Coral Disease Diagnostics: What's between a Plague and a Band?

Recently, reports of coral disease have increased significantlyacross the world's tropical oceans. Despite increasing effortsto understand the changing incidence of coral disease, veryfew primary pathogens have been identified, and most studiesremain dependent on the external appearance of corals for diagnosis.Given this situation, our current understanding of coral diseaseand the progression and underlying causes thereof is very limited.In the present study, we use structural and microbial studiesto differentiate different forms of black band disease: atypicalblack band disease and typical black band disease. Atypicalblack band diseased corals were infected with the black banddisease microbial consortium yet did not show any of the typicalexternal signs of black band disease based on macroscopic observations.In previous studies, these examples, here referred to as atypicalblack band disease, would have not been correctly diagnosed.We also differentiate white syndrome from white diseases onthe basis of tissue structure and the presence/absence of microbialassociates. White diseases are those with dense bacterial communitiesassociated with lesions of symbiont loss and/or extensive necrosisof tissues, while white syndromes are characteristically bacteriumfree, with evidence for extensive programmed cell death/apoptosisassociated with the lesion and the adjacent tissues. The pathologyof coral disease as a whole requires further investigation.This study emphasizes the importance of going beyond the externalmacroscopic signs of coral disease for accurate disease diagnosis.

INTRODUCTION

Top

Introduction

Coral disease is considered an important factor in the recentdecline of coral reefs worldwide (29, 30, 38, 64). Reports ofdisease and disease-like syndromes in reef-building corals haveincreased substantially since first being reported in 1973 (10).This increase in the incidence of disease is due in part toa better awareness of coral health but is also linked to theincreased environmental stresses affecting coral reefs (30,31, 35, 44). Between 18 and 30 diverse coral diseases and syndromesare described worldwide (29, 64, 65) on the basis of macroscopicfeatures. Coral disease diagnosis is primarily macroscopic,taking into account characteristics such as the extent of tissueloss, tissue color, and exposure of coral skeleton. These macroscopicdisease signs then become the basis for nomenclature and diagnosis.While these characteristics allow broad descriptions of changeon reefs, they are unreliable for accurate disease diagnosisand do not increase our understanding of the causes of coraldisease and disease progression.

The reefs of Eilat have endured several decades of high levelsof anthropogenic impact (38, 39, 66). As a result, coral diseaseand coral mortality have increased (39, 67), resulting in decreasesin coral abundance and diversity over the past 3 decades. Increasingstress from global warming and ocean acidification has alsobeen associated with the overall declining health of Red Seacorals (40). Al-Moghrabi (5) has reported outbreaks of blackband disease (BBD) in the northern Red Sea, and Antonius andRiegl (13) have reported white syndrome on the reefs of theSinai Peninsula, with the most abundant reef-building species(Acropora hemprichii) suffering the heaviest losses. Ben-Haimet al. (18) also reported a new disease of Pocillopora damicornisfrom the Gulf of Aqaba. In addition, anecdotal reports of blackband disease became common in the early 1990s in the Eilat sideof the Gulf of Aqaba, although no study was conducted (Y. Loya,personal communication). Recently, Barash et al. (17) reportedthat many of the massive corals of Eilat were affected by aninfectious white plague-like disease. Despite reports and fieldobservations of these diseases and syndromes within the region,relatively little is known about the pathology, cytology, microbialecology, and disease processes of corals in the Gulf of Aqaba.

