GROWTH A N D DEVELOPMENT OF “ A N T I G E N - F R E E ” G R A F T S O F FOETAL MOUSE INTESTINE*

ANNE FERGUSON AND DELPHINE M. V. PARROTT Department of Bacteriology and Immunology,

Western Infirmary, University of Glasgow

PLATES LII-LIV

THERE is indirect evidence that immune stimulation by normal food antigens affects the mucosal layer of the small intestine in two ways. In suckling and germ-free rodents, not only are there fewer lymphoid cells in the small-intestinal mucosa (Dubos, 1966; CrabbC et al., 1970), but crypt-cell mitotic rates and the rate of epithelial cell loss are slower than in adult, conventionally reared and fed animals (Abrams, Bauer and Sprinz, 1963; O’Connor, 1966).

In the course of investigations designed to examine the nature and functions of lymphoid cell populations associated with the gut, and the effects of immune reactions on intestinal morphology, we found it difficult to interpret results obtained from normal gut of convention- ally reared animals, The large numbers of lymphoid cells already present in the intestinal mucosa made it very dficult to detect minor changes in their numbers or morphology in the experimental situation.

Ideally, one would have preferred to use a completely antigen-free preparation of intestine, but it is only in foetal life that the intestine is truly antigen-free and to maintain this state in the adult it would be necessary to isolate a length of intestine, as a Thiry-Vella loop, before or immediately after birth. Such a procedure is feasible in large animals, but would be very difficult in rats and mice because of the narrow calibre of the gut at birth; we were restricted to the use of rats and mice because of the availability of the inbred strains necessary for cell-transfer experiments. We therefore approached the problem of preparing antigen-free gut in an entirely different way.

It is alreadyknown that endocrine glands (Everett, 1956; Stone andKennedy, 1962) and thymus (Pepper, 1961 ; Parrott and East, 1964) will survive and grow when grafted into heterotopic sites, and there seemed no reason why such a technique should not also succeed when applied to the intestine, especially if the graft was foetal, antigen-free intestine. We found that grafts of foetal mouse intestine did not survive and grow when implanted into the peritoneal cavity within a millipore chamber. However, when fragments of foetal intestine were grafted under the kidney capsule of adult syngeneic hosts the grafts took well, grew and developed normally.

This paper describes and evaluates this simple and reliable technique for the creation of short, “ antigen-free ” segments of intestine in mice. Growth and development of grafts have been assessed morphologically with the dissecting microscope, and histologically. Also, preliminary studies of graft epithelial cell mitotic rate have been carried out by using autoradiography.

Receiued 10 Mar. 1971; accepted 5 July 1971. * Departmental Publication no. 7103.

1. PATH.-VOL. 106 (1972) 95

96 ANNE FERGUSON A N D DELPHINE M. V. PARROTT

MATERIALS AND METHODS

Animals. Mice of inbred CBA and BALB/c strains, maintained in this department, were used throughout these experiments. In most cases CBA foetal gut was grafted into adult CBA recipients, or BALB/c foetal gut was grafted into adult BALB/c recipients (i.e., as iso- grafts). A smaller number of homografts were carried out, CBA foetal gut being grafted into BALB/c recipients. The table gives details of the numbers of isografts and homografts.

Preparation of donor intestine. A mouse, judged to be about 18 days pregnant, was anaesthetised with ether; the abdomen was opened and the foetuses were removed from the uterus (the exact gestational age could be determined by histological examination of one of the foetuses). Each foetus was pinned out on a cork board, the intestinal tract was removed from the abdomen and placed in ice-cold phosphate-buffered saline (PBS) at pH 7.2. The coils of intestine were teased apart and the fragile mesentery was removed. Stomach, small intestine, caecum and colon could be easily identified morphologically. The material to be grafted (usually small intestine) was selected, transferred to another di8h of cold PBS and cut into fragments about 5 mm in length.

Implantation of grafts under the kidney capsule. Recipients aged 3-6 mth were anaes- thetised with an intraperitoneal injection of the neuroleptanalgesic Hypnorm (0.1 mg per 100 g body weight), and light ether anaesthesia was used, if necessary, during the operation. The skin was shaved, the mouse placed on its side on a cork board and immobilised with strips of adhesive tape. A vertical incision was made in the lumbar region and the under- lying kidney gently pulled out of the abdomen and held in place by grasping the lower part of the renal mesentery with the tips of fine Spencer-Wells forceps.

