Cytosolic iron–sulfur protein assembly system identifies clients by a C-terminal tripeptide

Significance The maturation of eukaryotic cytosolic and nuclear iron–sulfur (Fe–S) proteins is mediated by the cytosolic iron–sulfur protein assembly (CIA) pathway. Unlike other metalloprotein maturation systems with a one-to-one correspondence between the metallochaperone and its apo-client, CIA must identify >30 clients, raising the question of how specificity of cofactor delivery is encoded. The discovery that a tripeptide motif at the C-terminal tail of CIA clients is necessary and sufficient to direct Fe–S cluster delivery from the CIA system not only provides fundamental insights into biochemistry of this vital process but it also unlocks the potential for bioengineering Fe–S cluster delivery to nonnative enzymes, a bottleneck that often frustrates the use of these versatile metalloproteins for synthetic biology applications.


Figures S1 to S13 Tables S1 to S8
Legends for Datasets S1 to S2 SI References Other supporting materials for this manuscript include the following: fusions, roughly 3-fold molar excess over bait protein concentrations were used.In all experiments, a "no-bait" negative control was completed in parallel to detect any nonspecific binding to the resin.The samples were incubated for 1 h 4˚C and then chromatographed through streptactin resin.Input and Elution fractions were analyzed SDS-PAGE (12% for experiments with Leu1, 15% for SPCfusions, 20% for Apd1).For Western blot analysis, proteins were transferred using a Trans-Blot Turbo Mini PVDF Transfer pack and Transfer System with the mixed molecular weight protocol in the manufacturer's instructions (Bio-Rad).The membrane was blocked with 5% casein in TBST and probed with Anti-His antibody (Cell Signaling Technology).Bands were detected via chemiluminescence, according to manufacturer's instructions.
Yeast Strains and Cell Growth.The Saccharomyces cerevisiae strain W303-1A (MATa, ura3-1, ade2-1, trp1-1, his3 -11,15, leu2-3,112) was used as the wild-type strain.Yeast strains are summarized in Table S4.To convert the W303-1A strain into a LEU2 + /Δleu1 strain, first homologous recombination with a leu2 PCR fragment followed by selection on SC glucose medium lacking leucine was carried out.The leu2 PCR product was obtained by fusion PCR of leu2 fragments A, B and C (see Table S5).Yeast strains, Δleu1, LEU2 + /Δleu1, Gal-NFS1/Δleu1, Gal-NAR1/Δleu1, Gal-CIA1/Δleu1, and Gal-CFD1/Δleu1, were constructed by homologous recombination in which the coding region of the leu1 gene was replaced by a natNT2 cassette amplified from pYM17, omitting the 6xHA tag (5).Gal-POL3 was constructed by homologous recombination in which the nucleotide-upstream promoter region was replaced by the natNT2 cassette from pYM-N27 (5), including a GAL promoter sequence.The correctness of the insertion was assured by PCR analysis of genomic DNA.Cells were grown on rich (YP) or minimal (SC) media containing 2 % (w/v) galactose or glucose as carbon source with the appropriate markers.
Yeast expression plasmid construction.For POL3 cloning, the E. coli strain NEB5 F'I q was used.All other genes were cloned using the E. coli strains NEB5 or NEB10.E. coli genomic DNA was used as the template to amplify leuC and leuD.For the other genes (POL3, NAR1, LEU1, APD1), S. cerevisiae genomic DNA was the template for PCR.Information, including primers (Table S6) and plasmids (Table S7), are provided and when not specified otherwise, standard cloning procedures were used.Variants were generated by site-directed mutagenesis with primer design according to Zheng et al. (6).For C-terminal truncations, the appropriate codon was replaced by a stop codon.Successful construction of all plasmids was verified by Sanger sequencing.
The POL3 coding region and the 500 bp at the 5' end of the start codon was inserted into the SacI and EcoRI sites of pRS416-MET25prom-MCS-CYC1term (maintaining the CYC1 terminator) (7).During cloning and mutagenesis of POL3, growth for plasmid minipreps required incubation of agar plates at room temperature for 2 days until colonies appeared and cultivation in liquid media at 30 °C.