While there have been numerous ecological and field-based studiesof corals exhibiting disease (5, 10, 11, 49, 50, 51, 54, 64),few primary pathogens have been identified to date (11, 16,18, 24, 31, 32, 34, 36, 48, 53). Also, few studies have examinedthe histopathological and microbial characteristics of diseasedcorals. BBD is considered one of the major diseases impactingcoral reefs worldwide. The first report of BBD was by Antoniusin 1973 (10). Since that time, BBD has been observed to affectcorals worldwide, especially in polluted environments (5, 13).The black band that is typical of this disease is composed ofa mixed microbial mat that is dominated by cyanobacteria andcomprises sulfur-reducing and sulfur-oxidizing bacteria as wellas a number of other microorganisms. The mat overgrows coraltissues, creating a toxic environment, and tissue loss is attributedto the presence of high sulfide levels (0.8 mM) in the tissuesadjacent the black band (22, 54). The distinct macroscopic signsof this disease and the growth pattern of the mat in a top-downmanner across polyps and tissue structures (10, 28, 54) areused as the primary method of the disease identification. BBDhas been described as one of the main diseases in the Caribbeanand Florida Keys, where it has been responsible for reef decline(37). Most active during the summer months, BBD is one of themost widespread and destructive coral diseases due to its highimpact on massive and framework-building corals, with ratesof tissue loss of up to 2 cm per day (28, 54).

White band and white plague diseases have also been describedfor many regions worldwide, including the Red Sea and the Gulfof Aqaba (5, 12, 13, 14, 17, 59). Field-based surveys and visualdescriptions of these diseases have shown that they typicallyhave a lesion of white, recently exposed coral skeleton. Itis believed that the white band diseases have had a major rolein the community structure shift occurring in the Caribbean(15). Reefs off the coast of Florida have experienced increasingoccurrence of white diseases, with patterns of disease spreadsuggesting a highly infectious nature (49). While pathogenshave been identified for some of the white diseases (24, 46,48, 51), others have been described as having no observablemicrobial community (13, 20, 21, 47). The majority of casualfactors underpinning the rise in coral disease have remainedelusive, particularly for many of the white diseases and whitesyndromes (21, 64). Richardson et al. (52) have rightly calledfor an integrated approach to disease diagnosis that incorporatesfield and laboratory studies.

In the present study we demonstrate that integrating observationson microbial diversity characteristics with specific cytologicalobservations may be a useful tool for understanding the diseaseprocess of corals and improving the basis on which disease isdiagnosed. This study also points to some significant problemsassociated with the simplistic use of macroscopic signs as theonly means of identifying coral disease, as observed in thecourse of studying the reef of Eilat, Gulf of Aqaba, in thesummer of 2005.

MATERIALS AND METHODS

Top

Materials and Methods

Sample collection and disease identification.
Corals were surveyed visually for signs of disease in June (earlysummer) 2005 at depths of 1 to 18 m on coral reefs near theMarine Biological Laboratory at Eilat in the Gulf of Aqaba.Massive and branching corals were assessed for the typical macroscopicsigns of disease, including tissue loss and the apparent rapidexposure of coral skeleton, which are indicative of white diseasesand white syndromes, as well as general paling from bleachingand the presence/absence of black bands. All colonies exhibitingsigns of disease were photographed using a Nikon CoolPix 5000(Nikon, Inc.) digital camera inside a Subal (Steyr, Austria)underwater housing prior to sampling. Replicate tissue and skeletonsamples (n = 3) were collected from corals displaying each theobserved diseases and syndromes to investigate microbial populationsand histopathological features of these diseases. Of the massive(favid) coral colonies observed with diseases, six classifiedas having BBD and six classified as having white band diseasewere sampled in both early and late June 2005. For the whitediseases and white syndromes of other massive corals and ofbranching corals, replicate samples of each type (n = 3) werecollected.

Sample preservation and tissue processing.
Coral samples were fixed in 4% (wt/vol) paraformaldehyde insterile phosphate-buffered saline (pH 7.4) (20) for 12 h priorto decalcification with 20% (wt/vol) EDTA (pH 8) (63) and standardprocessing for paraffin embedding. Tissues were processed throughwashes as follows: one wash each of 70% and 80%, two of 95%,and three of 100% ethanol for 40 min each; three xylene washesfor 40 min; and then three paraffin washes under vacuum for40 min each prior to embedding in paraffin. Serial tissue sections(4 µm) were collected onto Superfrost Plus slides (Menzel,Braunschweig, Germany) for use in fluorescence in situ hybridization(FISH) and histopathology.

