Identification of G2/M phase transition by sequential nuclear and cytoplasmic changes and molecular markers in mice intestinal epithelial cells

ABSTRACT

Although the regulatory network of G2/M phase transition has been intensively studied in mammalian cell lines, the identification of morphological and molecular markers to identify G2/M phase transition in vivo remains elusive. In this study, we found no obvious morphological changes between the S phase and G2 phase in mice intestinal epithelial cells. The G2 phase could be identified by Brdu incorporation resistance, marginal and scattered foci of histone H3 phosphorylated at Ser10 (pHH3), and relatively intact Golgi ribbon. Prophase starts with nuclear transformation in situ, which was identified by a series of prophase markers including nuclear translocation of cyclinB1, fragmentation of the Golgi complex, and a significant increase in pHH3. The nucleus started to move upwards in the late prophase and finally rounded up at the apical surface. Then, metaphase was initiated as the level of pHH3 peaked. During anaphase and telophase, pHH3 sharply decreased, while Ki67 was obviously bound to chromosomes, and PCNA was distributed throughout the whole cell. Based on the aforementioned markers and Brdu pulse labeling, it was estimated to take about one hour for most crypt cells to go through the G2 phase and about two hours to go through the G2-M phase. It took much longer for crypt base columnar (CBC) stem cells to undergo G2-prophase than rapid transit amplifying cells. In summary, a series of sequentially presenting markers could be used to indicate the progress of G2/M events in intestinal epithelial cells and other epithelial systems in vivo.

Abbreviations

Brdu

5-bromo-2-deoxyuridine

PCNA

Proliferating cell nuclear antigen

CBC

crypt base columnar

TA cell

Transit amplifying cell

INM

interkinetic nuclear migration

pHH3

histone H3 phosphorylated at Ser10

Fucci

Fluorescent Ubiquitination-based Cell Cycle Indicator

Introduction

The cell division cycle consists of four phases (G1, S, G2, M), during which a cell goes through a series of events to duplicate DNA and produce two daughter cells. The entrance and timeline of cell cycle phases are thought to be tightly regulated by intrinsic properties of the cell and signals from the local microenvironment. Any abnormalities in these processes may lead to genomic instability and impairment of tissue homeostasis [1,2]. G2 phase is a specific period when DNA duplication stops, protein synthesis progresses, and Golgi complex dissociates in preparation for mitosis [3,4]. Although a genetic network regulating the G2 phase entrance and G2-M phase transition has been intensively studied [1,5–7], the significance and regulation of the G2 phase remains undefined and unclear. Recently, several cycle track systems were developed, based on modified ubiquitination-based cellcycle indicator (Fucci) or endogenously tagged endogenous proliferating cell nuclear antigen (PCNA), to report G2 phase and G2-M phase transitions [8,9]. However, the in vivo cell cycle phases usually present some specific morphological changes which depend on the cell type and tissue organization modules, which cannot be fully recapitulated by in vitro cell line studies.

In vivo cell cycle progression may be coordinated with the movement of the cell nucleus. For instance, inter kinetic nuclear migration (INM) has been identified as a common feature in pseudostratified epithelium, especially in the developing vertebrate brain [10–12]. The onset of the G2 phase in these tissues is characterized by rapid apical migration of the nucleus, while the M phase starts and finishes at the apical surface [11]. Adult intestinal epithelium is a good model for investigating cell cycle regulation because of its high turnover speed and specific structures. The single layer of epithelium is organized into a large number of self-renewing crypt-villus units. The villi present as finger-like protrusions and consist of functional postmitotic cells. The crypts, surrounding the villi, present as epithelial invaginations that contain all the progenitors and stem cells which fuel the self-renewal of the whole epithelium [13,14]. The daily cycling crypt cells include crypt base columnar (CBC) stem cells located at the bottom of the crypt that are sandwiched by Paneth cells, and progenitor cells/transit amplifying cells, which are located in the middle and upper crypt, which may have different signal modes to control cell dividing programs [13,14]. Although Brdu labeling and cell cycle marker staining have been widely used to identify the proliferation and cycling state of crypt cells [13,15,16], the morphology and timeline of crypt cell G2-M phase remains undefined.

