Cell-surface-bound extracellular DNA (csbDNA) is present on the outer membrane of blood cells, including both red blood cells and leukocytes. Although less well characterized than cell-free DNA (cfDNA) in plasma and serum, leukocyte and red blood cell csbDNA form a considerable fraction of the blood extracellular nucleic acids pool, with typically at least comparable amount of DNA occurring bound to the outer surface of cells as compared with circulating free DNA in plasma. The cellular origin of csbDNA is not clear; however, as with cfDNA, in patients with cancer a proportion is derived from the tumor, thus making it potentially a useful source of DNA for cancer diagnosis, prognosis, and monitoring. © 2019 IUBMB Life, 71(9):1201–1210, 2019
- cell-free DNA
- cell-surface-bound DNA
Nucleic acids on the outer cell surface have been known for over four decades. Aggrawal was among the first researchers who, in 1975, directly observed cell-surface-associated nucleic acids when studying cell lines, primary cultures, and cells obtained directly from animals 1. He used DNA-binding, platinum–pyrimidine complexes for nucleic acid detection by electron microscopy, and discovered prominent electron-dense patches on the surface of tumorigenic cells. These patches appeared removable by neuraminidase and DNase I treatment, but not by RNase, suggesting that the particular substance labeled on the cell surface was DNA. In further studies using cells cultivated in the presence of tritiated thymidine, Aggrawal demonstrated that the bound DNA was of endogenous origin 2. Furthermore, a mild treatment of Chinese hamster ovary cells with trypsin or papain (causing no cell damage) had previously been shown to release nucleic acids from the cell surface 3, and treatment of mouse Sarcoma-180 cells with nucleic acid-hydrolyzing enzymes (DNase I or RNase A) decreased the cells’ electrophoretic mobility by 10% and 19%, respectively, reflecting a change in the surface charge 4. Somewhat later, in 1981, it was shown that DNA could also be associated with surface of blood cells 5, 6.
Despite these early findings, further studies in this area ceased for a long time. This was partly due to technical problems with the methods used; in particular, the control of mycoplasma contamination in the assayed cultures 7, as well as the unclear quantity, origin and role of the observed DNA complexes, which considerably diminished the scientific value of the early results.
An increase in the number of studies into DNA binding to the cell surface and natural nucleic acids transport into cells commenced in the 1980s–1990s with the advent of complementary addressed approaches 8, discovery of immunologically active DNA 9, 10, and DNA immunization 11. Thus, it was found that almost all cells were able to bind DNA (or oligonucleotides) and that DNA could bind to membranes both without involvement of membrane proteins 12 and with the help of receptors.
Another challenge for the studies of cell-surface-bound DNA was the pathogenesis of autoimmune diseases, which, as is known, are accompanied by production of the antibodies to DNA (or histones) and, correspondingly, circulation of DNA-containing immune complexes in the blood 13. The anti-DNA antibodies allowed for the discovery of DNA exposed on the surface of monocytes and B-lymphocytes 14. The fact that the DNA present in the blood is packaged in nucleosomes 15 dates back to the same period; it has been also shown that the immune complexes with DNA are able to form the antibodies to both DNA and histones 13, 16. Similar immune complexes and nucleosomes have been observed not only in a free form, but also bound to the cell surface 17.
However, with technical advances, numerous cell-surface proteins able to bind DNA were progressively discovered 18, and by the time it was found that extracellular DNA could also be generated by tumor cells and research into circulating DNAs became more clinically meaningful 19. The relevant literature already contained convincing data demonstrating the presence of DNA on the cell surface, including the surface of blood cells 14. DNA's characteristics of tumor cells are detectable among the csbDNA on the blood cell surface 20 demonstrating diagnostic potentialities of csbDNA.
Here we review the results of all studies of the general properties of сell-surface-bound DNA, mechanisms of DNA binding to cell surface, and analyzed the prospects of this DNA as a biomarker compared to cell-free DNA.
GENERAL PROPERTIES OF csbDNA AND THE FACTORS INFLUENCING ITS CONCENTRATION
csbDNA on blood cells differs in a number of aspects from cfDNA in plasma. Amount of csbDNA is at least comparable with amount of cfDNA in plasma. Even when cfDNA is underrepresented in the blood plasma of patients, csbDNA is eluted in amounts sufficient for the assay from the surface of blood cells 21. There are two fractions of csbDNA on red blood cells and leukocytes: a weakly associated fraction, the binding of which is dependent on divalent cations and which is easily removed by ethylendiaminetetraacetic acid (EDTA) treatment, and a tightly associated fraction, the binding of which is dependent on protein receptors and which requires trypsin to detach it from cells 22. In healthy subjects, over 95% of the DNA occurs in the tightly associated fraction 22; however, this decrease to around 80% in patients with prostate cancer 23. Gel electrophoresis of the weakly bound fraction shows that it is composed mainly of high-molecular-weight DNA (~20 kb). However, there is also a population of smaller fragments migrating somewhat below 500 bp 24. The tightly associated fraction is lacking in the low-molecular-weight fragments and contains only DNA of ~20 kb. As visualized by agarose gel electrophoresis, there is no detectable difference between the csbDNA on the surface of red blood cells and leukocytes 24. The absence of a ladder pattern in either fraction suggests that the csbDNA is not derived from apoptotic cells.
The length of circulating DNA (both cfDNA and csbDNA) extracted from blood specimens of healthy women and breast cancer subjects has also been assessed using capillary electrophoresis 25. The cfDNA comprised both low- and high-molecular-weight DNA fragments, while the DNA associated with the blood cell surface was mainly composed of high-molecular-weight variants (2 kbp and longer) 25. The increased length of csbDNA fragments is explainable by longer DNA fragments being more stably bound to the cell surface (for example, because of multiple binding points and the binding of extended DNA regions via additional proteins, such as hydrophobic histone H1) as well as an increased resistance of DNA molecules associated with the cell surface to nuclease activities.
