chylous ascites mice* - journal of medical genetics · chylous ascites 0 92 14 8i 187 17 3% 49-2%...

J. med. Genet. (I964). Is IO.

Chylous Ascites in Newborn Mice*B. M. HERBERTSON and MARGARET E. WALLACE

From the Departments ofPathology and Genetics, University ofCambridge

In I958, during breeding experiments in theDepartment of Genetics at Cambridge, a small-proportion of newbom offspring of backcrossmatings of the gene ragged (Ra, Slee, I957) de-veloped a swollen belly. The affected animalsappeared normal at birth but about 24 hours afterthey had begun to suckle their abdomens becameswollen with an accumulation of milk-like fluid inthe peritoneal cavity. During the first 5 days afterbirth the abdominal wall ofnormal mice is relativelytransparent, and the stomach, intestines, and otherviscera are clearly visible from the ventral surface.

* Received December 23, 1963.

In the affected mice, as the condition developed,milky fluid could be seen lying between the loopsof the intestines (Fig. i). In most of these animalsonly a small amount of fluid collected and the dis-tension was correspondingly slight; in some, how-ever, a larger amount collected and the viscera werepartially or completely hidden (Fig. 2), the abdomeneventually becoming immensely distended.The milky fluid was sterile and consisted princi-

pally of a fine suspension of fat in the form of in-numerable minute globules but also contained somemacrophages and a few other inflammatory cells.The term chylous ascites is used to describe such anaccumulation of chylous fluid in the peritoneal

FIG. I. Mouse (3 days old) with chylous ascites (the short tail and polydactyly are due to the genes Sd and Py), viewed from ventralsurface. (a) Before dissection: loops of intestine and free margin of liver clearly visible through abdominal wall. (b) Skin and abdominalmuscle reflected, peritoneum intact: loops of intestine contrast sharply with milky fluid. (c) Peritoneal cavity opened: milky fluid lyingbetween loops of intestine. (All x 2.)

IO

copyright. on N

ovember 29, 2020 by guest. P

rotected byhttp://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.1.1.10 on 1 Septem

ber 1964. Dow

nloaded from

http://jmg.bmj.com/

Chylous Ascites in Newborn Mice

1

b C

FIG. 2. Mouse (I6 days old) with chylous ascites, viewed from ventral surface. (a) Before dissection: loops of intestine visible throughabdominal wall. This mouse was stunted, had very little subcutaneous fat, and hair growth was retarded. In a normal I6-day-old mouseadipose tissue would prevent the viscera being seen. (b) Skin and abdominal muscle reflected, peritoneum intact: a few coils of intestine arevisible; the rest of the viscera is hidden by the miLky fluid. (c) Peritoneal cavity opened: milky fluid lying between loops of intestinewhich, in places, were matted together. The proximal portion of the small intestine was considerably distended. Intestinal obstructionappeared to be developing. (All x 2-75.)

cavity and it seemed an accurate description of thisdisorder.The fate of the affected animals varied with the

severity of the condition. Those in which only asmall volume of fluid was present seemed just asvigorous as their healthy litter mates and gainedweight normally. At about 7 days old the milkyfluid was absorbed and they could not then bedistinguished from their unaffected sibs. On theother hand, those in which a large amount of fluidcollected were not nearly as active as the others andfailed to gain weight at the usual rate. Some of thesemore severely affected animals recovered during thesecond, third, or fourth weeks of life and thendeveloped normally. In others the amount of peri-toneal fluid increased still further and at this stagethey became too weak to suckle and died.Those defective animals that survived until

classification of the gene Ra was possible in thisstock (about 7 days) were invariably found to beragged heterozygotes (Ra+), i.e. to have the sparseand retarded hair growth typical of the expressionof this gene. However, the majority of Ra+ in thecolony, while capable of transmitting the abdominaldefect to later generations, were not themselves

defective. Moreover, only about i in IO raggedsshowed chylous ascites. This suggested that thecondition was due either to an imperfectly penetra-ting dominant closely linked to Ra, or to an occa-sional pleiotropic effect of Ra+.An inherited tendency to chylous ascites does not

appear to have been previously recorded in anyspecies. This account describes an investigationof its morbid anatomical and physiological features(B. M. Herbertson) and of its genetical aspects(M. E. Wallace).

Materials and MethodsExamination of Peritoneal Fluid. The peritoneal

fluid from 22 mice with chylous ascites was collectedpost mortem and measured. Wet and dry preparationsof the fluid were examined microscopically, and 4 sam-ples of the fluid were cultured aerobically and anaero-bically to discover whether bacteria were present.The lipid composition of the peritoneal fluid and

stomach contents of four abnormal mice was analysedand compared. The peritoneal fluid from the 4 mice waspooled and provided about 50 mg. of material; thestomach contents were similarly pooled. The twospecimens were extracted with methanol :chloroform(2:I v/v), ifitered, evaporated to dryness in vacuo, and

II

copyright. on N



g.bmj.com

/J M



ber 1964. Dow

nloaded from

http://jmg.bmj.com/

Herbertson and Wallace

then dissolved in o04 ml. chloroform. 0o2 ml. aliquotswere investigated using thin-layer chromatography(Bowyer, Leat, Howard, and Gresham, I963). Afterspraying each thin-layer plate with 2', 7'-dichlorofluores-cein in ethanol, the approximate percentage and positionof the lipids was evaluated visually under ultravioletlight. The fatty acids of the triglyceride fraction werethen converted to their methyl esters and analysed by gasliquid chromatography.

