All figures contain electrophoretic patterns (right) and the corresponding X-ray patterns obtained by exposure to the same agarose blocks upon drying (left).

A series of denaturing electrophoreses was performed to clarify whether the increase in linear sizes of extracellular DNA was related to joining of the fragments into a continuous chain. We assumed that if the slowing of labeled material migration were connected with formation of stable complexes, the denaturing conditions would detect the monomers of the complexes.

All the processes were considerably less intensive compared with the stage of cell culture logarithmic growth. The cytoplasmic fraction displays a distinct label accumulation. In the interchromosomal fraction, the linear size of the initial DNA fragment increases successively with the increase in the time of exposure to labeled DNA. However, this pattern becomes distinct only at the point of 120 min. The zero point contains no label; the point of 60 min displays the DNA fragment of initial size. Similar to previous experiments, α-dATP* as a monomer is absent in all the cell compartments.

The agarose blocks stained with ethidium bromide demonstrate that the nucleic acid material is present in the samples; however, it is nonuniform depending on the compartment wherefrom the sample originates.

The quantitative estimates made based on comparison of the counts at experimental points for various cell compartments are listed in below tables. The dynamics of accumulation of the labeled material in the cell compartments is presented on the Fig. 5 (Charts 1 and 2). Diagram 1 from Fig. 5 illustrates the dynamics of accumulation and the maximum amount of the extracellular DNA that is revealed in the free form in the nucleus (interchromosomal fraction). Note that the labeled material during all the DNA-purification procedures is present in the absolute amount; therefore, it is appropriate to compare and relate it to the label initially introduced into the medium.

When analyzing the changes in the fragmented genomic DNA after it entered various cell compartments, we discovered the fact of processing of the extracellular DNA delivered to the nucleus and deposited with the interchromosomal space. The linear DNA size during processing increased to about 10000 bp and was retained under denaturing conditions. We hypothesized that the increase in question may result from either ligation of various fragments or a certain variant of synthesis – involving fragments as primers for polymerization on a chromosome template or using the labeled material hydrolyzed to monomers. Note that in no experiment we observed the labeled material corresponding to the precursor monomer (α-dATP*). To test what possibility was actually realized, we prepared two DNA substrates and incubated MCF-7 cells with these DNAs for various times. One substrate was the plasmid Carnegie 20-λ1.4(× 8), comprising the vector Carnegie 20 and eight copies of D. melanogaster 1.4-kbp M/SAR SalGI fragment [18] [EMBL, GenBank, DDBJ: Z37541], hydrolyzed with SalGI restriction endonuclease. This hydrolysate was labeled with α-dATP* using Klenow fragment in the presence of the three rest cold triphosphates and completed with cold dATP to form blunt ends. The other substrate was represented by 506- and 230-bp PCR fragments of human caspase-3 gene [GenBank: NM_004346]. Both fragments were labeled directly during PCR. As is known, the thermophilic polymerase attaches one surplus nucleotide, predominantly dA, at the 3'-ends of the fragments synthesized. We tested the feasibility of ligation of both the first and second substrates in an in vitro system. Carnegie 20-λ1.4(× 8)SalGI ligated with the efficiency characteristic of the fragments with blunt ends. As for the 506- and 230-bp fragments, we succeeded in detecting ligation products only upon blunting the 3'-ends of these fragments with S1 nuclease (data not shown). These manipulations produced two DNA substrates – the substrate with blunt ends and the substrate with one-nucleotide tagged 3'-ends. Using these substrates, we planned to determine which of the processes occurred – ligation or synthesis. In the first case, we would observe a ladder of bands multiple to the lengths of the substrate fragments. In the second case, we should anticipate a smeared pool of DNA with increasing lengths of fragments. In addition, we tried to answer the question on whether the physicochemical structure of the fragments' ends was important. Two blocks of electrophoretic patterns, described below, resulted from these experiments. MCF-7 cell culture at the stage of logarithmic growth and native condition of DNA isolation and electrophoresis were used in both experiments. The cells were incubated with substrates without and in the presence of competitor DNA. Salmon sperm DNA at a concentration of 200 μg/ml was used as a competitor. About 3 μg of the labeled substrate was taken per one experimental point.

