Interstitial orchitis with impaired steroidogenesis and spermatogenesisin the testes of rabbits infected with an attenuated strain of myxoma

virusS. Fountain, M. K. Holland, L. A. Hinds, P. A. Janssens

and P. J. KerrCooperative Research Centre for Biological Control of Vertebrate Pest Populations. CSIRO Division ofWildlife and Ecology, PO Box 84, Lyneham, ACT 2602, Australia; and 2Division of Biochemistry and

Molecular Biology, Australian National University, ACT, Australia

The pathogenesis of infertility associated with acute viral orchitis was investigated in a

rabbit model. Infection of postpubertal male European rabbits with an attenuated strain ofmyxoma virus caused systemic disease with viral replication to high titres in the testes.Infected rabbits developed an interstitial orchitis and epididymitis. This was associated withdegeneration of the seminiferous epithelium, decreased serum testosterone concentrationsand increased serum LH concentrations. Virus was localized within the interstitial cells.Clearance of infectious virus from the testis occurred between day 20 and day 30 afterinfection, although viral DNA could still be detected by polymerase chain reaction at 120days. Viral clearance was associated with a return to normal serum testosterone and LHconcentrations. Anti-sperm antibodies were present in the serum as early as 5 days afterinfection and declined during the recovery phase. Rabbits were infertile at 60 days butreturned to normal fertility 60\p=n-\90days after infection.

Introduction

Viruses that spread to and replicate in the testis are relativelyuncommon (Ladds, 1993). One of the best known examplesis mumps virus infection in humans. In postpubertal malesthis virus replicates in the testis (Bjorvatn, 1973) and can

induce an acute interstitial orchitis. In severe cases, tubularsclerosis, hyalinization and chronic infertility may result (Gall,1947; Charny and Meranze, 1948). A model for studying thepathogenesis of viral orchitis is provided by myxoma virus(MV) in European rabbits. However, there has been nodetailed study of the pathogenesis of this virus in the testis.This is of interest both as a model for virus-induced damageto the testis and also as a means to examine the nature ofthe response to an immunosuppressive virus (McFadden andGraham, 1994) replicating in an immune privileged site suchas the testis (Maddocks and Setchell, 1990). We were

particularly interested in whether long term viral persistencecould occur at this site.

Myxoma virus is a member of the family Poxviridae; genusLeporipoxvirus (Murphy et al, 1995). First isolated in Brazil, theSouth American virus is essentially avirulent for its host the

tapeti (Sylvilagus brasiliensis). However, in European rabbits(Oryctolagus cuniculus) MV causes myxomatosis, an acute andrapidly fatal disease transmitted by biting arthropods (Fennerand Ratcliffe, 1965).

In male rabbits, virulent MV spreads to the testes afterintradermal inoculation and replicates to high titres (Fennerand Woodroofe, 1953). Histologically this is accompanied byan early acute inflammatory response in the interstitium,necrosis of tubular cells and epididymitis (Hurst, 1937). InHurst's study, rabbits died or were killed during the earlyphase of the disease. In another study (Sobey and Turnbull,1956), male domestic rabbits that recovered from infectionwith a moderately attenuated (90% lethal) strain of MV wereinfertile and in some cases showed no sexual interest for 12months or longer. However, the pathogenesis of this infertil¬ity was not investigated. Male wild rabbits also showedevidence of infertility, as judged by the absence of spermato¬zoa in the epididymis, after epidemics of myxomatosis(Poole, 1960), although the impact of male infertility inducedby MV on female fecundity in the wild appears to beminimal (Marshall et al, 1955).

This paper provides an overview of the histological, viro-logical, immunological and endocrinological parameters associ¬ated with infertility in domestic rabbits infected with a highlyattenuated strain of MV, which causes clinical myxomatosis butis generally non-lethal. The use of an attenuated virus alsoenabled the return to fertility after virus clearance from thetestis and epididymis to be studied.

*Present address: Research School of Biological Sciences, Australian NationalUniversity, ACT, Australia.Correspondence.Received S August 1996.

Materials and Methods

Virus stockUriarra/2-53/l (Mykytowycz, 1953) had been plaque puri¬

fied on CV-1 cells (Russell and Robbins, 1989). Working stocksof this virus used for inoculation of rabbits were passaged threetimes in Sire cells. Virus was prepared from infected cellscultured in minimal essential medium with Earle's salts (MEM,Gibco, Gaithersburg, MD) supplemented with 8% newbornbovine serum (NBBS), 2 mmol L-glutamine 1~ , 21 mmolNaHC03 1

"\ 200 units penicillin ml

"

1 and 100 pg strepto¬mycin ml ~

, at 32°C. Virions were released from cells byfreeze-thawing the culture twice. Stocks were titrated byplaque assay on Vero cell monolayers. Virus titres were

expressed as plaque forming units (pfu) g ~

: of tissue or ml~

J

of culture.

