Vel. Pathol. 15 : 237-2 48 (1 978)

Escherichia coli and Di arrhoea in the Rabbit J . F.

P RESCOTT

Dep artm en t o f Clinica l Vet e rin ar y Medi cin e , Madin gle y Ro ad , Ca mb ridge , E ng la nd

A bstract. A n o utb rea k of se ve re diarrh o e a a nd dea th in yo ung rabb its wa s associa te d with man y non enterot o xigeni c Escherichia coli in th e caecu m . The sev ere clinica l, pat ho logical a nd bact eri o logical fea tures o f the di esa se , ac ute di arrh o ea associa ted wit h typh litis a nd man y E. coli in the caec u m, coul d be reproduce d e ithe r by the in tr ain test inal inoc ulation of many bacte ria recove red aerobica lly or anaerobica lly from the caecum of these ra bbits o r by the int e st in al in o cul a tio n of large nu mbers of a serogro u p of E . coli , 0 153 , reco vered fro m th e caec um . Further ex pe rime n ts sho wed th at this se rogro up of E . coli, as we ll as a no ne nte rop athogen ic se ro ty pe recovered from hum an faece s , wou ld cause ty phlitis a nd di a rrh o ea if ino cul a te d in la rge numb er s int o the jejunum; pa th o logica l cha nges a lso welt: see n in th e live r a nd kid ney . Si mila r cha nges a lso coul d be induce d by intrave nou s ino cula tio n of a free ze -t ha w (endo tox ic) extract p repa red from th ese st ra ins . A ny fac to r th a t a llo ws ra pid multip lica tion of E. coli in th e ra b bit caecu m may be fo llo we d by a bso rp tio n o f endot o xin a nd subs e q ue nt typh litis a nd so me times by se ve re d iarr ho ea; thi s effect is see n in so me fie ld cases of di arrh oe a in t he rabbit.

E. coli is an unu su al inhabitant of th e rabbit int estine [1 7J but is commonly found in la rge numbers in the caecum of dia rr hoeic rabbits a fte r weaning [11 , 13 , 16 , 22, 25, 27 J. The ro le of E . coli in th e e nte ritis comp lex of rabbits [27J has no t been elucid ated . Descriptions of the isolation of e nte ro toxin-p rod ucing E . coli se ro ty pes [7 , 14 J a re inc omplete ; different se ro ty pes ha ve been isolat ed fro m t he intesti ne s of diarrhoeic rabbits [26 ]. Som e workers [8J prod uced a fatal dia rr hoe ic d ise ase in rabbits aft er o ra l inocu latio n of 1Q7 oocyst s of se ve ra l species of coccidia with 10 10 E . coli of se rogro up 085 ; the main e ffect in body organ s wa s dissemin ated in trava scular co agu lation a ttrib uted to th e effe ct of a bso rbe d e ndo to xin . Infect io n with e ithe r coccidi a or E . coli a lo ne prod uced o nly a mi ld a nd brief diarrhoea . Ot hers [13 J inoc ulated 1011 E . coli into the sto mach of yo ung rab bit s with no ill effects ; E . coli of the sa me se roty pe subs eq ue ntly we re recove red from the fa ec es . Thi s paper describes a n o utbrea k o f diarrho ea in yo ung we aned rabbits a nd the reproduction of th e d isease. Materials and Methods An o ut break o f di a rrhoe a in a hyst erect om y-d eri ved coccid ia-f ree herd o f 50 br eeding do es was inve sti gat ed ; th e herd was designa ted sou rce H . D ia rrh o e a was see n ma inl y in 5= 237

