Reprogramming of fibroblast nuclei in cloned bovine embryos involves major structural remodeling with both striking similarities and differences to nuclear phenotypes of in vitro fertilized embryos

Abstract

Nuclear landscapes were studied during preimplantation development of bovine embryos, generated either by in vitro fertilization (IVF), or generated as cloned embryos by somatic cell nuclear transfer (SCNT) of bovine fetal fibroblasts, using 3-dimensional confocal laser scanning microscopy (3D-CLSM) and structured illumination microscopy (3D-SIM). Nuclear landscapes of IVF and SCNT embryonic nuclei were compared with each other and with fibroblast nuclei. We demonstrate that reprogramming of fibroblast nuclei in cloned embryos requires changes of their landscapes similar to nuclei of IVF embryos. On the way toward the 8-cell stage, where major genome activation occurs, a major lacuna, enriched with splicing factors, was formed in the nuclear interior and chromosome territories (CTs) were shifted toward the nuclear periphery. During further development the major lacuna disappeared and CTs were redistributed throughout the nuclear interior forming a contiguous higher order chromatin network. At all stages of development CTs of IVF and SCNT embryonic nuclei were built up from chromatin domain clusters (CDCs) pervaded by interchromatin compartment (IC) channels. Quantitative analyses revealed a highly significant enrichment of RNA polymerase II and H3K4me3, a marker for transcriptionally competent chromatin, at the periphery of CDCs. In contrast, H3K9me3, a marker for silent chromatin, was enriched in the more compacted interior of CDCs. Despite these striking similarities, we also detected major differences between nuclear landscapes of IVF and cloned embryos. Possible implications of these differences for the developmental potential of cloned animals remain to be investigated. We present a model, which integrates generally applicable structural and functional features of the nuclear landscape.

Keywords: 3D-CLSM, 3-dimensional confocal laser scanning microscopy; 3D-SIM, 3-dimensional structured illumination microscopy; B23, nucleophosmin B23; BTA, Bos taurus; CDC, chromatin domain cluster; CT, chromosome territory; EM, electron microscopy; ENC, embryonic nuclei with conventional nuclear architecture; ENP, embryonic nuclei with peripheral CT distribution; H3K4me3; H3K4me3, histone H3 with tri-methylated lysine 4; H3K9me3; H3K9me3, histone H3 with tri-methylated lysine 9; H3S10p, histone H3 with phosphorylated serine 10; IC, interchromatin compartment; IVF, in vitro fertilization; MCB, major chromatin body; PR, perichromatin region; RNA polymerase II; RNA polymerase II-S2p, RNA polymerase II with phosphorylated serine 2 of its CTD domain; RNA polymerase II-S5p, RNA polymerase II with phosphorylated serine 5 of its CTD domain; SC-35, splicing factor SC-35; SCNT, somatic cell nuclear transfer.; bovine preimplantation development; chromatin domain; chromosome territory; embryonic genome activation; in vitro fertilization (IVF); interchromatin compartment; major EGA, major embryonic genome activation; somatic cell nuclear transfer (SCNT).

Figure 1

(See previous page). Global reorganization of nuclear architecture during preimplantation development of in vitro fertilized bovine embryos studied with 3D confocal laser scanning microscopy. Panels A1–A8. Top, middle and bottom x/y-sections from image stacks of 8 DAPI-stained nuclei recorded with 3-dimensional confocal laser scanning microscopy (3D-CLSM) in an in vitro fertilized 8-cell embryo. All nuclei show the ENP phenotype (Embryonic Nucleus with a Peripheral chromatin arrangement; for a detailed description see Results). Panels B1–B8. Top, middle and bottom x/y-sections from representative DAPI-stained nuclei recorded with CLSM in an IVF 20-cell embryo. Panel B1 exemplifies one of 3 nuclei in this embryo with a persistent ENP phenotype. Panels B4–B8 provide examples for the nuclear phenotype, termed ENC, which was noted in 15 nuclei (Embryonic Nucleus with Conventional chromatin arrangement; for a detailed description see text). Two nuclei (Panels B2 and B3) apparently represent a transition stage between ENPs and ENCs, called ENP/C (for details see text). Bars: 10 μm in panel A1 is representative for all nuclei.

