AUTORADIOGRAPHIC CONFIRMATION OF THE MITOTIC DIVISION OF EVERY MOUSE JEJUNAL CRYPT CELL LABELLED WITH 3H-THYMIDINE

The question was investigated of whether for crypt epithelia of the jejunum of the mouse all cells labelled after a single injection of ³H-TdR subsequently divide or whether cells exist in the crypt which synthesize metabolic DNA and, therefore, do not undergo division after labelling.
A double labelling experiment was performed with a first injection of ³H-TdR followed 1 hr later by an injection of ¹⁴C-TdR. Then from double emulsion autoradiographs of isolated squashed crypts the number of ³H-only, ¹⁴C-only and double labelled cells and mitoses were counted.
The double labelling produced a narrow, 1 hr wide sub-population of ³H-only labelled cells. This subpopulation of S cells completed its division before labelled cells were lost from the crypts by migration onto the villi. The results showed that this subpopulation of ³H-only cells completely doubled within 3 hr and then remained constant through 6 hr. From this result it was concluded that every cell labelled after a single injection of ³H-TdR divides.
From the same autoradiographs the flow rate through the end of mitosis was measured. From the flow rate and the mitotic index a mitotic duration of 0·5 hr was determined. The agreement of this measured mitotic time with the value calculated from the labelling index, mitotic index and S duration is also strong evidence that every labelled cell divides.
Both experiments show that the intestinal crypt does not contain cells synthesizing metabolic DNA.     

CELL POPULATION KINETICS IN THE RAT JEJUNAL CRYPT

Cell kinetics in the jejunal crypt of the male Wistar rat were studied using autoradiographic techniques with tritiated thymidine and a stathmokinetic technique with vincristine. The migration rate measured by following the movement of the 50 % peak on the labelling index distribution curve with time after injection of tritiated thymidine gave a value of 1.43 ± 0.14 (SE) cell positions per hour, compared with a value from a cumulative birth rate of 1.78 cell positions per hour. The crypt column length was 32.9 ± 0.2 cells and the column count was 22.3 ± 0.2. This measurement gave a total crypt population of 734 cells, compared with an estimate of 650 ± 6 from direct observation of squashed, microdissected crypts. In each crypt 22.5 ± 0.5 mitoses were present, and the crypt cell production rate was 32 cells per crypt per hour; this latter value was confirmed using two independent techniques. The crypt growth fraction calculated from the durations of phases of the cell cycle and the labelling index was 0.62. A value of 0.61 was found from the labelling index distribution curve. As assessed from crypt squashes, there were 403 proliferating cells per crypt.     

ALTERATION OF NEONATAL RAT PAROTID GLAND ACINAR CELL PROLIFERATION BY GUANETHIDINE-INDUCED SYMPATHECTOMY

Newborn rats were injected with guanethidine-sulfate (20 μg/g body weight) every 48 hr from 12 hr after birth until day 14 (eight injections per animal). The guanethidine treatment resulted in an 86% absolute reduction in cell number in the superior cervical ganglia of 15 day old rats. The cells which remained after guanethidine treatment showed destruction of mitochondria and an extensive decrease in endoplasmic reticulum. Chemical sympathectomy with guanethidine induced a 3.1 hr lengthening of the acinar cell generation cycle time (17.4 hr to 20.5 hr), resulting from a longer G₁ period (6.9 hr in the control group as compared to 10.5 hr in the guanethidine-treated group), as well as a decrease in the mean percentage of [³H]thymidine-labeled acinar cells (22.3 ± 0.5% to 19.3 ± 0.5%) and mean acinar cell mitotic index (2.6 ± 0.2% to 2.1 ± 0.1%). A circadian rhythm was found to exist in parotid gland acinar cell mitotic activity of 15 day old rats and the amplitude of the rhythm was reduced from 26.5% to 14.9% in guanethidine-treated rats. This study indicates that the diminution of sympathetic influence on the developing parotid gland results in a slight, but significant alteration in acinar cell proliferation.     