The ends of csbDNA molecules show distinct types of breakage, which differ between cancer patients and healthy controls. The ends of csb and cfDNAs from healthy women and breast cancer patients were characterized by ligation to biotinylated double-strand oligonucleotide adapters with either blunt ends or random 1–3 base overhangs of both chains. DNA ligated to each type of adapter was isolated by binding to streptavidin beads and quantitated by PCR for LINE-1 repeats/α-satellite 25. The csbDNA of healthy women predominantly had blunt or 3′-protruding ends, with fewer DNA fragments with 5′-protruding ends. In contrast, in breast cancer patients, the vast majority of csbDNA fragments had 5′-protruding ends, and of these, most had a 2-nucleotide overhang, suggesting the presence of specific enzymes overrepresented in cancer. It is known that blunt-ended breaks bearing 5′-phosphates are present in apoptotic but not in early necrotic cells. Over-representation of 5′-overhangs versus blunt ends and 3′-overhangs has been shown to be typical of early necrosis 26 suggesting that, in cancer patients, a prominent part of csbDNA is released from necrotic cells. It is not clear whether these cells are true tumor cells, untransformed cells from tumors, or surrounding cells; however, the presence of tumor-specific DNA in this fraction (discussed below) confirms the contribution of tumor cells to the generation of csbDNA.
The occurrence of csbDNA is dependent on the release of genomic DNA into the extracellular space, the interaction of this DNA or its complexes with biopolymers of the extracellular medium and blood plasma, and specific features in the structure of blood cell surfaces. Since DNA circulates in the blood bound to nucleosomes 27 or within membrane enclosed complexes (apoptotic bodies and microvesicles) 28, 29, histones and histone-binding blood biopolymers also contribute to the DNA binding to the blood cell surface, as do interactions between the membranes of blood microvesicles and cells, cellular receptors for the proteins within the nucleoprotein complexes, and so forth. The number of DNA molecules that can bind to different cell types and the Kd have been measured using a plasmid DNA labeled with a fluorescent dye, YOYO-1, and amounted to 22,000–84,000 DNA molecules per cell and a Kd of 0.08–0.29, respectively 30. Taking into account the fact that the cell surface area of 1 mL blood is at least 0.76 m2 and that a multipoint interaction of long molecules or supramolecular structures provides tight binding even when each binding center has a low affinity, blood cells can transport a considerable amount of biopolymers and their complexes, including DNA.
The concentration of cfDNA to a certain degree depends on the total nuclease activity in the blood 21. Assessment of the concentration of cfDNA and nuclease activity in the blood of healthy donors and cancer patients demonstrated that a decrease in DNase activities is accompanied by an increase in the cfDNA (Spearman's coefficient −0.57, P < 0.01); thus, DNases are among the factors negatively regulating the cfDNA concentration 23. Presumably, DNase activity in blood is also one of the factors that determine csbDNA concentration. A study of a primary endothelial cell culture demonstrated that DNase I treatment removes approximately half of the extracellular DNA from the cell surface 31. However, another study, using human blood cells, demonstrated that DNases almost fail to hydrolyze the csbDNA within the protein complexes on the cell surface and do not significantly contribute to csbDNA fragmentation and dissociation from the cell 23.
Another factor potentially influencing the csbDNA concentration is blood protease activity. Indeed, it is known that a high protease activity in the serum of breast cancer patients weakly correlates with the presence of metastases and increased serum nucleosome content (r = 0.268, P = 0.008) 32. However, it is unclear whether they can influence hydrolysis of the surface proteins involved in binding cfDNA to the cell surface.
Finally, disturbance of the total metabolism (lipid metabolism included), which leads to fragility of erythrocyte membranes and is observed at late cancer stages 33, may also influence csbDNA association with cells because this interferes with membrane structure in general.
GENERAL MECHANISMS OF DNA BINDING TO CELL SURFACE
A number of studies have attempted to clarify the nature of the interactions between nucleic acids and the surface of various cells by studying the binding of both exogenous and endogenous DNAs in in vitro and in vivo experiments. The findings suggest several variants of DNA binding to the cell surface. It has been found that almost all cells are able to bind DNA (or oligonucleotides) and that DNA can bind to membranes both without involvement of membrane proteins 12 and with the help of membrane receptors. These receptors can be polyanion receptors 34, nonspecific receptors (scavenger receptors) 35, and specialized receptors predominantly binding nucleic acids 36. Numerous cell surface proteins able to bind DNA have been discovered and partially identified 18 and DNA binding to the cell membrane has been demonstrated to be a necessary stage in the DNA transport into cells 37.
Nonspecific DNA Binding to Membrane Surface
As is typical of polyanions, part of the csbDNA molecule is likely to interact with the cell membrane phospholipids in a nonspecific manner with the involvement of divalent metal cations, primarily Ca2+ and Mg2+ 12. Cell surface proteins containing lysine and arginine are also able to nonspecifically interact with nucleic acids and contribute to their binding to the cell surface. Indeed, P3X6 Ag8.653 myeloma cells and A9 fibroblasts have been shown to adsorb phage T7 DNA and pSKA2 plasmid in the presence of magnesium ions, with DNA binding to the cell surface dependent on Mg+2 concentration in the medium 12. The maximum quantity of DNA bound to the cell surface in the presence of 20 mM MgCl2, and amounted to 2,500 phage T7 DNA molecules per cell. The authors have observed that addition of 0.015 M NaCl to the cells after DNA sorption in the presence of MgCl2 causes release of the bulk of the adsorbed DNA, while 10%–15% of the DNA remains irreversibly bound to the cell surface 12. Fluorescence and X-ray scattering experiments have shown that Mg2+ and Ca2+ ions influence DNA binding to lipid membranes, while Cu2+ ions destabilize DNA double helix and are not directly involved in binding. The data demonstrated that the nature of the divalent cation influences DNA binding to membranes by modifying both the lipid membrane and DNA secondary structure 38. According to our own observations, 10 mM phosphate buffer (pH 7.5) containing 0.15 M NaCl and 5 mM EDTA is able to elute part of the extracellular nucleic acids nonspecifically associated with the cell membrane of blood cells 22.
DNA Binding Mediated by Cell Surface Proteins
Both protein 39, 40 and lipoprotein 41 components can mediate specific binding of the nucleoprotein complexes circulating in the blood to cells. As mentioned above, the treatment of cultured ovarian cells with trypsin or papain under conditions not interfering with their viability releases nucleic acids from the cell surface 3. The trypsinization of blood cells with preservation of plasma membrane integrity also results in release of extracellular nucleic acids 22. Indeed, abundant proteins involved in stable and selective DNA binding have been discovered, described, and partially identified on the surface of cells of various types 18.
Polyanion receptors are also able to mediate the binding of extracellular DNA to the cell surface. In particular, the Mac-1 protein, a heparin-binding integrin present on the surface of leukocytes, monocytes, macrophages, and natural killers, has been shown to bind phosphodiester oligonucleotides 34. Both Mac-1 subunits, of 95 (β2) and 195 (αM) kDa, bind DNA in the presence of Ca2+ or Mg2+ ions with dissociation constants of 8.8 × 10−6 and 1.7 × 10−5 M, respectively. Fibrinogen, a natural ligand for Mac-1, and monoclonal antibodies to Mac-1 inhibit the binding of this protein with DNA.