Morbid Anatomy and Histology. Twenty-twomice with chylous ascites and 20 normal controls of bothsexes and comparable ages, ranging from I to 27 days,were killed with chloroform and dissected shortly afterdeath. The peritoneal fluid from the abnormal micewas collected and examined (see above). The carcasseswere fixed in 4% formaldehyde saline. Transverseblocks of the abdomen and thorax at various levels wereembedded in paraffin wax, and sections were stained byEhrlich's haematoxylin and eosin and by Weigert'sresorcin-fuchsin method for elastic fibres and vanGieson's mixture for collagen fibres. Other stainingmethods used included Mallory's phosphotungstic acidhaematoxylin and McFarlane's modification ofMallory'strichrome method. Serial sections of the abdomen andthorax of 2 abnormal and 2 control mice were also pre-pared. In addition frozen sections of the small intestineof 9 abnormal and 7 control mice were cut and treatedwith a mixture of Sudan III and IV or with Sudanblack. A more thorough microscopical examination ofthe gastro-intestinal tract of 3 abnormal and 2 controlmice was made, comparable paraffin and frozen sectionsbeing prepared of the stomach, duodenum, jejunum,ileum, and colon.Three adult mice, 8 to I2 months old, that had de-

veloped chylous ascites shortly after birth and hadrecovered were also killed with chloroform and ex-amined.

Feeding Experiment. Three 2-day-old mice withchylous ascites were given a few drops of a carbon sus-pension (particles size not greater than 3ji) by mouth.The passage of the carbon along the alimentary tract wasobserved from the ventral surface of the abdomen duringthe subsequent two days.

Breeding Stocks. The stock in which chylous ascitesappeared in January I958 had been formed from twoclosed foundation stocks. One had been segregating for4 years for the linkage group V markers ragged (Ra), tan(at), and wellhaarig (we). Some i,6oo mice were ex-amined from linkage backcrosses, of which 848 wereraggeds (Parsons, I958). The number of raggedsexamined in the whole experiment is probably over I,ooowhen preparation matings are included. The otherfoundation stock had also been segregating for 4 yearsfor these markers and for the two further linkage groupV markers fidget (fi) and Danforth's short tail (Sd). It isestimated that some I,300 raggeds were examined. Theamalgamated stock segregated for a year in these 5markers before chylous ascites was seen.

The technical assistants in charge of all these stocksexamined all litters, irrespective of age, as a weeklyroutine, so that most young were scrutinized 2 or 3 timesin the nest. Since many cases of chylous ascites aregrossly abnormal, often for over a week, the chance thatany cases occurred before January I958 and were missedis extremely remote.The new linkage group V stock was established to

investigate genetical interference in the Ra-Sd part ofthe chromosome, and to this end the I6 possible types offivefold linkage backcross were gradually made. Theobservations on chylous ascites were incidental to thisprogramme, which is still under way. However, itsdesign and the even spacing of the markers (recombina-tion percentages are approximately Ra-25-at12 ze-22-fi-22-Sd, Wallace, I958) make it ideal for the dis-cernment of any loci, whether of large or small effect onthe chylous ascites phenotype, that may exist anywhere inthese segments.

Indeed, it was at once apparent that the area markedby Ra has a major effect, for the first nine cases were allRa+. No obvious correlation was found with sex orwith the other 4 markers, but any weaker correlationsthat exist will emerge from a full analysis of the completeexperiment. Meanwhile, two samples of the data,considering chylous ascites and ragged only, have beentaken. The first, for I958 to I960, and the second, forI96I to I963, appear as data (a) and (b) of Table I.

It was impossible to decide which matings segregatingin Ra but producing no chylous ascites animals werenevertheless genotypically capable of producing them.This was because affected animals nearly always camefrom normal parents and because the incidence of chy-lous ascites was low, only I or 2 cases occurring in amating: initially too, affected animals seldom survivedlong enough to breed. However, the confinement ofchylous ascites to the Ra+ phenotype continued. Forthese reasons, the data from all matings segregating inragged, irrespective of whether or not they producedaffected animals, are pooled.

Nevertheless, the possibility that affected parents wereoccasionally being used and that their use could alter theincidence of chylous ascites in their offspring wasworth investigating. The affected animals of Table Iare therefore classified in Table II according to whethertheir Ra+ parent had itself been classified as showingchylous ascites or not. Since only one parent was Ra+and this could be either the father or the mother, thedata are subdivided according to the sex of the Ra+parent.A small experiment on the genetic control of incidence

was also started in June I962. A single affected raggedwas crossed to the standard inbred strain AG/Cam.From its progeny, two series of backcrosses of Ra toAG/Cam were made. In the first, only affected raggedswere used in each generation, in the second only un-affected raggeds. The former series has reached thethird backcross, the latter the fifth. This process of'grading up' is expected to make both stocks graduallyisogenic with AG/Cam, theoretically complete isogenesisbeing reached after the fifth cross. The stocks differ

12

copyright. on N



g.bmj.com

/J M



ber 1964. Dow

nloaded from

http://jmg.bmj.com/

TABLE ICLASSIFICATION OF PROGENY FROM BACKCROSSES OF Ra, WITHOUT SELECTION

ASCITES (CHYLOUS ASCITES IS SYMBOLIZED 'Ch.As.')IN RESPECT OF CHYLOUS

Non-Ra RaData Total Ra Ch.As. All Ra

With Without With Without Divided By Divided ByCh.As. Ch.As. I Ch.As. Ch.As. All Ra Total

(a) Jan. I958 to July I960 (2?)* 1248 127 IOxI6 2391 III% 478%(b) Oct. 196I to Sept. I963 0 1043 60 741 1944 17-8% 46 3%

X' Analysis Non-Ra Ra Ra With Ch.As. Ra Without Ch.As.

