Disease prevalence and bacterial isolates associated with Acropora palmata in the Colombian Caribbean

Study area

Six localities were selected and sampled based on their oceanographic influence in the Colombian Caribbean (Díaz et al., 2000): (i) Uraba gulf, where the reefs are influenced by discharge from the Atrato river and streams from Serrania del Darién, (ii) Isla Fuerte, where the reefs are grouped on shallow terraces that are influenced by the Sinú River, (iii) San Bernardo Islands, where the reefs have developed above the sedimentary bottoms, which are a byproduct of karstic depressions, (iv) Rosario Islands, where the reefs are in a dynamic interaction with the high sedimentary loads coming from el Canal del Dique (Alvarado et al., 1986), (v) Isla Arena, which is a diapiric islet that is under a strong influence from the sedimentary charges of the Magdalena River, and (vi) Tayrona National Natural Park (TNNP), which is characterized by the presence of seasonal upwelling systems; however, in their absence, the coral communities interact with river discharges from the Magdalena river and from the Cienaga Grande de Santa Marta (Díaz et al., 2000) (Fig. 1).

Colony size structure and disease prevalence

During 2017 and 2018 once a year in each locality, A. palmata colonies were measured individually to construct the size class distribution calculating the area considering the maximum diameter (L) and the perpendicular diameter (l), assuming a ellipsoidal geometric shape of the colony followed by the equation: A = (L/2 * l/2) * π (Yap, Aliño & Gomez, 1992). At each colony, disease prevalence assessments were performed by analyzing photographic records and field notes taking from a diver approach. White-band disease was identified by decolored tissue at the base progressing to the tip of the colony, while White-pox was identified by irregular decolored spots on the colony tissue. Disease prevalence was obtained considering the number of colonies with each disease sign in the population divided by the total number of colonies in each locality. A population size structure histogram was constructed using the area data log₁₀ transformed to obtain a good resolution for the small colonies and a normal distribution (Bak & Meesters, 1998). Importantly, except for the Tayrona locality, we could not establish an exact seasonal sampling in the other areas, therefore, the climatic influence was not accurately considered. Additionally, to test the null hypotheses that each disease was associated to the locality a chi-square analysis of contingency table was performed. Furthermore, a multidimensional contingency table was carried to test whether the disease occurrence, the locality and the size-class distribution were mutually independent, due to the results a partial independence contingency table was done to confirm if the combined disease occurrence was related to the size-class (Zar, 2010).

Sample collection

At each location, through a roving approach mucus samples of two healthy colonies, two colonies with apparently WPX and two colonies with apparently WBD within the size class 4.29 cm2 log¹⁰-transformed, were collected from the coral surface using 5-mL sterilized syringes (needleless) (Barash et al., 2005; Gil-Agudelo et al., 2007; Koren & Rosenberg, 2008). Samples were transferred to 15-mL Falcon tubes and stored at 0–4 °C in darkness until laboratory processing. It is important to note that for Uraba gulf, Isla Fuerte and San Bernardo the transportation took more than three days until laboratory processing. The collected material was approved by the national authority “Parques Nacionales Naturales de Colombia with the entry permit 001 of March 6, 2017—PIC-DTCA 001-17”.

Laboratory analysis

A volume of one mL of mucus was enriched in 10 mL of 0.1% peptone water adjusted to pH 9.0 (Donovan & Van Netten, 1995). Subsequently, the samples were incubated at 36 °C for 18 h, and 0.1 mL of the enriched solution was inoculated on glycerol marine agar and thiosulfate citrate bile sucrose agar (TCBS) (Lipp et al., 2002; Sampai et al., 2022). The cultures were incubated overnight at 36 °C, and each plate was cultured once more and incubated until the isolated colonies were pure (Patterson, Porter & Ritchie, 2002; Ritchie, 2006). It is pertinent to note that the selection of the temperature and culture media was tailored to isolate enteric bacteria, as it constituted the primary objective of the research.

