Pseudocapillaria tomentosa, Mycoplasma spp., and Intestinal Lesions in Experimentally Infected Zebrafish Danio rerio

Abstract

Intestinal neoplasms and preneoplastic lesions are common in zebrafish research facilities. Previous studies have demonstrated that these neoplasms are caused by a transmissible agent, and two candidate agents have been implicated: a Mycoplasma sp. related to Mycoplasma penetrans and the intestinal parasitic nematode, Pseudocapillaria tomentosa, and both agents are common in zebrafish facilities. To elucidate the role of these two agents in the occurrence and severity of neoplasia and other intestinal lesions, we conducted two experimental inoculation studies. Exposed fish were examined at various time points over an 8-month period for intestinal histopathologic changes and the burden of Mycoplasma and nematodes. Fish exposed to Mycoplasma sp. isolated from zebrafish were associated with preneoplastic lesions. Fish exposed to the nematode alone or with the Mycoplasma isolate developed severe lesions and neoplasms. Both inflammation and neoplasm scores were associated with an increase in Mycoplasma burden. These results support the conclusions that P. tomentosa is a strong promoter of intestinal neoplasms in zebrafish and that Mycoplasma alone can also cause intestinal lesions and accelerate cancer development in the context of nematode infection.

Introduction

The zebrafish has emerged as a very important model in biomedical research,¹ second only to the mouse.² As with other laboratory animals, the impacts of underlying infections can seriously compromise research.³ Consequently, it is important to define the etiology and pathogenesis of common diseases of laboratory zebrafish. To this end, we have previously examined data available through the Zebrafish International Resource Center (ZIRC) diagnostic program (https://zebrafish.org/health/index.php) and reviewed histopathology records of ∼18,000 zebrafish as part of over 1,300 submissions from over 300 research laboratories spanning the periods from 2000 to 2019. These efforts reveal two common intestinal diseases: epithelial carcinomas^4,5 and infections by the nematode Pseudocapillaria tomentosa.^6,7 Similar to other capillarid nematodes, the worm invades intestinal tissues and causes severe inflammatory changes. Intestinal neoplasms and worms have occurred in about 17% and 15% of the facilities, respectively.⁸ The retrospective surveys^3,8 also showed no correlations with sex or strain of fish, further supporting an infectious etiology. The high prevalence of these intestinal neoplasms underscores the importance in determining their cause and the role the worm may play with progression of these neoplasms.

Infectious agents are increasingly being implicated as initiators or promoters of neoplasm, and certain helminths are recognized as the cause or at least a promoter for neoplasia.⁹ Kent et al.⁶ reported an association of the intestinal neoplasms in zebrafish with P. tomentosa based on review of the data from a carcinogenesis study conducted by Spitsbergen et al.,¹⁰ where zebrafish exposed to both DMBA and P. tomentosa demonstrated a higher prevalence of intestinal tumors than uninfected fish exposed to this carcinogen. We recently conducted a retrospective study of the ZIRC database and found a strong statistically significant association of the occurrence of the worms, by both case and submitting laboratories, with intestinal neoplasia.⁸ However, the worm is unlikely the primary cause because among the laboratories with intestinal neoplasia, the worms were absent in 70% of the fish. In addition, Burns et al.¹¹ reported that intestinal neoplasms could be transmitted in the absence of the worm by water borne exposure after about 9 months, and the disease was associated with a specific Mycoplasma strain related to Mycoplasma penetrans. The time of occurrence of neoplasms in this laboratory study was consistent with observations from zebrafish facilities, where intestinal neoplasms occur most often in fish older than 1 year. Hence, the cancers may have been initiated as larvae or postlarvae fish, but they were seen in fish initially exposed as adults.¹¹ Following that study, Gaulke et al.¹² performed infection studies with P. tomentosa and observed development of intestinal tumors after only 3 months exposure. Microbiome profiling revealed presence of a bacterium related to M. penetrans in individuals with the neoplasms. The preneoplastic lesions and intestinal neoplasms in the intestine of fish from research facilities⁴ and those observed in these transmission experiments^11,12 are indistinguishable at a histologic level, and hence, we assume at this time that the majority of them have a similar underlying cause. In our present investigation, we conducted two exposure experiments to elucidate the roles of P. tomentosa and Mycoplasma sp. in intestinal neoplasia in zebrafish.

Methods

Two separate inoculation experiments were conducted to elucidate the roles of Mycoplasma sp. and P. tomentosa in intestinal neoplasia formation in zebrafish. The Mycoplasma sp. was originally described as associated with the tumors in Burns et al.¹¹ The Mycoplasma sp. used in the present study was isolated in culture from affected fish from recipient group E of that study and, henceforth, is referred to as Mycoplasma E. The first experiment (Experiment 1) included three experimental groups; exposed to Mycoplasma E, exposed to Mycoplasma E and P. tomentosa, and unexposed (controls). At that time, we did not appreciate the abundance of members of the genus Mycoplasma in the microbiomes of our zebrafish population. Once we became aware of the extent of these bacteria in our zebrafish population that were not intentionally inoculated with Mycoplasma, we conducted a second experiment to validate previous findings about the impact of P. tomentosa alone, in the context of our fish population's endogenous microbiota. This experiment (Experiment 2) included two groups: P. tomentosa exposed and controls.

Fish and husbandry

Both experiments used AB line zebrafish from the Sinnhuber Aquatic Resource Center (SARL), Oregon State University. Fish from this facility are free from important zebrafish pathogens since it was established in 2007, including P. tomentosa.^13,14 Moreover, as part of their disease screening protocols, they routinely examine sentinel and retired fish from various lines by histology conducted by one of us (M.L.K.), and neither the intestinal neoplasm nor preneoplastic lesions have been seen in over some 3,000 fish that were examined by histology from that laboratory since 2007. Our vivarium contains flow through water (∼100 mL/min/tank), derived from charcoal filtered city water. Temperature was maintained at 27°C–28°C, with conductivity at 115–125 μS and at pH ∼7.5. Light in the vivarium is provided for 14 h/day. The experiments were conducted by approval of the Oregon State University and University of Oregon Institutional Animal Care and Use Committees, and both universities are accredited by the American Association for Accreditation of Laboratory Animal Care.

