Stimulation of cell division in pronephros of embryonic grafts following partial nephrectomy in the host (Xenopus laevis)

Embryonic tissues (Willis, 1962) or whole embryos (Kolodziejski, 1933; Galtsoff & Galtsoff, 1959; Simnett, 1966) grafted into mature hosts of the same species continue to grow and a range of histologically normal differentiated tissues can subsequently be identified. The rate of cell proliferation in certain adult (Goss, 1964; Bullough, 1965) and embryonic tissues (Simnett & Chopra, 1969) is controlled by tissue-specific growth regulatory factors and since the implanted embryo is linked to its host via the vascular supply it is possible that the growth rate of the embryonic tissues may be regulated by the growth factors present in the host circulation. If this assumption is correct the extirpation of a host organ would be expected to result in an alteration of the level of growth regulatory factors which in turn would modify the development and differentiation of the homologous organ in the embryonic implant.

This concept was examined in the present work by implanting embryos into hosts in which the kidney had been partially extirpated and by studying the subsequent development of the kidney in the embryonic implant. It seemed probable that the postulated influence of a partially nephrectomized host on implants would be determined by the time of vascularization of the implant, the pattern of development of the implant kidney and the time of maximum mitotic activity in the regenerating host kidney. Separate experiments were therefore carried out to establish these points.

Juvenile metamorphosed Xenopus laevis (weight 0·6–0·8 g) reared in the laboratory were anaesthetized in 0·1 % MS 222 and a small incision was made in the ventral abdominal cavity overlying the kidneys. Using fine tungsten needles attached to a crystal-controlled diathermy apparatus one of the exposed kidneys was partially extirpated by cauterization without any significant loss of blood. The total amount of kidney tissue thus removed was approximately 40 %. Control animals were subjected to a similar surgical intervention but without any extirpation (sham operation). After recovery from the anaesthetic the animals were maintained at 16–17 °C; postoperative mortality was approximately 10 %.

In order to investigate the mitotic activity in the regenerating kidney both partially nephrectomized and control animals were anaesthetized and injected into the ventral abdominal cavity with the metaphase-arresting agent colcemid (0·03 ml of a 100 mg % solution) 4 h before they were sacrificed. To avoid possible errors due to diurnal variation in mitotic incidence (Bullough, 1948) all animals were killed at the same time of day. The kidneys were fixed in Worcester’s fluid (10 % acetic acid and 10 °/₀ formaldehyde in saturated mercuric chloride), embedded in paraffin wax, sectioned at 5μ and stained in haematoxylin and eosin for histological examination.

Mitotic incidence (MI) was measured by counting the number of arrested metaphases in a sample of 4000–6000 nuclei of kidney tubule cells per animal. A value of MI, expressed as the number of metaphases per 10⁵ cells, was thus obtained for each animal and from these individual values the mean MI and standard deviation for each experimental group (five animals) were calculated. Consequently the MI represents the mean proportion of cells entering mitosis during the 4 h period following the colcemid injection.

Embryos were obtained by the injection of chorionic gonadotrophin into adult male and female Xenopus laevis and reared to the required stage identified according to the normal table of Xenopus laevis (Nieuw-koop & Faber, 1967). Juvenile metamorphosed animals (weight 0·6–0·8g) were used as hosts. Three embryos were implanted into each host—two in the dorsal lymph sac and one in the ventral lymph sac—using the method described earlier (Simnett, 1966). To avoid extrusion of implanted embryos the incisions were closed with two fine silk sutures.

The mitotic activity in the implants was measured at different intervals by injecting the hosts with colcemid 4 h before sacrifice. The implants were fixed in Worcester’s fluid, embedded in paraffin wax, sectioned serially at 7μ and stained in haematoxylin and eosin for histological examination.

The number of arrested metaphases was counted in 1000 nuclei of pronephric tubule cells from each implant. The MI for each implant and for each experimental group was again expressed as the number of arrested metaphases per 10⁵ cells.

In order to investigate the question of tissue specificity following partial nephrectomy of the host similar mitotic counts were made on 1000 nuclei from epidermal cells of each implant.