(a) Jan. I958 to July I960 1250 IOI6 127 Ioi6(b) Oct. I96I to Sept. I963 1043 741 I60 741

Contingency x' 4-42 18-4Probability <0'05 4-OOI

* The two possible cases of chylous ascites in non-Ra are doubtful because they occurred before the closeness of the association with Rawas realized; their classification was not, therefore, checked.

TABLE IICLASSIFICATION OF THE PROGENY WITH CHYLOUS ASCITES (FROM TABLE I) ACCORDING TO THE CHYLOUS

ASCITES PHENOTYPE OF THE Ra PARENT (CHYLOUS ASCITES IS SYMBOLIZED 'Ch.As.')

Mother Ra Father Ra Number From Ra Ch.As.Data Total Parent Divided

With Without With Without By TotalCh.As. Ch.As. Ch.As. Ch.As.

(a) Jan. 1958 to July 1960 I 41 1 84 127 2 =157%127

(b) Oct. 1961 to Sept. 1963 2 48 x6 94 i60 x8IiS 25%

X' Analysis Parent Ra

With Ch.As. Without Ch.As.

(a) Jan. 1958 to July I960 2 125 Corrected contingency x = 8-78(b) Oct. I96I to Sept. I963 I8 142 Probability = < o-Oor

TABLE IIICLASSIFICATION OF PROGENY FROM SUCCESSIVE BACKCROSSES TO AN INBRED STRAIN, WITH SELECTION IN

RESPECT OF CHYLOUS ASCITES (CHYLOUS ASCITES IS SYMBOLIZED 'Ch.As.')

Non-Ra RaData Total Ra Ch.As. All Ra

With Without With Without Divided By Divided ByCh.As. Ch.As. Ch.As. Ch.As. All Ra Total

(a) From Ra parents withchylous ascites 0 92 14 8i 187 17 3% 49-2%

(b) From Ra parents withoutchylous ascites 0 46 I 49 96 2-0% 52I°%

X' Analysis Non-Ra Ra Ra with Ch.As. Ra Without Ch.As.

(a) From Ra parents withchylous ascites 92 95 14 8I

(b) From Ra parents withoutchylous ascites 46 50 I 49

Contingency Xs o-oo6 4-27Probability > 0°95 <0'05

copyright. on N



g.bmj.com

/J M



ber 1964. Dow

nloaded from

http://jmg.bmj.com/


,9'$~.7'X,.,e,A,E,,,6,S, X,

k

,' 4 w g v !'< 4 f

xWX vNN;.W..

W; '",',s1'''.'s''','1''12W

FIG. 3. Peritoneal fluid of mouse (2 days old) with chylous ascites.Macrophage (about I2tu diam.) containing droplets of fat; a red bloodcell is also present. Innumerable refractile chylomicrons in the back-ground. (Wet preparation, x i,ooo.)

ultimately only for those genes selected by the choice ofthe ragged parent according to its phenotype in respectof chylous ascites. The data are given in Table III.

ResultsPeritoneal Fluid. In mice killed during the

first i6 days the peritoneal fluid was always milk-like in consistency and appearance. Later, it becameprogressively more 'watery', and in 2 mice killedduring the fourth week the fluid was only slightlyopalescent but had a curiously shimmering surface.The amount varied from a trace to I-4 ml., but inmost it was between o03 and 0-5 ml. Although suchvolumes seem small they appear as substantialamounts in the abdominal cavity of suckling miceand in some accounted for as much as I5 % of thetotal body weight. Towards the end of the firstweek some fragments of a cream-coloured cheesymaterial were often found floating freely in the milkyfluid.

Initially the most impressive microscopical

FIG. 4. Peritoneal fluid of mouse (6 days old)' with chylous ascites.A group of large macrophages (up to about 301± diam.) containingabundant droplets of fat. (Wet preparation, x I,ooo.)

features were an abundance of minute globules offat (chylomicrons) and some macrophages containingdroplets of fat (Fig. 3). A few polymorphonuclearleucocytes and red cells were also present. Thenumber, size, and fat content of the macrophagesthen gradually increased (Fig. 4), and a smallnumber of multinucleate cells appeared. Towardsthe end of the first week a few colourless anisotropiccrystals also began to develop. Most of thesecrystals were needle-like and were often clusteredtogether in the form of tufts (Fig. 5). A fewrectangular plate-like crystals also appeared. Duringthe second and third weeks the fluid still containedabundant chylomicrons but in the fourth week moreand more crystals formed, and the fatty globulesbecame fewer. At this stage fat-containing macro-phages also became less numerous. There can belittle doubt that the 'milky' appearance of the fluidin the first 2 to 3 weeks was due to the abundanceof chylomicrons during that period and that theopalescence and shimmering surface of the fluid

- ,.-

4,'

0

14

copyright. on N



g.bmj.com

/J M



ber 1964. Dow

nloaded from

http://jmg.bmj.com/


FIG. 5. Peritoneal fluid of mouse (S days old) with chylous ascites. A tuft of needle-likecrystals; a phagocyte and a few red blood cells are also present. (Wet preparation, x i,ooo.)

noticed particularly during the fourth week was dueto the high crystal content.No bacteria were isolated from the aerobic and

anaerobic cultures of the 4 specimens of peritonealfluid.The results of the analysis of the lipids in the

peritoneal fluid and stomach contents of 4 mice

with chylous ascites are shown in Tables IV and V.The general lipid composition of the two materialsis remarkably similar, both consisting chiefly oftriglycerides. The fatty acid content of the tri-glycerides in the peritoneal fluid and the stomachcontents also corresponds within the limits of theexperimental error for the method used.

