Automated matching of genomic structures in microscopic images of living cells using an information theoretic approach.

Advances in microscopic imaging technology have revolutionized biology in recent years by enabling the study of dynamic processes inside living cells. Time-lapse microscopy produces large numbers of sequential images of living cells taken over time. In this paper we describe the novel approaches we have developed to automate and introduce high accuracy to the process of identifying genomic structures in living cells and matching them between consecutive time-points. We derive control points from landmarks within the structures and use the Kulback-Leibler divergence as an information-theoretic approach to correctly resolve potential close matches within the iterative closest point (ICP) algorithm. We also describe the steps needed to extend our techniques to analyze three dimensional voxel images. The approaches we describe are widely applicable in the analysis of timelapse microscopy data.

Fractal analysis of replication site images of the human cell nucleus.

Epi-fluorescent microscopic images of the mammalian cell nucleus taken during the early, mid and late S (synthetic) phase of the cell cycle suggest that the mass of replicating DNA that belong to the cell nucleus can be characterized as a space filling fractal curve. We reason from a biological standpoint and our understanding of naturally occurring fractals that our microscopic images reveal portions of the spatially complex DNA molecule and present methods for computing the fractal dimensions of the images. Results presented here suggest that our methodology based on fractal properties can distinguish replication of DNA occurring in early versus mid or late S-phase.

Targeting of PCNA to sites of DNA replication in the mammalian cell nucleus.

We have examined the targeting of proliferating cell nuclear antigen (PCNA), an integral component of the mammalian replicative enzyme DNA polymerase delta, with sites of DNA replication by using confocal microscopy and computer image analysis. Labeling (5 min pulse) of DNA replication sites in normal human diploid fibroblast cells (NHF1) with BrdU was followed by immunostaining with PCNA antibodies. A striking degree of colocalization was seen between PCNA and the characteristic patterns of DNA replication sites of early, middle and late S-phase (Nakayasu and Berezney [1989] J. Cell. Biol. 108:1-11). These observations were confirmed by quantitative computer image analysis which revealed that approximately 90% of the PCNA-stained area overlapped with DNA replication sites in early S-phase. Pulse-chase experiments, involving in vivo labeling for replication followed by PCNA staining at later time points, suggested that PCNA disassembles from previously replicated sites and targets to newly active sites of DNA replication. To further study this phenomenon in living cells, stable GFP-PCNA transfectants under the control of a tetracycline-inducible promoter were created in mouse 3T6 cells. Like the endogenous PCNA, GFP-PCNA targeted to sites of replication (approximately 80% colocalization) and demonstrated similar dynamic changes following pulse-chase experiments in fixed cells. Studies of living cells revealed progressive changes in the GFP-PCNA distribution that mimic the replication patterns observed in fixed cells. We conclude that GFP-PCNA targets to DNA replication sites in living cells and is an effective marker for tracking the spatio-temporal dynamics of DNA replication as cells transverse the S-phase.

Heterogeneity of eukaryotic replicons, replicon clusters, and replication foci.

According to the current paradigm, replication foci are discrete sites in the interphase nucleus where assemblies of DNA replication enzymes simultaneously elongate the replication forks of 10-100 adjacent replicons (each approximately 100 kbp). Here we review new results and provide alternative interpretations for old results to show that the current paradigm is in need of further development. In particular, many replicons are larger than previously thought - so large that their complete replication takes much longer (several hours) than the measured average time to complete replication at individual foci (45-60 min). In addition to this large heterogeneity in replicon size, it is now apparent that there is also a corresponding heterogeneity in the size and intensity of individual replication foci. An important property of all replication foci is that they are stable structures that persist, with constant dimensions, during all cell cycle stages including mitosis, and therefore likely represent a fundamental unit of chromatin organization. With this in mind, we present a modified model of replication foci in which many of the foci are composed of clusters of small replicons as previously proposed, but the size and number of replicons per focus is extremely heterogeneous, and a significant proportion of foci are composed of single large replicons. We further speculate that very large replicons may extend over two or more individual foci and that this organization may be important in regulating the replication of such large replicons as the cell proceeds through S-phase.

Association of chromosome territories with the nuclear matrix. Disruption of human chromosome territories correlates with the release of a subset of nuclear matrix proteins.

