Soma–germ contact across the basement membrane in the ovary

Abstract

Epithelial cells interact with other cells and environments at their apical side, while the basement membrane typically impedes such interaction at the basal surface. One notable instance is communication between soma and germ cells within the ovary across numerous bilaterian taxa. This contact underlies proper oogenesis and subsequent embryogenesis. Throughout the history of morphology and cell biology, there has been an emphasis on this heterocellular interaction primarily occurring at the apical side of epithelial cells. Contrary to this long-standing understanding, we uncover that ovarian follicle cells in two myriapod species belonging to phylogenetically basal myriapod clades extend their cytoplasmic processes, penetrating the basement membrane to establish direct contact with oocytes. These discoveries demonstrate that the ovarian soma–germ interaction transverses the basement membrane, suggesting that the basal matrix is not always a physical barrier to soma–germ communication. Furthermore, we find that the ovarian somatic cells in a myriapod directly connect with the oogonia or young oocyte before forming their basement membrane. These results encourage reconsidering the conventional view of soma–germ interaction and suggest an overlooked construction manner of heterocellular communication in epithelial cells.

1. Introduction

Epithelium, a cellular layer covering animal organs, commonly exhibits polarity along the apicobasal axis [1–3]. Along the basal domain, these cells are consistently bordered by a layer of basement membrane—an extracellular matrix serving as a barrier between the epithelium and adjacent tissues [4–7]. In contrast, the apical domain of most cells is free from such a matrix and features microvilli or cytoplasmic projections to face the external environment directly, including coelomic fluid and neighbouring tissues [8–10]. Consequently, contacts and communication between the epithelial and other cells often occur at the apical surface.

One prominent example of epithelial cell contact with other cells is soma–germ communication in the ovary. Female somatic cells near germ cells, referred to as accessory, follicle or granulosa cells, frequently establish physical connections with developing oocytes in various bilaterians, i.e. one of the animal groups, such as arthropods, vertebrates and annelids, characterized by the bilaterally symmetrical body [11–16]. This contact underlies proper oogenesis and subsequent embryogenesis through facilitating nutrient supply to gametes, exchange of molecules and ions and participation in signal transduction for primary axis formation [9,17–25]. The polarized epithelial cells face the oocytes at their apical side. Consistent with the general canon, these polarized cells contact germ cells at their apical side through their cytoplasmic projections and heterocellular junctions.

In the context of the positional relationship between epithelial polarity and oocytes, the ovarian structure of Myriapoda holds significance. The ovarian structure in Myriapoda is characterized by the basement membrane between the follicle epithelium and the oocyte, with oocyte growth occurring within a haemocoelic space—an extraovarian environment [26–30]. Therefore, the follicle cells face oocytes at their basal side. There have been no reports of ovarian soma–germ connections in Myriapoda, despite extensive information on ovarian morphology in numerous species over the past century [26–45]. Furthermore, it was observed that a diplopod species lacks heterocellular connection [28]. These studies suggest that the basement membrane serves as a barrier against the soma–germ interaction in accordance with the general canon.

In contrast to such conventional knowledge, here, we unexpectedly uncover that follicle cells contact oocytes across the basement membrane in Thereuonema tuberculata, a species of the earliest branching centipede, Scutigeromorpha [46–53] and Hanseniella caldaria, a member of Symphyla, the most basal myriapod class in certain phylogenetic inferences [46,47]. In a broader sense, this finding suggests novel perspectives on ovarian soma–germ communication and unique modes of heterocellular interaction. Specifically, our novel finding is the soma–germ contact completely through the basement membrane of the somatic cells, which deviates from the accepted sense in zoology and cell biology.

2. Material and methods

(a). Animals

Thereuonema tuberculata (Wood, 1862) (Chilopoda: Scutigeromorpha: Scutigeridae) adults were collected in Sakuragawa City, Ibaraki Prefecture, Japan, between January and October 2016. Hanseniella caldaria (Hansen, 1903) (Symphyla: Scutigerellidae) adults were collected in Tsuchiura City, Ibaraki Prefecture, and Ibusuki City, Kagoshima Prefecture, Japan, between May 2015 and March 2016. Some individuals were kept in plastic cases at 25°C.

