VIBRIO CORALLIILYTICUS RE22 TYPE VI SECRETION SYSTEMS CONTRIBUTE TO TEMPERATE CORAL LYSIS AND ENDOSYMBIONT DEATH

Astrangia poculata is a facultatively symbiotic temperate coral that is being explored as a model system for studying the physiology and ecology of cnidarian-microbe symbiosis. Vibrio coralliilyticus is a known causative agent of a class of coral diseases called “white syndromes” that result in bleaching in tropical coral species. It is an effective pathogen due to a wide array of virulence factors including two Type 6 Secretion Systems (T6SS1 and T6SS2). In this study, we investigated the pathogenic potential of V. coralliilyticus RE22Sm in A. poculata and in cultures of its endosymbiont, Breviolum psygmophilum. To independently gauge the antagonistic effects of each of the two T6SSs, allelic exchange mutants of the hcp genes were utilized. In the A. poculata challenge, V. coralliilyticus RE22Sm caused tissue lysis in coral samples. Both aposymbiotic and symbiotic corals were susceptible to infection, and aposymbiotic corals displayed tissue lysis faster than symbiotic corals. Mutation of the T6SS1 hcp1 gene resulted in the greatest attenuation of virulence in the coral system. Coral survival increased from 12% in the wild-type challenged samples, to 60% for those challenged with Δhcp1. Virulence was also attenuated in corals challenged with Δhcp2, with 30% survival. Similarly, B. psygmophilum challenged with the Δhcp1 strain had a 20% increase in both cell survival and chlorophyll a content, compared to cultures exposed to wild type V. coralliilyticus RE22Sm. An hcp1 hcp2 double mutant resulted in minor attenuation of virulence in both coral and endosymbiont trials. Revertant strains with restored wild-type copies of the hcp genes displayed comparable virulence to wild-type V. coralliilyticus RE22Sm. These results suggest that Type 6 Secretion is a major component of pathogenesis against the temperate coral A. poculata and B. psygmophilum. Heightened susceptibility of aposymbiotic coral samples to bacterial challenge is consistent with literature that suggests symbiotic A. poculata is more effective than aposymbiotic colonies at mitigating of environmental stress. The data are consistent with bacterial challenges in an oyster larval system, which indicate that T6SS1 is primarily involved in eukaryotic antagonism.

Thank you for pushing me to be the best that I could be in my education and career to prepare me for the future. Sometimes you never know how much a grounding influence helps until you need it and I can safely say I couldn't have completed this endeavor without your support.

PREFACE
The following thesis has been prepared in manuscript format according to the guidelines of the Graduate School of the University of Rhode Island. This thesis contains a literature review and one manuscript.
The manuscript "Vibrio coralliilyticus RE22 Type VI Secretion Systems Contribute to Temperate Coral Lysis and Endosymbiont Death" has been formatted according to ASM guidelines and will be submitted to AEM.
viii ix

Introduction
Coral bleaching events have been increasing in magnitude and scope since the early 1980s [3], damaging both biodiversity associated with coral reef environments and commercial industries reliant on continued coral survival such as fisheries and tourism [1,2]. Coral reef environments are essential for biodiversity with a conservative estimate of 2,594,000 unique reef associated species dependent on the environment [4]. Coral bleaching has a wide variety of potential causes due to the delicate symbiosis between coral and their dinoflagellate endosymbionts, which, for most tropical corals, is obligate in nature [5]. Ocean acidification and the gradual increase of ocean temperature have been shown to contribute to significant coral stress and loss of symbiosis [6,7]; however, another cause of coral bleaching is bacterial pathogenesis. One prominent type of coral infection is referred to as White Syndrome (WS), occurring in a number of tropical coral species including Pocillopora damicornis, Montipora capitata, and Acropora cytherea [18,21,32]. Although the exact causes of WS have not been identified, WS often occurs in conjunction with an increase in levels of Vibrio species in infected coral microbiomes [22]. Vibrio species including Vibrio shiloi and Vibrio coralliilyticus have been demonstrated to cause bleaching and tissue lysis in coral samples under experimental conditions [17].
Because of the difficulties involved in prevention, avoidance, and treatment of coral reef infections long-term, a greater understanding of the mechanisms behind such widespread collapse is essential [35].
The pathogen V. coralliilyticus is the primary focus of this review due to its wide array of virulence factors and previously described interactions with coral hosts [30]. Virulence factors include multiple secretion systems including a type I, type II, type III, and 2 type VI secretion systems (T6SS) that are capable of translocation a wide variety of anti-eukaryotic and anti-bacterial effector compounds [60-64].