Animals and experimental designDomestic European rabbits (Oryctolagus cuniculus) were bred

in the Gungahlin animal facility. Experimental animals were

housed in individual cages in a temperature regulated room at22°C on a 12 h light cycle. Standard rabbit pellets and waterwere available ad libitum and green vegetables were fed as a

supplement.Seventeen mature male rabbits were inoculated intrader-

mally in the right thigh with 100 pfu of MV at day 0. Thesewere allocated randomly to sampling days and two animalswere killed (by intravenous barbiturate) at 5, 10, 15, 20, 30, 60and 90 days after infection. One animal was killed on day 7 andone on day 120. One animal died of myxomatosis at day 18.For virus localization studies, another rabbit was inoculatedwith the same dose of virus and killed 9 days after infection.Control tissues were obtained from uninfected rabbits or

archival tissue blocks.At each time point, including day 0, blood was collected

from all of the surviving animals, including those to be killed,and clinical symptoms of myxomatosis were recorded. On day7, blood samples were collected from only four animals.

All animal experiments were approved by the CSIRODivision of Wildlife and Ecology, Gungahlin, Animal Exper¬imentation Ethics Committee.

Tissue collection, fixation and processingBlood was collected by venepuncture from the marginal ear

vein into heparinized syringes. Plasma was separated andstored at

-

20°C From each rabbit, one testis and one

epididymis were collected aseptically into preweighed vialsand frozen at

—

70°C for analysis of virus titres. Transverseslices through the anterior pole, midsection and posterior poleof the other testis and epididymis were fixed in 10% neutralbuffered formalin, trimmed, dehydrated through a gradedethanol series (70—100%) and cleared in chloroform beforebeing embedded in paraffin. For histological analysis, 5 pm and7 pm sections were mounted and stained with haematoxylinand eosin.

For virus isolation, the testis and epididymis from eachanimal were aseptically minced with scissors and homogenized

in hand held glass homogenizers. Homogenates were freeze—thawed twice to ensure virus release from cells. Virus titreswere measured by plaque assay, in duplicate, on Vero cellmonolayers.

Testosterone assayPlasma testosterone concentrations were assayed using

extracted samples (50 pi in duplicate) and the single antibodyradioimmunoassay procedure (Jones et al, 1988). Intra-assayvariation was 6.5%; interassay variation was 10.5% and thesensitivity of the assay was 0.2 ng ml

~

.

Luteinizing hormone assayPlasma LH concentrations were assayed using a double

antibody radioimmunoassay (Sutherland el al, 1980) validatedfor the domestic European rabbit. Rabbit LH (NIH-AFP7818C)was used as standard and for radioiodination by the chloramine method (Sutherland et al, 1980). The primary antibody was

guinea pig anti-rabbit LH (NIH-AFP-8-1-28). Duplicate 200 pisamples of plasma were assayed. Interassay variation was 12%;intra-assay variation was 6% and the sensitivity of the assaywas 0.1 ng ml" .

Antibody measurements

Anti-sperm antibodies were measured by ELISA. As antigen,50 000 caudal epididymal spermatozoa suspended in 50 picarbonate buffer (14 mmol Na2C03 , 25 mmol NaHC03 1 " ')were adsorbed to each well of round bottomed 96-well plates(Maxisorp®1 Immunopiates, Nunc, Roskilde) for 2 h at 37°Cand stored at

-

20°. Before use, wells were blocked for 60 minwith 200 pi of 5% (w/v) skim milk powder in PBS (pH 7.2) andrinsed three times with PBS. Sera for assay were centrifuged at18 000 # for 10 min and serially diluted twice in PBS-0.25%(w/v) BSA (pH 7.2); 50 pi of the diluted serum was used perwell. Day 0 sera from each animal served as controls. Afterincubation at 37°C for 60 min, wells were washed three timeswith PBS-0.02% (v/v) Tween 20 (PBS-Tween) and 50 pi of a

1:3000 dilution in PBS-BSA of goat anti-rabbit IgG (Bio-Rad,Richmond VA) conjugated to horseradish peroxidase wasadded to each well and the plates incubated for a further60 min at 37°C Plates were washed three times with PBS-Tween and 50 pi 0.001 mg ABTS ml "1 (2,2-Azino bis3-ethylbenzthiazoline sulfonate, Sigma, St Louis, MO) inacetate buffer (100 mmol NaCH3COOH \ 50 mmolNaH2P04 1

" \ pH 4.0), and 0.01% (v/v) H202 were added toeach well. The colour reaction was allowed to develop for20 min at room temperature and then terminated by theaddition of 10% (w/v) SDS and the absorbance at 405 nm

determined with a Bio-Rad model 3350 microplate reader.Anti-viral antibodies were measured by ELISA using con¬

centrated whole virus as antigen as described by Kerr (1997).For immunoblots, detergent extracts of caudal epididymal

spermatozoa, made by the procedure of Rankin el al (1989),(10 pg per lane) were separated under reducing or non-

reducing conditions on SDS polyacryiamide gels (Laemmli,1970). Proteins were transferred to polyvinylidene difluoride

membrane (Immobilon, Millipore, Bedford, MA). The mem¬

brane was probed with undiluted rabbit serum collected fromone rabbit at 0, 15 and 20 days after infection. As a positivecontrol, a 1:5 dilution of rabbit polyclonal serum raised againstcaudal epididymal spermatozoa (M. Holland, unpublished) was

used.

