Letter | Published:

Rubisco condensate formation by CcmM in β-carboxysome biogenesis

Naturevolume 566pages131135 (2019) | Download Citation


Cells use compartmentalization of enzymes as a strategy to regulate metabolic pathways and increase their efficiency1. The α- and β-carboxysomes of cyanobacteria contain ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco)—a complex of eight large (RbcL) and eight small (RbcS) subunits—and carbonic anhydrase2,3,4. As HCO3 can diffuse through the proteinaceous carboxysome shell but CO2 cannot5, carbonic anhydrase generates high concentrations of CO2 for carbon fixation by Rubisco6. The shell also prevents access to reducing agents, generating an oxidizing environment7,8,9. The formation of β-carboxysomes involves the aggregation of Rubisco by the protein CcmM10, which exists in two forms: full-length CcmM (M58 in Synechococcus elongatus PCC7942), which contains a carbonic anhydrase-like domain8 followed by three Rubisco small subunit-like (SSUL) modules connected by flexible linkers; and M35, which lacks the carbonic anhydrase-like domain11. It has long been speculated that the SSUL modules interact with Rubisco by replacing RbcS2,3,4. Here we have reconstituted the Rubisco–CcmM complex and solved its structure. Contrary to expectation, the SSUL modules do not replace RbcS, but bind close to the equatorial region of Rubisco between RbcL dimers, linking Rubisco molecules and inducing phase separation into a liquid-like matrix. Disulfide bond formation in SSUL increases the network flexibility and is required for carboxysome function in vivo. Notably, the formation of the liquid-like condensate of Rubisco is mediated by dynamic interactions with the SSUL domains, rather than by low-complexity sequences, which typically mediate liquid–liquid phase separation in eukaryotes12,13. Indeed, within the pyrenoids of eukaryotic algae, the functional homologues of carboxysomes, Rubisco adopts a liquid-like state by interacting with the intrinsically disordered protein EPYC114. Understanding carboxysome biogenesis will be important for efforts to engineer CO2-concentrating mechanisms in plants15,16,17,18,19.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Data availability

The crystallographic models and structure factors for SSUL1ox and SSUL1red have been deposited to wwPDB under accession codes 6HBA (oxidized) and 6HBB (reduced), respectively. The electron density reconstructions and final 2RbcL–2RbcS–SSUL model have been deposited in the Electron Microscopy Data Bank (EMDB) and wwPDB under accession codes EMD-0180 and 6HBC, respectively. Source data for graphs in Figs. 13 are provided with the online version of the paper and the source data for the gels shown in Extended Data Fig. 5b are provided in Supplementary Fig. 1.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


  1. 1.

    Hinzpeter, F., Gerland, U. & Tostevin, F. Optimal compartmentalization strategies for metabolic microcompartments. Biophys. J. 112, 767–779 (2017).

  2. 2.

    Espie, G. S. & Kimber, M. S. Carboxysomes: cyanobacterial RubisCO comes in small packages. Photosynth. Res. 109, 7–20 (2011).

  3. 3.

    Rae, B. D., Long, B. M., Badger, M. R. & Price, G. D. Functions, compositions, and evolution of the two types of carboxysomes: polyhedral microcompartments that facilitate CO2 fixation in cyanobacteria and some proteobacteria. Microbiol. Mol. Biol. Rev. 77, 357–379 (2013).

  4. 4.

    Kerfeld, C. A. & Melnicki, M. R. Assembly, function and evolution of cyanobacterial carboxysomes. Curr. Opin. Plant Biol. 31, 66–75 (2016).

  5. 5.

    Dou, Z. et al. CO2 fixation kinetics of Halothiobacillus neapolitanus mutant carboxysomes lacking carbonic anhydrase suggest the shell acts as a diffusional barrier for CO2. J. Biol. Chem. 283, 10377–10384 (2008).

  6. 6.

    Price, G. D. & Badger, M. R. Expression of human carbonic anhydrase in the cyanobacterium Synechococcus PCC7942 creates a high CO2-requiring phenotype: evidence for a central role for carboxysomes in the CO2 concentrating mechanism. Plant Physiol. 91, 505–513 (1989).

  7. 7.

