CD133/Prom1 marks proximal mouse oviduct epithelial progenitors and adult epithelial cells with a low generative capacity

ABSTRACT The epithelium lining the oviduct or fallopian tube consists of multiciliated and secretory cells, which support fertilization and preimplantation development, however, its homeostasis remains poorly understood. CD133/Prom1 expression has been used as a marker to identify adult stem cell populations in various organs and often associated with cancer cells that have stem-like properties. Using an antibody targeted to CD133 and a Cre recombinase-based lineage tracing strategy, we found that CD133/Prom1 expression is not associated with a stem/progenitor population in the oviduct but marked predominantly multiciliated cells with a low generative capacity. Additionally, we have shown that CD133 is disparately localised along the oviduct during neonatal development, and that Prom1 expressing secretory cells in the ampulla rapidly transitioned to multiciliated cells and progressively migrated to the ridge of epithelial folds.


Introduction
Prom1 (Prominin-1) is a transmembrane glycoprotein whose expression was first characterized on the surface of hematopoietic stem cells using the antigen CD133 Miraglia et al. 1997). Prom1/CD133 is now recognized as an adult stem cell marker and has been used to isolate various tissue-specific stem cell populations across multiple organs (Uchida et al. 2000;Sagrinati et al. 2006;Richardson et al. 2004). CD133 has also been identified on the surface of cancer cells with stem cell like properties in many tumors including ovarian, liver, brain, prostate, colon, hepatocellular and lung cancer (Glumac and Lebeau 2018). The specificity of Prom1/CD133 as an adult stem cell marker has however come into question due to conflicting reports of CD133 specification, potentially due to the use of different antibodies between studies, and the broader expression pattern seen in Prom1 reporter mouse lines, indicating potential differences between CD133 antibody stainings and Prom1 expression (Weigmann et al. 1997;Pfenninger et al. 2007;Florek et al. 2005;Shmelkov et al. 2008;Snippert et al. 2009;Zhu et al. 2009Zhu et al. , 2016. In a recent study Zhu and colleagues performed a lineage tracing experiment of Prom1 expressing cells and found variation in the generative and proliferative capacity of Prom1 expressing cells between organs (Zhu et al. 2016). It was found that the Prom1 expressing cells with a high generative capacity had a higher risk of tumor formation after an oncogenic insult, indicating that the stem cell characteristics of Prom1 expressing cells varies between different organs and that locating those populations with a high generative capacity may identify highly susceptible cells to malignant transformation.

Biology Open • Accepted manuscript
The oviduct epithelium is constituted of multiciliated and secretory cells, that aid the transport, function and survival of gametes and embryos (Li and Winuthayanon 2017). It has been shown by lineage tracing studies that secretory cells are proliferative and can differentiate into multiciliated cells, suggesting that secretory cells represent a bipotent progenitor (Ghosh, Syed, and Tanwar 2017). Mounting evidence suggests that these cells are also the cell-of-origin in many cases of high-grade serous ovarian carcinoma (HGSOC), the most common and aggressive form of ovarian cancer (Kroeger and Drapkin 2017). The presence of an adult stem cell, ie, an undifferentiated cell whose progeny replenish dying cells, has not been characterised in the oviduct. Multiple studies have however eluded to the presence of a resident adult stem cell by the identification of label-retaining cells and organoid forming cells, enriched in the distal oviduct (Patterson and Pru 2013;Wang et al. 2012;Xie et al. 2018;Paik et al. 2012).
The expression of several common adult stem cell markers have also been reported such as Cd44, Prom1 and Lgr5 (Almasry and Elfayomy 2018;Alwosaibai et al. 2017;Ng et al. 2013). However, there have been no in vivo characterisations of these populations that confirm their status as a progenitor or adult stem cell.
In this study we investigate the stem cell properties of CD133/Prom1 expressing cells in the mouse oviduct epithelium using Cre-based lineage tracing techniques and CD133 antibody staining in adult and neonatal mice. We find that CD133/Prom1 expression is limited to proximal epithelial progenitors and then becomes restricted to a sub population of oviduct epithelial cells, that are predominantly multiciliated cells, along the length of the oviduct. Our lineage tracing experiments revealed that Prom1 expressing secretory cells in the ampulla rapidly differentiated to multiciliated cells and progressively moved from the base to the ridge of epithelial folds but remain restricted to the base of epithelial folds in the isthmus. Taken together our results highlight the developmental and homeostatic differences between distal and proximal populations and provide no evidence of a resident adult stem cell population in the mouse oviduct epithelium.