Figure 2

(See previous page). Global reorganization of nuclear architecture during preimplantation development of cloned bovine embryos studied with 3D confocal laser scanning microscopy (3D-CLSM). Panels A1–A7. Top, mid and bottom x/y-sections from image stacks of 7 DAPI-stained nuclei recorded with 3D-CLSM in a cloned 7-cell embryo. All nuclei show an ENP-like phenotype (compare Fig. 1, Panels A1-A8). Panels B1–B7. Top, mid and bottom x/y-sections from representative DAPI-stained nuclei recorded with CLSM in a cloned 25-cell embryo. One nucleus (Panel B1) has retained the ENP phenotype. The other nuclei (Panels B2–B7) represent the ENC phenotype (compare Fig. 1, Panels B4–B8). Nucleoli surrounded with dense chromatin were noted in all ENCs but not in ENPs. Green signals in all Panels represent immunostained kinetochores. Enlarged views from boxed areas in Panels A1 and B1 with kinetochores are shown at the bottom of this Figure (Panels a and b). Note the location of kinetochore clusters in modestly DAPI-stained areas. Bars: 10 μm in panel A1 applies to all Panels A1-7 and B1-7; 1 μm for enlarged boxes a1/a2, and b1/b2.

Figure 3

(See previous page). Comparison of painted chromosome territories in nuclei of bovine fetal fibroblasts and cloned embryos. Panels A–D. Midplane nuclear sections from imaging stacks recorded with CLSM from a bovine fetal fibroblast nucleus (panel A), an ENP (panel B), an ENP/C (panel C) and an ENC (panel D) from cloned embryos after 3D-FISH with a BTA 13 paint probe (green). Chromosome painting demonstrates that individual CTs 13 correspond to individual major chromatin bodies (MCBs) with variable configurations, occasionally including nearby chromatin clusters (arrow). Three further MCBs painted with the BTA 13 probe were detected in other light optical sections from this nucleus (see Supplementary Fig. S1 in accompanying article by Popken et al.³⁵). Bars: 3 μm for A1–3, B1, B2, C1, C2, D1, D2; 2 μm for the enlarged box in A1 and for B3, B4; 1 μm for C3, C4.

Figure 4

(See previous page). Arrangements of splicing speckles and de novo synthesized RNA in nuclei of in vitro fertilized and cloned embryos. Panels A and B. Midplane sections recorded by 3D-CLSM in an ENP (panel A) and an ENC (panel B) from IVF embryos stained with TO-PRO-3 DNA (red) show the enrichment of the splicing factor SC-35 (green) in splicing speckles both in the major lacuna of the ENP and distributed throughout the interchromatin compartment of the ENC. Panels C–E. Immunocytochemical detection of bromine-labeled RNA after incubation for 45 minutes with BrUTP or BrU precursors. Panel F. Projection of a confocal image stack from a 15-cell cloned embryo stained with TO-PRO-3 (red) following 45 minutes incubation with BrU shows nuclei with strikingly different phenotypes, including a pyknotic nucleus (arrow). Note that the visibility of all nuclei in this projection is precluded by nuclear overlays. Panels G–I. Enlarged views of nuclei, framed in F1 by boxes G, H and I, include an ENP-like nucleus (G), an ENP/C-like nucleus (H) and an ENC (I) (for definition of these nuclear phenotypes see Results and compare Figs. 1 and 2). These Panels also present evidence for de novo RNA synthesis in nuclei of this cloned embryo independent of differences between nuclear phenotypes. Note strong RNA synthesis in nucleoli of ENCs of both fertilized (panel E) and cloned embryos (panel I). Bars: 5 μm in A1 representative for Panels A–E, 25 μm in Panel F; 5 μm in G1 for Panels G–I.

Figure 5.

Representative confocal midplane sections demonstrate essential features of ENPs and ENCs in in vitro fertilized embryos. Panel A. ENP at G2/ prophase, characterized by strong immunostaining of H3S10p. Panel B. ENP in S-phase, characterized by pulse-labeling (30 min) of replication foci with EdU. Panel C. EdU pulse labeling pattern of a typical ENC in mid S-phase. Panel D. ENP with several nucleolar precursor bodies (NPBs), immunostained with the nucleolar marker B23. Panel E. Large nucleoli lined with B23 in the interior of an ENC. Panels F, G. ENP (F) and ENP/C (G) with positive H3K4me3 immunostaining of chromatin, clearly enriched in some MCBs (compare Supplementary Fig. S1). Panels H, I, J. Immunocytochemical detection of H3K9me3 in an ENP (H), an ENP/C (I) and an ENC (J). Bar: 4 μm in A1 representative for all Panels.

Figure 6.

Problems and consequences of image thresholding in 3D structured illumination microscopy. Midplane SIM sections from a DAPI-stained ENP (panel A) and an ENC (panel B) without any threshold (A1, B1) and after application of a moderate threshold (A2, A3 and B2, B3). Contrast and brightness of images A2, B2 was chosen to emphasize a pronounced pattern of concentric rings, which represents an artifact of structured illumination microscopy. C and D. DAPI intensity profiles recorded along dashed lines in A2 and B2. Maintenance of these patterns in unthresholded images resulted in erroneous local increases (A2, C, green dotted line, circles and arrows) or decreases (A2, B2, C, D, red dotted line, circles and arrows) of DAPI intensities. Application of a threshold just above average background (C and D) largely removed the cocentric-ring like patterns, but at the expense that part of the real DAPI signal was also lost (A3, B3). See Supplementary Figures S2 and S3 for problems and consequences of threshold application to immunostained signals. Bars: 3 μm in A3 representative for all Panels.