Additionally, V. coralliilyticus is capable of secreting 2 primary extracellular zincmetalloproteases, VcpA and VcpB, the former of which has close evolutionary similarities to the EmpA protease found in V. anguillarum [57,86]. High levels of proteolytic activity produced in high temperatures conditions within the host are primarily thought to be the main factors responsible for coral pathogenesis, likely due to a decline in photochemical efficiency of the coral endosymbiont when exposed to the pathogen [21]. However, recent studies have shown that bacterial knockout mutants of V. coralliilyticus affecting protease production can display a minimal effect on isolated cultures of coral zooxanthellae [68]. Genes commonly associated with Type VI Secretion (T6S) are also upregulated in such mutants potentially indicating the involvement of multiple virulence factors in coral pathogenesis [68]. While there are numerous strains of V. coralliilyticus currently categorized as coral pathogens [25], V. coralliilyticus RE22, a primary infectious agent of oyster aquaculture [83,84], is relatively unexplored as a potentially infectious strain to coral despite having minimal differences between their virulence repertoires [30]. V. coralliilyticus RE22 has been shown capable of infecting M. capitata coral fragments but its pathogenicity against other coral species could indicate the ability to produce a broad-spectrum of virulence factors that necessitate further study [87].

Tropical coral bleaching and White Syndrome (WS)
Coral reefs have been significantly declining in health and structural integrity for the past 35 years [3]. While reef-associated tourism is estimated to be valued at 36 billion USD globally as of 2017 [1] and reef-associated fisheries are estimated to be valued at 6.8 billion USD globally [8], both industries are threatened with decline due to the current loss of reef environments. Additionally, there are a number of tropical communities largely reliant on tourism and goods derived from local coral reefs despite the declining population of reef associated species [9,10]. Coral reef environments are host to an immense level of biodiversity concentrated in an area covering less than 1% of the ocean floor, creating a highly vulnerable environment currently threated by increasing oceanic stressors [2]. Much of the biodiversity associated with coral reef environments is in sharp decline with as many as 75% of reef-associated fish species exhibiting as much as a 50% loss in abundance with marine reserves unable to perpetually insure the conservation of threatened species [26]. Primary coral stressors include ocean acidification, global warming, increased microplastic density, other pollutants like human sewage, and bacterial pathogens all of which have contributed to the declining health of coral reefs and have been increasing over time [11][12][13]. As these factors are not isolated in a natural environment, the decline in coral health reflects simultaneous exposure to several of these stress factors [14]. Additionally, stressors such as global warming also impact reef environment recovery due to the decline in larval viability making the issue of maintaining reef health difficult to approach [15]. However, a factor that is somewhat understudied is the direct pathogenic potential of marine microbes as the cause of either primary or secondary infections of coral.
When approaching the issue of coral bleaching, it is important to consider that coral pathogens are numerous and produce a wide array of phenotypically different conditions based on the organism responsible, so specialized solutions will likely be necessary to combat each type of infection [16]. A bacterial infection of Acropora palmata referred to as white pox disease (WPD) is caused by the fecal bacterium Serratia marcescens. This is an example of a highly virulent coral pathogen, able to cause a high rate of tissue loss of approximately 2.5 cm 2 per day [33]. Originally, white pox disease was thought to be highly contagious as neighboring colonies were rapidly infected shortly after the initial infection; however, emerging research has demonstrated that spatial relationships in situ were not essential to disease progression and, instead, innate genomic susceptibility was a greater factor [34]. The emergence of WPD is attributed solely to the exposure of A. palmata colonies to high levels of human sewage containing the unique strain of S. marcescens, PDR60, as the strain commonly found in waste was isolated from infected colonies near offshore septic systems [88]. While not bacterial, fungal aspergillosis of soft gorgonian corals caused by Aspergillus sydowii is also a possible source of coral infection presenting another potential microbial threat in addition to those already affecting scleractinian coral [27].
Black band disease (BBD) is also a common coral disease primarily effecting scleractinian coral within the genera Montastrea, Colpophyllia, and Diploria which are typically categorized as boulder/brain coral [28]. Primarily caused by mixed populations of cyanobacteria, sulfide-oxidizing bacteria, and sulfate-reducing bacteria, the tissue decay attributed to BBD is due to a sulfide-rich environment creating hypoxic conditions as well as the black coloration [29]. Additionally, BBD associated bacteria were found to be unrelated to any terrestrial bacteria indicating that tourism is not a likely a contributing stressor [28]. Yellow band disease (YBD) is a coral disease associated with high levels of Vibrio spp. colonizing pale yellow tissue lesions [39].
YBD infected coral samples often display a marked decrease in chlorophyll a and c2 content likely indicating targeted degradation of the intracellular zooxanthellae causing coral tissue death through starvation [39]. Vibrio spp., primarily V. natriegens and V. parahaemolyticus are also the cause of Porites ulcerative white spot syndrome (PUWS) in Porites cylindrica. This coral species seems much more sensitive to infection than other coral species, as inocula of 1 × 10 4 CFU/mL were sufficient to achieve bleaching and tissue lysis under laboratory conditions [54].
