Crashing by design: utilizing DNA damage for MCC differentiation

Abstract

Multiciliated cells (MCCs) represent a complex cell type with a range of unique features. Jewett et al. identify a novel DNA damage response (DDR) that occurs during MCC differentiation. Inhibition of DDR blocks the formation of MCCs indicating this damage response is likely an important step in MCC differentiation.

Among the wide diversity of cell types that are found in biology, multiciliated cells (MCCs) are quite distinct [1]. Unlike most cells that have a single primary cilium, these cells have dozens to hundreds of motile cilia, whose concerted beating generates fluid flow across numerous tissues including the lungs, the oviduct, and the brain ventricles. In order to generate these numerous cilia, MCCs must generate dozens to hundreds of centrioles, which become the ciliary basal bodies. Centriole biogenesis is typically tightly regulated by the cell cycle, so to generate these centrioles, MCCs must enter into a dramatically elongated pseudo-S-phase that is permissible for centriole generation. Furthermore, these cells have an elaborately organized cytoskeletal architecture that surround the basal bodies and facilitate the rapid beating of their cilia. In addition to all these unique features, we can now add another oddity. A recent paper by Jewett et al. describes a massive DNA damage response (DDR) required for MCC differentiation [2]. A previous report observed DNA damage downstream of the MCC inducing factor geminin coiled-coil domain containing (GemC1) in ependymal cells [3]. The important advance in Jewett et al. is that the inhibition of DDR leads to a failure of MCC differentiation indicating DDR as an integral feature in MCC progression.

DDR describes multiple mechanisms to fix naturally occurring breaks in DNA. In particular, repairing double stranded breaks (DSB) is crucial for tissue survival as it can delay progression of the cell cycle or trigger apoptosis if a cell is too damaged [4]. During DNA damage, the phosphorylated histone variant γH2AX localizes to sites of damage and recruits components of the DDR including ATR checkpoint kinase (ATR), ATM serine/threonine kinase (ATM) and DNA-dependent protein kinase catalytic subunit (DNA-PK) [5]. Similar to mouse ependymal MCCs, Jewett et al. observed that the nuclei of MCCs undergoing differentiation (as marked by forkhead box J1 (FoxJ1)) had large amounts of γH2AX foci [2,3]. They confirmed this result by showing that MCCs had higher amounts of DNA fragmentation and were enriched in factors related to DDR such as tumor protein p53 binding protein 1 (TP53BP1) and the activated kinases pATM and pDNA-PK but not pATR. Critically, in cells where ATM and DNA-PK were chemically inhibited there was a failure of MCC differentiation, suggesting that this DNA damage and the subsequent DDR was beneficial to MCC maturation.

The authors propose two hypotheses for why DDR is important in MCC differentiation (Figure 1), although they are not mutually exclusive. The first is that DDR is a natural response to the heavy transcriptional load that is associated with MCC differentiation [6]. Centrioles, cilia and the associated cytoskeletal network are comprised of thousands of distinct proteins many of which need to be found in high concentrations. Given the broad distribution of the genes encoding these proteins across the genome, the transcriptional activity required to achieve this would be considerable, and DDR could be a biproduct of this burden. In support of this theory, auxin-induced depletion of the critical MCC transcriptional regulator, GemC1, resulted in no observable DNA damage, however these cells also fail to differentiate. Importantly, the authors did not observe substantial γH2AX in the telomeric region, indicating the damage was not due to replication stress. Interestingly, GemC1 is a paralogue of the cell cycle-regulated protein Geminin, which is an inhibitor of DNA replication. Although not fully understood, GemC1 (and multiciliate differentiation and DNA synthesis associated cell cycle protein (MCIDAS)) appear to have somewhat opposing roles compared to Geminin in that they are expressed only as Geminin is lost [7]. Is the loss of DNA damage in the GemC1 depleted cells due to some unknown alternative DNA replication-like role, or is it simply due to its role in initiating MCC transcription? Interestingly, the authors observed R-loops, RNA-DNA structures indicative of high transcriptional activity, associated with the observed yH2AX, pATM, and pDNA-PK foci [2,3]. Additionally, non-specific DNA damage using etoposide did not induce MCC differentiation in these cells indicating that the sites of DNA damage are likely critical. Future studies using chromatin immunoprecipitation sequencing (Chip-Seq), DNA-RNA immunoprecipitation followed by sequencing (DRIP-Seq) or RNA:DNA immunoprecipitation followed by sequencing (RDIP-Seq) will certainly be important to identify characteristics and specificity to this process [8].

Figure 1:

The role of DNA Damage Response (DDR) in multiciliated cell (MCC) differentiation. MCCs experience significant DNA damage during the window of centriole amplification. Two hypotheses are posed for this observation; 1. DDR is a response to massive transcriptional load associated with MCC differentiation and 2. DDR is essential to pause the cell cycle allowing for centriole amplification. Created in BioRender. Mitchell, B. (n.d.) https://BioRender.com/t470g9c.

In contrast to this DNA damage being the result of heavy transcriptional load the authors offer an intriguing alternative. Critical to proper MCC differentiation is the generation of dozens of centrioles that must occur in a “stalled” pseudo-S phase of the cell cycle [9]. Although many of the key molecular regulators of this alternative cell cycle have been identified, the control of timing for this process remains a mystery [10–13]. In mouse trachea and the air-liquid-interphase (ALI) cultures (used in this study) the differentiation of MCCs is a drawn-out process lasting many days. One of the main functions of the DDR is to pause the cell cycle while DNA is repaired so as not to proceed through cell division with faulty DNA. Could this feature be hijacked by MCCs to put an extended pause on the cell cycle to allow MCCs to complete their centriole amplification? This exciting possibility is supported by the result that MCCs in the lung (which contain 200–300 centrioles) have considerably more DNA damage than MCCs in the brain that only contain 50–100. Additionally, even within the ALI MCCs there is considerable variation in the number of centrioles, and the amount of DNA damage positively correlates with the amount of centriolar components. Therefore, the timing of this DNA damage seems critical and importantly is significantly higher during the window of MCC differentiation associated with centriole biogenesis (as opposed to ciliogenesis). One way to test of this idea moving forward would be to determine if this DNA damage is also observed in other species where the MCC differentiation process is accelerated compared to mammals. In systems with accelerated MCC differentiation, the transcriptional load would presumably be increased, whereas the need for an extended cell cycle delay would be lessened, potentially allowing these two possibilities to be teased apart. Regardless, the findings in Jewett et al. are certainly an exciting addition to the field and one more example of uniqueness of these cells. A minor limitation is that the bulk of the experiments were performed in culture leaving in vivo relevance uncertain. However, past connections with DNA damage in the MCCs of the ependyma coupled with previously observed MCC linked hydrocephalus phenotypes observed in DDR mutations are consistent with these findings [3,14,15]. Although additional work needs to be done to determine the scope and roles of different cell cycle and DDR components, this new work extends our understanding of MCC differentiation and opens new possibilities for understanding the cell cycle variation in the context of development.