induction of branchial (gill) neoplasms in the medaka fish ... · the japanese medaka (oryzias...

7
[CANCER RESEARCH 45, 3209-3214, July 1985] Induction of Branchial (Gill) Neoplasms in the Medaka Fish (Oryzias latipes) by /V-Methyl-AT-nitro-A/-nitrosoguanidine Mavis R. Brittelli, Hans H. C. Chen,1 and Carl F. Muska Haskell Laboratory for Toxicology and Industrial Medicine, E. l. du Pont de Nemours and Company, Inc., Newark, Delaware 19711 ABSTRACT Juvenile medaka were exposed to A/-methyl-A/'-nitro-A/-nitro- soguanidine in water under static renewal conditions for 28 days. Two groups of 134 fish each were pulsed 3 times weekly at nominal concentrations of 1.0 and 0.5 mg/liter with /V-methyl-AT- nitro-A/-nitrosoguanidine dissolved in dimethylformamide. A third group of 134 fish was exposed to the solvent control, 0.01% dimethylformamide in water. Following the 28-day exposure, and during the recovery period, fish were sampled at intervals of approximately 0, 3, 6, and 9 months and examined grossly. Selected tissues were evaluated microscopically. Many tumor types developed in both A/-methyl-/V'-nitro-/v- nitrosoguanidine exposure groups, but only the gill lesions will be discussed. Approximately 50% of the fish in both treatment groups died from gill damage in the second to third month of the recovery period. More than 90% of the surviving treated fish displayed gill lesions, which progressed from mild epithelial hy- perplasia of gill filaments at 0-months recovery to epithelioma- tous hyperplasia at 3 months and advanced to a more focal nodular appearance of gill filaments at 6 months. Eight to 9 months after the treatment period, at least four fish displayed branchial blastomas. The control fish had no gill lesions. Chemi cally induced gill tumors have not been previously observed in fish. Even gill tumors of unknown origin are very rare. INTRODUCTION Natural fish populations have been found to have tumors of almost all tissue systems, but most of these tumors are of unknown etiology (7). A recent review of Hendricks (9) discusses the use of aquatic models in cancer research. Most studies with fish have been conducted with chemicals that induce liver cancer in mammals. Although experimentally induced neoplasms or hyperplasias in fish are primarily observed in the liver, abnormal ities observed in kidney, spinal cord, testis, thyroid, intestine, and skin are also noted. These abnormalities are induced by such agents as testosterone, aflatoxin, /3-aminopropionitrile, se same oil, 2-acetylaminofluorene, and nitrosamines, not all of which are classic carcinogens in mammals (7, 9,10,13,17). Tumors of the respiratory system in fish have been extremely rare. To date, one malignant gill tumor has been reported in an Amazonian fish (Cichlidae, Chaetobranchus semifasciatus) which was thought to be related to metacercaria of trematodes (20), and 3 spontaneous neoplasms have been reported in rainbow trout (8, 18). Many cases of gill changes, such as hyperplasia and lamellae distended with blood, have been reported, however, and these are related to physical and chemical agents in salmo- nids, carp, bluegills, and the white sucker (3, 4, 6,15). 1To whom requests for reprints should be addressed. Received 1/9/84; revised 2/22/85; accepted 4/9/85. Fish use gills (branchia) for respiration and for the excretion of salt and nitrogenous wastes. The gills of a typical teleost com prise 2 sets of 4 holobranchs; each set forms one side of the pharynx and is covered by an operculum. Each holobranch consists of 2 hemibranchs projecting from the posterior edge of the branchial (gill) arch. The hemibranchs consist of a row of long thin filaments which project from the gill arch like the teeth of a comb. The respiratory unit in fish is the gill filaments that contain a border of long respiratory lamellae. The lamella is simply an extension of specialized capillaries from the efferent branchial arteriole that is covered by simple, flattened, respiratory epithe lium. In addition to the covering flattened epithelium and capillar ies, pillar cells and eosinophilicly stained chloride cells are also found in the lamellae. The pillar cell is a supporting cell that is arranged in rows 9 to 10 Mm apart lying between capillaries in the lamellae. The chloride cell is a highly specialized excretory cell that regulates electrolyte balance of fish and is mainly ob served at the roots of lamellae, i.e., gill filament-lamellar junction (18). In order to have a truly versatile fish model to study carcino- genesis, the ability to chemically induce branchial neoplasms is very important. MNNG2 seemed to be a likely candidate, because it is a carcinogen in mammals, it is direct acting, and it induces neoplasms at the site of administration; i.e., administered in drinking water, it induces malignant tumors of the glandular stomach in rats (11). Hendricks ef al. (10) exposed rainbow trout embryos to MNNG, which caused a high incidence of hepato- cellular carcinomas and a lower but significant incidence of nephroblastomas. The Japanese medaka (Oryzias latipes) has been used as a tool in biological research for many years but not until recently in cancer research (2, 12). The medaka is a small freshwater killifish indigenous to parts of Japan, Taiwan, and southeastern Asia, where it is common in rice paddies. Adult medakas are 2 to 4 cm long, very hardy, and easily maintained at ambient temperatures, having a life span of 4 or more years (12,14, 21). The purpose of our overall program was to evaluate the feasibility of using fish as models in cancer studies. However, most previous studies (1, 7, 9, 17, 19) were isolated ones that used a single species of fish and a single carcinogen. Our program was designed to study the effects of 2 carcinogens on 3 species of fish: channel catfish; rainbow trout; and medaka (2). We wanted to compare species with respect to tumor develop ment, progression of tumor development, time to tumor devel opment, and target organs in order to identify a model system for further evaluation. In this paper, a study of the effects of MNNG on the respiratory system of the medaka is presented. 2The abbreviations used are: MNNG, A/-methyl-A/'-nitro-N-nitrosoguanidine; DMF, dimethylformamide. CANCER RESEARCH VOL. 45 JULY 1985 3209 on March 22, 2021. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Upload: others