Fluorescence in situ hybridization.
Oligonucleotides were Cy3 labeled by Thermo Electron CorporationPty Ltd and used in a conventional FISH protocol (6, 7, 8, 9,41, 42, 43, 58). Paraffin sections were dewaxed through three5-min washes in xylene and four 5-min washes in fresh 100% ethanol.The hybridization was conducted in hybridization buffer (0.9M NaCl, 0.01% sodium dodecyl sulfate, 0.01 M Tris-HCl, pH 7.2)for 1.5 to 2 h at 46°C and was followed by a 10- to 20-minwash in prewarmed (48°C) wash buffer (0.08 M NaCl, 0.01%sodium dodecyl sulfate, 0.01 M Tris-HCl, 0.05 M EDTA) (3). AZeiss Meta 510 confocal scanning laser microscope (Zeiss, Germany)and spectral profile imaging via the Zeiss Image Browser softwarewere used to visualize tissue autofluorescence and probe-conferredfluorescence. Coral tissue sections treated with the FISH protocolwithout the application of probe were used for spectral profiling(3). This was repeated on several tissue sections for each coralspecies to determine the variability in fluorescence profiles.FISH probes used within this study include a universal bacterialprobe mix (EUBmix) and specific group probes for ${delta}$ -proteobacteria(GAM42A), Cytophaga-Flavobacterium (CF319), and Vibrio sp. (MV)(Table 1). These probes were selected since the correspondingbacteria represent major bacterial groups within oceanic communitiesand since many Vibrio spp. are pathogenic.

View this table:
[in this window]
[in a new window]

TABLE 1. Bacterial probes used in FISH

Histopathology.
The general tissue condition associated with the lesion andthe adjacent tissues of each of the diseases was ascertainedfollowing staining using Harris's hematoxylin and eosin (withphyloxine B) (HHS32 and HT110-1-32; Sigma-Aldrich Pty Ltd).The extent of mass tissue necrosis (swelling and lysis of cellsand disruption of cell structure) was recorded. In situ labelingof the 3' ends of DNA fragments was used to investigate thepresence and extent of programmed cell death (1, 4, 25, 26)by use of an ApopTag in situ apoptosis detection kit as perthe manufacturer's recommendations (S7101; Chemicon International,Inc.). This procedure has been shown to distinguish apoptosisfrom necrosis by specifically detecting DNA cleavage and chromatincondensation associated with apoptosis and confirmed by thelack of necrotic morphology. Cells were defined as apoptoticif the nuclear areas of the cells were positively labeled asindicated by red staining, as opposed to the blue hematoxylin-counterstainednonapoptotic nuclei (1).

RESULTS

Top

Results

Black band disease.
A simultaneous outbreak of black band disease and white banddisease in massive corals (from the genus Favia) on the reefsof Eilat (Red Sea) was observed during June 2005. Black banddiseased corals showed typical signs of the disease and werecharacterized by a thick black microbial mat that appeared togrow over and into underlying coral tissues (Fig. 1A1 and a1).The typical macroscale signs of white band/plague disease inFavia included an apparently clear lesion border between recentlyexposed bare white skeleton and tissues. The lesion lacked anyobservable black, mixed microbial band on the coral surface(Fig. 1B1 and b1). Newly exposed clean white coral skeletonwas visible at the lesion interface and away from the diseaselesion (Fig. 1b1). Based on macroscale signs, these corals wouldbe classified/diagnosed as suffering a white disease or whiteplague.

View larger version (97K):
[in this window]
[in a new window]

FIG. 1. Typical (panels labeled A and a) and atypical (panels labeled B and b) black band disease of Eilat reef corals in the Red Sea. Shown are underwater photographs of infected black band diseased colonies displaying typical (A1 and a1) and atypical (B1 and b1) signs of disease. Tissue necrosis and microbial penetration and disruption of tissue layers in black band diseased colonies is evident by histopathology, using hematoxylin and eosin staining of tissue sections of typical (A2 and a2) and atypical (B2 and b2) diseased colonies, and FISH of typical (A3 and a3) and atypical (B3 and b3) diseased colonies is also shown. Scale bars, 50 µm. Abbreviations: Po, polyp; Ep, epithelium; Ga, gastroderm; Ne, necrosis; Bac, bacteria. FISH and spectral codes: blue, coral tissue; green, Symbiodinium sp.; red, bacteria. Uppercase panel labels demonstrate lower-magnification images; lowercase panel labels denote higher magnification.

