The function of Mak16 in ribosome biogenesis depends on its [4Fe-4S] cluster

Abstract

Mak16 and its interacting partner Rpf1 play a critical role at an early step in the maturation of the ribosomal 60S subunit of eukaryotes, as revealed by cryoelectron microscopy structures. While these studies suggested no metal participation or the presence of a Zn²⁺ ion in Mak16, we identify a previously unexplored iron-sulfur (Fe/S) cluster in yeast Mak16 through both in vivo and in vitro methods. We demonstrate a functional link between mitochondrial and cytosolic Fe/S protein biogenesis and ribosome assembly, highlighting an overlooked aspect of 60S ribosomal biogenesis. Characterization of human and yeast Mak16 revealed a redox-active [4Fe-4S]^2+/1+ cluster with a midpoint potential below -500 mV. Oxidative stress destabilizes Mak16 and disrupts its interaction with Rpf1 in vivo, while in vitro H₂O₂ causes [3Fe-4S]¹⁺ cluster formation. Our findings also reveal that upon binding to rRNA expansion segment 7 the redox properties of the nearby Fe/S cluster largely remain unchanged. However, disruption of Fe/S cluster coordination destabilized Mak16, impaired the Mak16-Rpf1 complex formation and decreased the 25S rRNA level. The critical role of Fe/S proteins in eukaryotic DNA replication and repair, mitoribosomal function, and maturation has now been extended to nuclear ribosomal assembly. Relying on a vulnerable cofactor comes at a cost, as cluster loss can severely disrupt essential cellular processes. The inherent biosynthetic complexity and instability of the Fe/S cluster of Mak16 allows it to function as sensor for redox imbalance, creating the possibility to regulate cellular homeostasis under stress.

Keywords: iron–sulfur; metallocofactor; ribosome biogenesis.

Fig. 1.

Assembly of Fe/S clusters on Mak16 and efficient rRNA synthesis relies on the ISC and CIA machinery pathways. (A) Incorporation of ⁵⁵Fe in Mak16 in yeast strains expressing or depleted for ISC- or CIA-factors. The 416MET25-HA-Mak16 plasmid was transformed into W303, Gal-NFS1, Gal-NAR1, and Gal-CIA2 yeast cells, grown in iron-poor SC medium supplemented with galactose or glucose and ⁵⁵FeCl₃. HA-Mak16 was immunoprecipitated with anti-HA beads from cell extracts. Data are mean value ± SEM (n = 4). Significance of differences between groups, as indicated by crossbars, were determined using multiple comparisons one-way ANOVA with Šidák correction (*P < 0.05, ****P < 0.0001). (B) ⁵⁵Fe incorporation in Mak16 homologues expressed in W303 yeast cells in glucose containing SC medium as in (A). Abbreviations: S. cer., S. cerevisiae; H. sap., Homo sapiens; E. cun., Encephalitozoon cuniculi; T. bru., Trypanosoma brucei. ****P < 0.0001. (C) Total RNA isolated from W303 and yeast strains expressing or depleted for ISC- or CIA-factors. Yeast cells were cultured for 24 h in SC medium containing either galactose or glucose. Total RNA was then extracted from cells after two doubling times. n = 3, ***P < 0.001, **P < 0.01. (D) From the same samples used in (C) the RNA was separated by agarose gel electrophoresis and detected by GelRed staining.

Fig. 2.

Yeast and human Mak16 contain a [4Fe-4S]²⁺ cluster. (A) UV-Vis spectrum of anaerobically purified human Mak16 (0.037 mM, black line, as isolated) in 50 mM sodium phosphate, 300 mM NaCl, pH 8.0, and after incubation with 4 mM sodium dithionite (red line, reduced). Inset: Nonheme (Fe^2+/3+) and acid labile sulfide (S²⁻) contents (mean ± SEM, n = 5). (B) Mössbauer spectra at 77 K in the absence of an applied field of as isolated and partially reduced human Mak16 as in (A) but at ~0.4 mM concentration. Fits shown as lines: blue, [4Fe-4S]²⁺; orange/red, [4Fe-4S]¹⁺; black, sum of components; dotted gray, sum of components for as isolated for comparison. Parameters in SI Appendix, Table S1. (C) Compilation of Mössbauer parameters for all-cysteine coordinated [4Fe-4S]²⁺ (blue) and [4Fe-4S]¹⁺ clusters (orange, Fe^2.5+ pair; red Fe²⁺ pair; open for systems simulated with a single quadrupole doublet); Human Mak16, triangles; yeast Mak16, inverted triangles. Values and references in SI Appendix, Tables S2 and S4. (D) EPR spectra of 0.5 mM potassium ferricyanide oxidized or 4 mM sodium dithionite treated human Mak16 (0.128 mM) or yeast Mak16/Rpf1-Δ58 complex (0.074 mM). EPR conditions: 9.353 GHz, 10 K, microwave power 0.21 mW. Parameters for simulations (red lines) in SI Appendix, Table S3. (E) UV-Vis spectrum of anaerobically purified as isolated (black line), dithionite-reduced (red line), and aerobically purified yeast (dashed line) Mak16/Rpf1-Δ58 complex (0.011 mM), as in (A). Fe^2+/3+ and S^2- contents for anaerobically purified complex (mean ± SEM, n = 3). (F) Mössbauer spectroscopy of anaerobically purified as isolated and reduced yeast Mak16/Rpf1-Δ58 complex (~0.6 mM) as in (B).