Virus localization

Tissue sections on slides were dewaxed in xylene andrehydrated through a graded ethanol series (100-70%); sec¬

tions were blocked by incubation in 3% (w/v) BSA in PBS(pH 7.2) for 30 min at room temperature. A mouse monoclonalantibody that binds to a 43 kDa viral protein in MV infectedcells (H. Bults and P. J. Kerr, unpublished) (100 µ of a 1:100dilution of ascites fluid in PBS—BSA) or non-immune mouse

ascites fluid was added to each section and the slides incubatedin a moist box overnight at 4°C Slides were washed andantibody binding was detected by addition of horseradishperoxidase-conjugated sheep anti-mouse immunoglobulin, fol¬lowed by addition of 0.67 mg ml

~diaminobenzidene in

PBS-1.2 mmol H202 1 . Sections were counterstained withhaematoxylin. Viral DNA was detected by in situ hybridizationusing EcoRl X fragment of myxoma virus DNA (Russell andRobbins, 1989) labelled with digoxigenin by incorporationof dUTP digoxigenin, as described by the manufacturer(Boehringer, Mannheim). Probe binding was visualized byaddition of anti-digoxigenin horseradish peroxidase-conjugatedantibody or alkaline phosphatase-conjugated antibody andeither diaminobenzidine—H202 as above or 4-nitro bluetetrazolium chloride (NBT)—X-phosphate (Boehringer).

PCR detection of virus DNA was performed on total DNAisolated from 100 mg of homogenized testis by extraction in250 pi lysis buffer (100 mmol Tris 1" : pH 7.5, 5 mmol EDTA1 " \ 200 mmol NaCl 1

"

1, 0.2% (w/v) SDS, 0.1 mg proteinase ml- ). After overnight digestion at 58°C, the supernatantswere extracted with an equal volume of phenobchloroform(1:1) and nucleic acids precipitated with one volume isopropa-nol and 0.25 volumes of 10 mol ammonium acetate 1_I for60 min at room temperature, and then centrifuged at 18 000 gfor 15 min. The resulting pellets were washed in 70% ethanol,dried under vacuum and resuspended in 50 pi TE (10 mmolTris 1, 1 mmol EDTA \'\ pH 8) plus 10 pg RNase A.Uninfected testis and lysis buffer were extracted in parallel ascontrols. PCR primers were designed from the publishedsequence (Upton et al, 1987) to flank the myxoma growthfactor (MGF) gene producing a 640 bp product. Specificity ofthe PCR product was confirmed by Southern blot (Reedand Mann, 1985) using an internal SnaBl fragment, derivedfrom the cloned MGF fragment, labelled with digoxigenin(Boehringer) as a probe.

Evaluation of fertility by test matingThe rabbits allocated to days 60, 90 and 120 were each test

mated, before tissue collection. Each rabbit was mated with twofemales over a 48 h period. If successful copulation was notobserved then an extra female was used. Mating success was

determined by the number of females giving birth and the littersize.

Results

Gross pathological observationsAll rabbits developed symptoms of myxomatosis between

day 7 and day 10 after infection. These included swelling of theears, eyelids and anogenital region. A 3—4 cm diameter hard,red lesion developed in the skin at the inoculation site, startingat days 3—4. Between day 10 and day 20 after infection, smallsecondary lesions 1—2 cm in diameter were evident on the ears

and face and there was prominent mucopurulent dischargefrom the nose and eyes and, in some cases, respiratorydifficulty. Scrotal oedema was prominent at days 10-30 afterinfection. One rabbit died 18 days after infection. The remain¬ing rabbits were starting to recover by day 30. This was

manifested as a reduction in swelling of the ears, eyelids andanogenital region, together with resolution of primary andsecondary lesions. By day 90 the surviving rabbits were

clinically normal apart from some scarring around the ear

margins.At autopsy, body fat content was reduced in all sympto¬

matic rabbits, compared with uninfected rabbits. The internalorgans appeared normal throughout the course of the infectionwith the exception of the spleen, lymph nodes and testes. Atdays 10, 15 and 20 after infection, the spleens were visuallyenlarged 1.5—2.0 times normal size. There was a generalizedenlargement of lymph nodes, in particular the popliteal lymphnodes draining the primary inoculation site were 3—4 timesnormal size. The testes were initially enlarged at days 5—7, butby day 10 they were reduced in size compared with normal.This was accompanied by gross scrotal oedema up to 1 cm inthickness. At day 20 and day 30 after infection, the testes were

approximately 50% of normal size. The appearance and size ofthe testes had returned to normal by day 60-90, although on

autopsy at day 120, one rabbit still had scrotal adhesions.

HistopathologySections of testis and epididymis were examined from each

rabbit killed. The most obvious pathological changes associatedwith virus replication were a cellular proliferation, includingpatches of acute inflammation, in the interstitial tissue of thetestis and epididymis, and degeneration of the seminiferoustubules. These changes were most pronounced at days 15—20after infection. Regeneration of the seminiferous tubules com¬

menced between day 20 and day 30 and morphology was

essentially normal by days 60-90. Comparisons betweentimepoints and rabbits were facilitated by categorizing thepercentage of seminiferous tubules from each rabbit falling intosix different stages determined by the degree of degenerationor regeneration (Table 1).