Microbial and histopathological investigation of both BBD andwhite band/white plague corals identified similar distinctivecyanobacterium-dominated microbial mats. In the BBD coral, theblack encircling mat was easily visible on the colony surface.In contrast, for the white disease/plague Favia coral, a similarblack mat was found deep within the polyp structure and underneaththe disease lesion interface. This same type of mat was identifiedin all sampled white band colonies in early and late June 2005(n = 12). Given the similarity in cytological responses andmicrobial communities, we refer to this type of white band asatypical BBD (aBBD). Similar distinct cyanobacterium-dominatedmicrobial communities were observed in both BBD and aBBD coralsand appeared to cause tissue lysis and necrosis (compare Fig.1A2 and a2 with Fig. 1B2 and b2). Imaging of the microbial communitiesidentified cyanobacteria of similar morphotypes dominating themicrobial communities in both the typical (Fig. 1A3 and a3)and atypical (Fig. 1B3 and b3) black band diseased colonies.

View larger version (97K):
[in this window]
[in a new window]

FIG. 2. Diseased Hydnophora sp. (A1) with evident rapid tissue loss and clear lesion borders between the coral tissue and the recently exposed skeleton (a1). Histopathology showed extensive tissue necrosis of cell layers (A2) including epithelium and gastroderm (a2). FISH revealed an extensive microbial population by use of the EUB universal bacterial probe (A3 and a3) associated with tissue necrosis and adjacent tissues. Scale bars, 50 µm. Abbreviations: Ep, epithelium; Ga, gastroderm; Zx, Symbiodinium sp.; Ne, tissue necrosis; Bac, bacteria. FISH and spectral color codes: blue, coral tissue; green, Symbiodinium sp.; red, bacteria. Uppercase panel labels demonstrate lower-magnification images; lowercase panel labels denote higher magnification.

View larger version (40K):
[in this window]
[in a new window]

FIG. 3. FISH identified bacterial communities associated with diseased Hydnophora sp. identified with EUBmix (a) as belonging to ${delta}$ -proteobacterium (b) and filamentous Cytophaga-Flavobacterium (c) groups. Populations of a Vibrio species were also identified (d). Scale bars, 50 µm. Abbreviations: Ep, epithelium; Ga, gastroderm; Zx, Symbiodinium sp.; Ne, tissue necrosis; Bac, bacteria; GAM, ${delta}$ -proteobacteria; CF, Cytophaga-Flavobacterium; VB, Vibrio sp. FISH and spectral color codes: blue, coral tissue; green, Symbiodinium sp.; red, bacteria.

White disease and white syndrome.
Both massive and branching corals on the Eilat reefs appearedto be suffering from white disease, typified by abrupt lesionsand areas of exposed skeleton adjacent to otherwise normallypigmented tissues. White disease was identified for the massivecolonies of Hydnophora sp. (Fig. 2A1 and a1) and Porites sp.as well as for the branching coral Stylophora pistillata (seeFig. 4A1 and a1). White syndrome was identified for coloniesof the plating coral Acropora sp. (see Fig. 5A1 and a1) in theregion.

View larger version (107K):
[in this window]
[in a new window]

FIG. 4. White disease of Stylophora pistillata displaying tissue loss and partial bleaching in patterns starting at the branch base and progressing rapidly up the branches and through the colony (A1 and a1), with 50% mortality of the colony evident (A1). Hematoxylin and eosin staining shows that tissue structures remain intact at the lesion border (A2 and a2) with some regions of structural loss of the gastroderm adjacent to the lesions (aa2), and FISH using the EUB general bacterial probe identified large populations of bacteria associated with tissues adjacent to the lesion border (A3), a mixed bacterial population dominated by ${delta}$ -proteobacterial taxa (a3). Scale bars, 50 µm. Abbreviations: Ep, epithelium; Ga, gastroderm; Zx, Symbiodinium sp.; Ne, tissue necrosis; Bac, bacteria; L, lesion; PB, partial bleaching. FISH and spectral color codes: blue, coral tissue; green, Symbiodinium sp.; red, bacteria. Uppercase panel labels demonstrate lower-magnification images; lowercase panel labels denote higher magnification.

