Ontogenesis of the tear drainage system requires Prickle 1-controlled polarized basement membrane (BM) deposition

In terrestrial animals, lacrimal drainage apparatus evolved to serve as conduits for tear flow. Little is known about the ontogenesis of this system. Here, we investigated tear duct origin, developmental course, genetic and cellular determinants in mouse. We report that primordial tear duct (PTD) originates from junction epithelium of the joining maxillary and lateral nasal processes, which reshapes into future tear duct branches. We identified Prickle 1 as a hallmark for tear duct outgrowth, ablation of which stalled duct elongation. In particular, the disruption of basement membrane (BM) with cytoplasmic accumulation of laminin suggests aberrant protein trafficking. Mutant embryoid bodies (EBs) derived from iPSCs recapitulate BM phenotype of the PTD exhibiting defective visceral endoderm (VE), which normally expresses high level of Prickle 1. Furthermore, replenishing mutant VE with Prickle 1 completely rescued BM but not cell polarity. Taken together, our results reveal a distinct role of Prickle 1 in regulating polarized BM secretion and deposition in precedently uncharacterized tear drainage system and VE, which is independent of apicobasal polarity establishment.


Introduction
A main function of tear flow is to lubricate ocular surface preventing it from drying. Disturbing tear production or flow would lead to unhealthy ocular surface including dry eye syndrome, which has a global prevalence ranging from 5%~ 50% 1 . In many tetrapods, orbital glands and excretory/drainage conduits coevolved to produce and passage tears over the ocular surface 2 .
A few comparative studies demonstrated considerable phylogenetic variations in drainage ducts 2,4,[6][7][8][9][10] . The human NLD is of ectodermal origin and starts to develop at about 5.5 weeks of gestation 11 . Epithelial maxillary and nasal processes form a lacrimal groove, which then pinches off as a solid cord extending toward both orbital and nasal directions 11 . The rabbit NLD shares many similar anatomical aspects with that of humans and is considered a suitable model for human NLD study 2,4 . However, unlike human, the rabbit NLD originates at the subcutaneous region of the lower eyelid extending unidirectionally to the naris 7 . Similar observation was reported in a rodent animal, Mongolian gerbil 8 . Mouse NLD is assumed to develop similarly as humans based on a scanning electron microscopy study (Tamarin and Boyde, 1977). However, such a notion faces challenges in that both the mouse and Mongolia gerbil are taxonomically of Rodentia, evolutionarily much closer to rabbits than humans. Thus, an examination of ontogeny of the mouse NLD shall resolve such contradiction.
To date, developmental and genetic determinants of the tear duct are still lacking. CL and NLD defects are seen in several syndromic diseases with known causal mutations [12][13][14][15][16][17][18] . Among those, mutations in FGF signaling components (FGF10, FGFR2 and FGFR3) cause hypoplasia of NLD and lacrimal puncta often with conjunctivitis as part of Lacrimo-auriculo-dento-digital (LADD) syndrome 18 . Consistent with these findings, mutations in p63, upstream of FGF signaling, leads to disorders overlapping with LADD, in which lacrimal outflow obstruction can occur with agenesis of NLD and CL 19 . With congenital nasolacrimal duct obstruction (CNLDO) being present in up to 20% of newborn infants 20,21 , it is conceivable that many more genetic and/or genetic risk factors being involved in the tear duct obstruction and development.
The Wnt/PCP pathway plays key roles in morphogenesis of diverse tissues during development [22][23][24] . Particularly for tubulogenesis, tubular branching, elongation and migration require PCP-driven convergent extension (CE) and oriented cell division 25,26 . A set of six proteins including Prickle, Frizzled, Disheveled, Vangl, Diego, and Flamingo executes Wnt/PCP signaling to control cell polarity and oriented cell migration 27,28 . Mutations in Vangle, Frizzled, Prickle and a non-core PCP component, protocadherin Fat, all cause malformed renal tubules 25,29,30 . Because tear duct development is a process of tubulogenesis, we hypothesized that it would require Wnt/PCP signaling, as demonstrated in other tubular organs. To test our hypothesis, we investigated the full course of tear duct ontogenesis in mouse and identified a crucial role of Prickle 1 in tear duct elongation. We further show a general role of Prickle 1 in regulation of polarized secretion and deposition of basement membrane (BM) in tear drainage system and embryoid body (EB) organoids, which is likely independent of establishment of apicobasal polarity.

