN had important roles in the semi- lytic direction) [30–32]. Unidirectional bifunctional anabolic phosphorylating branch of ED pathway. In contrast to FBPA/ase (Mcup_2006), with 82% and 78% identity to these organisms, the corresponding genes have different ST0318 and SSO0286, was found to decrease 2.07 folds ORF arrangements in Metallosphaera spp. (Fig. S3). Phos- (p > 0.05) for autotrophic cells, the function of its homol- phoglycerate mutase (PGM) is also the specific enzyme ogous proteins ST0318 and SSO0286 had been verified in for the semi-phosphorylating ED branch. Like other S. tokodaii[33] and S. solfataricus[34,35].Inproteomicsof T Archaea, Metallosphaera spp. possess two kinds of PGMs Ar-4 , another protein, D-fructose 1,6-bisphosphate aldol- [27], cofactor-independent PGM (iPGM, Mcup_0070) and ase (FBPA, Mcup_1834) also showed 1.88 fold (p > 0.05) cofactor-dependent phosphoglycerate mutase (dPGM, decrease; being identical to SSO3226, its enzymatic func- Mcup_1229). In our research, levels of iPGM (Mcup_0070) tion still remains to be confirmed in Sulfolobales [16].Other and dPGM were unequally affected by growth condition. enzymes, acting for both glycolysis and gluconeogen- iPGM (Mcup_0070) decreased more than 5fold under esis, were iPGM (Mcup_0070) and dPGM (Mcup_1229). autotrophic condition, but dPGM (Mcup_1229) levels Triosephosphate isomerase (TIM, Mcup_0783), bifunc- remained almost constant. In recent research, iPGM tional phosphoglucose/phosphomannose isomerase (PGI/ (SSO0417) of S. solfataricus was found to be up-regulated PMI, Mcup_0425), and phosphoglucomutase/phospho- when grown on yeast extract and trypton vs. growth on mannomutase (PGM/PMM, Mcup_1849), were quantified, glucose [28], but the activity of dPGM (SSO2236) had not but did not respond to growth mode. From these results, it been detected [29]. The work of Johnsen showed that iPGM could be concluded that although thermophilic Archaea from the hyperthermophilic sulfate reducer Archaeoglobus Metallosphaera spp. did not use classical EMP pathway fulgidus was transcribed during growth on lactate/sulfate for carbon catabolism, they might utilize reversed EMP for gluconeogenesis direction [27]. Therefore, further pathway (gluconeogenesis pathway) for carbon anabolism. molecular and biochemical experiments under different Since enzyme amounts were mostly unaffected by growth growth conditions are necessary to define the physiolog- condition, flux control by post-translational modification ical role of both dPGM and iPGM in Archaea. or metabolic feedback could be present to control the rate In summary, our proteomics data confirmed that M. cuprina of gluconeogenesis (Fig. 1). metabolizes glucose and/or galactose via the non- (iii) Tricarboxylic acid (TCA) cycle: phosphorylative and the semi-phosphorylative branches In this comparative proteome study, most proteins of ED pathway. involved in the TCA cycle were identified, except for (ii) Proteins for gluconeogenesis: some subunits of pyruvate ferredoxin oxidoreductase.