There were minimal histological changes in the testis at days5—7 after infection, although occasional sections had a mono-

nuclear cell infiltrate in the interstitium. At day 10 afterinfection, there was an increase in the cellular component of theinterstitium. Inflammatory cells, mostly neutrophils, were

Table 1. Percentage of seminiferous tubules showing various stages of degeneration and regeneration inpostpubertal rabbits after infection with an attenuated strain of myxoma virus

Days afterinfection Stage I Stage II Stage III Stage IV Stage V Stage VI

Uninfected 87 9 4 0 0 0Day 5 77; 79 22; 16 1; 5 0; 0 0; 0 0; 0Day 7 78 18 4 0 0 0Day 9 55 26 18 0 0 0Day 10 79; 56 15; 29 6; 15 0; 0 0; 0 0; 0Day 15 0; 0 1; 8 1; 49 34; 25 64; 19 0; 0Day 20 0; 0 0; 0 0; 0 1; 4 63; 83 36; 13Day 30 0.5; 0 0; 0 6; 0 0; 3 2; 52 86; 46Day 60 0; 47 0; 0 3; 4 0; 0 6; 0 91; 49Day 90 89; 79 0; 0 4; 2 0; 0 0; 0 7; 19Day 120 87 1 12 0 0 0

Stage I, normal tubule with all germ cells present; stage II, first stage of degeneration, in which all germ cells except spermatozoawere present; stage III, tubules contained germ cells and also multinucleated bodies; stage IV, tubules contained spermatogonia,Sertoli cells and multinucleated giant cells; stage V, tubules contained only spermatogonia and Sertoli cells; stage VI, tubules were

undergoing regeneration with germ cells dividing and maturing and spermatozoa had not developed at this stage. Oncespermatozoa were present, the regenerating tubules were classified as stage I.For each rabbit, 400 tubules were assessed from different sections of the testis. Two rabbits were examined at each timepoint andthe results for each are shown, except at days 7, 9 and 120 when only one animal was available.

prominent in some sections in the tissue surrounding theperipheral tubules and around the rete testis (Fig. la) but muchof the interstitium had only slightly increased numbers of cells.Between 6% and 15% of the seminiferous tubules containedmultinucleated giant cells with 2-10 nuclei (Table 1). At thisstage, large quantities of spermatozoa were still present in theepididymis which also contained many sloughed spermato¬cytes and spermatids.

At 15 days after infection, the diameter of the seminiferoustubules was obviously decreased compared with earlier time-points and normal rabbits. This correlated with the decrease intestis volume noted at autopsy. Many tubules contained giantcells with up to 20 nuclei (Fig. lb). Spermatid nuclei werevacuolated with clumped and marginated chromatin. In sometubules spermatozoa could be seen near the basement mem¬

brane, possibly indicating breakdown of Sertoli cell function.Small numbers of ruptured tubules were present but there waslittle generalized inflammatory response within the tubules,suggesting that virus replication and the associated inflamma¬tory response were largely confined to the interstitium.

The number of cells in the interstitium was greatly increased.Cells observed included macrophages, fibroblasts and neu¬

trophils. There was a prominent infiltration of mononuclearcells. In some sections, these cells appeared to be proliferating

from blood vessel walls and migrating into the interstitium(Fig. lc). It was not possible to distinguish Leydig cells.

Spermatozoa were present in the distal epididymis but therewere few or none in the caput and corpus epididymis. Theepididymal lumen contained inflammatory cells, red blood cellsand germ cells (Fig. id).

The maximum degeneration of seminiferous tubules was

observed at day 20 after infection. Most tubules contained onlyspermatogonia and Sertoli cells at this stage (Fig. le; Table 1).In some tubules, Sertoli cells were also missing. In those tubulesthat contained germ cells, the primary and secondary sperma¬tocytes were very large with decondensed chromatin

—

thisfeature was more obvious at 30 days after infection. Theinterstitium was extremely cellular (mostly neutrophils andmononuclear cells) especially around the periphery of the testisand the rete testis. No spermatozoa were present in theepididymal lumen. Large, pale staining stellate cells appeared inthe connective tissue of the epididymis (Fig. if). These cellsresembled the myxoma cells described by Hurst (1937).

Regeneration of the seminiferous epithelium occurred atdays 20-30 after infection. This was not uniform in the samplesfrom the two animals examined (Table 1). Dividing germ cellswere present in the lumina of most tubules in one rabbit andin approximately half of the tubules from the second animal

Fig. 1. Histopathology of postpubertal rabbit testes (a) 10 days after infection with an attenuated strain of myxoma virus, inflammatory cells(arrows show examples) and oedema (arrowheads show examples) within the interstitium (scale bar represents 100 pm); (b) 15 days after infection,multinucleated giant cells (arrows) within seminiferous tubules (scale bar represents 25 µ ); (c) 15 days after infection; mononuclear cells (arrows)proliferating from blood vessel walls (scale bar represents 25 pm); (d) 15 days after infection, caudal epididymis showing inflammatory cells andabsence of spermatozoa within the lumen (scale bar represents 25 pm); (e) 20 days after infection, seminiferous tubules containing Sertoli cells(arrowheads indicate cytoplasm of Sertoli cells) and spermatogonia (arrows) (scale bar represents 25 pm). (f) 20 days after infection, myxoma cells(arrow) within the interstitial tissue of the proximal epididymis (scale bar represents 25 pm); (g) 30 days after infection, regenerating seminiferousepithelium (scale bar represents 25 pm); (h) 60 days after infection, maturing spermatids (arrows) within a seminiferous tubule (scale bar represents25 pm).