Tear duct origin and the timings of crucial events during development
To gain developmental insights into tear drainage system, we first determined when tear duct initiates by performing 3D-reconstruction of primordial tear duct (PTD) at E11 using iDISCO technology 31 (Fig. 1A, Supplemental movie 1).
From different perspectives, tear duct was seen to initiate from the epithelial junction of fusing maxillary and nasal plate ectoderm (Fig. 1A) to form an epithelial cord. This was likely achieved partially through epithelial-mesenchymal transition (EMT), which generated a cell mass with multipolar protrusions observed on sections ( Fig.1B-E, F-I). A thinning stalk connecting with maxillary/nasal surface ectoderm was detected at E12 ( Fig.   1F-I). The anterior/nasal portion of the PTD had already extended a considerable distance at this age but could not be visualized together with orbital PTD on same section (see later).
To find the timing of PTD reaching target tissues, we first sectioned mouse head coronally to determine when NLD reaches the nasal cavity. NLD extended close to the nasal cavity but did not reach it at E13.5 (Supplemental Fig. 1A-F). At E14, NLD reached the nasal cavity and merged with nasal epithelium (Fig. 1J-M). Because the positioned mouse head for sectioning was not completely symmetrical to vertical axis, only one side of NLD was shown.
The opposite side was beyond the joining point of NLD and the nasal epithelium. We next prepared parasagittal sections to define the timing of canaliculi joining the conjunctiva. The lower CL (LCL) started to join conjunctival epithelium at E14 (Fig. 1N-Q), when the upper CL (UCL) still had a distance to reach the upper eyelid (Supplemental Fig. 1G-I). By E16.5, both LCL and UCL joined the conjunctival epithelium of the lower (Fig. 1R-U) and the upper eyelid ( Fig. 1V-Y), respectively. Thus, the data together demonstrated that the tear duct originated from maxillary/nasal epithelium starting at around E11 and completely reached target tissues by E16.5.

Prickle 1 is a hallmark for tear duct ontogenesis
We next investigated whether Wnt/PCP signaling components, which play essential roles in tubulogenesis in a variety of tissue contexts, were expressed in PTD using in situ hybridization. A number of Wnt/PCP genes were expressed in developing tear duct (not shown). Among those, Prickle 1 is expressed exclusively in PTD but not conjunctival epithelium. Taking advantage of a knock-in eYFP reporter in Prickle 1 heterozygous null allele 32 , we examined Prickle 1 expression at a series of embryonic ages. Prickle 1 was found strongly expressed in initiating/primordial tear duct at E11 on parasagittal sections ( Fig. 2A-C, green). On horizontal sections, Prickle 1 was also seen to be strongly expressed at the frontal area of maxillary process, and moderately but specifically in initiating tear duct (Fig.2D, E). Similar expression was found at E11.5 and E12 parasagittal and horizontal sections, when the tear duct continued growing out and the orbital/posterior branched out ( Fig.   2F-O). Bifurcation of canaliculi (CL) and anterior NLD were clearly seen at E13 on separate parasagittal sections strongly expressing Prickle 1 (Fig. 2P-U), which continued through until CL and NLD reaching target tissues (Supplemental Fig. 2A, frontal, B-D, parasagittal). The data thus defined Prickle 1 as a bona fide marker for tear duct ontogenesis.