A longitudinal incision was made in the renal capsule with a rounded scalpel blade. A soft polyethylene cannula attached to a syringe (e.g., a Braunula intravenous cannula size 0) was used to lift a piece of donor intestine from the PBS. Surface tension held the graft on the tip of the cannula (fig. 1). The edge of the incised capsule was lifted up with fine forceps and the foetal intestine placed under the capsule (fig. 2). The soft tip of the cannula was used to push the graft well away from the incision. If there was any bleeding from the capsular incision, a small piece of Surgicel was placed over the bleeding point. The kidney was replaced within the peritoneal cavity and the abdominal muscle layer closed with one or two fine silk sutures. The skin incision was closed with a metal clip. The time required for implantation of a single graft was about 5 min.

Cylindrical millipore chambers were constructed from two millipore disks (pore size 0.22 pm) and a plexi- glass ring. A fragment of CBA foetal intestine, with 0.1 mi PBS, was put inside each chamber.

At laparotomy, under Hypnorm and ether anaesthesia, a single millipore chamber, with contained intestine, was placed within the peritoneal cavity of an adult CBA mouse. In all, 17 such grafts were carried out. Recipients were killed 3-18 days later; grafts were examined with the dissecting microscope and by routine histological methods.

Groups of recipient mice were killed at intervals from 1 to 110 days after operation, and the grafts were easily identified as rounded white swellings on the surface of the kidney (fig. 3).

For dissecting microscope examination the grafts, which were adherent to the renal capsule, were dissected free of underlying renal tissue and cut longitudinally. The cheesy material within the graft lumen was removed by shaking the tissue in a beaker of saline. The material was then fixed in formol-saline to which a few drops of crystal violet had been added.

For routine histological examination the grafts, with a small amount of underlying renal tissue, were fxed in formol-saline; 5pm sections were made and stained with haematoxylin and eosin (HE) and methyl green-pyronin (MGP).

Autoradiographic studies. Epithelial cell proliferation kinetics can be studied by means of autoradiography (Leblond and Messnier, 1958). Accordingly, 3H-thymidine labelling of dividing cells and subsequent autoradiography of grafts were carried out in four adult CBA mice, each with two isografts of CBA foetal intestine.

Implantation of grafts in millipore chambers within the peritoneal cavity.

Examination of grafts implanted under the kidney capsule.

FERGUSON AND PARROTT

GRAFTS OF " ANTIGEN-FREE " FOETAL MOUSE INTESTINE

PLATE LII

FIG. 1.4mplantationofagraft underthe kidney capsule of an adult mouse. One edge of an incision in the renal capsule is lifted with fine forceps. Material to be grafted is held by surface tension at the tip of a flexible poly- ethylene cannula. x c. 1.8.

FIG. 2.-The soft cannula is used to push the graft under the renal capsule, well away from the incision. x c. 2.7.

FIG. 3.-Isografts of foetal small intestine in CBA mice. Left, 80 days after implantation; right, 1 hr after implantation. x c. 4.

FERGUSON AND PARROTT

GRAFTS OF ‘‘ ANTIGEN-FREE ” FOETAL MOUSE INTESTINE

PLATE LIII

FIG. 4.-Dissecting-miscroscope appearance of an isograft of foetal small intestine in a CBA mouse; 3s days after implantation. Some debris from the graft lumen is still adhering to the mucosal surface. Crystal violet. x 1 1 .

FIG. 5.--Isograft of foetal small intestine in a BALB/c mouse; 37 days after implantation. Haematoxylin and eosin (HE). x 125.

FIG. 6.-Autoradiograph of an isograft of foetal small intes- tine in a CBA mouse; 38 days after implantation, 2 hr after an intraperitoneal injec- tion of 3H-thymidine. Label- led epithelial cells are present in the crypts. Mucosal lym- phocytes are also labelled. Methyl green-pyronin (MGP). x 365.