Microscopic examinations

For both experiments, multiple time points over several months were surveyed, with about 6 fish from each replicate tank sampled per time point (Fig. 1 and Table 1). Moribund fish were included in histological analysis (Table 1), whereas those fish that died were not examined because rapid postmortem autolysis made them unsuitable for histologic examination. Fish were euthanized by hypothermia,¹⁵ then gavaged with 4% paraformaldehyde,¹⁶ an incision was then made in the flanks near the abdomen, and then the fish were preserved in the same fixative for at least 2 weeks before processing. Fish were then processed into paraffin blocks using standard methods, and about 10 serial slides were prepared and 3 ribbons were mounted on each slide, resulting in about 15 fish sections/slide. Alternate sections were stained with hematoxylin and eosin (H&E) and evaluated by one of us (M.L.K.). Unstained adjacent slides were retained for fluorescence in situ hybridization (FISH) staining, and these were selected if the corresponding H&E slide had abundant intestinal tissue.

FIG. 1.

Time line for exposure and sampling of zebrafish; Experiment 1, exposed to either Mycoplasma E, Mycoplamsa E with Pseudocapillaria Tomentosa, or Negative Controls and Experiment 2, P. tomentosa and Negative Controls. DPF, days postfertilization; EPC cells, Epithelioma Papulosum Cyprini cells previously infected with Mycoplasma E. Color images are available online.

Table 1.

Two Experiments of Zebrafish Exposed to Mycoplasma sp. and Worms (Pseudocapillaria tomentosa)

Treatment	WP-M	WP-W	Tank No.	Sample size	W-Prev	Dys-Prev	Dys-Sev	Inflam-Prev	Inflam-Sev	Tumors
Experiment 1
Control	9	3	1	6	0	0	0	0	0	0
			2	6	0	0	0	0	0	0
			3	6	0	0	0	0	0	0
			4	6	0	0	0	0	0	0
Control	15	9	1	6	0	0	0	0	0	0
			2	6	0	0	0	0	0	0
			3	6	0	0	0	0	0	0
			4	6	0	0	0	0	0	0
Control	26	19	1	6	0	0	0	0	0	0
			2	6	0	0	0	0	0	0
			3	6	0	0	0	0	0	0
			4	6	0	0	0	0	0	0
Control	32	25	1	6	0	0	0	0	0	0
			2	6	0	0	0	0	0	0
			3	6	0	0	0	0	0	0
			4	6	0	0	0	0	0	0
Myco	9	3	1	6	0	2	3	1	1	0
			2	6	0	1	1	2	2	0
			3	6	0	0	0	0	0	0
Myco	15	9	1	6	0	2	2	0	0	0
			2	6	0	2	2	1	1	0
			3	6	0	1	1	0	0	0
Myco	26	20	1	6	0	2	2	0	0	0
Myco	32	25	2	6	0	2	2	0	0	0
			3	6	0	2	2	0	0	0
Myco	32	24	1	5	0	1	2	1	1	0
			2	5	0	1	2	0	0	0
			3	5	0	2	2	2	2	0
Myco	35	27	1	5	0	0	0	0	0	0
			2	5	0	3	3	0	0	0
			3	4	0	2	2	0	0	0
Myco+W	9	3	1	6	4	4	8	5	7	0
			2	6	5	5	7	5	9	0
			3	5	3	3	3	3	3	0
Myco+W	15	9	1	6	6	3	9	5	11	0
			2	6	5	4	7	3	7	0
			3	6	5	5	8	4	7	0
Myco+W	17	11	1a	1	1	1	3	1	2	1
Myco+W	18	12	3a	1	1	1	1	1	3	0
Myco+W	19	13	2a	1	1	1	3	1	3	1
Myco+W	22	16	1a	1	1	1	2	1	2	1
Myco+W	23	18	2a	1	1	1	2	1	2	0
Myco+W	24	18	3a	1	0	1	3	1	3	1
Myco+W	26	20	1	5	5	5	15	5	15	4
			2	5	5	5	10	5	11	2
			3	5	4	4	10	4	10	2
Myco+W	47	41	1a	1	1	1	3	1	3	0
Myco+W	53	47	1	4	3	3	6	4	5	0
			2	2	1	0	0	1	2	0
			3	2	2	2	5	2	4	1
Experiment 2
Control	NA	3	1	6	0	0	0	0	0	0
			2	6	0	0	0	0	0	0
Control	NA	8	1	6	0	0	0	0	0	0
			2	6	0	0	0	0	0	0
Control	NA	29	1	6	0	0	0	0	0	0
			2	6	0	0	0	0	0	0
Worms	NA	3	1	6	6	5	8	4	5	0
			2	6	3	3	3	3	3	0
Worms	NA	9	1	6	2	2	2	2	3	0
			2	5	0	0	0	1	1	0
Worms	NA	16	1	7a	6	7	16	6	14	1
			2	4	4	4	8	4	8	0
Worms	NA	30	1	3	2	3	5	3	3	0
			2	2	2	2	2	2	4	1

Experiment 1 had three groups; Mycoplasma sp. exposed (Myco), Mycoplasma with worms (Myco+W), and controls. Experiment 2 had two groups; exposed to worms and controls. Time designations pertain to time after exposure for groups that were exposed to pathogens. Severity of dysplasia/hyperplasia (Dys) or inflammation (Inflam) is scored 0–3. Neoplasia is positive (+) or negative (0). Dys-Sev and Inflam-Sev are the severity from an additive score from all fish in the sample.

^aMoribund fish.

NA, not applicable; Prev, prevalence; WP-M, weeks postexposure to Mycoplasma; WP-W, weeks postexposure to worms.