    Price, G. D., Coleman, J. R. & Badger, M. R. Association of carbonic anhydrase activity with carboxysomes isolated from the cyanobacterium Synechococcus PCC7942. Plant Physiol. 100, 784–793 (1992).

  8. 8.

    Peña, K. L., Castel, S. E., de Araujo, C., Espie, G. S. & Kimber, M. S. Structural basis of the oxidative activation of the carboxysomal γ-carbonic anhydrase, CcmM. Proc. Natl Acad. Sci. USA 107, 2455–2460 (2010).

  9. 9.

    Chen, A. H., Robinson-Mosher, A., Savage, D. F., Silver, P. A. & Polka, J. K. The bacterial carbon-fixing organelle is formed by shell envelopment of preassembled cargo. PLoS One 8, e76127 (2013).

  10. 10.

    Long, B. M., Badger, M. R., Whitney, S. M. & Price, G. D. Analysis of carboxysomes from Synechococcus PCC7942 reveals multiple Rubisco complexes with carboxysomal proteins CcmM and CcaA. J. Biol. Chem. 282, 29323–29335 (2007).

  11. 11.

    Ludwig, M., Sültemeyer, D. & Price, G. D. Isolation of ccmKLMN genes from the marine cyanobacterium, Synechococcus sp. PCC7002 (Cyanophyceae), and evidence that CcmM is essential for carboxysome assembly. J. Phycol. 36, 1109–1119 (2000).

  12. 12.

    Hyman, A. A., Weber, C. A. & Jülicher, F. Liquid–liquid phase separation in biology. Annu. Rev. Cell Dev. Biol. 30, 39–58 (2014).

  13. 13.

    Banani, S. F., Lee, H. O., Hyman, A. A. & Rosen, M. K. Biomolecular condensates: organizers of cellular biochemistry. Nat. Rev. Mol. Cell Biol. 18, 285–298 (2017).

  14. 14.

    Freeman Rosenzweig, E. S. et al. The eukaryotic CO2-concentrating organelle is liquid-like and exhibits dynamic reorganization. Cell 171, 148–162.e19 (2017).

  15. 15.

    Hanson, M. R., Lin, M. T., Carmo-Silva, A. E. & Parry, M. A. Towards engineering carboxysomes into C3 plants. Plant J. 87, 38–50 (2016).

  16. 16.

    Ort, D. R. et al. Redesigning photosynthesis to sustainably meet global food and bioenergy demand. Proc. Natl Acad. Sci. USA 112, 8529–8536 (2015).

  17. 17.

    Price, G. D. et al. The cyanobacterial CCM as a source of genes for improving photosynthetic CO2 fixation in crop species. J. Exp. Bot. 64, 753–768 (2013).

  18. 18.

    Zarzycki, J., Axen, S. D., Kinney, J. N. & Kerfeld, C. A. Cyanobacterial-based approaches to improving photosynthesis in plants. J. Exp. Bot. 64, 787–798 (2013).

  19. 19.

    Long, B. M. et al. Carboxysome encapsulation of the CO2-fixing enzyme Rubisco in tobacco chloroplasts. Nat. Commun. 9, 3570 (2018).

  20. 20.

    Bracher, A., Starling-Windhof, A., Hartl, F. U. & Hayer-Hartl, M. Crystal structure of a chaperone-bound assembly intermediate of form I Rubisco. Nat. Struct. Mol. Biol. 18, 875–880 (2011).

  21. 21.

    Hauser, T. et al. Structure and mechanism of the Rubisco-assembly chaperone Raf1. Nat. Struct. Mol. Biol. 22, 720–728 (2015).

  22. 22.

    Aigner, H. et al. Plant RuBisCo assembly in E. coli with five chloroplast chaperones including BSD2. Science 358, 1272–1278 (2017).

  23. 23.

    Cuevas-Velazquez, C. L. & Dinneny, J. R. Organization out of disorder: liquid–liquid phase separation in plants. Curr. Opin. Plant Biol. 45 (Pt A), 68–74 (2018).

  24. 24.

    Wang, J. et al. A molecular grammar governing the driving forces for phase separation of prion-like RNA binding proteins. Cell 174, 688–699.e16 (2018).

  25. 25.

    Alberti, S. et al. A user’s guide for phase separation assays with purified proteins. J. Mol. Biol. 430, 4806–4820 (2018).