238

Prescott

to 8-wcck -o ld rabbits ; morbidi ty a nd mo rtal ity avc ragcd 70 % . Occas io na lly . o lde r rabbits and ad ults were affected . Rabbits of ten were fo und de ad or died wit hi n 12 hours after signs of diarr hoea . Bacte ria we re recovered from the caecum and je ju n um o n ho rse blood aga r (Oxoid Ltd ., Lo ndon , En gland) and inc u bated aerobica lly o r o n re d uce d rei nforce d clo st ridia l medi um (Oxoid Ltd .) a nd inc ubated a nae ro bica lly . E. coli we re identified by met hod s described [3 J. Serot ypin g of E. coli isol a tes was done by met hods de scri bed [19 ]. Sa mp les were examined for sa lmo ne llae by st reak ing them o nto de o xych ol a te citra te aga r (O xo id Ltd .) bot h d irect ly a nd after inc ubatio n in se le nite brot h for 24 hours . T he productio n of hea t-l a bile e nterotoxin wa s tested by the stero idoge nic effect and p rod uctio n of mo rp ho logica l chan ges in cultured ad re na l ce lls [5] a nd he a t-sta bl e e nte ro toxin p rod uct ion b y the suck ling mo use te st [4] . Isol a tes T 26 , T4 1 a nd 05 we re te sted in this wa y . Fo r ligh t microscop ic exam inat ion, sectio ns of live r , kid ne y, duode n um , jejun um , ile um a nd caecum were fixed in b uffe red 10 % formo l sa line and tissues processed routi ne ly for pa raffi n sec tio ns . Section s we re sta ine d with ha e mat o xyl in a nd eo sin (HE) or by a ra pi d silve r sta in ing met ho d [6] . For e lectro n micro sco p ic exam inat ion, sma ll piece s of cae ca l e pithe lium were fixe d in 4 % glu ta ra lde hyde in p ho sphate b uffer , pH 7 .2 , then in osmi um tetrao xid e a nd we re dehydrated in et ha no l a nd e mbedded in a ra ld ite . Sectio ns fro m rab bit s H 6 , H 7 , 1-1 8 , H I 0 , T 4 1, T42 a nd T43 were exa mined . Bacteria l culture pla te s were inocul a te d wit h 0 .0 6 mi llilite rs of a 10- fol d di lution (in ste rile p hos p ha te-buffered sa line , p l-l 7 .2) of the caecal co ntents from dia rrhoeic rab bits 1-1 7 , H 8 a nd H I O. Pla te s were inc uba ted for 3 d ays at 3 T" C e it her ae ro bica lly or a naero bica lly . Bac teri a we re was he d o ff the pla tes wit h th e ste rile sa line , a nd the was hi ngs o f 20 blo o d ag ar plates or of 20 clo st ridi al pla te s eac h were ino cul a ted int o th e jeju n um of yo u ng ra bb its . T he nu mbe r of bacte ria th us inocul a te d in to eac h ra b bit was ab o ut 10' 2.5. E xper ime nt rab bit s , 6- 8 we e ks o ld , we re ob ta ined fro m a loc al com mercia l so ur ce . R ab bit s we re anaesthet ized with a combinatio n of a neuro lepta na lgesic age n t , Hypn o rrn " (Cro wn C he mica l Compa ny Ltd ., Lamber h urst , Ke nt , E ngland) a nd hal o th a ne (l C I Ltd ., Billingham , E nglan d) . T he jejunum was e xpo se d and pla te wa sh ings inoc ulat ed in a tot al vo lume of 20 m illiliter s of ph o sph a te-bu ffere d sa line . A fte r recov e ry , rabbits were g ro u pe d in ca ges in thre e roo ms . T he th re e gro ups we re th o se ino cul at e d wit h aero bic a lly recovered bact eria , th o se inocula ted wit h a na e ro bica lly gro wn bacteria a nd u ninocu late d co n tro ls . R a b bit s were observed da ily . D ia rr hoea was cla sse d as mil d , mo der a te or se ve re. D iarrh o e a was described as se ve re wh e n ra bbits lo o ked d ull , had watery d iarr hoe ic sta ining o f th e perine um a nd hin d legs an d whe n a " cae ca l sp las hing" so und wa s he a rd on ha nd ling th e ra b bits; this so und was ca used by free ga s in the liqu id co nten ts of the caecum . Diarrhoea wa s desc ribed as mild wh e n faece s wer e sem iso lid; the intermedi a te stage wa s co ns ide red as mo derat e d ia rr hoea . R ab bi ts usu all y were kille d wh en seve re d iarr hoea de ve loped beca use th e y we re d ying o r to sto p th e m from dying ove rnig ht, whic h wo uld pre ven t im me d iate hist ol ogical a nd bacterio log ica l proced u re s . E . coli se rog ro up 0 153, recovered from the di a rrh oei c rabbits described o r sero type 0 11 1: 8 4 :H1 2 , a no ne n teropathogenic serotype de rive d from the hum an inte stin e , we re grow n at 37 ° C for 24 hours in 1% pe pt on e- ye a st- glu cose med ium [9 ], wh ile being shake n mechani call y at 70 shakes per minute . T he bacte ria were ha rve st ed by cen trifuga tion to give a fin al co ncentrat ion of a bo u t 10 12.5 viab le E . coli wh ich were suspe nde d in 20 millilite rs of p hos p hate b uffered saline (p H 7 .2 ). T his sus pe ns io n was inoc ulated into th e je j unum of e ac h of 6- to 8-we e k-o ld rabbits . Afte r in ocul atio n , rabb its were killed at se ria l intervals a nd caeca l E . coli co u nts made by di lutio n o nto agar plat e s . T he caecal wet -to- dry ma tt er rat io was e sti ma ted by dry ing th e we igh e d caecal co nte nts to con stan t we ight a t 120° C. T he caecal E . coli co u nts we re