CD133 is disparately localised along the mouse oviduct epithelium on the apical surface of secretory and multiciliated cells
To determine the precise localization pattern of CD133 in mouse oviduct epithelial cells, we performed whole-mount 3D imaging of Fltp-H2B-mVenus (Flattop-driven H2B-Venus) transgenic mice with an antibody targeted to the CD133 antigen (n = 6 mice).
Fltp-H2B-Venus mice contain a LacZ-2A-H2B-Venus cassette knocked in to the Fltp locus resulting in Venus expression in the nucleus of multiciliated cells (Gegg et al. 2014). In the infundibulum and ampulla of adult mice, CD133 was observed extensively at the luminal side of the epithelium with strong staining marking both individual and clusters of cells ( Fig. 1A and C). In transverse sections, CD133 was observed at the apical surface of both multiciliated (white arrow) and secretory cells (yellow arrow) (Fig. 1B and D). In the ampulla-isthmus junction, CD133 was primarily co-localised with multiciliated cells (Fig. 1E and F). The distribution patterns of CD133 positive cells was generally scattered, with some coherent and continuous clusters seen in 3D reconstruction. By opening the oviduct and 3D imaging from the luminal surface we found CD133 to be covering the apical surface of multiciliated cells and located basal to cilia (Fig. 1G). In the isthmus, CD133 was enriched on the apical surface of multiciliated cells located at the base of transverse epithelial folds ( Fig. 2A, B and D). No CD133 was detected in the uterotubal junction, which also contained no multiciliated cells ( Fig. 2 C and E).

CD133 becomes absent in proximal oviduct secretory cells and emerges in distal regions during multiciliated cell differentiation in neonatal oviducts
To determine how the pattern of CD133 positive cells is established in the oviduct epithelium, we stained for CD133 in neonatal oviducts and performed 3D whole-mount imaging using Fltp-Venus::Sox17-mCherry mice to label multiciliated and epithelial cells respectively (Burtscher et al. 2012) (n = 3 mice per stage). We have previously shown Sox17 to be expressed in secretory cells of the distal oviduct and all epithelial cells in Biology Open • Accepted manuscript the proximal region . At the postnatal stages analysed, Sox17 is initially expressed in all epithelial cells but begins to be lost in some distal epithelial cells that are undergoing multiciliogenesis (Fig. S1A). On postnatal day 1, prior to multiciliated cell differentiation, we identified uniform CD133 staining on the apical surface of epithelial progenitors in the isthmus and uterotubal junction but absent in the infundibulum and ampulla, with a sharp boundary at the ampulla-isthmus junction (white arrows in Fig. 3A and Fig. S1B), consistent with our previous observations of distinct distal and proximal epithelial populations Ford et al. 2021). At Similarly, to postnatal day 1, we identified a sharp boundary at the ampulla-isthmus junction with proximal regions continuing to have a more uniform pattern of CD133 staining with some areas of stronger staining and only scattered multiciliated cells (boundary marked by white arrows in Fig. 3B, Fig. S1C and F). As the proportion of multiciliated cells increased in the infundibulum and ampulla on postnatal day 6, we found a significant increase in the number of CD133 positive cells, (Fig. 3C, Fig. S1D and G). An increase in the number of multiciliated cells was also seen in the ampullaisthmus junction and isthmus. Clear CD133 positive cells were identified within these ciliated regions with some uniform staining still present in the isthmus (Fig. 3C, Fig. S1D and G). No ciliogenesis was identified at the uterotubal junction and CD133 remained relatively uniform along the epithelium. By postnatal day 10, the adult pattern of the CD133 positive cell distribution in the infundibulum and ampulla was established, with enrichment on the apical surface of multiciliated and secretory cells (Fig. 3D, Fig. S1E and H). In the ampulla-isthmus junction and isthmus, CD133 had become absent from many epithelial cells and was predominantly associated with multiciliated cells (Fig. 3D, Tamoxifen is an estrogen receptor modulator and has been reported to induce hyperplasia in the mouse oviduct epithelium after acute prenatal and sustained adult exposure (Diwan, Anderson, and Ward 2014;Niwa et al. 1998). To determine the effects of an acute adult exposure of tamoxifen we administered five doses of tamoxifen and collected oviducts 6 hours after the final dose. Compared to homeostatic levels of Biology Open • Accepted manuscript proliferation, measured by ki67 staining, we identified no significant difference in proliferation in the isthmus after tamoxifen exposure, and the proportion of ki67+ cells fell within the normal range seen during the estrous cycle in the ampulla ( Fig. 4H and expressing Tdtomato+ cells to 1% after 1 month which was sustained (Fig. 5F). In the isthmus, however, we did not observe this due to the ubiquitous expression of PAX8 in all epithelial cells.
To determine the movement of Prom1 expressing cells, each labeled cell during lineage tracing was assigned a position in relation to the base, side and tip of epithelial folds in transverse sections (Fig. 5H). In the isthmus labeled cells remained restricted to clusters of cells at the base of epithelial folds (Fig. 5I). In the ampulla however, we detected a progressive increase in the proportion of labelled cells at the ridge of epithelial folds and a decrease at the base (Fig. 5J), suggesting a base to ridge drift of Biology Open • Accepted manuscript epithelial cells over time. In keeping with this observation, the majority of proliferating cells (Ki67+) in the ampulla were detected at the base of epithelial folds (Fig. 5K).

