Piscine mycobacteriosis – Involvement of bacterial species and reflection in pathology

Mycobacterium species shown to be present in a fish.

Currently, there is no accepted treatment protocol for mycobacteria infections in fish.Antibiotic resistance testing on fish is rarely performed.If a resistance test was done, susceptibility of mycobacteria towards different antibiotics seems to be dependent on the infecting Mycobacterium species and strain. 7It has been shown in experimentally infected yellowtails (Seriola quinqueradiata) that were orally treated with different antibiotics (rifampicin, streptomycin or erythromycin) that the treatment failed to eradicate the mycobacteria. 15 the presented study, we selected fish that were formerly diagnosed or suspected to be infected by mycobacteria from two zoological gardens in Switzerland (Zoo Basel and Tierpark Bern), with the goal to identify involved species of mycobacteria.Specifically, we addressed the following questions: (i) which mycobacteria species are involved in piscine mycobacteriosis and (ii) are there clinical or pathological differences between cases induced by M. marinum versus cases induced by other mycobacteria species?To investigate these questions, we performed a retrospective study using histological material from cases sent to the Centre for Fish and Wildlife Health (CFWH), University of Bern, Switzerland during the last ten years (2007-2016).

Sample selection
Archived material from 2007 to 2016 from previous fish cases were selected based on histological findings that confirmed or suspected piscine mycobacteriosis, originating from the Zoo Basel or Tierpark Bern.Confirmed cases are defined as the presence of acid-fast bacteria in histology, while suspected cases were selected based on the presence of indicative pathology without acid-fast bacteria in histology.In total, 30 cases from the Zoo Basel and 17 cases from the Tierpark Bern were selected for further examination.

Introduction
Piscine mycobacteriosis (fish tuberculosis) is a worldwide disease that affects an array of different fish species and has been associated with multiple Mycobacterium spp. 8,16 The course of this disease is known to be chronic progressive with varying clinical signs.Most lesions are unspecific, as weight loss, non-healing skin ulcerations, scoliosis, loss of colour, and exophthalmia.3 The mortality rate is usually low, while the development of multiple granulomas in the internal organs and in the skin is a common histological finding in routine fish necropsies.7,22 The histopathology of these granulomas often presents as a chronic inflammatory response with epitheloid macrophages, pigmented macrophages, an increased amount of fibroblasts and a necrotic centre where these bacteria are typically located.23 The bacteria located within these lesions can be identified using special stains, like Ziehl Neelsen.Mycobacteria are acid-fast aerobic organisms which react to the carbol fuchsine of the Ziehl Neelsen stain.5 Advanced technical molecular methods have now made it possible to differentiate Mycobacterium down to the individual species level. A a result of these molecular advancements, it was possible to identify the three leading pathogenic agents responsible for fish tuberculosis.8,20 From these three known pathogenic agents, M. marinum is supposed to be the most pathogenic one, and is known to occur both in marine and freshwater fish.8,23,27,32 Additionally, M. marinum has the potential to infect other vertebrates, including humans.3,14,17,24 The transmission of M. marinum is still not completely understood, but infection is thought to be spread by ingestion of infective material, such as bacteria loaded skin cells from an infected host, or through dermal injury if the density of bacteria in the environment is high enough.9,12 The prevalence of Mycobacterium spp. in European aquaria is reported to range between 41,7% and 46,8% 31,40 , while in wild populations little is known about its prevalence and impact.10 Once the infection is established, many factors make it challenging to eradicate these bacteria from the aquatic systems.Three main factors are responsible for the difficulties of eradication 1) longevity of the bacteria, up to two years in the environment, 2) difficulty to detect in standard screening tests, 2, 26 and 3) disinfection methods are not practical in a running aquatic system.23 All these factors contribute to an almost impossible eradication of some mycobacteria species. 23 Toavoid a potential introduction of the bacteria, imported fish should undergo a quarantine period, however, due to the long incubation period of the infection, it is difficult to establish an adequate quarantine time.spacer was performed with primers Sp1 (5´-ACC TCC TTT CTA AGG AGC ACC-3´) (AAGGA corresponds to the beginning of the spacer sequence) and Sp2 (5´-GAT GCT CGC AAC CAC TAT CCA-3´) as described.28 A positive control sample obtained from confirmed cases in a group of guppies (M.marinum, 218 bp, ACCTCCTTTCTAAGGAGCACCACGAGAAA CACTCCAATTGGTGGGGTGTAAGCCGTGAGG GGTTCTCGTCTGTAGTGGACGGAAGCCGGGTG CACAACAACAAGCAAGCCAGACACACTATTG GGTCCTGAGGCAACATCTCTGTTGGTTTCGG GATGTTGTCCCACCATCTTGGTGGTGGGGTGT GGTGTTTGAGAATTGGATAGTGGTTGCGAGCAT) and a negative control using water were included in the PCR procedure.To confirm the specificity of the PCR products, they were purified with WIZAR RD ® SV Gel and PCR Clean-Up System (Promega AG, Dübendorf, Switzerland). Th products were checked on a 1,5 % agarose gel for amplification and molecular weight.Samples with a weak band were reamplified before being sent to sequencing (Microsynth AG, Balgach, Switzerland).Sequencing results were determined by BLAST-n based on a search in the GenBank database (www.ncbi.nlm.nih.gov).

