Chromosome duplication and structure as determined by autoradiography

Discovering non-random segregation of sister chromatids: the naïve treatment of a premature discovery

The discovery of non-random chromosome segregation (Figure 1) is discussed from the perspective of what was known in 1965 and 1966. The distinction between daughter, parent, or grandparent strands of DNA was developed in a bacterial system and led to the discovery that multiple copies of DNA elements of bacteria are not distributed randomly with respect to the age of the template strand. Experiments with higher eukaryotic cells demonstrated that during mitosis Mendel's laws were violated; and the initial serendipitous choice of eukaryotic cell system led to the striking example of non-random segregation of parent and grandparent DNA template strands in primary cultures of cells derived from mouse embryos. Attempts to extrapolate these findings to established tissue culture lines demonstrated that the property could be lost. Experiments using plant root tips demonstrated that the phenomenon exists in plants and that it was, at some level, under genetic control. Despite publication in major journals and symposia (Lark et al., 1966, 1967; Lark, 1967, 1969a,b,c) the potential implications of these findings were ignored for several decades. Here we explore possible reasons for the pre-maturity (Stent, 1972) of this discovery.

Sister chromatid differentiation and isolabeling of chromosomes

Isolabeling observed during sister chromatid differentiation (SCD) was studied from human skin fibroblasts by the fluorescence-plus-Giemsa (FPG) technique. Bromodeoxyuridine (BrdU) was fed to exponentially dividing cells for 52 h to enable completion of two consecutive cycles of DNA replication. During this period, the late-replicating regions of some chromosomes were able to go through three replication cycles. These chromosome regions had evidently incorporated BrdU bifiliarly in both chromatids and hence, on staining with FPG, appeared isostained (isolabeled). Thus, incubation of exponentially dividing cells with BrdU for a period longer than that required for two cell cycles appears to be a suitable method for revealing the late-replicating regions of the genome, such as the X chromosome in a human female, as isolabeled. In another experiment with Indian muntjac chromosomes, isolabeled segments were darkly stained, which suggested unifilar incorporation of BrdU. In this case, unequal crossing-over or an unequal distribution of thymine residues probably is responsible for the isolabel.

Chromosomal isolabelling caused by three rounds of synthesis in late replicating regions

Isolabelling only occurs in CHO cells that have been allowed to replicate for more than 2 but less than 3 cell divisions in the presence of BrdU. The isolabelling is confined to late replicating regions of the chromosomes. The staining patterns obtained indicate that BrdU was incorporated three times in these regions and that the isolabelling did not come from the segregation of label in polynemic chromosomes.

Isolabeling of the long arm of the human Y chromosome demonstrated by the FPG technique

Isolabeling segments were found in the distal region of the long arm of Y chromosomes derived from human leukocytes grown through two replication cycles in medium containing BrdU and stained by the FPG technique. Three main types of Y chromosome staining patterns were demonstrated: I-Y chromosome with typical SCD, II-Y chromosome with weakly stained distal regions of long arms (isolabeling segments), III-Y chromosome with both terminal regions displaying SCD interrupted by one isolabeled segment. The existence of different types of Y chromosome staining patterns was explained on the basis of the previously described hypothesis of unequal distribution of thymine residues between two DNA polynucleotide chains in the distal part of the long arms of human Y chromosomes.

Isolabelling is a radiation-induced phenomenon

Human lymphocytes were incubated during two mitotic cycles in the presence of 5-bromodeoxyuridine and differentiation between chromatids was obtained with combined "Hoechst 33258" and azur-eosine staining. Analysis of non-irradiated cells revealed numerous sister chromatid exchanges (SCE) and no abnormalities of "harlequine" appearance of chromosomes. When, however, the cells were irradiated, an identical staining (IS, isostaining) of some chromosomes or chromosome segments were observed. Production of IS was accompanied by decrease of the frequency of SCE, the total frequency of SCE+IS remained, however, the same as in control. An antagonism between SCE and IS was established: the frequency of SCE decreased in the cells with multiple IS, and chromosomes with both SCE and IS were only rarely observed. Thus, IS is neither an artifact nor a physiologic event but a phenomenon induced by radiation. The reliable existence of IS is considered as an evidence for binemic structure of chromatid. It is suggested that some mechanism of lateral spread of genetic information is involved in the production of SCE. If delayed by radiation, the spread could be restricted only to a fraction of chromosome cross-section resulting in IS.