The band morphology changed in the medium immediately at the point of 30 min. The initial labeled material was detectable neither after 30 min nor at any other time point in any cell compartment. (Experiment V).

In the medium, the labeled fragments changed their mobility, forming smeared spots of the processed heavy and light fragments (Fig. 6). Both fragments (in an altered form) initially displayed a decreased mobility in the gel. The heavy fragment at the last point (120 min) migrated similarly to the initial fragment but was visible as a smeared spot. The short fragment degraded to a length of approximately 200 bp by 120 min of incubation.

The DNA processed similarly is found in the cytoplasmic fraction. We failed to detect any signs of discrete fragments in the interchromosomal fraction. Similarly to the experiments of the first section, the chromatin fraction contained a considerable amount of the labeled material. The overall pattern displayed by the labeled material remained the same as that obtained during the first set of experiments. An increase in the amount of labeled material in both compartments of the nucleus with the incubation time of cells with DNA substrate was evident.

Note that two individual fragments were used as a substrate. However, in no point of experiment we detected two initial labeled fragments. Presumably, all DNA fragments in the culture medium degraded during incubation.

To determine the effect of competitor DNA on physical parameters of the substrate (Experiment VI), we incubated two types of fragments – with blunt ends (Carnegie 20-λ1.4(× 8)SalGI) and with tag 3'-ends (506- and 230-bp) – with MCF-7 culture in the presence of 200 μg/ml salmon sperm DNA (Fig. 7). Since the initial labeled material in the previous experiment was undetectable already at the point of 30 min, we took the zero point (0 min) as the first experimental point in this experiment. Fig. 7 demonstrates the obtained distribution pattern of the labeled fragments and changes in their parameters. Similarly to the previous experiment, the labeled material in the absence of competitor DNA in the medium commenced degrading already at the zero point independently of the substrate. The size of completely degraded fragments amounted to about 200 bp.

As for the variant with 200 μg/ml of salmon sperm DNA in the medium, intermediate degradation phases were observed at the zero point. The light fragments (1.4 kb, 506, and 230 bp) remained intact. The heavy fragment (Carnegie 20-λ1.4(× 8)SalGI) degraded completely in the medium already at the zero point. With further increase in incubation time, all DNA in the medium degrade to fragments with a length of approximately 200 bp, which are then engulfed by the cells. Analysis of the cytoplasmic and, especially, interchromosomal fractions brought the following surprising phenomenon. Despite that the heavy fragment was undetectable in the medium already at the point of 0 min, it was present at the same time moment in the interchromosomal fraction. Together with heavy fragment, the rest labeled fragments – 1.4 kb, 506 bp, and 230 bp – were present in an intact state at the zero point in both the cytoplasm and interchromosomal space. Presumably, the presence of competitor DNA blocks the initial nuclease activities, thereby sparing some time for a certain amount of the native fragments to enter the cell before any contact with the nucleases in the medium; this took one-two minutes (the time required to add lysing buffer and start gradient centrifugation). At the point of 0 min, both native monomers and the fragment multiple to two 1.4-kbp monomers were observed for Carnegie 20-λ1.4(× 8)SalGI substrate in the presence of competitor DNA (Fig. 7, interchromosomal fraction). However, we did not find any fragments with lower mobility that were multiple to the fragments of 506 and 230 bp labeled with the same intensity. This suggests that ligation occurs inside the nucleus and that the structure of the fragment's ends is important for this process. In the case of longer incubation times, the already degraded fragments whose lengths are about 200 bp are delivered into the cell. Presumably, ligation of heterogeneous degrading fragments leads to the smeared tail of the labeled material. Staining of gels with ethidium bromide demonstrated that a considerable amount of competitor DNA was also delivered to the interchromosomal compartment of the nucleus in addition to the tracer fragment.