In this study, the coordination of sequentially presenting markers with the events of G2-M phase transition was evaluated in adult intestinal epithelium crypt cells in vivo. It was found that these markers could be used to identify cell cycle progress in crypt cells in vivo.

Results

pHH3 in G2/M phase progression

Histone H3 phosphorylated at Ser10 (pHH3), as a mitosis marker [17], was firstly used to stain the normal small intestinal crypt. Consistent with previous reports that pHH3 initiates nonrandomly in pericentromeric heterochromatin in G2 phase cells [18], weak and scattered foci of Histone 3 phosphorylation was observed at the nuclear periphery, and the nuclear shape and location were not phenotypically different to the neighboring pHH3 negative cells (Fig. 1A), a pattern indicative of the early G2 phase. The pHH3 foci became larger and much more prominent when the nuclei of the intestinal epithelial cells maintained regular nuclear borders and were elongated, a pattern which is thought to be an indicator of early prophase (Fig. 1A, E). Compared with the cells in the upper crypt, the elongation of nuclei in the CBC stem cells presented a pike-like shape at the crypt bottom (Fig. 1A,B). The pHH3 positive nucleus finally moved across the IEC nuclear apical line to form condensed chromosomes with the highest level of anti-pHH3 immunostaining (Fig. 1C), indicating clear metaphase without nuclear envelope constraint (Fig. 1A, E). However, the anti-pHH3 immunoreactivity was very weak in the anaphase and undetectable in the telophase (Fig. 1A, B, E), indicating a dramatic decrease in phosphorylation of Histone 3 in the late mitotic stage of intestinal crypt cells, which is consistent with observations in other cell line studies [18,19]. Moreover, We found that most pHH3 positive cells existed in the prophase and metaphase (Fig. 1D). A similar dynamic pattern in the phosphorylation of Histone 3 was observed in colonic crypt cells(data not shown). Together, these data indicate that intestinal epithelial cells at the G2-M phase present a series of nuclear transformation and movement events, accompanied by a specific Histone 3 phosphorylation pattern (Fig. 1C,E).

Figure 1.

pHH3 staining. Panel A:Representative morphology of pHH3 positive cells by immunohistochemistry. Panel B:Representative morphology of pHH3 positive cells by immunofluorescence. Panel C: Relative expression level of pHH3 in the G2 phase, prophase, metaphase and anaphase (n = 12cells/phase; * p< 0.05; ** p< 0.01,compared with the G2 phase; ^#p< 0.05; ^##p< 0.01, with one-way ANOVA and Tukey's HSD test for multiple comparisons). Panel D:The percentages of cells at G2, prophase, metaphase, and anaphase-telophase in the pHH3 positive cells (Counted 2579 cells from 12 mice). Panel E: Schematic diagram depicting sequential changes in crypt cell nuclear morphology from the S phase to the M phase. Arrow heads in panel A and B indicate the pHH3 stained nuclei at G2, prophase, prometaphase, metaphase, anaphase, and CBC stem cells, respectively (scale bar = 10µm).

