Regulation of Cajal bodies: Unexpected connection to 60S ribosomes

Does the ribosome regulate the production of spliceosomes? In this work (1), Arias Escayola and Zhang et al. present new data identifying novel components of Cajal bodies and show how those components affect Cajal body assembly. A surprising observation is that the depletion of some 60S ribosomal proteins (RPLs) perturb the normal numbers of Cajal bodies and levels of dimethylarginine (DMA) modifications. This identifies a new, and unexpected, connection between ribosomes and Cajal bodies.

The Cajal body is a nuclear condensate organelle that forms at the site of transcription, processing, and assembly for many small nuclear and nucleolar ribonucleoproteins (snRNPs and snoRNPs) (2, 3, 4). Some of these RNPs go on to participate in the spliceosome and ribosome biogenesis in the nucleolus, while others have roles elsewhere in nuclear RNA metabolism.

The Cajal body is an attractive research subject—it was first observed over 100 years ago by its namesake Santiago Ramón y Cajal (5). Under the microscope it can be nearly as prominent as the nucleolus in cells such as neurons or embryos, and genetic studies have suggested it plays a special role in development: depletion of its essential proteins manifests as a developmental block in fish and mice (6, 7). Cajal bodies are an example of transcriptionally nucleated condensates that have evolved to enhance the processing and assembly of nuclear RNPs, though we do not know precisely why these functions are so indispensable during early development.

To study a cellular structure, one must know its composition. Determining bona fide protein and RNA components of condensates is difficult because resident molecules partition and rapidly exchange between the condensed and dilute phases (8). Arias Escayola and Zhang et al. (1) have tackled this problem by using the APEX2 proximity labeling method with carefully chosen controls. Coilin is the canonical scaffold of the Cajal body and possesses an N-terminal oligomerization domain, which is required for coilin to enter and form Cajal bodies (9, 10). By conjugating the labeling enzyme to coilin, both with and without its N-terminal domain, they can distinguish true Cajal body components from those that also bind to coilin in the dilute nucleoplasm. Thus, the authors identified several new proteins that are associated with coilin and could be Cajal body components.

Importantly, the authors then examined how the depletion of their hits affects Cajal body formation and morphology. Using a systematic knockdown screening approach, they describe three effects of depletion: (1) a full or partial loss of Cajal bodies; (2) an increase in the number of Cajal bodies; and (3) fusion of Cajal body components with two other nuclear organelles, gems or the nucleolus.

The first observation has an important implication for how we understand the assembly of Cajal bodies and other transcriptionally nucleated structures. Though some molecules are required because they scaffold Cajal bodies (i.e., coilin), depleting proteins involved in RNA transcription (e.g., POLR2A) or snRNP components (e.g., SNRPD3) may induce disassembly because the processes of RNA synthesis and snRNP assembly appear to be requirements for Cajal body formation (2, 11).

A surprising observation was that depletion of certain protein components of the 60S ribosome, or a defect in 60S biogenesis, increased Cajal body number up to almost twofold (Fig. 1). Digging deeper, the authors attribute this effect to a reduction in RPL24 since it is reduced in all RPL knockdowns with an impact on Cajal bodies. Since other defects in 60S subunits can cause stronger translation effects than RPL24 but have weaker effects on Cajal body number, the parsimonious model is that RPL24 limits Cajal bodies independently of its translation function.

Reduced RPL protein levels also led to the merging of coilin-rich Cajal bodies and survival of motor neuron protein (SMN)-rich gems. The fusion of these two domains is probably due to increased levels of DMA modification activity since DMA increases with siRNA KD of RPL proteins. Moreover, prior work has shown that appropriate DMA modification is required to maintain the distinct Cajal body and gems domains (12). The mechanism by which RPL24 reduction alters DMA levels is not yet known. If RPL24 reduction broadly increases DMA, one anticipates other regulatory changes since DMA modification affects many biological pathways (13).

The merging of Cajal bodies and gems is an example of how related condensates can “collapse” into one another depending on the balance of homotypic interactions within and heterotypic interactions between organelles (8). Interestingly, the knockdown of 11 other proteins led to coilin relocalization to the nucleolus, which might reflect a similar merger of nucleoli and Cajal bodies. Understanding how the fusion of these condensate domains affects snRNA, snoRNA, and rRNA processing may yield insight into the precise reason for which they are normally separated.

The mechanism by which RPL24 knockdown leads to increased Cajal bodies is unclear. Though the mass spectrometry evidence suggests that RPL24 interacts with coilin, examination of GFP-tagged RPL proteins suggests they are not present in Cajal bodies. One possibility is that some RPL proteins, or a surface of the nascent 60S subunit, could interact in the nucleoplasm with the N-terminal domain of coilin and thereby limit coilin self-assembly and Cajal body formation. Alternatively, the increase in Cajal bodies could be due to changes in DMA levels, cell signaling, expression of key Cajal body components, or even increases in snRNA or snoRNA transcription. An exciting area of future work will be to determine the range of consequences and their mechanisms, following RPL24 reduction.

Small RNAs can sense ribosome biogenesis, so it is fitting that the components of ribosomes might influence snRNA or snoRNA biogenesis. For example, SNORA13 promotes senescence by interacting with a large ribosomal subunit (RPL23) and leads to the accumulation of free ribosomes (14). The authors suggest that feedback between different biogenesis pathways may be more pervasive than previously recognized, allowing the widespread coordination of gene expression components. An important future question will be to determine if RPL24 reduction leads to changes in snRNA/snoRNA biogenesis, which could explain the increase in Cajal bodies.

It is tempting to speculate that Arias Escayola and Zhang et al. (1) have identified an unexpected signaling pathway between the ribosome and the Cajal body. Their datasets expand the list of Cajal body components and assign a role in Cajal body morphology to each new protein. Looking forward, we anticipate future work elucidating mechanisms and consequences of the link between the 60S ribosome, DMA modification, and Cajal body abundance.