Pyruvate produced by CO2 fixation, TCA cycle or other Moreover, out of the 8 main enzymes of TCA, 5 enzymes metabolic pathways might be channeled into the gluco- were less abundant when grown autotrophically on

neogenesis for anabolism. The enzymes specific for sulfur and CO2 (Fig. 1), strongly indicating the presence gluconeogenesis in Archaea include phosphoenolpyr- of a functional oxidative cycle under heterotrophic uvate synthase (PEPS), phosphorylating GAP dehydroge- growth conditions. Hypothetical pyruvate ferredoxin nase (GAPDH) together with phosphoglycerate kinase oxidoreductase complexes (α, β, γ, and δ subunits) are (PGK), and fructose-1,6-bisphosphate aldolase/phospha- encoded in duplicate with 40%–50% sequence identity tase (FBPA/ase) [15,16], which are in charge of catalyzing in the genome of Metallosphaera spp. (mcup_1586–1589 the irreversible reactions of the gluconeogenesis process. and mcup_1758–1761 in M. cuprina), and together with Mcup_0533, Mcup_0577/0576 and Mcup_1834/2006 were dihydrolipoamide dehydrogenase (coded by three annotated as PEPS, GAPDH/PGK, and FBPA/FBPA/ase in candidate ORFs: mcup_0313, mcup_0680, or mcup_1163) M. cuprina, respectively, all of them were detected in the might have the function to oxidize pyruvate to proteome analysis. PEPS (Mcup_0533) decreased 1.77 folds acetyl-CoA, preparing for TCA entry. Only three subunits under autotrophic condition, another protein Mcup_1655, of hypothetical pyruvate ferredoxin oxidoreductase com- 82% similar to pyruvate:water dikinase (Msed_0431) of plex (Mcup_1587, 1759 and 1761) decreased less than 3 M. sedula, also detected by proteomics, although only 1.73 folds (p > 0.05) for autotrophic cells. The increase of JOURNAL OF PROTEOMICS 109 (2014) 276– 289 285

Mcup_0313 and Mcup_0680 or inconspicuous change Vera [39] a bi-functional protein (Msed_0399), homolog of (lower than 1.6 folds) of isocitrate dehydrogenase Mcup_1680, performed the activity of CCH and 3-HBCD in (Mcup_1421), fumarate hydratase (Mcup_0749), and ma- M. sedula. The function of other enzymes still awaits late dehydrogenase (Mcup_1632) when grown autotrophi- definition. Several gene candidates were annotated as cally versus heterotrophically was contradictory to our putative acetyl-CoA acetyltransferase (Mcup_0582, 0956, expectations and demands further verification of their 1437, 1683, 1961, 1798, 1799). However, only Mcup_1437 function. was up-regulated under autotrophic versus heterotrophic (iv) Autotrophic fixation pathway: conditions, like the orthologous Msed_0656 which was the With genomic, transcriptomic and enzyme activity anal- only ORF that exhibited up-regulation under autotrophic yses, it was established that members of the order conditions in M. sedula[4]. Hence, we infer Mcup_1437 as

Sulfolobales fix CO2 by 3-hydroxypropionate/4-hydro- putative ACK for converting acetoacetyl-CoA to acetyl- xybutyrate (3-HP/4-HB) cycle [4,36,20], although the func- CoA in 3-HP/4-HB cycle. Other proteins (Mcup_0582, 0956, tion of few proteins has still not been confirmed. Our 1683, 1961, 1798, 1799) were more abundant under comparative proteomics of M. cuprina revealed that most heterotrophy and might be involved in amino acid enzymes related to 3-HP/-4HB cycle, which had been metabolism when cultivated on yeast extract. found in some members of the Sulfolobales, were highly Upon converting acetoacetyl-CoA into two acetyl-CoA,