Nuclear transformation and migration in the G2/M phase

In pseudostratified epithelium, mitosis is coordinated with nuclear movement [10,20]. The crypts were isolated and stained with antibodies against E-cadherin and pHH3 to more precisely display the G2-M phase nuclear morphology. It was initially found that the nuclei of pHH3 positive cells enlarged and elongated in situ (Fig. 2A). Then, the transformed nuclei left the base line, moved towards the apical surface, and assumed a rounded form (Fig. 2A). Even in the anaphase/telophase, the cell still has a connection with the basal membrane (Fig. 2A). Previous studies suggest that the initiation of prophase requires a particular level of CDK1/cyclinB1 activation, which cannot be accomplished until cyclinB1 nuclear translocation occurs [6,7]. The isolated crypts were then stained with antibodies against cyclinB1 and pHH3. The cells with high levels of cyclinB1, mainly located in the cytoplasm, exhibited similar nuclear morphology to cyclinB1 negative cells (supplemental Fig. 1). However, crypt cells with nuclear transformation and/or apical movement showed obvious cyclinB1 nuclear translocation (Fig. 2B, supplemental Fig. 1). Increased maximum diameter was observed in the nuclei of prophase cells compared to their neighboring pHH3 negative cells (Fig. 2B, C, D, E). If taking the longest dimension of cell nucleus as the length, and the width as the direction perpendicular to the length [21], the aspect ratio of the nuclei of prophase cells is higher than that of neighboring pHH3 negative cells (Fig. 2F). Whereas, no significant increase of nuclear aspect ratio was observed in G2 phase cells compared to their neighboring pHH3 negative cells(data not shown). These data indicate that nuclear transformation, usually elongation may be taken as a morphological marker for the end of the G2 phase and the start of prophase. In summary, these observations indicate that cycling cells in adult intestinal epithelium exhibit a pattern of nuclear transformation and apical movement, which may coordinate with G2-M phase progression.

Figure 2.

Nuclear transformation and movement during the division of crypt cells. Panel A: Isolated intestinal epithelial crypts were fluorescently stained with anti-E-cadherin antibody and anti-pHH3 antibody. Arrow heads show the nuclear elongation in prophase, nuclear movement in prometaphase, and apical location of a nucleus in late anaphase. Panel B: Isolated intestinal epithelial crypts were fluorescently co-stained with cyclinB1 and pHH3. Arrow heads show translocation of cyclinB1 into an elongated nucleus of a prophase cell (scale bar = 10µm). Panel C:DAPI-stained nuclei of prophase cells and its neighboring pHH3 negative cell were shown in continuous slides with 2µm interval. Panel D: Schematic characterization of the length and width of the cell nucleus; E. Averaged maximum nuclear diameter of prophase cells and their neighboring pHH3 negative cells. F. Averaged nuclear aspect ratio of prophase cells and their neighboring pHH3 negative cells. Student's t-test, 2-tailed, n = 6, **p< 0.01, compared with the neighboring pHH3 negative cell.

Golgi and G2/M phase

Not only nuclear DNA has to be condensed and packaged to prepare for cell division, cytoplasm components such as Golgi also need to be disassembled for partitioning into the daughter cells [3,4,22]. Previous studies show that in mammalian cells the Golgi ribbon, a continuous membranous system localized in the perinuclear area, goes through sequential steps to fragment in the G2-M phase [4,22]. As a Golgi component and cell polarity marker [23], the expression and distribution of GM130 was detected in the intestinal epithelial cells at different cell cycle states. As expected, the expression of GM130 showed a typical ribbon pattern, locating above nuclei with a high cellular polarity in most cycling crypt cells and postmitotic villous cells (Fig. 3A). In the G2 phase, the Golgi complex maintained its morphological integrity overall; only the non-compact zones of the Golgi ribbon broke to generate small groups of isolated stacks. Isolated stacks only undergo further disassembly during prophase [4,22,24]. Consistently, it was found that compared with the neighboring pHH3 negative cells, the G2 phase cells, which were identified by the marginal pHH3 positive signals and regular nuclear shape, showed relatively condensed GM130 stained clustering (Fig. 3B). As soon as cells entered prophase, the GM130 stained clustering above nucleus disappeared (Fig. 3C and supplemental Fig. 2), indicating the quick disassemble of Golgi complex. Then, similar pattern of GM130 immunostaining was observed in the metaphase cells (Fig. 3D and supplemental Fig. 2). These results indicate that Golgi stacks undergo fragmentation during G2-M phase transition in intestinal epithelial crypt cells.

Figure 3.

