Perhaps the most exciting aspect of attending any scientific meeting is the privilege of becoming aware of novel research findings in our fields of interest, prior to their appearance in published literature – and this begins as soon as we have the abstract book in hand! Sitting in my hotel room in Suzhou, and browsing through the abstracts of the first Cold Spring Harbor meeting on cilia and centrosomes which was held in the spring of 2017, I chanced upon work from Chengtian Zhao’s group (Ocean University, Qindao, China) that described the analysis of gene expression changes in zebrafish embryos deficient in zymnd10, a gene encoding a ciliary dynein assembly factor (1). Since my lab had also performed similar analyses with two other cilia mutants, ccdc103 and lrrc50 (also encoding dynein assembly factors (2,3)), I decided that I must meet up with Chengtian and discuss whether we should collaborate or at the least, co-ordinate our studies. That evening, as I settled down for a sumptuous dinner in the esteemed company of two prominent cilia researchers, Hiroshi Hamada and Cecilia Lo, I was pleasantly surprised when a Chinese gentleman came forward and introduced himself to me as Chengtian Zhao, and asked if he could join us for the meal. Conversation flowed effortlessly and the rest, as they say, is history.
Over the years, a number of studies had observed that zebrafish embryos with mutations in many different cilia genes exhibit ventrally curved body axis. More recently, work from the Burdine and Ciruna labs had implicated defects in cilia motility, causing abnormal cerebrospinal fluid (CSF) flow, in curvatures of the vertebral column in adult zebrafish (4), making this system a good model for uncovering the etiological basis of the common human spine disorder idiopathic scoliosis (IS) that affects up to 3% of children and adolescents world-wide (5). Abnormal 3D curvatures of the spine in IS can cause a considerable degree of morbidity in patients, manifest in symptoms like difficulties in breathing, postural issues and gait problems. Despite these important findings, the molecular mechanism operating downstream of cilia-driven CSF flow that ensures proper spine development had remained unclear. With the collaboration that sparked off at the dinner table of the Suzhou meeting, Chengtian’s and my lab published a collaborative paper just before Christmas of 2018, describing a possible molecular pathway by which ciliary motility, within the brain ventricles and spinal canal of the zebrafish, ensures that they develop with a straight body axis (6). We showed that cilia beating transports catecholamines (like epinephrine) in CSF, and they stimulate the expression of Urotensin-related peptides (Urp) in CSF-contacting neurons (CSF-cNs) that differentiate along the spinal canal. The Urp family consists of small cyclic neuropeptides, homologous to Urotensin II, which is known to function as a potent vasoconstrictor (7). Indeed, it was the significant decline in expression levels of the urp genes in our respective gene expression analyses of cilia mutants that prompted us to follow up on Urp signalling to uncover the mechanism acting downstream of cilia-driven CSF flow. Our current model posits that Urp peptides, secreted from CSF-cNS, brings about a certain degree of tension or contraction of the somitic slow-twitch muscles of the dorsal somites, and this is the biomechanical force for axial straightening. In keeping with this view, we also found that a Urp receptor, Uts2r3, a G-protein coupled receptor, is expressed in the dorsal slow muscles, and when it was rendered non-functional using a targeted lesion at the locus, the adult uts2r3 mutants developed with severely curved spines, resembling humans with IS (6).
More or less concurrent with our publication, Claire Wyart, Pascal Bardet and colleagues reported a very intriguing story that ascribed, for the first time, a definitive biological role for the enigmatic Reissner fiber (RF) (8). Discovered in the 19th century by the German anatomist Ernst Reissner, RF is a filamentous structure that remains suspended in CSF along the brain ventricles and the spinal canal in all vertebrates examined, including humans. RF is polymerized from a large glycoprotein, SCO-spondin (Sspo), secreted from the subcommissural organ (hence SCO) and the floor plate. RF is thought to participate in many functions of the nervous system, but the lack of genetic mutations in Sspo has precluded a firm association of RF with any of these proposed roles. Wyart et al. showed that in zebrafish embryos, RF is required for proper axial development since sspo mutants developed ventrally curved body axes (and perished at the early larval stage), closely mimicking the curved bodies of cilia mutants (8). They also demonstrated that cilia motility is required for RF biogenesis, by somehow facilitating the polymerization of the protein into the fiber. Since biochemical studies with mammalian RF had already shown that it can efficiently bind and facilitate the transport of catecholamines present in CSF (9), it dawned on me that cilia, RF, and CSF catecholamines could all be functioning via Urp signaling from CSF-cNs.