AUTORADIOGRAPHIC INVESTIGATION ON THE CELL KINETICS OF CRYPT EPITHELIA IN THE JEJUNUM OF THE MOUSE °CONFIRMATION OF STEADY STATE GROWTH WITH CONSTANT FREQUENCY DISTRIBUTION OF CELLS THROUGHOUT THE CYCLE

Cell kinetic parameters of cells in the crypt of the jejunum of the mouse were obtained autoradiographically. A number of different methods used in cell proliferation studies were applied to the same animal strain kept under constant conditions. In order to avoid effects of geometrical factors, squashes of isolated crypts were used.
The generation time was determined by the per cent labelled mitoses method of Quastler, modified by double labelling with ³H- and ¹⁴C-TdR. This modified method permits a more exact determination of the generation time. The duration of the cycle was 14 hr.
Double labelling experiments in which an injection of ³H-TdR was followed by an injection of ¹⁴C-TdR after 1 hr showed that the cell flux was 7.0%/hr at the beginning of the S-phase and 7.68%/hr at the end. Assuming steady state growth a constant cell flux of 7.15%/hr within the whole cycle can be derived from the measured generation time of 14 hr. These results clearly show that the crypt epithelia constitute a steady state system with constant frequency distribution of the cells throughout the cycle.
The per cent labelled mitoses method after a single injection of ³H-TdR as well as double labelling experiments with ³H- and ¹⁴C-TdR give an estimate of the S-phase of 8.0 or 7.4 hr respectively. Double determinations lead to a value of 0.54 or 0.52 hr respectively for the duration of mitosis and to values of 77% and 72%     

Visualization of chromosome assembly during the S and G2 stages of the cycle and chromosome disassembly during the G1 stage in semithin sections of mouse duodenal crypt cells and other cells

Previous examination of dividing cells in the isthmus of the mouse pyloric antrum by using semithin (0.5-micron-thick) Epon sections revealed that the prophasic condensation of chromosomes began early in the DNA-synthesizing (S) stage. In order to examine whether the same observation could be made in other proliferating cell types, the crypt base columnar cells in mouse duodenum and the hepatocytes of the rat 48 hr after partial hepatectomy were investigated by morphologic and radioautographic techniques. When crypt base columnar cells were studied in semithin Epon sections, the four phases of mitosis showed the characteristic features described by classical cytologists. Moreover, the proportion of cells in prophase and telophase was high. To relate the mitotic phases to the stages of the cell cycle, the "frequency of labeled mitoses method" provided the duration of the cell cycle, 12.3 hr, and of the S stage, 7.3 hr. From the frequency of the occurrence of mitotic phases, it was estimated that metaphase lasted 0.3 hr and anaphase 0.11 hr, in line with previous estimates. However, the durations of prophase and telophase were long, 5.9 and 1.9 hr, respectively. The whole mitotic process took over 8 hr. From the duration of prophase and cycle stages, it was calculated that 67% of the S stage was occupied by prophasic cells. In fair agreement with this estimate, 68% of the labeled cells 10 min after a 3H-thymidine injection were found to be in prophase. In regenerating hepatocytes, the morphological features and frequency of prophase and telophase cells were similar to those observed in duodenal crypt cells. While the cycle time was not measured and, therefore, the duration of cycle stages and mitotic phases could not be estimated, it is likely that their duration would be of the same order of magnitude. In conclusion, the mitotic process in duodenal crypt cells takes over 8 hr. Moreover, the crypt cells, like antral isthmal cells, show features of early prophase soon after they enter the S stage of the cycle.     

Comparative study of the level of mitotic activity and duration of mitosis over a 24 hour period in mouse tumors and normal tissue]

Circadian rhythm of cell division in the forestomack epithelium proved to be largely similar to that in the transplantable (continuous) carcinoma of the forestomack; the duration of mitosis in these tissues changed in the course of 24-hours. The mean 24-hour mitotic activity in the tumour was double that in the forestomack contrary to this, colchamine (colcemide) accumulated in the course of the 24-hours 121.1% mitoses in the forestomack and 83.8% mitoses in the carcinoma. A greater number of mitoses in the tumour with the usual count is attributed to the fact that the mean 24-hour duration of mitosis of the remen carcinoma was 2.7 greater than in the epithelium of the forestomack.     