GM130 staining. Paraffin sections of mice intestinal epithelial crypts were fluorescently co-stained with GM130 antibody and pHH3 antibody and imaged under confocal microscope. pHH3 staining was used to identify the cell phase state. Three-dimensional confocal reconstruction was processed to illustrate the GM130 staining pattern in multiple Z stacks. Panel A: Representative expression and distribution pattern of GM130 in intestinal crypt-villi units. The ribbon shape just above the nucleus was labeled by stash lines. Panel B: Representative expression and distribution pattern of GM130 in the G2 phase; arrow heads indicate condensed clustering GM130 staining in a G2 phase cell whose nucleus had marginal weak pHH3 immunostaining. Panel C: Representative expression and distribution pattern of GM130 in prophase; arrow heads indicate the absence of condensed clustering GM130 staining in a prophase cell with intense pHH3 immunostaining. Panel D: Representative expression and distribution pattern of GM130 in metaphase; arrow heads indicate the absence of condensed clustering GM130 staining (n = 5 mice, scale bar = 10µm).

Expression of common cell proliferation markers in G2/M phase

As the phosphorylation of Histone 3 could not be used to mark crypt cells in late anaphase and telophase (Fig. 1), the expression pattern of several cell proliferation markers, including PCNA and Ki-67, was investigated. Recent studies demonstrated that PCNA could fluorescently label the replication foci in the S phase, with the signal reaching a maximum at the end of the S phase, dramatically decreasing in the G2 phase, and finally redistributing to the whole cell in the mitotic phase, which could be used to classify cell line S/G2 and G2/M transition [9]. In the present study, we did not find PCNA foci in the nuclei of S phase cells by immunostaining, indicating the foci presentation may depend on endogenous fluorescence labeling (Fig. 4A,B). Moreover, the immunostaining of PCNA in G2 and prophase crypt cells overlapped with nuclei and did not exhibit the expected sharp down regulation when compared to neighboring S phase cells (Fig. 4A, B, C). These results indicate that PCNA expression pattern should not be used to distinguish S/G2 and G2/M transition in vivo. However, PCNA staining decreased and distributed to the whole cell when the nuclear envelope broke down in the metaphase (Fig. 4D) and anaphase (Fig. 4E), which is consistent with the cell line pattern of fluorescently tagged endogenous PCNA [9].

Figure 4.

PCNA staining. Panel A: Representative expression and distribution pattern of PCNA during S phase and G2 phase; BrdU incorporation was used to determine the S phase state of cells; stars indicate an S phase cell which had a pHH3 negative, PCNA positive, and BrdU positive nucleus; arrow heads indicate a G2 cell which had a pHH3 positive and PCNA positive nucleus, with less BrdU incorporation. Panel B: Representative expression and distribution pattern of PCNA in the S phase and prophase; stars indicate an S phase cell; arrow heads indicate a prophase cell. Panel C: Relative expression level of PCNA in the G2 phase and prophase (n = 10 cells/phase, p > 0.05). Panel D: Representative expression and distribution pattern of PCNA in the metaphase; arrow heads indicate a metaphase cell exhibiting dispersed PCNA staining. Panel E: Representative expression and distribution pattern of PCNA in the anaphase; arrow heads indicate an anaphase cell showing more dispersed PCNA staining (scale bar = 10µm).

Ki67 is observed in all phases of cycling cells, reaching peak levels in the metaphase [25]. Moreover, as an important mitotic chromosome periphery component, Ki-67 may play critical roles in the distribution of nucleolar material, chromosome dispersion and maintenance of chromosome structure [26–28]. Accordingly, a higher expression of Ki67 was found in cells in the mitotic phase than cells ininterphase (Fig. 5D). The prominent expression of Ki-67 was localized in the perichromosomes in metaphase (Fig. 5B), instead of prophase, anaphase, or telophase (Fig. 5A,C). Taken together, these common proliferative markers may be used to more precisely identify sub-phases in mitosis.

Figure 5.

Ki67 staining. Paraffin sections of mice intestinal epithelial crypts were fluorescently co-stained with Ki67 antibody and pHH3 antibody. Panel A: Representative expression and distribution pattern of Ki67 in G2 phase and prophase; stars indicate a G2 phase cell and arrow heads indicate a prophase cell. Panel B: Representative expression and distribution pattern of Ki67 in the metaphase; arrow heads indicate a metaphase cell showing evident perichromosome localization of Ki67. Panel C: Representative expression and distribution pattern of Ki67 in the anaphase; arrow heads indicate an anaphase cell whose chromosome was detectable by pHH3 or Ki67. Panel D: Relative expression level of Ki67 in the G2 phase, metaphase and anaphase. (n = 8 cells/phase, ** p< 0.01, compared with the S phase, ^#p< 0.05; ^##p< 0.01, with one-way ANOVA and Tukey's HSD test for multiple comparisons, scale bar = 10 µm).