  26. 26.

    Li, P. et al. Phase transitions in the assembly of multivalent signalling proteins. Nature 483, 336–340 (2012).

  27. 27.

    Banani, S. F. et al. Compositional control of phase-separated cellular bodies. Cell 166, 651–663 (2016).

  28. 28.

    Cai, F. et al. Advances in understanding carboxysome assembly in Prochlorococcus and Synechococcus implicate CsoS2 as a critical component. Life (Basel) 5, 1141–1171 (2015).

  29. 29.

    Catanzariti, A. M., Soboleva, T. A., Jans, D. A., Board, P. G. & Baker, R. T. An efficient system for high-level expression and easy purification of authentic recombinant proteins. Protein Sci. 13, 1331–1339 (2004).

  30. 30.

    Liu, C. et al. Coupled chaperone action in folding and assembly of hexadecameric Rubisco. Nature 463, 197–202 (2010).

  31. 31.

    Saschenbrecker, S. et al. Structure and function of RbcX, an assembly chaperone for hexadecameric Rubisco. Cell 129, 1189–1200 (2007).

  32. 32.

    Baker, R. T. et al. Using deubiquitylating enzymes as research tools. Methods Enzymol. 398, 540–554 (2005).

  33. 33.

    Riddles, P. W., Blakeley, R. L. & Zerner, B. Reassessment of Ellman’s reagent. Methods Enzymol. 91, 49–60 (1983).

  34. 34.

    Brinker, A. et al. Dual function of protein confinement in chaperonin-assisted protein folding. Cell 107, 223–233 (2001).

  35. 35.

    Maharana, S. et al. RNA buffers the phase separation behavior of prion-like RNA binding proteins. Science 360, 918–921 (2018).

  36. 36.

    Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods 9, 676–682 (2012).

  37. 37.

    Gupta, A. J., Haldar, S., Miličić, G., Hartl, F. U. & Hayer-Hartl, M. Active cage mechanism of chaperonin-assisted protein folding demonstrated at single-molecule level. J. Mol. Biol. 426, 2739–2754 (2014).

  38. 38.

    Woodger, F. J., Badger, M. R. & Price, G. D. Sensing of inorganic carbon limitation in Synechococcus PCC7942 is correlated with the size of the internal inorganic carbon pool and involves oxygen. Plant Physiol. 139, 1959–1969 (2005).

  39. 39.

    Price, G. D. & Badger, M. R. Ethoxyzolamide inhibition of CO2-dependent photosynthesis in the cyanobacterium Synechococcus PCC7942. Plant Physiol. 89, 44–50 (1989).

  40. 40.

    Long, B. M., Tucker, L., Badger, M. R. & Price, G. D. Functional cyanobacterial β-carboxysomes have an absolute requirement for both long and short forms of the CcmM protein. Plant Physiol. 153, 285–293 (2010).

  41. 41.

    Price, G. D. & Badger, M. R. Evidence for the role of carboxysomes in the cyanobacterial CO2-concentrating mechanism. Can. J. Bot. 69, 963–973 (1991).

  42. 42.

    Price, G. D., Sültemeyer, D., Klughammer, B., Ludwig, M. & Badger, M. R. The functioning of the CO2 concentrating mechanism in several cyanobacterial strains: a review of general physiological characteristics, genes, proteins, and recent advances. Can. J. Bot. 76, 973–1002 (1998).

  43. 43.

    Wyatt, P. J. Light scattering and the absolute characterization of macromolecules. Anal. Chim. Acta 272, 1–40 (1993).

  44. 44.

    Evans, P. Scaling and assessment of data quality. Acta Crystallogr. D Biol. Crystallogr. 62, 72–82 (2006).

  45. 45.

    Evans, P. R. & Murshudov, G. N. How good are my data and what is the resolution? Acta Crystallogr. D Biol. Crystallogr. 69, 1204–1214 (2013).

  46. 46.

    French, G. & Wilson, K. On the treatment of negative intensity observations. Acta Crystallogr. A 34, 517–525 (1978).

  47. 47.

    Potterton, E., Briggs, P., Turkenburg, M. & Dodson, E. A graphical user interface to the CCP4 program suite. Acta Crystallogr. D Biol. Crystallogr. 59, 1131–1137 (2003).