DNA synthesis during meiosis of eight-spored strains of Chlamydomonas reinhardi

In strain 137F of Chlamydomonas reinhardi, the zygospores undergo one round of nuclear DNA replication followed by three divisions to produce octospores. The third division without replication has been interpreted by Sueoka et al. (1967, 1969) to mean that the gametes and vegetative cells have at least binemic chromosomes. We have repeated their experiments using the same strain. However, the meiotic products were inviable--unable to undergo postmeiotic vegetative growth, DNA replication or division. On the other hand, using a variant of strain 137C which also has three divisions during germination we have shown that meiosis is normal. Zygospores from this strain undergo two rounds of nuclear DNA replication prior to the formation of octospores. These meiotic products are viable and capable of postmeiotic vegetative growth, replication and division. Since the third division without DNA replication subsequent to the two meiotic divisions leads to inviable products, and the strain which has viable products after three divisions does not lack the additional replication, meiosis in Chlamydomonas reinhardi provides no evidence of a bineme chromosome structure.

The human leukocyte test system. VI. The use of sister chromatid exchanges as possible indicators for mutagenic activities

The trifunctional alkylating chemical mutagen trenimon increases the frequency of sister chromatid exchanges in human leukocyte chromosomes in vitro, as revealed by a BUdR-Giemsa method. Treatment with lead acetate exhibited negative results in this respect. The use of sister chromatid exchanges as possible indicators for mutagenic activities is discussed.

Resolving the enigma of multiple mutant sectors in stamen hairs of tradescantia

Mutant sectors in stamen hairs of Clone 02 Tradescantia are designated as "multiple sectors" when two or more occupy the same hair, separated by non-mutant cells. Statistical analyses show that most multiple sectors do not arise as chance associations of independent events: when the frequency of stamens with two or more sectors is lowest, the probability that the sectors will be located in the same, rather than in different, hairs is highest. Ontogenetically, the ratio of sector pairs in different hairs to pairs in the same hair is highest in that period of response to acute irradiation prior to the appearance of entire-hair sectors; thereafter, the ratio subsides, approaching that of spontaneous mutation and indicating that the initiating event takes place early in hair development. Most mononemic chromosome models will not account for the production of multiple mutant and non-mutant sectors, dispersed along a linear structure such as a stamen hair, following a single mutational event. Consideration is given to two models (one mononemic, and the other dinemic) which will readily provide the possibilities for either the immediate segregation of mutant and non-mutant cells, or for the perpetuation in daughter nuclei of a "heterozygous" chromosome capable of segregation at some later mitosis. The dinemic model is preferred because it affords operation of the mutation mechanism (including breakage and deletion) at either the DNA molecule or subunit level.

Replication of DNA in the chromosomes of eukaryotes

The evidence that each chromatid of a eukaryotic organism contains only one DNA double helix comes from a variety of observations. It begins with the autoradiographic demonstration by J. H. Taylor that tritiated thymidine, incorporated into chromosomes during one round of DNA synthesis, is present in both chromatids at the first division after labelling, but in only one chromatid after a further round of DNA synthesis accomplished in the absence of label. Further evidence comes from those experiments which demonstrate that when two sister chromatids break and fuse one with the other, each chromatid behaves as though it contained two chains of opposite polarity, fusion between chains being restricted to those of like polarity. J. G. Gall’s study of the kinetics of digestion of lampbrush chromosomes by pancreatic DNase also supports the view that each chromatid contains only two polynucleotide chains which are cleaved by this enzyme independently of one another; while O. L. Miller’s observations on the dimensions of the fibres remaining after lampbrush chromosomes have been digested by trypsin only allow for there being two polynucleotide chains per chromatid. By means of the technique of DNA fibre autoradiography devised by J. A. Huberman and A. D. Riggs, the units involved in replicating the chromosomal DNA of somatic cells ofXenopushave been compared with those ofTriturus. Both these organisms have initiation points for DNA replication that are arranged in tandem, and from each initiation point replication proceeds in opposite directions at divergent forks. The intervals between initiation points inXenopusrange from about 20 to 125µm apart, whereas those ofTriturusare much more widely separated. At 25 °C replication of DNA inXenopussomatic cells proceeds at 9µm per hour one-way at each fork, whereas the corresponding rate inTriturusis 20µm per hour.Triturussomatic cells take about 4 times longer than comparable cells ofXenopusto replicate their DNA. TheTriturusgenome contains about 10 times as much DNA as theXenopusgenome, and comparison of the replication process in these two organisms indirectly adds weight to the view that theTriturusgenome is 10 timeslongerthan that ofXenopus, rather than that it contains 10 times as many DNA double helices per chromatid. DNA fibre autoradiography has also been used to study replication inTriturusspermato-cytes. The round of DNA synthesis just before meiosis inTriturusis an exceptionally long-drawn-out process, taking 9 to 10 days for completion at 16 °C. This lengthy S-phase is not occasioned by abnormally slow replication, the rate being 12µm per hour one-way at 18 °C, nor is it the result of an exceptional staggering of replication starts. Instead it appears to be correlated with a gross reduction in the number of initiation points for replication. i.e. with an increase in the lengths of the replicating units. A rough calculation suggests that each meiotic chromomere may correspond to a unit of replication during the pre-meiotic S-phase.

