- Email: [email protected]
The biology of conformation in the regulation of the senescent and transformed cell phenotypes
seminars in CANCER BIOLOGY, Vol. 12, 2002: pp. 165–171 doi:10.1016/S1044–579X(02)00020-2, available online at http://www.idealibrary.com on
The biology of conformation in the regulation of the senescent and transformed cell phenotypes A. Macieira-Coelho
time. Towards the end of the cell population life span, cell size increases again just before a postmitotic state is reached.1 In human fibroblasts the percentage of cells synthesizing DNA during the first 24 h after seeding them on a plastic surface, declines progressively through the cell population life span. On the other hand, the maximal percentage of cells synthesizing DNA during a 24-h period between seeding the cells and the time when they reach resting phase decreases pronouncedly during the last divisions, coinciding with a sudden increase in cell volume.2 These modifications follow the evolution of the density and spacing of the chromatin fibers.3 The density declines progressively while the spacing increases only during the last divisions coinciding with the pronounced increase in cell volume. These changes at the supramolecular level are translated morphologically through an increase of the nucleolar and nuclear areas.4 Pertinent to the phenomenon dealt herein is the fact that these changes are accompanied by modifications of cell contractility evaluated through the capacity to retract a plasma clot.2 There is a decline in contractility coupled with the decline in the maximal number of cells capable of synthesizing DNA during a 24-h period. The results show that the increase in size and decreased contractility are due to a different organization of the cell scaffold accompanied by a reorganization of the chromatin fibers, which in turn is coupled with an increasing difficulty in initiating DNA synthesis. The relationship between changes in cell volume and cell division was approached using different technologies. Tritiated thymidine labeling,5 direct analysis by cinematography,6 and comparison of cell volume with population doubling time,7 all led to the conclusion that larger cells have a decreased probability of entering the division cycle. With the increase in cell volume, higher growth factor concentrations and larger substratum areas become necessary for proliferation,8 the cells become flat with few
The cytoskeleton and the composition of the cytoplasmic membrane of normal somatic cells are modified during proliferation in vitro. The loss of the proliferative potential during serial divisions is due in part to these structural modifications that induce a decline in the cell conformational flexibility. During viral transformation, the changes in the affinity of the cell to its matrix and to neighboring cells increase the cell migratory capability maintaining the conformational flexibility; this way the cells can proliferate to densities where normal cells stop dividing. Cell proliferation, the transformed phenotype, and differentiation could be modulated by changing the electric charge of a substratum. Results support the view that the biology of conformation is crucial for the expression of these cell properties. Key words: contact inhibition / transformation / cancer / oncogenes / oncogenic viruses / malignancy / differentiation © 2002 Elsevier Science Ltd. All rights reserved.
Conformational flexibility and initiation of DNA synthesis during cell aging During serial proliferation normal fibroblasts go through modifications of the cytoskeleton and affinity towards the supporting matrix that are expressed in variations in morphology and volume coupled with declines in the probability of cycling. In chicken fibroblasts the first detectable increase in cell volume, which occurs around the fifteenth cell population doubling, coincides with a prolongation of the time needed to reach the maximal DNA synthetic activity, and is followed by a prolongation of the doubling From the INSERM, 73 bis rue Maréchal Foch, 78000 Versailles, France. E-mail: [email protected] © 2002 Elsevier Science Ltd. All rights reserved. 1044–579X / 02 / $ – see front matter
165
A. Macieira-Coelho
microvilli, show little ruffling activity and are almost completely devoid of macropinocytosis.9 The need of cell movements for the initiation of the division cycle has been known since the early observations of Loeb and Fleischer10 who described that movements were more active in areas of higher mitotic activity. As suggested by Puck11 the cytoskeleton could be the link through which flows the growth-regulatory information from the membrane. But what could be the nature of this information reaching the nucleus and triggering the initiation of the division cycle? It should be reminded that mammalian DNA is a molecule more than 1 m long, confined in a sphere, the nucleus, with a diameter of about 5 m. This implies an elaborate folding, which is regulated inter alia by the anchorage of this makeup to a protein matrix. The scaffold upon which chromatin is anchored plays a crucial role in the organization of the high-order structure of DNA. There is indeed a protein framework, called the nuclear protein matrix, with which DNA is associated. DNA synthesis-initiating sites are preferentially located at the borders between condensed chromatin and interchromatin areas,12 suggesting an important role for the nuclear matrix, in particular the peripheral nuclear region, in the initiation of DNA replication. The anchorage of DNA is crucial not only for replication but also for transcription, since nascent RNA is associated with the nuclear cage.13 The nuclear lamina, a filamentous protein meshwork lining the