ALTERATION OF CELLULAR PROLIFERATION IN THE ILEAL EPITHELIUM OF SUCKLING AND WEANED RATS: THE EFFECTS OF ISOPROTERENOL

The purpose of this study was to analyse cell proliferation in the ileum before and after neonatal closure to macromolecules and to determine the effects of the sympathomimetic amine, isoproterenol (IPR), on cell cycle parameters. 8- and 28-day-old rats were employed in this study representing the neonatal periods of pre- and post-closure ileum respectively. The duration of the cell cycle phases was determined by the percentage of labeled metaphases technique (PLM) with computerized analysis of the curves. The generation cycle time was longer in 8-day-old suckling rats (18.63 hr) as compared to the older weaned rats (11.85 hr), and most of this 6.78 hr difference was in the G₁-period (4.5 hr). Other proliferative indices were also lower in the suckling rats—mitotic index (1.9 ± 0.4% as compared to 6.7 ± 0.9%), labelling index (26.8 ± 2.5% versus 44.2 ± 2.6%) and migration rate measured as per cent labeled villus cells (8.9 ± 3.1% versus 45.2 ± 3.4%). IPR was found to inhibit cellular proliferation in the ileal epithelium of both age groups. The cell cycle of the ileal epithelium of 8-day-old rats was lengthened from 18.63 to 21.07 hr and 28-day-old rats from 11.85 to 13.98 hr. IPR produced a decrease in mitotic index from 1.9 ± 0.4% to 1.6 ± 0.4% in pre-closure ileum and from 6.7 ± 0.9% to 5.1 ± 0.6% in post-closure ileum. Labeling index decreased from 26.8 ± 2.5% to 20.0 ± 2.0% in 8-day-old rats and from 44.2 ± 2.6% to 31.1 ± 3.0% in 28-day-old rats after IPR administration. There were also significant differences in growth fraction between age groups and a significant decrease in growth fraction after IPR-treatment. From the results of this study it appears that β-adrenergic stimulation has an inhibitory effect on neonatal ileal epithelium.     

Radioprotection to small intestine of the mice against ionizing radiation by semiquinone glucoside derivative (SQGD) isolated from Bacillus sp. INM-1

Ionizing irradiation induces severe damage to the intestinal crypt cells which are responsible for renovation and maintenance of the intestinal cellular architecture. Therefore, protection of intestinal cells and tissue against lethal irradiation using a semiquinone glucoside derivative (SQGD) isolated from radioresistant bacterium Bacillus sp. INM-1 is the prime focus of the present investigation. BALB/c mice were administered by SQGD (50?mg/kg.b.wt. i.p.) 2?h before whole body irradiation (10?Gy), and histological analysis of the jejunum section was carried out and compared to the irradiated mice. Significant (p?<?0.0001) increase in villus length, number of cells per villus, crypts numbers per villus section, total cells counts and mitotic cell counts per crypt and low goblet cells per villus section, and low apoptotic index per crypt section were observed in the irradiated mice pre-treated by SQGD at 48–168?h. Significant induction in NF-kβ at 24?h and Bcl-2/Bax ratio was observed in irradiated mice pre-treated by SQGD compared to only irradiated animals. SQGD pre-treatment before irradiation was found instrumental to reverse the radiation-induced degenerative changes by replenishment of the damaged cells by enhancing mitotic, proliferating, pro-survival, and apoptosis inhibitory activities probably through modulation of cell cycle arrest in G₁/S phase in the intestinal cellular milieu.     

Nycthemeral variations of mitotic activity in the adrenal cortex of the rat]

Working on histologic sections 7 microns thick in the equatorial part of male rat adrenal cortex, we found two high mitotic activity periods. The first one stays at midnight and shows an average number of 42 mitoses per section; the second one occurs at 5 a.m. with an average number of mitoses of 70,1 per section. If we except these particular points, the mitotic activity is low, the average number of mitoses being 13,2 per section.     

Intestinal crypt proliferation. II. Computer modelling of mitotic index data provides further evidence for lateral and vertical cell migration in the absence of mitotic activity

The position-dependent mitotic index before, and 1, 2 and 3 h after vincristine was scored. The accumulation of cells in mitosis leads to an increase in the mitotic index from 0.06 to 0.34 at crypt positions 8-12. Surprisingly, the leading edge of the position-related mitotic index distribution moves to higher crypt positions although cell division was stopped. In addition, the vertical clustering of mitotic figures in sections was recorded. The data were examined using a previously described computer crypt model. We conclude: the average mitotic phase duration is about 0.7 h (40 min) and varies little with cell position; the geometrical correction factor for overscoring mitoses in crypt sections is about 0.6-0.7 and adjacent cell columns can merge. Lateral cell displacement after mitosis, as predicted in a previous model analysis, would be a mechanism to counteract other forces that tend to reduce the crypt circumference. In the normal steady state merging and expansion processes would just balance each other. This would not follow if one mechanism was blocked. Thus we propose a new concept in which the crypt geometry would be dynamically determined by cell proliferative activity in connection with lateral positioning of new cells on one hand and contracting forces on the other hand.     

