The commander complex. [New image from Institute for Molecular Bioscience The University of Queensland]
Machine learning algorithms, such as Alphabet’s neural network-based model AlphaFold2, are steadily transforming medical research, shedding light on complex biological structures. A recent case in point involves research using the technology to explore the Commander complex, a 16-protein complex crucial for cellular protein transport processes. This complex is not only vital for normal cellular function, but it’s also associated with several diseases, making it a potential target for novel therapies. Scientists at the Universities of Bristol and Queensland (Australia) and the Medical Research Council Laboratory of Molecular Biology in Cambridge, collaborated on the research.
The renowned journal Cell published the study, titled “Structure of the endosomal Commander complex linked to Ritscher-Schinzel syndrome.”
Navigating the intricacies of the commander complex
University of Bristol Professor Peter J. Cullen, a lead author of the study, outlined the challenges they encountered and the role of AI in surmounting these: “We first needed to develop insect cell-based multigene expression technology to purify stable recombinant human Retriever and Cullin-RING-based BCR3-Cul3Ubiquitin ligase (CCC) complexes in sufficient amounts for structural analysis. The next challenge came from the resolution of these complexes by by cryo-electron microscopy (cryo-EM). This is where AI, in the form of AlphaFold2, provided invaluable assistance.”
Eventually, they were able to determine a high-resolution structure of the core COMMD ring of the dodecameric CCC complex and how this associated with the amino-terminal regions of CCDC22 and CCDC93. But they were unable to resolve the extensive coiled-coil regions of CCDC22/CCDC93 — “presumably because of flexibility within the dimeric coiled-coils,” Cullen added. “Furthermore, the 3D resolution of the retriever complex was limited because of an issue with preferred orientation.”
AlphaFold2 continues to make strides in bioinformatics
Since it was launched in November 2020, the protein structure prediction system AlphaFold2 has found an expanding set of applications in bioinformatics. Capable of predicting protein sequences with near-experimental accuracy, AlphaFold2 allows researchers to explore the form and function of proteins, helping unravel the intricacies of biological processes and disease mechanisms.
In the commander complex research, AlphaFold2 initially facilitated deriving a model of the Retriever heterotrimer, a component of the protein complex. The Retriever heterotrimer, an assembly of three different proteins, forms part of the backbone of the commander complex. This prediction by AlphaFold2 was a significant breakthrough, as it “nicely fitted into our low resolution cryo-EM envelope,” Cullen explained.
“AlphaFold2 played a crucial role in overcoming these challenges. As Cullen explained, ‘AlphaFold2 predicted not only the structure of the COMMD ring and the association with CCDC22 and CCDC93 (beautifully consistent with our cryo-EM structure), but also the overall organisation of the coiled-coil regions of the CCDC22/CCDC93 heterodimer. From this, AlphaFold2 predicted how Retriever bound to the carboxy-terminal coiled-coil regions of CCDC22/CCDC93 to assemble these sub-complexes. Finally, AlphaFold2 predicted how an additional subunit DENND10, associated to form the complete sixteen subunit Commander complex.” DENND10 is another protein involved in the 16-subunit Commander complex
Commander complex and disease implications
The commander complex is involved in normal cellular functioning but also plays a significant role in several diseases. One example is Ritscher-Schinzel syndrome, also known as cranio-cerebello-cardiac (3C) syndrome. This disorder involves growth delays, intellectual disability and heart defects. Mutations in the commander complex are responsible for the disease as they disrupt normal cellular protein transport. The rare disease affects about one in 1,000,000 people.
“Ritscher-Schinzel syndrome is a multi-system developmental disorder,” Cullen said. “The commander machine functions to orchestrate the sorting and transport of key integral membrane proteins through the endosomal network.” Scientists know of an expanding number of mutations in the protein complex that are behind the disease. “We consider that the disease therefore principally stems from a defect in the efficiency of sorting and transport of key integral membrane proteins for the development of the tissues and organs affected in the syndrome,” Cullen added.
Verifying the Ritscher-Schinzel model
Verifying the functional aspects of their model was another vital aspect of the study. “A key feature of our study is the extensive utilization of site-directed mutagenesis to quantify effects on the assembly of the complexes which we complemented through validating key interactions by atomic-level X-ray crystallography,” Cullen said.
The implications of understanding the commander complex extend beyond Ritscher-Schinzel syndrome. “We now know what the commander complex looks like, how it is assembled and how these features are conserved through eukaryotic evolution,” Cullen said. “This gives us, and others, fantastic information for formulating hypotheses as to its functional role within these disease contexts. We are now actively studying some of these questions.”
AI in molecular biology research poised to continue gaining ground
The study exemplifies the potential of AI in molecular biology research. Cullen further explained this point by adding, “AlphaFold2 is providing a revolution in our understanding of protein structure and the structure and assembly of multiprotein complexes.”
“We consider that our analysis of the commander complex serves as an example of how AI can be incorporated into classical biochemical and cell-based experimentation to provide exciting new insight into the structure of such complexes.”
Indeed, we and others have published initial findings to this effect,” Cullen noted, highlighting the far-reaching implications of their research.