Post on 17-Oct-2020

3 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Induction of Branchial (Gill) Neoplasms in the Medaka Fish ... · The Japanese medaka (Oryzias latipes) has been used as a tool in biological research for many years but not until

[CANCER RESEARCH 45, 3209-3214, July 1985]

Induction of Branchial (Gill) Neoplasms in the Medaka Fish (Oryzias latipes) by/V-Methyl-AT-nitro-A/-nitrosoguanidine

Mavis R. Brittelli, Hans H. C. Chen,1 and Carl F. Muska

Haskell Laboratory for Toxicology and Industrial Medicine, E. l. du Pont de Nemours and Company, Inc., Newark, Delaware 19711

ABSTRACT

Juvenile medaka were exposed to A/-methyl-A/'-nitro-A/-nitro-

soguanidine in water under static renewal conditions for 28 days.Two groups of 134 fish each were pulsed 3 times weekly atnominal concentrations of 1.0 and 0.5 mg/liter with /V-methyl-AT-nitro-A/-nitrosoguanidine dissolved in dimethylformamide. A third

group of 134 fish was exposed to the solvent control, 0.01%dimethylformamide in water. Following the 28-day exposure, and

during the recovery period, fish were sampled at intervals ofapproximately 0, 3, 6, and 9 months and examined grossly.Selected tissues were evaluated microscopically.

Many tumor types developed in both A/-methyl-/V'-nitro-/v-

nitrosoguanidine exposure groups, but only the gill lesions willbe discussed. Approximately 50% of the fish in both treatmentgroups died from gill damage in the second to third month of therecovery period. More than 90% of the surviving treated fishdisplayed gill lesions, which progressed from mild epithelial hy-perplasia of gill filaments at 0-months recovery to epithelioma-

tous hyperplasia at 3 months and advanced to a more focalnodular appearance of gill filaments at 6 months. Eight to 9months after the treatment period, at least four fish displayedbranchial blastomas. The control fish had no gill lesions. Chemically induced gill tumors have not been previously observed infish. Even gill tumors of unknown origin are very rare.

INTRODUCTION

Natural fish populations have been found to have tumors ofalmost all tissue systems, but most of these tumors are ofunknown etiology (7). A recent review of Hendricks (9) discussesthe use of aquatic models in cancer research. Most studies withfish have been conducted with chemicals that induce liver cancerin mammals. Although experimentally induced neoplasms orhyperplasias in fish are primarily observed in the liver, abnormalities observed in kidney, spinal cord, testis, thyroid, intestine,and skin are also noted. These abnormalities are induced bysuch agents as testosterone, aflatoxin, /3-aminopropionitrile, sesame oil, 2-acetylaminofluorene, and nitrosamines, not all of

which are classic carcinogens in mammals (7, 9,10,13,17).Tumors of the respiratory system in fish have been extremely

rare. To date, one malignant gill tumor has been reported in anAmazonian fish (Cichlidae, Chaetobranchus semifasciatus) whichwas thought to be related to metacercaria of trematodes (20),and 3 spontaneous neoplasms have been reported in rainbowtrout (8, 18). Many cases of gill changes, such as hyperplasiaand lamellae distended with blood, have been reported, however,and these are related to physical and chemical agents in salmo-

nids, carp, bluegills, and the white sucker (3, 4, 6,15).