To study the possible role of the nuclear matrix in chromosome territory organization, normal human fibroblast cells are treated in situ via classic isolation procedures for nuclear matrix in the absence of nuclease (e.g., DNase I) digestion, followed by chromosome painting. We report for the first time that chromosome territories are maintained intact on the nuclear matrix. In contrast, complete extraction of the internal nuclear matrix components with RNase treatment followed by 2 M NaCl results in the disruption of higher order chromosome territory architecture. Correlative with territorial disruption is the formation of a faint DNA halo surrounding the nuclear lamina and a dispersive effect on the characteristically discrete DNA replication sites in the nuclear interior. Identical results were obtained using eight different human chromosome paints. Based on these findings, we developed a fractionation strategy to release the bulk of nuclear matrix proteins under conditions where the chromosome territories are maintained intact. A second treatment results in disruption of the chromosome territories in conjunction with the release of a small subset of acidic proteins. These proteins are distinct from the major nuclear matrix proteins and may be involved in mediating chromosome territory organization.

Three-dimensional visualization of transcription sites and their association with splicing factor-rich nuclear speckles.

Transcription sites are detected by labeling nascent transcripts with BrUTP in permeabilized 3T3 mouse fibroblasts followed by laser scanning confocal microscopy. Inhibition and enzyme digestion studies confirm that the labeled sites are from RNA transcripts and that RNA polymerase I (RP I) and II (RP II) are responsible for nucleolar and extranucleolar transcription, respectively. An average of 2,000 sites are detected per nucleus with over 90% in the extranucleolar compartment where they are arranged in clusters and three-dimensional networklike arrays. The number of transcription sites, their three-dimensional organization and arrangement into functional zones (Wei et al. 1998) is strikingly maintained after extraction for nuclear matrix. Significant levels of total RP II mediated transcription sites (45%) were associated with splicing factor-rich nuclear speckles even though the speckles occupied <10% of the total extranucleolar space. Moreover, the vast majority of nuclear speckles (>90%) had moderate to high levels of associated transcription activity. Transcription sites were found along the periphery as well as inside the speckles themselves. These spatial relations were confirmed in optical sections through individual speckles and after in vivo labeling of nascent transcripts. Our results demonstrate that nuclear speckles and their surrounding regions are major sites of RP II-mediated transcription in the cell nucleus, and support the view that both speckle- and nonspeckle-associated regions of the nucleus contain sites for the coordination of transcription and splicing processes.

Spatial and temporal dynamics of DNA replication sites in mammalian cells.

Fluorescence microscopic analysis of newly replicated DNA has revealed discrete granular sites of replication (RS). The average size and number of replication sites from early to mid S-phase suggest that each RS contains numerous replicons clustered together. We are using fluorescence laser scanning confocal microscopy in conjunction with multidimensional image analysis to gain more precise information about RS and their spatial-temporal dynamics. Using a newly improved imaging segmentation program, we report an average of approximately 1,100 RS after a 5-min pulse labeling of 3T3 mouse fibroblast cells in early S-phase. Pulse-chase-pulse double labeling experiments reveal that RS take approximately 45 min to complete replication. Appropriate calculations suggest that each RS contains an average of 1 mbp of DNA or approximately 6 average-sized replicons. Double pulse-double chase experiments demonstrate that the DNA sequences replicated at individual RS are precisely maintained temporally and spatially as the cell progresses through the cell cycle and into subsequent generations. By labeling replicated DNA at the G1/S borders for two consecutive cell generations, we show that the DNA synthesized at early S-phase is replicated at the same time and sites in the next round of replication.

Segregation of transcription and replication sites into higher order domains.

Microscopy shows that individual sites of DNA replication and transcription of mammalian nuclei segregate into sets of roughly 22 and 16 higher order domains, respectively. Each domain set displayed a distinct network-like appearance, including regions of individual domains and interdigitation of domains between the two networks. These data support a dynamic mosaic model for the higher order arrangement of genomic function inside the cell nuclei.

Matrin CYP, an SR-rich cyclophilin that associates with the nuclear matrix and splicing factors.