Figure 7

(See previous page). Nuclear phenotypes in in vitro fertilized preimplantation embryos studied with 3D structured illumination microscopy. A–I. Midplane sections from typical DAPI-stained nuclei recorded with 3D-SIM at different preimplantation stages, male pronucleus from a zygote (A), female pronucleus from the same zygote (B), 2-cell stage (C), 4-cell stage (D), 8-cell stage (E), 16-cell stage (F), morula (G), and blastocyst (H from inner cell mass (ICM) and I from trophectoderm (TE)). The 2 pronuclei in a zygote were recorded with 3D confocal laser scanning microscopy (3D-CLSM) due to their distance to the cover glass. J–Q. Top row: Midplane-SIM sections from all DAPI-stained nuclei recorded in an IVF 8-cell embryo. Bottom row: Corresponding x/z-sections. With rare exceptions nuclei have a roundish shape and show the ENP phenotype (for definition see Results and Fig. 1). R1-R10. Midplane SIM sections present typical phenotypes of DAPI-stained nuclei at higher magnification taken from embryos with 4 cells (R1 – R3), 7 cells (R4), 8 cells (R5, R6), 9 cells (R7, R8), 19 cells (R9) and a morula with about 32 cells (R10). These nuclei were classified as pre-ENP (R1), transition state from the pre-ENP phenotype to ENP phenotype (R2) or early ENPs (R3, R4), fully developed ENPs (R5, R6), ENP/Cs (R7, R8), and ENCs (R9, R10). Typical nucleoli surrounded by intensely DAPI-stained chromatin (examples marked by asterisks in R7 – R10) were typically noted in ENP/Cs and in all ENCs. Arrows in R2 – R8 point to wide interchromatin compartment (IC) channels, pervading from the major lacuna toward the nuclear envelope between major chromatin bodies. They were first noted during the pre-ENP/ENP transition (R2), prominent in ENPs and ENP/Cs (see enlargements of boxed areas a-c in R6, R7) and disappeared together with the major lacuna during the development of ENCs (R9, R10). Major lacunas often adopted a central position in the nuclear interior of ENPs (R6), but acentric positions were also observed resulting in broad regions of direct contact of a major lacuna with the nuclear envelope (R3, arrowheads). At all stages of preimplantation development major chromatin bodies likely representing individual chromosome territories (compare Fig. 3) were built up from smaller chromatin domain clusters (CDCs) pervaded by small IC channels (see enlargements of boxed areas a-c in R5 and R10). S. Volumes recorded for a sample of nuclei recorded with 3D-CLSM or 3D-SIM from embryos with 4-cells (E4; 15 nuclei), 5 to 8 cells (E5–8; 27 nuclei), 9–16 cells (E9–16; 36 nuclei) and more than 18 cells (E > 18; 30 nuclei). Scatter plots are combined with box plots presenting mean values, quartiles and whiskers. Each dot respresents the volume of an individual nucleus. Black dots respresent ENPs, red dots ENP/Cs and blue dots ENCs. A highly significant volume decrease (P < 0.006) was noted for nuclei in embryos advancing from 5–8 cells to 9–16 cells. Bars: 5 μm in I representative for A–I; 3 μm in J representative for J–Q; 3 μm in R1 representative for R1-R10; 1 μm side length of enlarged boxes a, b and c in R5-R7, R10.

Figure 8

(See previous page). Nucleolar precursor bodies and nucleoli in nuclei of in vitro fertilized embryos. A1–E1. Midplane SIM sections of a DAPI-stained nucleus approaching ENP (A1) and of 4 ENP/Cs (B1–E1). A2–E2. Corresponding sections with immunostained H3K4me3 (green) and B23 (red) in A2 and B2 or H3K4me3 (green) and H3K9me3 (red) in C2 – E2. Boxed areas in Panel A frame a nucleolar precursor body or early nucleolus and likely mature nuclei in Panels B–E. Enlarged images of boxed areas are presented below: A3–E3, DAPI; A4, B4, B23; C4–E4, H3K9me3; A5–E5, H3K4me3; A6, B6, colored overlays of DAPI, H3K4me3 and B23; C6–E6, colored overlays of DAPI, H3K4me3 and H3K9me3. Note that the nucleolar precursor body or early nucleolus shown in Panel A is lined by B23 (A4) but lacks a surrounding rim of intensely DAPI-stained DNA (A3) in contrast to mature nucleoli (B3–E3). Unexpectedly, most of the nucleoli shown in Panels B, D and E present variable parts of their DAPI-stained rims labeled with H3K4me3 (B5, C5, D5, see arrow heads in C5 and D5), considered as a marker for transcriptionally competent chromatin, but little detectable staining with H3K9me3, considered as a marker for silent chromatin (C4, D4). The DAPI-stained rim of the nucleolus shown in Panel E is strongly marked by H3K9me3 (E4) but shows little H3K4me3 staining (E5, arrow heads). Bars: 5 μm in A1 is representative for A1–E2, 1 μm in A3 is representative for enlarged boxes A3–E6.