Each fish was scored for hyperplasia, dysplasia, neoplasia of the epithelium, and inflammation of the lamina propria as described previously.¹² In brief, neoplasms were diagnosed when anaplastic epithelial cells invaded through the lamina propria, and nests or pegs of these cells were observed in the muscularis. Scores ranged from zero (absent) to three (severe) (Fig. 2). For proliferative lesions, those with neoplasia were scored as 3 as they all also had severe preneoplastic changes (hyperplasia or dysplasia). Severity scores at each sample time for each tank were calculated by adding all scores together and dividing the results by number of fish in the sample (Table 1). Our scoring system for these changes, 0–3, is not “ordinal,” in that fish with a score of 3 have severe lesions that were consistently far greater than thrice as severe as a score of 1.

FIG. 2.

Intestine histology. H&E. Bar = 50 μm. (A, B) Fish exposed to Mycoplasma E (Exp. 1); arrows and square = region of hyperplasia/dysplasia. Note loss of bipolar, columnar appearance of epithelial cells. (C, D) Intestinal neoplasia in a fish exposed to Mycoplasma E and Pseudocapillaria tomentosa. Arrows = pegs or nests of neoplastic epithelial cells in lamina propria and in (D) nests within the muscularis. (E) Low magnification of fish exposed to Mycoplasma E and P. tomentosa (W). The epithelium (D) is severely dysplastic, flatten, or absent in some regions. LP exhibits severe chronic inflammation, and nests of neoplastic occur throughout the muscularis (arrows) and into the serosa/coelom (S). E, worm eggs; H&E, hematoxylin and eosin; LP, laminia propria. Color images are available online.

Fluorescence in situ hybridization

Unstained sections from selected fish were processed for FISH following the protocol described in our previous studies.⁶ Slides were selected from multiple fish from the various groups, control fish, as well as exposed fish, with or without lesions as determined by M.L.K. These slides were then processed for FISH and examined by one of us (E.S.W.) with no knowledge of the fish's status by groups or lesions.

Slides with paraffin embedded sections were deparaffinized with Clear-Rite 3 (Richard-Allan Scientific, Kalamazoo, MI) and rehydrated through decreasing concentrations of ethanol into sterile ddH₂0 following our general laboratory protocol.¹⁶ Prewarmed (70°C) hybridization buffer (20 mM Tris-HCl [pH 7.4], 0.9 M NaCl, 0.1% SDS, 35% v/v formamide) was applied to slides to equilibrate sections for 1 min and then removed. Prewarmed (70°C) hybridization buffer mixed with probes at a working concentration of 25 pmol was then applied to equilibrated slides, and a glass coverslip was placed on top to disperse probe mix and cover all sections. Slides were placed in a light excluding staining tray (with sterile water to maintain humidity) and incubated overnight at 48°C. After incubation, slides were soaked for 10 min at 48°C in sterile prewarmed 1 × phosphate buffered saline (PBS), cover slips were removed, and slides were subsequently soaked two more times in fresh sterile 1 × PBS for 10 min each, gradually bringing slides and PBS to room temperature.

Sections were processed as described above with a mixture of Mycoplasma genus specific oligonucleotide probes to 23s ribosomal DNA (rDNA) with Cy3 attached to 5′ and 3′ ends (Mycoplasma_295-3: AAGGAACTCTGCAAATTAACCCCGTA; Mycoplasma_295-4: AAGGAACTCTGCAAATTCATCCCGTAAG). We selected these 23s rDNA probes because they successfully detected M. penetrans in samples using a DNA microarray.¹⁷ As a positive control to visualize all bacteria, including Mycoplasma, a mixture of eubacterial oligonucleotide probes with fluorescein attached to 5′ and 3′ ends was used (Eub338-I: GCTGCCTCCCGTAGGAGT; Eub338-II: GCTGCCACCCGTAGGTGT; and Eub338-III GCTGCCACCCGTAGGTGT).¹⁸ 4′,6-diamidino-2-phenylindole (DAPI) was used to visualize DNA in eukaryotic and some prokaryotic cells (Integrated DNA Technologies Coralville, IA; www.microbial-ecology.net/probebase). Slides were mounted with VECTASHIELD Antifade Mounting Medium with DAPI (Vector Laboratories, Burlingame, CA) and sealed with a coverslip. Slides were assessed and imaged using a Nikon Eclipse Ti inverted microscope (Nikon Instruments, Inc., Melville, NY) equipped with an Andor iXon3 888 camera (Andor Technology, South Windsor, CT).

Sections were graded blindly with no knowledge of group status or histology. However, slides were selected for FISH testing that had a large amount of intestinal tissue based on observations from adjacent slides stained with H&E. The entire slide was evaluated based on a 0–6 scale by identifying areas within the intestines that were labelled with both Mycoplasma and eubacterial probes (Fig. 3). Scoring was as follows: 0 = no double positive signal in any sections on a slide; 1 = 1–2 particles, or 1–2 aggregates; 2 = 3–5 particles, or 3–5 aggregates; 3 = most, or every, section on slide has 1–2 particles or aggregates; 4 = most, or every, section on slide has 3–4 particles or aggregates; 5 = most, or every, section on slide has 5–10 particles or aggregates; and 6 = almost all or all sections have >10 particles or aggregates per section. Scores for each fish are then recorded in Table 2, with a total score adding particles and aggregates. Particles are singular points of fluorescence corresponding in size to single bacterial cells. Aggregates are particles of fluorescence that were spatially connected, corresponding in size to an aggregate of many bacterial cells.

FIG. 3.

Intestines with Mycoplasma and Pseudocapillaria tomentosa stained with either FISH or H&E. Bars = 50 μm for (A, B, D, F), or 10 μm for (C, E). W = worms, arrows = red/orange particles or aggregates, positive for Mycoplasma. (A, B) Adjacent slides stained with either H&E or FISH, with cross sections of worms either in the epithelium or in lumen. (B) Particles, including one within a worm. (C) FISH showing worms with adjacent particle. (D) Small aggregate, intracellular at tip of epithelial cell. (E) Intracellular particle adjacent to goblet cell. (F) Multiple intercellular aggregates. FISH, fluorescence in situ hybridization. Color images are available online.

Table 2.