100000

10000-

1000

Q-

100

10-

1

m

1 IIm

1

1^

1

ND ND ND ND

5 7 9 10 15 20 30 60 90 120Days after infection

Fig. 2. Myxoma virus titres in testis and epididymis determined byplaque assay of tissue homogenates. Titres are the mean of two rabbitsexcept for day 7 and day 9 when only one rabbit was sampled. Errorbars show the range. ND, not detected; pfu, plaque forming units; limitof detection was 100 pfu g ~

.

(Fig. lg). Elongating spermatids were also present in some

tubules. There was an obvious decrease in the cellular responsein the interstitium in these sections. Regeneration was largelycomplete in one rabbit at day 60, with 90% of tubulescontaining spermatozoa. In the second rabbit at 60 days, about50% of tubules contained spermatozoa or elongating sperma¬tids (Fig. lh). The interstitium was almost normal but some

extra cells were still present. Spermatozoa were present in theproximal epididymis.

At day 90 and day 120, 80-90% of tubules were normal andcontained spermatozoa. Vast quantities of spermatozoa were

present in each region of the epididymis together with smallnumbers of spermatids, spermatocytes, erythrocytes andinflammatory cells.

Virus growth in the testis

The replication of MV in the testis was measured from day5 to day 120 after infection by plaque assay of homogenizedtissue. Virus titres peaked at about 106 pfu g ~

1 at day 10 andremained at this value until day 20 (Fig. 2). No virus wasdetected by plaque assay at day 30 or thereafter (limit ofdetection: one plaque equals 100 pfu g ~ 1). Separate rabbitswere inoculated with approximately 0.5 g testis suspensionfrom each time point to determine whether infectious virus was

present at day 30 and day 60. The rabbit inoculated with theday 30 suspension developed myxomatosis; the rabbit inocu¬lated with the day 60 material remained normal.

As an additional test, total DNA was extracted from testissamples at each timepoint and a specific viral sequence ampli¬fied by PCR. Control experiments with infected tissue culturecells showed that this method could detect DNA from theequivalent of one infected cell in 107 uninfected cells. Thismethod was used to detect viral DNA in the testis samples

from day 5 to day 120, even though infectious virus was notdetected after day 30 (data not shown).

Virus location in the testis

The localization of virus in the testis was examined byinfecting a separate rabbit and killing it 9 days later. Thelocation of virus in the testis was examined using eitherimmunoperoxidase staining to detect viral antigens or in situhybridization to detect viral DNA. Both methods indicatedthat viral replication was occurring predominantly in theinterstitial tissue, although a few cells within tubules close toareas of acute inflammation showed the presence of viral DNAor antigen (data not shown). Much of the inflammation at thistime was occurring in the cranial pole of the testis, the retetestis and the fat at the cranial pole of the testis and viralantigen, and DNA was also localized to these regions of acuteinflammation.

Testosterone and luteinizing hormone concentrations in plasmaThere was a significant decrease in plasma testosterone

concentration at days 10, 15 and 20 after infection comparedwith concentrations at day 0 (P < 0.05, unpaired t test; Fig 3a).By day 30 plasma testosterone concentrations were similar tothose at day 0 and remained so for the remainder of theexperiment.

In contrast, plasma LH concentrations showed a significant(P < 0.05, unpaired f test) increase at days 10, 15, 20 and 30after infection. At day 60, plasma LH concentrations were

similar to those at day 0 (Fig. 3b).

Anti-sperm and anti-viral antibodiesIt was hypothesized that damage to the seminiferous tubules

and inflammatory reactions in the rete testis and epididymismay have exposed sperm antigens to immunological surveil¬lance. This hypothesis was assessed by measuring serum

anti-sperm antibodies by ELISA. All rabbits followed the same

general trend; anti-sperm antibodies were detected betweenday 5 and day 10 after infection and reached a peak at day 20(Fig. 4). Titres declined to values of 50—100 between day 30and day 60 in most animals.

In comparison, anti-viral antibodies were present in serum

from day 7 to day 10 after infection and remained at high titres(of the order of 104—10s) throughout the course of the study(data not shown).

The specificity of the antisperm reactivity was examined byimmunoblot. The serum used was from a rabbit with low(1:100) antibody titres. Antibodies detected a protein ofapproximately 21 kDa in caudal epididymal spermatozoa inserum from day 15 and day 20 (Fig. 5). Two bands weredetected under non-reducing conditions. The 21 kDa proteinwas not detected by polyclonal rabbit antisera raised to caudalepididymal spermatozoa.