View larger version (86K):
[in this window]
[in a new window]

FIG. 5. White syndrome of Acropora sp. (A1) showing a clear lesion boundary between the healthy tissue and the recently exposed skeleton (a1). Investigation of the tissue structure by use of hematoxylin and eosin shows that the tissues appear intact (A2) with no evidence for mass loss of tissue structure or integrity (a2). There is a distinct lack of microbial communities associated with the disease lesion (A3) and no bacterial penetration of colonization of the tissue layers (a3). Scale bars, 50 µm. Abbreviations: Ep, epithelium; Ga, gastroderm; Zx, Symbiodinium sp.; Ne, tissue necrosis; Bac, bacteria; Agg, bacterial aggregates. FISH and spectral color codes: blue, coral tissue; green, Symbiodinium sp.; red, bacteria. Uppercase panel labels demonstrate lower-magnification images; lowercase panel labels denote higher magnification.

There were distinct cytological and microbial differences betweenthese macroscopically similar syndromes. The white disease ofHydnophora sp. was characterized by extensive tissue breakdownand mass tissue necrosis within the lesion areas (Fig. 2A2 anda2). The lesions were also extensively populated by bacteria,which were found to penetrate all tissue layers and infiltrateadjacent regions of tissue (Fig. 2A3 and a3). Investigationof several tissue sections showed that the bacteria were fromboth ${delta}$ -proteobacterium (Fig. 3b) and Cytophaga-Flavobacterium(Fig. 3c) groups, and communities of Vibrio sp. were also foundwithin the lesion (Fig. 3d).

In contrast, the white disease observed for the branching coralS. pistillata (Fig. 4A1) exhibited patches of bleaching closeto the lesion border and regions of tissue loss (Fig. 4a1).For this white disease or form of white disease, extensive bacterialcommunities were found in tissue layers associated with thelesion (Fig. 3A3 and a3), and a variety of bacterial groupswere evident, with no single bacterial group being dominant.There was a lack of evidence for extensive tissue breakdownand necrosis (Fig. 3A2 and a2), with only small regions of necrosisin the lesion of white disease of S. pistillata; this necrosiswas limited to regions of the gastroderm (Fig.4aa2) and presumablywas associated with macroscopic areas of bleaching.

Disease signs consistent with descriptions of white syndromewere also observed for the plating Acropora sp., with clearlesions evident between apparently healthy tissues and recentlyexposed skeleton. The white syndrome identified on Acroporasp. (Fig. 5A1) had a distinctly clear lesion border that appearedto be moving quickly across affected corals, as determined bythe lack of macroscopic algal overgrowth over the large exposedareas of skeleton. The exposed skeleton at the lesion was freeof any evidence for macroscopic algal overgrowth (Fig. 5a1),and tissues at the lesion border appeared healthy despite diseaseprogression (Fig. 5B1 and b1). Microscopic analysis of the lesionborder showed it to be distinctly devoid of any significantbacterial populations, with tissue adjacent to the lesion lackingany evidence of tissue breakdown or necrosis associated withthe disease (Fig. 5A3 and a3). This provides clear evidenceof distinct microbial community differences and morphologicaldifferences between white syndrome and white diseases (Table2); these differences require further investigation.