Disruption of Prickle 1 stalled tear duct outgrowth
The unique Prickle 1 expression pattern prompted us to query its function in tear duct development. Using Prickle 1 null mutants that were created previously 30 , we first performed 3D-reconstruction of serial age progression of TD from both wild type and mutant embryos from E11 to E14 at half-day intervals. This offered us an additional opportunity to have an overall look at full course of tear duct development. Consistent with the previous observations ( Fig. 1 & 2), on lateral views, wild type tear duct anlage/primordial tear duct (PTD) was detected as early as embryonic day 11 (E11) (Fig. 3A). As the PTD elongated, the anterior became thinner and advanced nasally, whereas the posterior was thickened with a narrow stalk connected to the original surface ectoderm (Fig. 3B, C). By E12.5, PTD completely separated from the surface ectoderm and was transformed into three tubular branches (Fig. 3D). The long anterior branch was the future nasolacrimal duct (NLD), and the shorter posterior branches were prospective upper and lower canaliculus (UCL and LCL), respectively. The three tubules continued growing toward their prospective targets-the conjunctival and nasal epithelia ( Fig. 3E-G). The whole tear drainage system --the canaliculi and NLD, and its positional relationship to conjunctiva and nasal cavity was clearly recognized after E12.5 ( Fig. 3E-G).
The mutant PTD was initiated at about the same time as the wild type ( Fig.   3H), however, it was shorter and wider at all ages examined (Fig. 3I-N).
Separation of the epithelial stalk from the presumptive conjunctiva was accomplished at E13 (Fig. 3L), half a day later than that of the wild type, which was complete at E12.5 (Fig. 3D). Collective movement/migration of the outgrowing epithelial cells was hampered ( Fig. 3I-N), and characteristics of canaliculi and NLD branches were barely identifiable in the mutants ( E-G). Further examination of -and p120 catenins revealed a regional difference for their localization of in that the advancing tip in wild type mice exhibited more cytoplasmic staining than rest of the tube (Supplemental Fig.   3K, L, S, T). This regional difference also existed in basement membrane components (see later). Regardless, like -catenin, both -and p120 catenins were ectopically located in cytoplasm throughout the mutant duct along with patchy E-cadherin in majority of areas (Supplemental Fig. 3M, N However, the mutant EBs failed to form basement membrane (Fig. 6G, H, J).
The loss of BM was further confirmed by mislocalization of Col. IV and Perlecan (Supplemental Fig. 7). Phalloidin-stained actin filaments clearly outlined apical surface of the control endodermal cells (Fig. 6K, L, N). In the mutant, actin bundles were still present but thinner, without obvious asymmetrical distribution (Fig. 6O, P, R). Microtubule tracks labeled by acetylated tubulin were roughly parallel to apicobasal axis in control VE cells

Rescue of BM phenotype but not AB polarity of the mutant EBs by expressing Prickle 1
The severe BM and cell polarity phenotypes in both mutant PTD and EBs implied a conserved role of Prickle 1 in these molecular events. We next sought to rescue BM and/or cell polarity by adding back Prickle 1 to mutant EBs. Using a lentiviral system with built-in tetracycline inducible components, we infected mutant iPS cells to control Prickle 1 expression by addition of doxycycline and harvested EBs at different time points (Fig. 7A-C). The mutant EBs showed rescue of BM starting from differentiation D2 and continuing to increase by D5 (Fig. 7C). Nearly 60% (Fig. 7C) of mutant EBs were rescued with integral BM upon Prickle 1 induction (Fig. 7F, G, H), in contrast to less than 10% EBs having BM in control groups with either induced Cherry or uninduced Prickle 1 (Fig. 7D, E). Surprisingly, although the mutant outer layer endodermal cells expressing Prickle 1 were generally sparse (Fig. 7F, G, H