A large number of coral bleaching diseases have been found to be caused by various bacteria within the genus Vibrio including V. alginolyticus, V. shiloi, and V. coralliilyticus among others [17]. Many of these coral pathogens are responsible for a class of diseases referred to as white syndrome (WS) or white band disease (WBD) due to the gradual outward progression of a band of bleached tissue to appear on affected coral in contrast to the black band of BBD [18]. White syndrome has been known to affect a variety of tropical corals but it is particularly devastating to reef environments composed of plate corals of the Acropora spp. with prevalence among colonies as high as 50% in some regions [18].  [20], a number of identified coral pathogens also increase in number and virulence at heightened temperatures [21]. Additionally, it was found that increased seawater temperatures were capable of increasing innate Vibrio populations associated with samples of P. damicornis potentially indicating an increase in ocean temperatures resulted in both an increase in coral stress as well as an increase in potential for opportunistic infections [22]. An increase upwards of 4 orders of magnitude in V.
coralliilyticus populations specifically associated with heat-treated samples of P.
This indicates that, though WS is not caused exclusively by Vibrio species [31], a primary cause of coral disease is V. coralliilyticus necessitating further study on its interactions with its coral host. and 48 h after inoculation (C). The spacing of the plastic supports of the grids under the coral fragments is 1 cm by 1 cm [32].
Understanding the pathogens involved in coral disease is essential due to the limitations associated with management measures currently being used to protect coral environments. While some coral infections may be overcome due to selection pressures against genetic susceptibility such as white pox disease in A. palmata [34], more direct measures may be necessary to combat WS. Currently, antibiotics such as amoxicillin are being tested as potential protective measures against coral bleaching in M. cavernosa colonies affected by SCTLD outbreaks [35]. While lesions are reduced by 95% in treated colonies, the introduction of antibiotics likely alters the composition of the coral microbiome, potentially weakening it and allowing for future re-infections [35]. Additionally, the introduction of antibiotics into the water will promote further antibiotic resistance making future infections harder to combat and this antibiotic solution largely temporary. Coral mucus-associated Vibrio species have been shown capable of rapid genetic exchange of antibiotic resistance genes increasing the spread of antibiotic resistance among potentially pathogenic populations [36]. Ocean environments serve as a collection of mobile antibiotic resistance genes (ARG) that has been steadily growing with overuse of antibiotics making it imperative to carefully consider the use of antibiotics for coral treatment to prevent further expansion of the oceanic resistome [49,50]. Alternative solutions to managing the spread of coral pathogens include treating affected coral with potential probiotic organisms, but that requires knowledge concerning the effectiveness of the targeted probionts against infectious species. Putative novel probiotic organisms isolated from P. damicornis colonies and the surrounding water column have recently been tested to determine their effectiveness at limiting coral bleaching from both pathogenic and thermal stressors [37]. In experiments testing the protective effects of several innate Pseudoalteromonas spp. against V. coralliilyticus in P. damicornis it was found that probiotic organisms successfully stymie the progression of coral bleaching at 30˚C, reducing the decline in photochemical efficiency (Fv/Fm) [37]. While probiotic solutions to Vibrio-induced pathogenesis have been widely explored in other systems [38], use of probiotics for coral protection and disease prevention is still largely unexplored. Introduction of new probiotic organisms to specifically target bacterial coral pathogens could potentially be a long-term solution to the problem of coral bleaching. However, extensive information about the pathogen itself as well as its interactions with a prospective probiont are necessary to best make use of a probiotic option due to the extant extensive anti-microbial and antibiotic resistance of coral associated V. coralliilyticus [42].

Vibrio coralliilyticus
Among potential coral pathogens, Vibrio coralliilyticus is among the most commonly identified during isolation of bacteria associated with infected coral tissue [30,40]. V. coralliilyticus is a gram-negative marine Gammaproteobacteria and a member of the genus Vibrio within the family Vibrionaceae [41]. Characteristics that are typically associated with Vibrio spp. include a high level of flagellar motility, a curved-rod like shape, and the ability to be facultatively anaerobic [43]. The curved Vibrio shape typically determined by the CrvA protein is primarily identified as an adaptation allowing for bacterial tunneling through mucus with minimal resistance [44]. Vibrio spp. are near ubiquitous in the ocean with the distribution area of infectious species enlarging with increased ocean temperatures [45]. Additionally, they can be found in a wide variety of habits and environments with some isolates exhibiting extremophilic characteristics and some in association with hosts as either pathogens or members of the core microbiome [46]. Because a large number of Vibrio harveyi) [53], and humans (e.g. V. alginolyticus and V. parahaemolyticus) [46]. Most pathogenic Vibrio spp. are capable of infecting a wide variety of hosts thanks to their immense array of virulence factors that function against both prokaryotic and eukaryotic prey organisms [56].
Vibrio coralliilyticus displays a number of identified virulence factors consistent with those found among other pathogenic Vibrio spp. [57]. Multiple secretion systems have been shown to be characteristic of infectious Vibrio spp. and are often present in multiple copies [58,59]. Virulence factors associated with type I secretion (T1S) include enterotoxins, cytotoxins such as the multifunctional-   As most of the virulence factors associated with coral pathogenesis are regulated by quorum sensing, a sufficient cell density may need to be reached before the progression of coral disease such as WS [53]. While quorum sensing associated transcripts are up-regulated in the presence of coral mucus, reliance on a significant cell density would potentially make V. coralliilyticus more effective as an opportunist than a primary pathogen due to its ability to rapidly populate a new niche [81,30].  isopods, and human-associated pathogens [30].