Partially nephrectomized metamorphosed animals were sacrificed in groups of five at intervals of 2, 4, 6, 8, 10, 12 and 16 days after the operation. Control animals were killed at intervals of 2, 4, 6 and 8 days after the sham operation. Mitotic counts were made in two sample areas from each animal and each estimate of MI in a group of five animals thus represents the mean of ten samples. The data for MI in control and nephrectomized animals are shown in Fig. 1.

During the period of observation there was no significant change in the MI of the kidney in the control animals. The mean MI for the whole period was 235 ± 142/10⁵.

The first change in the MI of the partially nephrectomized animals was noted after 4 days (MI 344 + 105/10⁵). The maximum value of the MI (1000 ± 352/10⁵ or approximately 4 times the control value) was reached on the 6th day. On the 12th day the MI in nephrectomized animals (306 ± 98/10⁵) was still higher than in controls but by the 16th day there was no significant difference between the MI in nephrectomized (239 ± 151/10⁵) and control animals.

These results suggest that following partial nephrectomy changes occur in the level of circulating factors which may determine the rate of mitosis in the kidney. It would seem reasonable to suggest that if these changes have any influence on the renal tissue of implanted embryos the effect would be most obvious during the period of increased mitotic activity in the regenerating host kidney.

Observations were made at three different intervals (4, 8 and 16 days) after implantation of embryos at four developmental stages (stages 12, 32, 38 and 45).Each of the twelve experimental groups contained a maximum of five host animals which received either two or three implants each.

In agreement with a previous study (Simnett, 1966) the majority of implants examined after 4 days were found already to contain blood vessels which were connected to the vascular system of the host. At the end of the experimental period all implants of stages 32, 38 and 45 were present but only a proportion of stage 12 (gastrula implants) was recovered: four out of six in the 8-day group and one out of six in the 16-day group. Since the hosts were checked at regular intervals during the whole of the experimental period for extruded embryos and none were found it was concluded that a proportion of implanted gastrulae was reabsorbed by the host.

Implants of stage 45 were found to contain necrotic areas which were first observed after 4 days and which became progressively more extensive. This observation agrees with previous findings (Simnett, 1966). There was no necrosis at any other stage.

Tissues from stage 32 and 38 implants examined after 4 days were poorly differentiated and still contained abundant yolk platelets but implants examined after 8 and 16 days contained a number of well-differentiated tissues, including pronephric kidney situated at the level of the third somite. The pronephros of stage 38 implants was more compact and had a better developed tubular structure than that found in the pronephros of stage 32 implants. Mesonephros was also present in both stages but was more difficult to recognize because of its dense structural units which resembled gastric glands. In view of this it was decided to confine subsequent observations to the pronephros of stage 38 implants.

Partial nephrectomy is a potentially debilitating operation and it was therefore considered advisable to allow a 2-day recovery period before implantation. Eight days after implantation (or 10 days after nephrectomy) the host kidney, as we have shown above, would still have an increased MI while by 16 days after implantation (18 days after nephrectomy) it would have returned to normal. From a comparison of 8- and 16-day implants it should therefore be possible to decide whether there is any correlation between the MI of host and implant kidney which would be expected if humoral regulatory factors were exchanged via the common circulation established 4 days after implantation.

Twelve metamorphosed Xenopus laevis were partially nephrectomized and, after a recovery period of 2 days, each host received three implants of stage 38, two in the dorsal lymph sac and one in the ventral lymph sac underneath the lower jaw. The implantation procedure was repeated in twelve sham-operated control hosts. Two animals from each group died during the postoperative period. Five partially nephrectomized and five control hosts were sacrificed on the 8th day and the same numbers on the 16th day after implantation. In all cases colcemid was injected 4 h before the animals were killed. After sacrifice of the animals the implants were removed and treated for histological examination in the manner earlier described. In some implants the pronephros was difficult to identify and mitotic counts were therefore made in only ten from the total of fifteen implants recovered from each group of five host animals. Each estimate for MI thus represents the mean of ten samples.