expressed when grown autotrophically on CO2 with sulfur pyruvate synthase transforms the acetyl-CoA to pyruvate. compared to heterotrophically on yeast, only 3-hydroxy- Two multienzyme complexes (Mcup_1586–1589 and propionyl-CoA dehydratase (3-HPCD, Mcup_0286) and Mcup_1758–1761) were annotated as pyruvate ferredoxin acryloyl-CoA reductase (ACR Mcup_0809) did not show oxidoreductases, for which their dual function for cata- distinct response to growth condition in the first part of lyzing the oxidative decarboxylation of pyruvate to acetyl- 3-HP/4-HB cycle (Fig. 1). For many homologous enzymes CoA and inverse reaction from acetyl-CoA to pyruvate in the function has been verified in M. sedula,suchas some Archaea had been proven under different growth acetyl-CoA carboxylase/propionyl-CoA carboxylase com- conditions [40,41], but in our work, only Mcup_1587 and plex (ACC/PCC, Mcup_0858, 1925 and 1926/Msed_1375, 1759 showed unequivocal abundance changes. Thus, they 0148 and 0147), malonate semialdehyde reductase (MSR, most likely form the active complex under our growth Mcup_0293/Msed_1993), 3-hydroxypropionyl-CoA synthe- conditions. tase (3-HPCS, Mcup_0744/Msed_1456), methylmalonyl- It is worth mentioning that in our shotgun proteome CoA epimerase (MCE, Mcup_1517/Msed_0639), methyl- analysis, several enzymes, including Mcup_0858, 1926, malonyl-CoA mutase (MCM, Mcup_0235/Msed_2055), 0293, 0744, 0811 and 0888, were found to be enriched in the and methylmalonyl-CoA mutase α subunit (Mcup_1516/ membrane fraction and we attempted their heterologous Msed_0638) [24,25,19]. Starting from succinyl-CoA, the expression in different hosts, but failed to obtain them in second part of 3-HP/4-HB cycle was completed by succinic soluble form. semialdehyde reductase (SSR), 4-hydroxybutyryl-CoA synthetase (4-HBCS), 4-hydroxybutyryl-CoA dehydratase 3.3. Amino acid metabolism (4-HBCD), crotonyl-CoA hydratase (CCH), and acetoacetyl- CoA β-ketothiolase (ACK), of which. SSR (Mcup_0811) and Commonly yeast extract has been used as carbon and energy 4-HBCD (Mcup_0888) were up-regulated more than 4 folds. source for heterotrophic growth of Sulfolobales, therefore, Abundance increases under autotrophic growth could be amino acid metabolism was important for central carbon used to narrow-down candidate enzymes and to verify metabolism. From proteomics data of M. cuprina in this genome annotation and biochemical data from M. sedula. research, many enzymes involved in amino acid metabolism In case of putative 4-HBCD, only one protein (Mcup_0888) (annotated in KEGG database) exhibited different expression in M. cuprina corresponded to Msed_1321, and no protein for cells growing on sulfur autotrophically vs. growing on was homologous to another candidate 4-HBCD from M. yeast extract heterotrophically (Table S1). More than 30 sedula (msed_1220). Recently, 4-hydroxybutyrate-CoA syn- enzymes were up-regulated under heterotrophic growth, thetase (4-HBCS) function was described for Msed_0406 especially enzymes involved in , , and Msed_0394 in M. sedula[5].Inourproteomedata, and degradation, like Mcup_1792, 1679, 1693, 1691, Mcup_1674 with 73% amino acid identity to Msed_0406 0582, 1798 and 1799. Most of the enzymes involved in ala- showed no significant change under both growth condi- nine, aspartate, glutamate and tryptophan metabolism tions, Mcup_1685 with 67% amino acid identity to (mainly including: Glu/Leu/Phe/Val dehydrogenase, Mcup_ Msed_0394, decreased 4.5 folds under autotrophic growth. 0214; /-glyoxylate transaminase, Mcup_1305; On the contrary, Mcup_0813, similar to Msed_1422, aspartate carbamoyltransferase Mcup_0320; 4-aminobutyrate annotated as coenzyme F390 synthetase-like protein, aminotransferase, Mcup_1793; enoyl-CoA hydratase, Mcup_ increased in the autotrophic cells. Previously, Msed_1422 1693; acetyl-CoA acetyltransferase, Mcup_1691, 1798, 1799,) wasannotedas4-HBCS[37],butintheworkofHanetal., were also strongly regulated under both growth conditions. no activity was detected on 4-HB [38]. Thus, the function Some proteins annotated either in amino acid metabolism or