Strategy to investigate timeline of G2-M phase in crypt cells

The G2 phase in some mammalian cell lines is reported undetectable, indicating a direct transition from DNA replication to mitosis [29]. Herein, taking advantage of the morphological and molecular identification of G2 and prophase, we directly measured the timeline of the G2-M phase in intestinal epithelial cells in vivo. As illustrated in Fig. 6, over a given period, the Brdu incorporation rate may reflect the instant cell cycle state, when cells are subjected to Brdu administration. The pHH3 negative cells with whole nuclei labeled by Brdu indicated cells that were in early or middle S phase (Fig. 6). The pHH3 positive cells with partial Brdu foci labeling were cells that had undergone a transit from late S phase to G2-M phase. The pHH3 positive cells without Brdu labeling were the cells that were in the middle of the G2-M phase. The pHH3 negative and Ki67⁺ cells without Brdu labeling marked cells that were under a late M phase-G1 phase transit. Thus, by changing the Brdu incorporation time, it is possible to define the least time for crypt cells to transit from the late S phase to any given period in the G2-M phase.

Figure 6.

Schematic diagram depicting the Brdu labeling and pHH3 strategy for measuring the duration of S-G2-M phase transition. The given labeling period was indicated by Xh. Brdu incorporation extent may reflect the time cells stayed in S phase. The pHH3 fluorescence levels combined with nuclear shape changes indicate the marker of sub-phases. (Refer to results for detail deducing method).

Diversity of timeline of G2-M phase in crypt cells

Thirty minutes after Brdu administration, a few cells in G2 phase, rather than prophase, showed tiny positive Brdu foci immunostaining in the nuclei, confirming the existence of cells in the G2 phase in the crypt (Fig. 7C). The labeling efficiency for G2 phase cells dramatically increase done hour after administration, when more than 40% of prophase cells showed Brdu positive incorporation(Fig. 7A,C), indicating that it took at least one hour for most crypt cells to complete the G2 phase. Ninety minutes after Brdu administration, all G2 phase cells and over 90% of prophase cells were labeled by Brdu (Fig. 7A,C), indicating most cycling crypt cells could finish G2-prohase transition within this time. Two hours after Brdu administration, almost all crypt cells in the late mitotic phase, including anaphase and telophase, were labeled by Brdu to a different extent when compared with prophase and metaphase cells, indicating that most crypt cells had completed G2-M phase in about two hours (Fig. 7A,C). Although the Brdu incorporation rate, indicative of time spent by cells in the S phase, is hardly quantified, the full and partial incorporation patterns can be easily distinguished by the extent of Brdu overlap with DAPI and pHH3 immunostaining (Fig. 7A,B). After the labeling time was extended to three hours, partial Brdu incorporation plus pHH3 immunostaining was only found in prophase CBC stem cells. Meanwhile, even peripheral anaphase cells in the upper crypt generally showed full Brdu labeling (Fig. 7B), indicating that compared to most progenitors/transit amplifying cells, the CBC stem cells took much longer to go through-G2- prophase, though the reason for this is still unknown.

Figure 7.

G2/M phase timeline diversity in crypt cells. Mice were administrated with BrdU for 0.5 h, 1 h, 1.5 h, 2 h and 3 h before sacrifice. Paraffin sections were made and fluorescently co-stained with pHH3 and Brdu antibodies. The total and the number of Brdu labeled cells at G2 phase, prophase, metaphase, and anaphase/telophase were counted and the percentages of the Brdu labeled cells in the total number of the cells at each phase were deduced. Panel A: pHH3 and Brdu co-staining in sections of intestinal tissues frommice administrated with Brdu for differing time periods; arrow heads indicate some typical cells, from which the duration of the cell cycle phase could be deduced. Panel B: Representative Brdu incorporation extent in pHH3 positive CBC stem cells after Brdu was administrated for 3 h. Yellow arrowhead, pHH3 and Brdu co-staining of CBC stem cells (crypt bottom); white arrowhead, TA cells (upper crypt); CBC stem cells were partially labeled by Brdu in prophase while TA cells were fully labeled by Brdu in anaphase. Panel C: The percentage of Brdu labeled cells out of the total number of cells at the corresponding sub-phases, after administration of Brdu over defined times (n = 3 mice/timepoint, scale bar = 10µm).