  48. 48.

    Vagin, A. & Teplyakov, A. Molecular replacement with MOLREP. Acta Crystallogr. D Biol. Crystallogr. 66, 22–25 (2010).

  49. 49.

    Emsley, P. & Cowtan, K. Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 (2004).

  50. 50.

    Murshudov, G. N. et al. REFMAC5 for the refinement of macromolecular crystal structures. Acta Crystallogr. D Biol. Crystallogr. 67, 355–367 (2011).

  51. 51.

    Chen, V. B. et al. MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr. D Biol. Crystallogr. 66, 12–21 (2010).

  52. 52.

    Gouet, P., Courcelle, E., Stuart, D. I. & Métoz, F. ESPript: analysis of multiple sequence alignments in PostScript. Bioinformatics 15, 305–308 (1999).

  53. 53.

    Mastronarde, D. N. Automated electron microscope tomography using robust prediction of specimen movements. J. Struct. Biol. 152, 36–51 (2005).

  54. 54.

    Zheng, S. Q. et al. MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nat. Methods 14, 331–332 (2017).

  55. 55.

    Scheres, S. H. RELION: implementation of a Bayesian approach to cryo-EM structure determination. J. Struct. Biol. 180, 519–530 (2012).

  56. 56.

    Kucukelbir, A., Sigworth, F. J. & Tagare, H. D. Quantifying the local resolution of cryo-EM density maps. Nat. Methods 11, 63–65 (2014).

  57. 57.

    Kremer, J. R., Mastronarde, D. N. & McIntosh, J. R. Computer visualization of three-dimensional image data using IMOD. J. Struct. Biol. 116, 71–76 (1996).

  58. 58.

    Hrabe, T. et al. PyTom: a python-based toolbox for localization of macromolecules in cryo-electron tomograms and subtomogram analysis. J. Struct. Biol. 178, 177–188 (2012).

Download references


We thank S. Gärtner, R. Lange, N. Wischnewski and L. Rourke for technical assistance; D. Balchin and R. H. Wilson for critically reading the manuscript; M. Strauss, J. Plitzko, F. Beck, S. Albert and Q. Guo (MPIB cryo-EM facility) for input on data collection and processing; the Centre for Advanced Microscopy (ANU) for assistance with electron transmission microscopy of cyanobacteria; and the staff at the MPIB Crystallization and Imaging facilities, and at the European Synchrotron Radiation Facility (ESRF) in Grenoble, France. This work was supported by a grant from the Deutsche Forschungsgemeinschaft (DFG) (SFB1035) to M.H.-H. and F.U.H., funding to G.D.P. from the Australian Government through the Australian Research Council Centre of Excellence for Translational Photosynthesis (CE1401000015), and the Minerva foundation of the Max Planck Society (M.H.-H.).

Reviewer information

Nature thanks J. Dinneny, T. Yeates and the other anonymous reviewer(s) for their contribution to the peer review of this work.

Author information

Author notes

    • H. Aigner

    Present address: KWS SAAT SE, Einbeck, Germany

  1. These authors contributed equally: H. Wang, X. Yan


  1. Department of Cellular Biochemistry, Max Planck Institute of Biochemistry, Martinsried, Germany

    • H. Wang
    • , X. Yan
    • , H. Aigner
    • , A. Bracher
    • , F. U. Hartl
    •  & M. Hayer-Hartl
  2. Australian Research Council Centre of Excellence for Translational Photosynthesis, Research School of Biology, The Australian National University, Canberra, Australian Capital Territory, Australia