Among non-blood cells, DNA/RNA-binding proteins with molecular weights of 79 and 90 kDa and mean Kd = 2 × 10−7 M have been observed on the surface of mouse fibroblasts L929 in amounts of 1.2 × 105 molecules per cell 36. Ku70 and Ku80, the DNA-binding components of DNA-dependent protein kinase, bind DNA with a high affinity (Kd, 5–25 × 10−10 M) and have been found on the surface of endothelial cells, that is directly contacting blood 42.
Leukocyte Cell-Surface DNA Binding Proteins
A number of DNA binding proteins have been identified on white blood cells. Using monoclonal antibodies to DNA photo-adducts, three high-affinity DNA-binding proteins of 28, 59, and 79 kDa have been found on the outer surface of lymphocytes 43. In addition, DNA binding to monocytes, T- and B-cells, and neutrophils has been demonstrated using phage λ DNA 44. The binding of phage λ [3H]-DNA to blood leukocytes was saturable and could be inhibited by preliminary cell trypsinization as well as incubation with unlabeled DNA, but not with RNA, poly[d(A)·d(T)], or mononucleotides 44. The treatment of cells with pepsin, neuraminidase, phospholipase, or RNase had almost no effect on DNA binding. An increase in Ca2+, Mg2+, or SO4 concentration in the medium to 1 mM elevated the amount of cell-bound DNA. A 30 kDa protein was found to be involved in the DNA binding to the surface of human neutrophils, monocytes, and lymphocytes 44, with a dissociation constant of 10−9 M for all studied cell types. The amounts of bound DNA varied in the range of 0.81 × 103–2.6 × 103 molecules per cell depending on the cell type. The amount of the 30-kDa protein on the surface of lymphocytes was assessed as 5 × 105 molecules per cell using a monoclonal antibody to this protein. Using flow cytometry, a DNA receptor has been found in 67% of lymphocytes and 98% of monocytes. A study of the distribution of this receptor in a population of lymphocytes using a double labeling technique demonstrates its expression in 90% of B-cells and 50% of T-cells 45.
The RC3H2 gene, initially designated MNAB, or membrane-associated nucleic acid-binding protein encodes a protein with a molecular weight of about 130 kDa, capable of binding DNA on the cell surface 46. RC3H2 was discovered by screening a human monocyte cDNA library with vector λgt 11 and the serum of systemic lupus erythematosus patients, which has the ability to competitively inhibit DNA binding to the cell surface by interacting with DNA-binding proteins. The binding constant to DNA was approximately 4 × 10−9 M, while a point mutation in the conserved region of the zinc finger decreased this value by 50%. In addition to being present in the perinuclear membranes and endoplasmic reticulum, RC3H2 protein also localized to the cell surface by flow cytometry 46.
Platelet and Red Blood Cell-Surface DNA Binding Proteins
Despite encouraging results about leukocyte cell-surface DNA binding proteins, DNA-binding proteins on the surface of platelets and red blood cells have not yet been identified. However, platelets, which are 10–15-fold smaller in size than erythrocytes, can also bind DNA 4. The maximum amount of bound DNA is 0.5–2.75 ng of bacterial single-strand DNA per million platelets, preliminarily washed to remove DNA. The DNA bound to the cell surface is not degraded by cell treatment with DNase I and prevents cell aggregation after platelet activation with aspirin 5.
A DNA-binding membrane protein with a molecular weight of 42 kDa has been detected on the surface of red blood cells, which represent about 40% of the total blood volume 6. A functional deficiency of the receptor binding ability is characteristic of systemic lupus erythematosus patients as compared with healthy subjects and, presumably, accounts for an increase in extracellular DNA in this pathology.
DNA Binding Via Receptors to Nucleic Acid Associated Proteins
Along with the receptors involved in nucleic acid binding, receptors for the proteins of nucleoprotein complexes also participate in the binding of these complexes to the cell surface. Endogenous DNAs circulating in the blood are as a rule packaged in nucleosomes; however, such circulating complexes can also contain amyloid P 47 and several blood serum proteins displaying rather high affinity for DNA, in particular, lysozyme 48, lactoferrin 49, 50, C1-reactive protein 51, immunoglobulins 52, and albumin 53, allowing for interaction via a broad range of receptors.
Receptors for nucleic acid–binding proteins are frequently expressed on the surface of blood cells. In particular, the receptors for immunoglobulin Fc fragments of all five classes of antibodies have been described 54 and are present on all types of immune cells (T and B lymphocytes, monocytes, granulocytes, etc. 55). A specific receptor (57–60 kDa) for human serum albumin enhancing DNA binding has been discovered on endothelial cells (BPAEC culture) 56. In addition, membrane immunoglobulins and immunoglobulin-like receptors can bind DNA on the surface of lymphocytes 57.
Another mechanism for DNA binding to cells may involve nucleosomal DNA binding to the cell surface or binding through histones exposed on the cell surface. DNA is present on the surface of Raji cells, a cell line of hematopoietic origin, as is shown by the presence of anti-DNA antibodies in the serum of systemic lupus erythematosus patients; treatment of such antibodies with DNase I or addition of histones significantly decreases the efficiency of their binding to the cell surface. In addition, anti-histone antibodies form complexes with cell surface proteins, thus demonstrating the presence of histones on the cell surface. These findings suggest that the anti-DNA antibodies bind to the cell surface through interaction with histones exposed to the cell surface in complexes with DNA 58.
Two receptor proteins (29 and 69 kDa) binding both DNA and nucleosomes with equal efficiencies have been detected on the surface of human peripheral blood mononuclear and murine T cells (S49). It was shown that addition of nucleotides inhibits the binding of biotin-labeled DNA to the cell surface, and addition of histones increases this binding. That the discovered proteins are nucleosome-/DNA-binding components of a more complex molecule rather than independent proteins could not be concluded; histones in this case may act as mediators in DNA binding to cell receptors 59.
The presence of a nucleosome receptor on the surface of CV-1 fibroblasts has been demonstrated 60. Using 125I-labeled mononucleosomes, the cell surface was shown to have two binding sites (approximately 9 × 107 sites per cell): a high-affinity binding site with a Kd of 7 nM and a low-affinity binding site with a Kd of 400 nM. As was demonstrated, a 50-kDa protein is responsible for nucleosome binding to the surface of CV-1 fibroblasts. Moreover, the nucleosomes associated with the cell surface were able to react with antibodies to histones and DNA.