1To whom requests for reprints should be addressed.

Received 1/9/84; revised 2/22/85; accepted 4/9/85.

Fish use gills (branchia) for respiration and for the excretion ofsalt and nitrogenous wastes. The gills of a typical teleost comprise 2 sets of 4 holobranchs; each set forms one side of thepharynx and is covered by an operculum. Each holobranchconsists of 2 hemibranchs projecting from the posterior edge ofthe branchial (gill) arch. The hemibranchs consist of a row of longthin filaments which project from the gill arch like the teeth of acomb. The respiratory unit in fish is the gill filaments that containa border of long respiratory lamellae. The lamella is simply anextension of specialized capillaries from the efferent branchialarteriole that is covered by simple, flattened, respiratory epithelium. In addition to the covering flattened epithelium and capillaries, pillar cells and eosinophilicly stained chloride cells are alsofound in the lamellae. The pillar cell is a supporting cell that isarranged in rows 9 to 10 Mm apart lying between capillaries inthe lamellae. The chloride cell is a highly specialized excretorycell that regulates electrolyte balance of fish and is mainly observed at the roots of lamellae, i.e., gill filament-lamellar junction

(18).In order to have a truly versatile fish model to study carcino-

genesis, the ability to chemically induce branchial neoplasms isvery important. MNNG2 seemed to be a likely candidate, because

it is a carcinogen in mammals, it is direct acting, and it inducesneoplasms at the site of administration; i.e., administered indrinking water, it induces malignant tumors of the glandularstomach in rats (11). Hendricks ef al. (10) exposed rainbow troutembryos to MNNG, which caused a high incidence of hepato-

cellular carcinomas and a lower but significant incidence ofnephroblastomas.

The Japanese medaka (Oryzias latipes) has been used as atool in biological research for many years but not until recentlyin cancer research (2, 12). The medaka is a small freshwaterkillifish indigenous to parts of Japan, Taiwan, and southeasternAsia, where it is common in rice paddies. Adult medakas are 2to 4 cm long, very hardy, and easily maintained at ambienttemperatures, having a life span of 4 or more years (12,14, 21).

The purpose of our overall program was to evaluate the

feasibility of using fish as models in cancer studies. However,most previous studies (1, 7, 9, 17, 19) were isolated ones thatused a single species of fish and a single carcinogen. Ourprogram was designed to study the effects of 2 carcinogens on3 species of fish: channel catfish; rainbow trout; and medaka (2).We wanted to compare species with respect to tumor development, progression of tumor development, time to tumor development, and target organs in order to identify a model systemfor further evaluation. In this paper, a study of the effects ofMNNG on the respiratory system of the medaka is presented.

2The abbreviations used are: MNNG, A/-methyl-A/'-nitro-N-nitrosoguanidine;

DMF, dimethylformamide.

CANCER RESEARCH VOL. 45 JULY 1985

3209

on March 22, 2021. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

Page 2: Induction of Branchial (Gill) Neoplasms in the Medaka Fish ... · The Japanese medaka (Oryzias latipes) has been used as a tool in biological research for many years but not until

GILL NEOPLASMS IN MEDAKA FISH INDUCED BY MNNG

MATERIALS AND METHODS

Fish. Golden strain medakas were bred at the Haskell Laboratory.Brood stock originated from the Carolina Biological Supply Co., Burlington, NC. Brood stock and embryos were maintained at 25°Cin aerated,

filtered well water. Newly hatched fry were fed 3 times daily. The dietconsisted of a combination diet for 1 week [Liquifry No. 1 for egg layers,Tetra Min Baby Fish Food "E" for egg layers, and freshly hatched San

Francisco type brine shrimp (Artemia salina)], which was gradually re

placed with freshly hatched brine shrimp nauplii. Fish were 4 to 5 weeks

old when the study began.Water and Water Quality. Water, from a 350-ft well cased to bedrock

to prevent surface water contamination, was aerated and filtered througha 10-Mm filter to remove particulates and then distributed to each

aquarium through aged Polyvinylchloride pipe. Flow rate was at least 6aquaria volumes per 24 h. Back-up air stones were used as a safety

feature in the event of a power failure. Water quality parameters includingpH, Do, hardness, conductivity, and alkalinity were measured at leastmonthly. These values have been consistent and acceptable for aquaticlife. Ammonia and nitrite concentrations were monitored sporadicallyprior to and during the test period. The concentration of total ammoniawas approximately 0.2 ppm at pH 7.2 to 7.8. The concentration of nitritewas barely detectable. More recent routine analyses of the well waterhave confirmed these low concentrations. (Additionally, throughout thestudy period, at least 7 other groups of fish were held under these

conditions and developed no gill lesions.)Chemicals. MNNG was obtained from Sigma Chemical Co., St. Louis,

MO. DMF was certified ACS grade. Tricainemethanesulfonate was obtained from Crescent Research Chemicals, Inc., Paradise Valley, AZ. Allchemicals used to prepare fixatives were reagent grade.