We report the identification and cloning of a nuclear matrix protein termed matrin cyclophilin or matrin CYP. The derived sequence of matrin cyp encodes a protein of 752 amino acids with a predicted mass of 88 kDa. A 172-residue stretch at the amino terminus shows high identity with the ubiquitous family of cyclophilins. Clustered throughout the carboxyl half of the protein are a series of serine-arginine (SR) repeats that are a characteristic feature of many RNA splicing factors. Antibodies raised against matrin CYP recognize a 106-kDa antigen that is detected in isolated nuclei and quantitatively subfractionates in the nuclear matrix. Laser scanning confocal microscopy localizes most of the anti-matrin CYP-specific antigen within the nucleus in a pattern of large bright speckles that co-localize with splicing factors and diffuse nucleoplasmic staining. A strikingly similar pattern of staining is observed in cells extracted for in situ nuclear matrices. A fusion protein containing the cyclophilin domain of matrin CYP exhibits cyclosporin A (CsA)-sensitive, peptidylprolyl cis-trans-isomerase activity that is characteristic of native cyclophilins. Although total rat liver nuclei contains predominantly CsA-resistant PPIase activity, the corresponding activity in the nuclear matrix is largely CsA-sensitive.

A nuclear matrix protein interacts with the phosphorylated C-terminal domain of RNA polymerase II.

Yeast two-hybrid screening has led to the identification of a family of proteins that interact with the repetitive C-terminal repeat domain (CTD) of RNA polymerase II (A. Yuryev et al., Proc. Natl. Acad. Sci. USA 93:6975-6980, 1996). In addition to serine/arginine-rich SR motifs, the SCAFs (SR-like CTD-associated factors) contain discrete CTD-interacting domains. In this paper, we show that the CTD-interacting domain of SCAF8 specifically binds CTD molecules phosphorylated on serines 2 and 5 of the consensus sequence Tyr1Ser2Pro3Thr4Ser5Pro6Ser7. In addition, we demonstrate that SCAF8 associates with hyperphosphorylated but not with hypophosphorylated RNA polymerase II in vitro and in vivo. This result suggests that SCAF8 is not present in preinitiation complexes but rather associates with elongating RNA polymerase II. Immunolocalization studies show that SCAF8 is present in granular nuclear foci which correspond to sites of active transcription. We also provide evidence that SCAF8 foci are associated with the nuclear matrix. A fraction of these sites overlap with a subset of larger nuclear speckles containing phosphorylated polymerase II. Taken together, our results indicate a possible role for SCAF8 in linking transcription and pre-mRNA processing.

Growth-related changes in phosphorylation of yeast RNA polymerase II.

The largest subunit of RNA polymerase II contains a unique C-terminal domain (CTD) consisting of tandem repeats of the consensus heptapeptide sequence Tyr1-Ser2-Pro3-Thr4-Ser5-Pro6-Ser7. Two forms of the largest subunit can be separated by SDS-polyacrylamide gel electrophoresis. The faster migrating form termed IIA contains little or no phosphate on the CTD, whereas the slower migrating II0 form is multiply phosphorylated. CTD kinases with different phosphoryl acceptor specificities are able to convert IIA to II0 in vitro, and different phosphoisomers have been identified in vivo. In this paper we report the binding specificities of a set of monoclonal antibodies that recognize different phosphoepitopes on the CTD. Monoclonal antibodies like H5 recognize phosphoserine in position 2, whereas monoclonal antibodies like H14 recognize phosphoserine in position 5. The relative abundance of these phosphoepitopes changes when growing yeast enter stationary phase or are heat-shocked. These results indicate that phosphorylation of different CTD phosphoacceptor sites are independently regulated in response to environmental signals.

The new paradigm: integrating genomic function and nuclear architecture.

A new view of the cell nucleus is emerging based on the functional dynamics of nuclear architecture. The striking structural preservation of a variety of genomic processes on the nuclear matrix provides an important approach for correlating nuclear form and function. In situ labeling coupled with three-dimensional microscopy and computer imaging techniques shows that DNA replication and transcription sites are organized into higher-order units, or "zones," in the cell nucleus. The dynamic interplay and "re-zoning" of replication and transcription regions during the cell cycle may form the structural basis for the elaborate global coordination of replicational and transcriptional programs in the mammalian cell.