TABLE IVAPPROXIMATE COMPOSITION (%) OF LIPID EXTRACTED FROM PERITONEAL FLUID AND STOMACH CONTENTS

OF MICE WITH CHYLOUS ASCITES

Cholesterol Free FattySpecimen Triglycerides Cholesterol Esters Acids Cephalin Lecithin Sphingomyelin

Peritonea fluid 70 5 5 IO Trace 5 5Stomach contents 6o 5 5 10 10 5 5

Note: The figure before the colon denotes the number of carbon atoms and that after it the number of double bonds.

I5

TABLE VFATTY ACID COMPOSITION OF TRIGLYCERIDES (MOLES %) FROM PERITONEAL FLUID AND STOMACH

CONTENTS OF MICE WITH CHYLOUS ASCITES

copyright. on N



g.bmj.com

/J M



ber 1964. Dow

nloaded from

http://jmg.bmj.com/

Morbid Anatomy and Histology.GENERAL. In the youngest mice the abnormalities

were confined to the abdomen and consisted of theaccumulation of fluid in the peritoneal cavity andchanges in the small intestine and mesenteric lymphvessels. The more severely affected animals aged2 to 4 weeks showed similar abdominal changes butin addition were, smaller than their normal littermates, and had a rather wizened appearance and ageneral deficiency of adipose tissue. In some ofthese creatures hair growth was more retarded thanusual in Ra+.The 3 adult mice that had developed and re-

covered from neonatal chylous ascites were normalapart from a few fibrous adhesions between the coilsof the intestine and the other viscera.

SMALL INTESTINE. The wall of the gastro-intestinal tract of normal newborn mice is very thinand the contents of the various portions can readilybe seen through it. After suckling, the stomach andproximal part of the duodenum contain milk andappear creamy white. On the other hand, the distalportion of the duodenum, the jejunum, and ileumwith their bile-stained contents are usually a muchdeeper colour. In most of the abnormal mice thewhole length of the duodenum and the proximalpart of the jejunum were creamy white and thedistal portion of the small intestine was unduly pale.When the gastro-intestinal tracts of the abnormalmice were opened the walls of the small intestines,particularly of the jejunal portion, were found to bethicker and paler than usual but their contentsappeared normal.

Microscopically, the wall of the small intestinewas abnormal in all the mice with chylous ascites.In those killed during the first IO to I2 days theprincipal alteration was an accumulation of dropletsof fat in the submucosa. The amount of fat variedboth from animal to animal and in a particularmouse from one part of the small intestine toanother. In a few mice only a small quantity of fatwas present but in others there was a substantialamount, and the submucosa, normally an insignifi-cant structure, formed a broad layer bloated withfat (Fig. 6, 7, 8, and 9). The largest amount of fatwas always to be found in the jejunum but even inthe least affected animals fat was usually also presentin the submucosa of the duodenum and the ileum.At first most of the fat was in the form of dropletsbut towards the end of the first week acicular crys-tals began to appear. In frozen sections much of thefat was anisotropic (Fig. io) and some showed'maltese-cross' birefringence. There was remark-ably little cellular response to the presence of the

V,.I,;r ' Zii

4C

.r.-;

FIG. 6. Small intestine of mouse (3 days old) with chylous ascites.Broad 'oedematous' submucosa: frozen sections showed that the'oedematous' appearance was due to the presence of fat. (H. and E.X 52.)

FIG. 7. Small intestine, abdominal wall, and part of stomach of nor-mal mouse (3 days old) The submucoss of the normsl small intestineis an insignificant layer. (H. and E. x 52.)

copyright. on N



g.bmj.com

/J M



ber 1964. Dow

nloaded from

http://jmg.bmj.com/

FIG. 9. Jejunum of normal mouse (5 days old). Insignificant sub-mucosa of jejunum about 3 5 cm. from pylorus. (H. and E. x 52.)

FIG. 8. Jejunum ofmouse (5 days old) with chylous ascites. 'Oedema-tous' appearance of submucosa of jejunum about 3 5 cm. from pylo-rus. The diameter of the jejunum is greater than normal. (H. and E.

FIG. IO. Jejunum of mouse (5 days old) with chylous ascites. Bire- FIG. II. Jejunum of mouse (13 days old) with chylous ascites. Widefringent lipid in submucosa and in core of villi of jejunum about sinusoidal channels containing blood in submucosa. (H. and E.4 cm. from pylorus. From same mouse as Fig. 8. (Polarized light, x 52.)X 52.)

copyright. on N



g.bmj.com

/J M



ber 1964. Dow

nloaded from

http://jmg.bmj.com/


fat in the submucosa and although some was en-gulfed by macrophages most was lying freely. Afew small clusters of polymorphonuclear leucocyteswere also occasionally present in the submucosa.