nucleoplasmic surface of the nuclear envelope, probably provides an anchoring site at the nuclear periphery for interphase chromatin.14 When the nuclear shell is isolated, it contains chromatin structures made of packed nucleosomes 28–32 nm thick (identical to the high-order solenoid DNA structure) that are associated with the three nuclear lamins.15 The lamina is composed of proteins called lamins, which seem to be intermediary structures between DNA-binding proteins and the cytoskeleton. Indeed the lamina is tightly bound to chromatin since it can be dissociated from chromatin only by high salt solution, which also extracts the tightly bound histones in the nucleosome cores.15 On the other hand, lamins have a striking sequence homology with intermediate filaments, a component of the cytoskeleton.14 Thus, the anchorage of chromatin seems to be fulfilled with the preservation of the continuity with the cytoplasmic scaffold. In this way, DNA is linked to the cytoskeleton through its anchorage to the nuclear cage and via the former to the cell membrane and
the extracellular matrix. The cytoskeleton and the nuclear matrix act as integrators not only of space but also of function within the cell. This whole structure can be seen as a three-dimensional manifold16 constituted by linked cranks, where information flows to a great extent through changes in molecular configurations. Cell behavior is determined by the way in which this network is connected, i.e. by its topology. The creation of new topological constraints allowing the flow of information through the cell depends on conformational flexibility. Conformational modifications in this network induced by membrane and cytoplasmic movements carry the message to the nucleus for initiation of DNA synthesis, helping to create the chromatin conformation favorable for the expression of the genes that have to be activated during the G1 period. Moreover the steric position of initiating sites must be crucial for entrance into the S period;23 contraction of nuclear structures or conversely too much stretching may lead to the arrest of cell division. Contraction takes place during contact inhibition of growth and excessive stretching is found in cellular senescence. In both instances the probability would be low for the chromatin structure to meet the right conformation and DNA synthesis initiating sites the right steric position for the commencement of the division cycle and progression along the cycle. Conversely, rapid movements should increase the probability for the initiating sites to reach the spatial configuration favorable for DNA synthesis.
Dependency of the transformed cell phenotype on conformational flexibility Figure 1 illustrates the morphology of normal human fibroblasts, identical cells transformed by Rous sarcoma virus (RSV), bovine fibroblasts also transformed by RSV, and human fibroblasts transformed by SV40. When the growth curves of these cell populations were compared with the percentage of labeled interphases during each 24 h after subcultivation, it was found that in normal human fibroblasts the percentage of labeled cells during a 24-h period fell to very low levels when the growth curve reached a plateau.17 The same pattern was observed after infection with RSV, but resting phase was reached at a cell density twice that of the controls. In RSV transformed bovine fibroblasts, the growth curve never reached a plateau and the percentage of labeled interphases never fell below 50%. In SV40 transformed human 166
Cell senescence and transformation in vitro
Figure 2. Schematic representation of the growth of the same cells illustrated in Figure 1.
fibroblasts the maximum cell density was not higher than in the controls but assumed a sigmoid shape, the percentage of labeled cells never falling below 50%. All these transformed cells had in common the fact that when they reached the density where the respective controls entered resting phase the percentage of dividing cells was still significantly high, but differed in the maximal densities reached, in cell attachment, and in the pattern of proliferation.17 The different patterns of proliferation were due to different affinities of the cells towards neighboring cells and towards the substratum. Human fibroblasts infected with RSV (Figure 2) have very little tendency to overlap but have a strong affinity towards the solid substratum. They reach very high densities with a plateau because DNA synthesis stops when each cell is attached to the minimum area possible. Bovine fibroblasts infected with RSV adhere well to the solid substratum and have a great tendency to overlap neighboring cells. The cell density increases progressively never reaching a plateau because DNA synthesis does not stop since the cells growing on top of one another have new substratum available to proliferate. Human fibroblasts transformed by SV40 do not have a marked tendency to overlap, but cells are detaching all the time leaving new areas available for cell proliferation. DNA synthesis goes on at high levels with low maximal cell densities. These results on cell overlapping can be explained on the basis of the experiments of Carter18 showing that cells in culture tend to migrate to the substratum with which they have a better affinity, a property he named haptotaxis. Thus, although contact inhibition of division can be dissociated from contact inhibition of movement,19 the latter can influence indirectly the former making available a larger substrate area for cell attachment.
Figure 1. Normal human fibroblasts (A), human fibroblasts transformed by Rous sarcoma virus (B), bovine fibroblasts transformed by Rous sarcoma virus (C), human fibroblasts transformed by SV40 virus (D).