of Mcup_0831 requires further verification. Six ORFs CO2 fixation pathway, exhibited conflicting abundance changes (Mcup_0293, 0812, 1680, 1688, 1692 and 1693) were anno- for their respective growth condition, such as Mcup_0235, 0858, tated as putative 3-hydroxyacyl-CoA dehydrogenase 1427, 1437, 1516, 1517 being more abundant in autotrophic (3-HBCD) or enoyl-CoA hydratase. According to Ramos- cells, and the opposite for Mcup_1683, 1691, 1693, 1798, 1799. 286 JOURNAL OF PROTEOMICS 109 (2014) 276– 289

Altogether our observations were in concordance with former plex (mcup_0684–0689), similar to A. ferrooxidans (AFE_2555–2550) research that found that M. cuprina could grow on these amino [45]; the proteins in the cluster might be involved in sulfur acids [9], but demonstrate the necessity of further experimental transfer and reversible reduction of the disulfide bond X-S-S-X. works to clarify enzyme function and metabolic networks for Mcup_0686, identical to AFE_2553 (Hdr A) together with amino acids. Mcup_0685, Mcup_0688 and Mcup_0689, annotated as HdrB subunit, HdrC1-like protein, and HdrB1, respectively, increased 3.4. Oxidation of reduced inorganic sulfur compounds (RISCs) their abundance in autotrophic cells. Other members of this under autotrophic culture conditions cluster, i.e., Mcup_0681 (annotated as DsrE family protein), Mcup_0682 and Mcup_0687 (hypothetical protein), Mcup_0683 According to the enzymes detected in the proteome as well as (SirA family protein, TusA), exhibited indeterminate differences putative enzymes annotated in the genome and additional under either autotrophic or heterotrophic conditions. Their exact homology searches, we have proposed a hypothetical model for function in Metallosphaera remains to be established. NADPH: the oxidation of reduced inorganic sulfur compounds (RISCs). sulfur oxidoreductase (NSR, Mcup_0209), with 43% identity to (Fig. 2). Of major importance, prior research had not found genes PF1186 of Pyrococcus furiosus, decreased 4 folds under autotrophic homologous to sulfur oxidoreductase (SOR) in the genome conditions. NSR had been found with the S0 reduction activity of Metalloshaera spp.[3,10], thus their mechanism of oxidizing dependent on NADPH and coenzyme A in several anaerobically elemental sulfur is unknown. Quantitative proteomics showed growing Archaea like Pyrococcus and Thermococcus[46,47],butthe that sulfide:quinone oxidoreductase (SQR, Mcup_0231) and function of it in the aerobic Archaea is unknown — its decrease in thiosulfate:quinone oxidoreductase (TQO) complex (Mcup_1712 amount during autotrophic growth doubts an imaginable reverse and Mcup_1713) increased under autotrophic growth conditions. function, i.e. the oxidation of sulfur, which could explain the For an SQR homolog of A. ambivalens, SQR activity (i.e. oxidation of observed oxidation of elemental sulfur in the absence of an SOR sulfide to polysulfides and electron transfer to quinone) was homolog. demonstrated [42]. Another protein, Mcup_1702, with 25% identity to DoxD2 CAC86936 of A. ambivalens, was absent in the 3.5. Transporters proteome. However, functions of these putative Dox enzymes have not been experimentally verified so far. We failed to detect We sought to identify transporter proteins that most likely TetH (Mcup_1281), for which tetrathionate-hydrolyzing ability play pivotal roles for growth under the respective conditions has been confirmed for ferrooxidans and based on their abundance changes (Fig. S4). Two putative Acidithiobacillus caldus [43,44]. Interestingly, the failure to detect sugar MFS systems (Mcup_1006 + Mcup_1623) increased their TetH (Mcup_1281) during growth with elemental sulfur as the abundances during autotrophic growth. Two ABC transport energy source agrees with the fact that the activity of TetH could systems, which might be involved in trehalose or maltose not be detected in A. ferrooxidans and A. caldus,too,whensulfur transport were affected by the growth condition. For one ABC was the [43,44]. In the genome of M. cuprina,there transporter (Mcup_0957–0960), we obtained conflicting re- existed a cluster encoding heterodisulfide reductase (Hdr) com sults, i.e. Mcup_0958 and Mcup_0959, increased in abundance