(b). Histology

The whole body of T. tuberculata females was cut into pieces and fixed with Bouin’s solution (picric acid : formaldehyde : glacial acetic acid = 15 : 5 : 1) overnight at room temperature. The fixed specimens were dehydrated with 90% ethanol and cleared in a graded ethanol-n-butanol series. Then, the specimens were immersed in paraffin fluid and subsequently embedded in paraffin (Paraplast Embedding Media Paraplast Plus, Sigma-Aldrich, MO, USA). The embedded specimens were cut into 5-µm-thick serial sections using a microtome (Model 820, American Optical, NY, USA) with stainless knives (Microtome Blade Stainless Steel S35, Feather Safety Razor, Osaka, Japan). The sections were visualized by azan (Azocarmine G, Aniline Blue-Orange G) staining and observed with a light microscope (BH2, Olympus, Tokyo, Japan).

(c). Transmission electron microscopy

In T. tuberculata, the ovary was dissected from the body and sectioned. In H. caldaria, the whole body, including the ovary, was cut into some blocks of 1 mm in length. Each sample was pre-fixed with 2.5% glutaraldehyde in 0.1 M phosphate buffer (PB) (pH 7.0−7.2) for 3 h at room temperature and then post-fixed with 1% osmium tetroxide in the same buffer for 1 h at 4°C. After washing with PB, the fixed samples were dehydrated in a graded acetone series and then immersed and embedded in the Quetol 651 resin (Nisshin EM, Tokyo, Japan). The resin blocks containing the samples were polymerized for 36−60 h at 60°C and then cut into ultrathin sections of 50−80 nm with a diamond knife on an ultramicrotome (EM UC7, Leica Microsystems, Wetzlar, Germany). The sections were then viewed under a transmission electron microscope (H-7650, Hitachi, Tokyo, Japan) equipped with a charge-coupled device camera (Velta 2k × 2k, Olympus, Tokyo, Japan) at 80 kV.

3. Results

First, we observed the follicle cells to confirm the Myriapoda-type ovarian structure in T. tuberculata. A single ovary was located on the dorsal side along the anteroposterior axis (figure 1a,b). The follicle epithelium continued with the ovarian epithelium, i.e. the ovarian wall, and consisted of a single layer of flattened cells (figure 1c). There was a lining in a homogeneous mono-laminated matrix between the follicle cells and oocyte (figure 1d,e). The matrix beneath the follicle cells was a continuous layer with that of the ovarian epithelial cells (figure 1d). Electron-dense structures, like hemidesmosomes, were found between the follicle cells and the matrix (figure 1d,e). These features were consistent with those of a basement membrane described in previous studies on myriapod ovaries [26–30]. There was a homogeneous matrix covering vitellogenic oocytes. Since the oocyte-covering matrix was separate from the basement membrane, it was identified as a vitelline membrane (figure 1d,e). Haemocytes were found between the basement and vitelline membranes (figure 1d). The basement membrane was situated between the cells and oocytes (figure 1e). On the opposite side, adjacent cells adhered to each other through adherens junctions (asterisks in figure 1f). Thus, oocytes grew within the haemocoel, i.e. outside of the ovary, in T. tuberculata.

Figure 1.

(a) Schematic of the location of the ovary in T. tuberculata. (b) Cross section of the ovary of the frame in (a). This image was made by merging multiple images. Paraffin sections sections visualized by azan staining. (c) The enlargement view of the white frame in (b). (d) The TEM image of the white frame in (a). (e) The basic morphology of the follicle epithelium in T. tuberculata. (f) The adherens junction of the follicle cells. Asterisks and arrowheads indicate the adherens junction and hemidesmosomes, respectively. Abbreviations: bm: basement membrane, fc: follicle cell, fe: follicle epithelium, h: haemocyte, hc: haemocoel, oe: ovarian epithelium, ol: ovarian lumen, ovm: outer vitelline membrane, pvo: previtellogenic oocyte, vm: vitelline membrane, vo: vitellogenic oocyte. Light microscopy (LM) (b–c). TEM (d–f). Scales: 50 µm (b,c), 2 µm (d), 0.5 µm (e), 1 µm (f).

We then elaborately observed the space between oocytes and follicle cells to explore a soma–germ connection. Our observation on many serial sections revealed that some follicle cells of T. tuberculata extended their cytoplasmic processes to the oocyte (figure 2a–d). The basement membrane persisted in these cells with the hemidesmosome-like structures (figure 2a,b). Thus, the cytoplasmic processes penetrated the basement membrane of the follicle epithelium, reaching the surface of the oocyte. The cytoplasmic processes bordered the oocyte with narrow gaps (figure 2c). Electron-dense materials were also distributed along the gaps. These features of the contact zone are consistent with those of a gap junction. Therefore, the cytoplasmic process of the follicle epithelium was directly in contact with the oocyte through heterocellular junctions. Furthermore, some vesicular structures existed in the follicle cells in contact with oocytes (figure 2c).