Mycoplasma in Zebrafish Intestines Examined with Fluorescence In Situ Hybridization Genus Specific Probe in Two In Vivo Trials

Treatment	WP-M	WP-W	Tank No.	Fish No.	FISH P	FISH A	FISH total	W	Dysplasia	Inflam	Neoplasia
Experiment 1
Control	9	3	1	1	1	2	3	0	0	0	0
			1	3	2	0	2	0	0	0	0
			1	4	1	3	4	0	0	0	0
			1	1	2	1	3	0	0	0	0
			2	4	3	0	3	0	0	0	0
			3	1	2	2	4	0	0	0	0
			3	2	1	1	2	0	0	0	0
Control	15	9	1	8	2	2	4	0	0	0	0
			2	11	1	1	2	0	0	0	0
Control	25	19	2	14	2	1	3	0	0	0	0
Control	32	24	1	22	2	0	2	0	0	0	0
			3	22	1	4	5	0	0	0	0
			3	23	4	2	6	0	0	0	0
			2	24	0	4	4	0	0	0	0
Myco	9	3	1	1	2	0	2	0	1	0	0
			1	2	2	3	5	0	2	1	0
Myco	15	9	1	7	0	2	2	0	0	0	0
			1	8	3	2	5	0	0	0	0
			1	10	1	1	2	0	1	0	0
			2	12	2	2	4	0	0	0	0
			2	10	2	1	3	0	0	0	0
			3	7	1	2	3	0	1	0	0
			3	12	2	1	3	0	2	1	0
Myco	26	20	1	15	1	1	2	0	1	0	0
			1	17	1	3	4	0	0	0	0
			2	16	1	3	4	0	1	0	0
			2	17	3	3	6	0	0	0	0
			3	15	2	3	5	0	0	0	0
Myco	32	24	1	22	2	2	4	0	1	0	0
			3	19	1	1	2	0	1	0	0
Myco+W	9	3	1	3	4	0	4	0	0	1	0
			1	4	0	0	0	0	0	0	0
			1	6	1	1	2	2	3	2	0
			2	4	1	0	1	1	1	2	0
			2	6	0	0	0	1	1	1	0
			3	2	0	2	2	2	1	1	0
			3	5	2	3	5	1	1	1	0
			3	6	3	1	4	2	x	x	X
Myco+W	15	9	1	8	0	3	3	3	2	3	0
			1	10	3	0	3	1	0	1	0
			2	7	4	1	5	2	2	2	0
			2	8	4	0	4	2	0	2	0
			2	9	3	1	4	1	1	0	0
			2	10	2	1	3	0	0	0	0
			3	7	3	1	4	2	0	1	0
			3	9	4	0	4	1	1	1	0
			3	10	2	0	2	3	2	2	0
			3	12	1	0	1	1	1	0	0
Myco+W	26	19	1	21	3	3	6	0	3	3	0
			2	18	1	3	4	0	0	0	0
			2	19	2	4	6	1	3	3	1
			2	21	1	5	6	3	3	3	1
Myco+W	32	26	3	20	2	1	3	2	1	0	0
			3	21	1	5	6	3	3	3	1
Myco+W	37	28	1	24	3	2	5	2	2	1	0
			1	25	1	1	2	0	0	1	0
			2	25	1	2	3	0	0	0	0
			3	23	3	0	3	2	3	2	1
Experiment 2
Control	NA	8	1	7	1	0	1	0	0	0	0
			1	8	1	1	2	0	0	0	0
			1	9	2	1	3	0	0	0	0
			1	10	1	0	1	0	0	0	0
			1	11	2	0	2	0	0	0	0
			1	12	1	1	2	0	0	0	0
			2	8	2	2	4	0	0	0	0
Control	NA	29	1	13	1	0	1	0	0	0	0
			1	16	0	1	1	0	0	0	0
			1	18	2	1	3	0	0	0	0
			2	13	0	0	0	0	0	0	0
			2	15	0	3	3	0	0	0	0
			2	18	2	2	4	0	0	0	0
			2	20	3	1	4	0	0	0	0
Worms	NA	3	1	1	0	2	2	1	2	2	0
			1	6	3	0	3	1	2	0	0
			2	1	1	1	2	0	0	0	0
			2	4	1	1	2	1	1	1	0
Worms	NA	9	1	8	1	4	5	0	0	1	0
			1	9	0	2	2	0	2	1	0
			1	12	2	2	4	1	0	0	0
			2	12	2	1	3	0	0	0	0
			1	19	2	1	3	3	3	3	0
			1	20	4	1	5	2	1	2	0
			2	17	2	3	5	1	1	1	0
Worms	NA	16	2	21	2	0	2	1	3	1	1
Worms	NA	30	1	24	3	2	5	1	3	3	1

Time designations pertain to time after exposure for groups that were exposed to pathogens. Time designations pertain to time after exposure for groups that were exposed to pathogens. FISH results are presented as individual particles or as aggregates of presumptive Mycoplasma cells. Severity worm infection (W), dysplasia/hyperplasia (Dys), or inflammation (Inflam) is scored 0–3. Neoplasia is positive (+) or negative (0).

FISH, fluorescence in situ hybridization; Fish No., individual fish number; FISH total, addition of scores for both particles (P) and aggregates (A); Myco, exposed to Mycoplasma; Myco+W, exposed to Mycoplasma and worms (Pseudocapillaria tomentosa).

Statistics

Statistical analyses were conducted in R (R Core Team 2020; https://www.r-project.org). To test the association between FISH and histological scores, the following procedure was conducted. We built ordinal logistic regression models of the ordered categorical histological score data using the function polr from the MASS package.¹⁹ We note that ordinal regression may underestimate histological variation given that differences in histological score categories increase along the scoring range. We started with a base model that only included Experiment ID and built increasingly complex models by adding parameters to this base model, which enabled us to determine whether the additional parameters better explained the variation in either FISH or histological scores than Experiment ID did alone. Our progressive model construction approach added parameters one by one and included interaction terms in the most complex models. We sequentially applied likelihood ratio tests to these models using the anova function from the base stats package to select the most optimal model. For all histology measures, optimal models were consistently identified as the following when considering aggregate scores:

Moreover, all histology measures manifested a consistent optimal model when considering particle scores:

However, while these were determined to be optimal models, not all parameters in these models were necessarily significantly associated with the histological response variable. Using the selected model, we calculated odds ratios for each term and estimated their 95% confidence intervals (function confint in the base stats package). Terms with confidence intervals that overlapped zero were considered not significant and therefore not included in the reported table (Table 3). All other statistical comparisons were conducted using the two-sample Wilcoxon rank sum tests using the function wilcox.test from the base stats package.