Classification of histological findings
Histological lesions were classified into three different groups: acute, chronic or chronic-active.Acute lesions were defined as the presence of macrophage accumulation without circumscribed granuloma formation, no fibrotic capsule, pigmented macrophages present or absent, multifocal central necrotic areas in the macrophage aggregations.Chronic lesions were classified based on the presence of a well circumscribed granuloma with central necrosis, surrounded by low numbers of macrophages and a rim of fibroblasts.Finally, chronic-active lesions were determined if both components (acute and chronic) were present in one organ.The histopatholog-ical changes were graded on a scale from 0-6, with 0 meaning no lesions and 6 severe.Additionally, the distribution was judged from focal to systemic.

Classification of clinical findings
Only confirmed cases of piscine mycobacteriosis were included, which resulted in six reports from the Zoo Basel and five reports from the Tierpark Bern.The clinical signs were classified according to the following criteria: dyspnea, variation of coloration, skin ulcerations, apathy, anorexia, motoric incoordination, others.Additionally, the origin of the affected fish was documented and classified according to fresh-or seawater and cold or warm water.

Cause of disease
As in many cases several infectious agents were present, mycobacterial infections were classified as either (i) the most probable cause for clinical signs and mortality, or as (ii) possible cause for clinical signs and mortality, or as (iii) a secondary finding.

Origin of cases
Forty-seven cases indicative or suspicious for mycobacteria infections were included in this study (Figure 1).Twenty-two cases were diagnosed as piscine mycobacteriosis, confirmed by acid fast bacteria indicative for mycobacteria detectable in the histological slides (totalling 47% of all studied cases).In 18 cases, we detected multiple well circumscribed granulomas in different inner organs or the skin, without detectable acid fast bacteria.In one additional case, the granulomatous inflammation was diffuse without well demarcated granuloma formation.In five cases, an increased amount of pigmented macrophages was present.In one case, Photobacterium damselae was isolated on a bacteriology plate, however, it was undetermined if it was the responsible agent for the granulomatous inflammation.Thirty cases originated from the Zoo Basel with four different water sources (cold/warm freshwater and cold/warm seawater) used in 45 aquaria (Figure 2).The majority of all examined cases (64%) originated from seawater with half of these cases belonging to the cold seawater group.The majority of the remaining cases belonged to the warm freshwater group (Figure 2a).
The majority of confirmed piscine mycobacteriosis cases originated from the warm freshwater group (63%), while only 25% belonged to the cold seawater group (Figure 2b).The Tierpark Bern cases were only differentiated between fresh-and seawater groups because only warm water is used in the Tierpark Bern.Out of the 17 submitted cases, nine cases originated from the freshwater group (53%).

Identification of involved bacteria
Of the initial 47 cases, mycobacteria could be confirmed in 17 cases.Fourteen out of the 22 histologically positive cases were also positive by PCR and sequencing.In three out of the 25 suspicious submissions, the involvement of mycobacteria was confirmed by molecular techniques (Table 1).Additionally, positive results were further subdivided into M. marinum and other Mycobacterium spp.
(Table 1), where it was determined that 86% of the M. marinum isolated cases belonged to fish originating from freshwater.Often multiple mycobacteria species were identified within a single case, which resulted in the total number of bacteria exceeding the total number of positive cases (Table 2).

Pathology
Lesions caused by M. marinum were dominated by multifocal infiltration of high numbers of macrophages and small necrotic areas (Figure 3, 4) which were interpreted as acute.Two cases, where M. marinum involvement was identified, showed chronic-active changes, while only one case presented with chronic lesions (Figure 3, 4).
On the other hand, cases that exclusively consisted of other mycobacteria species, the pathology consisted of more chronic alterations (30%) with the lesions showing a high proportion of fibrosis (Figure 4).
Cases associated with M. marinum were always classified as severe, while cases involving other Mycobacterium spp.were classified to be less severe (Figure 4b).

Clinical signs and cause of death
Within the eleven reports, we discovered that the majority of clinical signs consisted of apathy, variation of Three bacteria species (M.marinum, M. ulcerans and M. chelonae) dominated in seven out of the eight sequenced cases, where fish tuberculosis was diagnosed as cause of death (Table 4).coloration, skin ulcerations and anorexia.M. marinum cases never presented skin lesions, but all other clinical symptoms were similar to cases involving other Mycobacterium spp.(Table 3).