Diarrhoea in the Rab b it

239

exp ressed pe r gram of dry weight of caeca l co nte nts . T wo ex pe rime nts were don e over 8 an d 7 days, respectively . Eight rabbits were inocu lated intra je ju nally with \{)1 2.5 E . coli of se rogro up 0 153 produced by the met hod described above and killed by heating at 60° C for 1 hou r. Th e inoc ulation was repea ted 4 days la te r. Rabbit s were o bserved for signs of dia rr hoe a . Plasma glucose was est imated by t he glucose oxid ase met hod [21 1afte r col lection of blood into sodium fluor ide . An e ndotox in extrac t was prepar ed by the free ze-t haw met hod [I OJ from both the se rogro ups of E . coli described . One millilite r of endotoxi n ex trac t of ea ch se rog ro up was inoculated intravenously into eac h of two groups of five rab bits . T he effect of inocul at ion o n the rab bits was not ed ; plasma glucose was measure d immed iat e ly when the rabb it died. Sections o f liver , caec um a nd kidney wer e pro cessed for light microscop y.

Results T he co n te nt s of the caec u m in a ll ra bbit s fro m so urce H were fo ul-s mellin g , wat ery a nd brown . A gas la ye r rapid ly accum u lated in t he caeca l co n te nts . In rabbit H4 , t he re were " pa in tb r us h" ha em orrha ges in th e se ros a . In ot her rabbi ts , howe ver , t he caeca l wall was t ra ns pa rent . T he co lo n cont ai ned wa tery brown fa e ce s . T he s ma ll int e stin e lo o ke d norm al. Live rs a nd sp lee ns were swo lle n, t u rgi d a nd con ge st ed . T he kidn e ys were co nges te d a nd t he m e senteric lymp h nod e s were severa l tim e s t he ir norm al size ; th e re was a slig ht excess of a se rosa ng u ino us peri ton e a l fluid . In t he caecu m , th e re was se ve re oedem a of th e lam in a p ro pri a and sub m ucosa, dil a tion of crypt s, lo ss o f surface e p it he lium, vac uola tio n a nd ne crosis of e p it he lia l ce lls and in filt ra tion by pol ymorphonucl e ar le uco cytes into th e oe de mato us lamin a pro pria . Blood vesse ls and lym p ha tics we re co ngested a nd t he re was ext ravasa tio n of e ry t h ro cy tes into t he la m ina p ro pri a a nd s u b m ucosa . Ne ut ro p hi l in filtrati on in to su be p it he lia l ti ssues was o fte n g rea t. Exte ns ive a ttac h me nt by I- m icro me te r di am et er co cco bac illi to the ca eca l e p it he liu m ofte n was see n; a ll ra b b its had so me degre e of thi s ba ct eri al attachm ent a lt ho ug h th e e x te n t see me d to va ry with t he seve rity o f th e ca ecal le sion s . Hi st o lo gica l c ha nges in th e ileum a nd col on we re sim ila r to , but less se ve re t han , t ho se se e n in th e ca ecum ; t he re wa s littl e c ha nge in the d uoden um a nd jej unum . Va c uo la tio n and cloud y swe lling o f he pa tocyte s wa s a fe a t u re of th e he pa tic le sion s . T here wa s a lso diffuse ne utrop hil infilt ra t io n and con ge stio n a nd h yd ro pi c c han ge in the Kupffer ce lls. In t he kidn ey s th ere was hyd ropic c hange in t he pro xim a l t u b u la r e p it he liu m. Sta in ing wi th silve r sho we d th in vibrio -li ke b act eri a (2 .3 x O.3 micro meters) in t he cry p ts o f th e caeca of seve ra l rabbit s. R abbit H1 0 had t he most orga ni s ms whi ch in so me areas see me d to be in vadin g th e e p it he liu m . In o t he r rabbits t he y were co nfine d to th e lum en ; no vib rio - like o rga n is ms were d et e ct ed in rabbit s H7 a nd H 8 . Exa m inat io n w it h th e e lectro n m icrosco pe a lso sho wed th at le sion s were most seve re in th e caec u m . Da maged epit he lia l cells ha d lo st th eir microvilli a nd had aggregat ions of e lectron-de nse ro u nd foci of 300 x 10 - 7 rnrn di am et er. The cocco-