FERGUSON AND PARROTT

GRAFTS OF ‘‘ ANTlGtN-FREE ” FOETAL MOUSE INTtSTINE

PLATE LIV

Frti. 7.-Autoradiograph of an iso- graft of foetal small intestine in a CBA mouse; 39 days after implanta- tion, 24 hr after an intraperitoneal injection of 3H-thymidine. Labelled epithelial cells are present in the crypts and in the basaI parts of the villi. MGP. x 365.

FIG. 8.-Isograft of foetal colon in a CBA mouse; 31 days after implantation. HE. x 125.

FIG. 9.-Rejection of a homograft of CBA foetal intestine in a BALBjc mouse; 4 days after implantation. MGP. x 125.

GRAFTS OF '' ANTIGEN-FREE '' FOETAL MOUSE INTESTINE 97

Days after implantation

Thirty-eight days after implantation of the grafts each mouse was given an intraperitoneal injection of 100 pCi 3s-thymidine (Radiochemical Centre, Amersham). Two of the mice were killed 2 hr later; the other two were killed 24 hr after the 3H-thymidine injection.

For each mouse, host intestine and grafts of intestine were fixed in formol-saline,embedded in para& wax and sectioned at 5 pm. The mounted sections were coated with photographic emulsion (Ilford KS) by the " dipping " technique of Kopriwa and Leblond (1962), stored at 4°C for 2 wk and after development stained through the emulsion with methyl green- pyronin.

TABLE Numbers of grafts of foetal mouse intestine examined at different

times after implantation

isografts of small intestine homografts of

- small intestine in millipore under capsule under renal capsule chambers

I Number of

I 1-5 6-10

11-20 21-30 31-40 41-50 51-60 91-110

1 7 9 ... ... ... ... ...

4 3

26 28 18 8

8 ...

10 8

16 6

I6 2 2 4

34 22 16 ... ... ... ... ...

isografts of colon under renal capsule

(CBA+ (BALB/c+ CBA) BALB/c)

I ;: 1 ...

...

...

... ...

RESULTS When the technique was first used there was some mortality due to the

anaesthetic and to bleeding from unnecessarily deep incisions in the renal capsule, which damaged the renal cortex. However, in the last 100 mice operated on (all CBA male recipients) there has been no mortality. At the most recent 50 post-mortem examinations (with bilateral grafts in all animals), 99 established grafts have been found. The number of grafts available for examination at different times after implantation is given in the table.

It is interesting to note that although infection has not occurred when foetal donor tissue is used, on the occasions when intestine was taken from postnatal mice (8 days old) some grafts were clearly infected and histological examination showed abscess formation.

Growth and development of normally sited, small-intestinal mucosa In newborn mice the small intestine is very immature, with low villi and no

well-defined crypts of Lieberkuhn. By 3-4 days an orderly pattern of crypts and villi has been established, with mitosis taking place in the crypts of Lieberkuhn, and newly formed cells are passing from the crypts up the sides

J. PATH.-VOL. 106 (1972) H

98 ANNE FERGUSON AND DELPHINE M. V. PARROTT

of the villi, to be extruded from the villous tip. Goblet cells and Paneth cells are present soon after birth, but lymphocytes and plasma cells do not appear in the loose connective tissue of the lamina propria until 16-18 days, at the time of weaning. At this time, lymphocytes also appear between the epithelial cells that cover the villi. After this there are no further changes in general morphology, although the Peyer’s patches continue to enlarge and germinal centres appear around 5 wk.

Growth and development of isografts of small intestine implanted under the kidney capsule

Immediately after implantation some grafts show small areas of haemor- rhage-presumably due to handling-but at no time have we found evidence of disintegration of the grafts. Four or 5 days after implantation, villi and crypts form, just as in normally sited small intestine, and Paneth and Goblet cells appear. Later, blood vessels and lacteals can readily be identified and smooth muscle develops below the mucosa. Aggregates of lymphocytes, resembling immature Peyer’s patches, also appear in the mucosa, but at no time have we seen germinal centres in the grafts.

Just as in normal small intestine, epithelial cell mitosis is confined to the crypts of Lieberkuhn and epithelial cells are extruded from the villous tips, accumulating as debris in the graft lumen. This debris can be washed away revealing a completely normal appearance of the villi as judged by examination with the dissecting microscope (fig. 4). About 18 days after grafting, lympho- cytes and plasma cells appear in the loose connective tissue of the lamina propria, and some lymphocytes are present between the intestinal epithelial cells covering the villi.