Three-dimensional laser scanning two-photon fluorescence confocal microscopy of polymer materials using a new, efficient upconverting fluorophore.

Three-dimensional confocal imaging of polymer samples was achieved by the use of two-photon excited fluorescence in both positive and negative contrast modes. The fluorophore was a new and highly efficient two-photon induced upconverter, resulting in improved signal strength at low pumping power. Because of the relatively long wavelength of the excitation source (798 nm from a mode-locked Ti:Sapphire laser), this technique shows a larger penetration depth into the samples than provided by conventional single-photon fluorescence confocal microscopy. Single-photon and two-photon images of the same area of each sample show significant differences. The results suggest the possibility of using two-photon confocal microscopy, in conjunction with highly efficient fluorophores, as a tool to study the surface, interface, and fracture in material science applications.

A hyperphosphorylated form of the large subunit of RNA polymerase II is associated with splicing complexes and the nuclear matrix.

A hyperphosphorylated form of the largest subunit of RNA polymerase II (pol IIo) is associated with the pre-mRNA splicing process. Pol IIo was detected in association with a subset of small nuclear ribonucleoprotein particle and Ser-Arg protein splicing factors and also with pre-mRNA splicing complexes assembled in vitro. A subpopulation of pol IIo was localized to nuclear "speckle" domains enriched in splicing factors, indicating that it may also be associated with RNA processing in vivo. Moreover, pol IIo was retained in a similar pattern following in situ extraction of cells and was quantitatively recovered in the nuclear matrix fraction. The results implicate nuclear matrix-associated hyperphosphorylated pol IIo as a possible link in the coordination of transcription and splicing processes.

The nuclear matrix: a structural milieu for genomic function.

While significant progress has been made in elucidating molecular properties of specific genes and their regulation, our understanding of how the whole genome is coordinated has lagged behind. To understand how the genome functions as a coordinated whole, we must understand how the nucleus is put together and functions as a whole. An important step in that direction occurred with the isolation and characterization of the nuclear matrix. Aside from the plethora of functional properties associated with these isolated nuclear structures, they have enabled the first direct examination and molecular cloning of specific nuclear matrix proteins. The isolated nuclear matrix can be used for providing an in vitro model for understanding nuclear matrix organization in whole cells. Recent development of high-resolution and three-dimensional approaches for visualizing domains of genomic organization and function in situ has provided corroborative evidence for the nuclear matrix as the site of organization for replication, transcription, and post-transcriptional processing. As more is learned about these in situ functional sites, appropriate experiments could be designed to test molecular mechanisms with the in vitro nuclear matrix systems. This is illustrated in this chapter by the studies of nuclear matrix-associated DNA replication which have evolved from biochemical studies of in vitro nuclear matrix systems toward three-dimensional computer image analysis of replication sites for individual genes.

Nuclear matrins: identification of the major nuclear matrix proteins.

A preparative two-dimensional polyacrylamide gel system was used to separate and purify the major Coomassie blue-stained proteins from the isolated rat liver nuclear matrix. Approximately 12 major proteins were consistently found. Of these, 5 proteins represented identified proteins, including nuclear lamins A, B, and C, the nucleolar protein B-23, and residual components of core heterogeneous nuclear ribonucleoproteins. The remaining eight major proteins termed the nuclear matrins consisted of matrin 3 (125 kDa, slightly acidic), matrin 4 (105 kDa, basic), matrins D-G (60-75 kDa, basic), and matrins 12 and 13 (42-48 kDa, acidic). Peptide mapping and two-dimensional immunoblot studies indicate that matrins D-G compose two pairs of related proteins (matrins D/E and F/G) and that none of the matrins resemble the nuclear lamins or any of the other major proteins detected on our two-dimensional gels. Subfractionation immunoblot experiments demonstrated the nearly exclusive localization of matrins F/G and other matrins to the nuclear matrix fraction of the cell. These results were further supported by indirect immunofluorescence microscopy that showed a strictly interior nuclear localization of the matrins in intact cells in contrast to the peripherally located nuclear lamins. We conclude that the nuclear matrins are a major class of proteins of the nuclear matrix interior and are distinct from the nuclear lamins.