A comparison of the number of nucleotides per unit length in Escherichia coli and phage T4 chromosomes

The grain density produced in radioautographs by fully labeled bacteriophage T(4) chromosomes was compared with the grain densities produced by fully labeled and half-labeled exponential phase Escherichia coli chromosomes. Taking into account the differing molar proportions of thymidine in E. coli and T(4) chromosomes, it was seen that the two types of chromosomes have approximately equal numbers of nucleotides per unit length, indicating that their structures are the same. Using molecular weight estimates in the literature for the T(4) and related T(2) chromosomes, and the lengths obtained in the radioautographs, the T(4) chromosome was estimated to have between 1.9 and 3.6 nucleotides/3.4 A. On the basis of these values alone, the E. coli and T(4) chromosomes could be either one Watson-Crick helix in a form about equal to or more condensed than the B form, or two helices more stretched out than the B form. The length of the T(4) chromosome was 48.7 +/- 4.1 mu when dried on the dull side of Millipore membranes and 42.3 +/- 4.8 mu on the shiny side, under the conditions used. Thus, the supporting surface apparently affects the configuration of a chromosome. Further evidence is also presented in support of the conclusion that the E. coli chromosome undergoes semi-conservative replication.

The strandedness of chromosomes: evidence from chromosomal aberrations

If cells are irradiated late in the mitotic cycle (late G2or early prophase), at the following anaphase they frequently exhibit characteristic chromosomal configurations known as sidearm bridges. These are often interpreted as sub-chromatid aberrations and are taken as evidence that chromosomes are multi-stranded. This interpretation, although recently challenged, is supported by experiments based upon the normal replication that converts chromatids to full chromosomes. The rationale is that aberrations involving only one chromatid of a chromosome are converted by replication to chromosome aberrations involving both chromatids. After replication, therefore, there should be no chromatid aberrations remaining unless the initial aberration involved less than a full chromatid. The results show that chromatid aberrations do appear after chromosomal replication: at the second mitosis after irradiation. Another experiment shows that most such chromatid aberrations are not the result of errors in the replication of previous chromatid aberrations.

MEIOSIS IN ORNITHOGALUM VIRENS (LILIACEAE): MEIOTIC TIMING AND SEGREGATION OF H3-THYMIDINE LABELED CHROMOSOMES

The timing of meiosis and the segregation of tritium-labeled chromosomes has been studied in Ornithogalum virens. Tritiated thymidine was administered to the inflorescences by a wick feeding process. The length of time required for labeled cells to pass from premeiotic DNA synthesis to the tetrad stage of meiosis is approximately 4 days. The labeled chromosomes display a semiconservative segregation pattern when labeled one division prior to meiosis; however, chromatid exchange events occur prior to, or during meiosis, or both. These exchange events are most likely a combination of crossing over between homologous chromosomes and sister chromatid exchange. The reductional separation of the centromere region at the first meiotic division is confirmed. The observations do not support Sax's two plane theory of crossing over, but are consistent with the theory that chiasmata are the cytological consequences of genetic exchange.

Chromosome structure--a model based on DNA replication

A model of chromosome structure is proposed which assumes: (i) that DNA replication is accomplished via right-hand (RH) rotation; and (ii), that where replicating DNA segments arc very long RH-rotation will not proceed with absolute freedom. It is expected that inhibition of rotation in daughter molecules will lead to the formation of left-hand-individual (LH-I) coiling systems in the two daughter molecules. It is also expected that inhibition of rotation in the parental molecule will cause the LH-I coiled daughters to be held together in a right-hand-relational (RH-R) association. The interaction between LH-I and RH-R coiling is expected to cause separation of the daughter molecules. Multistranded DNA-containing structures are expected to show, in addition to the LH-I coiling heirarchies formed within individual strands, an RH-R coiling heirarchy formed by the complex as a whole. An LH-I coiling heirarchy was looked for, and found, in Cleveland's drawings of flagellate chromosomes. Evidence for the existence of RH-R coiling was also found. Results of electron-microscope studies on chromosome structure were briefly examined, as were the structures of lampbrush chromosomes, salivary chromosomes and "normal" chromosomes. These studies provided additional, though less direct, evidence in favor of the replication hypothesis and the predictions developed from it.