Carcinogen Exposures. Exposures were conducted in a chemicalfume hood with an average face velocity of 2.8 eu m/min. Because thehood temperature was 21 ±1°C,the fish were gradually acclimatized

from 25°to 21 °Cprior to initiation of the study. Three groups of 134

four- to five-week-old medakas were placed in three 19-liter aquaria,

each containing 16 liters of aerated water. The aquaria waters werechanged 3 times/week. After changing, the aquaria were pulsed witheither DMF at 0.1 ml/liter (control), or MNNG at 0.5 mg/liter and DMF at0.1 ml/liter (low concentration), or MNNG at 1.0 mg/liter and DMF at 0.1ml/liter (high concentration). MNNG is quite unstable in water with a half-

life of h (8), so the dry compound was stored frozen and dissolved in

DMF immediately prior to use. The fish were fed freshly hatched brineshrimp nauplii twice daily during the exposure period. After 28 days,each group was rinsed thoroughly with clean water and placed intoseparate 57-liter flow-through systems. The fish were maintained in

these recovery systems and fed brine shrimp nauplii for up to 9 months.Sampling. Fish were sampled at approximately 0, 3, 6, and 9 months

into recovery. However, many fish were found dead or were sacrificed"in extremis" between scheduled sacrifices. These fish were processed

for histological examination and included in the results if the tissues werenot autolyzed. At each sampling interval or unscheduled sacrifice, thefish were anesthetized with tricainemethanesulfonate, if necessary, andexamined under a Wild M5A stereomicroscope at x6 to x50 with the

aid of microdissecting tools. The animals were examined externally, andthe abdominal cavity was carefully opened to examine the internal organswithout damaging tissues or organs.

Histological Studies. The entire fish was placed in Bouin's or for-

malin:alcohol (ethanol):acetic acid (2:18:1) (16) fixative for subsequenthistomorphological evaluation. Generally, horizontal sections were made

through the whole animal. However, if lesions were apparent from grossobservations, sagittal sections and especially transverse sections weremade through the gill region. Sections were stained with hematoxy-

lin:eosin and other stains as necessary.

RESULTS

During the 28-day exposure period, no overt signs of toxicity

to MNNG were observed. Gills of the animals sacrificed at 28days (0-day recovery) had no gross changes. However, microscopically, the 5 fish sampled from each MNNG-treated grouphad very subtle epithelial hyperplasia in filament-lamellae compared to the control fish gills (Figs. 1 to 3). At the filament-

lamellar junction, i.e., along the length of gill filaments, therewere minute clumps of hyperplastic and hypertrophie squamousepithelial cells. These cells were hyperchromatic with large nucleibearing a prominent nucleolus. Hyperplastic change was moreobvious around the tips of the filaments (filament knobs) (Fig. 3).

After approximately 1 to 4 months of the recovery period,animals in both treatment groups began to die, and many couldnot be examined histologically because the tissues autolyzed.Tissue autolysis seems to occur rapidly, because the fish wereobserved twice daily, and any dead or dying fish were removedand examined. Of 268 fish in both treatment groups, only 97were suitable for histological evaluation. Control mortality wasvery low for the entire 9-month period (one of 134 fish died).

Gross observations of fish found dead or killed in extremis andmost of those sampled at 3 months of recovery were similar.The most obvious findings were: congested liver with distendedgall bladder; enlarged hearts; and pale, almost white edematousgills. The cause of death appeared to be respiratory and/orcirculatory failure. The periphery of many swollen gills was hy-

pertrophied and appeared to be fused together, leading to asomewhat consolidated appearance of holobranchs (Fig. 4).Many red foci and white nodules with or without red spots werealso observed in the consolidated holobranchs or enlarged gillfilaments (Fig. 5). Red nodules were, in many cases, visiblethrough the opercular flaps while the animals were swimming.(The red foci nodules ranged in size from pinpoints to as largeas 1 mm or more; the average length and weight of the test fishwere 2.4 cm and 0.2 g.)

