**3. Results**

#### *3.1. Ecological Analysis*

On the surveyed reefs, benthic coverage was dominated by hard corals (36 ± 8.6%), followed by dead corals (23.3 ± 16.1%) and coral rubble (20 ± 6.4%), and the same trend was observed in both sites (Figure S1). A total of 39 scleractinian genera were recorded (Table S3) and among them the genus *Pectinia* showed the highest cover percentage (~11%), followed by *Dipsastraea* and *Merulina* (~9% and 8%, respectively). All the other coral genera displayed a cover percentage close to or less than 5% (Table S3). However, the two sites at Pulau Hantu and Kusu Island showed a distinct abundance and diversity of the various coral genera (Table S3). Indeed, in Pulau Hantu 33 genera were observed, while Kusu Island showed a higher coral diversity with 38 genera recorded (Table S3). In addition, Pulau Hantu reef was dominated by corals belonging to the genera *Pectinia* (~13%), *Merulina* (~10%), *Goniopora*, and *Dipsastraea* (both ~8%), while Kusu Island displayed a greater heterogeneity in terms of coral genera coverage, with *Dipsastraea* (~10%), *Heliopora* and *Pectinia* (all ~8%), *Favites* (~7%), and *Pachyseris*, *Montipora* and *Platygyra* (all ~6%) showing the highest abundance (Table S3).

Overall, 1044 coral colonies were observed, and the overall prevalence of the coral– *Waminoa* sp. association was 10.4 ± 2.3%, with the site on Kusu Island showing a significantly higher prevalence compared to Pulau Hantu (ANOVA, F(1,34) = 3.1, *p* = 0.012; Figure 3). In total, 17 out of the 39 scleractinian genera recorded in the study area were found in association with *Waminoa* sp. and significant differences in the association prevalence were detected among the host coral genera (ANOVA, F(12,143) = 2.39, *p* = 0.008; Figure 4A). In particular, *Lobophyllia* clearly displayed the highest prevalence, followed by *Goniastrea* and *Favites*, and later by *Mycedium*, *Platygyra*, *Oxypora*, and *Pachyseris*, while for the five other host genera the prevalence recorded was lower than 10% but higher than 5% (*Pectinia*, *Echinopora*, *Fungia*, *Ctenactis*, and *Podabacia*, Figure 4A). However, significant differences in the association prevalence were observed only between *Lobophyllia* and *Echinopora*, and *Fungia*, *Ctenactis* and *Podabacia* (Figure 4A). In addition, the five other scleractinian genera, namely *Merulina*, *Porites*, *Dipsastraea*, *Hydnophora*, and *Montipora*, were found associated with *Waminoa* sp. but with prevalence < 5% (4.2, 3.7, 3.5, 2.5, and 1.9%, respectively). The prevalence patterns recorded in both sites were not uniform and did not fully reflect those recorded in the whole study, with *Goniastrea*, *Pachyseris* and *Lobophyllia* showing the higher prevalence in Pulau Hantu, and *Lobophyllia*, *Favites*, *Mycedium*, and *Oxypora* in Kusu Island (Figure S2). The selectivity coefficient Ei allowed the comparison of the relative abundance of coral genera colonized by *Waminoa* sp. with the relative abundance of the same coral genera recorded in the whole study area (Figure 4B). The analysis was performed only for coral genera showing an association prevalence > 5%. It revealed that *Mycedium* was the preferred host for *Waminoa* sp., followed by *Lobophyllia*, *Oxypora*, and, surprisingly, *Ctenactis*, which was among the coral genera that showed the lowest prevalence of the association (Figure 4B). Moreover, *Waminoa* sp. showed a marked avoidance for high/medium-prevalence genera, such as *Platygyra, Pachyseris* and *Pectinia* (Figure 4B).

**Figure 3.** Prevalence (%) of *Waminoa* sp.–corals associations in Pulau Hantu and Kusu Island. Numbers above each bar indicate the total number of coral colonies (both with and without *Waminoa*) analyzed per site. Data are expressed as the mean ± SEM. One-way ANOVA was performed between sites. Letters on the bars denote significant difference among sites. Different letters indicate significant difference (*p* < 0.05), while same letter indicates no significant difference (*p* ≥ 0.05).