Discussion

In the present study, changes in nuclear shape and size were analyzed to assess coordination with the onset and progress of prophase in crypt cells. Although the nuclear volume in mammalian cells is thought to double during the cell cycle, the exact size of cell nuclei has not been measured due to difficulties in evaluating the volume of nuclei during cell division [30–32]. It has been reported that nuclear shape is modulated by nuclear laminas [33,34]. Nuclear laminas contain conserved CDK1 consensus motifs, the phosphorylation of which is responsible for nuclear envelope break down (NEBD) in mitosis [35]. However, it has been reported that NEBD occurs during prometaphase, while nuclear shape and size does not change in the prophase of Hela cells [35], supporting the theory that CDK1/cyclinB1 activation in the prophase may not be the cause of morphological nuclear changes in the intestinal crypt cell model. Further studies are required to elucidate the mechanism underpinning changes in nuclear shape and size in prophase. Additionally, our data indicate that, as well as Xenopus egg cells and C. elegans [30–32], mouse crypt cells are a good model for investigating cell cycle events that influence nuclear morphology.

For investigations into the coordination of nuclear movement with cell phases, most studies have been carried out on the developing tissues of pseudostratified epithelium [10,11,36]. During the G2 phase, apical nuclear movement is reported to be driven by the concerted action of actomyosin constriction and dynein-mediated motility to the minus end of the microtubules [11]. Herein, we described mitotic cell nuclear movement progression accompanied by nuclear transformation in the prophase and metaphase of single layer epithelial cells, which is reminiscent of the apical movement of the apoptotic enterocyte nucleus in our previous studies, which might be driven by the F-actin assembly and accumulation inside the apoptotic enterocytes, rather than the contraction of the actomyosin ring in live neighboring cells [37]. Further study is required to determine if nuclear movement in mitotic crypt cells actually employs a similar mechanism. Moreover, it would be interesting to test if nuclear migration is essential for cell cycle progression single layer epithelium.

The disappearance of endogenous PCNA fluorescence foci from the S phase to G2 phase has been used to report S-G2 transition, and the occurrence of whole cell distribution of PCNA in metaphase was taken as a sign of G2-M transition in transgenic cell lines [8,9,38]. Interestingly, we found that there were no changes in the PCNA immunostaining pattern in vivo during the S-G2 and G2-prophase transitions, indicating it is not appropriate to use the PCNA immunostaining pattern to report S-G2 transition in vivo. It is also inappropriate to use the whole cell distribution of PCNA [9,38] and highly condensed chromosomes [8] to report the onset of prophase in vivo, when the nuclear envelope exists and chromosome condensation has just started. Instead, based on our observations, the nuclear translocation of cyclinB1 and the whole cell distribution of GM130 or other Golgi complex components seem to be appropriate markers to report G2-M phase transition. The change from a nuclear localized pattern to a whole cell distributed pattern in PCNA immunostaining after prophasepoint to an appropriate marker for identifying metaphase and anaphase in vivo.

