Nematocyst sequestration represents an excellent opportunity to investigate how ubiquitous cellular mechanisms like phagocytosis can be modified to selectively internalize and store extrinsic structures. However, laboratory models for nematocyst sequestration have long remained elusive, largely due to difficulties in culturing nudibranch species in the laboratory, including low levels of settlement and metamorphic success in the absence of specific environmental cues [37–39] and limitations on culturing the prey for a given species . With Berghia stephanieae, these challenges have been overcome due to its lecithotrophic development [41, 42] and dietary preference for Exaiptasia diaphana [40, 42], a cnidarian species that is simple to culture under laboratory conditions . Here, we show that the nudibranch species B. stephanieae is an excellent model for nematocyst sequestration, in part due to the ease of access to early developmental stages where juveniles are more transparent and manipulable. These tools have allowed us to address the “when” and “where” of selectivity for nematocysts in B. stephanieae. Our results clearly indicate that in B. stephanieae, nematocyst sequestration begins shortly after feeding and prior to ceras formation, and nematocyst selectivity occurs inside the cnidosac, likely within the cnidophage cells. We discuss the importance of these findings below.
The when - cnidosac development
The morphology of the adult cnidosac in Berghia stephanieae is broadly consistent with previous work in this species, and in the family Aeolidiidae [25, 44]. This morphology includes a short, simple entrance into the cnidosac from the digestive gland, a thick (multi-layered) musculature surrounding the cnidosac, a putative proliferation zone of cnidophage cells at the proximal end of the cnidosac , multiple nematocysts housed within cnidophage cells along the lining of the cnidosac, and an exit. However, we did not identify a clear epithelial lining inside the cnidosac toward the tip (i.e., a cnidopore), which is a structure commonly found in members of Aeolidiidae, including Aeolidia papillosa , Anteaeolidiella chromosoma, and Cerberilla amboinensis . The cnidopore has been hypothesized to be a special adaptation for releasing the exceptionally long and narrow nematocysts sequestered from anemones . The lack of this structure in B. stephanieae suggests that this is not a necessary feature for managing such nematocysts.
Sequestered nematocysts appear within fully-formed cnidosacs within 2-4 days after feeding. Cnidosacs are defined as a “muscular prolongation of the digestive gland located in the apex of each ceras” . In B. stephanieae juveniles, digestive diverticula (branches) and cnidosacs develop before the beginning of cerata formation (Figures 3B-B”, 4B-B”). These early stage cnidosacs, prior to ceras formation, already have a clear musculature (Figure 4B”, C”). Furthermore, nematocysts are also clearly bound inside what appear to be cnidophages (Figure 3B’-B”), indicating that, if not fully mature, these cnidosac structures are at least fully functional. These results differ from regeneration work in other nudibranch species, including Hermissenda crassicornis [41, 45, 46] and Pteraeolidia semperi . In both species, experiments following the regeneration of cerata after autotomy (or removal) suggested that cerata and the digestive gland form first, followed by the cnidosac [45, 46]. In Pteraeolidia semperi, cnidosacs were documented regenerating from a cell aggregation at the tip of the regenerating digestive tract . A prior study on B. stephanieae, focused on neuromuscular development, also incorrectly identified the timing of cnidosac development after ceras formation, based on the presence of filled cnidosacs under light microscopy . These inconsistencies may be due to differences among species [45, 46], differences between regeneration and development [45, 46], or lower resolution imaging techniques [45, 46].
Our results show that nematocyst sequestration in B. stephanieae begins much earlier than previously thought. This fact will be useful for studies in B. stephanieae in the long term, because the cerata and cnidosacs at these early juvenile stages are more accessible visually due to less external pigmentation relative to later stages . Similarly, our ability to detect nematocyst sequestration this early in juvenile development means that in the future we could score gene knock-out phenotypes (e.g. generated by CRISPR/Cas9 genome editing) at juvenile stages. Clarification of the timing of nematocyst sequestration may also allow for the use of transient knock-down methods in B. stephanieae, such as morpholinos or RNA interference . B. stephanieae juveniles therefore promise to be an excellent tool for investigating the cell and molecular processes involved in nematocyst sequestration.