The main goal of this work was to attempt tracing the mere possibility and the changes that could occur in the extracellular exogenous DNA transported to various cell compartments. We assumed that the DNA fragmented to a size multiple to 1–10 nucleosome units and composed of the fragments representing the complete human genome would be the most "physiological" for its recognition by DNA delivery and utilization mechanisms. This DNA size corresponds to the size of the apoptotic DNA present in the blood plasma. As was mentioned repeatedly, many examples reporting delivery of DNA of various compositional complexities to the cell are described. We believed possible to trace the route of DNA and the changes it underwent when entering the cytoplasm and nucleus via a natural delivery mechanism used by the cell. Two sets of experiments were performed. In the first set, the genomic DNA fragmented to a length of about 500 bp and labeled with α³²P* by nick translation, was used. In the second set, individual labeled fragments were used for analysis of the events taking place during transportation of nucleic acids into the cell. Various phenomena were found indicating that the extracellular DNA that entered the cell became an active component of the processes occurring in the cell.

The first set of experiments gave an unanticipated result that, first, α-dATP* was conveyed to all the cell compartments not as a monomer, but within DNA fragments of about 200 bp long. Further, this DNA is involved in all the processes similar to the DNA added to the medium. In all respects, the activity of α-dATP* utilization is severalfold less intensive compared with the utilization of labeled extracellular DNA. Then it was found that extracellular DNA was virtually immediately (less than during 1 min if the cells were actively dividing) delivered to all the cell compartments analyzed (cytoplasm and nuclear space) and label appears in the chromatin. In the cytoplasm of actively dividing cells, approximately equal amount of labeled DNA was present at all the experimental points except for zero point (Fig. 5, Chart 1). The label accumulated in the nuclear compartments depending on the time of exposure to labeled DNA present in the culture liquid. DNA was present in the interchromosomal space of the nucleus as fragments of various lengths. The size of the fragments at the zero point in the actively dividing cells corresponds to the size of extracellular DNA added to the medium. In all the rest variants, the linear DNA size increased to approximately 10 kb. Moreover, the "ladder" of the bands detected (Fig. 1, left lower block), suggests that a part of the fragments were ligated with one another. In cells of the culture at the stage of contact inhibition, DNA processing (denaturing conditions) displayed a distinct dynamics (Fig. 3). Presumably, the metabolic processes at this stage of cell culture growth are slowed down, and we succeeded in detecting the dynamic stage of the processes of accumulating the labeled material in the cytoplasm (separation zone of gel; freely migrating DNA) and forming multimers of the initial DNA fragments in the nucleus. Increase in the linear size of the DNA molecules delivered to the nuclear space may result from the specific DNA ligase IV activity, involved in repair events [19].

In the second set of the experiments, it was found, first, that any DNA added to the culture medium of MCF-7 cells was degraded virtually immediately. Only the fraction of tracer DNA, which was protected from nuclease activities by a high concentration of competitor DNA and thereby succeeded in binding to surface cytoplasmic factors and entering cell compartments,

When using a high concentration of competitor DNA, a certain amount of initial fragments is virtually immediately delivered to cell compartments in an intact state. At zero point when using the fragments with blunt ends and excess of competitor DNA, we detected in the interchromosomal fraction emergence of a weak but distinct fragment that could be a dimer produced by ligation of the 1.4-kbp fragment (Fig. 7, left lower block; indicated by an arrow). No discrete bands with a low electrophoretic mobility were detected for the variant with tag 3'-ends.

Cytological examination of the behavior of extracellular DNA during its contact with cells demonstrated that the labeled material appeared in the cytoplasm within vacuoles already after several minutes. The label was virtually undetectable in the nuclear space. The entire label was localized to the nuclear compartments after a 14-h incubation. In the nucleus, the labeled material was located in a form of conglomerates and displayed an intensive fluorescence. It is known that pinocytosis is a general mechanism for engulfing any extracellular material. The corresponding hypothesis states that the exchange of macromolecules or their fragments between the cell and environment is bidirectional. This means the possibility of an "intravital" exchange of genetic information between individual cells of the organism via releasing the DNA fragments that appeared as a result of extracellular DNA turnover by reverse pinocytosis. This concept may entail another hypothesis assuming that a certain part of the DNA molecules present in the cell comprises the cell DNA carrying the information about metabolism of this cell itself, whereas the other part is the population of molecules that could have been conveyed from the ambient medium. The latter molecules may carry the information about the overall organism, the organism as a species (and as any other entity), and have little in common with the information about the overall metabolism of this particular cell [4]. A continuous sorting of such molecules may form the informational mechanism that transfer the data about the general state of the organism to individual cells.