Although the accumulation offat was always muchmore conspicuous in the submucosa than in the otherlayers of the intestinal wall, the core of the vilialso often contained more fat than usual. In addi-tion, small droplets of fat were frequently presentin smooth muscle cells, particularly those in theinner portion of the circular muscle layer adjoiningthe submucosa. In contrast, no abnormality wasdiscovered in the epithelial cells lining the smallintestine. The cells were of normal size, shape, anddistribution, and they contained a normal amountof fat.Towards the end of the second week the amount

of fat in the wall of the small intestine rapidlyiminished, and during the third and fourth weeksonly very small quantities remained. At this stagethe most notable change was the presence in thesubmucosa of widely distended sinusoidal channelscontaining blood (Fig. ii).

MESENTERIC LYMPH VESSELS AND NODES. If themesentery of the small intestine of a suckling mouseis gently spread out the mesenteric lymph vesselsfilled with chyle can readily be seen coursing fromthe intestine towards the root of the mesentery. Aparticularly good view of these vessels can beobtained by using a dissecting microscope. In 5mice with chylous ascites killed during the firstweek the mesenteric lymph vessels appeared undulydistended and tortuous, and milky fluid could beseen oozing from the surface of the mesentery andalso from the peritoneal surface of the jejunum.This alteration was seldom observed in abnormalmice killed during the second and subsequent weeks.Moreover, the distension of the lymph vesselsobserved during the first week was never as promi-nent in paraffin sections as it had seemed duringnecropsy.The size and number of lymph nodes at the root

of the mesentery seemed to be similar in both theaffected and normal mice, but in those with chylousascites the sinuses in some of the mesenteric nodeswere uniformly filled with blood, and a few mega-karyocytes were often present among the lympho-cytes in the denser lymphatic tissue. These unusualnodes were probably haemal nodes or haemal lymphnodes. None of the nodes showed any evidence ofinfection or neoplasia.

PERITONEUM. Flakes of cream-coloured cheesymaterial began to form on the peritoneal surface ofthe viscera towards the end of the first week. This

material was particularly prominent in the neigh-bourhood of the liver and pancreas and consistedof lipid-filled cells, droplets of fat, acicular andplate-like crystals, and cell debris closely packedtogether (Fig. I2 and 13). Most of the cells weremacrophages, some being multinucleate, but a fewpolymorphonuclear leucocytes were also presentAt first, this material was only loosely attached to

the peritoneum, but during the second and thirdweeks it became more firmly adherent and inseverely affected animals the loops of intestine andother viscera became matted together. During thisperiod the number of recognizable cells diminishedand the material tended to resemble the pultaceousmush seen in human atherosclerosis. In places itbecame invaded by granulation tissue and thefibrous adhesions seen in the 3 adult mice wereprobably the end stage of this process.

LIVER. In most of the more severely affectedanimals the liver cells immediately beneath Glisson'scapsule contained abundant fat in their cytoplasm(Fig. 14). The rest ofthe liver appeared normal andthe haemopoietic tissue present at birth diminishedat about the same rate in both affected and normalanimals.

Feeding Experiment. The carbon suspensionfed to the three 2-day-old mice blackened the con-tents of the stomach and intestines but none es-caped into the peritoneal cavity. When viewedthrough the ventral wall of the abdomen the blackloops of the intestine contrasted sharply with thewhite chylous fluid.

Genetical DataCONFINEMENT OF CHYLOUS ASCITES TO Ra+. In

both samples of the linkage group V stock, cases ofchylous ascites remained confined to ragged hetero-zygotes, only 2 doubtful cases in non-raggeds beingrecorded (Table I). In the small isogenesis experi-ment also, only raggeds were affected (Table III).

CHANGES IN INCIDENCE OF CHYLOUS ASCITES INRa+. In the I958-60 sample, only 2/I27 raggedswith chylous ascites came from a similarly affectedparent (Table II), but in the i96i-63 sample theproportion rose to I8/I60 (Table II). The differ-ence is significant (X2 = 8'78 for 1 d.f., probabilityless than o-ooi).The incidence of chylous ascites in the ragged

progeny also showed an increase, namely fromII-I% in the earlier data to I7 8% in the later(Table I). This increase is even more signifi-cant (X2 = I8.4 for 1 d.f., probability much lessthan o-ooi).

I8

copyright. on N



g.bmj.com

/J M



ber 1964. Dow

nloaded from

http://jmg.bmj.com/

Chylous Ascites in Newborn Mice I9

~ ~ * ~

*..--~~~~~~~~~~~~~~~~~~-- w ~ ~ ~ ~ ~ ~ ~--..f-w~ ~ ~ ~~~~~~~~~~~~~~~~4

~~~~~~~~~~~- -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~T

4 - *. . ~~~~~~~~~~~~~~~~~~ --~~~~T

2%~ ~ ~ ~ ~

*~~~~~~~~~l Tt*0P.

--'4.~~~~~~~~~~~4'jN~~~~,~A--

~~~~~~~~Ai, 4 ;

A .4,Z.V ~ *~VW. W~~~ ~ ~ ~ ~~ ~~~~ * 1-i

K.**..

FIG.~~~~12ie n iprg fmue(ody l)wt hlu IG 3 aceao os asod ihcyosactsasits.Anacumlaio olpi-cntinn marohae andsm Colcin flpd-ntnng arphes nd ter elsipoyopoucerluoctsi eitna ait ewe ia prtnalcvt djiigpnres H ndE 0.phragmabove)andlier. (H.and E.x zoo.

