Pregnant guinea-pigs were exposed to an environmental temperature of 42·0–42·5 °C for 1 h on day 21 of gestation. Their embryos were removed at periods from 45 min of heating to 48 h following exposure. Histological preparations of embryos showed clumping of nuclear chromatin and subsequent death of cells which were at about the stage of mitosis. Affected cells were particularly numerous in the central nervous system. Further mitotic activity was inhibited for 6–8 h. Squash preparations of the telencephalon at 1 h after heating showed an increase from 3 to 86 % in the number of mitotic cells showing damage in the form of nuclear clumping; this number fell progressively to 30% by 24 h after heating. The proportion of cells in various stages of mitosis changed considerably at 1–8 h after heating, but had returned to pre-heating values by 24 h. The proportion of cells in prophase fell markedly, while the proportion of metaphase cells was doubled at 4 h after heating, indicating blocks to the cell generation cycle before prophase and in metaphase.
Newborn guinea-pigs which have been exposed to maternal heating during days 20–23 of gestation usually have a number of serious developmental defects, including severely retarded development of their brains and less severely retarded general bodily development (Edwards 1969 a, b). The smaller brains of heated young contain fewer cells and probably fewer neurones (Edwards, Penny & Zevnik, 1971) than control animals. The deficit in brain development is not compensated for in post-natal growth, and at maturity these micrencephalic guinea-pigs perform poorly in learning experiments, compared with control animals (Lyle, Jonson, Edwards & Penny, 1973). Both at birth and at maturity, the shape of the smaller brain is generally normal, and the histological appearance does not show any marked deviation from that of controls.
The present experiments were designed to observe the effect of heat on the developing guinea-pig embryo, and particularly on the dividing cells of its central nervous system soon after the heat stress was applied. Preliminary observations showed that after a day or more post-heating few significant changes were to be found, so examinations were concentrated within the first few hours after heating.
MATERIALS AND METHODS
Details of the methods used in determining the stage of gestation and the induction of hyperthermia have been described elsewhere (Edwards, 1967, 1969a). On day 21 of gestation guinea-pigs were exposed for 1 h in a forceddraught egg-incubator set at 42 · 0 – 42 · 5 °C dry bulb and 23 · 0 – 27 · 5 °C wet bulb. In the histological study some embryos were also recovered after 45 min of maternal heat exposure. Deep rectal temperatures were recorded immediately before and after exposure. The mean elevations in body temperature of the heated groups were 3 · 1 – 4 · 0 °C. Embryos were obtained from control mothers and from mothers after 45 min of exposure, or at short intervals between 0 and 48 h following exposure. The mothers were anaesthetized by pentobarbital sodium solution given by intraperitoneal injection, and the embryos were placed immediately in Bouin’s fixative. After 24 h they were washed in tap water and stored in 70 % alcohol. Sections were cut at 6 μm and stained with haematoxylin and eosin or cresyl violet.
Brain squash preparations were made by obtaining embryos at 21 days of gestation as before, immersing them in isotonic citrate and quickly excising the walls of the telencephalic vesicles. The excised portions were cut into pieces of less than 1 mm diameter, fixed in methanol: acetic acid (3:1) for 10 min and stained on a glass slide with lacto-acetic orcein for 15 min. They were then squashed by firm pressure under a glass coverslip, which was sealed to the slide with molten paraffin.
Embryos were sampled from control mothers and from groups of heated mothers at 1, 4, 8, 12,16 and 24 h after the end of heat exposure. There were five mothers in each group and one embryo was examined from each mother. A number of squash preparations was made, and a total of 500 mitotic nuclei counted from each embryo. Nuclei were classified as prophase, metaphase, anaphase, telophase or damaged.
Sections of control 21-day guinea-pig embryos showed many cells in mitosis in the ependyma adjacent to the lateral, third and fourth ventricles of the brain, and adjacent to the central canal of the spinal cord (Figs. 1, 2). The mitotic cells were not evenly distributed throughout the ependyma. They appeared most numerous around the lateral ventricles and in the dorsal part of the ependyma of the spinal cord. Fewer were found in the third and fourth ventricles, the retina adjacent to the choroid, in the cranial and spinal ganglia, and fewer still in sites other than the nervous system.