Grafts maintained a normal appearance up to 110 days (the longest time studied) and duodenum (with Brunner’s glands), jejunum (with tall villi and few goblet cells) and ileum (with lower villi and many goblet cells) could all be recognised from their general morphological features. Figure 5 illustrates the typical histological appearance of an established graft.

Examination of the autoradiographs confirmed that epithelial cell mitosis was confined to crypt cells (fig. 6) and that 24 hr later the leading edge of the column of labelled cells had moved to the basal part of the villous layer (fig. 7). However, when the autoradiographs of the grafts were compared with the auto- radiographs of the host mouse’s normally sited intestine two differences were apparent. At 2 hr the dividing crypt cells in the grafts were less heavily labelled than the crypt cells in the host intestine. Also, at 24 hr the leading edge of labelled cells in the grafts had reached only the basal third of the villous layer, whereas in the normally sited intestine it was more than halfway up the villi.

Growth and development of isografts of small intestine in millipore chambers within the peritoneal cavity

Examination of the contents of the millipore chambers showed that none of the grafts had enlarged. At dissecting-microscope examination recognisable

GRAFTS OF " ANTIGEN-FREE *' FOETAL MOUSE INTESTINE 99

" villi " were present in only four of the 17 grafts; all the other chambers contained friable debris. Histological examination confirmed that the millipore chambers had remained intact throughout the experiment, for there was no infiltration of host cells. The areas that had been considered to have " villi " were found, histologically, to consist only of the connective-tissue cores of submucosa and villi.

Growth and development of isografts of colon Grafts of colon also grow and develop normally, but the colonic mucosal

cells secrete mucus into the lumen, so that the graft enlarges to become a cystic swelling, sometimes as large as the kidney itself. However, as can be seen from fig. 8, the morphology of the grafted colonic mucosa is completely normal.

Growth and development of homografts of foetal small intestine When CBA foetal gut was grafted into BALB/c hosts lymphocytes and

pyroninophilic blast cells were present in the grafts from 3 days after implanta- tion (fig. 9), and the mucosa was destroyed by 6-7 days; the smooth muscle of the grafts was also heavily infiltrated with lymphocytes, but retained its general morphological features for several days after the disintegration of the mucosa.

DISCUSSION Recent studies on germ-free animals have illustrated the usefulness of

comparatively antigen-free gut for investigation of gastro-intestinal immunology (Cooper et al., 1968; CrabbC et al., 1969, 1970). However, germ-free animals are neither completely free from extrinsic antigens in food nor readily available.

We have shown that isografts of foetal mouse intestine grow well when grafted under the kidney capsule of adult hosts, though not when implanted in millipore chambers within the peritoneal cavity. This technique is simple, appropriate to conventional laboratory conditions and gives a preparation of intestine that has never been exposed to intraluminal antigens. However, it should be emphasised that the grafts differ from normally sited intestine not only in the absence of antigenic stimuli, but also in that they are not functioning in the absorption of food, are not exposed to bile and pancreatic secretions and do not have normal vagal and sympathetic innervation. Nevertheless, their morphological development parallels that of normally sited intestine of the same age. This includes the early differentiation of Paneth and goblet cells and the appearance of mucosal lymphocytes and plasma cells at a time corresponding to a postnatal age of 16 days. Furthermore, preliminary studies on disaccharidase enzymes in grafts (Ferguson, Russelland Gerskowitch, unpublished) have shown that the graft enzyme contents resemble those of normally sited intestine of the same age.

The rate of graft epithelial cell proliferation would also seem to be a feature inherent in the graft itself, rather than related to the adult host environment.

100 ANNE FERGUSON AND DELPHINE M. V. PARROTT

Our preliminary autoradiographic studies suggest that in respect of crypt- cell mitotic rates, and rates of cell-transit along the villi, the grafts resemble germ-free intestine (Abrams et al., 1963) rather than the host animal's own intestine.