167
A. Macieira-Coelho
Figure 3. Mouse L cells growing on polyglutamic acid (left) and on polylysine (right) covered surfaces.
polymers.20 The polymerization of BSA is due to crosslinkage originated by bonds between amino groups of proteins and active aldehyde groups of glutaraldehyde. After polymerization of the protein, active aldehyde groups from glutaraldehyde remain free in the polymer. Substances possessing free amino groups (e.g. polylysine) will link by covalent bonds to these aldehyde groups. Substances like heparin or DEAEdextran, which do not possess free amino groups are presumably fixed to the BSA polymer by noncovalent bonds; heparin binds to positively charged groups (e.g. amino groups) and DEAE-dextran to negatively charged groups (e.g. carboxyl groups) of the BSA. In
These results suggested that one can modify cell proliferation and the expression of the transformed phenotype through the modulation of the attachment of cells towards the substratum.
Modulation of the transformed cell phenotype through changes in cell attachment to the substratum A protein polymer made of bovine serum albumin (BSA) was adapted to cell culture using a method available for the preparation of insoluble protein
Figure 4. Electron microscopy of mouse L cells attached to a surface covered with polyglutamic acid (left) and with polylysine (right). The triangle points to the cell-matrix interface.
168
Cell senescence and transformation in vitro
this way, the protein polymer can be covered with a variety of substances that give different physicochemical properties to the surface used to support cell growth. The transformed mouse cell line, which on conventional cell culture flasks grows as a monolayer, was seeded on negatively charged polymers covered with polyglutamic acid.20 On this surface, the cells proliferated in clusters, reaching a higher number than when spreading on positively charged surfaces covered with polylysine (Figure 3). Figure 4 illustrates the attachment of the cells on a negatively and a positively charged surface at the ultrastructural level; there is a solution of continuity between the cell membrane and the substratum on the polyglutamic charged surface while adhesion of the cell membrane to the polylysine coated surface is characterized by continuity. Aggregation or monolayering is due to the electric charge of the substratum, since all neutral or negatively charged substances caused rounding of the cells, while all positively charged ones induced spreading, independently of other properties such as the pH at which each polyaminoacid is isoionic or the l and d forms of the molecules. Furthermore, neither the degree of binding of the polyaminoacid to the polymer nor structural differences influenced cell behavior. The critical nature of the electric charge could also be demonstrated covering the BSA with different proportions of lysine and alanine. The number of attached cells had a direct relationship with the amount of lysine. The polymer had no direct effect on cell metabolism. Indeed, substances bound to the substratum did not enter the cell. Moreover, substances that are able to influence cell metabolism when in solution, such as polyornithine, which stimulates the transport of proteins, lose this property when in an insoluble form bound to the substratum.21 Results also showed that polymerized serum used as a substratum loses its nutritional and growth-promoting properties on cultivated cells,21 and that growth factors lose their mitogenic property when attached to the substratum.22 Figure 5. Mouse leukemia L1210 cells growing in suspension (top), progressively attaching to a substratum (middle), and the culture being invaded by anchorage-dependent cells (bottom).
Modulation of malignancy through changes in cell attachment to the substratum We have obtained further evidence showing how the matrix to which cells attach can influence the cell phenotype, when L1210 mouse leukemia cells were subcultivated on a BSA polymer treated with 169
A. Macieira-Coelho
Figure 6. Survival of mice injected with L1210 wild-type leukemia cells grown in suspension (—) and with the same cells adapted to grow in a monolayer (- - -). The number of cells injected per animal in each group is indicated.
polylysine.20 A few cells eventually lost the rounded shape and spread out assuming a fibroblastic form (Figure 5). The fibroblastic cells progressively invaded the whole surface while the round cells were lost during trypsinization. These cells that acquired adhesive properties had lost the malignant potential (Figure 6). Indeed, 105 wild-type cells injected per mice killed all animals, while all animals survived when 108 revertant cells were injected. The L1210 cells were serially passaged during a year corresponding to more than 200 doublings; after this time, round cells grown in suspension could again be recovered and carried separately. This population had reacquired their malignant potential, although not fully since 106 cells took longer to kill the animals and one survived.