Most of the coral colonies were moderately (from 11 to 25% of the coral surface) or severely (26–50%) covered by *Waminoa* sp., while a few colonies showed an extreme colonization of worms (>50%) on their surface (Figure 5A). However, no significant differences in the abundance percentages were recorded among the *Waminoa* sp. coverage categories (ANOVA, F(3, 44) = 1.83, *p* = 0.158, Figure 5A). Therefore, the distribution of the flatworms on the coral colonies was heterogeneous, although *Waminoa* sp. mostly occupied less than 50% of the coral's surface (Figure 5A). This pattern was also found in the coral genera showing the highest prevalence of the association, such as *Lobophyllia*, *Goniastrea*, and *Favites*, as well as in the preferred genus *Mycedium* (Figure 5B). In particular, in *Lobophyllia* corals the flatworms mostly colonized from 26 to 50% of the colony surface, while in both *Goniastrea* and *Favites* about 40% of the colonies had less than 10% of their area occupied by *Waminoa* sp. individuals. Almost all the colonies of *Mycedium* hosting the worms had less than 10% of their surface covered, while the *Waminoa* sp. infestation on the coral surface was mostly severe in *Pachyseris* and extreme in *Fungia* (Figure 5B).

**Figure 4.** (**A**) Prevalence (%) of *Waminoa* sp.–corals associations by genus in the whole study area. Data are expressed as mean ± SEM. Letters denote Tukey's significant differences among the different groups (*p* < 0.05); the same letter indicates no significant difference (*p* ≥ 0.05). (**B**) Host preferences of *Waminoa* sp. according to the Van der Ploeg and Scavia selectivity coefficient Ei (−1 = complete avoidance; 0 = random choice; +1 = exclusive preference) for each coral genus. In both graphs, only the coral genera with an association prevalence > 5% are reported.

#### *3.2. Molecular Analysis*

All the sampled coral portions occupied by worms showed no visible surface damage or lesions. All the candidate genes showed reliable amplification, since the efficiency of amplification was within the range of acceptable values of 1.49–2.2 (Table S4, [38]).

Significant differences in biomarker levels among portions of coral tissue colonized or not by *Waminoa* sp. were recorded (PERMANOVA: df = 1, F = 3.372, *p* = 0.007). All the genes showed a lower expression level in samples collected underneath *Waminoa* sp. compared to samples of the same coral colony taken at least 5 cm apart from the flatworms (Figure 6A). Therefore, the presence of *Waminoa* sp. on coral caused a down-regulation of the expression of all the investigated genes in the portion of coral tissue directly in contact with the flatworms. The effect of the presence of *Waminoa* sp. on the gene expression was significant for *C3* (pMCMC = 0.03), *Hsp70* (pMCMC = 0.01), and *Actin* (pMCMC = 0.005), but not for *C-type lectin* (pMCMC = 0.13), (Figure 6A). The highest fold change in expression levels was observed for *Actin*, followed by *Hsp70*, while the lowest change was observed for *C-type lectin*, which showed a non-significant down-regulation (Figure 6B).

**Figure 6.** (**A**) Changes in expression levels of the analyzed genes (log2-transformed) between coral fragments not infested (W/0) and infested with *Waminoa* sp. (W). Significant differences in the gene expression abundance are indicated with asterisks. (**B**) Modulation of each gene as fold change. Fold changes were calculated with respect to levels detected in "W" fragments and were log2-transformed.In both graphs, data are expressed as means (*n* = 5).