    • N. D. Nguyen
    • , W. Y. Hee
    • , B. M. Long
    •  & G. D. Price


  1. Search for H. Wang in:

  2. Search for X. Yan in:

  3. Search for H. Aigner in:

  4. Search for A. Bracher in:

  5. Search for N. D. Nguyen in:

  6. Search for W. Y. Hee in:

  7. Search for B. M. Long in:

  8. Search for G. D. Price in:

  9. Search for F. U. Hartl in:

  10. Search for M. Hayer-Hartl in:


H.W. and X.Y., together with M.H.-H. and F.U.H., designed the experiments. H.W. performed most of the structural analyses by cryo-EM and tomography, including sample screening, grid preparation, data acquisition and single-particle/tomography reconstructions. X.Y. performed most of the biochemical experiments. H.A. cloned M35 and some of the SSUL modules, performed initial biochemical experiments, and crystallized SSUL1. A.B. solved the crystal structures of SSUL1ox and SSUL1red and supervised the structural analysis by H.W. G.D.P, B.M.L, W.Y.H. and N.D.N. designed the cyanobacterial experiments and generated the pHUE-SSUL1 plasmid, with data collection conducted by N.D.N., W.Y.H. and B.M.L. M.H.-H. conceived the project, participated in data interpretation and wrote the manuscript with contributions from F.U.H. and other authors.

Competing interests

The authors declare no competing interests.

Corresponding author

Correspondence to M. Hayer-Hartl.

Extended data figures and tables

  1. Extended Data Fig. 1 Crystal structure and sequence analysis of the SSUL domains.

    a, Ribbon representation of the asymmetric unit in the crystal lattice of the SSUL1 domain in the thiol-oxidized state. A view along the approximate dyad axis is shown. The two copies of SSUL1 with nearly identical conformations (with Cα r.m.s.d. 0.43 Å) are shown in magenta and cyan. Secondary structure elements and chain termini are indicated. b, SEC–MALS analysis of purified SSUL1. The red dotted line across the peak indicates the molar mass and homogeneity of the protein sample. The molar mass is indicated (calculated theoretical mass is 10,537.7 Da). A representative experiment is shown (n = 2 independent experiments). c, Sequence alignment of SSUL domains 1–3 in CcmM and RbcS from S. elongatus PCC7942. Amino acid sequences were aligned using the EBI Clustal-Ω server. Secondary structure elements for the SSUL1 domain are indicated above the sequences. Residues that are similar between SSUL (group 1) and RbcS (group 2) are boxed with blue frames; identical residues are shown in white on a red background. Triangles below the sequence indicate the contact residues of RbcS with the RbcL8 core in the Rubisco complex. The mutation site in the SSUL1 module (R251–R252) is indicated. The stars in magenta above the sequence indicate the contact residues of SSUL in the complex with Rubisco. Note that these residues are generally highly conserved in other β-cyanobacteria. The cysteine residues conserved in SSUL1 and SSUL2 are shown in bold with a yellow background. SSUL1 and SSUL2 share 84% identity, while SSUL3 is more divergent (58% identity with SSUL1). d, Surface properties of the SSUL1 domain. Hydrophobic side chains are indicated in yellow. Red and blue represent negative and positive charges, respectively. Dashed boxes indicate two areas of charge clusters. e, Ribbon representation of the asymmetric unit in the crystal lattice of the SSUL1 domain in the thiol-reduced state. A view along the approximate dyad axis is shown. The two copies of SSUL1 are shown in gold and blue, respectively. Secondary structure elements and chain termini are indicated. f, Overlay of SSUL1 structures in the oxidized (magenta) and reduced (gold) forms.