240

Prescott

Table I. E. coli co lon y count in diarrhoeic rabbit s , original so urce (H) : count on horse blood agar per gra m of int e stinal content s R abbit

Jejunum, in th ou sand s

Caecum , in million s

H2

NO ' 10000 27 0 0 23 0 0 .8 30000

15 000 NO NO 5 500 4000 7 90 0

H4 H5 H6 H7

H8 I

NO = not do ne .

bacilli previously described were often found attached to damaged epithelium in the areas where microvilli were absent ; attached cell s were se pa ra te d by a 300 x 10 - 7 mm gap . Coccobacilli occasionally were found within a cell. Vibrio-like bacteria , which possessed a single pol ar flag ellum, usu ally were in larg e numbers parallel to the epithe lial surface rather th an invading it ; in rabbit HI 0 vibrio-lik e organisms occasionally were found in degenerate cells . No coccidi a were detected . Counts of E. coli recovered on agar are shown in table I. There also were typ es of spreading bacilli but becau se there were few they a re not shown in the tabl e . No salmonellae were recovered. The E . coli recovered from these rabbits all belonged to serogroup 0153 . They did not appear to produce heat-labile or he at -stable e nte ro tox ins as te sted in the adrenal cell culture or suckl ing mouse test. The result of inoculatin g many bacteria recove red ae ro bically and ana e ro bically from the caecum of rabbit s H7 and H8 into the jejunum of rabbits is shown in table II . A simi lar experim ent involvin g the inocul ation of bacteria recovered from the caecum of rabbit HI0 gave the following results: of six rabb its inoculated with bacteri a recove red ae ro bically, four developed severe diarrhoea (T40- T44) on day 8 after inocul ation a nd one developed severe diarrhoe a on day 9 (T45) . Six rabbits were inoculated with an aerobically recovered bacteria and one developed severe diarrhoea on day 9. Six un inoculated controls remain ed unaffected . Nin e rabbits were inoculate d with 10 12 . 5 viable E . coli se ro gro up 015 3 . Severe diarrhoea developed in three rabbits on days 6 and 7 a fte r inoculation ; o n day 8 two rabbits had severe diarrhoe a and one was killed (T4 6) ; on day 9 one had severe diarrhoea and four had severe diarrhoea 10 days afte r inoculation . No diarrhoea was noted after thi s time. Control rabbits were un affected . Gross lesions were similar to those in the ori ginal diarrhoeic rabbits . Th e main chan ges were con fined to the caecum . Caecal histological chan ge s (fig . 1, 2 ) wer e also similar to those in th e original rabbits . Attachm ent by coccobacilli to the damaged caecal epithelial cell s was aga in a prominent feature of the dise ase (fig . 2 , 3). Changes in the liver a nd kidneys were also similar to those in t he ori ginal ra bbits . Many vibrio-like or gani sms were see n in about half of th e rabbits with induced