Fig. 2 – Hypothetical model of sulfur metabolism in M. cuprina derived from enzymes detected in the proteome or possible enzymes annoted in the gemone: SAOR (Mcup_1714); SR, sulfite reductase (Mcup_1109); SQR (Mcup_0231); TQO (Mcup_1712

&1713); NSR (Mcup_0209); TetH (Mcup_1281); HdrC (Mcup_0684); TusA (Mcup_0683); DsrE (Mcup_0681); SOR; Q, quinone; QH2, hydroquinone; APS; APAT; APSR; AK, adenylate kinase. The up-regulated proteins for autotrophy with p value <0.05 were given in red, and with p value >0.05 were given in pink. The down-regulated proteins with p value <0.05 were in green, with p value >0.05 were in blue, and the proteins with no significant response to growth mode were given in black. The proteins not identified with proteomics were given in purple; ?: stand for no functional genes being predicted in the genome. JOURNAL OF PROTEOMICS 109 (2014) 276– 289 287 for autotrophic cells, but the other subunits Mcup_0957 and So far, important aspects about the carbon metabolic Mcup_0960 for heterotrophic cells. The second ABC transport- networks and their regulation in hyperthermophlic Archaea er (Mcup1624_Mcup1627), was more abundant during he- are unknown, yet they are of general interest and often terotrophic growth. A type II secretion system component important in biotechnological applications. This comprehensive (Mcup_1528–1529) increased upon autotrophic growth, sug- proteomics analysis discovered several enzymes on which gesting higher demands for secreted proteins. Since the further experimental research and verification should be con- secretome was not analyzed, no further conclusions towards centrated. The obtained data and developed proteomics the possible physiological demand (e.g. sulfur mobilization) workflow would provide a good basis for subsequent studies of can be drawn. Because a Glu/Leu/Phe/Val dehydrogenase metabolic mechanisms in Archaea under different growth (Mcup_0214) was detected in high amounts during growth modes, such as autotrophy, heterotrophy, or mixotrophy. on yeast extract, concomitant increase of amino acid or The mass spectrometry proteomics data have been depos- oligopeptide importers appears reasonable. Indeed, several ited to the ProteomeXchange Consortium (http://www.proteo transporters were quantified, but only one transporter mexchange.org) via the PRIDE partner repository with the (Mcup_0101–0105) increased accordingly. Quite intriguingly, dataset identifier PXD000676. in another ABC peptide transport system (Mcup_1641–1646), The details: subunit Mcup_1644 was more abundant under autotrophic Project name: Metallosphaera cuprina cultivation on sulfur growth; perhaps increase of peptide transporters could be or yeast extract implicated with a general starvation response. Project accession: PXD000676

Supplementary data to this article can be found online at 4. Conclusion http://dx.doi.org/10.1016/j.jprot.2014.07.004.

Members of Metallosphaera genus belong to Sulfolobales that grow aerobically at low pH and high temperatures. Beside Conflict of Interest autotrophic growth on carbon dioxide by oxidation of ele- mental sulfur, ferrous sulfate, potassium tetrathionate or The authors declare no conflict of interest. pyrite, all known strains also showed heterotrophic growth on various organic compounds. Some biochemical studies have suggested that when hyperthermophilic Archaea grow het- Acknowledgments erotrophically on organic compound(s), they metabolize – glucose via the ED (Entner Doudoroff) pathway to produce We gratefully acknowledge the help from Tanja Bojarzyn with pyruvate [21]. Nevertheless, more researches concentrated data deposition at ProteomeXchange. The work was support- on autotrophic carbon fixation in via the ed by the Ministry of Science and Technology (973 program 