Table 3.

Pathologic Changes, Treatments, Fluorescence In Situ Hybridization Results (Particles or Aggregates), and Pathology Endpoints in Two Zebrafish Laboratory Transmission Experiments in Which Groups of Fish Received Various Treatments: Pseudocapillaria tomentosa, Mycoplasma E, or Unexposed Controls

Pathology score	Selected model	Significant terms	Odds ratio	95% CI
Inflammation	Experiment # + Aggregate score + Worm exposure	Aggregate score	1.62	1.09–2.47
		Worm exposure	191.00	30.8–3880
Hyperplasia	Experiment # + Aggregate score + Worm exposure	Worm exposure	11.80	4.45–34.7
Neoplasm	Experiment # + Aggregate score + Worm exposure	Aggregate score	2.04	1.12–3.73
		Worm exposure	1.8 × 108	7.64e+07–4.39 × 108
Inflammation	Experiment # + Particle score + Worm_exposure	Worm exposure	118.00	21.5–2230
Hyperplasia	Experiment # + Particle score + Worm exposure	Worm exposure	11.20	4.28–32.5
Neoplasm	Experiment # + Particle score + Worm exposure	Worm exposure	1.74 × 108	7.62 × 107–3.99 × 108

CI, confidence interval; Experiment #, Experiment 1 or 2; Hyperplasia, hyperplasia and/or dysplasia.

Experiment 1

The Mycoplasma sp. used in the present study was isolated from affected fish from our previous transmission study.¹¹ The isolate used here was from recipient group E from that study and, hence, is referred to as Mycoplasma E. It was isolated from zebrafish on SP4 solid media (50 g/L BHI, 10 g/L Peptone, 5 g/L NaCl, 15 g/L Yeast extract, 15 g/L Agar, 200 mL/L sterile heat inactivated horse serum, 20 mg/L Amphotericin B, 250 mg/L Ampicillin, 250 mg/L Polymyxin B).²⁰ Colonies consistent with Mycoplasma were cut from agar and inoculated in SP4 broth (as above, without agar) for ∼7 days at 37°C with shaking. Then 2 mL of culture was transferred to a cryotube and centrifuged at 16,000 × 6 at 4°C for 30 min to pellet Mycoplasma cells, supernatant was aspirated, and the pellet was suspended in 750 μL fresh SP4 broth (room temperature) and 750 μL sterile 20% glycerol. The suspension was mixed and allowed to equilibrate at room temperature for 15 min; it was then flash frozen in liquid nitrogen and stored at −80°C. The similarity of Mycoplasma E 16s rDNA to the original sequence described by Burns et al.¹¹ from donor fish and M. penetrans from humans (ATCC 55252, GenBank JN935872.1) and the sequence from Mycoplasma BHJA from salmon²¹ GenBank number AY065998.1. 16S ribosomal RNA (rRNA) candidate sequences were aligned using the align.seqs command in mothur v.1.44.3²² and the SILVA SEED v132 database²³ as a template. A distance matrix was then generated using the resulting alignment and the dist.seqs (options: calc = nogaps, output = square) command in mothur. Uncorrected pairwise distance values were subtracted from 1.0 and converted to percentages to find the percent similarity.

Mycoplasma spp. grow very well in cell culture,²⁴ and Mycoplasma E was also grown in cell culture, using a common fish cell line used in virology; Epithelioma Papulosum Cyprini (EPC) cells from fathead minnow (Pimephales promelas). The cells were cultured with regular media changes, at 25°C in Leibovitz L-15 media supplemented with 10% sterile Heat Inactivated Fetal Bovine Serum, Pen/Strep (1:1000 dilution; Sigma), and Gentamycin (10 μg/mL). Cells were split twice in a 7-day period, and with the second split, half the cells were suspended in the abovementioned media, while the other half were suspended in Leibovitz L-15 media supplemented with 10% sterile Heat Inactivated Fetal Bovine Serum, Amphotericin B (2.5 mg/mL), Ampicillin (100 mg/mL), and Polymyxin B (50 mg/mL), or Mycoplasma media. Flasks containing Mycoplasma media were inoculated with Mycoplasma E glycerol stock using wooden sticks; flasks containing regular EPC growth media were exposed to sterile wooden sticks. Inoculated and control flasks were maintained at 30°C, with regular media changes. Twenty-four hours before inoculation into fish flasks, the medium was changed, but all cells were washed with sterile 1 × PBS and then replaced with Mycoplasma media. On the day of inoculation into fish cell line, Mycoplasma cells were harvested by centrifuging cells and the medium at 1000 g for 5 min, medium was removed, and then Mycoplasma cells were suspended in sterile 1 × PBS. Harvested EPC cells were counted and inoculated into each flask containing 4 days postfertilization (dpf) larvae at a density of ∼2.2 × 10⁶ EPC cells/flask.