The idea that I wanted to explore is whether sspo mutants develop curved bodies because in the absence of RF, they are unable to bind CSF catecholamines and present to CSF-cNs to activate urp gene expression. It is with this view in mind that we approached Claire and Pascal, who most generously shared with us the sspo mutants. In a paper that we have just published in Biology Open (10), we now show that consistent with our expectations, sspo mutants exhibit a loss of urp gene expression from CSF-cNs: urp1 is significantly reduced, while urp2 is almost completely absent. As we had demonstrated previously for cilia mutants, culturing sspo mutants in the presence of exogenously added epinephrine in embryo medium, restored urp expression in CSF-cNs and also rescued their ventrally curved axial defects. However, one issue that continued to confound us was whether abnormalities of the embryonic axis in sspo mutants have any connection with scoliosis of the spine in adult zebrafish. If there is, our work will have relevance for furthering our understanding of the etiology of IS. The Burdine and Ciruna et al. paper, that had initially linked cilia motility and CSF flow with spine curvature, used a clever strategy to bypass the severe embryonic body curvature and associated lethality of the cilia mutants that they studied (4). They injected the corresponding (in vitro synthesized) wild-type sense mRNA into mutant eggs: this rescued the axial defects as well as lethality, and the mutants developed into adults but they exhibited severely curved spines. Unfortunately, we could not utilize this strategy to rescue sspo mutants as the Sspo protein consists of more than 5000 amino acids, and in vitro synthesis of an mRNA to encode such a large protein is not feasible. To circumvent this problem, we prematurely dechorionated the sspo mutants with the hope that their severe ventrally curved axis could be partially rescued when freed off the confines of the spherical, non-elastic chorion. Indeed, this simple trick ameliorated the strong axial curvature of sspo mutants to varying degrees, and many of them matured into adults with severe spine malformations reminiscent of cilia mutants rescued of their embryonic lethality and mutations in the Urp receptor. Thus, in the zebrafish, ventral curvature of the embryonic axis and scoliotic malformations of the adult spine represent linked morphogenetic anomalies. Most satisfyingly, we found that restoring expression of Urp2 exclusively in CSF-cNs, rescued not only the embryonic and larval body curvature of sspo mutants, but also allowed one such mutant to develop into an adult with an apparently normal spine! Additional findings that we report in our Biology Open paper include data showing that it is a definite threshold of Urp signaling that is critical for the morphogenesis of a straight body axis. Too little signaling causes ventral curvature as in sspo and cilia mutants, while exaggerated signaling (for instance, by over-expression of the urp gene or the protein in the muscle cells themselves) causes the converse effect of profound dorsal curvature of the axis. Finally, using mutations in Smoothened (Smo), an essential component of the Hedgehog pathway that directs slow-twitch muscle cell differentiation in the zebrafish somites, we could show that lack of the slow muscles make these mutants refractory to Urp signaling (10). Smo mutants have strong ventrally curved bodies that could not be rescued by over-expression of the Urp proteins.
How do we take these findings forward, especially for furthering our understanding of the mechanistic basis of pathogenesis in IS? First, with reference to the zebrafish, we need to better understand how Urp signaling-induces activity of the slow-twitch fibers of the somites, and how this activity feeds back to the growing spine to ensure that it develops along a straight axis. Secondly, while RF and CSF-cNs exist in mammals and Urp signaling has also been shown to operate there, we will need to examine whether the circuitry that we have been able to dissect in the zebrafish is also conserved. In this regard, one caveat that we need to bear in mind is that traditional experimental mammals like mice and rats are quadrupeds, and they have not turned out to be effective for modeling human spine disorders (hence the promise of the zebrafish)(11). Finally, based on what we glean from all of these investigations, we can begin to parse the underlying mechanisms driving spine malformations in IS. There is already accumulating evidence that ciliary dysfunction could be causative of the disease (12,13). Moreover, presence of RF has been reported in human embryos and a teenager (14), implying that defects in this structure could be responsible for IS in some of the individuals afflicted with IS. And of course, the real benefit of all this research effort will be if we can invent effective therapeutic strategies for IS by pharmacologically manipulating the Urp pathway, since current treatment options are largely limited to managing the disorder with physiotherapy and braces, and in severe cases, the rectification of severe spine deformities with invasive surgery.
1. ZMYND10 functions in a chaperone relay during axonemal dynein assembly. Mali GR et al. Elife 2018 7. pii: e34389.
2. CCDC103 mutations cause primary ciliary dyskinesia by disrupting assembly of ciliary dynein arms. Panizzi JR et al. Nat. Genet. 2012 44:714-9.
3. Deletions and point mutations of LRRC50 cause primary ciliary dyskinesia due to dynein arm defects. Loges NT et al. Am. J. Hum. Genet. 2009 85(6):883-9.
4. Zebrafish models of idiopathic scoliosis link cerebrospinal fluid flow defects to spine curvature. Grimes DT et al. Science 2016 352:1341-4.
5. Genetics and pathogenesis of idiopathic scoliosis. Grauers A et al. Scoliosis Spinal Disord. 2016 11:45.
6. Cilia-driven cerebrospinal fluid flow directs expression of urotensin neuropeptides to straighten the vertebrate body axis. Zhang X et al. Nat. Genet. 2018 50:1666-73.
7. International Union of Basic and Clinical Pharmacology. XCII. Urotensin II, urotensin II-related peptide, and their receptor: from structure to function. Vaudry H et al. Pharmacol. Rev. 2015 67:214-58.
8. The Reissner fiber in the cerebrospinal fluid controls morphogenesis of the body axis. Cantaut-Belarif Y et al. Curr. Biol. 2018 28:2479-86.
9. Reissner fiber binds and transports away monoamines present in the cerebrospinal fluid. Caprile T et al. Brain Res. Mol. Brain Res. 2003 110:177-92.
10. Reissner fiber-induced Urotensin signaling from cerebrospinal fluid-contacting neurons prevents scoliosis of the vertebrate spine. Lu et al. Biol. Open 2020 9:pii: bio052027.
11. Understanding idiopathic scoliosis: A new zebrafish school of thought. Boswell CW and Ciruna B. Trends Genet. 2017 33:183-96.
12. Functional variants of POC5 identified in patients with idiopathic scoliosis. Patten SA et al. J. Clin. Invest. 2015 125:1124-8.
13. New associations of primary ciliary dyskinesia syndrome. Engesaeth V et al. Pediatr Pulmonol. 1993 16:9-12.
14. Reissner’s fibre: the exception which proves the rule, or the devil according Charles Baudelaire? Olry R and Haines DE. J. Hist. Neurosci. 2003 12:73-5.