Previous quantitative analyses of G2 phase duration, whether based on a fluorescent biosensor-based cell track system or calculating the percentage of cells labeled with radioactive thymidine, are not accurate, as they cover the period from the end of the S phase to the metaphase, which includes the prophase [8,9,29]. Herein, a more accurate measurement of G2 phase duration was made in crypt progenitors/transit amplifying cells based on morphological and cell cycle markers, which showed that there is diversity in the duration of G2 phase among crypt cells. The intestinal crypt is a prototype stem cell compartment where stem cells and progenitors and their descendent TA cells sequentially seat from the bottom to the top. The CBC stem cells, located at the crypt bottom, have a longer cell cycle time than the progenitor cells and TA cells in the upper crypt [13,39]. The percentage of CBC stem cells in the G2/M phase is particularly low, which makes it difficult to evaluate their exact G2 phase duration. Escobar et al. reported the length of the G2/M phase of mouse CBC stem cells to be 3.4 h, which was calculated by using the percentage of pHH3 positive CBC stem cells multiplied by their cell cycle time (28.5h) [15]. Consistently, we found that CBC stem cells took much longer to go through the G2-prophase than most TA cells. Future study will focus on investigating the mechanism underpinning the observed diversity in G2 phase duration among different crypt cells.

In conclusion, G2/M phase transition was coordinated with distinct nuclear morphology changes and movement, Golgi complex fragmentation, pHH3 levels, BrdU incorporation, and in terms of sub-phases, with PCNA and Ki67, in mice intestinal epithelial crypt cells. These sequential markers could be used to track G2/M transition and mitotic sub-phases, and to measure the duration of G2/M transition in intestinal crypt cells. These results also lay the foundation for further investigations into G2/M events, including cell cycle arrest in intestinal epithelial cells and other organ/systems consisting of epithelium in vivo.

Materials and methods

Mice and treatment

Male C57BL/6 mice with ages of 8 to10 weeks were purchased from Chengdu Dossy Biological Technology Company and Model Animal Research Center of Nanjing University. For BrdU incorporation test, 50 mg /kg body weight of BrdU (Sigma, St. Louis, MO) was administrated by intraperitoneal injection. After indicated time including 0.5, 1h, 1.5h, 2h, 3h, 4h, animals were sacrificed by decapitation, and small intestine, colon and stomach tissues were excised, and fixed in Zinc formalin fixative (Sigma, St. Louis, MO) for 24 hr. For all other tests using paraffin sections, untreated mice were sacrificed, and tissues were collected and fixed in the same way as mentioned above. After fixation, the intestinal tissues were further processed and embedded in paraffin. Sections of 3 µm were prepared using a microtome, and laid on to glass slides.

Immunofluorescent and immunohistochemistry staining

Immunofluorescent staining was performed following the procedure as previously described [37,40]. Briefly, after the paraffin sections were deparaffinized and hydrated, they were boiled at 95℃ in citric acid antigen retrieval buffers (Maixin; MVS-0100; Fuzhou, China) for 15 min. After cooling down, they were rinsed in PBS and blocked with 5 % bovine serum albumin (Beyotime, China), and detected with the primary antibodies at optimal dilutions overnight at 4℃. The primary antibodies used include: rabbit monoclonal anti Phospho-Histone H3(Ser10) (1:300, 9701, CST, Massachusetts, USA), rat monoclonal anti BrdU (1:200, OBT0030, BIO-RAD, California, USA), mouse polyclonal anti BrdU (1:300, 555627, BD, Cambridge, USA), rabbit monoclonal anti-mouse/human Ki-67 (1:200, AB16667, Abcam, San Francisco, USA), rat monoclonal anti-mouse/human Ki-67 (1:150, 151202, Biolegend, San Diego, USA), and mouse monoclonal anti PCNA (1:250, 60097-1-lg, protein tech, Rosemont, USA). The sections were rinsed in PBST and incubated with secondary antibody: Donkey Anti-Mouse IgG (H+L) Secondary Antibody, Alexa Fluor® 555 conjugate (1:300, Thermo, USA); Alexa Fluor® 488 AffiniPure Donkey Anti-Rabbit IgG (H+L)(1:300,Jackson,USA); or Alexa Fluor 647-conjugated AffiniPure Donkey Anti-Rat IgG (H+L) (1:300, Jackson, USA) for 1.5 h at room temperature. The slides were then rinsed several times and counter stained with 4,6-diamidino-2-phenylindole (DAPI) (Beyotime, China). After the slides were mounted with ProLong gold antifade reagent and covered with coverslips fluorescent images were acquisitioned with a fluorescence microscope (Olympus, BX63, Japan and Zyla 4.2, ANDOR, England) .Some images were captured by Zeiss LSM 780 and 880 confocal microscope and 3D images were constructed by ZEN 2.1 (blue edition) software (3DXL was powered by Arivis). For Immunohistochemistry, after primary antibody incubation, sections were incubated with secondary antibody conjugated with HRP and developed with DAB substrate solution according to the manufacture's instruction (rabbit polymer detection system PV-6001; ZSGB-BIO, Beijing, China).