3-hydroxypropionate/4-hydroxybutyrate cycle [37,48,7]. Little 2010CB630903, 863 program 2012AA061501) and National is known about the energy metabolism for Metallosphaera Science Foundation of China (31070042, 31171234). genus at autotrophic and heterotrophic growth conditions. Shotgun comparative proteome analysis from cells cultured autotrophically on sulfur vs. those cultured heterotrophically REFERENCES on yeast extract indicated that the proteome changes were extensive, and provided the first insights into carbon and sulfur metabolism on the protein level. [1] Zaparty M, Esser D, Gertig S, Haferkamp P, Kouril T, Manica A, et al. “Hot standards” for the thermoacidophilic archaeon Sulfolobus solfataricus. 2010;14:119–42. • When grown autotrophically, highly expressed proteins [2] ChenL,BruggerK,SkovgaardM,RedderP,SheQ, Metallosphaera indicated that spp. fix CO2 through 3-hydro- Torarinsson E, et al. The genome of Sulfolobus acidocaldarius, xypropionate/4-hydroxybutyrate (3-HP/4-HB) cycle, and oxidize a model organism of the .JBacteriol the RISCs through a series of new unusual enzymes that are 2005;187:4992–9. different from other acidophilic Archaea and bacteria. [3] Auernik KS, Maezato Y, Blum PH, Kelly RM. The genome • When grown heterotrophically, Metallosphaera spp. could sequence of the metal-mobilizing, extremely utilize yeast extract as energy sources, through amino acid thermoacidophilic archaeon Metallosphaera sedula provides insights into bioleaching-associated metabolism. Appl metabolism gaining access into central carbon metabolism, Environ Microbiol 2008;74:682–92. catabolize glucose via the branch ED pathway and obtained [4] Auernik KS, Kelly RM. Physiological versatility of the the energy through TCA cycle. extremely thermoacidophilic archaeon Metallosphaera sedula • Although Metallosphaera spp. do not use classical EMP supported by transcriptomic analysis of heterotrophic, pathway for carbon catabolism, they may use reversed autotrophic, and mixotrophic growth. Appl Environ Microbiol – EMP (gluconeogenesis) pathway for carbon anabolism, like 2010;76:931 5. [5] Hawkins AS, Han Y, Bennett RK, Adams MW, Kelly RM. Role in Sulfolobus spp. of 4-hydroxybutyrate-CoA synthetase in the CO fixation • Changing abundances of the proteins related to different 2 cycle in thermoacidophilic archaea. J Biol Chem carbon sources and sulfur imply that the metabolic regula- 2013;288:4012–22. tion of M. cuprina might be on many levels, ranging from [6] Auernik KS, Kelly RM. Impact of molecular hydrogen on transcription to metabolic feedback. chalcopyrite bioleaching by the extremely thermoacidophilic 288 JOURNAL OF PROTEOMICS 109 (2014) 276– 289

archaeon Metallosphaera sedula. Appl Environ Microbiol Entner–Doudoroff pathway in Sulfolobus solfataricus. FEBS Lett 2010;76:2668–72. 2004;576:133–6. [7] Estelmann S, Hugler M, Eisenreich W, Werner K, Berg IA, [24] Liu B, Wu L, Liu T, Hong Y, Shen Y, Ni J. A MOFRL family Ramos-Vera WH, et al. Labeling and enzyme studies of the glycerate kinase from the thermophilic crenarchaeon, central carbon metabolism in Metallosphaera sedula. J Bacteriol Sulfolobus tokodaii, with unique enzymatic properties. 2011;193:1191–200. Biotechnol Lett 2009;31:1937–41. [8] Maezato Y, Johnson T, McCarthy S, Dana K, Blum P. Metal [25] Reher M, Schonheit P. Glyceraldehyde dehydrogenases from resistance and lithoautotrophy in the extreme the thermoacidophilic euryarchaeota Picrophilus torridus and thermoacidophile Metallosphaera sedula. J Bacteriol Thermoplasma acidophilum, key enzymes of the 2012;194:6856–63. non-phosphorylative Entner–Doudoroff pathway, constitute [9] Liu LJ, You XY, Guo X, Liu SJ, Jiang CY. Metallosphaera cuprina a novel enzyme family within the aldehyde dehydrogenase sp. nov., an acidothermophilic, metal-mobilizing archaeon. superfamily. FEBS Lett 2006;580:1198–204. Int J Syst Evol Microbiol 