Figure 2.

(a) The follicle cells extending cytoplasmic processes (white arrows) in T. tuberculata. (b) The enlargement view of the white frame in (a). (c) The oocyte-contact zone of the follicle cell. Note that the cytoplasmic process was largely interwoven, resulting in an image that is not continuous with the follicle cells. (d) The adherens and septate junctions between the oocyte-contact cell and its neighbouring cell. (e) The follicle cells with the cytoplasmic process (white arrow) in H. caldaria. (f) Schematic of the summary of the ovarian soma–germ contact. Asterisks, arrowheads and black arrows indicate the adherens junction, hemidesmosomes and heterocellular junction, respectively. Abbreviations: bm: basement membrane, h: haemocyte, hc: haemocoel, mv: microvilli, nfc: neighbouring follicle cell, ofc: oocyte-contact cell, pvo: previtellogenic oocyte, sj: septate junction, v: vesicle, vm: vitelline membrane, vo: vitellogenic oocyte. Scales: 1 µm (a,c,e), 0.5 µm (b,d).

To confirm the epitheliality of the oocyte-contact cells, we elaborately observed the continuity of these cells to the neighbouring epithelial cells. We found that the adherens junction was detected between the oocyte-contact cell and the neighbouring cell on the surface opposite the oocyte-contact zone (figure 2d). The septate junction was also found on the basolateral domain of the oocyte-contact cell (figure 2d). Consistently, the basement membrane was continuous between the oocyte-contact cell and neighbouring cells (figure 2d).

To investigate the ovarian soma–germ contact in the other myriapod group, we observed the follicle cells of H. caldaria, a symphylan species whose ovarian structure was reported to be the myriapod type in a previous study [27]. As in the case of T. tuberculata, the follicle cell of H. caldaria possessed a cytoplasmic process, penetrating its basement membrane and reaching the surface of the oocyte (figure 2e). We summarized the nature of the ovarian soma–germ contact in these myriapod species in figure 2f.

Finally, we observed the contact zone between the follicle cells and oocytes during oogenesis in T. tuberculata to elucidate how ovarian somatic cells pierce the basal matrix in Myriapoda. We found that somatic cells closely contacted germ cells, either oogonia or young oocytes (figure 3a–c). At this stage, we could not detect a basement membrane in the somatic cells, indicating that the cells were unpolarized. During the previtellogenic stage, before the formation of the vitelline membrane, the basement membrane could be observed underneath the follicle cells with hemidesmosome-like structures, indicating that the cells were likely polarized (figure 3d). The follicle cells extended cytoplasmic processes across the basement membrane (figure 3d). These cytoplasmic processes persisted during the vitellogenic stage, penetrating the basement membrane and massive double-layered vitelline membranes (figure 3e,f). Therefore, the soma–germ contact occurred before the basement membrane formation was completed and existed throughout oogenesis in T. tuberculata.

Figure 3.

(a) The LM image of the earlier oogenetic stage, i.e. oogonia or young oocyte stage. Paraffin sections sections visualized by azan staining. (b) The TEM image of the earlier oogenetic stage. (c) The enlargement view of the white frame of (b). The blue and red areas indicate the female germ cell and the ovarian somatic cell, respectively. (d) The previtellogenic stage. (e) The vitellogenic stage. (f) The enlargement view of the white frame of (e).The black arrows and arrowheads indicate the soma–germ connection and hemidesmosomes, respectively. Abbreviations: bm: basement membrane, fc: follicle cell, ivm: inner vitelline membrane, mt: mitochondria, mv: microvilli, fgc: female germ cell, oo: oocyte, osc: ovarian somatic cell, ovm: outer vitelline membrane, v: vesicle. Scales: 20 µm (a), 2 µm (b), 0.5 µm (c,f), 1 μm (d,e).

4. Discussion

We demonstrated that the cytoplasmic processes of follicle cells penetrate the basement membrane, establishing connections with oocytes in two myriapods: T. tuberculata and H. caldaria. Close contacts between female germ cells and the cytoplasmic processes or microvilli of ovarian somatic cells have been observed across various ecdysozoan groups, including Nematoda [54,55], Chelicerata [56–64] and Pancrustacea [65–72]. However, Myriapoda has traditionally been considered an exception because such contact has not been previously described, in addition to a previous report indicating the absence of heterocellular connections in a diplopod species [28]. Furthermore, certain features of the myriapod ovary, such as a persistent basement membrane between follicle cells and germ cells [26–30], have supported this interpretation. Contrary to this conventional belief, our findings represent evidence of soma–germ communication in Myriapoda, highlighting the ubiquity of ovarian soma–germ contact among Arthropoda.