### **4. Discussion**

#### *4.1. Ecology of the Waminoa-Coral Association in Singapore Reef*

Our data contributed to extend the geographic distribution of the *Waminoa*–coral association. In fact, to date it has been recorded only in the reefs of the Red Sea [9,11,12,17], Indonesia [13,20,75], Micronesia [76], Australia [77,78], Japan [8,14], and Taiwan [79]. In addition, our results indicate that the association appeared to be abundant in the study area, with a prevalence greater than 10%. In areas close to Singapore, such as Taiwan, less than 1% of the corals analyzed were found colonized by *Waminoa* sp. [79], while in Wakatobi (Sulawesi, Indonesia) a total of 4.8% of all observed hard corals were associated with the acoel worm in 2006 and 2.6% of hard and soft corals in 2007 [13]. However, a comparison between the prevalence obtained in these studies may not be completely reliable, since the investigated geographic areas were characterized by diverse habitat structure and ecological traits, and the survey methods and approaches used were not entirely the same.

Although the presence of the flatworms was reported in both the investigated sites, a significantly higher prevalence of the association was observed on Kusu Island compared to Pulau Hantu. This difference could be explained by the greater diversity of coral genera, rugosity, and reef complexity on Kusu Island [59], which may have contributed to available niches for *Waminoa* sp., as well as by the different environmental and physical characteristics of the sites. Pulau Hantu is sheltered by adjacent and heavily developed islands in an area

of intense industrialization and ship traffic, while Kusu Island experiences comparatively lower anthropogenic impacts and higher exposure to wave action [80,81]. This generates a significantly higher average turbidity, sedimentation, and light attenuation rate in Palau Hantu, resulting in an overall lower light intensity and shallower euphotic depth than Kusu Island [81]. These conditions potentially affect the presence of photosynthetic dinoflagellatehosting organisms such as *Waminoa* flatworms, which may also have more difficulty in colonizing sediment-covered surfaces and may themselves be vulnerable to environmental disturbances.

The *Waminoa* sp. host range was updated with additional scleractinian genera, many of which are new records. Indeed, in Singapore, *Waminoa* sp. was in association with 17 coral genera belonging to six families, namely Lobophylliidae, Merulinidae, Agariciidae, Poritidae, Fungiidae, and Acroporidae. Among them, the family Merulinidae was largely the most represented, as also recorded in Taiwan, despite only six coral genera in total being found infested by *Waminoa* sp. [79]. In Sulawesi (Indonesia), the association with *Waminoa* was confirmed for 21 coral taxa (Wakatobi [13]), but in Bangka Island it was recorded for only 4 coral genera (*Gardineroseris*, *Platygyra*, *Porites*, *Turbinaria* [75]). In the Red Sea, 13 coral genera were found infected [11]. In Japan, *Waminoa* individuals were found on 4 coral genera only, namely *Cycloseris*, *Echinomorpha*, *Echinophyllia*, and *Pachyseris* [14], and 13 scleractinian hosts all belonging to Lobophylliidae [8]. Therefore, *Waminoa* sp. in Singapore coral reefs showed a wide host range. However, only few coral genera such as *Lobophyllia*, *Mycedium, Oxypora*, and to a lesser extent *Goniastrea* and *Ctenactis*, appeared to be preferred hosts. These coral genera are characterized mainly by massive/submassive or laminar/encrusting colony morphologies, while corals without *Waminoa* sp. are typically branching or columnar. On the contrary, in previous studies *Waminoa* sp. was predominantly observed on branching *Acropora* and *Stylophora* corals, as well as in the columnar *Tubastrea* [11,13]. The question of why *Waminoa* sp. colonizes and/or prefers only specific coral taxa remains largely unanswered. We hypothesize that the coral skeleton morphology could represent a factor driving the choice of the flatworms, given that some structures could favor protection from predators, allowing the worm to hide. Since coral mucus represents a possible food source for *Waminoa* spp. [17], the different mucus production among different coral taxa could also represent an additional host selection factor. This is even more relevant in Singapore's turbid reefs, as some corals can increase or decrease mucus production as a defense mechanism in response to persistent sediment stress [82–85]. In this regard, it would be interesting to analyze the mucus production and composition of the different coral taxa to explore possible correlations with the *Waminoa* presence. Finally, since *Waminoa* also feed on zooplankton caught by corals [18], the ability of a coral species to capture zooplankton, which is determined by its morphology, coral polyp size, and the type of tentacles and nematocysts (reviewed in [86]), may play an important role in the *Waminoa* host selection. However, in addition to these hypotheses, we cannot exclude that the *Waminoa* individuals analyzed here, albeit being of a single morphotype, did not belong to a single species but represented a complex of cryptic species, each of which specialized in a different host.