FIG. 14. Liver of mouse (6 days old) with chylous ascites. The liver cells immediatelybeneath Glisson's capsule are enlarged and vacuolated. Frozen sections showed that thesecells contained abundant fat. Normal amount of haemopoietic tissue for age. (H. and E.x IOO.)

copyright. on N



g.bmj.com

/J M



ber 1964. Dow

nloaded from

http://jmg.bmj.com/


In the isogenesis experiment which started with asingle affected ragged, the series in which onlyaffected raggeds were used in each generation hasan over-all incidence of I7'3 % (Table III). This isclose to the incidence of the stock from which it wasderived, namely I7-8 %. In the series in which onlyunaffected raggeds were used (after the first affectedone), the over-all incidence is much lower, 20%(Table III). The difference is just significant (X2 =4X27 for I d.f., probability less than o0os).

CHANGES IN VIABILITY OF RAGGED. The increasein incidence of chylous ascites in raggeds from thetwo samples of the linkage group V stock wasaccompanied by a slight decrease in viability ofRa+. From 47-8% of all progeny, a frequencyclose to the expected 50% for full viability, the pro-portion of raggeds fell to 46 3 % (Table I); thedifference is just significant (X2 = 4-42 for 1 d.f.,probability less than o0o5). Clearly this is a smallerchange than the change in incidence.

In the isogenesis experiment, the greater incidencein the selected series over that in the unselected isalso accompanied by a difference in the viability ofragged (Table III). It is again in the same directionas in the larger data, but it is insignificant (X2 =o-oo6 for 1 d.f., probability greater than o095); infact the proportion of raggeds is in both cases closeto the 50% expectation for full viability.

DiscussionMorbid Anatomical and Physiological As-

pects. Milky fluid of various kinds may collect inthe peritoneal cavity of man and other animals. Inchylous ascites the fluid consists of chyle that hasleaked from lymphatic vessels into the peritonealcavity as a result of rupture or obstruction of thethoracic duct or its main abdominal tributaries. Inman the most frequent causes of this condition areinjury, tuberculosis, filariasis, and malignant tu-mours involving the lymphatic system. Apart fromtrue chylous ascites, fluid with a milky appearancemay be found in certain other circumstances. Forexample, a chronic inflammatory effusion in aserous cavity sometimes becomes milky due, it isclaimed, to the liberation of fat from disintegratinginflammatory cells. In the present context it is alsoconceivable that in newborn animals a substantialdevelopmental defect of the stomach or duodenummight allow milky liquid to escape directly into theperitoneal cavity.

If all aspects of the disorder in the young mice aretaken into account there can be little doubt that the

condition is a true chylous ascites due to defectivelymphatic drainage. The features favouring thisview are the close time relation between the onset ofsuckling and the appearance offluid in the abdominalcavity, the excessive distension of the mesentericlymph vessels, the oozing of milky liquid from themesentery, and the chylous nature of the fluid. Onthe other hand, the natural history of the disorderand the composition of the fluid provide no supportfor the idea that a chronic inflammatory process isthe fundamental abnormality. Moreover, the failureduring dissection to find any gross communicationbetween the gastro-intestinal tract and the peri-toneal cavity makes it most unlikely that the milkyfluid consists of milk and digestive juices that haveleaked through a developmental fault in the wall ofthe alimentary tract. This opinion is supportedby the results of the carbon feeding experiment. Inaddition, if a mechanism of this kind were respons-ible, the affected mice would almost certainly havedeveloped acute peritonitis, and this was neverobserved.Although inadequate lymphatic drainage appears

to be the likely mechanism of this condition, aconvincing anatomical explanation for this de-ficiency has not been satisfactorily demonstrated.The absence of chylothorax and lymphatic oedemaof the trunk and extremities suggests that the defectis limited to the abdominal lymph nodes and vessels,and the frequent recovery of affected mice indicatesthat obstruction to the passage of chyle is often atemporary phenomenon. Examination of the lymphnodes and main lymphatic ducts in the abnormalanimals revealed no evidence of an inflammatory orneoplastic disorder that might account for theobstruction. However, some of the mesentericnodes had blood in all or most of their sinuses andappeared to be haemal or haemal lymph nodesrather than true lymph nodes. If this interpretationis correct it is conceivable that unusual communica-tions between blood vessels and lymphatic channelshad formed in this region during foetal life and, byconverting potential lymph nodes into haemalnodes, had reduced effective lymphatic drainage.Although this idea may be incorrect it is likely thatin the affected animals this portion of the lymphaticsystem is inadequately developed at the time ofbirth and is unable to cope with the substantialdemands suddenly placed upon it when sucklingbegins. In other words, this disorder may be acounterpart in the lymphatic system of the varioushypoplastic conditions of other systems that causefunctional embarrassment at or after birth.

There has been much controversy about variousaspects of fat absorption including the route by

20

copyright. on N



g.bmj.com

/J M



ber 1964. Dow

nloaded from

http://jmg.bmj.com/


which fat is transported from the small intestine.For many years after the classical study of Munkand Rosenstein (I89I) it was believed that about6o0% of absorbed fat passed from the intestinalepithelium into the lacteals, the remaining 40%going directly into the blood-stream. Morerecently, however, Bloom, Chaikoff, Reinhardt,Entenman, and Dauben (1950), Bloom, Chaikoff,Reinhardt, and Dauben (195I), and others haveprovided convincing evidence that in normalcircumstances after feeding common fats, such asthe triglycerides of the longer chain fatty acids,practically all the fat absorbed from the intestineenters the lymphatics with very little passing intothe intestinal blood capillaries. Although investi-gation of the abnormal mice yields no informationabout the actual amounts of the various lipidsentering the lymphatics or blood capillaries thesimilarity of the lipid composition of the peritonealfluid and of the milk taken from the stomach isimpressive.