The where - nematocyst selectivity
Our results show that Berghia stephanieae selectively sequesters nematocysts (over spirocysts or dinoflagellates) inside the cnidosac. Most studies on nematocyst sequestration in nudibranchs have assumed, but not demonstrated, that only nematocysts are sequestered, or have not clarified the type of cnidocysts found within nudibranch cnidosacs . Three types of cnidocysts are produced in the phylum Cnidaria: nematocysts, spirocysts, and ptychocysts. Nematocysts are the most widespread across the phylum, with spirocysts and ptychocysts only known from Hexacorallia (and ptychocysts only from the hexacorallian order Ceriantharia) . The lack of clarification regarding which cnidocysts are sequestered may be due to the fact that few nematocyst-sequestering nudibranchs feed on cnidarians known to possess ptychocysts or spirocysts (i.e., Hexacorallia) . However, nudibranchs in Aeolidiidae feed on anemones (including Exaiptasia diaphana), which possess high concentrations of spirocysts in their tentacles for prey capture (e.g., Figure 2C) [50, 51]. We know of only one study that suggested the presence of spirocysts inside the cnidosac, which were identified in specimens of an undescribed species of Spurilla from Argentina . However, no images of these structures were provided, so independent confirmation is impossible. The selectivity for nematocysts we find in B. stephanieae, however, may be due to the more defensive (and offensive) function of nematocysts in cnidarians. Nematocysts deliver venom as a means of defense and offense , unlike spirocysts  and most likely ptychocysts . Therefore, selectivity of nematocysts over less potent structures like spirocysts provides support for a defensive function of nematocyst sequestration in nudibranchs, which is a long-standing debate (see refs in ).
In addition to differentiating cnidocyst types (spirocysts versus nematocysts), we further identified the types of nematocysts sequestered in the B. stephanieae cnidosac, including basitrichous isorhizas, large microbasic p-mastigophores, and microbasic p-amastigophores of varying sizes (Supplementary Figures 2 and 3). Based on previous work in Exaiptasia diaphana (Exaiptasia pallida in ) and the size and type of the nematocysts found in B. stephanieae (Figure 1C-D; Supplementary Figures 2 and 3), we have evidence of the sequestration of nematocysts from multiple tissues in Exaiptasia (including the acontia and tentacles). In some nudibranch species, researchers have found that certain nematocyst types are preferentially sequestered in the cnidosacs [20, 21, 55, 56]. For example, larger, more penetrant nematocyst types [57, 58] such as mastigophores and long isorhizas, appear to be preferred in some species. We found no direct evidence suggesting B. stephanieae select for particular types of nematocysts, though large microbasic p-mastigophores were prevalent in the cnidosacs. Future studies performing more quantification of nematocyst types compared to their distribution across prey tissues will be necessary to test this hypothesis further in B. stephanieae.
In B. stephanieae, our results support the hypothesis that selectivity for nematocyst occurs within the cnidosacs, most likely with the cnidophages . This support stems from the fact that both nematocysts (Figure 6) and dinoflagellates (Figure 7) can enter the cnidosac during digestion, without first being internalized within digestive cells. Other hypotheses suggest that nematocysts are selected for by modified digestive cells inside the digestive gland, and move along the digestive tract to the cnidosac , or that the entrance of the cnidosac (which is surrounded by a muscular sphincter ) is a barrier to all structures except for nematocysts. Given our evidence for selectivity occurring inside the cnidosac, we now know to target cnidophages when investigating the molecular mechanisms of nematocyst sequestration. In cnidarian-dinoflagellate endosymbiosis, recognition and phagocytosis appear to be mediated by microbe-associated molecular pattern (MAMP)-pattern recognition receptors (PRR) that target a variety of molecular patterns on the cell surface of potential symbionts . Similar interactions may also be used to identify plastids in Sacoglossa sea slugs that sequester functional choloroplasts . But in nudibranchs, nematocysts that are selected for and phatocytosed by cnidophages have been stripped of their nematocyte (the cell where nematocysts form) [4, 30]. This raises questions as to whether cnidophages can also use MAMP-PRR interactions to select for nematocysts over other tissues and structures from their prey (e.g., spirocysts or dinoflagellates). B. stephanieae would be an excellent system for investigating how phagocytosis has been modified to select for structures that have been stripped of their original cell.