  2. Extended Data Fig. 2 Properties of the Rubisco–M35 network.

    a, SDS–PAGE of recombinantly expressed and purified proteins. A representative SDS gel is shown (n = 3 independent experiments). b, Rubisco carboxylation activity of the reactions shown in Fig. 1e, including RbcL8 core complexes and RbcS in the absence of M35. The activity of purified RbcL8S8 is set to 100%, equivalent to a catalytic constant (kcat) of ~4 s–1 per active site (dark grey bar). Data are mean ± s.d. (n = 3 independent experiments). c, Kinetics of M35-mediated network formation of Rubisco and Rubisco assembly intermediates by turbidity assay in buffer A at 25 °C. RbcL8 core complexes (0.25 µM) were incubated with RbcX from Anabaena sp. CA (8 µM), or the cognate Raf1 (2 µM) or BSD2 from Arabidopsis thaliana (4 µM) in buffer A for 15 min to generate the respective Rubisco assembly intermediates. M35 (2 µM) was then added to start the turbidity measurement at 340 nm. Turbidity of Rubisco with M35 is shown as control. A representative experiment is shown (n = 3 independent experiments). d, Kinetics of network formation of Rubisco (0.25 μM) with M35 (2 μM) or with combined M13-1, M13-2 and M13-3 (2 μM each) by turbidity assay in buffer A at 25 °C. A representative experiment is shown (n = 3 independent experiments). e, Kinetics of network formation by turbidity assay at 25 °C of unlabelled Rubisco (0.25 μM) with NT650–M35 (1 μM) and of unlabelled M35 (1 μM) with A532–Rubisco (0.25 μM). A reaction containing unlabelled Rubisco (0.25 μM) and unlabelled M35 (1 μM) is shown as control. A representative experiment is shown (n = 3 independent experiments). f, dcFCCS of 0.25 μM Rubisco (containing 2 nM A532–Rubisco) and 1 μM M35 (containing 2 nM NT650–M35). Measurements were taken for 30 min after 5 min incubation in buffer A. Cross-correlation curves are shown. After formation of the Rubisco–M35 network, unlabelled M13-2 (25 μM) was added and cross-correlation measured for 30 min. Auto-correlation curves of 0.25 μM Rubisco (containing 2 nM A532–Rubisco) and 1 μM M35 (containing 2 nM NT650–M35) are also shown. The diffusion coefficients (D) are indicated. The relatively slow rate of diffusion of M35 suggests that the protein is expanded and not as compact as would be expected for a protein of ~35 kDa. To exclude the possibility that the slow diffusion rate of M35 is due to interactions between M35 molecules, we also measured the auto-correlation of M35 at very low concentration (2 nM NT650–M35), where protein–protein interactions are unlikely, and found the same rate of diffusion. Data are mean ± s.d. (n = 3 independent experiments).

  3. Extended Data Fig. 3 Salt sensitivity of M35–Rubisco interaction.

    a, Kinetics of Rubisco network formation by turbidity assay at different salt concentrations at 25 °C. Rubisco (0.25 μM) in the presence of 2 μM M35red in buffer (50 mM Tris-HCl pH 8.0, 10 mM Mg(OAc)2, 5 mM DTT) containing different concentrations of KCl at 25 °C. A representative experiment is shown (n = 3 independent experiments). b, Kinetics of Rubisco network formation as in a at 50 mM and 150 mM KCl, and higher concentrations of Rubisco (0.5 or 1.0 μM) and M35 (4.0 or 8.0 μM). A representative experiment is shown (n = 3 independent experiments). c, Kinetics of Rubisco network formation by turbidity assay at different salt concentrations at 25 °C as in a but in the presence of 2 μM M35ox and in buffer without DTT. A representative experiment is shown (n = 3 independent experiments). d, M35 function is redox-sensitive. Network formation of Rubisco (0.25 μM) by M35ox (2 μM) was monitored in buffer (50 mM Tris-HCl pH 8.0, 10 mM Mg(OAc)2) containing 100 mM KCl for 5 min, followed by addition of 5 mM DTT for further 10 min. Rubisco (0.25 μM) with M35red (2 μM) in buffer containing 100 mM KCl and 5 mM DTT is shown as control. A representative experiment is shown (n = 3 independent experiments). e, f, Kinetics of Rubisco network formation by turbidity assay at different salt concentrations at 25 °C as in a in the presence of 2 μM reduced M24-1/2 (e) or M24-2/3 (f). A representative experiment is shown (n = 3 independent experiments).

  4. Extended Data Fig. 4 LLPS of Rubisco by M35red and M35ox.

    a, Size distribution of liquid droplets (n = 275) formed by unlabelled Rubisco (0.25 μM) and M35 (1 μM, containing 10% NT650–M35). The Feret’s diameter for the main peak (indicated by arrow) and average size are 1.33 µm and 1.32 µm, respectively. b, Time-lapse images of droplet fusion. Reactions containing labelled M35red (1 μM, containing 10% NT650–M35) and unlabelled Rubisco (0.25 μM) were observed over time. Droplets undergoing fusion are indicated by white arrowheads. Scale bars, 10 μm. A representative experiment is shown (n = 3 independent experiments). c, LLPS of Rubisco–M35ox (left) and Rubisco–M35red (right). NT650–M35ox or NT650–M35red was mixed with unlabelled M35ox or M35red, respectively, at a ratio of 1:10 (1 μM total) and imaged by fluorescence microscopy at 25 °C in combination with unlabelled Rubisco (0.25 μM). Scale bars, 10 μm. A representative experiment is shown (n = 3 independent experiments). d, Rubisco–M35red condensates were generated as in c and imaged by fluorescence and bright-field microscopy at 25 °C (panels 1 and 2). Dissociation of droplets was observed upon addition of unlabelled M13-2 (25 μM) (panels 3 and 4) to preformed droplets as in panels 1 and 2. Scale bars, 10 μm. A representative experiment is shown (n = 3 independent experiments).