Fertility of recovered rabbitsThe fertility of recovered rabbits was assessed directly by

mating each of the day 60 and day 90 rabbits and the day 120

Eco

4 (a)

17

2

"coo

1

liliI

ili¡I"i IIliofiliil I

Ili HI'Il1If · "I "— -1-

7 10 15 20 30 60 90Days after insemination

1.5

1-

0.5

(b)

17 17

»

14

11

"liCà¬

io. 15 30 60Days after insemination

Fig. 3. Plasma concentrations (means ± sem) of (a) testosterone inrabbits from day 0 to day 90 after infection and (b) LH from 0 to 60days after infection. The single day 120 testosterone value has beenpooled with the three values for day 90. The results for all animalssampled at each time point have been pooled. The number of animalssampled at each timepoint is indicated above the bars. *Significantdifference to the day 0 values (P< 0.05).

rabbit with two females and counting the number of offspringborn to the females. All males mated enthusiastically at eachopportunity, indicating a normal libido. One of four femalesmated to the two males at 60 days after infection produced a

litter, indicating limited fertility of one male rabbit. At day 90and day 120, fertility was normal, and all females producedlitters except for one which was not observed to mate. Thenumber of offspring in all litters was within the normal rangefor the breeding colony. The mating success rate in this colonyover the previous 6 months was > 90%.

Fig. 4. Distribution of serum anti-sperm antibody titres in rabbits fromday 0 to day 90 after infection. Twofold dilution intervals were usedin titrations. Individual rabbits are not identified.

Fig. 5. Immunoblot of detergent-extracted caudal epididymal spermproteins separated under non-reducing (lanes 1-4) and reducingconditions (lanes 5-8). Undiluted serum from a rabbit at day 0 (lanes4 and 8), day 15 (lanes 3 and 7) and day 20 after infection (lanes 2 and6) was used to probe the membrane. Lanes 1 and 5 were probed witha 1:5 dilution of rabbit polyclonal antiserum generated againstepididymal spermatozoa as a positive control. Large arrows indicateproteins in the non-reducing conditions detected by serum from day15 and day 20 but not day 0, and by the positive control serum. Smallarrow indicates a protein band at approximately 21 kDa detected,under reducing conditions, by antiserum from day 15 and 20 but notby the positive control serum.

Discussion

Domestic rabbits infected with a highly attenuated strainof MV developed an acute interstitial orchitis associated withdegeneration of the seminiferous epithelium and epididymitisand decreased serum testosterone concentrations. The immuneprivileged nature of the testis did not prevent viral clearance.

Broadly, the disease process could be divided into threephases. A proliferative, interstitial phase occurred between day5 and day 10 after infection. This was characterized by fever,virus replication in interstitial tissue and peritubular cells and an

acute interstitial inflammatory response. Decreased concen¬

trations of testosterone and increased concentrations of LH inthe plasma showed that the endocrine function of the testis was

perturbed. Antibodies to MV and to sperm antigens were

present in the serum. The second phase, between day 10 andday 20 after infection, w? characterized by degeneration of theseminiferous epithelium. Virus titres in the testis and epidi¬dymis were high and stable. Between day 20 and day 60 afterinfection, infectious virus was cleared from the testis andepididymis, there was regeneration of the seminiferous epithe¬lium, the serum titres of antisperm antibodies declined andtestosterone and LH concentrations in plasma returned tonormal.

The major functions of the testis are steroidogenesis andspermatogenesis. These are under the control of thehypothalamic-pituitary-gonadal axis and local intratesticularfactors (reviewed by Maddocks et al, 1990; Spiteri-Grech andNieschlag, 1993). Both functions of the testis were affected byMV infection; however, the hypothalamic-pituitary axisremained responsive, as indicated by the increase in plasma LHconcentrations when plasma testosterone concentration was

depressed. This implies that factors acting within the testiswere responsible for the pathogenesis of testicular damagerather than testicular damage being a secondary consequenceof a centrally perturbed endocrine system.

Physical mechanisms for seminiferous epithelium damageassociated with myxomatosis include increased body tempera¬ture and the loss of thermal regulation of the testis owing tooedema and inflammation of the scrotum and its associatedtunics (Ladds, 1993; Mieusset and Bujan, 1995). Heat damageto the testis has also been associated with a moderate decline intestosterone production (Setchell, 1978; Mieusset and Bujan,1995). Pressure necrosis of tubules may also occur as a result ofincreased numbers of cells and oedema in the interstitial tissueinterfering with fluid exchange. Viral infection of intratubularcells was demonstrated in only a few cells in the sectionsexamined, suggesting that most of the degeneration of theseminiferous epithelium is not due to direct virus infection. Theabsence of inflammatory cells within tubules indicated that, in

general, the Sertoli cell barrier was not compromised.At the molecular level, it is likely that disruption of complex

paracrine and autocrine regulatory networks (Skinner, 1991;Spiteri-Grech and Nieschlag, 1993) and decreased testosteronesynthesis within the testis are largely responsible for degenera¬tion of the seminiferous epithelium. Inflammatory mediatorsmay be important, for example the cytokine tumour necrosisfactor (TNF- ) has been shown to antagonize the action ofFSH on Sertoli cells in vitro (Mauduit et al, 1993) and Mealyet al (1990) have postulated a direct toxic effect of TNF-aon adluminal germ cells.