There are multiple currently characterized strains of V. coralliilyticus that are known as aggressive coral pathogens. Strain BAA-450, also known as strain YB1, is primarily identified as a pathogen of P. damicornis causing a rapid progression of bleaching and tissue lysis with symptoms fully apparent in as little as ten days but largely avirulent at temperatures under 22˚C [21]. The primary strain responsible for M. capitata WS is identified as OCN008, which causes bleaching of coral fragments exposed    Overall, A. poculata has been used to model the response of a temperate facultative coral to a variety of potential contributors to tropical coral stress and bleaching such as extreme temperature stress, microplastic consumption, and ocean acidification.
Currently, infection against A. poculata has not been observed in marine environments potentially due to the non-reef forming nature of the coral as well as the greater skeletal surface area, which could obscure disease with algal fouling. However, as E.
pallida, another facultative Anthozoan, and multiple tropical scleractinian corals have been successfully infected with V. coralliilyticus [30,32,102], it is likely that A.
poculata will be able to serve as a model for coral stress from Vibrio bacterial infection as well.

Goals of this study
The overall goal of this study was to examine the bacterial pathogenic potential of V. coralliilyticus RE22 against temperate coral species A. poculata. An emerging model organism, A. poculata could be beneficial in modeling coral disease progression typically associated with bleaching diseases such as White Syndromes (WS) [18] or Stony coral tissue loss disease (SCTLD) [30] while also minimizing environmental impact due to being a temperature non-reef forming scleractinian coral [47].
Additionally, while the role of V. coralliilyticus in tropical coral pathogenesis is partially understood, the role of many of its individual virulence factors remains unexplored necessitating further research to combat the spreading pathogen.
The first aim of this study was to establish a protocol for bacterial challenge of The third aim of this study was to characterize the involvement of the V.
coralliilyticus T6SS in coral and dinoflagellate pathogenesis. Knockout mutants of several essential components of the T6SS were constructed and confirmed using PCR and protease testing to confirm expected phenotype. Mutants were then tested in both coral and endosymbiont challenge systems to examine any attenuation in virulence associated with removed or impeded T6S functionality. Focus was on the components of the T6SS essential for translocation of effectors and pathogenic potential rather than assembly and anchoring to the inner membrane. suggests symbiotic A. poculata is more effective than aposymbiotic colonies at mitigating of environmental stress. The data are consistent with bacterial challenges in an oyster larval system, which indicate that T6SS1 is primarily involved in eukaryotic antagonism.

Importance:
The rapid decline of coral reefs is a grave threat to ocean biodiversity as well as tourism and the fisheries industry. Reef-forming coral populations have declined by at least 50% across species since 1995 but this decline is especially pronounced in branching and table corals. Understanding the primary pathogens and mechanisms of virulence involved in bacterial antagonism of coral is essential to design potential protective management tools such as microbial colonization of coral by probionts. This study additionally aims to establish A. poculata as a model organism for testing bacteria-induced coral bleaching and pathogenesis in a temperate host. Additional benefits to the A. poculata coral system include its widespread availability, which enables use and collection without further disrupting the declining tropical coral ecosystem and that it is facultatively symbiotic with B. psygmophilum.

INTRODUCTION:
The collapse of reef environments due to coral tissue bleaching and lysis has profound economic impacts on tourism and fishery industries while also contributing to a massive decline in biodiversity with the loss of reef associated species [1][2]. A conservative estimate of reef associated species would be 2,594,000 unique organisms [3] with Caribbean reef biodiversity alone accounting for ~8-9% of potential reef species. Coral populations have declined by approximately 50% since 1995 due to a wide array of factors including ocean acidification and temperature increase [15,16], but another primary cause includes bacterial pathogenesis. One of the primary bacterial threats to reef health is a class of coral diseases called White Syndromes (WS) [17]. with Vibrio coralliilyticus [6,7]. Moreover, investigations into causative agents of a tropical coral bleaching disease known as "white syndrome" revealed that coral species including Pocillopora damicornis, Montipora capitata, and Acropora cytherea were increasingly affected by various strains of V. coralliilyticus [9][10][11].
The progression of coral bleaching in infected corals occurs rapidly causing noticeable tissue lysis and bleaching in tropical corals in as little as 5-10 days [9].
Proposed blanket treatments of diseased tropical coral such as antibiotics use does not necessarily prevent reinfection [12] as it can disrupt the innate coral microbiome and create greater opportunities for potentially opportunistic pathogens like V.
coralliilyticus [14]. As a result, it has become imperative to increase our understanding of coral pathogenesis to develop long-term management measures.