Conclusions The term ‘conformational biology’ was coined by Ivanov et al.26 to define the properties of DNA conferred by different forms and transitions of the double helix. The works reviewed herein show that the term can be extended to cell biology to explain transitions in cell phenotypes. During proliferation normal somatic cells go through genomic reorganizations during the division cycle, which induce a progressive structural evolution that changes cell volume and adhesion to a substratum, diminishing the flexibility of molecular conformation. On the other hand, the genomic modifications introduced by oncogenic viruses create new adhesive properties that allow the cells to migrate under conditions where normal cells 170
Cell senescence and transformation in vitro
are immobilized. The preservation of conformational flexibility through migration maintains the cells in a proliferative state. The experimental system described above where the cell phenotype could be modulated seeding the cells on substrata with different physicochemical properties, illustrates how adhesion can send the information to the genome. Indeed, the loss of malignancy for several generations after the cells acquired anchorage properties, can only be explained by an effect of cell adhesion-dependent physical forces on gene expression. We could also demonstrate the effect of cell attachment on gene expression, modulating myoblast cell differentiation through variations in cell attachment.24,25 On polylysine coated surfaces the cells spread out and stopped differentiating while on polyglutamic acid the differentiation program was completed. Identical results were obtained by seeding polylysine-coated beads on the cells without changes in cell morphology. These results showed that the substratum acts on the cell phenotype through physical forces, and that cell shape is not critical for the expression of the different phenotypes. Motion and function are interdependent in a cell through the tension in a network of structures that extend from the periphery to the nuclear cage that constitute the cellular scaffold. The function of this network is regulated through the syntheses of molecules with the right steric configuration and through energy turnover. Adhesion to a matrix is the trigger to build up tension. Cell behavior is determined by the way in which this network of structures is connected, i.e. by its topology. The maintenance of conformational flexibility through variations in cell-matrix interactions, allows the creation of new topological constraints and the maintenance of the flow of information.
5. Bowman PD, Meek RL, Daniel CW (1975) Aging of human fibroblasts in vitro, correlation between DNA synthetic ability and cell size. Exp Cell Res 93:184–190 6. Absher PM, Absher RG (1976) Clonal variation and aging of diploid fibroblasts. Cinematographic studies of cell pedigrees. Exp Cell Res 103:247–255 7. Mitsui Y, Schneider EL (1976) Relationship between cell replication and volume in senescent human diploid fibroblasts. Mech Ageing Dev 5:45–56 8. Collins VP, Arro E, Blomquist E, Brunk U, Frederikson BA, Westermark BA (1979) Cell locomotion and proliferation in relation to available surface area, serum concentration and culture age. Scanning Electron Microsc 111:411–420 9. Blomquist E, Arro E, Brunk U, Westermark B (1978) Plasma membrane motility of cultured human glia cells in phase II and III. Acta Pathol Microbiol Scand A Pathol 86:257– 263 10. Loeb L, Fleischer MS (1919) Cell migration and division. J Med Res 40:405–515 11. Puck TT (1977) Cyclic AMP, the microtubule-microfilament system and cancer. Proc Natl Acad Sci USA 74:4491–4495 12. Berezney R, Coffey DS (1975) Nuclear protein matrix: association with newly synthesized DNA. Science 189:291–293 13. Jackson DA, McCready SJ, Cook PR (1984) Replication and transcription depend on attachment of DNA to the nuclear cage. J Cell Phys 1:59–79 14. Gerace L (1985) Structural proteins in the eukaryotic nucleus. Nature 318:508–509 15. Bouvier D, Hubert J, Seve AP, Bouteille M (1985) Characterization of lamina-bound chromatin in the nuclear shell isolated from HeLa cells. Exp Cell Res 93:184–190 16. Thurston PW, Weeks JR (1984) The mathematics of three-dimensional manifolds. Sci Am July:94–106. 