For APD1, its coding region, the 526 bp natural promoter, and 327 bp terminator regions were cloned into the SacI and KpnI sites of pRS416-MET25prom-MCS-CYC1term.The NAR1 coding region, along with its natural promoter (591 bp) and terminator (558 bp) were cloned into SacI and KpnI sites of pRS416-MET25prom-MCS-CYC1term.For the Nar1-encoding plasmid under control of the MET25 promoter, the NAR1 coding region and its terminator (558 bp) were cloned into SpeI and KpnI sites of pRS416-MET25 (7).For the Nar1 plasmid with NBP35 promoter, the promoter region of the NBP35 gene (508 bp) was inserted between the SacI and SpeI sites of pRS416-MET25prom-Nar1-NAR1term.
The LEU1 coding region, along with its natural promoter (552 bp) and terminator (512 bp) regions, were cloned as two PCR fragments employing the natural SalI site (883-888) in the LEU1 coding region.First, the SalI-NgoMIV fragment was cloned into pRS416-MET25prom-MCS-CYC1term, followed by the SacI-SalI fragment.Variants with other promoter regions were generated by introducing a SpeI site just before the ATG start codon of LEU1, followed by cloning PCR fragments of the promoter regions of the MET25 (412 bp), TDH3 (649 bp), RET2 (500 bp) and RPL18B (500 bp) genes between SacI and SpeI sites.For pRS416-TDH3prom-Leu1-TDH3term, an XhoI site was introduced immediately following the LEU1 stop codon and a BamHI site was inserted after the LEU1 terminator.Then, a PCR fragment of 449 bp corresponding to the TDH3 terminator was cloned into the XhoI and BamHI sites.
For generation of pRS416-TDH3prom-Leu1-TDH3term with the C-terminal 19 amino acids of the yeast Leu1 replaced by the C-terminal 19 amino acids of the homologs of S. pombe or A. nidulans, a BglII was introduced by mutagenesis at nucleotides 1847-1852 of the leu1 gene, creating a silent mutation.Then, synthetic genes corresponding to the BglII sequence plus nucleotides 1853 to 2280 of yeast LEU1, and yeast codon optimized nucleotides encoding the 19 C-terminal amino acids of the homologs plus a stop codon and the XhoI sequence were cloned into BglII/XhoI sites of pRS416-TDH3prom-Leu1-TDH3term.
For construction of the yeast expression vector for LeuCD (pRS426-FBA1prom-LeuC-linker-LeuD-Leu1-CT-FBA1term), E. coli leuC and leuD genes were each amplified and separately cloned into the SpeI/XhoI sites of the MCS of pRS424 and pRS426, respectively.To create the backbone of the shuttle vector, the TDH3 promoter of vector pRS426-TDH3prom-MCS-CYCterm (7) was replaced by 1000 bp of the S. cerevisiae FBA1 promoter, using SacI and SpeI sites.Then, the CYC1 terminator was exchanged for the FBA1 terminator (1000 bp), using XhoI and KpnI sites.To create pRS426-FBAprom-LeuD-Leu1-TCR, the DNA fragment encoding the yeast Leu1 C-terminus (last 30 amino acids, Leu1-TCR) was cloned into EcoRI/BamHI of the multiple cloning site of pRS426-FBA1prom-MCS-FBA1term followed by cloning of leuD from E. coli (using pRS426 with E. coli leuD as template) into pRS426-FBA1prom-Leu1-TCR-FBA1term.To introduce leuC and the linker region of Leu1 (amino acids 481-543) concomitantly, PCR overlap extension was used (8).In the first PCR reaction (I), leuC (in pRS424) was amplified such that the forward primers introduced a SpeI restriction site and the reverse primer introduced 15 nucleotides corresponding to the 5' end of the LEU1 linker.In the second PCR reaction (II), the LEU1 linker was amplified (from pRS416-TDH3prom-Leu1-TDH3term).The forward primer introduced 20 nucleotides corresponding to the 3'end of leuC and the reverse primer appended an SpeI restriction site.An equimolar amount of PCR-product, I and II, were mixed and hybridized by using a PCR protocol without dNTPs.In a final PCR reaction, the primers flanking the SpeI sites were used to amplify the product, which was subsequently cloned into the SpeI restriction site of pRS426-FBAprom-LeuD-Leu1-TCR, generating pRS426-FBAprom-LeuC-LeuD-Leu1-TCR-FBAterm.To test the effect of the Leu1-TCR, a stop codon was introduced after the leuD coding region by site-directed mutagenesis.To create the QDW tail, primers were used to insert the tripeptide coding region plus a stop codon immediately following leuD.
To shorten the TCR tail from 29 to 10 amino acids (770-779 amino acids of Leu1), Q5 mutagenesis was used to delete the region corresponding to amino acids 750-769 of Leu1-TCR and install an Xho1 site.The resulting PCR product was digested with XhoI cut and ligated.For the construct with the SSG linker sequence substituting the Leu1-TCR, a synthetic DNA sequence encoding for the eight SSG repeats and the codons for QDW was chemically synthesized and cloned into the XhoI/EcoRI sites of 426-FBA1prom-LeuC-linker-LeuD-Leu1-TCR-FBA1term.