Figure 9.

A transcriptome analysis suggests a major increase of intronic sequences from genes coding for ribosomal proteins at the 8-cell stage of in vitro fertilized embryos. Columns present the results of a transcriptome analysis of 83 genes coding for ribosomal proteins at germinal vesicle (GV) and at metaphase II stage bovine oocytes, as well as in IVF embryos with 4, 8 and 16 cells and in blastocysts. For all genes the parameter RINP was determined individually as a measure for the coverage of all intronic sequences in transcripts from each gene. RINP indicates the ratio of intronic read counts to not-covered intronic positions. A fold change ≥10 in RINP between subsequent replicates of the embryonic stages was considered as indicative of nascent transcription. Background was defined as the 75th percentile of RINP in the oocyte stages (for further details see Graf et al. 2014). A fold change ≥10 could only be confirmed for a fraction of ribosomal protein genes at each stage (for details see Results). Columns show average values of RINP for the entire pool of ribosomal protein genes.

Figure 10.

Topography of DAPI-stained chromatin, RNA polymerase II-S5p and H3K4me3 in nuclei from in vitro fertilized preimplantation embryos recorded with 3D structured illumination microscopy. Panels A and B. Midplane SIM sections from an ENP (panel A) and ENC (panel B). A1 and B1. Patterns of DAPI-stained DNA (gray). A2 and B3. Overlay of H3K4me3 (green) and RNA polymerase II-S5p signals (red) on the DAPI images. B2 shows only the H3K4me3 pattern, B4 only the RNA polymerase II-S5p pattern of this ENC. Rectangular boxes of the same size area marked in A1, A2, B1–B4 are presented as enlarged views in images a1, a3, a5, a7 and b1, b3, b5, b7. To allow a more detailed visual inspection at the level of individually recognizable pixels (39.5 nm² pixel size) rectangular boxes marked in these views are further enlarged in a2, a4, a6, a8 and b2, b4, b6, b8. An increase in color intensity of DAPI positive pixels (blue) is tentatively considered as a reflection of a local increase in the compaction of chromatin stained with DAPI (a1, a2, b1, b2). Black pixels in a3, a4 and b3, b4 denote a colocalization of DAPI positive pixels with H3K4me3 positive pixels. Green colored H3K4me3 positive pixels were noted in DAPI negative areas of thresholded images (compare Fig. 6) (tentatively considered as interchromatin compartment channels). Red colored pixels in a5, a6 and b5, b6 are taken as an indication for RNA polymerase II-S5p positive pixels located within the interchromatin compartment, whereas black pixels in these images denote a colocalization of RNA polymerase II-S5p positive pixels with DAPI positive pixels. Black pixels in a7, a8 and b7, b8 indicate a colocalization of H3K4me3 positive pixels with RNA polymerase II-S5p positive pixels, whereas green and red pixel clusters hint to separate H3K4me3 and RNA polymerase II-S5p signals. Note the predominance of positive pixels for H3K4me3 and RNA polymerase II-S5p clusters at the periphery of chromatin clusters stained with DAPI often expanding into the IC (white). RNA polymerase II-S5p clusters show side-by-side associations with clusters of H3K4me3 labeled chromatin, but are also frequently located remote from each other. Bars: 3 μm for A1, A2, B1–B4; 1 μm for a1, a3, a5, a7 and b1, b3, b5, b7; 300 nm for a2, a4, a6, a8 and b2, b4, b6, b8.