View this table:
[in this window]
[in a new window]

TABLE 2. Summary of microbial and disease characteristics of coral diseases in Eilat^a

Evidence of necrosis and apoptosis in disease.
Hematoxylin and eosin staining of tissue sections revealed extensiveevidence for necrosis associated with the tissue loss in favidcorals with black band disease. Tissue breakdown consistentwith necrosis was also evident for white disease in Hydnophorasp. and Porites sp. Comparatively few or no signs of necrosisor mass loss of tissue integrity were evident for the whitedisease of S. pistillata or Acropora sp. In situ end labelingof fragmented DNA, used as a marker of programmed cell death,showed no evident staining in tissues of massive corals witheither typical (Fig. 6a) or atypical (Fig. 6b and c) black banddisease or in any of the tissues associated with the necroticwhite disease of Hydnophora sp. (Fig. 6e to g). However, positivestaining was evident in large regions of tissues associatedwith white syndrome of Acropora sp. within both epithelial (Fig.6h) and gastrodermal (Fig. 6i) tissue layers, while no stainingwas evident in healthy tissue in the same colonies (Fig. 6j).Positive staining was also evident with the tissue layer samplesof white disease of S. pistillata (Fig. 6k and l). However,this staining was limited to cells directly adjacent to thetissue lesions; cells away from the lesion border were not sostained (Fig. 6m). Cells showing morphology associated withnecrotic cell death showed no staining when in situ end labelingfor fragmented DNA was used (Fig. 6a to f). Therefore, celldeath markers consistent with programmed cell death were identifiedonly in the apparently bacterium-free lesion for Acropora sp.white syndrome and in the lesion edge for the white diseaseof S. pistillata and were not associated with BBD or found wheremass tissue necrosis was evident (Table 2), again demonstratingthat white syndrome is distinctly different from white diseases.In this case, the differentiation between white syndrome andwhite diseases is based on the patterns of cell death associatedwith the necrotic diseases (Table 2).

View larger version (115K):
[in this window]
[in a new window]

FIG. 6. In situ end labeling of fragmented DNA evident of apoptotic cell death is not detected for atypical (a) or typical (b and c) black band disease or for white diseases of Hydnophora sp. (d, e, and f). Positive staining was evident at the lesion border for the white disease of Stylophora pistillata (g and h) but not for tissues away from the border (i), and a high density of apoptotic cells was associated with white syndrome of Acropora sp. at the lesion border (k and l) but not away from the lesion border (m).

DISCUSSION

Top

Discussion

While coral disease has been reported from the reefs of thenorthern Red Sea (5, 13, 14, 17, 39, 40, 59), this is the firststudy to explore histopathological changes and in situ microbiologyassociated with these diseases and syndromes. This approachto the study of coral disease has revealed that dependence onvisual macroscopic characteristics alone is unreliable for accuratediagnosis and overlooks critical information on the cellularand microbial processes associated with disease-like states.The incorporation of physiological assays, for example, descriptionsof photosynthetic assimilate translocation (27, 55), which demonstratecolonial activity and integration, and assays addressing hostmechanisms (33) are also useful in understanding the coral colonyresponses to different diseases and may provide important informationfor disease diagnosis. We suggest that studies going beyondmacroscopic disease signs by incorporating cytological, microbial,and physiological assays provide a greater level of detail fortruly understanding the processes of disease in corals and willprovide a better basis upon which to make accurate disease diagnosis(Fig. 7). Accurate diagnosis of coral disease can direct researchand management practices towards addressing and managing thetrue underlying cause of disease on reefs.

View larger version (60K):
[in this window]
[in a new window]

FIG. 7. Profiling the interaction of holobiont changes and multiple levels of disease signs can be used as a basis for identifying specific criteria for disease diagnosis and to improve understanding of divergent disease processes underlying coral disease progression.