Bacterial identification

Bacterial isolates for molecular identification of the 16S gene were selected based on basic characteristics, such as Gram staining and bacterial morphology. Each isolate was grown overnight in broth for further DNA extraction. Extraction of DNA was performed with the MasterPureTM kit (EPICENTRE), and the universal primers for bacteria 27F (5′-AGAGTTTGATCMTGGATCAG-3′) and 518R (5′-ATTACCGCGGCTGCTGG-3′) were used to amplify the hypervariable V1–V3 region (Chakravorty et al., 2007). The PCR mixtures were as follows: 0.4 mM dNTP mix (10 mM dNTP mix; Promega, Madison, WI, USA), 1X reaction buffer (ABM, 10X PCR Buffer), 2 mM MgSO₄ (ABM, 25 mM MgSO4), and 0.1 (U/mL) Taq Polymerase (MyTaq™ DNA Polymerase 50 (U/µL; Bioline, London, UK), each of which was completed to a volume of 25 µL. The PCR parameters used were 5 min at 95 °C; 1 min at 94 °C, 1 min at 63 °C, and 1 min at 72 °C for 35 cycles; 7 min at 72 °C; and 10 min at 4 °C. The amplified product was confirmed in 2% agarose and visualized in a UV chamber, these parameters were obtained after different trails were proven. The PCR products were sequenced in the forward direction using the dideoxy terminator cycle-sequencing method in an ABI 3730XL sequencer (Applied Biosystems, Foster City, CA, USA) at SSIGMOL, Universidad Nacional de Colombia, Bogotá, Colombia. DNA sequences were aligned and manually inspected for errors using BioEdit software (Hall, 1999). Bacterial identity was confirmed by free-access software BLAST, and accession numbers are given in Supplemental Information 1.

Colony size structure and disease prevalence

In general, a total of 1604 colonies from six localities were recorded by random sampling. WPX was more common than WBD, for WPX the disease was related to the locality (chi-square test, X${}_{\mathrm{0.05,3}}^2=11.07$; P = 167.7), on the contrary WBD was not related to the locality (chi-square test, X${}_{\mathrm{0.05,3}}^2$ = 11.07; P = 5.7), however, it is stated that this may be the result of the low occurrence of this disease. Overall, healthy colonies predominated, with an average of 81.85% among the colonies surveyed (Fig. 2). Isla Arena showed the higher percentage of healthy colonies (92.59%), the lowest values of healthy colonies were observed in the Rosario Islands (69.9%). Disease prevalence was dominated by WPX, which was highest in TNNP (26.54%), followed by the Rosario Islands (26.32%), and the other localities exhibited a disease prevalence of WPX below 20%. The prevalence of WBD was less than 10% in all localities (Fig. 2).

The size-class distributions were grouped into 12 groups (1.45–5.71 cm²), most of which were represented by larger colonies between the sizes of 2.87–4.29 cm² (74.42%), whereas the other classes were 25.5%. The highest prevalence of both diseases was found in colonies grouped in the second-class size which contains small colonies (73.91%), followed by large colonies within the tenth size class, colonies around 4.64–5.00 cm2 (47.36%). The lowest prevalence of both diseases combined was found in size class 6 (19.2%) (Fig. 3). The multidimensional contingency table showed that the disease occurrence, locality, and size-class were not mutually independent (chi-square test, X${}_{0.05,126}^2$ = 153.1; P = 421.05), concluding that the disease occurrence was present, no matter what class-size was evaluated (chi-square test, X${}_{0.05,1573}^2$ = 168; P = 154.9).

Molecular analysis

Sixty-four pure bacterial isolates were obtained from healthy and diseased mucus samples, but only 33 were selected for molecular identification. The selection criteria were that at least one of each strain obtained from a diseased colony (if present) and one from a healthy colony were successfully purified from each locality. Some samples could not be sequenced due to poor-quality DNA. From the 32 sequenced strains, nine belonged to healthy coral, 12 to WBD, and 11 to WPX.

Thirteen strains were affiliated with Vibrio sp., six with Bacillus sp., three with Enterococcus sp., two with Enterobacter sp., and two with Staphylococcus sp. One strain was affiliated with Salinicola sp., Exiguobacterium sp., Kocuria sp., Cobetia sp., Pseudomonas sp., and Pseudoalteromonas. From healthy mucus, Salinicola sp. Pseudoalteromonas sp., Enterobacter sp., Cobetia sp., and Kocuria sp. were present in WBD samples, whereas Exiguobacterium sp. and Pseudomonas sp. in the WPX samples (Table 1).

The total number of bacterial isolates is shown in Table 1. However, Vibrio sp. were documented in each locality, followed by Bacillus sp. which were retrieved from three localities. Specifically, in TNNP, Kocuria oceani was found in WBD. In Isla Arena, Vibrio sp. were present in the healthy mucus, and Enterobacter sp. were present in the WBD samples. Rosario islands were represented by three strains obtained only from samples with WPX disease. Strains of Enterococcus sp. were exclusively isolated from San Bernardo Islands in disease mucus. Regarding Isla Fuerte, more diverse bacterial isolates were found, such as Salinicola sp. in healthy tissue; Exiguobacterium sp. from WPX samples, and Cobetia sp. and E. coli cultured from WBD samples. Finally, in Uraba Gulf, a strain of Pseudoalteromonas sp. was identified in WBD.