Figure 1 provides a time line of the experimental design for exposure to Mycoplasma E and P. tomentosa. Approximately 800 embryos were transferred to University of Oregon from the SARL at Oregon State University, and bacterial loads were depleted using a previously described chorion disinfection method.²⁵ Surface sterilized embryos were placed in sterile T75 filtered tissue culture flasks (TPP, Trasadingen, Switzerland) containing 50 mL sterile embryo medium EM at a density of ∼1 embryo/2 mL. Flasks of zebrafish larvae were maintained at 28°C–30°C within a room with a 14-h light/10-h dark cycle. At 4 dpf, when zebrafish first start feeding, one half of the fish were exposed to Mycoplasma by feeding EPC cells from infected flasks, whereas the negative controls were fed uninfected EPC cells. Starting at 6 dpf, larvae from both groups were transferred to Petri dishes at density of ∼1 larvae/2 mL. Larvae were fed rotifers (Brachionus plicatilis) provided by the Animal Care Services, University of Oregon and given health checks twice a day, with water changes once a day. At 10 dpf, each Petri dish of larvae was transferred to a fresh T75 filtered tissue culture flasks with 150 mL sterile zebrafish embryo medium, and flasks were transported to the Kent vivarium in Nash Hall at Oregon State University for further maintenance. They continued to be fed rotifers from the University of Oregon, and Gemma Micro diet (Skretting, Westbrook, ME) was included in the diet starting at 21 dpf. Fish were first fed with the Gemma Micro 75 (75 mm) diet, and they were transitioned to Gemma Micro 300 (300 mm) diet. Fish were then fed twice daily with Gemma diet for the remainder of the study. Fish were held in two 9 L aquaria (exposed and controls) with a slow drip water exchange until 24 dpf. Flow was then increased to about 100 mL/min. At 35 dpf (31 days postexposure to Mycoplasma), fish were transferred to 9 L tanks with the same flow rate and set up into the three groups as follows: (1) Mycoplasma E only (three replicate tanks), (2) Mycoplasma with P. tomentosa (three replicate tanks), and (3) Negative Controls (four replicate tanks). Control fish were held in a separate adjacent room in the vivarium, but received water from the same system. Note that a fourth group, exposed to worms but not Mycoplasma, was not included as early in the experiment; we discovered that Mycoplasma spp. were present in controls by testing feces with the polymerase chain reaction (PCR) test as described in Burns et al.¹¹

For worm exposures, the Mycoplasma E+P. tomentosa group was exposed to 75 larvated worm eggs/fish as described in Kent et al.²⁶ when the fish were 42 dpf. The nematode eggs were disinfected with 30 ppm sodium hypochlorite for 10 min and then treated with sodium thiosulfate before exposure as previously described²⁶ to reduce concurrent bacteria from the donor fish for these parasites. Six fish from each tank were collected and examined by histology at 4 time points from 2 through 8 months. All groups of fish in tanks were inspected daily, and moribund fish were euthanized and processed for histology, whereas dead fish were not analyzed due to severe postmortem autolysis.

Experiment 2

We conducted a second experiment following the discovery that members of the genus Mycoplasma were more widespread in zebrafish and occurred in Experiment 1 controls. In addition, an experiment in our laboratory showed that fish exposed to P. tomentosa may develop tumors as soon as 3 months postexposure.¹² This experiment had two groups; exposed to P. tomentosa and negative controls, each with a replicate tank. Embryos from the same stock of AB zebrafish from SARL used as Experiment 1 were reared in the Kent vivarium. Fish were divided into 4 tanks, and 2 tanks received chlorine nematode eggs at 50 eggs/fish as described above. Similar to Experiment 1, fish from each tank were examined over an 8-month period (Fig. 1 and Table 1). As in Experiment 1, they were initially fed rotifers provided by the University of Oregon and then transitioned on to Gemma diet for the remainder of the study.

Results

Fish exposed to the nematode in both experiments exhibited profound pathologic changes in the intestine, including the occurrence of neoplasms in several fish (Fig. 2 and Table 1). Some fish exposed to Mycoplasma E alone exhibited mild-to-moderate hyperplastic or dysplastic changes (Fig. 2 and Table 1), whereas all controls in both experiments showed no histologic changes in the intestine. As observed in zebrafish facilities and other experiments,¹¹ the hyperplastic changes occurred for the most part in the anterior region of the intestine near the esophageal junction. Severe dysplastic changes and neoplasms in the fish infected with worms also occurred more anteriorly, but severe lesions extended through the mid intestine.

Experiment 1

A total of 23 of 83 fish in the Mycoplasma E only group had mild hyperplasia or dysplasia of the intestinal epithelium, which was first observed in the 9 weeks postexposure (weeks PE) sample. An additional four fish had equivocal proliferative changes in the epithelium, but they were scored as negative due to the uncertainty in diagnosis. Among the fish positive for epithelial changes in this group, six fish also exhibited mild, diffuse chronic inflammation of the intestinal lamina propria. No lesions were observed in the control fish. A Wilcoxon rank sum test showed that hyperplastic lesions were significantly more numerous in the fish exposed to only Mycoplasma E (W = 2,772; p = 0.0342).

With the fish exposed to both Mycoplasma E and the nematode, a total of 59 fish that were exposed to Mycoplasma E and the nematode were examined, and the nematode was observed in 91.0% of the fish (Table 1). A range of lesions from mild-to-severe hyperplasia and dysplasia was observed in 91.5% of the fish (Fig. 2). In addition, six fish had profound dysplasia of the epithelium to the extent that the folds of the intestinal plates were completely lost, and the underlying lamina propria was severely inflamed (Fig. 2E). Starting at 9 weeks PE to the worm (15 weeks PE to Mycoplasma E) 13 of 59 fish (22%) had lesions consistent with neoplasia, with pegs of neoplastic epithelial cells penetrating through the lamina propria and muscularis, as well as free nests of neoplastic cells in these locations (Fig. 2C–E). Seven moribund fish were included in examinations, and they consistently exhibited more severe lesions and most had the neoplasm (Table 1). An additional 12 fish died in the study and were not included due to severe postmortem autolysis. Wilcoxon rank sum tests showed that fish exposed to both Mycoplasma E and the nematode carried a higher prevalence of preneoplastic lesions (W = 2,556; p << 0.0001), as well as neoplasms (W = 4,716; p = 0.0029), compared to fish only exposed to Mycoplasma E.

Percent sequence identity of the 16S rRNA gene was used to gauge the taxonomic proximity of Mycoplasma E and the other notable Mycoplasma strains (Table 4). Mycoplasma E was actually more closely related to M. penetrans from humans (95.4%) compared to the original sequence reported in Burns et al.¹¹ (94.0%), whereas the sequence from Burns et al.,¹¹ was identical to the sequence from salmon.²¹

Table 4.