In addition to binding to the cell membrane, csbDNA can bind to the membranes of vesicles circulating in the blood, in particular, exosomes. As has been shown earlier, the exosomes circulating in the blood of pancreatic cancer patients contain tumor DNA 61. However, the contents of exosomes reflect the cytosolic composition of donor cells and it is difficult to imagine the mechanism by which genomic/mitochondrial DNA is transported into these vesicles. Similarities between plasma and exosomal membranes suggest that exosomes also carry some DNA-binding proteins on theirs surface. Indeed, DNase I treatment of the exosomes from tears (up to 100 nm in size) detected double-strand genomic DNA of 3.9 kb on the outer surface of the vesicular membrane 62. The mechanism by which genomic DNA is trafficked to the surface of exosomes remains unclear. DNA may bind to the surface of exosomes either outside (directly in biological fluids) or inside the cells.
Q-PCR of LINE-1 repeat 63 was used to estimate the concentration of exosomal DNA in blood in 10 healthy women and 10 women with breast cancer. DNA was present in the plasma exosomes of 57% of healthy women at an average concentration of 41 ± 6 pg/mL blood and of 71% of the breast cancer subjects at an average concentration of 64 ± 1.3 pg/mL. As is shown, the DNA concentration within the exosomes associated with the blood cell surface is higher as compared with the plasma exosomes in a statistically significant manner (Mann–Whitney test, P < 0.05) in both healthy and breast cancer subjects. Taking into account the cfDNA concentration in the blood, the share of the DNA on the exosome surface does not exceed 0.1%–0.5% of the total DNA circulating in the blood. Correspondingly, exosomal DNA lacks any prospects in the search for tumor markers because of its very low concentration 64.
ISOLATION OF csbDNA
As described above, considerable amounts of DNA can circulate bound to the surface of blood cells; moreover, the DNA binding can be through both weak interactions with the cell surface, including binding via divalent cation bridges 12, and through “tight binding” via, for example, DNA or nucleosomes specifically binding to protein receptors exposed on the plasma membrane 18. It is possible to isolate the weakly bound portion of csbDNA by treating blood cells (after plasma removal) with excess PBS containing EDTA (nine volumes of 10 mM phosphate buffer, 0.15 M NaCl, and 5 mM EDTA; pH 7.5). After this treatment of blood cells, the DNA concentration in PBS–EDTA eluate varies from 57 to 298 ng/mL blood 22. It is evident that the blood cell fraction after plasma removal still contains some amount of plasma. According to estimates, this amount is approximately 100 μL per 4–5 mL blood cells. Taking into account that the average cfDNA concentration in the blood plasma is ~20 ng/mL 19, the content of plasma DNA in csbDNA fraction is no more than 0.2 ng/mL blood; that is, the DNA associated with the cell surface via weak interactions and present in the PBS–EDTA eluate accounts for over 90% of the measured concentration.
Cell washing with physiological saline allows at least 90% of the blood plasma proteins to be removed and provides the conditions for efficient trypsin hydrolysis of the cell surface proteins that bind DNA and release of the “tight binding” component of csbDNA. For example, “tight binding” component of csbDNA in blood healthy donors and prostate cancer patients was found to contain 4 ± 1% and 17 ± 3% of the total amount of circulating DNAs bound to blood cells, respectively. Since the integrity of the cell membrane can be damaged at this stage, potentially releasing intracellular DNA, it is necessary to strictly limit the enzyme concentration (0.125%, 1800 BAEE units/mg), the enzyme-to-cell ratio, and the treatment time which must not exceed 5 min. Addition of a trypsin inhibitor is also a necessary step, required for a rapid enzyme inactivation and prevention of cell lysis. The obtained eluate specimens meet the criteria for the absence of hemolysis/blood cell lysis: hemoglobin content is equivalent to an absorbance of <0.175 at 414 nm and lactate dehydrogenase activity <0.179 U/mL initial blood (cut off, 0.179 U/mL) 65.
Isolation of the csbDNA fractions needs to be carried out at, or soon after, blood collection (within 2 h). This creates challenges for biobanking, as the csbDNA isolation protocol is relatively laborious compared to simply separating and freezing blood plasma or serum. Hence, while it is relatively common for academic and commercial biobanks to store frozen plasma, serum, and cells, and to make these accessible to researchers either collaboratively or for a fee, csbDNA is not routinely biobanked. Lack of appropriate biobanked specimens in turn limits clinical studies to those that are able to prospectively undertake blood collection specifically for the purpose of the study.
CONCENTRATION AND DISTRIBUTION OF csbDNA IN THE BLOOD OF HEALTHY AND ILL SUBJECTS
Since csbDNA was found in human blood, attempts have been made to correlate its concentration with the appearance and progression of different pathologies. The extracellular DNA concentration in the plasma and in blood cell surface eluates of healthy subjects and patients with diagnosed conditions was estimated by fluorescent assay or quantitative PCR for LINE-1 repeat after DNA isolation using glass-milk or microcolumn based protocols 22, 63. Since csbDNA assays show potential for improving clinical diagnosis, the extracellular DNA concentration in blood was measured in various pathologies, including virus infections, autoimmune diseases, and tumors (Table 1).
|Disease||Mean concentration of DNA, ng/mL (range)||References|
|Hepatitis B men||12 ± 8||345||11||631||66|
|n = 21||n = 21||n = 14||n = 14|
|Polyarticular rheumatoid arthritis women||15 ± 13||430||42 ± 37*||165 ± 116*||67|
|n = 20||n = 20||n = 10||n = 10|
|Gastric cancer men and women||15 ± 13||508||150||669||68|
|n = 22||n = 22||n = 20||n = 20|
|Colon cancer men and women||13 ± 13||283||115||504||69|
|n = 26||n = 26||n = 10||n = 10|
|Breast nonmalignant tumor women||15 ± 13||430||96||326||20|
|n = 20||n = 20||n = 15||n = 15|
|Breast cancer women||15 ± 13||430||285||52||20|
|n = 20||n = 20||n = 20||n = 20|
|Lung cancer men||17 ± 4||254 ± 21||18 ± 6||89 ± 11*||70|
|n = 23||n = 23||n = 42||n = 42|
|Lung cancer men||16 ± 7||1,021 ± 377||13 ± 7||86 ± 36*||71|
|n = 15||n = 15||n = 42||n = 42|
- All concentrations are recalculated to initial blood volume.