Corals of Singapore showed variable patterns of flatworm density, ranging from colonies that were densely and extremely infested to others that were only moderately and sparsely populated, as previously observed [11]. However, as also occurred in Okinawa [14], *Waminoa* sp. individuals were not equally distributed among different host taxa. In particular, in Singapore we detected that different *Waminoa* infestation rates could be related to the coral growth form. Indeed, in corals with massive growth forms (such as *Lobophyllia*, *Goniastrea*, *Favites*, and *Platygyra*), *Waminoa* sp. showed a heterogeneous pattern of distribution (but in general < 50% of the coral surface was occupied). Corals with a foliose and/or encrusting growth form (*Mycedium*, *Podabacia*, *Oxypora*, and *Echinopora*) were sparsely or moderately covered by flatworms, while Fungidae corals (*Ctenactis* and *Fungia*) appeared extremely colonized by *Waminoa* sp., as also previously observed [20].

#### *4.2. Effect of Waminoa sp. on Coral Putative Immune- and Stress-Response Genes*

*Waminoa* spp. can cause physiological damage to corals by inhibiting photosynthesis, reducing the coral tolerance to environmental stress, and impairing coral respiration and feeding [11,17,20]. Our analysis on coral gene expression produced a detailed description of the early response to stress at cell/tissue level, since changes at the molecular level occur before morphological and physiological impairment appear evident [87,88].

Our results show that *Waminoa* sp. affected the analyzed host molecular pathways associated with the coral's stress tolerance and immunity response, causing a uniform down-regulation of the expression of all the investigated genes. Moreover, this modulation was only observed in the physically undamaged coral tissue portions colonized by *Waminoa* individuals and not in those free from flatworms. This might sugges<sup>t</sup> that, as previously observed in corals infected by bacteria or protozoans [24,46,67], the stress response was confined in a restricted area just below the flatworm, even though polyps are linked together by common tissue in the coral colony.

The complement pathway of the immune system is triggered by lectins binding a pathogen-associated molecule and results in the activation of the complement component factor C3 and C4 [89]. Indeed, in corals, *C-type lectin* and C3 protein were usually up-regulated and activated in response to epizootic diseases [47,48,90,91]. However, their down-regulation may reflect suppression of host immunity, as previously observed for the association between corals and the microalga *Chromera* [92]. Mohamed et al. [92] also suggested that the down-regulation of some PRRs could reflect the host attempting to limit interactions with non-beneficial organisms, since both complement C3 and the *C-type lectin* have been implicated in symbiont recognition and in host-symbiont communication [34]. In addition, the down-regulation of *C-type lectin* and C3 has been observed in corals subjected to temperature/light stress, suggesting that these stresses might compromise the coral's immune defenses and therefore increase the coral's susceptibility to diseases [38,93–95]. Likewise, the decreased expression of the *C-type lectin* and C3 here suggests that the presence of *Waminoa* sp. individuals on coral tissue might interfere with the ability of the whole host coral to respond to the attack of various pathogens and at the same time could make it more vulnerable to environmental stressors. However, while the C3 appeared to be significantly down-regulated by the flatworm presence, the decrease in expression of the *lectin* was not significant. Considering that *C-type lectins* have been shown to respond immediately following an immune challenge [34,90] but may not show any significant response at later times [47], we hypothesize that the observed modulation could be influenced by the sampling times.