Microscopically, the gill filaments and the lamellae were swollen, thickened, congested, and edematous. The periphery of gillfilaments was hyperplastic and hypertrophie and had lost theirnormal individual filamentous character (Fig. 5). They were frequently fused together and were composed of proliferated cu-

boidal or squamous cells. The red foci and white nodules withred spots observed grossly had a somewhat unique appearancethat was marked by epitheliomatously proliferated tissue surrounding a malformed vascular core or telangiectatic hematocyst(Fig. 6). The telangiectatic hematocysts were characterized by acollection of irregularly formed vascular structures filled withblood or were composed of markedly dilated and tortuouslyconnected capillaries. The peripheral epitheliomatous tissueswere characterized by sheets of uniformly sized cuboidal orsquamous cells (Figs. 7 and 8). The cytoplasm was latticed andvesicular with centrally located nuclei containing a single prominent nucleolus. Necrosis of the individual cells was frequentlyobserved. The histological designation of these greatly proliferated gills was epitheliomatous hyperplastic nodules.

In addition to branchial epitheliomatous hyperplastic nodulesobserved in these treated fish, there was generalized mild tomarked epithelial hyperplasia in the lamellae of the gill filaments(Figs. 3,6,7, and 8). For classification purposes, mild hyperplasiawas designated to be 2 to 5 cell layers or a clump of cells up to

CANCER RESEARCH VOL. 45 JULY 1985

3210

on March 22, 2021. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

Page 3: Induction of Branchial (Gill) Neoplasms in the Medaka Fish ... · The Japanese medaka (Oryzias latipes) has been used as a tool in biological research for many years but not until

GILL NEOPLASMS IN MEDAKA FISH INDUCED BY MNNG

5 on the lamellae or up to 10 around the filament knobs (Fig. 3).Moderate hyperplasia was 5 or more cell layers or a clump ofhyperplastic cells greater than that observed in mild hyperplasia.Marked hyperplasia was similar to moderate, but more lamellaein the same filament were involved. Branchial epitheliomatoushyperplastic nodules described above were similar to a markedhyperplasia, but the hyperplastic change was extensive. Thenormal individual configuration of lamellae and filaments waslost. They were fused and covered by massively proliferatedsquamous or cuboidal epithelial cells. The proliferated cellularsheet filled the interfilamentous space, especially at the peripheral margin of the hypertrophie filaments which fused togetherforming a mass of epithelial sheets. The involvement not onlyoccurred in gill filaments in the same hemibranchia (intrahemi-branchial fusion of filament) but in filaments of 2 or more hemi-

branchs which caused fusion that could be observed easily atgross examination (Figs. 5, 7, and 8).

It was very difficult to set clear-cut criteria to differentiate

between gill hyperplasia and gill adenomatous nodules. This wasespecially true in the early days of the experiment, because thegills of young medaka are small and short. The plane of sectioning also makes histológica! interpretation difficult. However, gilltumors were classified based not only on the histological features, but on gross evidence of tumor-like enlargement of the

gills (Fig. 9).After 3 to 5 months of recovery, most fish showed either

marked gill filament epithelial hyperplasia or branchial epitheliomatous nodules (Table 2). Gross observation of these fish revealedfurther loss of normal gill structure, although gill color was nearnormal. Most gill filaments were indistinguishable because offusion. Large distinct clear nodules were common, presenting alumpy, irregular texture in a normally very uniform organ (Figs. 4and 5). (The clear nodules were as large as 3 mm; the averagelength and weight of the test fish were 2.6 cm and 0.25 g.)Histological appearance of the clear nodules was comparable tothose previously described.

After 8 to 9 months of recovery, the branchial epitheliomatoushyperplastic nodules began to have a malignant pattern (Table2). Grossly and histologically, the nodules were much larger(Figs. 9 and 10). Four fish had malignant tumors, 3 in the low-concentration exposure group and one in the high-concentration

exposure group. These tumors were composed of poorly differentiated anaplastic basophilic cells arranged in columns or cord-

like structures. The neoplastic cell cords proliferated randomlyand invaded adjacent tissue. These cells resembled more primitive stem or basilar cells and appeared to be originating from thesquamous epithelium along the gill filament, especially aroundthe filament knob; mitotic figures were common in these cordaland columnar neoplastic cells. Occasionally, trapped chondroidtissue remnants were found in the neoplastic tissue (Figs. 11and 12).

At the end of 9 months of recovery, less than 20 test fishremained alive, so the study was terminated.

A summary of gill lesions found in medaka is shown on Table1. Table 2 is a summary of gill epithelial lesions versus time.Seventy-nine control fish have been examined histologically to

date. None had gill lesions or any other remarkable lesions. Theremaining 50 fish are being held to maturity to examine forspontaneous tumors.

Gill sections from control and treated fish were examined for

Table 1

Summary of medaka gill lesions

No lesions were present in control fish gills. None of the MNNG-treated fish waswithout lesions.