- *DNA concentration differs between healthy and ill subjects (P < 0.05).
- **The csbDNA concentration was below the detection limit for fluorescence quantitation in 85% samples.
As shown in Table 1, the concentrations of cfDNA and csbDNA are similar between healthy individuals and those with chronic viral diseases 66; however, redistribution between these fractions does take place in autoimmune pathologies, such as polyarticular rheumatoid arthritis 67: the level of cfDNA increases, while the level of csbDNA decreases. However, the authors did not reveal a correlation between the concentrations of circulating cfDNA in the plasma and csbDNA on the surface of blood cells in the polyarticular rheumatoid arthritis subjects with systemic manifestation has been found 67. These data suggest that the concentration of extracellular DNA circulating in the blood (both cfDNA and csbDNA on the blood cell surface) may be used as a marker for risk of polyarticular rheumatoid arthritis.
Levels of cfDNA and csbDNA differ between individuals with cancer and healthy controls. As shown in Table 1, comparative studies have shown that cfDNA concentration in plasma is significantly elevated in gastric 68, colon 69, and breast 20 cancers; however, lung cancer patients display no difference from the healthy controls 70, 71. Conversely, a substantial lower level in comparison to healthy controls in csbDNA concentration has been observed in breast, prostate, and lung cancer but not in gastric and colon cancers and nonmalignant breast and prostate tumor patients. Summing up, the ratio of cfDNA to csbDNA in the blood depends on the state of the individual's health: a significant part of circulating DNA in healthy individuals is bound to cells, whereas cancer patients display different distributions of circulating DNA depending on the tumor type and location. This pattern is retained when using both fluorescence dyes and PCR assay for assessment of circulating DNA concentration in the blood; however, the latter significantly underestimates csbDNA concentrations. This may stem from DNA specimens possibly containing both PCR inhibitors and various substances boosting the fluorescence of intercalating dyes (e.g., lipopolysaccharides). In addition, inhomogeneous fragmentation of DNA molecules outside the cells and in the bloodstream, and consequent uneven representation of circulating DNA in the blood may be a factor 72.
SOLiD sequencing technology allowed revealing DNA repetitive sequence representation differences between apoptotic cfDNA and HUVEC genome DNA 73. Alu repeats were found to be overrepresented in the apoptotic cfDNA, whereas LINE-1 repeats are underrepresented in apoptotic cfDNA. It should be noted that LINE-1 elements are mainly located in the transcriptionally inactive, condensed constitutive heterochromatin and Alu repeats are mainly localized in the euchromatin regions that are enriched with coding (active) genes characterized by transcription activity. Our data on a LINE-1 and Alu repeats in the cfDNA are consistent with the results of several whole-genome studies demonstrating different representation of various classes of dispersed repeats within the cfDNA in the blood plasma. In particular, long LINE-1 repeats are underrepresented in the DNA circulating in the blood plasma/serum as compared with the genomic DNA versus short Alu repeats, which are overrepresented 73-76. It is necessary to emphasize that the concentration of circulating DNA fragments, being decreased for LINE-1 and elevated for Alu repeats, is significantly more pronounced in the circulating DNA of tumor subjects as compared with healthy donors 73, 77. In addition, a significant decrease in the concentration of region 2 LINE-1 fragments in the circulating DNA of non-small-cell lung cancer patients, as compared with healthy donors, may result from chromatin repression in LINE-1 regions with a decrease in DNA methylation, which is explainable by activation of the cell defense mechanisms 78.
The factors causing a decrease in the csbDNA concentration in the blood of cancer patients are unclear. Presumably, this phenomenon is associated with changes in the sequences representation of this DNA in comparison to cfDNA and genomic DNA or with changes in the membrane structure of blood cells. Indeed, it has been shown that cancer patients at an advanced disease stage display decreased membrane fluidity in their lymphocytes and erythrocytes as well as altered lipid composition and protein integration. This interferes with the cation transport systems and disrupts the cell surface architecture 33, potentially reducing DNA binding capacity.
csbDNA AS A TUMOR MARKER
PCR assays can detect cancer-specific sequences characteristic of tumor cells in blood plasma even when a tumor in the body comprises 100–1,000 cells 79. In addition to mutations, tumor DNA also harbors methylation changes. Analysis of the epigenetic status of DNA fragments circulating in the blood is attractive for diagnosis; however, the low concentration of plasma DNA molecules, the relatively low content of tumor-specific DNA as compared with the DNA from healthy cells, and the small size of DNA fragments make it difficult to use this method 80. Since most of the extracellular DNA circulating in the blood is associated with the blood cell surface, and since the csbDNA fragments are larger than the cfDNA fragments circulating in plasma 24, 25, we have focused on assessing the diagnostic value of csbDNA in different cancer types. We have shown that assaying markers in csbDNA elevates the detection rate of tumor DNA and as well as the sensitivity and specificity in discrimination between healthy donors and cancer patients. In particular, the methylation rate in the both promoter regions of RASSF1A and RARβ2 genes in stage I–III breast cancer subjects is considerably higher in the csbDNA on blood cells (95%) as compared with the cfDNA free in the plasma (30%). Analysis of the total circulating DNA in the blood (cfDNA circulating in the plasma and csbDNA on the blood cells) detected methylation of at least one of these genes in 95% of the patients 20. Analysis of the methylation status of the promoter regions in MGMT, p15, and hMLH1 genes within the csbDNA also improved the sensitivity of detecting stage II–IV gastric cancer from 13% in cfDNA to 80% in a total csbDNA pool 68.
In order to further improve the sensitivity of the epigenetic markers, quantitative methylation-specific PCR was used; this demonstrated a more than sevenfold increase in RARb2 gene methylation level in csbDNA and only a threefold increase in RARb2 gene methylation level in plasma cfDNA in lung cancer patients as compared with healthy donors. These data suggest that methylated gene fragments are more efficiently accumulated in the csbDNA fraction as compared with plasma cfDNA 81. Note that the level of unmethylated RARb2 gene showed no significant increase in the csbDNA of lung cancer subjects versus a decrease observed in the plasma cfDNA. The observed differences in methylation of the DNA fragments bound to the cell surface in cancer patients may result from the cancer-related changes in the structure of cell membranes, the content of circulating DNA-carrying complexes, and the composition of mediators for cell membrane binding. These findings suggest that analysis of the methylation profile of the csbDNA on blood cells and cfDNA in plasma can significantly elevate the accuracy of early diagnosis for cancer diseases as well as be useful in assessment of anticancer therapies.