In an investigation of the genetics, morphology,and development of ragged, Slee (1957) found thata high proportion ofRa Ra mice developed general-ized oedema in the embryo stage and died eitherin utero or shortly after birth. Ofthe various possiblecauses of this oedema, Slee considered that heartfailure or an endocrine abnormality were the mostlikely. However, there appear to be no features inhis account that refute the possibility of its beingdue to a fault in lymphatic drainage. Indeed, thegeneralized oedema and the chylous ascites mayboth be an expression of lymphatic insufficiency,the differences between them resulting from varia-tions in the severity and position of the defects.

Genetical Aspects. It is clear from all the gen-etical data that the incidence of chylous ascites couldnot have been maintained at I I-I7 % unless Ra, ora dominant close to it, were the main factors res-ponsible. A major factor independent of Ra, orseveral independent factors with cumulative effect,could not alone maintain this incidence in the faceof 5 years (5-20 generations) of natural selectionagainst a somewhat deleterious phenotype and 3crosses to a stock in which chylous ascites has neverbeen seen.

In the latter experiment a normal linkage group Vchromosome is supplied in each generation. Chy-lous ascites must therefore be due to a dominant:either to an imperfectly penetrating dominantclosely linked with Ra, or to an occasional pleio-tropic effect of the dominant Ra itself.

Imperfect penetrance often has an environmental

component. However, a genetic component isstrongly indicated when the penetrance variesaccording to the direction of genetic selection en-forced. This is the case here. For, in the linkagegroup V stock, an increased use of affected parentswas accompanied by an increased incidence ofchylous ascites in ragged progeny; while in the iso-genesis experiment the series in which only affectedparents were used in each generation had a greaterincidence of the condition in the ragged progenythan in the series where only unaffected parents wereused.The present analysis does not disclose where these

modifying loci are situated. But there is a suggestionthat part of the Ra-Sd segment itself may containone or more of them; in the isogenesis experimentwhere the incidence is I7 %, this whole segment,which came from the linkage group V stock, hassurvived intact; whereas in the series with 2%incidence, only the Ra-at segment has been main-tained. This possibility can be further tested in afuller analysis ofthe linkage group V stock where the5 markers of this segment segregate simultaneouslyin all combinations.

It seems very likely that modification of viabilityof the chylous ascites phenotype has accompaniedits increase in penetrance. In both sets of data thechange or difference in penetrance has been greaterthan the change or difference in ragged viability:indeed in the isogenesis experiment where theincidence is I7% there is no loss of raggeds at all,and the condition of the chylous ascites cases hasnot been extreme. In effect, selection operates sothat as more raggeds express the condition thisexpression is such that its deleterious effects arelessened. This course of events, predictable fromthe nature of selective forces, has probably not beendemonstrated before, owing to the lack in othergenetic situations of a fully penetrant gene identi-fying the genotype of the imperfectly penetrantfactor.

It was hoped that the present work might dis-criminate between the hypotheses of pleiotropy ofRa and of adjacent loci. However, as long as pene-trance of chylous ascites is imperfect, only theoccurrence of chylous ascites in a non-ragged wouldgive the preliminary evidence needed for thepossibility of crossing-over. In the present workonly 2 doubtful instances have been reported out ofsome 300 cases of chylous ascites examined.

Nevertheless, there are other considerations thatmake one hypothesis more plausible than the other.The first is the sudden appearance of chylousascites in the Cambridge stocks. As mentionedearlier, until January I958 at least 2,300 raggeds

21

copyright. on N



g.bmj.com

/J M



ber 1964. Dow

nloaded from

http://jmg.bmj.com/


were rigorously examined in the nest, and noneshowed chylous ascites. In the 1,I43 raggedsexanined in the 2 years after that date, the incidenceis i i %, a sudden and very significant increase. Thesecond is the fact that Slee (1957, I962) does notmention cases with milky fluid in the peritoneumand indeed has not seen the condition in any of some5,ooo or more ragged heterozygotes that he hasreared (personal communication).

It seems highly likely that a spontaneous mutationoccurred in I958. Since all the ragged stocks in theGenetics Department show the same hair defectand the same generalized oedema in homozygotes asdo those in Dr Slee's hands, the supposition of aseparate locus to account for the chylous ascitescondition is very plausible. The symbol Chy isproposed.A third consideration is the virtual or complete

absence of chylous ascites in Dr Slee's ragged homo-zygotes. In general, the homozygotes of an im-perfectly penetrating mutant show a very muchhigher incidence and expression of the defectscommon in the heterozygotes, and selection capableof reducing the penetrance in heterozygotes to zerois usually insufficient to reduce the penetrance inhomozygotes to anything like the same extent. Inunselected stocks ragged homozygotes die beforesuckling starts and so cannot show chylous ascites.In Dr Slee's selected stocks, where Ra Ra wereviable until after suckling, it is to be expected that,if Ra itself causes chylous ascites, the incidence ofthe condition, zero in the heterozygotes, might below in homozygotes but hardly zero. However, thisis the case. Of some 2,500 homozygotes, only 4 (i.e.less than o-2 %) showed anything that can be inter-preted as chylous ascites, namely milky fluid in theperitoneum, and these were at that time thought tobe cases of ruptured intestines (personal communi-cation). On the other hand, absence of chylousascites in Ra Ra is wholly to be expected if a secondmutant, Chy, controls the condition, and was absentin Dr Slee's stocks.Even if these 4 cases were indeed affected with