Percent Identity of 16s Ribosomal DNA Sequence Among Mycoplasma—Mycoplasma penetrans from Human, the Original Zebrafish Isolate from Burns et al. (Myco Burns), Myco E, and the Mycoplasma Sequence from Salmon from Hoblen et al. (Myco Salmon)

	Myco E	Myco Burns	M. penetrans	Myco Salmon
Myco E	670 bp	94.0	95.4	90.9
Myco Burns		200 bp	94.8	100
M. penetrans			686 bp	92.7
Myco Salmon				351 bp

Diagonal is the number of base pairs used in analysis.

Source: Burns et al.¹¹ and Hoblen et al.²¹

Experiment 2

Fish exposed to the nematode that were examined (n = 32) showed 64% prevalence of infection by the worm. As in Experiment 1, these fish exhibited a range of lesions from mild-to-severe hyperplasia and dysplasia (81%), as well as inflammation of the lamina propria (78%) (Table 1 and Fig. 4). Two fish were diagnosed with the intestinal neoplasm at 23 and 37 weeks PE, respectively. A total of eight fish among both tanks died throughout the experiment, starting 8 weeks PE to the worm, but these fish were not examined. As in Experiment 1, no lesions were observed in 36 control fish. Fish exposed to the nematode carried significantly greater number of lesions compared to these control fish as determined by a Wilcoxon rank sum test (Wilcoxon rank sum test: W = 490; p << 0.0001).

FIG. 4.

Histology scores (Inflammation or Hyperplasia/Dysplasia) for the two exposure experiments. Orange = Experiment 1, Blue = Experiment 2. X-axis indicates exposure condition: untreated (Myco−/Pcap−), exposed to Mycoplasma E (Myco+/Pcap−), exposed to only to Pseudocapillaria tomentosa (Myco−/Pcap+), or exposed to both pathogens (Myco+/Pcap). Black dots indicate means for the scores for the given treatment, and the black error bars indicate the bootstrapped 95% CIs for the means. CIs, confidence intervals. Color images are available online.

Fluorescence in situ hybridization (FISH)

Slides from zebrafish from all groups examined by FISH, including controls from both experiments, were positive for Mycoplasma spp. (Table 2). Mycoplasma signal often occurred associated with the intestinal epithelium as individual particles (corresponding in size to single bacterial cells) or aggregates of positive (red) staining (corresponding in size to an aggregate of many bacterial cells) (Figs. 2 and 3). In addition, FISH-positive particles were observed within worms (Fig. 3B), but not within worm eggs. The approximate abundance of Mycoplasma sp. based on FISH was similar among all groups, but the distribution of the Mycoplasma was associated with pathology. Specifically, increased abundance of aggregates was modestly but significantly associated with both inflammation and neoplasm scores (Table 3). No such associations were observed between the abundance of particles and pathology (Table 3).

Discussion

Infectious agents are increasingly being associated with neoplasia, ranging from oncogenic viruses, bacteria, protozoa, and helminths.^27,28 Development of cancer is usually considered to require at least two mutations in the host cell of origin and often results from multifactorial exposures and mutations. Thus clarification of the precise role of underlying chronic infections and concurrent inflammation, either as the initiator or a promoter, is challenging. With helminths, some well-recognized examples include Schistosoma haematobium in urinary bladder cancer and Opisthorchis viverrini, Opisthorchis felineus, and Clonorchis sinensis with liver cancer.^29,30 Regarding intestinal neoplasia, Schistosoma japonicum has been linked to colorectal adenocarcinomas in humans through case-controlled and epidemiological studies.³¹ Probably the best example of a nematode linked to cancer is Spirocerca lupi in dogs, which penetrates the esophageal lining and sarcomas develop at the site of infections.³² Bacteria have also been associated with cancers, with Helicobacter pylori being the best studied example of a carcinogenic bacterium and driver of gastric cancer.³³ In the much more densely colonized distal gastrointestinal tract, several bacteria have been implicated in the development of colorectal cancer, including Fusobacterium nucleatum, enterotoxigenic Bacteroides fragilis, and Escherichia coli, expressing the polyketide synthase genomic island.³⁴ Growing evidence finds associations between Mycoplasma spp. and cancer. Yang et al.³⁵ reported an association between Mycoplasma hyorhinis and gastric and lung cancers through immunohistochemistry and PCR examinations of human cancers and were supported by studies with mice and cell culture. There are also various lines of evidence of Mycoplasma spp. as a cause or exacerbating prostate cancer.³⁶ M. penetrans in particular has been shown to promote malignant transformations in the gastric epithelium of immunocompromised mice.³⁷

The mechanistic basis for microbial-associated carcinogenesis is likely to be diverse.³⁸ In some cases, a specific microbial toxin is implicated in driving cancer formation, as in the case of the H. pylori translocated effector protein CagA, which we showed could induce intestinal neoplasia in zebrafish³⁹ or the DNA alkylating genotoxin colibactin of E. coli.⁴⁰ In the case of nematode infections, specific metabolites from worms may lead to formation of DNA adducts, and thus, worms may actually be initiators, rather than just merely promoters, of cancer.⁹ In Mycoplasma spp., Yang et al.³⁵ identified a specific protein (p37) in M. hyorhinis that acted as a carcinogen, and Zella et al.⁴¹ reported a Mycoplasma fermentans DnaK with oncogenic properties. In other instances of microbial-associated cancers, specific microbes or microbial consortia promote generic carcinogenic processes, such as chronic inflammation, or inhibit protective processes such as immune clearance of cancer cells.²⁸ Chronic inflammation is a hallmark of chronic helminth infections, which have been linked to cancer by various mechanisms, including cytokine alterations.²⁹ In addition, liver fluke infections have been associated with increased burdens of H. pylori, suggesting that chronic inflammation can be caused by infection-induced changes in the host microbial ecology.²⁹ Certain Mycoplasma species are also well known to promote generic carcinogenic processes, such as upregulating inflammation and inhibiting p53 mediated cell cycle control and apoptosis.⁴² Another example of bacterial promoters in neoplasia with fish in research, Mycobacterium marinum, probably through chronic inflammation, increases the occurrence of liver cancers in a Japanese medaka (Oryzias latipes) exposed to benzo-a-pyrene.⁴³