MNNG exposure concentration(mg/liter)

HistologicallesionNone

Mild epithelial hyperplasiaModerate epithelial hyperplasiaMarked epithelial hyperplasiaTelangiectatic hematocystOther*

Epitheliomatous hyperplasiaBranchial blastoma0a73/79

0/730/730/730/730/730/730/730.50/50

10/5010/5013/5016/503/50

16/50C

3/501.00/47

13/4711/4715/4711/4716/474/471/47

Six of 79 animals processed had no cut through the gills; the remaining fishare still in recovery.

6 Crooked filament tips, stunted and misshapen filaments, fused lamellae, and

chloride cell proliferation.c One animal had both epitheliomatous hyperplasia and branchial blastoma.

Table 2G/7/lesions in MNNG-treated medaka versus time (test groups pooled; control

animals not included)Gill epithelial lesions progressed from mild hyperplasia at 0 to 2 months post-

exposure to moderate or marked hyperplasia at 3 to 9 months postexposure.Benign or malignant tumors were absent at 0 to 2 months but were present in 23and 30% of the animals by 3 to 5 and 5 to 9 months.

% of gill lesions at the followingrecovery times

LesionMild

epithelial hyperplasiaModerate epithelial hyperplasiaMarked epithelial hyperplasiaEpitheliomatous hyperplastic nodule or

branchial blastoma0-2

mo100

0003-5

mo14

2339235-9

mo11

352430

No. of fish8 13 43 37

The 97 animals either did not all have epithelial lesions or the lesions were notvisible in the section; consequently, only 93 animals appear on this table.

infectious agents. In 4 fish, including one control, a few longslender nonbranching filamentous fungal microorganisms, 1 ^min thickness, were found. Also rarely, protozoal cysts, the cystsfilled with basophilic granules and measuring 25 to 40 u.rr\ indiameter, were observed in a few control and treated fish. Wefeel these agents were not related to the gill alterations.

DISCUSSION

In this study, the development of gill tumors in the medakaappears to begin as a response to an irritant with epithelialhyperplasia, vasodilation of arterioles and capillaries, fusion oflamellae, and clubbing of filaments. This phenomenon was observed as early as 1910 by Osbum, as cited by Eller (3), and hasbeen associated with both physical and chemical agents andmicroorganisms. This phenomenon can be reversible, such as inthe case of exposure to the herbicide endothal (Hydrothol 191),the fish acclimated after 14 days, and tissue alterations reversedand gradually returned to normal (5).

A similar response has been observed from bacterial, fungal,and protozoal infections (3). In many cases, the fish will die ofother disease effects before the gills are severely enough damaged to cause death from anoxia. While it is difficult to predictthe number of survivors of these infections, it is probable thatthe survivors do not develop gill tumors, since gill tumors are sorare. Four recent reports of spontaneous gill neoplasms (8, 18,20) and these data are the first reports of gill tumors in fish.

CANCER RESEARCH VOL. 45 JULY 1985

3211

on March 22, 2021. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

Page 4: Induction of Branchial (Gill) Neoplasms in the Medaka Fish ... · The Japanese medaka (Oryzias latipes) has been used as a tool in biological research for many years but not until

GILL NEOPLASMS IN MEDAKA FISH INDUCED BY MNNG

The delayed mortality we observed in this study, at least forthe first 3 to 4 months, was a result of chemical insult or irritationwhich the fish could not overcome. They appeared to have diedof circulatory failure and/or anoxia directly related to respiratoryinsufficiency.

Later in the study, mortalities occurred, or fish were sacrificedin extremis between scheduled sacrifices. While this again is aresult of respiratory insufficiency, the actual lesions were nolonger those typically caused by an irritant. MNNG-induced gill

lesions were not reversible but had progressed to much moremarked epithelial hyperplasia and fusion of filaments. In somecases, benign branchial tumors were present by 3 to 5 months,but in most cases, the lesions did not progress to that point until5 to 9 months into recovery (Table 2). It is possible that, if theanimals had been held longer, more gill tumors would have beenobserved.

As stated previously, MNNG is primarily a local-acting carcin

ogen in mammals. In fish, the response was also local in nature.It is noteworthy that we observed several other tumor types inmedakas exposed to MNNG (or MNNG in combination withDMF), all of which can be considered a response to a local-actingcarcinogen. These include thyroid follicular hyperplasia, adenom-atous goiter and follicular carcinoma,3 and fibrosarcoma of soft

tissue on the underside of the opercular (gill) flap. All of these,along with the gill tumors, are tumors of tissues that were bathedin MNNG during the aqueous exposures.