CELL POPULATION KINETICS OF PLUCKED AND UNPLUCKED MOUSE SKIN I. UNIRRADIATED SKIN

A detailed study of the cellular proliferation kinetics in interfollicular plucked and unplucked mouse skin has been made in Swiss albino mice, using tritiated thymidine autoradiography. Diurnal variations in mitotic and labelling indices were demonstrated in both systems.
The mean cell cycle times for unplucked and plucked skin were estimated by four different methods and found to be 100 ± 10 and 47 ± 3 hr respectively. Most of the difference was due to the shortening of G₁ phase after plucking. Repeated labelling at intervals shorter than the DNA synthesis times resulted in all the basal layer cells becoming labelled, so that the growth fraction was unity, in unplucked and plucked skin.
A well-defined second wave of labelled mitoses was seen at about 100 hr after labelling the unplucked (i.e. normal) mouse skin.
A double labelling technique using ¹⁴C-TdR and ³H-TdR with a single layer of emulsion gave reasonable values for the duration of the DNA synthesis phase.     

Antimitotic activity of EA21b mammary-carcinoma extract

Many tumors produce factors that affect cell-cycle and cell proliferation. In the present study we have analyzed the effect of a mammary-tumor extract injection on the mitotic activity of several organs in young male C3H/S mice previously standardized for circadian periodicity. One-half of the animals received an intraperitoneal EA21b tumor extract dose at 16:00 h, while the other half received saline. Animals were sacrificed on the following day at 08:00, 12:00 or 16:00 h. 4 h after receiving an injection of colchicine by the same route. Samples of duodenum, kidney, liver, and submaxillary gland were excised and processed for hematoxylin-eosin staining. Mitotic indices, expressed as the number of colchicine-arrested metaphases per 1,000 nuclei, were assessed in convoluted tubule epithelium, duodenal crypt enterocytes, hepatocytes and submaxillary gland ductal and acinar sialocytes. All values were expressed as mean ± SEM. Statistical analyses were performed by ANOVA, Bonferroni and Student’s t-tests. In contrast to the mitotic indices reductions observed in renal convoluted tubules cells and duodenal crypt enterocytes, neither the submaxillary gland nor the liver were found to contain cell types whose mitotic activity was affected by the tumor extract. We conclude that EA21b mammary carcinoma contains one or more factors that inhibit the proliferation of selected populations of normal cells.     

Intestinal Crypt Proliferation. Ii. Computer Modelling of Mitotic Index Data Provides Further Evidence For Lateral and Vertical Cell Migration In the Absence of Mitotic Activity

Abstract. The position-dependent mitotic index before, and 1, 2 and 3 h after vincristine was scored. the accumulation of cells in mitosis leads to an increase in the mitotic index from 0.06 to 0.34 at crypt positions 8-12. Surprisingly, the leading edge of the position-related mitotic index distribution moves to higher crypt positions although cell division was stopped. In addition, the vertical clustering of mitotic figures in sections was recorded. the data were examined using a previously described computer crypt model. We conclude: the average mitotic phase duration is about 0.7 h (40 min) and varies little with cell position; the geometrical correction factor for overscoring mitoses in crypt sections is about 0.6-0.7 and adjacent cell columns can merge. Lateral cell displacement after mitosis, as predicted in a previous model analysis, would be a mechanism to counteract other forces that tend to reduce the crypt circumference. In the normal steady state merging and expansion processes would just balance each other. This would not follow if one mechanism was blocked. Thus we propose a new concept in which the crypt geometry would be dynamically determined by cell proliferative activity in connection with lateral positioning of new cells on one hand and contracting forces on the other hand.     