Protein expression and purification of Leu1 variants for activity measurements. Yeast Leu1
was cloned into pET28a (Novagen), fused to a N-terminal hexa-histidine tag for heterologous expression and purification.After transformation into BL21 (DE3) cells (NEB), an overnight preculture was diluted (2% inoculum) into 2 liters of LB medium containing ampicillin and 3% (v/v) ethanol p. a and allowed to grown at 37 °C and 200 rpm/min shaking.When the culture reached an OD600 of 0.5, the induction was started by addition of IPTG (0.5 mM), followed by incubation at overnight at 30 °C.Cells were opened with French Press and purified using Ni-IDA (Cube) according to the manufacturer's instructions.The protein eluates were desalted using a Sephadex G-25 column (GE Healthcare) equilibrated with 25 mM Tris-HCl, pH 8.0, 300 mM NaCl.The purified proteins were shock-frozen and stored at -80 °C until use.
Reconstitution of recombinant purified apo-Leu1 was performed in an anaerobic chamber (Coy Laboratory).Approximately 50 µM of Leu1 was reduced with DTT (15 mM, end concentration) for 45 min.After addition of 5 molar-equivalents of ammonium iron (III) citrate and development of a red color the same equivalents of lithium sulfide were slowly pipetted into the reaction tube.The sample was incubated for 2-4 h and the Leu1 activity was tested during this time without desalting the sample.

55
Fe incorporation into Leu1.Leu1 or mutants thereof cloned into pRS416 plasmids under the control of the endogenous LEU1 promoter were transformed into yeast cells, as indicated in figure captions.In vivo radiolabeling with 55 FeCl3 and determination of 55 Fe incorporation into Leu1 beads was carried out as described (9).Briefly, freshly transformed cells were grown on iron-poor SC medium for 40 h supplemented with the appropriated carbon source.After incubation with 55 FeCl3 and 1 mM ascorbate for 2 h in SC iron-poor medium, cell extracts were prepared by using glass beads.The Leu1 protein was immunoprecipitated by using polyclonal antibodies against Leu1 bound to Protein A-Sepharose beads.The amount of protein-associated radioactivity was assessed via liquid scintillation counting.
Growth complementation.Growth complementation assays used the W303, Gal-POL3, Δapd1, Gal-NAR1 and LEU2 + /Δleu1 strains transformed with appropriate plasmids (see figure legends).Transformants were grown in liquid SC medium supplemented with galactose or glucose (2 %) at 30 °C for 40 h (Δapd1, LEU2 + /Δleu1 and Gal-NAR1, including control W303), or for 16 h (series Gal-POL3, including control W303).At this point, the cultures were diluted in SC medium to an OD600 of 0.5 and 5 µL aliquots were submitted to a 10-fold serial dilution and spotted into SC agar plates supplemented with glucose or galactose carbon source.SC agar plates with methyl methane sulfonate or gallobenzophenone at the indicated concentration were freshly prepared.The plates were incubated for 48 h at 30 °C and photographed.At least three experiments using independently transformed yeast cells were performed.