The rationale for an ordered arrangement of chromatin in the interphase nucleus

Images

Nomarski interference contrast resolution of subchromatid structure

Images

Replicating units (replicons) of DNA in cultured mammalian cells

Exponentially growing L5178Y mouse leukemic cells were incubated in the presence of 5'-bromodeoxyuridine (BUdR) for about 4 hr, transferred to the nonBUdR-containing medium for a certain period (t hours), and then pulse-labeled with TdR-(3)H for 10 min. When DNA isolated from these cells was subjected to CsCl gradient centrifugation, the (3)H-activity was found to shift gradually from the heavy BUdR-containing peak to the light nonBUdR-containing peak with increasing time t. The average time required for the complete shift of (3)H-activity from the heavy to the light DNA fraction was 2.76 hr. Taking this as the average replicating time and the size of DNA fragments in the present preparation as 1.3 x 10(7) daltons, the rate of replication was found to be 2.1 nucleotides per strand per replicon per sec. By taking the upper limit of the average replicating time as the S period (7.3 hr), various characteristics of the replicating units, such as the lower and upper limits of average size, the average replicating time, the average number of replicating units, etc., were calculated (see Table I).

Distribution of chromatids at mitosis

The distribution of labeled chromatids at the second mitosis after labeling with tritiated thymidine is random in both Vicia faba (the broad bean) and Potorous tridactylis (the rat kangaroo). This finding is contrary to that predicted by the hypothesis that chromatids containing "grandparent" polynucleotide templates segregate from those containing "parent" templates.

Segregation of deoxyribonucleic acid in bacteria: association of the segregating unit with the cell envelope

Cells of the gram-positive organism Lactobacillus acidophilus R-26 were labeled with (3)H-thymine to measure the segregation of radioactive deoxyribonucleic acid (DNA) into daugher cells. Such cells were found to contain 8 conserved units of DNA which would correspond to two replicating chromosomes per cell. Fluorescent antibody (FA) against this organism was used to demonstrate that portions of the cell surface (2 to 4 units per cell) were conserved during growth and division. The permanent association of DNA with these conserved cell surface units was measured by combining autoradiography with FA techniques. DNA synthesized immediately before FA labeling was not associated with the fluorescent cell surface, whereas DNA synthesized a generation previously was. The results are consistent with a model in which DNA becomes permanently fixed to the cell surface when it is first used as a template.

The strandedness of meiotic chromosomes from Oncopeltus

Meiotic chromosomes were isolated from male Oncopeltus fasciatus by dissecting the testes under insect Ringer's solution and spreading the living cells on the Langmuir trough. After being dried by the critical point method, preparations were examined under the electron microscope. Chromosomes at all stages of prophase prove to be multistranded. A significant increase in the number of parallel 250 A fibers in the chromosomes occurs between zygotene and diakinesis. Parallel folding, rather than true multistrandedness, is interpreted as the mechanism responsible for this observed increase in multistrandedness. It has not been possible to determine whether the multistrandedness observed at leptotene represents true multistrandedness or is the result of parallel folding. Apparent multistrandedness is lost at metaphase when the 250 A fibers of the chromosomes become coiled more tightly. In preparations isolated by these methods, no structures other than the 250 A chromosome fibers are visible in the chromomeres, which appear as regionally coiled or folded areas of the fibers along the arm of the chromosome.

Double-strandedness of meiotic prophase chromatids to light microscope optics and its relationship to genetic recombination

Images

Strandedness of Vicia faba chromosomes as revealed by enzyme digestion studies

Chromosomes and nuclei isolated from neutral formalin-fixed Vicia faba lateral roots were treated with trypsin, pepsin, RNase, or DNase. Only trypsin affected the morphology of the chromosomes and nuclei. The appearance of the chromosomes after trypsin digestion indicated that each chromatid contained four strands that could be seen with an ordinary light microscope. The experiments are interpreted as indicating that mitotic chromosomes of Vicia faba are multistranded and that the linear continuity of the chromosome is dependent on protein.