Fig. 3.

Disruption of [4Fe-4S] coordination in Mak16 leads to impaired cell growth and rRNA synthesis. (A, Top) conserved cysteine residues and domain structure of Mak16 in comparison with human L28e. (Bottom) Structural alignment of yeast Mak16 (PDB 6EM1) with the human ribosomal protein L28e (PDB 8G5Y). (B) Gal-MAK16 yeast cells, transformed with 416 plasmids under control of the MAK16 promoter lacking an insert (empty), or encoding wild-type (WT) or cysteine variants of Mak16, were cultured for 16 h in SC medium with glucose. OD-normalized 10-fold serial dilutions were spotted on agar plates of the same medium and photographed after 48 h at 30 °C. (C) Northern blots of total RNA isolated from wild-type (W303) and Gal-MAK16 yeast cells cultivated for 40 h in SC medium with galactose (Gal) or glucose (Glu). rRNA was visualized with digoxigenin-ddUTP labeled 25S (Left) or 18S (Right) probes and anti-digoxigenin-peroxidase conjugate (SI Appendix, Fig. S10). (D) Quantification of total RNA isolated from cells grown for 24 h in galactose or glucose containing SC medium, followed by ~6 h of growth to the logarithmic phase. Data are mean value ± SEM (n = 3). Significance of differences between groups, as indicated by crossbars, were determined using multiple comparisons one-way ANOVA with Šidák correction (ns, not significant, *P < 0.05, ***P < 0.001, ****P < 0.0001). (E) Total RNA from samples in (D) was separated by agarose gel electrophoresis. rRNA bands were visualized with ethidium bromide (SI Appendix, Fig. S11).

Fig. 4.

Redox properties of Mak16. (A) EPR-mediated redox titration. Human Mak16 (16 µM) in pyrophosphate buffer pH 9.0 was mixed with redox mediators, and the reduction potential was adjusted with sodium dithionite. EPR spectra for samples taken at indicated potentials are shown. EPR conditions: 9.356 GHz, 10 K, microwave power 0.21 mW. (B) EPR-mediated redox titration of yeast Mak16/Rpf1-Δ58 complex (9 µM) as in (A). (C) Binding of the Mak16/Rpf1 complex to ES7. Purified Mak16/Rpf1-Δ58 was titrated with a fixed amount of ES7 (0.8 µg) in binding buffer. The protein/RNA complex was separated by 5% nondenaturing acrylamide gel electrophoresis, visualized with GelRed and recorded with a gel documentation system. The same gel was further stained with Coomassie dye (SI Appendix, Fig. S12). (D) Effect of ES7 binding on the redox potential of Mak16/Rpf1 complex. Purified Mak16/Rpf1-Δ58 complex and ES7 (7.3 µM each) were incubated in binding buffer, and subjected to a redox titration as in (A).

Fig. 5.

Effect of redox stressors on wild type Mak16, its C38A variant and Rpf1 binding. (A) Synthetic lethality between CIA machinery depletion and the Mak16 C38A variant. Indicated yeast strains were transformed with an empty, WT Mak16, or C38A Mak16 variant encoding plasmid, along with an empty plasmid or plasmid encoding the indicated CIA factor. After 16 h of growth on galactose or glucose, 10-fold serial dilutions were spotted on SC agar galactose or glucose plates and photographed after 48 h at 30 °C. (B) Disruption of Fe/S coordination in Mak16 impairs its interaction with Rpf1. W303 without plasmids or the Gal-MAK16/Gal-RPF1 strain transformed with plasmids encoding WT or cysteine variants of HA-Mak16 and Myc-Rpf1 were grown for 40 h on glucose. Immunoprecipitates of cell extracts with either HA or Myc beads were subjected to western blot analysis with indicated antibodies (SI Appendix, Fig. S13). (C) Conditional lethality caused by expression of the Mak16 C38A variant and treatment with redox stressors. Gal-MAK16 cells transformed with an empty plasmid, or a plasmid encoding WT, or the C38A Mak16 variant were grown on SC glucose for 16 h, then treated 2 h with indicated redox stressors, and 10-fold serial dilutions were spotted on SC glucose agar plates. After 48 h of growth at 30 °C plates were photographed. (D) Disruption of Fe/S of WT Mak16 by stressors in vivo impairs its stability and interaction with Rpf1. Gal-MAK16/Gal-RPF1 cells transformed with empty plasmids or plasmids encoding WT HA-Mak16 and Myc-Rpf1 were grown and processed as in (B). Full blots in SI Appendix, Fig. S14. (E) Relative intensities of the quadrupole doublets at 77 K in the Mössbauer spectra (SI Appendix, Fig. S15) of purified ⁵⁷Fe-enriched Mak16/Rpf1-Δ58 after treatment with indicated compounds (30 min., 23 °C). The sample used had a higher [3Fe-4S]¹⁺ content than that in Fig. 2F. (F) EPR spectra of purified Mak16/Rpf1-Δ58 frozen after treatment with indicated compounds (30 min, 23 °C). EPR conditions: 9.352 GHz, 10 K, microwave power 0.21 mW.