Gill tumors appear to be exceedingly rare, which is not unexpected, when the role of the gill is considered. Most chemicalsand known carcinogens, other than local-acting alkylating

agents, are unlikely to induce gill tumors. If they are irritating,many will cause gill lesions, but these will either kill the fish orheal after the insult has been removed. In this case, either scartissue may form, or complete healing may occur.

The medaka is sensitive to MNNG, time to tumor developmentis short, and several types of tumors are observed. The fact thatgill tumors are induced by MNNG (or MNNG in combination with

3H. C. Chen, M. R. Britlelli, and C. F. Muska. Thyroid hyperplasia and neoplasia

in medaka fish induced by MNNG, manuscript in preparation.

DMF) increases the versatility and, consequently, the usefulnessof the fish as a model for cancer studies.

REFERENCES

1. Aoki, K., and Matsudaira, H. Induction of hepatic tumors in a teleost (Oryzias(atipes) after treatment with methylazoxymethanol acetate. J. Nati. Cancerlnst.,59: 1747-1749,1977.

2. Chen, H. C., Brittelli, M. R., and Muska, C. F. Evaluation of tumors in medakafish induced by diethylnitrosamine or W-tnethyl-W'-nitro-W-nitrosoguanidine.

The Toxicologist, 3: 145 (abst. No. 580), 1983.3. Eller, L. L. Gill lesions in freshwater teleosts. In: W. E. Ribelin and G. Migaki

(eds.), The Pathology of Fishes, pp. 305-330. Madison, Wl: The University ofWisconsin Press, 1975.

4. Eiter, L. L. Histopathologic lesions in cutthroat trout (Salmo dark!) exposedchronically to the insecticide endrin. Am. J. Pathol., 64: 321-336,1971.

5. Eller, L. L. Pathology in redear sunfish exposed to hydrothol 191. Trans. Am.Fish. Soc., 98: 52-59,1969.

6. Harshbarger, J. C. Activities Report: Registry of Tumors in Lower Animals:1978 Supplement. Washington, DC: Smithsonian Institution, 1979.

7. Harshbarger, J. C. Role of the registry of tumors in lower animals in the studyof environmental carcinogenesis in aquatic animals. Ann. NY Acad. Sci., 298:280-289,1977.

8. Harshbarger, J. C. Lung tumors in nondomesticated animals. In: H. M. Reznick-Schulter (ed.), Comparative Respiratory Tract Carcinogenesis, Vol. 1, Chap.10, pp. 219-231. Boca Raton, FL: CRC Press, Inc., 1983.

9. Hendricks, J. D. Chemical carcinogenesis in fish. In: L. J. Weber (ed.), AquaticToxicology, Vol. 1, pp. 149-211. New York: Raven Press, 1982.

10. Hendricks, J. D., Scanlan, R. A., Williams, J. L., Sinnhuber, R. 0., and Grieco,M. P. Carcinogenicity of N-methyl-N'-nitrc-W-nitrosoguanidine to the livers and

kidneys of rainbow trout (Salmo gairdneri) exposed as embryos. J. Nati. CancerInst., 64:1511-1519,1980.

11. lARC Monographs on the Evaluation of Carcinogenic Risk of Chemicals toHumans, Vol. 4, pp. 183-195. Lyons, France: International Agency for Research on Cancer, 1974.

12. Ishikawa, T., Shimamine, T., and Takayama, S. Histologie and electron microscopy observations of diethylnitrosamine-induced hepatomas in smallaquarium fish (Oryzias ¡atipes).J. Nati. Cancer Inst., 55: 909-916,1975.

13. Khudoley, V. V. Spontaneous and induced neoplasms of aquarium fish. Usp.Sovrem. Biol., 74: 473-481, 1972.

14. Kirchen, R. V., and West, W. R. The Japanese Medaka: Its Care and Development. Burlington, NC: Carolina Biological Supply Company, 1976.

15. Kumaraguru, A. K., Beamish, F. W. H., and Ferguson, H. W. Direct andcirculatory paths of permethrin (NRDC-143) causing histopathological changesin the gills of rainbow trout, Salmo gairdneri. J. Fish Biol., 20: 87-92,1982.

16. Luna, L. G. (ed.), Manual of Histologie Staining Methods of the Armed ForcesInstitute of Pathology, Ed. 3, pp. 4-5. New York: McGraw-Hill, 1968.

17. Matoushima. T., and Sugimara, T. Experimental carcinogenesis in small aquarium fishes. Prog. Exp. Tumor. Res.. 20: 367-379,1976.

18. Roberts, R. J. Fish Pathology. London: Balilliere Tindall, 1978.19. Stanton, M. F. Diethylnitrosamine-induced hepatic degeneration and neoplasia

in the aquarium fish, Brachydanio rerio. J. Nati. Cancer Inst., 34: 117-123,1965.