The amounts of extracellular DNA delivered to cell compartments are estimated as tenths of a percent to percents of the DNA present in the culture medium. A simple estimation of the delivery dynamics of extracellular DNA to the cell space displays the following trends.

The label is quickly accumulated in the cytoplasm (Table 3; Fig. 5, Charts 1, 2). The maximal DNA amount is detected between 60 and 120 min, followed by a decrease by 180 min.

The increase in DNA amount in the interchromosomal space up to the point of 120 min followed either by a decrease. We estimated percentage of the DNA localized to the interchromosomal space with respect to the haploid genome. It appeared that in the case of the indicated DNA content in the culture medium, 2 000 000 to 66 000 000 bp (or 0.66–2% of haploid genome) is accumulated in the interchromosomal space (Fig. 5, Diagram 1). As it was mentioned above, the DNA is represented by 500-bp fragments, a part of which is joined in concatemers. This high number of free double-stranded ends is likely to activate drastically the repair and recombination machinery of the cell and may cause an active integration of extrachromosomal material into the chromatin.

The plot of labeled material accumulation in the chromatin is smoother. The labeled material is accumulated over 60 min followed by reaching the plateau that continues to the last experimental point of 180 min in all experiments. Presumably, a considerable amount of the labeled material detected in the chromatin is connected not with the integration, but with the re-utilization of labeled material. Nonetheless, our data on rescue of mutation in caspase-3 gene [20] may suggest that a certain part of the extrachromosomal fragments integrate into the genome in a legitimate manner via homologous recombination.

The data reported suggest that a certain amount of DNA under the specified experimental conditions (MCF-7 cell culture, DNA concentration, number of cells, and reaction media) is delivered to the nucleus, constantly presents in the interchromosomal space, and possibly integrates into the genome. In the interchromosomal space, DNA undergoes alterations that increase its linear size to about 10 kbp.

Analyzing the events that occur when the extracellular DNA appears in the nucleus, the following questions are of the greatest importance. Is the DNA flow into the cell constant and does it appear to be a part of the general cell mechanism of the extracellular DNA utilization? Does the DNA delivered to the nucleus induce recombination process? What is the intensity of this process and does this intensity depend on the amount of DNA delivered to the nucleus? What factors influence the ratio of legitimate to nonlegitimate recombination?

We assume that the extracellular DNA is an important component of the molecular processes running in the organism.

Human breast adenocarcinoma (MCF-7) cells were cultivated in the RPMI 1640 medium supplemented with 10 mM L-glutamine and 50 μg/ml streptomycin (Sigma, USA) at 37°C in the presence of 5% fetal bovine serum (FBS; Biolot, Russia) in an atmosphere of 5% CO₂ to a density of 0.7 × 10⁷ cells per 4 wells of a 24-well plate.

Human placental DNA (10 μg) fragmented to 500 bp was labeled by nick translation in the presence of Klenow fragment, three cold, and one hot dNTPs. The unincorporated precursors were removed by double precipitation with isopropanol according to the protocol described by Glover [21]. The yield of labeled DNA upon purification amounted to 7–15 μg. DNA was dissolved in an appropriate volume of distilled water and added to wells of the plate (no more than 20 μl per well). Each well contained 250–350 μl of the medium. A 1-μl aliquot was taken from the working volume to determine radioactivity.

The aliquot of α-dNTP* was diluted in the appropriate volume of water, and 1 μl of the diluted triphosphate was sampled to count radioactivity. An approximately similar amount of α relative to DNA (with respect to radioactive count and volume) was added per each point.

The amounts of DNA (μg and cpm) and α-dNTP* (cpm) for the three experiments are listed in Table 1.