17. Macieira-Coelho A (1967) Relationship between DNA synthesis and cell density in normal and virus transformed cells. Int J Cancer 2:296–303 18. Carter SB (1965) Principles of cell mobility: the direction of cell movement and cancer invasion. Nature 208:1183–1187 19. Macieira-Coelho A (1967) Dissociation between inhibition of movement and inhibition of division in RSV transformed human fibroblasts. Exp Cell Res 47:193–200 20. Macieira-Coelho A, Avrameas S (1972) Modulation of cell behavior in vitro by the substratum in fibroblastic and leukemic mouse cell lines. Proc Natl Acad Sci USA 69:2469–2473 21. Macieira-Coelho A, Berumen L, Avrameas S (1974) Properties of protein polymers as substratum for cell growth in vitro. J Cell Phys 83:379–388 22. Avrameas S, Ternynck T, Macieira-Coelho A (1976) Loss of mitogenic activity by immobilized lectins. Biochem Biophys Res Commun 72:790–795 23. Macieira-Coelho A (1983) Changes in membrane properties associated with cellular aging. Int Rev Cytol 83:183–220 24. Wahrmann JP, Delain D, Bournoutian C, Macieira-Coelho A (1981) Modulation of differentiation in vitro. Influence of the attachment surface on myogenesis. In Vitro 17:752–762 25. Sénéchal H, Wahrmann JP, Delain D, Macieira-Coelho A (1984) Modulation of differentiation in vitro. II. Influence of cell spreading and surface events on myogenesis. In Vitro 20:692– 698 26. Ivanov VI, Minchenkova LE, Minyat EE, Schyolkina AK (1983) Cooperative transitions in DNA with no separation of strands. Cold Spring Harbor Symp 47:243–250
References 1. Lima L, Macieira-Coelho A (1972) Parameters of aging in chicken embryo fibroblasts cultivated in vitro. Exp Cell Res 70:279–284 2. Macieira-Coelho A, Azzarone B (1990) Correlation between contractility and proliferation in human fibroblasts. J Cell Phys 142:610–614 3. Macieira-Coelho A (1991) Chromatin reorganization during senescence of proliferating cells. Mutat Res 256:81–104 4. BeMiller PM, Miller JE (1979) Cytological changes in senescing WI-38. A statistical analysis. Mech Ageing Dev 10:1–15
171
The biology of conformation in the regulation of the senescent and transformed cell phenotypes A. Macieira-Coelho
time. Towards the end of the cell population life span, cell size increases again just before a postmitotic state is reached.1 In human fibroblasts the percentage of cells synthesizing DNA during the first 24 h after seeding them on a plastic surface, declines progressively through the cell population life span. On the other hand, the maximal percentage of cells synthesizing DNA during a 24-h period between seeding the cells and the time when they reach resting phase decreases pronouncedly during the last divisions, coinciding with a sudden increase in cell volume.2 These modifications follow the evolution of the density and spacing of the chromatin fibers.3 The density declines progressively while the spacing increases only during the last divisions coinciding with the pronounced increase in cell volume. These changes at the supramolecular level are translated morphologically through an increase of the nucleolar and nuclear areas.4 Pertinent to the phenomenon dealt herein is the fact that these changes are accompanied by modifications of cell contractility evaluated through the capacity to retract a plasma clot.2 There is a decline in contractility coupled with the decline in the maximal number of cells capable of synthesizing DNA during a 24-h period. The results show that the increase in size and decreased contractility are due to a different organization of the cell scaffold accompanied by a reorganization of the chromatin fibers, which in turn is coupled with an increasing difficulty in initiating DNA synthesis. The relationship between changes in cell volume and cell division was approached using different technologies. Tritiated thymidine labeling,5 direct analysis by cinematography,6 and comparison of cell volume with population doubling time,7 all led to the conclusion that larger cells have a decreased probability of entering the division cycle. With the increase in cell volume, higher growth factor concentrations and larger substratum areas become necessary for proliferation,8 the cells become flat with few
The cytoskeleton and the composition of the cytoplasmic membrane of normal somatic cells are modified during proliferation in vitro. The loss of the proliferative potential during serial divisions is due in part to these structural modifications that induce a decline in the cell conformational flexibility. During viral transformation, the changes in the affinity of the cell to its matrix and to neighboring cells increase the cell migratory capability maintaining the conformational flexibility; this way the cells can proliferate to densities where normal cells stop dividing. Cell proliferation, the transformed phenotype, and differentiation could be modulated by changing the electric charge of a substratum. Results support the view that the biology of conformation is crucial for the expression of these cell properties. Key words: contact inhibition / transformation / cancer / oncogenes / oncogenic viruses / malignancy / differentiation © 2002 Elsevier Science Ltd. All rights reserved.
