A ZFYVE19 gene mutation associated with neonatal cholestasis and cilia dysfunction: case report with a novel pathogenic variant

Cholestasis affects approximately 1 in every 2500 term infants and it is defined as reduced bile formation or flow resulting in the retention within the liver of biliary substances normally excreted into bile and destined for elimination into the intestinal lumen. Neonatal cholestasis may stem from several conditions requiring either medical or surgical treatment. Some forms may remain of undetermined origin (so called, with descriptive term, “idiopathic neonatal hepatitis”) [1]. In a European study, biliary atresia was the most common diagnosis (41%), followed by progressive familial intrahepatic cholestasis (PFIC) and idiopathic cases (approximately 10% each), which thus represent a minor but still challenging and substantial group of disorders [2]. Advanced sequencing methods promise to further increase the diagnostic yield of genetic approaches. [3, 4].

While obstacles on the extrahepatic biliary tree [e.g. surgical emergencies such as Biliary Atresia (OMIM #210500)] are quite regularly characterized by high levels of gamma-glutamyl transpeptidase (GGT), assessment of intrahepatic cholestasis is more difficult. In this respect, patients can be usefully categorized by whether levels of serum GGT remain in normal ranges despite hyperbilirubinemia activity or rise together with serum concentrations of conjugated bilirubin [1]. High GGT values often reflect rare genetic cholangiopathies such as Alagille syndrome (OMIM #118450), Renal Cysts and Diabetes Syndrome (RCAD) (OMIM #137920), alpha-1-antitrypsin deficiency (OMIM #107400), ABCB4 disease (OMIM #171060, Phenotype MIM number 60234), and neonatal sclerosing cholangitis (OMIM #61739) (involving the CLDN1, DCDC2, KIF12, or PPM1F genes respectively). Recently, Luan et al. described a new and previously unidentified genetic cause, the zinc finger FYVE-type containing 19 (ZFYVE19), alias abscission/nocut checkpoint regulator, as a novel cause of high-GGT infantile cholestasis in a small series of Chinese children with a DCDC2-unrelated neonatal sclerosing cholangitis-like phenotype (Congenital Hepatic Fibrosis and Sclerosing Cholangiopathy) [5]. The ZFYVE19 protein controls cytokinesis and resides mainly on centrosomes in interphase and early mitosis (https://​www.​genecards.​org/​cgi-bin/​carddisp.​pl?​gene=​ZFYVE19). In post-mitotic cells, centrosomes move to the apical cell surface and contribute to the formation of primary cilia. Epithelial cells lining the lumen of intrahepatic bile ducts, i.e. cholangiocytes, display primary cilia consisting of (a) the microtubule-based axoneme that has nine peripheral microtubule doublets arranged around a central core that does not contain two central microtubules, and (b) the basal body, from which the axoneme emerges. Primary cilia are well established sensory organs and their physiologic significance in cholangiocytes remains unclear. Cholangiocyte cilia extending from the apical plasma membrane to the lumen of the bile duct are ideally positioned to detect changes in bile flow, composition and osmolality. These mechanosensory, osmosensory and chemosensory organelles have been proposed to control cholangiocytes’ functions such as the formation of ductal bile [6]. Interestingly, biliary specimens in chronic cholestasis due to syndromic and non-syndromic biliary atresia may show shorter cilia, abnormal in their orientation, and less abundant compared to controls. While this may result from the same severe cholestasis or inflammation themselves, it has been suggested that it may also reflect common mechanistic pathways in different forms of biliary atresia and may have implications for understanding the progression of the disease [7].

Our study offers a first confirmation of the link between ZFYVE19 and a neonatal-onset chronic cholestasis phenotype as well as further experimental evidence for a cilia involvement in this condition. It also endorses the concept that the rate of patients diagnosed with the descriptive term ‘‘idiopathic neonatal hepatitis’’ will continue to decline with the advancement in diagnostic evaluation and the use of next-generation DNA sequencing technologies [1, 3, 4]

Our patient with a high GGT cholestatic hepatobiliary disease was erroneously suspected of having extrahepatic biliary atresia on the basis of the clinical, laboratory and liver histological pictures. Subsequently, only WES analysis clarified the nature of the disorder by detecting mutations in the protein coding gene ZFYVE19, a putative key regulator of abscission step in cytokinesis, i.e., the very last phase of mitotic cell division that results in a physical separation of two daughter cells (https://​www.​genecards.​org/​cgi-bin/​carddisp.​pl?​gene=​ZFYVE19). This transcription unit has hitherto been described as a novel cause of high-GGT infantile cholestasis only in a series of 9 Chinese children with neonatal sclerosing cholangitis-like phenotype. On histopathologic study those patients had portal-tracts widening and fibrosis without interface activity, and ductular reaction consonant with a persistence of the ductal plate [5]. Both these histological features are in keeping with those of the portal tract of the liver biopsy of our patient (Fig. 1, panel 1a–d) and also of other patients with mutations in other genes encoding cilium-associated proteins. More specifically, the renal-hepatic ciliopathy due to Doublecortin domain-containing protein 2 (DCDC2) mutations results in a hepatic histological pattern resembling congenital hepatic fibrosis, but high GGT cholestasis in this condition appears independent from biliary ducts infections but rather is linked to sclerosing cholangitis. Other high GGT putative cholangio-ciliopathies caused by mutations in Kinesin family member 12 (KIF12), and protein phosphatase 1F (PPM1F) also have histopathologic features suggestive of biliary cirrhosis ± paucity of bile ducts and liver fibrosis [12]. Finally, variants in ANKS6 (Ankyrin Repeat and Sterile Alpha Motif Domain Containing 6), which encodes a protein that interacts with other proteins of the Inv compartment of cilium (NEK8, NPHP2/INVS, and NPHP3), were detected in other patients who had histological periportal fibrosis and cholestatic hepatopathy along with a picture of infantile nephronopthisis-related ciliopathy (NPHP-RC) [13, 14].