Atypical black band disease.
Rosenberg and Ben-Haim (59) and Barash et al. (17) recentlyreported an anomalous temperature spike in the Red Sea and asimultaneous outbreak of black band disease and white plaguein the summer months from 2001 to 2004. Attempts by these authorsto determine causative agents of the apparent white plague experienceddifficulties in pathogen culturing and transmission of a suspectedpathogen. In the present study, histopathological investigationrevealed the presence of cyanobacterium-dominated microbialmats deep within the coral polyp structure of Favia exhibitingcharacteristics of white disease/plague. These microbial populationsand the pattern of coral tissue disruption were consistent withthose observed for typical black band disease, where cyanobacterialfilaments were found penetrating tissues deep within the polypstructure. Based on the fact that the apparently white plague-infestedcorals all contained cyanobacterial mats deep in the tissues,this disease state appears to be an atypical form of black banddisease; in addition, it appears that the black band consortiumcan cause an apparent white disease or that there can be a progressionfrom a white disease into black band disease. We suggest theuse of the term "atypical black band disease" to describe thisdifferent form of black band disease. This conclusion is supportedby the results of Bythell et al. (21), who previously observedan apparent progression from white plague to black band diseasein the Caribbean. The difficulties in pathogen identificationand infection studies experienced by Barash et al. (17) maybe due to the fact that some of these diseased corals were actuallyinfected by atypical BBD, illustrating the confusion that relyingon macroscopic characteristics alone may incur. We also suggestthat the incidence of BBD on reefs may be underestimated ifthis type of atypical form is common. Further studies investigatingBBD should also attempt to determine the extent and impact ofatypical forms of this disease in other regions.

White diseases and white syndrome.
White diseases and white syndrome of corals showed marked differencesevident from both microbiological and cytological investigation;here we define white diseases and white syndrome of corals basedon specific microbial and cytological parameters. White diseasescharacteristically show extensive bacterial infiltration ofthe lesion tissue layers and tissues adjacent to the diseaselesion, as well as extensive necrosis, loss of tissue structure,and/or symbiont loss. White diseases also characteristicallyshowed little or no evidence for programmed cell death associatedwith the disease lesions. The term "white disease" may encompassa range of diseases or states of the disease, as seen in theexample of white disease of S. pistillata in this study. A similarmacroscopic pattern of tissue loss was apparent in the whitedisease of S. pistillata and the white disease of Hydnophorasp., yet no mass loss of tissue structure and only small regionsof necrosis of the tissue layers were apparent in the whitedisease of S. pistillata. However, dense and mixed microbialcommunities were evident within the tissue layers characterizingthe white diseases of both massive and branching colonies; bacteriaof Cytophaga-Flavobacterium and ${delta}$ -proteobacterial groups werepresent in both, as were communities of Vibrio sp., yet no singlebacterial group appeared dominant. The potential of secondarycolonizers involved in white diseases and the role of opportunisticpathogens in white disease progression may be very important,as evidenced by the diverse microbial community of these diseasesand also considering the source of highly diverse microbialenvironments of the coral holobiont (19, 28, 45, 60, 61) andthe coral reef.

In contrast, there was a distinct lack of microbial communityinteraction associated with the disease lesion of white syndrome.The disease lesion of white syndrome is characteristically freeof any evidence of mass tissue breakdown; with no loss of structuralintegrity or necrosis, apparently healthy tissues bordered exposedcoral skeleton at the disease lesion. Extensive in situ endlabeling of fragmented DNA, suggestive of programmed cell deathor apoptosis, was associated not only with the tissue layersof the white syndrome lesion but also with the adjacent tissuesnot directly associated with the disease lesion. This is consistentwith previous observations of white syndrome of tabular Acroporasp. on the Great Barrier Reef (4, 56, 57). Observations of extensiveprogrammed cell death associated with white syndrome and a lackof resource translocation to the white syndrome lesion sitesuggested the disease to be a host reaction progressing independentof an apparent pathogen (4, 56). We suggest that white syndromeof the corals of Eilat may be typical of a similar host reaction,or white syndrome, and that the syndrome is a disease progressiondivergent from that seen for necrotic and bacterial white diseases.We therefore define white syndrome of coral as a disease statethat progresses independent of an apparent pathogen or microbialcolonization, with little or no evidence for mass tissue necrosisand with extensive programmed cell death associated with thelesion. Previous studies of coral disease have also suggestedthat white syndrome is a distinct coral disease apparently linkedto a lack of an observable microbial population associated withthe disease progression (13, 21).