chylous ascites, the hypothesis of a separate locusfor Chy remains very plausible. It has been pointedout above that the oedema ofRa Ra could be due toa defect in the lymphatic system, and that chylousascites certainly is due to a fault in lymphaticdrainage. A very low incidence of chylous ascitesin Ra Ra without Chy would not be unexpected ifthe fault causing oedema were itself an enhancer ofChy. The situation has analogies in other geneticsystems. For example, sporadic cases of poly-dactyly have occurred in a number of stocks of mice.Selection of the residual genotype and of mutants at

the pallid and fidget loci can, however, in thepresence of the gene py homozygously, raise itsincidence to IOO %; if this stock is now made homo-zygous non-py, a low incidence of polydactylyremains (Bodmer, I960). It is thus possible that theRa locus with its oedematous effect is in the samerelation to the gene Chy as are the selected residualgenotype, and pallid and fidget mutants, to the genepy; namely, each can produce a high incidence ofthe defect in the presence of the gene concerned,and a low one in its absence.

If Ra is an enhancer of Chy or even necessary forits expression, then the chance of finding the latterin non-Ra is indeed low. Further genetic investiga-tion, along with further study of correlated defectsof the lymphatic system, is clearly needed for finaldiscrimination between pleiotropy of Ra and aseparate locus Chy. Meanwhile, on the currentevidence, the use of the symbol Chy appears bothjustified and convenient.

SummaryChylous ascites has been observed in i i to I7%

of newborn heterozygous ragged (Ra+) mice in aCambridge stock. The affected mice appearednormal at birth but, after suckling, milky fluidgradually accumulated in the peritoneal cavity. Thefluid was sterile, contained innumerable chylo-microns, and in its lipid composition resembledmilk. The wall of the small intestine was muchpaler and thicker than usual, its submucosa beingbloated with droplets of fat. There was no grosscommunication between the lumen of the alimen-tary tract and the peritoneal cavity. The mesentericlymph vessels were distended and tortuous and someof the mesenteric nodes appeared to be haemalnodes rather than lymph nodes. There was no evi-dence of infection or neoplasia. Many of the micerecovered completely but the more severely affectedbecame too weak to suckle and died. The disorderis believed to be a true chylous ascites due to in-adequate lymphatic drainage from the small intes-tine, but an anatomical explanation for the in-sufficiency has not yet been fully established.On data for Ra from Cambridge and Edinburgh

it appears that while an occasional pleiotropic effectof Ra cannot be completely excluded this inheritedtendency to chylous ascites is probably due to aclosely linked dominant. The symbol Chy is pro-posed. As penetrance of Chy increases, its deleteri-ous effects decrease. There is a possibility that Raenhances penetrance and that there are other modi-fiers in linkage group V.

22

copyright. on N



g.bmj.com

/J M



ber 1964. Dow

nloaded from

http://jmg.bmj.com/


The authors are grateful to Dr A. N. Howard andMr D. E. Boyer for the chemical analysis, to Mrs J. A.Bigelow and Mr J. H. A. Allen for technical assistance,and to Mr J. F. Crane and Mr S. W. Patman forphotography.

REFaRENcS

Bloom, B., Chaikoff, I. L., Reinhardt, W. O., and Dauben, W. G.(I95I). Participation of phospholipides in lymphatic transport ofabsorbed fatty acids. J. biol. Chem., 189, 26I.

-, -, Entenman, C., and Dauben, W. G. (I950). Thequantitative significance of the lymphatic pathway in transport ofabsorbed fatty acids. ibid., 184, I.

Bodmer, W. F. (I960). Interaction of modifiers: the effect of pallidand fidget on polydactyly in the mouse. Heredity, 14, 445.

Bowyer, D. E., Leat, W. M. F., Howard, A. N., and Gresham, G. A.(I963). The determination of the fatty acid composition of serumlipids separated by thin-layer chromatography; and a comparisonwith column chromatography. Biochim. biophys. Acta (Amst.),70, 423-

Munk, I., and Rosenstein, A. (I89I). Zur Lehre von der Resorptionim Darm, nach Untersuchungen an einer Lymph (chylus-) fistelbeim Menschen. Virchows Arch. path. Anat., 123, 230.

Parsons, P. A. (I958). Additional three-point data for linkage groupV of the mouse. Heredity, I2, 357.

Slee, J. (1957). The morphology and development of ragged-amutant affecting the skin and hair of the house mouse. J. Genet.,55, 570.

(1962). Developmental morphology of the skin and hairfollicles in normal and 'ragged' mice. J. Embryol. exp. Morph., Io,507.

Wallace, M. E. (1958). Experimental evidence for a new geneticphenomenon. Phil. Trans. B. 241, 211.

23

copyright. on N



g.bmj.com

/J M



ber 1964. Dow

nloaded from

http://jmg.bmj.com/

chylous ascites mice* - journal of medical genetics · chylous ascites 0 92 14 8i 187 17 3% 49-2%...

Documents