Figure 11

(See previous page). Topography of DAPI-stained chromatin, H3K4me3 and H3K9me3 in nuclei from in vitro fertilized preimplantation embryos recorded with 3D structured illumination microscopy. Panels A–C. Midplane SIM sections from an ENP (A), an ENP/C (B) and an ENC (C) present typical examples of the topography of DAPI-stained DNA (gray, A1–C1) together with overlays of DAPI-stained DNA with H3K4me3 (green) and H3K9me3 (red) (A2–C2). Enlargements of boxed areas a, d, e, g and h in these nuclei are shown on the right with DAPI-stained chromatin in blue (a1, d1, e1, g1 and h1) and corresponding overlays of DAPI images with pixels (pixel size 39.5 nm²) classified as immunopositive for H3K4me3 (green) or H3K9me3 (red) (a2, d2, e2, g2 and h2). Pixels and pixel clusters positive for both H3K4me3 and H3K9me3 are denoted in black. Circles in the corresponding images d1/d2, e1/e2, g1/g2, h1/h2 were drawn into these enlarged images either to mark areas of special interest (see below) or to facilitate comparisons between images showing only the DAPI pattern (left) and their corresponding counterpart showing additional immunolabeling of epigenetic markers (right). Panel A. In this ENP most of the core parts of the strictly peripherally located DAPI-stained major chromatin bodies (MCBs) are labeled with H3K9me3. In the periphery of some of these MCBs we note intensive labeling with H3K4me3 (compare a1 and a2). H3K4me3 positive clusters can be noted directly at the nuclear border. The MCB presented at the right side of b1 and b2 exemplifies a case, strongly labeled with H3K4me3 throughout with very little additional staining of H3K9me3, whereas the MCB on the left side of b1 and b2 shows overlapping signals of H3K4me3 and H3K9me3. Panel B. This nucleus was identified as ENP/C. Nucleoli surrounded by densely DAPI-stained chromatin were noted in other sections of the 3D-SIM image stack (for an example see box c). In comparison with the ENP shown in Panel A, the size of the major lacuna is apparently reduced by the invasion of MCBs toward the nuclear interior. A clear separation of neighboring MCBs by wide IC channels is no longer possible in this nucleus. Potential MCBs or alternatively major chromatin domain clusters with contributions from several neighboring CTs were preferentially labeled with H3K4me3, while others were preferentially labeled with H3K9me3 (see boxed area e in Panel B and enlarged views e1 and e2 on the right). Enlarged boxes d1/d2 show a peripheral nuclear region with intensely DAPI-stained chromatin enriched at the nuclear border. H3K9me3 labeled chromatin is noted in clusters both at the nuclear border and away from the border. These clusters, however, colocalize only in part with the densely DAPI-stained chromatin but also extend into modestly DAPI-stained regions. Clusters of H3K4me3 labeled chromatin are distributed between the H3K9me3 labeled clusters. The enlarged views f1/f2 show that not only H3K9me3 but also H3K4me3 positive clusters can be noted directly at the nuclear border (compare with a2). Panel C. Midplane SIM section from a typical ENC nucleus. In contrast to the ENP and ENP/C, the major lacuna has disappeared. Chromatin at the nuclear border is nearly exclusively marked with H3K9me3 and extends as an interconnected chromatin network throughout the nuclear interior with numerous large and small H3K9me3 labeled clusters, whereas enrichment of H3K4me3 label is only noted in dispersed small clusters (compare g1/g2 and h1/h2). Although these enlarged views suggest a preference of H3K9me3 for regions more densely stained with DAPI, whereas H3K4me3 appears preferentially located in less densely stained regions, detailed inspection shows an occasional extension of H3K9me3 positive pixels also into regions with weakly DAPI-stained chromatin and on the contrary an extension of H3K4me3 positive pixels into intensely DAPI-stained regions (see Results part 3 for quantitative analysis). Bars: 2 μm for A1–C2 and c; 1 μm for a1–b2 and d1–h2.

Figure 12.

Comparison of nuclear phenotypes from bovine fetal fibroblasts and cloned preimplantation embryos studied with 3D structured illumination microscopy. Panels A1–A8. Midplane SIM x/y-sections (left) and y/z-sections (right) from DAPI-stained fetal fibroblast nuclei indicate their common flat-ellipsoidal shape. DAPI-stained sections are shown together with immunostaining of nucleophosmin B23 (red) and of H3K4me3 (green) in A1, A3, A5 and A7. Corresponding sections with DAPI staining alone are presented in A2, A4, A6 and A8. Panels B1–B8. Midplane SIM x/y-sections (left) and y/z-sections (right) from DAPI-stained nuclei in a cloned 8-cell embryo demonstrate the transformation of the flat-ellipsoidal shape of fibroblast nuclei into a roundish shape with a more pronounced clustering of chromatin and the formation of a major lacuna, marked by asterisks (compare Figure 7, Panels J-Q). Panels C1–C4 and D1–D4. Four nuclei recorded by 3D-SIM from a cloned, non-hatched blastocyst (C1–C4) and 4 nuclei from a cloned, hatched blastocyst (D1–D4). These nuclei were present in cells, which maintained their connection during microdissection of the embryos for 3D-SIM (see Results and Extended Experimental Procedures) and likely represent trophectoderm nuclei. Bar: 4 μm in A1 representative for all Panels.