Enzyme activity determination.
Protein determination for activity in crude extracts was performed using the Microbiuret method with desoxycholate/trichloroacetic acid coprecipitation (10,11).Isopropylmalate isomerase and succinate dehydrogenase activity were determined in freshly prepared yeast or E. coli cell extracts.Yeast cell lysates were prepared using the glass bead method and clarified by centrifugation, whereas E. coli cell lysates were prepared using a French Press.For Leu1 activity, the reaction was conducted in a buffer containing 20 mM Tris-HCl, pH 7.4 and 50 mM NaCl.The formation of isopropylmaleate after addition of 0.2 mM 3-isopropylmalate was followed by the increase of absorption at 235 nm.To correct for variations in growth conditions and deviations from the ambient temperature (22 °C), a wild type Leu1 or LeuC-linker-LeuD-TCR (variant 12, Fig. 4C) control was measured in parallel and used to normalize the series measured on different days.
For the coupled assay of Leu1 activity, E. coli isopropylmalate dehydrogenase (leuB) was cloned into the BamHI and HindIII sites of MCS1 of pET-Duet1 and purified by Ni-NTA using standard procedures.Assays contained 20 mM Tris-HCl (pH 8), 50 mM KCl, 10 mM MgCl2, 1 mM DTT, 0.2 mM isopropylmaleate, 0.4 mM NAD + , 2 mM pyrazol and 0.4 U LeuB.The absorbance increase from NADH formation at 340 nm was followed.
The succinate dehydrogenase activity in isolated mitochondria was determined by measuring the reduction of cytochrome c at 550 nm in a reaction containing 8.9 mM succinate and 1 mg/mL bovine heart cytochrome c in a buffer containing 50 mM Tris, pH 8.0 and 50 mM NaCl.The background activity was assayed by measuring the increase of absorbance in the same reaction containing 14.5 mM malonate and 1 mM potassium cyanide.Bioinformatic analysis.Fe-S proteins and adaptor amino acid sequences were collected by NCBI BlastP searches using yeast und human sequences as query.Amino acid sequence alignment was carried out with ClustalO on the EMBL-EBI server (12).The same server was used for the download of reference proteomes for the analysis of C-termini (Data S1).For analysis of the C-termini of Homo sapiens (13), Saccharomyces cerevisiae (14), E. coli (15), Methanocaldococcus jannaschii (16), and Arabidopsis thaliana (17) the published Fe-S protein inventories were manually updated.Datasets for the generation WebLogos (18) were downloaded by selection of the appropriate taxonomic class of organisms in the OrthoDB catalogue of orthologs (19).Sequence fragments lacking N-terminal or C-terminal regions and excessively long sequences due to erroneous translation and/or wrong assignment of introns were manually removed.Then the unaligned Cterminal 20 amino acids were submitted to the Weblogo server for data in Fig. 1E, and fig.S3.The Apd1 collection was from the phylogenetic tree as described in Stegmaier et al. (2).

Fig. S1.
Bar graphs comparing enrichment in C-terminal amino acids in the Fe-S proteomes of yeast (A), humans (B), plants (C), and non-eukaryotic organisms (D).For Panels (A-C), the CIA proteome is in yellow, the mitochondrial Fe-S proteome is purple, and plastidial Fe-S proteome is red.In Panel D, the Methanocaldococcus jannaschii Fe-S proteome is in light gray and Escherichia coli Fe-S proteome is black.In each dataset, the C-terminal amino acid with the highest enrichment is marked with an asterisk.A terminating W (dark green) or F (light green) is exclusively enriched in the CIA proteome.

Fig. S3 .
Fig. S3.Conservation of a C-terminal W/F correlates with conservation of Fe-S cluster binding ligands.(A).Absence (top two Weblogos) and presence (bottom) of a C-terminal W or F residue correlates with the absence and presence, respectively, of Fe-S cluster binding signatures in eukaryotic glutamine phosphoribosylpyrophosphate amidotransferases (PUR1).Weblogo depictions for the C-terminal 20 amino acids of 570 yeast-like Ade4 proteins, which do not bind a [4Fe-4S] cluster, of 205 plant PUR1 proteins, which due to their plastidial localization lack a Cterminal W/F signature, and of 468 human-like PUR1 proteins, which based on conservation of 4 cysteine residues bind a [4Fe-4S] cluster.Sequences were collected via OrthoDB (19) and sorted according to their amino acid sequence identities with yeast, human and plant PUR1/Ade4 proteins and manually annotated for the Fe-S binding ligands.(B).Absence (top) and presence (bottom) of a C-terminal W or F residue correlates with the absence and presence, respectively, of Fe-S cluster binding signatures in eukaryotic RNA polymerases.WebLogo depictions for the C-terminal 20