Cellular abnormalities in the nervous system of heated embryos were most numerous in the regions corresponding to the areas which had shown the most marked proliferative activity in control embryos.
Some cells adjacent to the ventricles in the developing ependyma of the telencephalon, mid-brain and spinal cord showed a clumping of nuclear chromatin after 45 min of exposure to heat (Fig. 3). The nuclei of these cells later became irregular and pyknotic, and their cytoplasm became eosinophilic and then vacuolated (Fig. 4). By 6 – 8 h after the end of heat exposure the nucleus showed karyorrhexis (Fig. 5). At this stage a small number of cells situated some layers deep in the mantle zone, and well away from the ventricular lining, also showed pyknotic changes in the nucleus (Fig. 6). Some mesenchymal cells were seen with similar pyknotic changes, or with two or more nuclei or nuclear particles.
Normal mitotic figures were infrequently found in the ependyma until about 6 h after the end of heat-stress. After this period, numerous mitotic cells were found both in the ependyma lining the ventricles, and (Fig. 6) some were also found in the mantle layer.
A similar series of changes occurred in some nuclei of the proliferative zone of the retina, in the cranial and spinal ganglia and in the developing ear.
By 24 h after exposure, much of the nuclear debris had disappeared, but some small pieces of chromatin-like material could usually be found about 3 – 4 cells deep in the mantle zone. At this stage mitotic activity appeared to be proceeding briskly.
As the histological study revealed many cells with clumping of chromatin in the ependymal layer lining the ventricles, nuclei with this appearance in squash preparations were assumed to be at about the stage of mitosis, and were classified as damaged mitotic nuclei. In the control embryonic brains about 3 % of nuclei were damaged. The proportion was increased to 86 % 1 h after heat exposure, and progressively decreased to 30% at 24 h (Table 1). The apparently undamaged nuclei, classified according to stage of mitosis, showed a marked decrease in the proportion of prophase figures, particularly at 4 h, and a decrease of anaphase and telophase nuclei 1 h after exposure (Fig. 7). However, the proportion of metaphase figures doubled at 4 h. By 24 h the proportions of all phases of mitosis had returned to control levels.
Figure 8 shows the appearance of normal metaphase chromosomes in a squash preparation from a control embryo. The appearance of the damaged cells in squash preparations resembled that in the histological sections, and is illustrated in Figs. 9 – 12. There were few apparently normal mitotic figures until 6 h after the end of heat-stress. By 8 h there were numerous chromatin-like fragments amongst many apparently normal mitotic nuclei. A number of chromatin-like fragments were still present 24 h after the end of heating. In some squash preparations, clumped chromatin appeared to be extruded through the nuclear membrane leaving ‘ghost’ nuclei containing no chromatin-like material.
The results from the present study indicate that both cell death and delay in mitosis could contribute to the deficit of cells in the brains of newborn guineapigs following heating during early embryonic development.
Most cells which were killed were situated in the ependyma adjacent to the ventricles, and appeared to be at about the stage of mitosis. In the developing brain, mitosis normally occurs only at this site (Fujita, 1960; Berry, Rogers & Eayres, 1964). Cells with clumped chromatin were found in embryos exposed to 45 min of maternal heating. The prevalence of these cells appeared to increase up to 1 h after exposure and then to diminish up to 6 – 8 h after the end of exposure. At this stage, a small number of cells with clumped chromatin or groups of chromatin fragments were found about 7 – 8 cells deep in the mantle layer of the brain. Whether these lesions represent cells which have died in an intermitotic phase subsequent to heating, or represent cell remnants following heat damage at about mitosis, is uncertain.
Sections from embryos sampled at the end of the heat-stress period were usually devoid of cells in mitosis and further mitotic activity was inhibited for 6 – 8 h. This period of inhibition appeared to be followed by a period of intense mitotic activity, which might represent a degree of synchronization of mitosis - which can be induced by heat-shocks in cultures of unicellular organisms (Scherbaum & Zeuthen, 1954) or, less efficiently, with cold-shocks in mammalian cells (Newton, 1964).