Figure 13

(See previous page). Topography of DAPI-stained chromatin, nucleophosmin B23, RNA polymerase II-Ser2p, H3K4me3 and H3K9me3 in nuclei of bovine fetal fibroblasts studied with 3D structured illumination microscopy. Panels A–C. SIM midplane x/y-sections of DAPI-stained bovine fibroblast nuclei (A1, B1, C1) and x/z-sections (A3, B3, C3) are shown together with corresponding immunostainings of H3K4me3 (green) and nucleophosmin B23 (red) (A2, A4), H3K4me3 and RNA polymerase II-Ser2p (red) (B2, B4), H3K4me3 and H3K9me3 (red) (C2, C4). A fine line marked by pairs of arrows in A2/A4, B2/B4 and C2/C4 indicates the sites of the respective x/z- and x/y-sections. All nuclei contain major chromatin domain clusters with a modestly DAPI-stained core and an intensely DAPI-stained periphery. Examples are framed by white boxes. The yellow boxes in B2 and B4 indicate a major chromatin cluster in the nuclear interior connected to chromatin at the nuclear border. Panels a-d. Enlargements of white boxed areas in the 3 nuclei show major chromatin clusters with a modestly DAPI-stained core (light blue) and an intensely DAPI-stained periphery (dark blue) (a1–d1), overlays of pixels representing immunostained markers are shown in the corresponding images a2–d2. Black pixels represent colocalization of the 2 immunostained markers. Panel a presents a major chromatin cluster, which interrupts the B23 stained rim of a nucleolus. Note also the lining of an IC lacuna with H3K4me3. Panel b indicates that the core part of the chromatin domain cluster is modestly labeled with H3K4me3 but lacks RNA polymerase II-Ser2p. IC-lacunas are lined with both H3K4me3 and RNA polymerase II-Ser2p. Panel c and d show intense H3K9me3 label enriched at the intensely DAPI-stained periphery of major chromatin clusters. Bars: 3 μm for A1–C4; 2 μm for a1–d2.

Figure 14.

Topography of H3K4me3 and H3K9me3 observed with 3D structured illumination microscopy in nuclei from a cloned 8-cell embryo and a cloned blastocyst. Panels A and B. Midplane SIM sections from an ENP-like (Panel A) and an ENP/C-like nucleus (Panel B) recorded from a cloned 8-cell embryo with DAPI-stained chromatin (gray), H3K4me3 (green) and H3K9me3 (red). Panels C and D. Midplane SIM sections from 2 DAPI-stained ENCs of a cloned, hatched blastocyst with immunostaining of the 2 epigenetic markers. The 4 example nuclei show prominent major chromatin clusters, some of them intensely, others modestly stained with DAPI, but all prominently labeled with H3K9me3. Panels a to h. Enlargements of boxed areas demonstrate strongly H3K9me3 labeled major chromatin clusters, which cover both modestly and strongly DAPI-stained chromatin. In ENP- and ENP/C-like nuclei such clusters are surrounded by H3K4me3 labeled chromatin, but contain little H3K4me3 in their interior (Panels a-d). Black pixels denote colocalization events of H3K4me3 and H3K9me3 positive pixels, which mostly occurred at the periphery of a given major chromatin cluster. In ENCs of the cloned, hatched blastocyst we found numerous major chromatin clusters, modestly stained with DAPI and strongly labeled with both H3K4me3 and H3K9me3 (for examples compare e1/e2 and g1/g2). Panel h1/h2 shows 3 major chromatin clusters. Black pixels indicate colocalization of H3K4me3 and H3K9me3 throughout the interior of 2 modestly DAPI-stained clusters (upper left corner and bottom). An intensely DAPI-stained major chromatin cluster (upper right corner) shows only few black pixels indicating colocalization of H3K4me3 and H3K9me3 in its interior, and an enrichment at the periphery. Bars: 3 μm for A1-D2; 1 μm for a1–h2.

Figure 15.

H3K4me3 and RNA polymerase II-Ser2p arrangements observed with 3D structured illumination microscopy in nuclei from a cloned 8-cell embryo and a cloned, non-hatched blastocyst. Panels A and B. SIM midplane sections recorded from an ENP-like nucleus (panel A) and an ENP/C-like nucleus (panel B) of a cloned, DAPI-stained (A1, B1) 8-cell embryo following immunostaining of H3K4me3 (green) and RNA polymerase II-Ser2p (red) (A2, B2). Panels C and D. SIM midplane sections from 2 DAPI-stained nuclei (C1, C2) from a cloned, non-hatched blastocyst demonstrate dispersed RNA polymerase II-Ser2p together with dispersed H3K4me3, as well as large, H3K4me3 labeled chromatin clusters (C2, D2), particularly prominent in D2 (compare Fig. 14, Panels D, G and H). Enlarged images of boxed areas in these nuclei (Panels a-h) show major chromatin clusters strongly marked with H3K4me3 but lacking RNA polymerase II-Ser2p. An enrichment of RNA polymerase II-Ser2p together with H3K4me3 is noted in chromatin lining IC-channels/lacunas. Bars: 3 μm for A1–D2; 1 μm for a1–h2.

Figure 16.