The strain used in this study, V. coralliilyticus RE22, is a broad-spectrum pathogen commonly associated with vibriosis in oyster species Crassostrea gigas and C. virginica [18][19][20] but has also been shown to cause bleaching in tropical coral M.
capitata [21]. Vibrio coralliilyticus RE22Sm is a Gram-negative motile marine bacterium with a wide array of virulence factors including a Type I Secretion System (T1SS), a Type II Secretion System (T2SS), a Type III Secretion System (T3SS), two Type VI Secretion Systems (T6SS) and several extracellular zinc-metalloproteases [7,20]. Little is known about the direct mechanistic effect of V. coralliilyticus strains on tropical corals and whether or not pathogenesis is toxin-mediated, contact-mediated, or both. It has previously been hypothesized that the zinc-metalloprotease produced by V. coralliilyticus is the most likely causative agent of endosymbiont death in tropical coral [6]; however, it has been shown that a knockout mutation of the primary protease gene, vcpA, has no significant effect on virulence against a clade C1 Symbiodinium culture isolated from Acropora tenuis [13]. Additionally, loss of vcpA leads to the upregulation of other virulence factors including components of the T6SS [13]. This is a strong indicator of the involvement of additional virulence factors like type 6 secretion (T6S) in the progression of disease in A. tenuis.
The T6SS of V. coralliilyticus resembles an inverted T4 bacteriophage-like nanomachine and is primarily purposed for the cell contact mediated translocation of effector molecules [22,23] but may have a secondary role in quorum sensing [24].
The system is composed of several distinct structures assembled from thirteen conserved proteins including a baseplate complex anchored to the inner membrane, a hollow needle-like structure composed of hexomeric hemolysin co-regulated protein (Hcp), a VipA/B contractile sheath surrounding the Hcp barrel, and a valine-glycine repeat protein (VgrG) which has a hardened proline-alanine-alanine-arginine (PAAR) repeat designed to puncture eukaryotic or prokaryotic prey cell membranes [25].
Effectors either decorate the PAAR motif and are released upon activation of the mechanism or are translocated via the Hcp needle structure [26,27]. The V.
coralliilyticus RE22 T6SS on chromosome 1 (T6SS1) has been suggested to have divergent but overlapping function with the T6SS on chromosome 2 (T6SS2). T6SS1 has been shown to be vital for pathogenesis in eukaryotic oyster models compared to the primarily anti-bacterial activity associated with T6SS2 [20]. In this report, we focus on the involvement of the T6SS puncturing device in antagonism against a potential coral prey organism and its endosymbiont.
The coral species used in this study is the emerging model organism temperate coral Astrangia poculata. This coral has an ability to tolerate a wide range of temperatures and a facultative relationship with its dinoflagellate endosymbiont, Breviolum psygmophilum [28][29][30]. Differences in A. poculata stress responses between symbiont states are already well characterized with densely colonized (symbiotic) fragments recovering from stress events and wounding at a greater rate than sparsely colonized (aposymbiotic) fragments [31]. The symbiont state in A. poculata also has an impact on innate immune gene expression. Suppression of genes associated with the innate immune response has been observed in symbiotic samples when compared to aposymbiotic samples [32]. Additionally, the composition of the A. poculata mucosal microbiome has been shown to vary negligibly between coral fragments with differing symbiont states suggesting greater consistency in comparative experimental trials [33]. Establishing A. poculata as a model for tropical coral pathogenesis can minimize the impact of research on the tenuous state of reef environments while also providing opportunities to study host-symbiont interactions and their role in pathogenesis. The distinct differences between colonies will allow investigation into the impacts of bacterial challenge on coral health independent of the symbiont state, further elucidating the role of the endosymbiont in the progression of Vibrio associated coral disease.

V. coralliilyticus causes differing rates of tissue lysis in A. poculata by symbiont state
Killing potential of V. coralliilyticus RE22Sm against A. poculata was first examined by exposing isolated symbiotic coral samples to V. coralliilyticus at cell densities ranging from 1 × 10 5 to 1 × 10 8 CFU/mL. Coral fragments exposed to cell densities below 1 × 10 7 CFU/mL displayed 100% survival identical to no treatment control conditions (Fig. 1c-d). At cell densities of 1 x 10 7 (Fig. 1e) and 1 × 10 8 CFU/mL (Fig.   1f) coral fragment survival was reduced to 80% and 10% respectively. This would indicate an effective LD50 of approximately 5 × 10 7 CFU/mL for V. coralliilyticus RE22 against symbiotic A. poculata. In diseased coral fragments, tissue lysis was observed within 5-10 days of primary exposure to V. coralliilyticus RE22Sm with concurrent bleaching rather than bleaching before tissue decay (Fig. 1b). Progression of ill-health in coral began with a decrease in polyp activity coupled with an overproduction of coral mucus which became clouded in the water column. V.
coralliilyticus RE22Sm was present in high densities within clouded mucus samples.
Within 2-4 days of preliminary pathogenesis, coral tissue began to develop a pale color and dissociate from the skeleton upon disturbance of the water column. This indicated that coral bleaching and tissue lysis of the polyps were happening concurrently rather than sequentially. No-treatment control coral samples (Fig. 1a) exhibited high activity levels including polyp extension and responsiveness to feeding, consistently scoring 4-6 on the activity scale, that remained stable over the 20 day experiment. In contrast, A. poculata samples exposed to the pathogen became inactive within 5 days and largely remained inactive for the duration of the experiment regardless of their survival.