The relative contribution of death of cells at about mitosis, and of inhibition of mitosis, to the cell deficit which exists in the brain at birth is difficult to assess from the data available, as it is possible that the apparent burst of mitotic activity 6 – 8 h after the end of heating might compensate for the period of inhibition.
If it is assumed that most cells which were injured will show clumping of chromatin by 1 h after heating, and that few additional cells will show these changes after 1 h, the data of Fig. 7 indicate that the cell generation cycle is blocked before prophase and in metaphase. As the proportion of cells in prophase falls sharply by 1 h and is halved by 4 h, the block before prophase appears to be caused by some interference with a process occurring in the G 2 phase and perhaps in the S phase.
The block occurring in metaphase might result from a denaturation of the spindle. A reversible change in the nature of the spindle of chick embryo cells has been produced by exposure to incubation temperatures which were higher than normal (Lewis, 1933). Although it would be convenient to propose simply that the clumping of chromatin found in cells at about mitosis might result from such a denaturation of the metaphase spindle and collapse of the unsupported chromosomes, this mechanism does not appear to be likely in all cases, as cells in metaphase do not appear to be the only ones involved. For a period after the heat treatment, a nuclear membrane still enclosed the clumped chromatin of some affected cells. The nuclear membrane normally disappears toward the end of prophase and is re-formed during telophase (Mazia, 1961).
Fertilized sea-urchin eggs (Psammechinus miliaris) heated to temperatures which block cell division (25 – 27 °C) show a nuclear cycle in which the nuclear envelopes dissolve, but chromosomes do not split into chromatids and a single nucleus is reconstituted. Similar irreversible damage to mitotic cells occurs also in Limnaea and Schizosaccharomyces. It was considered that the damage to cells in mitosis and the set-back of interphase cells of these cultures were both due to damage to microtubules (Zeuthen, 1972). Cultures of Tetrahymena pyriformis can be prevented from entering mitosis by heat-shocks induced by changing the temperature of incubation from the optimum of 28 – 29 °C to a sublethal temperature, 32 – 34 °C (Scherbaum & Zeuthen, 1954). Rao & Engelberg (1966) found that Hela cells stopped dividing immediately after a change in incubation temperature from 37 to 41 °C. There was also a lag of 4 h, during which no divisions occurred, after a shift from 34 to 37 °C, which suggests that the relative change in temperature might be important in the delay imposed on mitotic activity.
Many other possible causes of the delay in mitosis following heat treatment of cultures have been advanced, including disintegration of the many preparatory processes required for division (Prescott, 1961); the inactivation of a specific division-linked protein (Scherbaum, 1963; Zeuthen, 1963); transitions in the state of intracellular water or changes in the properties of lipid membranes (Rao & Engelberg, 1966); prevention of synthesis and destruction of divisionassociated RNA (Moner, 1967) and ultrastructural lesions affecting the nucleolus (Amalric, Simard & Zalta, 1969; Simard, Amalric & Zalta, 1969; Love, Soriano & Walsh, 1970). However, heat-induced cell death is popularly believed to be due to protein denaturation. Rosenberg, Kemeny, Switzer & Hamilton (1971) showed a good numerical correlation between thermodynamic parameters of protein denaturation and cell death rates. Westra & Dewey (1971) believed that heat inactivation of cells results primarily from denaturation of proteins. They showed cultures of Chinese hamster ovary to be most sensitive to the lethal effects of heat during the S phase and mitosis. It was recently calculated that 0 · 2 % of cells cultured from lungs of Chinese hamsters and incubated at 37 °C were irreversibly lost from the proliferative population as a result of heat injury. The theoretical upper temperature limit for the growth of these cells was found to be 40-6 °C, and rapidly proliferating cells were more sensitive than slowly proliferating cells (Johnson & Pavelec, 1972).
The set-back to interphase cells by heat-shock might represent a mechanism of some importance in the protection of a proliferative cell population against the lethal effects of heat. The damage produced by moderate elevations of temperature during interphase appears to be largely reversible and it prevents cells from entering mitosis when this increased temperature might cause irreversible damage.
This study was supported by grants from the Wellcome Foundation, the Australian Research Grants Committee and the World Health Organization.