Crypt isolation and whole mount staining

Mouse Intestinal crypt-villus units were isolated and stained as previously described [37,40]. Briefly, the isolated small intestinal tract was opened and cleaned by washing with cold Hank's balanced salt solution (HBSS) without Ca2+/Mg2+, the jejunum was cut into 3- to 5-mm pieces and incubated in 30 mmol/L EDTA in HBSS for 20 minutes in an ice bath. After the EDTA solution was removed and the tissues were washed with HBSS they were incubated in HBSS for another 10 minutes on ice. After the 50ml incubation tubes containing 40ml cold HBSS were shaken at 2 to 3 times/ second with hands, the isolated crypt-villus units were collected by sedimentation. The crypt-villus units were fixed in 10% formalin at 37℃ for 10 min and then at 4℃ for 1h. After washed three times with PBS, the isolated crypt-villus units were permeabilized with 0.5% Triton X-100 in PBS for 30 min at room temperature. After blocked in PBS containing 5 % bovine serum albumin and 0.1% Triton X-100 for 2h at room temperature, the epithelial units were incubated with primary antibodies (Phospho-Histone H3(Ser10) (CST, 9701, 1:200), E-cadherin (BD, 610181,1:100), Phospho-Histone H3(Ser10) (CST, 9706, 1:200), or cyclinB1 (CST, 4138, 1:100) ) at 4℃ overnight with shaking. Donkey Anti-Mouse IgG (H+L) Secondary Antibody, Alexa Fluor® 555 conjugate (1:300, Thermo, USA) and Alexa Fluor® 488 AffiniPure Donkey Anti-Rabbit IgG (H+L)(1:300,Jackson,USA), were used as secondary antibodies. DNA was detected with 3mg/ml DAPI. The images of the stained crypt-villus units were captured by Zeiss LSM 780 confocal microscope.

Statistical analysis

For evaluating PCNA expression in G2 and prophase, PCNA fluorescence intensity of S phase cells in the same crypt was taken as internal control.For evaluating Ki67 expression in M phase, the fluorescence intensity of neighboring S phase cells was taken as internal control. For evaluating pHH3 expression in S,G2,metaphase and anaphase, the fluorescence intensity of prophase cells in the same crypt was taken as internal control. The fluorescence intensity was quantified by software Image-Pro PLUS 6. Data were analyzed with SPSS 18.0 (SPSS Inc., Chicago, IL, USA). Results are expressed as means ± standard deviation (SD). Statistical analysis for the data with homogenous variance was performed by one-way ANOVA and Tukey's HSD test for multiple comparisons. p < 0.05 was taken as statistically significant. The diameters of G2 phase and prophase were evaluated by software Image-Pro PLUS 6. The maximum diameters were selected based on continuous slides with 1.5-3µm interval. The width is the dimension perpendicular to the length. The aspect ratio of the nucleus for individual prophase cell and its neighboring pHH3 negative cell was assessed. (Results are expressed as means ± standard deviation (SD). Statistical analysis for the data was performed by T-test. p < 0.05 was taken as statistically significant.)

Funding Statement

The authors disclose receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the following grants: National Natural Science Foundation of China (No. 81472916); the Fund from PLA (AWS13J002, AWS14C002), and the Funds of Key Laboratory of Trauma, Burn and Combined Injury (No. SKLZZ201420).

Acknowledgments

We thank Xiao-Yang Zhou, Shuang Long, Qing Zhou, Li-Ting Wang and Min Jin for their excellent technical assistance and Yanqi Zhang for advice in statistical analyses.

Disclosure of potential conflicts of interest

The authors declare no conflict of interest.