BIOLOGICAL EFFECTS OF csbDNA
Chen 83 has formulated several hypotheses on the biological effects of cfDNA. They postulate that blood DNA is able to transfect healthy cells causing metastases (Fig. 1A). This hypothesis expands the concept of genometastasis, a model that implies that the DNA within apoptotic bodies is the source of cancer transformation. Moreover, since DNA-binding receptors are present on the cell surface 18, the authors suggest tissue-specific targeting in the development of metastases. In order to transform cells, DNA is required to reach the cell nucleus, association with the plasma membrane being a key point in the DNA transport 37.
In addition, it is proposed that cfDNA derived from normal cells (for example, the DNA released into the bloodstream by lymphocytes as a result of antigenic stimulation) can transfect tumor cells. In particular, integration of the cfDNA carrying cytokine-coding regions into the genome of a tumor cell can lead to elevated expression of various cytokines, such as interleukin 2, interleukin 12, and macrophage colony-stimulating factor; a cfDNA fragment carrying a non-mutated oncogene (for example, ras) or a non-mutated cancer suppressor gene (for example, wild-type p53) is able to knock out the corresponding mutant oncogene or cancer suppressor gene via homologous recombination within a cancer cell and, as a result, cause the apoptosis of tumor cell or even spontaneous cancer remission 83.
The genometastasis theory has its advocates and opponents, both proposing valid arguments 84. Chen 83 explains the presence of extracellular DNA on the surface of blood cells in healthy donors by the ability of circulating DNA to bind to receptors on the surface of leukocytes 85, and in cancer patients, the decrease in csbDNA during progression of a tumor 24, by the absence of the corresponding DNA receptor on the surface of blood cells in cancer patients or a loss of the DNA-binding properties because of mutation. In this model, cfDNA in the bloodstream is a signaling molecule and its binding to a specific receptor on the lymphocyte surface can activate cells and induce the antitumor immune response (Fig. 1B). Thus, a mutation of a DNA receptor on the lymphocyte surface can entail the tolerance of anticancer immune response.
The authors believe that a comprehensive study of the DNA binding to cells, namely, the search for the sequences most stably bound to cell surface proteins and identification of these proteins, will make it possible to gain the new data on the biological role of extracellular nucleic acids.
Since Yamamoto 10 and Krieg 11 some two decades ago described immunomodulatory CpG-rich DNAs and Veiko 86 discovered such endogenous CpG-rich DNAs, we have attempted to find the immunomodulatory DNAs on the surface of blood cells. However, the attempts have been unsuccessful. Nonetheless, we have observed when studying the effect of DNA on the inhibition of immune response induction by double-strand RNA that csbDNA is a more efficient inhibitor for the proinflammatory cytokine production by fibroblasts and endothelial cells as compared with the genomic or circulating DNAs 87. Although we have detected immunomodulatory motifs in the csbDNA pool, the mechanism underlying the effect of csbDNA on the double-strand RNA–induced immune response is still rather vague 88.
Extracellular DNA in the blood associated with the surface of red blood cells and leukocytes is at least as represented as circulating DNA in plasma. Stable nucleic acid binding to the cell surface can take place in a variety of ways, including via a charge-based interaction, binding of cell surface exposed proteins with the DNA target, binding of mediator molecules with nucleic acids and the plasma membrane, or a combination of these variants. csbDNA binding to cell membrane proteins can influence nucleic acid transport into cells and signal transduction from extracellular receptors. csbDNA is less fragmented than the DNA circulating in blood plasma and contains tumor cell DNA. The use of csbDNA in combination with cfDNA makes it possible to increase the sensitivity of cancer detection markers for various cancer types as compared with the cfDNA alone. This fact is the most apparent advantage of csbDNA as a source of diagnostic material. However, the laborious protocol for csbDNA extraction, requiring the control of membrane integrity, and a lack of biobanked samples from clinically characterized patient cohorts limit its active study and introduction to clinical practice.
We would like to thank Kristina Warton for critical reading of the manuscript. The work was supported by Russian State funded budget project of ICBFM SB RAS # AAAA-A17-117020210024-8.
- 1, , , and (1975) Cell-surface-associated nucleic acid in tumorigenic cells made visible with platinum-pyrimidine complexes by electron microscopy. Proc. Natl. Acad. Sci. USA 72, 928–932.
- 2 (1977) Further evidence in support of cell-surface-associated deoxyribonucleic acid in tumor cells: an autoradiographic study. J. Histochem. Cytochem. 25, 359–370.
- 3, and (1974) An "external" RNA removable from mammalian cells by mild proteolysis. Proc. Natl. Acad. Sci. USA 71, 4960–4964.
- 4, and (1982) Actions of cis-diamminedichloroplatinum on cell surface nucleic acids in cancer cells as determined by cell electrophoresis techniques. Cancer Res. 42, 3565–3573.
- 5 (1981) Binding of single-strand DNA to human platelets. Thromb. Res. 24, 119–129.
- 6 (1988) 42 kD erythrocyte surface membrane protein with binding capacity to polynucleotides shows functional lack in systemic lupus erythematosus. Immunobiology 178, 141–142.
- 7, , , , and (2009) Extracellular DNA in culture of primary and transformed cells, infected and not infected with mycoplasma. Bull. Exp. Biol. Med. 147, 63–65.
- 8, , , , , et al. (1974) Photoaffinity reagents for modification of aminoacyl-tRNA synthetases. FEBS Lett. 49, 159–162.
- 9 (2012) CpG still rocks! Update on an accidental drug. Nucleic Acid Ther. 22, 77–89.
- 10, , and (2000) Oligodeoxyribonucleotides with 5'-ACGT-3′ or 5'-TCGA-3′ sequence induce production of interferons. Curr. Top. Microbiol. Immunol. 247, 23–39.
- 11, , , , , et al. (1995) CpG motifs in bacterial DNA trigger direct B-cell activation. Nature 374, 546–549.
- 12, , , and (1988) Mg2+-dependent interaction of DNA with eukaryotic cells. Mol. Biol. (Mosk). (in Russian) 22, 1667–1672.
- 13 (2016) Anti-DNA antibodies--quintessential biomarkers of SLE. Nat. Rev. Rheumatol. 12, 102–110.