In this study, we differentiate white syndrome from white diseasesbased on specific cytological/morphological and microbial differences.Weil et al. (65) have defined a syndrome as a disease that hasyet to have a causative agent identified. However, we suggestadhering to a medical definition of a syndrome as a collectionof specific signs that occur together to characterize the particulardisease or abnormality. We suggest that microbial, cytological,and physiological characteristics are useful criteria for differentiatingcoral diseases (Fig. 7) and for determining the specific signssuitable for accurate disease diagnosis. This also allows forthe addition of other criteria that may further differentiatethese diseases, as shown in Fig. 7. Furthermore, we postulatethat the mechanisms of tissue loss and disease progression inwhite diseases and white syndrome are divergent and indicatethe importance of understanding the physiological mechanismsthat are associated with these diseases to truly identify theunderlying cause of their increases worldwide.

Primary versus opportunistic pathogens.
Opportunistic pathogens are defined as those which infect compromisedor previously stressed individuals, whereas primary pathogensare those that cause disease in an uncompromised host (62).The involvement of opportunistic as opposed to primary pathogenshas been overlooked in coral disease research and needs to beconsidered when attempting to understand disease and determinedisease causation, disease progression, and colony mortality.This is even more evident when the rapidly changing and highlyimpacted environment of coral reefs is considered. Increasedenvironmental stress and pollution may destabilize the coralholobiont, creating increased potential for possible opportunisticpathogens to affect the stressed corals. The question of differentiatingmicrobial pathogens involved in disease initiation from thoseinvolved in disease progression or sources of secondary infectionsis of great importance. Future work addressing disease causationshould consider the definitions of the two types of pathogens:a primary pathogen is the first infecting pathogen in an uncompromisedhost (62), and a secondary pathogen is one that causes a secondaryor subsequent infection by multiplying within already diseasedtissue but that is not the primary pathogen (2). Finally, anopportunistic pathogen is defined as an organism that is normallycommensal but gives rise to infection in compromised hosts (62).The use of a range of diagnostic techniques will allow us tobetter understand the processes of disease and address questionsof the underlying causes of disease progression in dense anddiverse microbial ecosystems.

CONCLUSIONS

Top

Conclusions

This study has demonstrated the importance of cytological andin situ microbiological studies in the investigation of coraldisease. Comparative analysis of disease states indicates thatthe current use of macroscopic disease signs is insufficientfor characterizing and understanding coral disease. A particularlygood example is that of black band disease within this study.The extent and impact of black band disease on reefs worldwidemay be underestimated if the lack of macroscopic disease signs,as observed for Favia in Eilat, is common. Black band diseaseetiology and epidemiology require further analysis, especiallyto determine if there is the potential for other diseases toprogress into this disease during summer months. Our study demonstratesthat research depending solely on macroscopic signs of diseaseruns the risk of misdiagnosing diseases among corals.

We strongly suggest that studies incorporating cytological,microbiological, and physiological investigations provide criticalinsights into the inception, progression, and causal factorsunderpinning the current global increase in coral disease. Accuratediagnosis of coral disease is vital in providing researchersand managers with a better basis for understanding disease causationon reefs. We therefore conclude that future studies of coraldisease, whether they be field or laboratory based, must examinefundamental cellular characteristics and physiological processesunderlying disease progression.

ACKNOWLEDGMENTS

We thank the Eilat Coral Beach Nature Reserve and David Zakaifor assistance and expertise in sampling and surveying coralsof the Eilat reefs. We thank the Centre for Advanced Light Microscopyat the University of Queensland for assistance with confocalmicroscopy and also Oded Yarden and Bill Leggat for conceptual,logistical, and editorial assistance.

We are grateful for support provided by the GEF Coral Reef TargetedResearch Program (www.gefcoral.org) and the ARC Centre of Excellencefor Coral Reef Studies (www.coralcoe.org.au).

FOOTNOTES

* Corresponding author. Mailing address: Centre for Marine Studies, University of Queensland, Brisbane 4072, Australia. Phone: 61 7 3365 3548. Fax: 61 7 3365 4755. E-mail: t.ainsworth{at}uq.edu.au.

Published ahead of print on 8 December 2006.

REFERENCES

Top