In experiments comparing the responses of aposymbiotic and densely symbiotic coral fragments to infection with a high dosage (1×10 8 CFU/ml) of V. coralliilyticus, a difference in the rate of death was observed. Within the first 5 days of the experiment 79.4% of all symbiotic coral samples were still alive, albeit mostly quiescent, compared to a survival of 37.8% in treated aposymbiotic samples (Fig. 2).
This difference was found to be significant (P < 0.02) only at the 5-day measure as survival of densely symbiotic fragments dropped by day 10 to 47.6%, which was not significantly different from the aposymbiotic survival rate of 31.1%. This trend continued for the remainder of the experiment with both aposymbiotic and symbiotic samples dying at comparable rates after day 10. All measures for both aposymbiotic and densely symbiotic samples were significantly different from the negative control after 0 days.

T6SS mutations attenuate virulence of V. coralliilyticus against coral host A. poculata
To assess the contributions of the T6SSs towards the coral pathogenicity of V.
coralliilyticus RE22Sm, coral challenge assays were performed using RE22Sm bacterial mutants deficient in major structural components of the T6SS. Coral treated with wild-type RE22Sm declined in health over 20 days until only 15.6% of fragments survived, which was significantly different (P < 0.001) from the 94.9% survival observed in the no treatment control (Fig. S1). Increased survival was observed in coral fragments treated with bacterial mutants of the hcp1 and hcp2 T6S components.
Survival increased significantly (P < 0.05) from 15.6% in the wild-type treated samples to 66.7% in the coral samples exposed to the Δhcp1 strain (Fig. 3e).
Bleaching but not tissue lysis was observed in 2 samples treated with Δhcp1 within 10 days, which were counted among surviving fragments (Fig. 3b). Bleached coral fragments were only observed in Δhcp1 treated samples. While virulence against coral was moderately attenuated in the Δhcp2 strain with survival increasing to 40%, the difference was not found to be significant to wild type (Fig. 3g). Treatment of coral samples with revertant strains of either the Δhcp1 and Δhcp2 mutant resulted in restored levels of virulence and were not significantly different from the wild-type treated positive control samples (Fig. 3f, 3h). Double mutant strains of T6SS genes hcp1 and hcp2 were also used to test whether or not activity from both T6SS1 and T6SS2 contributed towards an additive effect in regards to virulence. Surprisingly, inactivation of both hcp1 and hcp2 in double mutant strain RE22 Δhcp2 pDM5::hcp1 resulted in increased virulence with reduced survival of coral fragments treated with this strain compared to either of the single mutant strains with a 0% survival observed after 20 days (Fig. 3j). A double mutant strain for vgrG1 and vgrG2 was also used and, comparably, survival of treated coral samples reached only 20% (Fig. 3i). Inactivity of coral polyps was observed consistently in all samples treated with RE22 regardless of strain (Fig. 3d). Treated polyps were deeply retracted into the coral skeleton with minimal coenosarc visible and lacked response to stimuli.

B. psygmophilum is impacted by exposure to V. coralliilyticus RE22
The experiments described above showed coral bleaching prior to coral death, RE22Sm was 59.3% so the dosage was kept consistent for subsequent experiments as an approximate LD50. V. coralliilyticus RE22Sm was capable of inducing a reduction in membrane stability and discoloration in treated B. psygmophilum (Fig. 5).
Additionally the presence of regions of high auto-fluorescence, hypothesized to be chloroplasts, decreased in quantity and fluorescent intensity by 96 hours after exposure to RE22Sm. The hcp and vgrG double mutants were significantly different from both the no treatment control and the RE22Sm treated samples (P < 0.05).
Chlorophyll was also extracted and quantified for each sample and time point, normalized by initial B. psygmophilum cell density for each condition and represented as pg/cell (Fig. 6).

DISCUSSION:
The broad-spectrum antagonistic activity of V. coralliilyticus makes it an effective primary and opportunistic pathogen against a large variety of scleractinian coral species [9,10]. While current research has described potential virulence mechanisms active against tropical coral species through disruption of the standard coraldinoflagellate symbiosis [13,46], the involvement of T6S is largely unexplored. Prior work has demonstrated that exposure to the zinc-metalloproteases produced by V.
coralliilyticus has a pronounced effect on the survival of the endosymbiont within coral tissues [46]; however, direct exposure assays to the supernatant of a VcpA deficient mutant V. coralliilyticus produced negligible differences in photoinactivation [13]. With the data, we demonstrate that V. coralliilyticus is capable of infecting A. poculata at high doses and that aposymbiotic samples are more susceptible to infection than symbiotic samples. Additionally, we found that T6S is involved in pathogenesis against both A. poculata and its endosymbiont B.
psygmophilum, with the V. coralliilyticus RE22 T6SS1 having greater involvement in pathogenesis than T6SS2. Our results have significant implications by: 1) expanding potential host systems for modeling coral bacterial infection to include the nonendangered, facultative, temperate coral A. poculata and 2) characterizing the impact of a largely unexplored virulence mechanism in coral-endosymbiont pathogenesis.