For cytological examination, DNA was labeled with the modified precursor containing FITC fluorochrome.

Restriction fragments were labeled at the sticky end formed by hydrolysis with SalGI restriction endonuclease in the presence of Klenow fragment, three cold triphosphates, and hot dATP*. Upon attachment of cold dATP to make the ends blunt, DNA was purified from the unincorporated precursors by double precipitation with isopropanol from 0.3 M NaAc.

DNA of the PCR fragments of 506 and 230 bp was labeled during PCR using specific primers [20].

The labeled DNA treated as described above was added to the culture medium into the well where cells were cultivated. Cells were incubated with DNA at 37°C for the time required in an air thermostat. The following incubation times were chosen: 0 (labeled DNA was added to the medium in a titrator placed on ice; the medium was immediately sampled and supplemented with Triton X-100), 15, 30, 60, 120, and 180 min with variations in different experiments. The incubation time for α-dATP* was 180 min. The DNA amounts in the incubation medium per point are listed in Table 1.

After incubation, the plate was placed on ice. The medium was quantitatively sampled, and buffer A containing 2 mM CaCl₂ and 0.5% Triton X-100 [22], was immediately added to cells. Each well was supplemented with 250 μl of lysing buffer, being 1 ml per point. Cells were kept with lysing buffer on ice for 10 min. The lysed cells were resuspended several times and loaded onto 1 ml of sucrose gradient in buffer A with 2 mM CaCl₂ and centrifuged in conic tubes (25 ml) in a bucket rotor using a K23 centrifuge at 600 × g (2000 rpm) for 20 min. The supernatant, containing cytoplasmic fraction, was collected in a separate tube to precipitate with isopropanol from 0.3 M NaAc. The pellet of nuclei was washed twice with 250 μl of lysing buffer (buffer A containing 2 mM CaCl₂ and 0.5% Triton X-100). After each washing, the preparation was centrifuged for 10 min in a bucket rotor in a K23 centrifuge at 600 × g (2000 rpm). The nuclei were resuspended in 500 μl of buffer A containing 2 mM CaCl₂, examined cytologically, and transferred to the tubes of a JA-21 rotor of a J2–21 centrifuge. The suspension of nuclei was supplemented with 2 M NaCl and 1% SDS. Nuclear lysate was incubated at 65°C for 15 min until a complete clarification of the reaction mixture without any shaking. The nuclear lysate was then centrifuged at 21000 rpm (52000 × g) for 20 min at 35°C. The supernatant, being the nuclear sap or interchromosomal material, was collected preparatively with a yellow pipette tip and transferred to another tube. The supernatant was precipitated with 0.6 volume of isopropanol from 0.3 M NaAc. The lenticular precipitate, representing the chromosome fraction, was (1) supplemented with 100 μl of water and left to swell for 60 min before electrophoresis; (2) supplemented with 500 μl of water and left to swell for 10 min before precipitation with isopropanol; or (3) dissolved in the buffer containing 8 M urea and 8% glyoxal (denaturing conditions).

All the procedures were performed at 0°C. All the samples upon treatment were immediately precipitated with 0.6 volume of isopropanol from 0.3 M NaAc in 25-ml conic tubes to centrifuge in a bucket rotor in a K23 centrifuge at 4500 rpm for 20 min. The precipitate was dissolved in (1) 100 μl of water or (2) denaturing buffer at 65°C for 60 min. The amount of labeled material was determined routinely in Eppendorf tubes put into counting vials with a 1209 RackBeta counter (Finland). The samples were fractionated in 0.7% agarose gel. The DNA of chromosome fraction in the experiments under native conditions was not dissolved completely to reduce DNA degradation connected with dissolution but just left over 60 min for swelling. The "jellyfish" of undissolved, swollen precipitate was loaded into the well of agarose gel. During electrophoresis, chromosome DNA remained at the start. In the case of denaturing conditions, solution of nucleic acid material was loaded onto agarose gel without prior dialysis. Upon electrophoresis, the agarose gel was dried on a plate under a flow of hot air. X-ray film was exposed to the gel was overnight or for a period depending on the amount of labeled material.