Discussion
For many years, the identification of mycobacteria was determined by gathering phenotypic data and biochemical test results of isolated species in culture medium. 3,6 wever, in the last few decades, new methods, using advanced technology, have provided a tool that allows for rapid and accurate identification of mycobacterial species, thus replacing older, more laborious methods. 6,13,21,25 A articular method on the rise is multiplex polymerase chain reaction (PCR), which allows for the simultaneous amplification of more than one sequence of target deoxyribonucleic acid (DNA) in a single reaction, thus saving time and money.11 One drawback of using PCR is that the sensitivity of detection has not been fully confirmed yet.28,30,37 In previous studies that tested the efficiency of the PCR detection found a success rate between 67-77%, 25,30 while we had a slightly lower detection success rate of 64%.
As a first step in the diagnostic cascade, presence of clinical signs is often judged.However, clinical signs of fish tuberculosis are mostly unspecific and cannot lead to final diagnosis.Signs, such as apathy, variation of coloration and anorexia were present in our study and these signs have been well documented in other studies as well. 3,4,28 I addition to the aforementioned clinical signs, the presence of shallow irregular ulcerations have also been described as a typical sign for piscine mycobacteriosis. 3,4,8,28 Th presence of skin lesions can be interpreted as a sign of a chronic prolonged course of the disease, which can also be a path for shedding infectious material via the wounds.Interestingly, we determined approximately 30% of our cases induced by Mycobacterium spp.beside M. marinum to show chronic lesions with a high proportion of skin ulcerations.However, skin ulcerations were absent in all fish affected by M. marinum.Additionally, M. marinum infections mostly showed a systemic distribution and were characterized by acute changes.
M. marinum is a slow-growing atypical Mycobacterium that is commonly found in freshwater and marine fish. 1 It is the most prominent, pathogenic Mycobacterium sp.found in fish tuberculosis cases. 4,8,23,34 Th combination of severe and mostly acute lesions supports the high pathogenicity of this species.The majority of cases involving M. marinum infections are found in fish living in warm seawater. 28However, in this study, 86% of M. marinum cases were found in freshwater aquaria.
M. marinum is closely related to another important Mycobacterium species, M. ulcerans. 36In our study, five isolates were confirmed as M. ulcerans.This species is considered to have the highest risk to develop a zoonotic potential, and it was described as the causative agent for the Buruli ulcer in humans. 38,39 his disease has been recently categorized as a new emerging infectious disease in humans. 34Clinical signs in humans are thought to be related to the toxin mycolactone. 33In fish, no toxic effects of mycolactone have been reported so far. 8,20 n medaka and zebrafish experimentally infected with M. ulcerans, no mortality occurred and only a mild inflammatory response was seen in histopathology. 19n accordance to our cases, M. ulcerans was always diagnosed as co-infection with M. marinum.
Furthermore, a mycobacteria strain was identified not reported before to occur in fish.M. holsaticum was described so far to occur only in mammals, mostly humans, causing pulmonary tuberculosis. 18,35 e identified M. holsaticum in a silver mooney fish (Monodactylus argenteus) from a seawater aquarium, presenting multiple skin ulcerations.In the same animal, M. marinum and M. ulcerans were demonstrated.Therefore, the role of M. holsaticum for the health of the fish remains unclear.The potential of fish as reservoirs for this Mycobacterium species has to be further investigated.

Acknowledgments
This study was financed by the Zoo Basel and the Tierpark Bern and own finances of the CFWH.The authors would like to thank the diagnostic team of the CFWH for their assistance during necropsies.A sincere thank you to the native speaker Jessica Rieder for her diligent proofreading of this manuscript.

Figure 1 :Figure 2 :
Figure 1: Overview of the initial diagnoses according to the histological findings, shown are 47 cases sent to the CFWH between 2007 and 2016

Figure 3 :
Figure 3: Histological picture of acute (a, b) and chronic cases (c, d); a. spleen of a lyretail coral fish (Pseudanthias squa mipinnis) showing pathology interpreted as acute lesions, like infiltration with high numbers of macrophages and single cell necrosis in the centre of the infiltration (circle); b. acid fast bacteria are visible in high numbers intracellular in the macrophages and extracellular in the surrounding tissue (open arrowheads); c. liver of a lyretail coral fish (Pseudanthias squamipinnis) showing chronic lesions characterized by multiple well circumscribed granulomas with a central necrosis, a small rim of macrophages and peripheral thick rim of fibroblasts (closed arrowhead); d. acid fast bacteria are present only extracellular in the surrounding tissue (open arrowhead); bars = 50mm; HE stain (a, c), ZN stain (b, d)

Figure 4 :
Figure 4: a. Pathology of fish infected with M. marinum or with other Mycobacterium spp., blue bars = acute lesions, orange bars = chronic lesions, yellow bars = chronic-active lesions; b.Severity of lesions in fish infected with M. marinum or with other Mycobacterium spp., grey bars = severe lesions, black bars = moderate lesions, white bars = mild lesions.

Table 1 :
Correspondence of histological results and results obtained by PCR and sequencing; proportion of M. marinum in cases diagnosed as piscine mycobacteriosis based on histological findings and other diagnoses

Table 2 :
Identified Mycobacterium spp. in cases originating from freshwater or seawater aquaria, the accession numbers of sequences are given in parenthesis.

Table 3 :
Clinical signs described by the zookeepers before sending the cases to the CFWH for necropsy