  5. Extended Data Fig. 5 Role of disulfide bonds in SSUL in carboxysome function.

    a, Photosynthetic O2 evolution in response to external Ci by wild-type Se7942 and ccmM mutant strains. Shown is a representative set of Clarke-type oxygen electrode measurements of Ci-dependent O2 evolution by different strains of Se7942 from 10 µM to 250 mM Ci. Cysteine mutants (C279S, C395S and CcmM-4S) had an intermediate CO2-requiring phenotype, with CcmM-4S having the highest CO2 requirement. A representative experiment is shown (n = 3 biological replicates). b, Relative RbcL abundance in carboxysome-enriched pellet fractions in wild-type and ccmM mutant strains of Se7942. RbcL and CcmM proteins were detected using anti-RbcL and anti-M35 antibodies, respectively. A representative experiment is shown (n = 3 biological replicates). c, Length:width ratios of carboxysomes in wild-type and ccmM mutant strains. Carboxysomes were visualized using TEM, and their lengths and widths measured. These measurements were used to calculate the length:width ratio. Large increases in carboxysome length:width ratio were observed in cysteine mutants, particularly those expressing C279S and CcmM-4S. The numbers of carboxysomes analysed in the different strains are indicated. The horizontal line represents the mean of each data set. Whiskers indicate upper and lower data points of the range. The number of measured values (n) is indicated above each data set. d, Carboxysomes per cell section for wild-type and ccmM mutant strains analysed by TEM. Cell sections (wild-type n = 25; ΔccmM + ccmM n = 23; C279S n = 28; C395S n = 26; CcmM-4S n = 25) were analysed and the data expressed as a percentage of the number of carboxysomes in wild-type ± s.e.m. *P ≤ 0.05; ***P ≤ 0.001 (Tukey’s multiple comparisons test). P values indicated for wild-type and ΔccmM+ccmM cells versus cysteine mutants (wild-type versus CcmM-4S P = 0.0006, wild-type versus C395S P = 0.0008, ΔccmM+ccmM versus CcmM-4S P = 0.0397, ΔccmM+ccmM versus C395S P = 0.0464).

  6. Extended Data Fig. 6 Cryo-EM single-particle reconstruction.

    a, Representative micrograph of Rubisco–M35 complexes used for single-particle reconstruction (n = 4,723 micrographs from one EM grid). b, 2D class averages of complexes in a showing extra density of SSUL bound to Rubisco in some classes. c, Meshwork representation of unsharpened D4 symmetry-averaged electron density map of the Rubisco–M35 complex (at a contour level of 2.8σ, appropriate for Rubisco), with the D4-averaged final model in backbone trace representation superposed (top). Below, zoomed-in region. Note that there is only ~50% residual density for the two SSUL modules, except for the region in which the modules overlap in the centre (helices α1). d, The workflow for single-particle data processing. The left scheme of the flowchart identified only one SSUL domain bound between two RbcL subunits. The main scheme took the units consisting of two RbcL, two RbcS and two SSUL for focused classification and further refinement to improve the map quality. See Methods for details. e, Gold-standard FSC corrected curve (masked and B-factor sharpened) of the final 3D reconstruction. The resolution is ~2.77 Å at the FSC cutoff of 0.143.