24 1

Dia rr hoea in the Ra bbit

Table II. Effect of inoc ulation of bacteri a recove red fro m the cae cum of diarr hoeic rabb its ( 1-17, 1-1 8)

Inoculum

Aerobic bacter ia (1 2 rabbit s)

Anae rob ic bacteria (II rabbit s)

Uninoc ulated con trol (II rabbit s)

Days afte r inoculation I 2 3 4 5 6 7 8 9 10 II

6S ' 7S : kill 1'26, 0 1,02 5S : kill 1'29,1'3 1, 1'27 D 2S: kill 04 2S: kill 1'5 3 2m'

4S : 5S: I S: IS ,

4M' kill 03 kill 1'28,1'30 kill 05 1M : kill 1'3 2

I S: kill 06 I S: kill 07

12

, S, M, m severe , mode rate , mild diarr hoea; D = died ; no diarrh oea ; ra bbits with '1" prefix refer to those inoculated with bacter ia recovered from H7 and '0 ' prefix to those fro m H8 . d ia rr hoea but cou ld no t be fo und in o ther rab bit s . Twe nty-fo ur unin o cul at ed co ntro l rabbit s were exa m ine d histol o gicall y for vib rio-like o rg a n isms; th e se were fo und in lo w n umbers in th e m ucosa l crypts o f t he caec u m o f six rabbits . Table III s ho ws E . coli co u nts o n agar in ra bbit s inoc u la te d with b act e ria rec o ve red from th e caeca o f th e o rigi na l dia rr hoeic rab b its . T he result s o f inocu latin g 10 12 . 5 E . coli of se rogro up 0153 o r se ro ty pe OI I l :B4 :H I2 a re s ho w n in tabl e s IV a nd V . T he wet to d ry m a tt er rati o o f the norm al caecu m was fo u nd to b e le ss tha n 4.0 ; a ra tio be tw een 4 .0 a nd 5.0 repres e nt ed m ild or mod erate diarrhoea a nd over 5 .0 wa s ass o cia te d wit h inc re a sin gl y se ve re di ar rh o ea . T he hi stol o gical a p pea ra nce of th e caecu m, liver a nd kidn e y afte r intesti na l inocu lation o f E. coli is de scrib ed b el ow . T we n ty-fo u r hours afte r in o cul ati on the caecu m wa s histol o gicall y norm al. Fo rty-e ig ht a nd 7 2 hours after ino cu lat ion the caeca l epithe lia l s urface re ta ine d its re gu lar a p pea ra nce but th e re wa s e vide nce of cellul ar ch an ge in th e s upe rficia l e p it he lia l cell s wh ere t here was b alloo nin g dil ati on of cell s , a ccu m ula tio n o f in tr acytopl asm ic debris a nd nu cl e ar pykn o sis . Ne utrop h il pol ym o rph s were in the la m ina pr opri a . N ine ty-six hours a fte r in o cul ation th e cae ca l surface o u tline was irregu lar, th ere was vac uol ati on , necrosis a n d ulc erati on of epithe lial cell s wit h attach me n t of cocco baci lli a lo ng th e s urface e p it he lium , particu la rly in a reas o f ulc e rati on . T he lamin a pr opri a a nd s ub m ucosa were oede ma to us a n d t he re were ma ny neutrophi l po lymorph s . All c ha nges were not se e n in e ve ry rabbit. The description a bove is a co m posite b ased o n th e three expe rime nts .

Prescott

242

'

.

. l

~ , ~.

-

" ;.

-..-.

..

• 't }

, ""'f••

.,

. ". •