DNA synthesis, cell division and specific cytodifferentiation in cultured pea root cortical explants

Pea root segments cut 10–11 mm behind the tip of germinating seedlings were prepared by removal of the central cylinders with a tissue punch. These cortical explants were cultured aseptically on nutrient medium containing auxin with and without added cytokinin. In the absence of kinetin, the cortical cells enlarged and separated but failed to show DNA synthesis, mitosis, cell division or subsequent cytodifferentiation. In the presence of 1 ppm kinetin, cortical nuclei showed ³H-thymidine incorporation beginning between 24 and 32 hr; mitoses began about 48 hr, reaching a maximum of 6% at 60 hr. From an initial number of 8000 cells per segment, the cell count increased to 37,000 by day 7 and 140,000 by day 21. At the outset all mitoses were tetraploid; with time the proportion of tetraploid mitotic cells decreased and an octaploid population increased. A frequency of less than 10% diploid mitoses was observed after day 5. Only 25% of the cortical cells showed initial labeling. Beginning on day 7 tracheary elements differentiated from cortical derivatives. By day 14 about 25% and by day 21 about 35% of the total cell population had formed tracheary elements. As a system for analysis in biochemical and cytological terms, pea cortical explants represent an excellent system for the study of cytodifferentiation.     

Megakaryocytopoiesis in the mouse: Response to varying platelet demand

The response of murine megakaryocytopoiesis was studied under conditions of varying platelet demand. Twenty-four hours after mice were given a single injection of rabbit anti-platelet serum, megakaryocyte number and volume were increased, becoming maximal at 65 and 40 hr, respectively. Total body megakaryocytic colony-forming unit (CFU-M) numbers did not change until 90 hr, when a 35% increase in the experimental group was noted. The percentage of CFU-M in DNA synthesis in the experimental group was 38 ± 2% at 24 hr, 49 ± 1% at 40 hr, and returned to normal (11 ± 3%) at 90 hr. When mice were made thrombocytotic by platelet transfusions, both megakaryocyte number and volume were decreased compared to controls, while no difference was noted in the number and percentage of CFU-M in DNA synthesis. Finally, experiments were performed to examine the effect of platelet transfusions on regenerating marrow. Experimental mice were given platelet transfusions while control animals received platelet buffer solution. At sacrifice the number and volume of megakaryocytes in the experimental group (platelet count 2.568 × 10⁶/μl) were less than controls (platelet count 0.363 × 10⁶/μl), while the number and percentage of CFU-M in DNA synthesis were similar in both groups. These results demonstrate that CFU-M are not immediately responsive to acute changes in platelet demand. The data suggest that megakaryo-cytopoiesis is structured on at least two levels which are independently regulated.     

MODE OF GROWTH OF THE JEJUNAL CRYPT CELLS OF THE RAT: AN AUTORADIOGRAPHIC STUDY USING DOUBLE LABELLING WITH 3H- AND 14C-THYMIDINE IN LOWER AND UPPER PARTS OF CRYPTS

The frequency distribution of cells through the mitotic cycle in lower and upper portions of jejunal crypts of the rat was examined by the ³H-¹⁴C-thymidine double labelling technique. Isolated crypts were cut perpendicular to the longitudinal axis so that the percentage of cells in the lower portion varied from 16 to 74 %. The lower and upper portion of the same crypt were squashed separately on one microscope slide and the number of ³H- and ¹⁴C-only labelled cells were scored to determine the flow rate into and out of S for the two portions. The mitotic cycle and its phases of the crypt epithelial cells were also determined. For lower portions of crypts which contained less than 40 % of the total cell number in that crypt the flow rate into S was about 1–7 times that of the flow rate out of S indicating that nearly every mitosis in this region produced two proliferative daughter cells. As the proportion of cells in the lower part of the crypt increased the quotient of the flow rate into S divided by the flow rate out of S decreased, and approached the steady state value of 1 0 in lower portions containing 60–74 % of the cells. For upper portions of crypts which contained less than 40% of the total crypt cells the flow rate into S was about 0 2 times that of the flow rate out of S, indicating that in this region mitoses predominantly produced non-proliferative daughter cells. The results obtained were in good agreement with the model of crypt cell proliferation proposed by Cairnie, Lamerton & Steel (1965b).     