2011;61:2395–400. [26] Ettema TJ, Ahmed H, Geerling AC, van der Oost J, Siebers B. [10] Liu LJ, You XY, Zheng H, Wang S, Jiang CY, Liu SJ. Complete The non-phosphorylating glyceraldehyde-3-phosphate genome sequence of Metallosphaera cuprina, a metal dehydrogenase (GAPN) of Sulfolobus solfataricus: a key-enzyme sulfide-oxidizing archaeon from a hot spring. J Bacteriol of the semi-phosphorylative branch of the Entner–Doudoroff 2011;193:3387–8. pathway. Extremophiles 2008;12:75–88. [11] Haussmann U, Qi SW, Wolters D, Rogner M, Liu SJ, Poetsch A. [27] Johnsen U, Schonheit P. Characterization of Physiological adaptation of Corynebacterium glutamicum to cofactor-dependent and cofactor-independent benzoate as alternative carbon source — a membrane phosphoglycerate mutases from archaea. Extremophiles proteome-centric view. Proteomics 2009;9:3635–51. 2007;11:647–57. [12] Fischer F, Wolters D, Rogner M, Poetsch A. Toward the [28] Kort JC, Esser D, Pham TK, Noirel J, Wright PC, Siebers B. A complete membrane proteome: high coverage cool tool for hot and sour archaea: proteomics of Sulfolobus of integral membrane proteins through solfataricus. Proteomics 2013;13:2831–50. transmembrane peptide detection. Mol Cell Proteomics [29] Haferkamp P. Biochemical studies of enzymes involved in 2006;5:444–53. glycolysis of the thermoacidophilic crenarchaeon Sulfolobus [13] Niessen S, Mcleod I, Yates III JR. Digestion of protein solfataricus. Universität Duisburg-Essen: Fakultät für Chemie; complexes. In: Golemis E, Adam P, editors. Protein–Protein 2011. Interactions. 2nd ed. CSH; 2005. [30] Lorentzen E, Hensel R, Knura T, Ahmed H, Pohl E. Structural [14] Vizcaino JA, Cote RG, Csordas A, Dianes JA, Fabregat A, Foster basis of and substrate specificity of the JM, et al. The Proteomics Identifications (PRIDE) database and non-phosphorylating glyceraldehyde 3-phosphate associated tools: status in 2013. Nucleic Acids Res 2013;41: dehydrogenase from Thermoproteus tenax. J Mol Biol D1063–9. 2004;341:815–28. [15] Gerdes S, Overbeek R. Embden-Meyerhof glycolytic pathway [31] Schut GJ, Brehm SD, Datta S, Adams MW. Whole-genome and Gluconeogenesis (www.theseed.org/SubsystemStories/ DNA microarray analysis of a and an Glycolysis/Story.pdf). archaeon: Pyrococcus furiosus grown on carbohydrates or [16] Bräsen C, Esser D, Rauch B, Siebers B. Carbohydrate peptides. J Bacteriol 2003;185:3935–47. metabolism in archaea: current insights into unusual Brunner NA, Siebers B, Hensel R. Role of two different enzymes and pathways and their regulation. Microbiol Mol glyceraldehyde-3-phosphate dehydrogenases in controlling Biol Rev 2014;78:89–175. the reversible Embden–Meyerhof–Parnas pathway in [17] Ahmed H, Ettema TJG, Tjaden B, Geerling ACM, van der Oost J, Thermoproteus tenax: regulation on protein and transcript Siebers B. The semi-phosphorylative Entner–Doudoroff level. Extremophiles 2003;5:101–9. pathway in hyperthermophilic archaea: a re-evaluation. [32] Schafer T, Schonheit P. Gluconeogenesis from pyruvate in Biochem J 2005;390:529–40. the hyperthermophilic archaeon Pyrococcus [18] Esser D, Kouril T, Talfournier F, Polkowska J, Schrader T, furiosus — Involvement of Reactions of the Embden–Meyerhof Brasen C, et al. Unraveling the function of paralogs of the Pathway. Arch Microbiol 1993;159:354–63. aldehyde dehydrogenase super family from Sulfolobus [33] Fushinobu S, Nishimasu H, Hattori D, Song HJ, Wakagi T. solfataricus. Extremophiles 2013;17:205–16. Structural basis for the bifunctionality of fructose-1, [19] Kouril T, Wieloch P, Reimann J, Wagner M, Zaparty M, Albers SV, 6-bisphosphate aldolase/. Nature et al. Unraveling the function of the two 2011;478:538–41. Entner–Doudoroff branches in the thermoacidophilic [34] Say RF, Fuchs G. Fructose 1,6-bisphosphate aldolase/ Crenarchaeon Sulfolobus solfataricus P2. FEBS J 2013;280:1126–38. phosphatase may be an ancestral