Discussion
Scant attention has been paid to the tear drainage system despite its importance. A few studies primarily focused on comparative anatomy between species 2, 4, 7, 11 . The current study details ontogenesis of the mouse tear drainage system and outlines several crucial events -tear duct origin, elongation, and reaching final destinations. The study further offers genetic and molecular insights into tear duct tubulogenesis, highlighting polarized BM secretion controlled by Prickle 1 as a crucial event for duct elongation.
The formation of mouse tear duct is distinct from all other studied animals based on the current study. Origin wise, the mouse tear duct is more like that of humans, both of which initiate from joining maxillary and nasal process 11 .
However, initiation of the mouse PTD is continuous with orbital conjunctival epithelium rather than the nasolacrimal groove in humans 11 . Mouse tear duct appears to share a common origin with conjunctiva, as both are from joining surface ectoderm of maxillary and nasal plates. Fusion of the two is a zipping process from frontal/nasal to orbital/rare, leaving an orbital epithelial notch for PTD outgrowth. As the PTD separates from the conjunctival epithelium, the fusion is accomplished. The CL branches later reconnect to conjunctiva to serve as drainage conduits. Although mouse and Mongolia gerbil are both taxonomically Rodentia, origin of tear duct in the gerbil is more similar to the one of the rabbit, which is from the subcutaneous region of the orbital lower Tubulogenesis driven by PCP has been well studied 25,26 , though not from the perspective of ECM contribution. The current study, for the first time, demonstrates a crucial role of PCP-controlled ECM secretion in tubulogenesis.
Notably, the Drosophila egg chamber employs an unusual form of planar polarity with aligned basal actin bundles and BM providing "molecular corset" to direct chamber elongation 38 . The similar process may also be utilized for PTD elongation to restrict expansive forces and promote lengthening of the duct.
In summary, our study provides novel knowledge of ontogeny of tear drainage system, which is implicated in a wide range of ocular disorders. The ontogenesis of tear duct requires Prickle 1-regulated polarized basement membrane secretion and deposition. This process is likely through a role of Prickle 1 in intracellular trafficking and secretion, which is independent of general polarity machinery. Disruption of the intrinsic polarized BM secretion will be ultimately transformed into an amplified extrinsic impact on AB or tissue polarity, which may also directly involve Prickle 1.

Acknowledgements
We thank Dr. Tiansen Li from the National Eye Institute for providing valuable suggestions in preparation of the manuscript. The authors thank Tiansen Li, Rong Ju for critical reading the manuscript and helpful comments. We thank lab members Shujuan Xu, Shanzhen Peng, and Xinyu Gu for technical supports to the work.

Funding
This work was supported by grants from the National Natural Science   30 . Mouse genotyping was conducted as described previously 30,32 .

Mice and genotyping
A knock-in eYFP reporter under the control of endogenous Prickle 1 promoter was used to monitor Prickle 1 expression.

Generation of iPSCs from MEFs
Isolation, culturing, and maintenance of mouse embryonic fibroblasts (MEFs) followed standard protocols 41  The derived-iPSCs were first characterized by alkaline phosphatase staining according to the manufacturer's instructions (MA0197, Meilunbio, China), then subjected to immunohistochemistry for examination of a set of stem cell markers including Sox2, Nanog and SSEA1 (Supplemental Fig 6). Teratoma formation was used to functionally evaluate pluripotency of iPSCs by intraperitoneal implantation of 1x10 6 cells into immunodeficent mice 42 . Mice were sacrificed at 2.5 months after transplantation followed by H&E staining to identify tissue types (Supplemental Fig 6).

EB differentiation and rescue of the mutant EBs
The derived iPSCs were differentiated into embryoid bodies (EBs) according to the protocol described previously 34

Imaging and 3D-reconstruction
For P1 tear duct reconstruction, fresh frozen mouse heads were cut coronally at 30 m, stained with anti-p63 antibody to identify tear duct on each section.
For embryos from E11-14, sections were cut parasagittally at 100m, stained with E-cadherin antibody to identify developing tear duct.

Quantification and statistics
Cell axis orientation was quantified by angle between apicobasal axis (defined by Grasp65 staining, Fig. 6) and tubule axis indicated by E-cadherin staining.
Similarly, cell division orientation was defined by angle between mitotic spindle axis indicated by acetylated-tubulin staining and tubule axis. A total of 236 wild type and 215 mutant cells from 4 animals, and 61 wild type and 53 mutant cells from 8 animals were quantified for cell axis and cell division orientation, respectively.