DNA synthesis in human duodenal biopsies maintained in organ culture

Incorporation of 3H-thymidine during organ culture was studied in duodenal biopsies from 14 patients. Pulse-label at various intervals disclosed active incorporations during the first 2 h in culture. Labelling index declined to low levels at 3-4 h. Thereafter incorporation increased again and persisted throughout the rest of the culture period of 11 h. The DNA synthesis rate of crypt cells between 4 and 11 h in culture was calculated in 5 patients after pulse-label and continuous labelling of explants in parallel culture. The rate of entry into DNA synthesis was about 24 cells per 1,000 crypt cells per hour in flat, coeliac biopsies, versus 9-13 in controls, Gluten did not influence DNA synthesis rate, whereas wheat germ lectin inhibited DNA synthesis. Counting of the total number of mitoses and labelled fraction of mitoses disclosed active crypt cell renewal in flat, coeliac biopsies. In normal-appearing biopsies no mitoses were labelled, indicating delayed exit from S-phase or long duration of G2-phase in these explants.     

Circadian Rhythms In the Epithelial Cells and the Pericryptal Fibroblast Sheath In Three Different Sites In the Murine Intestinal Tract

Variations in percentage labelling (LI) and mitotic activity (mitoses per crypt) have been studied over a 24-hr period in the epithelial cells and pericryptal fibroblast sheath (PCFS) of the small intestine, caecum, and colon of the mouse. All three tissues displayed clear, synchronized circadian rhythms in DNA-synthetic activity in both the epithelial cells and PCFS. Peak values were coincident within a tissue, but staggered between tissues. In the epithelial cells, peak mitotic values were found between 3 and 6 hr after the peak LI values. the low level of mitotic activity in the PCFS appeared to be synchronized with the rhythms in the adjacent epithelia. the epithelial cells of the lowest crypt third displayed the clearest circadian rhythms. However, the PCFS cells at all levels produced similar curves. A craniocaudal wave of proliferative activity is proposed.     

THE CELL CYCLE TIME IN THE RAT JEJUNAL MUCOSA

The regional variation of the duration of cell cycle parameters was studied by constructing fraction of labelled mitoses curves at several levels in the jejunal crypt column of male Wistar rats. Prolonged T_c and T_s values were apparent only in the bottom eight cell positions, and these differences were shown to be significant compared with the remaining cell positions by analysing the data by the method of Gilbert (1972). Above cell position 8 the proliferating crypt cells showed effectively the same phase durations. For the whole crypt column T_c was 11.32 ± 0.14 (SE) and T_s 6.49 ± 0.10. Although variation in phase durations was confined to the basal portion of the crypt, the results essentially confirm the findings of Cairnie, Lamerton & Steel (1965a), and may be interpreted in terms of the slow cut-off model. The demonstration of prolonged T_c values in basal cell positions confirms the presence of a longer cycling subpopulation of cells at the bottom of the crypt.     

Scoring Mitotic Activity In Longitudinal Sections of Crypts of the Small Intestine

Abstract. Various counts have been made of the number of mitotic figures in whole crypts and sections of crypts of the small intestine of the mouse. Samples were analysed from animals killed at different times of the day and at different times after administration of vincristine. Measurements have been made of the size of mitotic and interphase nuclei and of the radial position of mitotic figures. the correction factor, f, which is required to take into account the enhancement of mitotic counts in sections as a consequence of their centripetal position has been investigated. the results indicate the following: (1) transverse sections of the crypt differ from longitudinal sections if they involve cutting the intestime before fixation which may result in a relaxation of the crypt and its widening by 25%; (2) columnar cell nuclei have a shape that resembles a sphere flattened so that the average diameter is 20% greater in crypt transverse sections; (3) mitotic nuclei tend to be about half-way between the crypt edge and the central axis of the crypt; (4) between about four and seven times more mitotic figures have their mitotic axis parallel to the long axis of the crypt; (5) about one-third of all mitotic figures in a crypt are seen in a longitudinal section of the crypt. If this is related to the number of cells in the crypt as a whole and in a section, a correction factor f_d for the mitotic index of 0.59 is obtained; (6) the correction factor f_T derived from the shape and position of the mitotic figures measured in 3 μm longitudinal sections is 0.53; (7) relating cell cycle and mitotic accumulation data using a computer-based model of the crypt also permits a correction factor f_mod to be estimated. This gives a value of 0.66. When sectioned material is used to calculate a mitotic index the most appropriate correction factor is f_D; for mouse small intestine it is 0.59.