The data presented provide evidence that A. poculata and its endosymbiont are both vulnerable to the pathogenic activity of marine bacteria. Dosage sensitivity ranging from 1 × 10 7 and 1 × 10 8 CFU/mL (Fig. 1e) is consistent with cell densities used in tropical coral challenges across the literature [9,40]. However, the cell density necessary for RE22Sm to exhibit vibriosis in oyster larval models is 1 × 10 4 CFU/mL, 10 -4 -fold of the infectious dose in coral systems [38]. A proposed explanation is that despite the involvement of several virulence factors in cross-species pathogenesis, infections in coral systems require a greater virulence repertoire than infections in the oyster system [39]. However, temperature and cell density also influence the production of virulence factors associated with both pathogenesis in larval C. gigas and in coral. The role of temperature in production of virulence factors by Vibrio spp.
was examined in Ben-Haim et al. [9] and suggested that the extracellular zincmetalloprotease production was greatly influenced by growth temperature of the organism. V. coralliilyticus also relies on quorum-sensing for the regulation of a large number of its virulence factors, and Kimes et al. [41] demonstrate that quorum-sensing can also be regulated by growth temperature. Since our experimental temperatures were set to 25-26˚C, slightly below the optimal growth temperature of 27˚C for the bacterial pathogen, a higher initial quorum may be necessary to establish virulence; however, the effect on growth is likely minimal as V. coralliilyticus is still pathogenic in the range of 24 -28˚C [9]. As virulence in V. coralliilyticus is tied to cell density due to quorum sensing mechanisms up-regulating several primary virulence factors, a higher dosage would be necessary to achieve optimal infectious potential. An additional factor that could contribute to the difference in susceptibility between adult A. poculata and larval C. virginica is the maturity of the host systems as oyster larvae have an immature immune response compared to adult oyster, which are resistant to infection, or coral [59].
The observed response to bacterial challenge in the A. poculata host system was also seemingly dependent not only on dosage of the pathogen but also the density of the endosymbiont population associated with the coral tissue. Coral fragments that exhibited an initially dense B. psygmophilum population were more resistant to infection than aposymbiotic samples in the beginning stages of coral infection. This is counterintuitive considering the observed immunomodulation used to selectively suppress genes of the innate immune system to accommodate endosymbiont in symbiotic samples compared to aposymbiotic samples [32]. Additionally, not only does B. psygmophilum have a greater sensitivity to temperature than its coral host [32], but also the endosymbiont population is thought to be the primary target of anti-coral Vibrio pathogenesis [42]. One potential explanation is the increase in metabolite availability in densely symbiotic coral that allows them to recover from initial bacterial antagonism. A. poculata with high endosymbiont density have been shown to recover from wounding at a much greater rate than aposymbiotic samples [31].
Considering significantly greater photochemical efficiency was observed in symbiotic corals compared to aposymbiotic corals [43] and that consistent consumption of heterotrophic food sources can increase the rate of photosynthesis in tropical coral Stylopora pistilla [44], it is likely that symbiotic A. poculata fragment have greater energy availability than aposymbiotic fragments. While suppression of the innate indicating a greater defensive response to infection but potentially a greater vulnerability to pathogen induced cell death as well [45]. Interactions between A.
poculata and its endosymbiont may negatively affect survival during prolonged infection due to suppression of the immune system; however, our data suggest that during primary exposure to pathogenic bacteria, aposymbiotic samples are more likely to exhibit signs of infection.
Our data also indicate that the two T6SS present in the RE22Sm genome contribute heavily to eukaryotic antagonism in the coral model system. In both A. poculata (Fig. 3h)  psygmophilum, increased survival of the corals and endosymbionts was still observed compared to the samples treated with RE22Sm. Some overlapping but still specialized functionality is a feature of other organisms with multiple T6SSs such as P.
aeruginosa, which uses the sigma factor RpoN (σ 54 ) to induce activity of one T6SS while suppressing the other [49]. It was surprising that both double mutant strains for the Hcp (hcp1 hcp2) and VgrG (vgrG1 vgrG2) components of the two T6SSs were more virulent in both coral and endosymbiont systems when compared to the tested single mutants. This could be due to some compensatory up-regulation of alternative virulence factors [13] or potential alternative interactions of the secretion apparatus resulting in an altered virulence profile when two or more genes are rendered nonfunctional [50]. Additionally, coral fragments challenged with RE22 displayed an inactive quiescence phenotype consistent with the phenotype described in A. poculata samples immersed in cold-water conditions of 6-8˚C which is typical of a general stress response [47]. This could be due to increased stress from bacterial infection causing a change in activity level as has been previously observed in tropical corals affected by SCTLD [12]. V. coralliilyticus is highly motile and attracted to coral mucus [55] so a retraction of the polyps is a well characterized defense to minimize production of chemo-attractants and intake of pathogenic bacterial from the water column [54].