Fig. S4 .
Fig. S4.Non-cropped images of droptests shown in Figure 2A-D (yellow boxes illustrate data in main text) along with controls showing growth under permissive conditions including: in the presence of glucose (A and C); in the presence of leucine (D); or in the absence of gallobenzophenone (B).For Nar1 (C) the strong MET25 or the weaker NBP35 and NAR1 promoters were used to drive expression of the WT and Δ1 inserts.For Leu1 (D) and Nar1 (C), the weak (RET2, NBP35 and RPL18B), moderate (LEU1) and strong (MET25 and TDH3) promoters were used.In both cases, the stronger promoters can mask the growth defect observed for the Δ1 variants in comparison to the weaker promoters.
Fig. S6.Western blot of cell free extracts of the indicated yeast strains transformed with pRS416 plasmids expressing wild type (WT) Leu1 and its C-terminally truncated variants from its natural promoter (corresponding to Fig.2E).Rabbit polyclonal antibodies raised against purified yeast Leu1 were used.

Fig. S7 .
Fig. S7.The loss of activity upon removal of W779 is not due to regulatory phenomena related to LEU1 promoter or terminator regions.Specific activities (using the regular assay) of wild-type and the Δ1 truncation of yeast Leu1 isopropylmalate isomerase in cell extracts from Δleu1 cells transformed with pRS416 plasmids with LEU1 under control of the indicated promoter or terminator.

Fig. S9 .BFig. S10 . 15 Fig. S11 .
Fig. S9.Full spot tests corresponding to Fig. 3D (yellow box).The left panel is a control plate onto which the same cell suspensions were spotted onto solid media supplemented with leucine.The doubling times measured in leucine deficient liquid medium are indicated on the right.

Fig. S13 .
Fig. S13.Somatic mutations associated with cancer for cytosolic and nuclear Fe-S proteins carrying a TCR.For six proteins mutations occur, which lead to loss of the TCR functionality (substitution in the tripeptide, of the stopcodon or C-terminal truncation of up to 20 amino acids; red star).The outputs of the COSMIC database (https://cancer.sanger.ac.uk/cosmic) depict the mutations in form of a histogram at the level of amino acids (numbering at the top, amino acid in the wild type protein indicated at the bottom).Grey rectangles correspond to nonsense mutations.

Table S1 .
Overview of failed expression of active Fe-S enzymes in heterologous hosts for biotechnological purposes.

Table S2 .
Proteomic data and structural data for TCR-containing proteins demonstrating that the TCR motif is not processed after Fe-S protein maturation.

Table S3 .
Mutagenesis primers for plasmids relating to in vitro interaction studies.

Table S6 .
Primers used for cloning and mutagenesis.Restriction sites are in bold.ATT AAC TAG TGA AAA ACA ACG TCC Leu1_mut_W779Y_rev TTC ACT AGT TAA TAA TCC TGG TGG ACT TTA TCG Leu1_C424A_for CTG GTT GTT CAA TAG CTT TAG GTA TGA ACC CTG Leu1_C424A_rev CCT AAA GCT ATT GAA CAA CCA GCT TCT CTC CAT Leu1_C421A/C424A_for CTG GTG CTT CAA TAG CTT TAG GTA TGA ACC CTG Leu1_C421A/C424A_rev CCT AAA GCT ATT GAA GCA CCA GCT TCT CTC CAT Leu1_D3_pET_for GTC CAC TAG TAT TGG TAG CTC GAG GAT CCG GC Leu1_D3_pET_rev GCT ACC AAT ACT AGT GGA CTT TAT CGA AAG TAG Leu1_1849_BglIIf GCT GAG ATC TTG GTT GTT ACT GGT GAC AAT TTC G Leu1_1849_BglIIr CAG TAA CAA CCA AGA TCT CAG CTT CCC TCC AAG G