gluconeogenic enzyme. [20] Kardinahl S, Schmidt CL, Hansen T, Anemuller S, Petersen A, Nature 2010;464:1077–81. Schafer G. The strict molybdate-dependence of [35] Kouril T, Esser D, Kort J, Westerhoff HV, Siebers B, Snoep JL. glucose-degradation by the thermoacidophile Sulfolobus Intermediate instability at high temperature leads to low acidocaldarius reveals the first crenarchaeotic molybdenum pathway efficiency for an in vitro reconstituted system of containing enzyme — an aldehyde oxidoreductase. Eur J gluconeogenesis in Sulfolobus solfataricus. FEBS J Biochem 1999;260:540–8. 2013;280:4666–80. [21] Ahmed H, Ettema TJ, Tjaden B, Geerling AC, van der Oost J, [36] Alber B, Olinger M, Rieder A, Kockelkorn D, Jobst B, Hugler M, Siebers B. The semi-phosphorylative Entner–Doudoroff et al. Malonyl-coenzyme A reductase in the modified pathway in hyperthermophilic archaea: a re-evaluation. 3-hydroxypropionate cycle for autotrophic carbon fixation in Biochem J 2005;390:529–40. archaeal Metallosphaera and Sulfolobus spp. J Bacteriol [22] Siebers B, Wendisch VF, Hensel R. Carbohydrate metabolism 2006;188:8551–9. in Thermoproteus tenax: in vivo utilization of the [37] Berg IA, Kockelkorn D, Buckel W, Fuchs G. A non-phosphorylative Entner–Doudoroff pathway and 3-hydroxypropionate/4-hydroxybutyrate autotrophic carbon characterization of its first enzyme, glucose dehydrogenase. dioxide assimilation pathway in archaea. Science Arch Microbiol 1997;168:120–7. 2007;318:1782–6. [23] Lamble FJ, Milburn CC, Taylor GL, Hough DW, Danson MJ. [38] Han Y, Hawkins AS, Adams MW, Kelly RM. Epimerase Gluconate dehydratase from the promiscuous (Msed_0639) and mutase (Msed_0638 and Msed_2055) convert JOURNAL OF PROTEOMICS 109 (2014) 276– 289 289

(S)-methylmalonyl-coenzyme A (CoA) to succinyl-CoA in the [44] Bugaytsova Z, Lindstrom EB. Localization, purification and Metallosphaera sedula 3-hydroxypropionate/ properties of a tetrathionate hydrolase from Acidithiobacillus 4-hydroxybutyrate cycle. Appl Environ Microbiol caldus. Eur J Biochem 2004;271:272–80. 2012;78:6194–202. [45] Quatrini R, Appia-Ayme C, Denis Y, Jedlicki E, Holmes DS, [39] Ramos-Vera WH, Weiss M, Strittmatter E, Kockelkorn D, Bonnefoy V. Extending the models for iron and sulfur Fuchs G. Identification of missing genes and enzymes for oxidation in the extreme Acidithiobacillus autotrophic carbon fixation in Crenarchaeota. J Bacteriol ferrooxidans. BMC Genomics 2009;10:394. 2011;193:1201–11. [46] Schut GJ, Bridger SL, Adams MW. Insights into the [40] Furdui C, Ragsdale SW. The role of pyruvate ferredoxin metabolism of elemental sulfur by the hyperthermophilic oxidoreductase in pyruvate synthesis during autotrophic archaeon Pyrococcus furiosus: characterization of a coenzyme growth by the Wood–Ljungdahl pathway. J Biol Chem A-dependent NAD(P)H sulfur oxidoreductase. J Bacteriol 2000;275:28494–9. 2007;189:4431–41. [41] Bock AK, Kunow J, Glasemacher J, Schonheit P. Catalytic [47] Toth A, Takacs M, Groma G, Rakhely G, Kovacs KL. A novel properties, molecular composition and sequence alignments NADPH-dependent oxidoreductase with a unique domain of pyruvate: ferredoxin oxidoreductase from the structure in the hyperthermophilic Archaeon, Thermococcus methanogenic archaeon Methanosarcina barkeri (strain litoralis. FEMS Microbiol Lett 2008;282:8–14. Fusaro). Eur J Biochem 1996;237:35–44. [48] Alber BE, Kung JW, Fuchs G. 3-Hydroxypropionyl-coenzyme A [42] Brito JA, Sousa FL, Stelter M, Bandeiras TM, Vonrhein C, synthetase from Metallosphaera sedula, an enzyme involved in – Teixeira M, et al. Structural and functional insights into autotrophic CO2 fixation. J Bacteriol 2008;190:1383 9. sulfide:quinone oxidoreductase. Biochemistry 2009;48:5613–22. [43] Kanao T, Kamimura K, Sugio T. Identification of a gene encoding a tetrathionate hydrolase in Acidithiobacillus ferrooxidans. J Biotechnol 2007;132:16–22.