After 8 days the MI in pronephros of implants from nephrectomized hosts (3288 ± 1409/10⁵) was significantly higher (P < 0 ·01) than in that from the control hosts (1589 ± 676/10⁵) but at 16 days no significant difference was found between the two groups (Table 1).

The MI in epidermis of implants from control and nephrectomized hosts removed 8 days after implantation were 5641 ± 809/10⁵ and 5591 ± 1658/10⁵ respectively and for those removed after 16 days the corresponding values were 5287 ± 1658/10⁵ and 4585 ± 2103/10⁵ respectively. There was no statistically significant difference between these values.

In adult mammals the maximum mitotic rate in the compensating kidney usually occurs 2 days after partial nephrectomy (Goss & Rankin, 1960) while in metamorphosed Xenopus maintained at 16–17 °C it was not reached until the 6th postoperative day (Fig. 1). This slower response may be a common feature of amphibian tissues since in the regenerating liver of Xenopus (at 16–17 °C) the maximum MI was observed 6 days after partial hepatectomy (D. P. Chopra, unpublished results) and in Triturus viridescens (at 24 °C) the maximum rate of DNA synthesis was not attained until 10 days after partial hepatectomy (MacDonald, Guiney & Tank, 1962). Although the rate of mitosis in amphibian tissues is temperature-dependent (Simnett & Balls, 1969) the slower mitotic responses mentioned above cannot be ascribed entirely to the fact that the body temperature of Amphibia is lower than that of mammals since very rapid responses have been observed in the compensating tissues of larval Amphibia. For example, the maximum mitotic activity in the pronephros and mesonephros of unilaterally pronephrectomized Xenopus tadpoles maintained at 16–18 °C (Chopra & Simnett, 1969a) and in the injured lens epithelium of Rana catesbiana tadpoles at 30 °C (Reddan & Rothstein, 1966) was already attained on the 2nd and 3rd postoperative day respectively.

In intact Xenopus larvae maintained at 16–17 °C the MI in the pronephros after 4 h of colcemid treatment was 550/10⁵ (Chopra & Simnett, 1969a) and this decreased to 130/10⁵ over a period of 10 days. In contrast, the MI of larvae implanted into non-pronephrectomized hosts maintained at 16–17 °C was 1589 and 3051 /10⁵ (8 and 16 days after implantation respectively). A similar increase was observed in the epidermis where the MI in intact larvae was 1289/10⁵ (authors’ unpublished results) compared with the values of 4000–5000/10⁵ observed in implants. It is not certain whether the higher MI in implant tissues is due to an increased supply of nutritional or hormonal factors or whether a specific disturbance of normal growth control mechanisms is involved.

Following partial nephrectomy in the host an increase in the mitotic rate in the kidney of implanted embryos was observed at the time of increased mitosis in the regenerating host kidney. Similar results were obtained in another series of experiments where the value of MI in the pronephros of experimental implants at 8 days after implantation (2088 ± 629) was significantly higher than that of control implants (740 ± 259) (authors’ unpublished results). No such response occurred in the implant epidermis. The tissue specificity of the response in implants confirms the view that a modification of the growth control system in one of the host organs can produce a specific alteration in the growth rate of the homologous embryonic implant organ. Embryonic implantation may be regarded as a form of artificial pregnancy and this type of experiment raises the important question whether the mammalian foetus is similarly exposed to tissue-specific growth factors present in the maternal circulation or whether it is shielded from their full effect by the placenta. Such a barrier has been discussed by Goss (1963) following his finding that no increased mitotic count was observed in foetal kidney 48 h after maternal unilateral nephrectomy. However, Goss (1963) did not examine the foetal kidney to see if there was any increase in weight and since the mitotic response in embryonic and immature tissues is more rapid than in adult tissues (Chopra & Simnett, 1969a) it is possible that the rate of mitosis in the foetal kidney did increase but that it returned to normal by 48 h. Rollason (1961) observed no increase in foetal kidney weight after maternal total nephrectomy, but this is not surprising considering the severe debilitating effects of the operation. He did, however, report an increase of approximately 50 % in the mitotic count of the foetal kidney 24 h after operation and it has also been shown that the liver of rat foetuses responds to partial hepatectomy of the mother (Ballantine, 1965). This suggests that, if it exists, the placental barrier to growth regulatory substances is incomplete. Unlike the situation in the non-placental vertebrates, where the embryo can be regarded as a self-contained unit from the point of view of tissue-specific growth control, it is possible that in the eutherian foetus growth regulation may, at least in part, depend on maternal factors transmitted via the placenta.