Bacterial isolates

The metabolic features of the holobiont are entirely influenced by environmental conditions, and any anomalies such as nutrient inputs, temperature variation, sediment suspension, and other anthropogenic tensors have negative impacts on coral health, resulting in the colonization of pathogenic and opportunistic bacteria or dysbiosis of the bacterial groups residing in the coral (Sunagawa et al., 2009). In particular, each locality in the surveyed Colombian reefs has its own oceanographic aspects that can affect the balance of coral health by changing the bacterial composition or the colonization of pathogens that have repercussions on coral diseases, which can be associated with the probiotic coral hypothesis (Reshef et al., 2006), which suggests that there is a dynamic relationship between the different symbiont microorganisms and the environmental conditions that interact with the holobiont. Next generation sequencing studies must be done to clarify support this statement.

Most of the bacterial isolates in this study belong to Enterobacteriaceae, which is considered a highly pathogenic family among biological systems (Brenner & Farmer, 2015). In marine organisms, its incidence and affectation trigger disease events in groups such as mollusks, fishes, and marine mammals (Guo & Ford, 2016). Although Enterobacteriaceae are a common member in different coral tissues, its role is unclear, but some data suggest that this family disrupts the holobiont wellbeing (Morrow et al., 2012; Roder et al., 2014). For example, Leite et al. (2018) found a relationship between the dissolved nitrogen concentration and the abundance of members of the family Enterobacteriaceae within different Brazilian coral populations, and they found a significant positive correlation for disease presence and polluted urban focuses. This finding indicates that this group may work as a proxy to establish control and conservation strategies against epizootic events in corals located near urban areas.

Uraba Gulf was characterized by the low presence of WBD. This locality receives contributions from the Atrato river characterized by the high abundance of total organic carbon and total nitrogen (Velez-Agudelo & Agudelo-Ramirez, 2016), and although WBD is not strictly related to river discharges, the presence of extra nutrients can favor the overgrowth and colonization of pathogenic bacteria that express virulence factors (Pollock et al., 2011). The causal agent for this disease is uncertain; however, a direct relationship has been found with representatives of the genus Vibrio sp. in different tissue isolations from diseased corals (Gil-Agudelo et al., 2007; Gignoux-Wolfsohn & Vollmer, 2015). Three strains of Vibrio sp. corresponded to WBD, which coincides with other studies, this also validated that the presence of Vibrio sp. may play a critical role in holobiont maintenance during the healthy and diseased state.

Many of the isolated and sequenced strains of the bacteria in this study coincide with those of a previous report (Sweet, Croquer & Bythell, 2014) in which bacteria of the genus Vibrio sp. and Bacillus sp. were isolated from healthy and diseased colonies of A. cervicornis, suggesting that there are bacterial groups that are closely linked to the appearence of diseases in acroporid corals. Pantos & Bythell (2006) used independent culture techniques in colonies of A. palmata with WBD and found a significant proportion of representatives of the Gammaproteobacteria class. In the present study, the sequenced bacteria were grouped primarily within this taxonomic level, indicating that members of Gammaproteobacteria may be a common group involved in advancement of the disease.

Exiguobacterium sp. were isolated from a colony with WPX in Isla Fuerte, representing a strain of interest, as it is part of the healthy microbiome in A. palmata and participates in a beneficial manner within the host; this strain retards the growth of pathogenic bacterial populations by inhibiting glycosylated compounds, such as those present in S. marcescens, presumably the causative agent of WPX (Krediet et al., 2013). The presence of this strain suggests that during the emergence of disease signs, some beneficial bacteria participate in the equilibrium and health of the host. Cobetia sp. were also observed in this locality, which participate in habitual metabolic processes such as the assimilation of carbohydrates, amino acids and other metabolic processes (Badhai, Ghosh & Das, 2016). Although both were isolated from diseased mucus, the strains are probably present because of a gradual change in the composition of bacteria during progression of the disease (Cárdenas et al., 2012; Roder et al., 2014). In A. palmata, this finding may suggest that the species harbors beneficial microbes with the potential to combat and stop advancement of the disease if environmental conditions are favorable but as the data presented here is preliminary in terms of microbiota, long-term studies combined with current methodologies should address this statement.

Kocuria sp. was isolated in TNNP, where the species were shown to be representative of the pathogenic microbiome that initiates signs of disease when combined with other bacteria (Sweet & Bulling, 2017). Enterobacter sp. were isolated in Isla Arena from WBD mucus. This pathogen and its representatives within the family Enterobacteriaceae contribute to the advancement of different diseases, such as black band and white plague; however, they have also been found in low proportions in healthy mucus (Frías-lópez et al., 2002; Morrow et al., 2012; Roder et al., 2014). The presence of Enterobacter sp. should be revised in future studies considering how the periodic contributions from the Magdalena River may affect the coral health. On the San Bernardo Islands, two strains of Enterococcus sp. were recovered from diseased mucus, and although their presence cannot be attributed to the emergence of disease, they appear to be associated with waters contaminated by surrounding human populations. Therefore, bacteria of fecal origins may contribute to and influence diseases in A. palmata (Sutherland et al., 2010).