More broadly, the contact between follicle cells and oocytes in myriapods challenges a traditional canon in soma–germ connections within the ovary. In the ovaries of many animals, including vertebrates, nematodes and insects, epithelial cells typically adhere to developing oocytes [9,11,12,14–16,55]. These cells face oocytes at their apical side, establishing soma–germ interactions through heterocellular junctions. To the best of our knowledge, there have been no reports of ovarian somatic cells contacting oocytes by penetrating their basement membrane. Our findings provide the first evidence of ovarian soma–germ interactions completely across the basement membrane, demonstrating that the basement membrane does not always serve as a barrier between somatic cells and germ cells.

It has been known that unpolarized cells generally exist within the epithelium in some epithelial tissues (e.g. [73]). Thus, it is possible that the oocyte-contact cells of T. tuberculata may be unpolarized even in the previtellogenic and vitellogenic stages. However, the oocyte-contact cells connected with the neighbouring epithelial cells via the apical adherence and basolateral septate junctions and adhered to the basement membrane through the hemidesmosome-like structure that is the general structure between the polarized epithelial cells and the basement membrane [74,75]. These results suggest that the oocyte-contact cells are likely polarized in the previtellogenic and vitellogenic stages. While cytoskeleton and molecular localization dynamics will examine the polarity of the oocyte-contact cells, it is reasonable that the follicle cells of T. tuberculata come into contact with the oocytes through the cytoplasmic processes growing from their basal domain.

In certain tissues, epithelial cells can communicate with other cells at the basal side in some distinct manners [6,76]. During some developmental processes, e.g. epithelial–mesenchymal transition and oncogenesis, epithelial cells sometimes undergo de-epithelization and break their basement membrane [77,78]. Also, cells near the epithelium, such as neural, muscle and immune cells, can insert projections like tunnelling nanotubules into epithelial cells through pores in the basement membrane [6,79]. These two ways are common in that de-epithelialized or non-epithelial cells breach the basement membrane. Additionally, it is hypothesized that epithelial cells and adjacent cells burrow filopodia-like thin projections, called cytonemes, into the basement membrane, establishing communication with each other in the matrix [80–82]. Therefore, epithelial cells can interact with other cells at their basal surface through loss of epitheliality, invasion by other cell types or utilizing the basement membrane as a scaffold rather than crossing it. In contrast, the follicle cells in the myriapod ovary keep their epithelial nature, extend their cytoplasmic processes and fully penetrate the basement membrane. In this context, it is suggested that the soma–germ interaction in the myriapod ovary may represent a simple but overlooked mode of heterocellular interaction across the basement membrane.

The soma–germ contact in T. tuberculata was established before forming the basal matrix. While further examination is needed to determine whether the earliest contacts are identical to those after the previtellogenic stage, this result suggests that the penetration of the basement membrane may be a secondary consequence of basal matrix secretion around the soma–germ contact rather than a result of the breakdown of the completed matrix. If so, the construction of the soma–germ contact in T. tuberculata differs from the previously known heterocellular connections at the basal domain that occur after basement membrane secretion [6,77–79]. In this study, we could not observe the polarization process of the follicle cells or the construction of the soma–germ connection in the other two species due to the limitation of the developmental stages of the samples. Further investigation into the mechanism and manner of soma–germ contact across the basement membrane in myriapod species is needed to test our concept.

The species in focus of this study belong to the basal centipede group [46–53] and Symphyla. Moreover, the soma–germ connection is generally involved in somatic cell contribution to oogenesis [83]. Consequently, it is inferred that the ovary with the soma–germ connection persisted in the common ancestor of Chilopoda and Symphyla, engaging in communication with oocytes and potentially playing a role in oogenesis. Further observations of taxa yet to be examined will enhance evolutionary history.

Overall, our findings unveil a more widespread prevalence of ovarian soma–germ contact among arthropods than previously believed and present evidence for ovarian soma–germ interaction across the basement membrane, an exception to the conventional view in zoology. The myriapod ovary, with its unique features, may be a goldmine of discoveries that can be exceptions to conventional canons.

Ethics

This work did not require ethical approval from a human subject or animal welfare committee.

Data accessibility

All data for this study are included in the main figures.

Declaration of AI use

We have not used AI-assisted technologies in creating this article.

Authors’ contributions

Y.C.: conceptualization, investigation, methodology, resources, visualization, writing—original draft, writing—review and editing; K.Y.: conceptualization, methodology, resources, writing—review and editing.

Both authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.