- 14, , and (1986) Anti-DNA antibodies react with DNA expressed on the surface of monocytes and B lymphocytes. J. Rheumatol. 13, 679–685.
- 15, , and (1999) Plasma nucleosome levels in node-negative breast cancer patients. Breast Cancer 64, 361–364.
- 16, , and (1998) Isolation of anti-nucleosome antibodies from the plasma of lupus nephritis patients. Clin. Rheumatol. 17, 189–194.
- 17, , and (2004) Nucleosomes are exposed at the cell surface in apoptosis. J. Immunol. 172, 6692–6700.
- 18, , and (2006) Proteins involved in binding and cellular uptake of nucleic acids. Biochemistry (Moscow) 71, 583–596.
- 19, and (2007) Circulating nucleic acids (CNAs) and cancer-a survey. Biochim. Biophys. Acta 1775, 181–232.
- 20, , , , , et al. (2006) Cell-free and cell-bound circulating DNA in breast tumors: DNA quantification and analysis of tumor-related gene methylation. Br. J. Cancer 94, 1492–1495.
- 21, , and (2008) Circulating DNA in the blood and its application in medical diagnosis. Mol. Biol. 42, 9–19.
- 22, , , , , et al. (2005) Circulating nucleic acids in blood of healthy male and female donors. Clin. Chem. 51, 1317–1319.
- 23, , , , , et al. (2015) Redistribution of free- and cell-surface-bound DNA in blood of benign and malignant prostate tumor patients. Acta Nat. 7, 115–118.
- 24, , , , , et al. (2004) Cell-surface–bound nucleic acids: free and cell-surface–bound nucleic acids in blood of healthy donors and breast cancer patients. Ann. N. Y. Acad. Sci. 1022, 221–228.
- 25, , , , and (2016) Features of circulating DNA fragmentation in blood of healthy females and breast cancer patients. Adv. Exp. Med. Biol. 924, 47–51.
- 26, , and (2003) Early necrotic DNA degradation: presence of blunt-ended DNA breaks, 3′ and 5′ overhangs in apoptosis, but only 5′ overhangs in early necrosis. Am. J. Pathol. 162, 1571–1578.
- 27, , , , , et al. (2001) DNA fragments in the blood plasma of cancer patients: quantitations and evidence for their origin from apoptotic and necrotic cells. Cancer Res. 61, 1659–1665.
- 28, , and (2000) Segregation of RNA and separate packaging of DNA and RNA in apoptotic bodies during apoptosis. Exp. Cell Res. 260, 248–256.
- 29 (2010) Microparticles in cancer. Semin. Thromb. Hemost. 36, 888–906.
- 30, , , , , et al. (2000) Quantification of DNA binding to cell-surfaces by flow cytometry. J. Immunol. Methods 241, 141–146.
- 31, , , , and (1991) A role for DNA in anti-DNA antibodies binding to endothelial cells. J. Autoimmun. 4, 463–478.
- 32, , , , , et al. (2011) Apoptosis-related degradation of proteolytic activities and high serum levels of circulating nucleosomes and DNA in blood correlate with breast cancer progression. BMC Cancer 11, 4.
- 33, , , , , et al. (2006) Molecular disturbances in erythrocyte membrane at different pathologies. Bull. Sib. Med. (in Russian) 2, 62–69.
- 34, , , , , et al. (1997) Mac-1 (CD11b/CD18) is an oligodeoxynucleotide-binding protein. Nat. Med. 3, 414–420.
- 35, , , , , et al. (1994) Binding of oligoguanylate to scavenger receptors is required for oligonucleotides to augment NK cell activity and induce IFN. J. Biochem. 116, 991–994.
- 36, , , , , et al. (1989) Mechanism of oligonucleotide uptake by cells: involvement of specific receptors? Proc. Natl. Acad. Sci. USA 86, 6454–6458.
- 37, , and (1996) Identification of two oligodeoxyribonucleotide binding proteins on plasma membranes of human cell lines. Biochem. Pharmacol. 51, 431–436.
- 38, , , and (2011) DNA condensation and its thermal stability influenced by phospholipid bilayer and divalent cations. Colloids Surf. B Biointerfaces 86, 212–217.
- 39, , and (1985) DNA binding to human leukocytes. Evidence for a receptor-mediated association, internalization, and degradation of DNA. J. Clin. Invest. 76, 2182–2190.
- 40, , and (1987) DNA receptor dysfunction in systemic lupus erythematosus and kindred disorders. Induction by anti-DNA antibodies, antihistone antibodies, and antireceptor antibodies. J. Exp. Med. 166, 850–863.
- 41, , , , , et al. (1991) Uptake of oligodeoxyribonucleotides by lymphoid cells is heterogeneous and inducible. Antisense Res. Dev. 1, 161–171.
- 42, , , , , et al. (2012) Ku protein as the main cellular target of cell-surface-bound circulating DNA. Expert Opin. Biol. Ther. 12, S35–S41.
- 43, , , , and (1990) Cell membrane DNA: a new target for psoralen photoadduct formation. Photochem. Photobiol. 52, 315–321.
- 44, and (1984) Biotin-labelled DNA: a novel approach for the recognition of a DNA binding site on cell membranes. Biochem. Biophys. Res. Commun. 122, 1034–1039.
- 45, , , , , et al. (1988) The production and characterization of murine monoclonal antibodies to a DNA receptor on human leukocyte. J. Immunol. 140, 2937–2942.
- 46, , , , , et al. (2000) A human gene coding for a membrane-associated nucleic acid-binding protein. J. Biol. Chem. 275, 33655–33662.
- 47, , and (1990) Pentraxinchromatin interactions: serum amyloid P component specifically displaces H1-type histones and solubilizes native long chromatin. J. Exp. Med. 172, 13–18.
- 48, , and (1992) Heparin-binding properties of lactoferrin and lysozyme. Comp. Biochem. Physiol. B 103, 889–895.
- 49, , , , , et al. (1997) N-terminal stretch Arg2, Arg3, Arg4 and Arg5 of human lactoferrin is essential for binding to heparin, bacterial lipopolysaccharide, human lysozyme and DNA. Biochem. J. 328, 145–151.
- 50, , , and (1999) Human milk lactoferrin binds two DNA molecules with different affinities. FEBS Lett. 451, 235–237.
- 51, and (1982) Interaction of C1q with DNA. Mol. Immunol. 19, 1179–1187.
- 52, , , , and (1974) Native DNA binding in rheumatoid arthritis. Ann. Rheum. Dis. 33, 357–360.