In the Chinese study of Luan, ZFYVE19 depletion in retinal pigmented epithelial cell line (hRPE1) showed supernumerary centrioles and separation/abnormal arrangement of the centriole pair(s). However, ciliogenesis was not affected and additional cilia protrude from supernumerary basal bodies/centrioles. Similar ciliary and centrioles phenotypes were observed in ZFYVE19-deficient fibroblast-like cells derived from patient iPSCs [5]. Our study on skin fibroblasts from the ZFYVE19-mutated patient confirmed both the centrioles abnormalities and the comparable number of ciliated cells observed between the patient’s sample and control, as described in Luan et al. Moreover, we observed that the ciliary phenotype was also characterized by discontinuous staining of ARL13b in primary cilia axonemes. The differences observed in cilia morphology may be explained by the different mutations studied or by the different experimental conditions as our study was conducted on primary fibroblasts while the Chinese study was performed in ZFYVE19-knockdown hRPE1 cells and patient-derived iPSCs [5]. Our results, however, confirm the observation of cilia dysfunction in ZFYVE19-mutated patients also in cells not subjected to chronic cholestasis. The abnormalities observed in patient-derived skin fibroblasts reflect the cilia abnormalities of the affected tissues and indicate that indeed ZFYVE19 plays a role in ciliogenesis. Conversely, the eventual absence of a ciliary phenotype in fibroblasts would have not ruled out a tissue-specific ciliary defect [15]. Table 1 summarizes the main findings in the Chinese series and our case. Out of a cohort of 25 children with undiagnosed high-GGT cholestasis, 9 showed ZFYVE19 mutation. Only in one parental consanguinity had been established, as in our case. Similarly, no child was icteric at enrolment in the study although four patients presented with neonatal jaundice. GGT, GOT, and GPT were elevated in all patients. Seven were treated with UDCA and 4 required liver transplantation.

Although Fibrocystic Liver Disease, a heterogeneous group of biliary disorders characterized by abnormal development of the embryonic ductal plate, secondary to genetically determined dysfunctions of cholangiocytes cilioproteins, is a well-established cilia-associated liver disorder, isolated and/or severe neonatal cholestasis is rarely recognized as a ciliopathy. However, the recent identification of biallelic mutations in Tetratricopeptide Repeat Domain 26 (TTC26) a well-known ciliary protein, in a rapidly progressive severe neonatal cholestatic condition highlighted the need of including ciliopathies in the differential diagnosis of severe neonatal cholestasis even in the absence of more typical features [16]. Generally, ciliopathies are multisystemic disorders and the liver is frequently involved CHF and CD are the most common liver manifestations of ciliopathies in children. Both the patient’s cohort described by Luan et al. and our case present only hepatic involvement.

Analysis of our patient plasma sterols in addition to increased levels of cholestanol, which is known to be also secondary to chronic cholestasis, showed increased campesterol and sitosterol levels. This was in agreement with the diagnosis of sitosterolemia (OMIM #210250) coincidental with the associated cholestatic pattern due to ZFYVE19 defect. Although high concentrations of serum phytosterols are correlated with liver disease and severity of cholestasis in Total Parenteral Nutrition-dependent children, liver involvement in sitosterolemia has been (rarely) reported only in adults and never of cholestatic origin [17]. Notably, also the series of Luan patients harbored unexpected variants in several other genes without established relationships with liver disease [5].

The case here reported offers at least three foods for thought. Firstly, it confirms that ZFYVE19 is a novel cause of high GGT neonatal-onset chronic cholestasis where cilia involvement plays a likely role. The latter specific aspect calls for a continuous monitoring of the patient to verify the liver disease progression (namely, overt sclerosing cholangitis and/or portal hypertension) and whether the phenotype will involve also other organs and systems not yet affected. The coexistence with sitosterolemia, a condition hitherto described having not effects on the liver, also will deserve attention.

Secondly, it further confirms that use of WES may improve the diagnostic yield in children with undiagnosed/undefined cholestasis or compound phenotypes with a likely inheritable background. The application of WES, although not routinely, enables the identification of new causative genes, widening the knowledge on the pathophysiology, eventually resulting in therapeutic implications.