Conformational flexibility and initiation of DNA synthesis during cell aging During serial proliferation normal fibroblasts go through modifications of the cytoskeleton and affinity towards the supporting matrix that are expressed in variations in morphology and volume coupled with declines in the probability of cycling. In chicken fibroblasts the first detectable increase in cell volume, which occurs around the fifteenth cell population doubling, coincides with a prolongation of the time needed to reach the maximal DNA synthetic activity, and is followed by a prolongation of the doubling From the INSERM, 73 bis rue Maréchal Foch, 78000 Versailles, France. E-mail: [email protected] © 2002 Elsevier Science Ltd. All rights reserved. 1044–579X / 02 / $ – see front matter
165
A. Macieira-Coelho
microvilli, show little ruffling activity and are almost completely devoid of macropinocytosis.9 The need of cell movements for the initiation of the division cycle has been known since the early observations of Loeb and Fleischer10 who described that movements were more active in areas of higher mitotic activity. As suggested by Puck11 the cytoskeleton could be the link through which flows the growth-regulatory information from the membrane. But what could be the nature of this information reaching the nucleus and triggering the initiation of the division cycle? It should be reminded that mammalian DNA is a molecule more than 1 m long, confined in a sphere, the nucleus, with a diameter of about 5 m. This implies an elaborate folding, which is regulated inter alia by the anchorage of this makeup to a protein matrix. The scaffold upon which chromatin is anchored plays a crucial role in the organization of the high-order structure of DNA. There is indeed a protein framework, called the nuclear protein matrix, with which DNA is associated. DNA synthesis-initiating sites are preferentially located at the borders between condensed chromatin and interchromatin areas,12 suggesting an important role for the nuclear matrix, in particular the peripheral nuclear region, in the initiation of DNA replication. The anchorage of DNA is crucial not only for replication but also for transcription, since nascent RNA is associated with the nuclear cage.13 The nuclear lamina, a filamentous protein meshwork lining the nucleoplasmic surface of the nuclear envelope, probably provides an anchoring site at the nuclear periphery for interphase chromatin.14 When the nuclear shell is isolated, it contains chromatin structures made of packed nucleosomes 28–32 nm thick (identical to the high-order solenoid DNA structure) that are associated with the three nuclear lamins.15 The lamina is composed of proteins called lamins, which seem to be intermediary structures between DNA-binding proteins and the cytoskeleton. Indeed the lamina is tightly bound to chromatin since it can be dissociated from chromatin only by high salt solution, which also extracts the tightly bound histones in the nucleosome cores.15 On the other hand, lamins have a striking sequence homology with intermediate filaments, a component of the cytoskeleton.14 Thus, the anchorage of chromatin seems to be fulfilled with the preservation of the continuity with the cytoplasmic scaffold. In this way, DNA is linked to the cytoskeleton through its anchorage to the nuclear cage and via the former to the cell membrane and
the extracellular matrix. The cytoskeleton and the nuclear matrix act as integrators not only of space but also of function within the cell. This whole structure can be seen as a three-dimensional manifold16 constituted by linked cranks, where information flows to a great extent through changes in molecular configurations. Cell behavior is determined by the way in which this network is connected, i.e. by its topology. The creation of new topological constraints allowing the flow of information through the cell depends on conformational flexibility. Conformational modifications in this network induced by membrane and cytoplasmic movements carry the message to the nucleus for initiation of DNA synthesis, helping to create the chromatin conformation favorable for the expression of the genes that have to be activated during the G1 period. Moreover the steric position of initiating sites must be crucial for entrance into the S period;23 contraction of nuclear structures or conversely too much stretching may lead to the arrest of cell division. Contraction takes place during contact inhibition of growth and excessive stretching is found in cellular senescence. In both instances the probability would be low for the chromatin structure to meet the right conformation and DNA synthesis initiating sites the right steric position for the commencement of the division cycle and progression along the cycle. Conversely, rapid movements should increase the probability for the initiating sites to reach the spatial configuration favorable for DNA synthesis.
Dependency of the transformed cell phenotype on conformational flexibility Figure 1 illustrates the morphology of normal human fibroblasts, identical cells transformed by Rous sarcoma virus (RSV), bovine fibroblasts also transformed by RSV, and human fibroblasts transformed by SV40. When the growth curves of these cell populations were compared with the percentage of labeled interphases during each 24 h after subcultivation, it was found that in normal human fibroblasts the percentage of labeled cells during a 24-h period fell to very low levels when the growth curve reached a plateau.17 The same pattern was observed after infection with RSV, but resting phase was reached at a cell density twice that of the controls. In RSV transformed bovine fibroblasts, the growth curve never reached a plateau and the percentage of labeled interphases never fell below 50%. In SV40 transformed human 166
Cell senescence and transformation in vitro
Figure 2. Schematic representation of the growth of the same cells illustrated in Figure 1.