Our first transmission experiment found a high abundance of Mycoplasma in populations with intestinal neoplasia based on high-throughput sequencing of the 16S rRNA gene, while this genus was found infrequently in the bacterial microbiome in control fish without tumors¹¹ The original isolate from donor fish with neoplasms from those experiments was related to M. penetrans, and our subsequent study showed that multiple sequence variants form a clade of these related bacteria.¹² Likewise, Mycoplasma E used in the present study came from a recipient group of fish from our previous transmission study,¹¹ and it showed minor sequence differences from the predominate sequence from donor fish. Gaulke et al.¹² and our study here demonstrate that Mycoplasma spp. are more abundant in the zebrafish intestinal microbiome than we previously thought based on our earlier study.¹¹ This extends to other fishes, as Hoblen et al.²¹ showed that 96% of the microbiome of certain populations of wild Atlantic salmon (Salmo salar) was composed of M. penetrans-like bacteria. Moreover, a recent study also showed that members of the Mycoplasmataceae were common in salmon.⁴⁴ Interestingly, the 16s rDNA sequence from the Mycoplasma sp. from salmon²¹ was identical to the sequence from zebrafish in Burns et al.¹¹ Hence it would be reasonable to assign all of these isolates from zebrafish and salmon to M. penetrans sensu lato given that the 16s rDNA differences between different species within the genus Mycoplasma are so dissimilar. Further taxonomic clarification on the relationships of isolates from fish with M. penetrans and related species will require analysis of more complete rDNA sequences, including more sequences from fish, and perhaps other genes.

Using our Mycoplasma FISH probe, which was specific only to the genus level, we observed similar levels of Mycoplasma among all zebrafish treatment groups, regardless of whether they were exposed to worms, Mycoplasma E, or neither. In the present study when the worm was absent, only fish deliberately exposed to Mycoplasma E developed hyperplastic or dysplastic lesions in the epithelium. Preneoplastic changes (hyperplasia leading to dysplasia) in the epithelium and inflammation in the lamina propria have been consistently observed in populations of zebrafish that have the neoplasm.^4,11,12 The earliest time in which tumors were observed following Mycoplasma exposure by cohabitation with infected fish or exposure to effluent from positive tanks was 8 months,¹¹ and most of the intestinal neoplasms in zebrafish from research facilities are not seen until fish are about 1 year old.⁴ In contrast, in Experiment 1 almost all of the fish were collected before 8 months, and perhaps tumors would have developed without exposure to P. tomentosa if they had been incubated with the Mycoplasma E isolate for longer. Now that it is recognized that the M. penetrans-like bacteria are actually a clade with some variation, it is possible that among this clade some strains are more carcinogenic.

Evidence suggests that both P. tomentosa and Mycoplasma E enhance development of the intestinal tumors commonly seen in zebrafish, including both in vivo laboratory experiments,^6,10–12 and in retrospective evaluation of hundreds of diagnostic cases submitted to the ZIRC diagnostic service.⁸ With regards to the nematode, it enhances both the incidence to intestinal cancer in fish exposed to a chemical carcinogen^6,10 and also shortens the time to onset of the tumors.¹² It was unlikely that the neoplasms seen in fish exposed to the worm were caused by an undetected agent that came from the source fish for the worms as we pretreated the worm eggs with chlorine at a relatively high concentration.

There is evidence that M. penetrans and P. tomentosa act synergistically to cause the lesions and ultimately tumors. Fish infected with both P. tomentosa and Mycoplasma E (in Experiment 1) had 22% tumors versus 4% in the fish in Experiment 2, which were exposed to the worm without deliberate exposure to the Mycoplasma E isolate. However, comparison of results between the two experiments should be done with caution as fish were exposed to a lower dose of worms and a lower prevalence of infection occurred in Experiment 2. Severity of disease is often increased with coinfections by bacteria and helminths. Both M. penetrans and P. tomentosa are closely associated with the epithelium, and we observed a signal for Mycoplasma inside of a worm. Capillarid nematodes and Trichuris spp. are taxonomically related, and both cause chronic penetrating infections of the gastrointestinal epithelium. Infections by Trichuris suis in pigs have been associated with an increase in Campylobacter bacteria,⁴⁵ and T. suis infections enhance the severity of infections by Campylobacter jejuni and associated lesions.⁴⁶ Interestingly, among the pathological changes that increased with pigs with coinfections by the whipworm and the bacterium, hyperplasia of the epithelium was particularly more severe,⁴⁵ and hyperplasia is a preneoplastic lesion that is consistently associated with the intestinal tumors in zebrafish.^4,11

Future experiments are planned to disentangle the relationships between Mycoplasma spp., specifically Mycoplasma E and related isolates, the nematode, and the intestinal cancers. For example, exposure to various strains of M. penetrans should be repeated as perhaps some isolates within this clade are more carcinogenic, with or without the worm as a promoter. Experimental fish should be held for about a year to reflect that time of tumor development observed in zebrafish laboratories⁴ and in vivo experiments.¹¹ Moreover, we cannot exclude the possibly at this time that the neoplasms are not caused by multiple agents in which the pathologic presentations are indistinguishable. For example, we described morphologically similar intestinal neoplasms in transgenic zebrafish that overexpressed the H. pylori virulence factor CagA that were homozygous for a loss-of-function allele of p53.⁹ We, therefore, plan to investigate candidate oncogenic viruses and other bacteria as possible causes of these transmissible neoplasms.

Disclosure Statement

No competing financial interests exist.

Funding Information

This study was supported, in part, by NIH ORIP 5 R24 OD010998 (M.L.K., T.J.S.), NIH NIAID R21 AI135641 (T.J.S., M.L.K.), and NIH R01CA176579 (K.G., M.L.K.). National Institute of Environmental Health Sciences supported Environmental Health Sciences Center P30ES000210 by providing zebrafish.