Differences between major chromatin clusters observed with 3D structured illumination microscopy in fetal fibroblast nuclei and nuclei from cloned, non-hatched and hatched blastocysts. Panel A. Corresponding midplane SIM sections from a typical DAPI-stained (A1) bovine fetal fibroblast nucleus with immunostained H3K4me3 (A2) and H3K9me3 (A3), overlay (A4; H3K4me3 green; H3K9me3 red) demonstrate numerous major chromatin domain clusters strongly labeled with H3K9me3 and sparsely with H3K4me3. Panel B. In contrast, the ENC recorded from a cloned, non-hatched blastocyst presents major chromatin domain clusters, strongly labeled with H3K4me3. Panels C and D. Two DAPI-stained ENCs (C1, D1) from a cloned, hatched blastocyst reveal major chromatin clusters strongly labeled with H3K9me3 (C3, D4), but differ with regard to H3K4me3 (C2, D2). Whereas this marker is apparently enriched in 2 of 3 clusters noted in one ENC (Panel C), no enrichment of H3K4me3 was detected in the major chromatin clusters present in the other ENC (panel D) similar to the pattern noted in the fibroblast nucleus (panel A). E. A quantitative assessment of the labeling patterns of major chromatin clusters in 57 fibroblast nuclei (A), 17 nuclei from cloned, non-hatched blastocysts (B) and 11 nuclei from cloned, hatched blastocysts (C, D) demonstrates the reproducibility of the similarities and differences of patterns described for the 4 example nuclei. Bar: 3 μm in A1 representative for all Panels.

Figure 17.

Quantitative assessment of different nuclear phenotypes recorded with 3D structured illumination microscopy during preimplantation development of in vitro fertilized and cloned embryos. A. Nuclei from 9 fertilized embryos with 2 cells (E2; 17 nuclei), 6 embryos with 3 or 4 cells (E3–4; 22 nuclei), 7 embryos with 7 or 8 cells (E7–8; 56 nuclei), 8 embryos between 9 and 15 cells (E9–15; 66 nuclei) and 6 embryos with more than 18 cells on their way toward the morula stage (about 32 cells) (E >18; 98 nuclei) were classified as pre-ENP, ENP, ENP/C and ENC according to the criteria described in Results and the examples presented in Figure 7. B. Classification of nuclei from 5 cloned 8-cell stage embryos (E8; 40 nuclei) and 2 blastocysts (28 nuclei) as pre-ENP-like, ENP-like, ENP/C-like and ENC was performed as described in Results.

Figure 18

(See previous page). Typical examples of nuclei with color-coded DAPI intensity classes recorded with 3D structured illumination microscopy in in vitro fertilized and cloned embryos. Panels A–E. Midplane SIM sections with color-coded DAPI intensity classes 1 to 7 from typical nuclear phenotypes observed in in vitro fertilized embryos (A, pre-ENP; B, transition from pre-ENP to ENP; C, ENP; D, ENP/C; E, ENC). Classes 1 to 7 present increasing DAPI intensity. The color code is presented at the top. Class 1 coded blue, class 2 purple, class 3 dark red, class 4 light red, class 5 orange, class 6 yellow and class 7 white. A1-E1 provide color-coded SIM sections after thresholding to remove patterns of concentric rings and diffuse background (see Fig. 6). A2–E2 show the same sections before thresholding. Note that the size of the areas occupied by the lowest DAPI density classes, in particular class 1, are strongly increased in thresholded compared to non-thresholded sections. For quantitative assessments of the nuclear topography of RNA polymerase II, H3K4me3 and H3K9me3 arrangements presented in Figure 19 with respect to these DAPI intensity classes we used non-thresholded, color-coded SIM sections (for further details see Supplementary Fig. S4, Results and Extended Experimental Procedures). Panels F–J. Midplane SIM sections with color-coded DAPI intensity classes from typical nuclear phenotypes observed in cloned embryos (F, pre-ENP-like; G, transition from pre-ENP to ENP-like state; H, ENP-like; I, ENP/C-like; J, ENC), F1–J1 and F2–J2, color-coded SIM sections after and before thresholding. Panels a-d. Enlarged views of boxed areas depicted in 2 example nuclei (E1/E2, H1/H2) are presented in panels a/b and c/d. The enlarged box from the ENP-like nucleus shown in c demonstrates particularly well chromatin domain clusters (CDCs), separated by IC-channels (blue). Red color in the periphery of CDCs signifies less compact DAPI-stained chromatin, yellow and white colors more compacted DAPI-stained chromatin in the core part of CDCs. The strong difference between thresholded and corresponding non-thresholded images emphasizes limitations of our current approach with regard to the resolution achieved with 3D-SIM and the sensitivity of DAPI staining to distinguish unequivocally between chromatin and DNA-free parts of the interchromatin compartment. Bars: 3 μm for A1–J2; 1 μm for a-d.

Figure 19.