The relatively late appearance of lymphoid cells in syngeneic " antigen- free " grafts is of considerable interest. There can be no question that the renal subcapsular site is in some way immunologically privileged, and that cells have difficulty in gaining access to the grafts, for in allogeneic grafts, pyroninophilic blast cells appeared within 3 days and the grafts were promptly rejected. The fact that lymphocytes, plasma cells and ill-developed Peyer's patches appear at all is intriguing, especially in view of some theories as to the function of the intra-epithelial or " thelio "-lymphocytes (Fichtelius, 1967).

It is anticipated that these grafts will be useful experimental tools, not only for the investigation of such theories but also for studies of cell-traffic in the intestine, of auto-immune reactions in the small intestine and colon and of experimentally induced food allergies.

SUMMARY Short segments of foetal mouse intestine were implanted orthotopically

in adult syngeneic and allogeneic host mice. Isografts grew well and developed normally when implanted under the kidney capsule of adult mice, but not when placed in a millipore chamber within the peritoneal cavity. Homografts of small intestine implanted under the kidney capsule were rapidly rejected.

In subcapsular isografts of small intestine, differentiation of Paneth and goblet cells occurred just as in normally sited intestine of the same age. Lymphocytes and plasma cells were present in grafts from the 3rd wk after implantation and, later, small Peyer's patches, without germinal centres, were found. Crypt-cell mitotic rates and rates of cell-passage along villi were lower in grafts than in the host mouse's normally sited intestine.

This technique gives, in conventionally reared and fed mice, preparations of intestine that have never been exposed to intraluminal antigens; such " antigen-free " grafts are likely to be useful experimental tools.

We are grateful to Mr H. Cairns, who prepared all the histological material. This work was aided by a grant from the Medical Research Council.

REFERENCES AEIRAMS, G . D., BAUER, H., AND SPRINZ, H. 1963. Influence of the normal flora on mucosal

morphology and cellular renewal in the ileum. Lab. Invest., 12, 355. COOPER, G. N., THONARD, J. c., CROSBY, R. L., AND DALBOW, M. H. 1968. Immunological

responses in rats following antigenic stimulation of Peyer's patches. 11. Histological changes in germ-free animals. Austral. J . Exp. Biol. Med. Sci., 46, 407.

CRABBB, P. A., NASH, D. R., BAZIN, H., EYSSEN, H., AND HEREMANS, J. F. 1969. Antibodies of the IgA type in intestinal plasma cells of germfree mice after oral or parenteral immunization with ferritin. J. Exp. Med., 130, 723.

CRABBB, P. A., NASH, D. R., BAZIN, H., EYSSEN, H., AND HEREMANS, J. F. 1970. Immuno- histochemical observations on lymphoid tissues from conventional and gerrn-free mice. Lab. Invest., 22, 448.

GRAFTS OF '' ANTIGEN-FREE " FOETAL MOUSE INTESTINE 101

DUBOS, R. 1966. The microbiota of the gastrointestinal tract. Gastroenterology, 51, 868. EVERETT, J. W. 1956. Functional corpora lutea maintained for months by autografts of

rat hypophyses. Endocrinology, 58,186. FICHTELIUS, K. E. 1967. The mammalian equivalent to Bursa Fabricii of birds. Expl

Cell Res., 46, 231. KOPRIWA, B. M., AND LEBLOND, C. P. 1962. Improvements in the coating technique of

radioautography. J. Histochem. Cytochem., 10, 269. LEBLOND, C. P., AND MESSNIER, B. 1958. Renewal of chief cells and goblet cells in the small

intestine as shown by radio-autography after injection of thymidine H3 into mice. Anat. Rec., 132, 247.

OCONNOR, T. M. 1966. Cell dynamics in the intestine of the mouse from late fetal life to maturity. Amer. J. Anat., 118, 525.

PARROTT, DELPHINE M. V., AND EAST, J. 1964. In The thymus in immunobiology, ed. by R. A. Good and A. E. Gabrielsen, New York, p. 523.

PEPPER, F. J. 1961. Transplantation of thymic tissue in inbred strains of mice. J. Endocr., 22, 349.

STONE, H. B., AND JSENNEDY, W. J. 1962. Survival of heterologous mammalian transplants. Ann. Surg., 155, 623.