20. Thatcher, V. E., and Varella, A. B. A malignant tumor of the gills related tometacercaria of a trematode. Acta Amazónica, JO: 651-656,1980.

21. Yamamoto, T. Medaka (Killifish) Biology and Strains. Tokyo: Keigaku Publishing Company, 1975.

Fig. 1. Histological section of gill hemibranch in control medaka. 8-month recovery. Note: individual gill arches (A); filaments (B); and filament knobs (C). H & E, x 50.

Fig. 2. Histological section of gill filaments in control medaka, 0-day recovery. Note: filaments (A); filament blood vessel (B): chondroid tissue (C); lamellae with outercovering of epithelial cells and inner lamellar capillary (D). H & E, x 320.

Fig. 3. Histological section of MNNG-treated medaka gills, 0-day recovery. Note: mild epithelial hyperplasia (A): filaments (B); filament blood vessel (C); lamella withouter covering of epithelial cells and inner lamellar capillary (D). H & E, x 320.

Fig. 4. Gross appearance of gills in control (left) and MNNG-treated (right) medaka. Note: opercular flaps removed. Eyes (A); abdomens (8); control gill showingstraight reddish filaments (C); treated gills showing pale, fused, swollen, and edematous filaments (D) with red foci (£).x 6.

Fig. 5. Gross appearance of gill in MNNG-treated medaka. Note: opercular flap is raised with pin (A) to expose the left gill. Nodular appearance is evident: red nodule(B); clear nodule (C); fused filaments (D); abdomen (E), x 12.

Fig. 6. Histological section of gill in MNNG-treated medaka. Note: section through a red nodule in the gill. Hyperplastic epithelium (A) with inner hematocyst (8). H &

E, x 125.Fig. 7. Histological section of gill in MNNG-treated medaka. Note: marked epithelial hyperplasia (A) with intraholobranchial fusion of lamellae (B); gill arch (C). Masson's

trichrome stain, x 51.Fig. 8. Histological section of gills in MNNG-treated medaka. Note: marked hyperplasia causing intra- and interhemibranchial fusion of gill filaments. Chondroid tissue

remnant (A). This is an enlargement of Fig. 7. H & E, x 320.Fig. 9. Gross appearance of gill in MNNG-treated medaka branchial (gill) tumor (A) pushing out the right opercular flap (B). Also shown are right eye (C) and abdomen

(D). x 12.Fig. 10. Histological section of gill in MNNG-treated medaka. Note: branchial blastoma. This photomicrograph is the tumor in Fig. 9. H & E, x 51.

Fig. 11. Higher magnification of Fig. 9. Note: branchial blastoma showing the anaplastic character of the tumor and remnants of chondroid tissue (A). H & E, x 320.Fig. 12. Histological section of MNNG-treated medaka branchial (gill) tumor. Note: cord-like structures throughout and remnants of chondroid tissue (A). This is a

different animal than shown in Figs. 9 to 11. H & E, x 320.

CANCER RESEARCH VOL. 45 JULY 1985

3212

on March 22, 2021. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

Page 5: Induction of Branchial (Gill) Neoplasms in the Medaka Fish ... · The Japanese medaka (Oryzias latipes) has been used as a tool in biological research for many years but not until

GILL NEOPLASMS IN MEDAKA FISH INDUCED BY MNNG

V> ?j»w*r ••'•!»jKtS

v 'Aifiífcj*?*. '-• VAÌ.XÌ..-;''.'^.^ r?A.V> -• *-.- f

CANCER RESEARCH VOL. 45 JULY 1985

3213

on March 22, 2021. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

Page 6: Induction of Branchial (Gill) Neoplasms in the Medaka Fish ... · The Japanese medaka (Oryzias latipes) has been used as a tool in biological research for many years but not until

GILL NEOPLASMS IN MEDAKA FISH INDUCED BY MNNG

. - ;•v

CANCER RESEARCH VOL. 45 JULY 1985

3214

on March 22, 2021. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

Page 7: Induction of Branchial (Gill) Neoplasms in the Medaka Fish ... · The Japanese medaka (Oryzias latipes) has been used as a tool in biological research for many years but not until

1985;45:3209-3214. Cancer Res   Mavis R. Brittelli, Hans H. C. Chen and Carl F. Muska 

-nitrosoguanidineN-nitro-′N-Methyl-N) by Oryzias latipesInduction of Branchial (Gill) Neoplasms in the Medaka Fish (

  Updated version

  http://cancerres.aacrjournals.org/content/45/7/3209

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  [email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/45/7/3209To request permission to re-use all or part of this article, use this link

on March 22, 2021. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from