fibroblasts the maximum cell density was not higher than in the controls but assumed a sigmoid shape, the percentage of labeled cells never falling below 50%. All these transformed cells had in common the fact that when they reached the density where the respective controls entered resting phase the percentage of dividing cells was still significantly high, but differed in the maximal densities reached, in cell attachment, and in the pattern of proliferation.17 The different patterns of proliferation were due to different affinities of the cells towards neighboring cells and towards the substratum. Human fibroblasts infected with RSV (Figure 2) have very little tendency to overlap but have a strong affinity towards the solid substratum. They reach very high densities with a plateau because DNA synthesis stops when each cell is attached to the minimum area possible. Bovine fibroblasts infected with RSV adhere well to the solid substratum and have a great tendency to overlap neighboring cells. The cell density increases progressively never reaching a plateau because DNA synthesis does not stop since the cells growing on top of one another have new substratum available to proliferate. Human fibroblasts transformed by SV40 do not have a marked tendency to overlap, but cells are detaching all the time leaving new areas available for cell proliferation. DNA synthesis goes on at high levels with low maximal cell densities. These results on cell overlapping can be explained on the basis of the experiments of Carter18 showing that cells in culture tend to migrate to the substratum with which they have a better affinity, a property he named haptotaxis. Thus, although contact inhibition of division can be dissociated from contact inhibition of movement,19 the latter can influence indirectly the former making available a larger substrate area for cell attachment.
Figure 1. Normal human fibroblasts (A), human fibroblasts transformed by Rous sarcoma virus (B), bovine fibroblasts transformed by Rous sarcoma virus (C), human fibroblasts transformed by SV40 virus (D).
167
A. Macieira-Coelho
Figure 3. Mouse L cells growing on polyglutamic acid (left) and on polylysine (right) covered surfaces.
polymers.20 The polymerization of BSA is due to crosslinkage originated by bonds between amino groups of proteins and active aldehyde groups of glutaraldehyde. After polymerization of the protein, active aldehyde groups from glutaraldehyde remain free in the polymer. Substances possessing free amino groups (e.g. polylysine) will link by covalent bonds to these aldehyde groups. Substances like heparin or DEAEdextran, which do not possess free amino groups are presumably fixed to the BSA polymer by noncovalent bonds; heparin binds to positively charged groups (e.g. amino groups) and DEAE-dextran to negatively charged groups (e.g. carboxyl groups) of the BSA. In
These results suggested that one can modify cell proliferation and the expression of the transformed phenotype through the modulation of the attachment of cells towards the substratum.
Modulation of the transformed cell phenotype through changes in cell attachment to the substratum A protein polymer made of bovine serum albumin (BSA) was adapted to cell culture using a method available for the preparation of insoluble protein
Figure 4. Electron microscopy of mouse L cells attached to a surface covered with polyglutamic acid (left) and with polylysine (right). The triangle points to the cell-matrix interface.
168
Cell senescence and transformation in vitro
this way, the protein polymer can be covered with a variety of substances that give different physicochemical properties to the surface used to support cell growth. The transformed mouse cell line, which on conventional cell culture flasks grows as a monolayer, was seeded on negatively charged polymers covered with polyglutamic acid.20 On this surface, the cells proliferated in clusters, reaching a higher number than when spreading on positively charged surfaces covered with polylysine (Figure 3). Figure 4 illustrates the attachment of the cells on a negatively and a positively charged surface at the ultrastructural level; there is a solution of continuity between the cell membrane and the substratum on the polyglutamic charged surface while adhesion of the cell membrane to the polylysine coated surface is characterized by continuity. Aggregation or monolayering is due to the electric charge of the substratum, since all neutral or negatively charged substances caused rounding of the cells, while all positively charged ones induced spreading, independently of other properties such as the pH at which each polyaminoacid is isoionic or the l and d forms of the molecules. Furthermore, neither the degree of binding of the polyaminoacid to the polymer nor structural differences influenced cell behavior. The critical nature of the electric charge could also be demonstrated covering the BSA with different proportions of lysine and alanine. The number of attached cells had a direct relationship with the amount of lysine. The polymer had no direct effect on cell metabolism. Indeed, substances bound to the substratum did not enter the cell. Moreover, substances that are able to influence cell metabolism when in solution, such as polyornithine, which stimulates the transport of proteins, lose this property when in an insoluble form bound to the substratum.21 Results also showed that polymerized serum used as a substratum loses its nutritional and growth-promoting properties on cultivated cells,21 and that growth factors lose their mitogenic property when attached to the substratum.22 Figure 5. Mouse leukemia L1210 cells growing in suspension (top), progressively attaching to a substratum (middle), and the culture being invaded by anchorage-dependent cells (bottom).