- 53, , , and (2006) DNA interaction with human serum albumin studied by affinity capillary electrophoresis and FTIR spectroscopy. DNA Cell Biol. 25, 63–68.
- 54 (1989) Structures and functions of low affinity Fc receptors. Chem. Immunol. 47, 1–258.
- 55 (1991) Fc receptors and immunoglobulin binding factor. FASEB J. 5, 2684–2690.
- 56, , and (1996) Isolation and characterization of a cell surface albumin-binding protein from vascular endothelial cells. Proc. Natl. Acad. Sci. USA 93, 250–254.
- 57, , , , and (1997) Interaction of oligonucleotides with barrier fluid proteins. Biochemistry (Moscow) (in Russian) 62, 613–618.
- 58, , and (1990) Interpretation of the cross-reactivity of anti-DNA antibodies with cell surface proteins: the role of cell surface histones. Immunol. Lett. 23, 187–193.
- 59, , , , , et al. (1992) Nucleosomes and DNA bind to specific cell-surface molecules on murine cells and inducecytokine production. Clin. Immunol. Immunopathol. 63, 245–251.
- 60, , , , , et al. (1996) Binding of nucleosomes to a cell surface receptor: redistribution and endocytosis in the presence of lupus antibodies. Eur. J. Immunol. 26, 472–486.
- 61, , , , , et al. (2014) Identification of double-stranded genomic DNA spanning all chromosomes with mutated KRAS and p53 DNA in the serum exosomes of patients with pancreatic cancer. J. Biol. Chem. 289, 3869–3875.
- 62, , , , , et al. (2016) Exosomes in tears of healthy individuals: isolation, identification, and characterization. Biochem. Moscow Suppl. Ser. B 10, 165–172.
- 63, , , , , et al. (2011) A reliable method to concentrate circulating DNA. Anal. Biochem. 408, 354–356.
- 64, , , , , et al. (2015) Blood cell-surface-bound exosomes as a new source of tumor-specific microRNA. Eur. J. Cancer S13, 59–60.
- 65, , , and (2016) What is the acceptable hemolysis index for the measurements of plasma potassium, LDH and AST? Ann. Biol. Clin. (Paris) 74, 323–328.
- 66, , and (2004) Extracellular DNA in blood plasma of healthy donors and acute viral hepatitis. Important issues of the military medical service of the Siberian Military District at the present time. Mol. Biol. (Mosk.) (in Russian) 3, 157–158.
- 67, , and (2008) In Circulating Nucleic Acids Concentration in Blood of Patients with Rheumatoid Polyarthritis. Collection Symposium Series, Modern Treatment and Diagnostics Assays for Medical Practice (in Russian), ISBN 978-5-89017-011-8. ( G. S. Soldatova, editor.) Siberian Branch of Russian Academy of Sciences, Novosibirsk; p. 258–259.
- 68, , , , , et al. (2008) Circulating DNA in the blood of gastric cancer patients. Ann. N. Y. Acad. Sci. 1137, 226–231.
- 69, , , , , et al. (2005) Circulating nucleic acids in blood of patients with malignant tumors of gastrointestinal tract. Biomed. Khim. 51, 321–328.
- 70, , , , , et al. (2008) Cell-surface-bound circulating DNA as a prognostic factor in lung cancer. Ann. N. Y. Acad. Sci. 1137, 214–217.
- 71, , , , , et al. (2016) Epigenetic probes for lung cancer monitoring: line-1 methylation pattern in blood-circulating DNA. Russ. J. Genet. Appl. Res. 6, 99–104.
- 72, , , , et al. (2011) Methylated cell-free DNA in vitro and in vivo. Circulating Nucleic Acids in Plasma and Serum. Ed PB Gahan, Springer, London. 2, 185–194.
- 73, , , and (2001) Circulating nucleic acids in plasma or serum. Clin. Chim. Acta 313, 143–146.
- 74, , , , , et al. (2012) A comparative study of cell-free apoptotic and genomic DNA using FISH and massive parallel sequencing. Expert Opin. Biol. Ther. 12, S11–S17.
- 75, , , , and (2009) Profile of the circulating DNA in apparently healthy individuals. Clin. Chem. 55, 730–738.
- 76, , , , and (2009) Characterization of circulating DNA by parallel tagged sequencing on the 454 platform. Clin. Chim. Acta 409, 21–27.
- 77, , , and (2010) Next generation sequencing of serum circulating nucleic acids from patients with invasive ductal breast cancer reveals differences to healthy and nonmalignant controls. Mol. Cancer Res. 8, 335–342.
- 78, , , , , et al. (2016) Impact of airflow limitation on comorbidities and postoperative complications in patients undergoing thoracic surgery: a retrospective observational study. Ann. Thorac. Cardiovasc. Surg. 22, 146–152.
- 79, , , , , et al. (2000) Fluorescent methylation-specific polymerase chain reaction for DNA-based detection of prostate cancer in bodily fluids. Cancer Res. 60, 5941–5945.
- 80, and (2011) Origin, translocation and destination of extracellular occurring DNA-a new paradigm in genetic behaviour. Clin. Chim. Acta 412, 806–811.
- 81, , , , , et al. (2011) RARβ2 gene methylation level in the circulating DNA from blood of patients with lung cancer. Eur. J. Cancer Prev. 20, 453–455.
- 82, and (2007) The origin of circulating free DNA. Clin. Chem. 53, 2215.
- 83, , , , and (2005) Circulation DNA: biological implications for cancer metastasis and immunology. Med. Hypotheses 65, 956–961.
- 84, , and (2016) The dark side of circulating nucleic acids. Aging Cell 15, 398–399.
- 85, and (1981) Capping of DNA on phytohemagglutinin-stimulated human lymphoblasts. J. Immunol. 126, 703–705.
- 86, , , and (2016) Enrichment of extracellular DNA from the cultivation medium of human peripheral blood mononuclears with genomic CpG rich fragments results in increased cell production of IL-6 and TNF-a via activation of the NF-kB signaling pathway. Biomed. Khim. (in Russian) 62, 331–340.
- 87, , , , and (2013) DNA inhibits dsRNA-induced secretion of proinflammatory cytokines by gingival fibroblasts. Immunobiology 218, 272–280.
- 88, , and (2016) Oligodeoxynucleotide analogues of circulating DNA inhibit dsRNA-induced immune response at the early stages of signal transduction cascade in a cell type-dependent manner. Adv. Exp. Med. Biol. 924, 105–107.