It has been suggested (Goss, 1964; Johnson & Vera Roman, 1968) that growth and mitotic homeostasis in the kidney are controlled by the physiological load placed upon the organ. An alternative hypothesis is that these are regulated by an organ-specific growth inhibitor or chalone (Bullough, 1965) produced in the kidney. Evidence in support of the latter hypothesis is provided by our observations that extracts of kidney from mature Xenopus laevis (Simnett & Chopra, 1969) and rat (Chopra & Simnett, 1969 ó) contain an organ-specific factor or chalone which inhibits mitosis in Xenopus laevis embryonic kidney (pronephros). The increased rate of mitosis in the implant pronephros could therefore be due to a decreased concentration of kidney chalone in the host blood resulting from a reduction in the mass of the tissue which produced the chalone. In a further series of experiments carried out in vivo (D. P. Chopra, unpublished results) injection of rat kidney extract caused a tissue-specific mitotic inhibition both in the regenerating kidney of partially nephrectomized, metamorphosed Xenopus and in the pronephros of embryos implanted in such hosts. It would appear that the chalone mechanism is involved not only in the mitotic homeostasis of mature tissues but also in the regulation of embryonic growth.

It has been demonstrated (Clayton, 1953) that in amphibian embryos some organ-specific antigens appear during gastrulation before the morphological differentiation of the corresponding organs. Similarly, it has been shown (Flickinger, 1962) that in chick and frog embryos proteins with the immunological characteristics of adult lens antigens appear before the differentiation of the embryonic lens. Consequently, it is possible that certain of these organspecific antigens may act as growth control factors.

Après néphrectomie partielle de Xenopus laevis récemment métamorphosés, l’activité mitotique du rein en régénération atteint son maximum au 6e jour et retourne à son niveau normal le 16e jour.

L’activité mitotique a été mesurée dans le pronephros et l’épiderme de larves de Xenopus (stade 38) à différents intervalles après leur implantation dans les sacs lymphatiques d’hôtes métamorphosés témoins et d’hôtes partiellement néphrectomisés.

Dix jours après néphrectomie partielle, pendant la période d’activité mitotique accrue dans le rein en régénération de l’hôte, l’activité mitotique du pronephros implanté est deux fois plus élevée que dans le pronephros implanté dans les hôtes témoins.

Dix-huit jours après néphrectomie partielle, quand l’activité mitotique du rein en régénération de l’hôte est redevenue normale, il n’y a pas de différence entre l’activité mitotique du pronephros des implantais dans les hôtes néphrectomisés et dans les hôtes témoins.

Il n’y a pas de différence significative entre l’activité mitotique de l’épiderme des implantais dans les hôtes témoins ou néphrectomisés, ni dans celle de l’épiderme d’implantats examinés 10 et 18 jours après la néphrectomie de l’hôte.

On en conclut que le (ou les) facteur circulant responsable du contrôle de l’activité mitotique du rein en régénération de l’hôte, pénètre dans l’implantat par ses voies vasculaires et influence l’activité mitotique du tissu embryonnaire homologue. Il est possible que ce facteur soit un inhibiteur mitotique, tissu-spécifique, synthétisé par le rein de l’hôte.

This work was supported by a grant from the North of England Council of the British Empire Cancer Campaign for Research. Thanks are due to Professor A. G. Heppleston for kindly providing research facilities in his Department, to Mr and Mrs H. Elliot for skilled technical assistance and to Mrs M. Jackson, our Editor of Research Publications, for valuable editorial help.