It is not known whether or not Vibrio-induced coral pathogenesis is contact mediated or toxin mediated, but our data suggest that the T6SS, a virulence factor that relies on direct contact to translocate toxins, is a primary component in coral pathogenesis. It is known that V. coralliilyticus is capable of invading tropical coral tissue providing the opportunity for cell-to-cell contact and subsequent activation of contact mediated virulence factors [9]. Work in other eukaryotic models reveal the ability for components and effectors of the T6SS to induce apoptosis or autophagy in prey cells [51,52], but as this work was done primarily in mammalian models its applications to marine pathogenesis may be limited. Additionally, little is known about the effectors translocated by T6S in V. coralliilyticus. While the up-regulation of apoptotic mechanisms in response to V. coralliilyticus infection has been observed in tropical corals [40] and facultative anthozoans [45], it is unknown if RE22 is capable of exploiting or regulating apoptosis to exacerbate infection. An alternative could be that T6S plays a broad role in pathogenesis due to its wide array of secreted effector proteins rather than a single specialized function. By increasing invasion via adhesion, increasing intracellular viability through innate immune regulation, inducing disruptions to the actin cytoskeleton, the anti-eukaryotic effectors associated with T6S [26] could induce a stress response of sufficient magnitude to lead to intact coral cell expulsion that occurs under temperature stress [56]. Among the many anti-eukaryotic effectors translocated through T6S are those that allow for the evasion of eukaryote innate immune response potentially making the difference between symbiotic and aposymbiotic regulation of immunity negligible [48]. V. shiloi adhesion and penetration of tropical coral tissue has been previously characterized [53], but V.
coralliilyticus adhesion and accumulation around coral tissue occurs minimally and largely around the polyp pharynx rather than the coenosarc suggesting intake through the coral gastrovascular cavity [54]. However, it is still possible that V. coralliilyticus is still able to penetrate coral tissue and become intracellular potentially providing the opportunity for the pathogen to disrupt the integrity of the symbiosome.
B. psygmophilum survival was reduced due to exposure to V. coralliilyticus, but pathogenic activity against B. psygmophilum was attenuated in all of the V.
coralliilyticus T6SS mutants. Despite T6SS mutants producing a measurable increase in endosymbiont survival, the involvement of the RE22Sm zinc-metalloproteases is unexplored in our system and would need to be tested in further research to assess which virulence factor has a more pronounced impact on survival. Both cell density and chlorophyll a content declined after bacterial challenge; however, chlorophyll a declined at a slower rate than cell density suggesting that chlorophyll a is potentially a delayed measure of cell survival due to natural degradation [57]. The pigment itself would be unaffected by pathogen exposure and, therefore, more established measures of endosymbiont health such as photochemical efficiency should be examined to support the data presented.  Merodiploid mutagenesis: Construction of allelic exchange mutants adheres to protocols previously described by Schuttert et al. [20]. Briefly, this study utilized a pDM4 plasmid modified with a kanamycin resistance gene (Km r ), pDM5, and linearized at a SacI restriction site within the multicloning region and constructed with overlapping 5' and 3' target gene fragments using the Gibson Assembly Reaction [34].

Competent E. coli cells were transformed via electroporation with the BioRad Gene
Pulser II in a 2 mm cuvette (2.5 kV; 25 µF; 200 Ω) after addition of ligation mixture.
The plasmid was then conjugated from E. coli into V. coralliilyticus RE22Sm [35].
Transconjugates were selected for on mYP30Sm 200 Cm 5 due to the chloramphenicol resistance conveyed by pDM5.
Coral husbandry and tank conditions: Samples of A. poculata were obtained from Fort Wetherill State Park (Jamestown, RI, USA) via manual fragmentation and transported in sea water before being placed in an aerated holding tank of 3% ASW at temperatures ranging from 13 -20˚C to mimic local seasonal water temperatures. A 50% water change was performed every 2 weeks to ensure continued coral health.
Fragments were further divided into fragments no larger than 4 cm 2 in size and allowed to recover for ten days before water temperatures were gradually increased by 0.5˚C per day up to 25 -26˚C to acclimate samples to experimental conditions. Salinity and pH were checked every five days to confirm optimal water conditions.
Fragments were fed with homogenized frozen brine shrimp every five days. P values of < 0.05 were considered to be statistically significant.    These data represent at least 10 replicates for each treatment.     A. poculata fragments exposed to different bacterial mutants of V. coralliilyticus RE22.
E) the dashed line with open triangles represents fragments exposed to mutant strain Δhcp1; F) the dashed line with closed diamonds represents fragments exposed to the in-cis revertant strain for Δhcp1; G) the solid line with open diamonds represents fragments exposed to mutant strain Δhcp2; H) the dashed line with closed circles represents fragments exposed to the in-cis revertant strain for Δhcp2; I) the solid line with closed triangles represents fragments exposed to double mutant strain ΔvgrG1 These data represent at least 10 replicates for each treatment.