It seems most probable that decreased plasma concentrationsof testosterone were caused by disruption or downregulationof interstitial Leydig cells. These could be damaged directly byMV replication within the cells or by cells of the immunesystem recognizing virally infected cells. Even in the absence ofdirect infection of Leydig cells by MV, products stemming

from the inflammatory response in the interstitium or disrup¬tion of interstitial macrophage function could affect Leydig cellsteroidogenesis. It is clear, for example, that the inflammatorycytokines, interleukin-1, interleukin-2, -interferon and TNF-a,can act on Leydig cells to suppress testosterone production invitro (Calkins et al, 1988; Orava et al, 1989; Mauduit et al,1991, 1992; Hales, 1992; Meikle et al, 1992; Xiong and Hales,1993). Cytokine-like and cytokine-binding proteins encoded byMV, including a functional homologue of epidermal growthfactor (Upton et al, 1987; McFadden and Graham, 1994), couldalso disrupt local cytokine interactions in the interstitium andwithin the seminiferous tubule and peritubular environment.

It is possible that the decreased plasma testosterone concen¬

trations could simply be the result of interference with venous

drainage from the testis, and intratesticular androgen concen¬

trations could remain normal (Galil and Setchell, 1987). How¬ever, this seems unlikely since at day 10 after infection, whenplasma testosterone concentrations were already decreased, thehistological damage within the testis was quite patchy and notextensive.

The pathology of autoimmune orchitis in rabbits (Tung andWoodroffe, 1978) has similarities to that observed with myxo¬matosis with the exception that mononuclear cells were themajor inflammatory cell in allergic orchitis, whereas in myxomavirus infection polymorphonuclear cells were a prominentfeature. It remains to be determined whether there is a

significant autoimmune component to the orchitis observedwith myxomatosis. An interesting finding was the presence ofserum antibodies reacting with spermatozoa during the acutephase of the infection. The contribution of these antibodies tothe inflammatory response in the testis is unclear.

In conclusion, the molecular causes of testis pathology andthe associated infertility after MV infection of male rabbits are

complex and probably involve direct and indirect effects of thevirus together with systemic and local inflammatory responses.The role of individual cytokines and the extent of perturbationof cytokine networks within the testis during viral infectionremains to be examined. The differences in cytokine secretionprofiles after activation of testicular macrophages comparedwith peritoneal macrophages (Kern et al, 1995) suggest thatstudies on inflammatory responses in the testis will lead toimportant insights into testicular pathology and the control ofinfectious agents in immune-privileged sites.

The help and technical expertise of S. Cooke, K. M. Saint, J. Grigg,H. Clarke and B. Shepherd are gratefully acknowledged. R Seamarkand A. ]. Robinson critically read the manuscript. S. Fountain was therecipient of a Vertebrate Biocontrol Centre Honours Scholarship. Thiswork was partially supported by a grant from the International WoolSecretariat.

References

Bjorvatn (1973) Mumps virus recovered from testicles by fine-needleaspiration biopsy in cases of mumps orchitis Scandanavian Journal of InfectiousDisease 5 3—5

Calkins JH, Sigel MM, Nankin HR and Lin (1988) Interleukin-1 inhibits Leydigcell steroidogenesis in primary culture Endocrinology 123 1605-1610

Charny CW and Meranze DR (1948) Pathology of mumps orchitis Journal ofUrology 60 140-146

Fenner F and Ratcliffe FN (1965) Myxomatosis Cambridge University Press,Cambridge

Fenner F and Woodroofe GM (1953) The pathogenesis of infectious myxoma¬tosis: the mechanism of infection and the immunological response in theEuropean rabbit {Oryctolagus cunkulus) British Journal of Experimental Pathol¬ogy 34 400-410

Galil KAA and Setchell BP (1987) Effects of local heating of the testis on

testicular blood flow and testosterone secretion in the rat international Journalof Andrology 11 73-85

Gall EA (1947) The histopathology of acute mumps orchitis American Journal ofPathology 23 637-651

Hales DB (1992) Interleukin-1 inhibits Leydig cell steroidogenesis primarily bydecreasing 17a-hydroxylase/Cl7-20 lyase cytochrome P450 expressionEndocrinology 131 2165-2172

Hurst EW (1937) Myxoma and the Shope fibroma. I. The histology of myxomaBritish Journal of Experimental Pathology 18 1-15

Jones RC, Stone GM, Hinds LA and Setchell BP (1988) Distribution of5a-reductase in the epididymis of the tammar wallaby (Macropus eugenii) anddependence of the epididymis on systemic testosterone and luminal fluidsfrom the testis Journal of Reproduction and Fertility 83 779-783

Kern S, Robertson SA, Mau VJ and Maddocks S (1995) Cytokine secretion bymacrophages in the rat testis Biology of Reproduction 53 1407—1416

Kerr PJ (1997) An ELISA for epidemiological studies of myxomatosis: persist¬ence of antibodies to myxoma virus in European rabbits (Oryctolaguscuniculus) Wildlife Research 24 53—65