  7. Extended Data Fig. 7 Cryo-EM single-particle reconstruction.

    a, Local resolution for the refined map of class v (2RbcL–2RbcS–SSUL) particles (see Extended Data Fig. 6d). The colour gradient from blue to red indicates local resolution from 2.0 to 4.0 Å. Right, zoomed-in views of the solvent-facing side and the buried interaction side of SSUL. As expected, the elements facing Rubisco have higher resolution. b, Quality of the cryo-EM density map of SSUL and the structural model of 2RbcL–2RbcS–SSUL in the vicinity of some aromatic side chains of SSUL1. The density is shown as a meshwork in cyan. The backbone of the structural model is in ribbon representation, and side chains are shown in stick representation. The cryo-EM map is shown at 1.2σ. c, d, Cryo-EM density map for SSUL at interface I (c) and interface II (d). c, Structural features of interface I. Critical interacting residues (see Fig. 4c): Phe253SSUL forms van der Waals contacts to His353RbcL-A/Glu355RbcL-A and Arg254SSUL forms interactions with Glu351RbcL-A and a backbone–backbone H bond with His353RbcL-A. The side-chain of Arg251SSUL is sandwiched between RbcS residues Gln36/Gly37 and Asn94/Ile95, and forms a salt bridge with Asp93RbcS. Arg252SSUL forms a salt bridge with Asp76RbcL-B. Thr255SSUL and Ser257SSUL form additional contacts with Arg79RbcL-B and Asp76RbcL-B. d, Structural features of interface II. Critical interacting residues (see Fig. 4d): Ile293SSUL contacts Thr30RbcL-B. Arg298SSUL makes van der Waals interactions with Tyr85RbcL-B and His86RbcL-B, and Arg300SSUL with Tyr29/Thr30/Pro31/Lys32 in RbcL-B. e, Circular dichroism wavelength scans of wild-type M35 and the mutant M35(R251D/R252D) (3.5 μM each) in buffer (50 mM KH2PO4 pH 7.5) measured at 25 °C with 0.1-cm cuvettes using a Jasco 715 CD spectrometer. f, Dependence of Rubisco–M35 interaction on interface I contacts and RbcS. RbcL8 (0.25 μM) and RbcS (2.5 μM) were mixed in buffer A and incubated for 10 min, followed by addition of M35 or M35(R251D/R252D) (1 μM each), and complex formation monitored by turbidity (black and green, respectively). The requirement for RbcS was analysed by addition of RbcS to premixed RbcL8 and M35. Arrow indicates time of RbcS addition. A representative experiment is shown (n = 3 independent experiments). g, Inability of M35 mutant R251D/R252D to induce phase separation of Rubisco. Unlabelled M35(R251D/R252D) (1 μM) was mixed with Rubisco (0.25 μM, containing 10% A532–Rubisco) and imaged by bright field (left) and fluorescence microscopy (right) at 25 °C (top). As control, unlabelled M35 (1 μM) was mixed with Rubisco (0.25 μM, containing 10% A532–Rubisco) (bottom). Scale bars, 10 μm. A representative experiment is shown (n = 3 independent experiments).

  8. Extended Data Fig. 8 Cryo-ET of Rubisco–35 and Rubisco–M24-1/2 condensates.

    a, Analysis of the Rubisco–M35 condensate by cryo-ET. A slice through a tomographic volume of the matrix (left) and Rubisco–SSUL particles, from single-particle reconstruction with C1 symmetry and low-pass filtered to 40 Å, rendered into a 3D matrix (right). The circled cluster of complexes is shown again in Fig. 4e. A representative tomogram is shown (n = 3 independent tomograms of the same biological sample). b, Distribution of nearest neighbour centre-to-centre distances between Rubisco particles in tomograms of Rubisco–M35 (2,654 particles analysed) and Rubisco–M24-1/2 (1,405 particles analysed). Data are from n = 3 tomograms for each sample. The schematic above the histogram shows the distance between two Rubisco particles relative to the maximum distance spanned by the linkers between SSUL modules in M35. c, Analysis of the Rubisco–M24-1/2 condensate by cryo-ET. A slice through a tomographic volume of the matrix (left) and Rubisco–SSUL particles, from single-particle reconstruction with C1 symmetry and low-pass filtered to 40 Å, rendered into a 3D matrix (right). A representative tomogram is shown (n = 3 independent tomograms of the same biological sample).

  9. Extended Data Table 1 X-ray crystallography, cryo-EM statistics and model validation
  10. Extended Data Table 2 Plasmids, proteins and strains used or generated for this study

Supplementary information

  1. Supplementary Information

    This file contains the full uncropped gels.

  2. Reporting Summary

Source data

About this article

Publication history




Issue Date




By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.