The carboxy‐terminus of the human ARPKD protein fibrocystin can control STAT3 signalling by regulating SRC‐activation

Abstract Autosomal recessive polycystic kidney disease (ARPKD) is mainly caused by variants in the PKHD1 gene, encoding fibrocystin (FC), a large transmembrane protein of incompletely understood cellular function. Here, we show that a C‐terminal fragment of human FC can suppress a signalling module of the kinase SRC and signal transducer and activator of transcription 3 (STAT3). Consistently, we identified truncating genetic variants specifically affecting the cytoplasmic tail in ARPKD patients, found SRC and the cytoplasmic tail of fibrocystin in a joint dynamic protein complex and observed increased activation of both SRC and STAT3 in cyst‐lining renal epithelial cells of ARPKD patients.

genetic variants specifically affecting the cytoplasmic tail in ARPKD patients, found SRC and the cytoplasmic tail of fibrocystin in a joint dynamic protein complex and observed increased activation of both SRC and STAT3 in cyst-lining renal epithelial cells of ARPKD patients.

K E Y W O R D S
cilia, genetic kidney diseases, polycystic kidney disease

| INTRODUC TI ON
Autosomal recessive polycystic kidney disease (ARPKD) is a severe and early-onset hepatorenal fibrocystic disease, mainly caused by variants in the PKHD1 gene. 1 PKHD1 encodes a 450 kDa protein of poorly understood function termed fibrocystin (FC), which consists of a large extracellular part, a single transmembrane domain, and a short C-terminal cytoplasmic tail. Post-translational proteolytic cleavage and nuclear translocation of a C-terminal FC-fragment have been described, 2,3 but an orthologous mouse model has recently questioned the functional relevance of the cytoplasmic tail. 4 ARPKD and the more common autosomal dominant polycystic kidney disease (ADPKD) show overlapping clinical and genetic characteristics. Increased activation of the pro-proliferative transcription factor STAT3 has previously been observed in preclinical PKD models and ADPKD patient samples. 5 Using patient samples, we here describe activation of the SRC-STAT3 axis in ARPKD cyst-lining renal epithelia and furthermore show that a carboxy-terminal fragment of human FC can control a SRC-STAT3-signalling module in cellular studies.

| Immunohistochemistry, coimmunoprecipitation and western blot
For immunohistochemical staining, formalin-fixed, paraffin-embedded patient kidney tissue was used. Immunohistochemistry, co-Immunoprecipitation and Western blot analyses were performed using standard methods. For details and antibodies, please see supplemental methods.

| Stat3 luciferase reporter assay
HEK293T cells were seeded in a 96-well plate. Plasmids were transfected using Lipofectamine 2000 (Invitrogen) and treated with 1 µM Forskolin as indicated. After cell lysis, firefly luciferase activity was detected using the Dual-Luciferase® Reporter assay system (Promega) and the EnSpire® plate reader (Perkin Elmer).

| Statistics
Data are expressed as datapoints of single n's with mean ± SEM of n experiments. Statistical evaluation was performed by using 2-tailed Student's t test or repeated measures ANOVA with Tukey's post-hoc analyses. P values less than 0.05 were considered significant.