Ladds PW (1993) The male genital system. In Pathology of Domestic Animals pp471-529 Eds KVF Jubb, PC Kennedy and Palmer. Academic Press, SanDiego

Laemmli UK (1970) Cleavage of structural proteins during the assembly of thehead of bacteriophage T4 Nature 227 680-685

McFadden G and Graham (1994) Modulation of cytokine networks bypoxviruses: the myxoma virus model Seminars in Virology 5 421-429

Maddocks S and Setchell (1990) Recent evidence for immune privilege in thetestis Journal of Reproductive Immunology 18 9—18

Maddocks S, Parvinen M, Soder O, Punnonen J and Pollanen (1990) Regulationof the testis Journal of Reproductive Immunology 18 33—50.

Marshall ID, Dyce AL, Poole WE and Fenner F (1955) Studies in the epidemiol¬ogy of infectious myxomatosis of rabbits. IV. Observations of diseasebehaviour in two localities near the northern limit of rabbit infestation inAustralia, May 1952 to April 1953 Journal of Hygiene (Cambridge) 53 12—25

Mauduit C, Hartmann DJ, Chauvin MA, Revol A, Morera AM and Benahmend M(1991) Tumor necrosis factor- inhibits gonadotropin action in culturedporcine Leydig cells: site(s) of action Endocrinology 129 2933—2940

Mauduit C, Chauvin MA, Hartmann, DJ, Revol A, Morera AM and Benahmed M(1992) Interleukin-lct as a potent inhibitor of gonadotropin action in porcineLeydig cells: site(s) of action Biology of Reproduction 46 1119-1126

Mauduit C, Jaspar JM, Poncelet E, Charlet C, Revol A, Franchimont andBenahmed M (1993) Tumor necrosis factor- antagonizes follicle-stimulatinghormone action in cultured Sertoli cells Endocrinology 133 69—76

Mealy , Robinson , Millette CF, Maxjoub J and Wilmore DW (1990) Thetesticular effects of tumour necrosis factor Annals of Surgery 211 470—475

Meikle AW, Cardoso de Sousa JC, Dacosta N, Bishop DK and Samlowski WE(1992) Direct and indirect effects of murine interleukin-2, gamma interferon,and tumour necrosis factor on testosterone synthesis in mouse Leydig cellsJournal of Andrology 13 437-443

Mieusset R and Bujan L (1995) Testicular heating and its possible contributionsto male infertility: a review International Journal of Andrology 18 169—184

Murphy FA, Fauquet CM, Bishop DHL, Ghabrial SA, Jarvis AW, Martelli GP, MayoMA and Summers MD (1995) Virus taxonomy: classification and nomen¬clature of viruses. In Sixth Report of the International Committee for theTaxonomy of Viruses Springer—Verlag, Wien

Mykytowycz R (1953) An attenuated strain of the myxomatosis virus recov¬ered from the field Nature 172 448-449

Orava M, Voutilainen R and Vihko R (1989) Interferon- inhibits steroido¬genesis and accumulation of mRNA of the steroidogenic enzymes P450sccand P450cl7 in cultured porcine Leydig cells Molecular Endocrinology 3887-89A

Poole WE (I960) Breeding of the wild rabbit, Oryctolagus cuniculus (L), inrelation to the environment Australian Wildlife Research 5 21-43

Rankin TH, Holland MK and Orgebin-Crist M-C (1989) Lectin binding charac¬teristics of mouse epididymal fluid and sperm extracts Gamete Research 24439-451

Reed KC and Mann DA (1985) Rapid transfer of DNA from agarose gels tonylon membranes Nucleic Acids Research 13 7207—7221

Russell RJ and Robbins SJ (1989) Cloning and molecular characterization of themyxoma virus genome Virology 170 147-159

Setchell BP (1978) The Mammalian Testis Cornell University Press, IthicaSkinner MK (1991) Cell—cell interactions in the testis Endocrine Reviews 12

45-77

Sobey WR and TurnbuII (1956) Fertility in rabbits recovering from myxoma¬tosis Australian Journal of Biological Sciences 9 455—461

Spiteri-Grech J and Nieschlag E (1993) Paracrine factors relevant to theregulation of spermatogenesis

—

a review Journal of Reproduction and Fertility9$ 1-14

Sutherland RL, Evans SM and Tyndale-Biscoe CH (1980) Macropodid marsupialluteinizing hormone: validation of assay procedures and changes in concen¬

trations in plasma during the oestrus cycle in the female tammar wallaby(Macropus eugenii) Journal of Endocrinology 86 1—12

Tung KSK and Woodroffe AJ (1978) Immunopathology of experimental allergicorchitis in the rabbit Journal of Immunology 120 320—328

Upton C, Macen JL and McFadden G (1987) Mapping and sequencing of a

gene from myxoma virus that is related to those encoding epidermalgrowth factor and transforming growth factor alpha Journal of Virology 611271-1275

Xiong Y and Hales DB (1993) The role of tumor necrosis factor- in theregulation of mouse Leydig cell steroidogenesis Endocrinology 132 2438-2444