The cytoplasmic chaperonin Hsp70 is involved in assembly of newly synthesized proteins and in the refolding of misfolded or aggregated proteins, contributing to the protein transfer to different cellular compartments or to the proteolytic machinery and acting as cellular defensive mechanism [96]. Up-regulation of the *Hsp70* has been proposed as an activator of other components of the coral effector immune systems, such as the prophenoloxidase cascade, in corals infected by pathogens [47]. On the contrary, the down-regulation of Hsp70 in corals reflected the impairment of the cellular defense mechanisms that is due to severe and intolerable stress [97–99], and may indicate a reduced activity of the immune system because of diseases [24]. In addition, since Hsps are ATPdependent chaperones, the decrease of *Hsp70* expression may be related to the high-energy expenditure necessary to reduce the deleterious effects of *Waminoa* sp. and restore cellular damage. However, it is important to underline that, since the roles of *Hsp70* in organismal function are broad, changes in expression of this gene could also reflect changes in other physiological processes.

Actin was the most responsive gene, showing the greatest down-regulation. In addition to being fundamental for cell motility, contractibility, mitosis, and intracellular transport, Actin is also an important part of the nuclear complex, being required for the transcription of RNA polymerases and in the export of RNAs and proteins from the nucleus [100]. Down-regulation of Actin has previously been observed in corals subjected

to thermal stress and acidification [38,68,101,102]. Since Actin is a major cytoskeletal component involved in growth, down-regulation of this gene could be indicative of growth inhibition caused by the presence of *Waminoa* sp. Moreover, the reduced expression of *Actin* may reflect a change in the regulation of gene transcription of proteins involved in cytoskeletal interactions and may imply changes in intracellular transport and cell shape/integrity, as previously suggested [102]. The overall down-regulation of all the analyzed genes may reflect a negative effect of the acoelomate ectosymbiont *Waminoa* sp. on the host coral *L. radians*. However, alternative scenarios should also be considered. For example, it could be possible that *Waminoa* did not cause detectable cellular stress to hosts, preferentially colonizing polyps with reduced defense responses, or interfering with polyp feeding and causing the observed gene down-regulation, as reduced resources would lead to reduced investment in defense.

In conclusion, our study demonstrated that *Waminoa* sp. showed a high prevalence and wide host range in Singapore coral reefs and its distribution patterns were specific to certain scleractinian host genera. Moreover, *Waminoa* sp. could impair both the cellular homeostasis and components of the immune system of the host, thus representing a potential further threat for coral communities living in an area already subjected to multiple stresses, such as sedimentation and light limitation. However, further studies analyzing more genes and biomarkers in different hosts are necessary to have a more complete picture of the association.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/d14040300/s1, Figure S1: Coverage percentage of the different benthic categories in the two sites analyzed and in the whole study area. Data are expressed as the mean ± SEM; Figure S2: Prevalence (%) of *Waminoa* sp.-corals association by coral genus in the two sites. Data are expressed as the mean ± SEM; Table S1: Symbiodiniaceae nucleotide databases; Table S2: RT-qPCR Mastermix used for determining gene expression and efficiency; Table S3: Coverage percentage of each coral genus in the two sites and in the whole study area; Table S4: Gene efficiency (E) for each analyzed gene (SD: standard deviation)

**Author Contributions:** G.M. and D.S. wrote the manuscript (original draft preparation); D.H., D.M., S.M. and E.M. reviewed and edited the manuscript; D.M., S.M., E.M. and D.S. analyzed the results; D.H., G.M. and D.S. conceived and designed the research and experiments; D.H., S.S.J. and R.Z.B.Q. secured funding for this research and for all reagents and materials; G.M., S.S.J., R.Z.B.Q. and R.C.P.-D. performed and supervised the lab activities. G.M. and S.S.J. conducted the field activities. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research is supported by the Temasek Foundation under its Singapore Millennium Foundation Research Grant Programme.

**Institutional Review Board Statement:** All applicable permits and institutional guidelines required to perform the work were followed. Collections were made under permit NP/RP16–156.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The datasets generated during and/or analyzed during the current study are available from the corresponding authors on reasonable request.

**Acknowledgments:** We are grateful to members of the Reef Ecology Lab, National University of Singapore, for their support throughout this project, both in the field and in the laboratory. Special thanks to Nicholas Yap for his support throughout the project. Thanks to Daisuke Taira, Andrea Leong, and Ng Zhi Sheng for their support during diving activities.

**Conflicts of Interest:** The authors have no relevant financial or non-financial interest to disclose.