| RE SULTS
In order to identify proteins interacting with the cytoplasmic tail of To study mechanisms of STAT3-activation in ARPKD, we therefore used a C-terminal fragment of human FC that contains a short extracellular part, the transmembrane domain and the cytoplasmic tail (FCm; Figure S1D) to allow posttranslational cleavage of FC. 2,3 The expression of FCm neither affected STAT3-dependent transcription per se nor a constitutively active, dimerizing mutant of STAT3 (STAT3_CA; Figure 1B We then focussed on SRC-STAT3-signalling. An increase in intracellular cAMP concentration can enhance SRC activity and is considered to be a major driver in PKD. 1 FCm could inhibit SRC-STAT3-activation by forskolin, a pharmacological activator of adenylate cyclase increasing cAMP levels ( Figure 1C). The effect was lost for two constitutively active SRC variants (SRC Y530F and SRC D102N ) suggesting a direct FCm effect on SRC-activation ( Figure 1D). FCc could be found in joint protein complexes with SRC and JAK2 ( Figure 1E). DZIP1L, a recently described bona fide ARPKD protein, 1 was also detected in protein complexes with SRC and STAT3, but without effects on SRC-STAT3-dependent transcription ( Figure S2).
SRC, but not JAK2, induced a double band of human FCc ( Figure 1E). To study potential tyrosine phosphorylation, we  Figure 2C). Interestingly, FCc co-precipitation was more pronounced with activated SRC variants ( Figure 2D).
We therefore speculated that FCm might induce its effect on F I G U R E 1 A C-terminal fragment of human fibrocystin can inhibit STAT3-activation by SRC. (A) Immunohistochemical staining of renal ARPKD and control tissue show increased STAT3 phosphorylation at tyrosine 705 (brown) in ARPKD cyst-lining epithelia. Nuclei are counterstained in blue. Kidney lysates of three independent ARPKD patients and controls were stained for pSTAT3 Y705 and STAT3, and signal intensity was quantified. ARPKD patients show significantly increased STAT3 phosphorylation (t(4) = 2,890; P = .0223; *P < .05). (B-C) STAT3 luciferase reporter assay showing that co-expression of FCm reduced the activation of STAT3-dependent transcription by SRC or JAK2 (N = 4; Repeated Measures ANOVA analysis demonstrated significant difference in STAT3 activation between groups (F(7,21) = 28.99, P < .0001; Tukey's post-hoc ***P < .001)) (B). The effect was seen even after increase of intracellular cAMP concentrations by forskolin (N = 5; Repeated Measures ANOVA analysis demonstrated significant differences in STAT3 activation between groups (F(7,28) = 60.15, P < .0001; Tukey's post-hoc *P < .05, **P < .001)) (C). There was no effect of FCm on a dimerizing STAT3 mutant (STAT3_CA) (B). (D) Constitutive active SRC variants (Y530F and D102N) activate STAT3 stronger than SRC WT. This activation cannot be reduced by FCm (N = 5; Repeated Measures ANOVA analysis demonstrated significant differences in STAT3 activation between groups (F(7,28) = 45.51, P < .0001)). #T test comparing the groups SRC WT with SRC WT + FCm only confirms significant difference in SRC-induced STAT3 activation (t(4) = 5.779, P = .0022). (E) After exogenous expression FCc co-precipitates with JAK2 and SRC, but not with a control protein.
Co-precipitating FCc appears as double band in the presence of SRC. In luciferase assays, different shades of grey represent independent experiments STAT3-dependent transcription by inhibiting SRC-activation.
Indeed, expression of FCm resulted in a significantly reduced signal for the activating phosphorylation of SRC at tyrosine 419 with and without forskolin ( Figure 2E). Strikingly, this phosphorylation was strongly increased in cyst-lining epithelia of ARPKD kidneys ( Figure 2F).

| D ISCUSS I ON
The pathophysiology of cystogenesis in ARPKD and the molecular function of fibrocystin are not well understood. Overlapping functions with the ADPKD proteins have been suggested. [6][7][8] We show that the C-term of human FC can regulate SRC-activation and that SRC and STAT3-activation can be observed in ARPKD cyst-lining epithelial cells. Interestingly, pharmacological inhibition of SRC-activation results in amelioration of the phenotypes in various preclinical models of PKD. [9][10][11] First clinical trials on SRC inhibitors for PKD have been initiated (eg ClinicalTrials.gov identifiers NCT03096080 and NCT03203642). 12 SRC can be activated through multiple mechanisms, including increased intracellular cAMP concentration, 5 but details in PKD remain elusive.
We found SRC and FCc in a common protein complex fitting to previous data linking FC to focal adhesion complexes. 13

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.