Quantitative analysis of topographical relationships between DAPI-stained chromatin, H3K4me3, H3K9me3 and RNA polymerase II in nuclei from in vitro fertilized embryos, cloned embryos and fetal bovine fibroblasts. A. ENPs from in vitro fertilized embryos (n = 20). B. ENCs from in vitro fertilized embryos (n = 20). C. ENP-like nuclei from cloned embryos (n = 15). D. ENCs from cloned embryos (n = 20). E. Nuclei from growing, non-synchronized fetal bovine fibroblast cultures (n = 20). For each sample A – E our analysis comprised 10 nuclei immunostained for H3K4me3 and H3K9me3 (cloned ENP-like n=7) and 10 nuclei immunostained for H3K4me3 and RNA polymerase II (cloned ENP-like n=8). In each nucleus 3 to 6 consecutive mid-nuclear SIM sections were analyzed. For each nucleus 7 DAPI intensity classes were distinguished in unthresholded midplane sections covering the entire range from lowest to highest pixel intensity values (Fig. 18; for details see Supplementary Information: Extended Experimental Procedures). The height of columns presents a relative measure for the overrepresentation (+) or underrepresentation (-) of H3K4me3 (green), H3K9me3 (red) and RNA polymerase II (blue) positive pixels, respectively, calculated as the relative frequency of such pixels in comparison to the number of DAPI positive pixels attributed to classes 1 to 7. For all tested combinations of DAPI, H3K4me3 and RNA Polymerse II thresholds (compare Supplementary Figs. S2, S3 and S4) the null-hypothesis of a random class assignment was rejected (P < 10⁻¹⁵). A comparison of A-E indicates striking similarities of local RNA polymerase II, H3K4me3 and H3K9me3 assignments to the 7 DAPI intensity classes despite the striking differences between the global nuclear architectures of ENPs and ENP-like nuclei compared with ENCs and fibroblast nuclei. The overrepresentation of RNA polymerase II signals in class 4 (P < 10⁻¹⁴) of ENPs, ENP-like nuclei, ENCs and fibroblast nuclei supports the hypothesis that the major transcriptionally active nuclear compartment is located at the periphery of chromatin domain clusters, called the perichromatin region (PR) (compare Fig. 18). In ENPs and ENCs of in vitro fertilized embryos (A, B) the immunostained RNA polymerase II carried a C-terminal domain (CTD) with a phosphorylated serine 5 (Ser5p), whereas the polymerase immunodetected in ENP-like nuclei and ENCs from cloned embryos (C, D), as well as fibroblast nuclei (E) carried a CTD with a phosphorylated serine 2 (S2P). The topography of RNA polymerase II in relation to the 7 DAPI intensity classes was essentially the same for both serine 2 and 5 phosphorylations.

Figure 20.

A high-resolution model view of the functional nuclear landscape. Evidence stems in part from the present study and in part from the literature (see Discussion). (A and B) Representative landscape of an ENC. B shows an enlargement of the boxed area in A. CTs are built up from interconnected basic chromatin domains with a DNA content in the order of a few hundred kbp, which form higher order chromatin structures, such as chromatin domain clusters (CDCs), chromosome arm and band domains (for further details see Discussion). Identification of individual structures in microscopic images requires their individual visualization. CDCs carry transcriptionally silent chromatin (red) in their interior. The less dense, transcriptionally competent chromatin (green) at the periphery of CDCs presents the perichromatin region (PR). It is enriched with transcription factories (blue) and represents the nuclear sub-compartment, where genes are transcribed. The PR lines the interchromatin compartment (IC), which starts with channels at nuclear pores and forms a 3-dimensional network throughout the nuclear interior. It extends between neighboring CTs but also throughout CTs. At numerous sites the IC forms larger lacunas, which contain nuclear bodies, such as splicing speckles. Considering the potential role of the IC in import-export functions, as well as functional interactions between nuclear bodies and machineries for transcription, co-transcriptional splicing, chromatin replication and repair acting within the PR, we propose that the PR and the IC provide the active nuclear compartment (aNC), whereas the compact interior of CDCs forms the inactive nuclear compartment (iNC). The nucleolus in A is marked with ‘n’. (C). This model presents a hypothetical topography at still higher resolution envisaged for a landscape at the nuclear periphery with nuclear pores connected to IC channels. Little chromatin loops invade these channels, their size and hypothetical arrangements minimize problems of chromatin entanglements in the perichromatin region and interchromatin compartment. A dotted circle comprises a chromatin domain with a DNA content of about 500 kbp. Chromatin domains attached to the nuclear lamina (yellow) are connected with domains extended into the nuclear interior (D). Comparison of the nuclear landscape shown in C with little streams and ponds pervading a natural landscape. Note some vegetation expanding from embankments into the inner part of streams and ponds (images recorded in the Spreewald southeast of Berlin and provided by courtesy of Marion Cremer). Bars: 2 μm for A; 1 μm for B; 200 nm for C.