Conclusion
CD133/Prom1 is a well characterised adult and cancer stem cell marker in many tissues Miraglia et al. 1997;Uchida et al. 2000;Sagrinati et al. 2006;Richardson et al. 2004;Glumac and Lebeau 2018). In the presented study we  Harwalkar et al. 2021). In our lineage tracing study, we find that the distribution pattern of Prom1 expressing cells and the drift of these cells is distinct between the ampulla and isthmus.
In addition, only epithelial cells in the ampulla responded to changes during the oestrus cycle resulting in an increased proportion of proliferating during estrus. These results support the notion of independent homeostatic mechanisms within the morphologically distinct regions of the oviduct.

Mouse stains and injections
All

Whole-mount immunostaining, tissue clearing, and 3D confocal imaging
Mouse female reproductive tracts were collected and straightened by removing the mesosalpinx. The straightened oviduct was fixed with DMSO (Dimethyl sulfoxide, Sigma Aldrich D8418): methanol in the ratio 1:4 and cut into three to four pieces prior to placing at −20°C overnight. For the surface view of multiciliated cells, a single longitudinal cut was made along the length of the oviduct and the oviduct opened to expose the luminal surface which was then fixed as described above. The antibody staining protocol has been previously described in (Arora et al. 2016

Imaging and statistics
All imaging was performed on an LSM 800 confocal microscope (Zeiss) at the advanced bioimaging facility (McGill University). Image analysis was conducted with FIJI using custom designed software (Schindelin et al. 2012). Statistical analysis was then performed either using Excel (Microsoft) or R-studio.

Declaration of interests
The authors declare no competing interests. staining was detected on postnatal day 6 in the infundibulum and ampulla. More extensive multiciliogenesis was also detected in the ampulla-isthmus junction and isthmus which was coupled with areas of higher and more discrete CD133 staining.

Author contributions
Diffuse CD133 staining was still seen in the isthmus and uterotubal junction. (D) By postnatal day 10 CD133 in the infundibulum and ampulla resembled staining patterns seen in adult tissue. In the ampulla-isthmus junction and isthmus the pattern of CD133 and multiciliated cells had yet to form while CD133 was now barely detectable in the uterotubal junction. Scale bars = 100µm.

Biology Open • Accepted manuscript
Biology Open • Accepted manuscript