Modulation of malignancy through changes in cell attachment to the substratum We have obtained further evidence showing how the matrix to which cells attach can influence the cell phenotype, when L1210 mouse leukemia cells were subcultivated on a BSA polymer treated with 169
A. Macieira-Coelho
Figure 6. Survival of mice injected with L1210 wild-type leukemia cells grown in suspension (—) and with the same cells adapted to grow in a monolayer (- - -). The number of cells injected per animal in each group is indicated.
polylysine.20 A few cells eventually lost the rounded shape and spread out assuming a fibroblastic form (Figure 5). The fibroblastic cells progressively invaded the whole surface while the round cells were lost during trypsinization. These cells that acquired adhesive properties had lost the malignant potential (Figure 6). Indeed, 105 wild-type cells injected per mice killed all animals, while all animals survived when 108 revertant cells were injected. The L1210 cells were serially passaged during a year corresponding to more than 200 doublings; after this time, round cells grown in suspension could again be recovered and carried separately. This population had reacquired their malignant potential, although not fully since 106 cells took longer to kill the animals and one survived.
Conclusions The term ‘conformational biology’ was coined by Ivanov et al.26 to define the properties of DNA conferred by different forms and transitions of the double helix. The works reviewed herein show that the term can be extended to cell biology to explain transitions in cell phenotypes. During proliferation normal somatic cells go through genomic reorganizations during the division cycle, which induce a progressive structural evolution that changes cell volume and adhesion to a substratum, diminishing the flexibility of molecular conformation. On the other hand, the genomic modifications introduced by oncogenic viruses create new adhesive properties that allow the cells to migrate under conditions where normal cells 170
Cell senescence and transformation in vitro
are immobilized. The preservation of conformational flexibility through migration maintains the cells in a proliferative state. The experimental system described above where the cell phenotype could be modulated seeding the cells on substrata with different physicochemical properties, illustrates how adhesion can send the information to the genome. Indeed, the loss of malignancy for several generations after the cells acquired anchorage properties, can only be explained by an effect of cell adhesion-dependent physical forces on gene expression. We could also demonstrate the effect of cell attachment on gene expression, modulating myoblast cell differentiation through variations in cell attachment.24,25 On polylysine coated surfaces the cells spread out and stopped differentiating while on polyglutamic acid the differentiation program was completed. Identical results were obtained by seeding polylysine-coated beads on the cells without changes in cell morphology. These results showed that the substratum acts on the cell phenotype through physical forces, and that cell shape is not critical for the expression of the different phenotypes. Motion and function are interdependent in a cell through the tension in a network of structures that extend from the periphery to the nuclear cage that constitute the cellular scaffold. The function of this network is regulated through the syntheses of molecules with the right steric configuration and through energy turnover. Adhesion to a matrix is the trigger to build up tension. Cell behavior is determined by the way in which this network of structures is connected, i.e. by its topology. The maintenance of conformational flexibility through variations in cell-matrix interactions, allows the creation of new topological constraints and the maintenance of the flow of information.
5. Bowman PD, Meek RL, Daniel CW (1975) Aging of human fibroblasts in vitro, correlation between DNA synthetic ability and cell size. Exp Cell Res 93:184–190 6. Absher PM, Absher RG (1976) Clonal variation and aging of diploid fibroblasts. Cinematographic studies of cell pedigrees. Exp Cell Res 103:247–255 7. Mitsui Y, Schneider EL (1976) Relationship between cell replication and volume in senescent human diploid fibroblasts. Mech Ageing Dev 5:45–56 8. Collins VP, Arro E, Blomquist E, Brunk U, Frederikson BA, Westermark BA (1979) Cell locomotion and proliferation in relation to available surface area, serum concentration and culture age. Scanning Electron Microsc 111:411–420 9. Blomquist E, Arro E, Brunk U, Westermark B (1978) Plasma membrane motility of cultured human glia cells in phase II and III. Acta Pathol Microbiol Scand A Pathol 86:257– 263 10. Loeb L, Fleischer MS (1919) Cell migration and division. J Med Res 40:405–515 11. Puck TT (1977) Cyclic AMP, the microtubule-microfilament system and cancer. Proc Natl Acad Sci USA 74:4491–4495 12. Berezney R, Coffey DS (1975) Nuclear protein matrix: association with newly synthesized DNA. Science 189:291–293 13. Jackson DA, McCready SJ, Cook PR (1984) Replication and transcription depend on attachment of DNA to the nuclear cage. J Cell Phys 1:59–79 14. Gerace L (1985) Structural proteins in the eukaryotic nucleus. Nature 318:508–509 15. Bouvier D, Hubert J, Seve AP, Bouteille M (1985) Characterization of lamina-bound chromatin in the nuclear shell isolated from HeLa cells. Exp Cell Res 93:184–190 16. Thurston PW, Weeks JR (1984) The mathematics of three-dimensional manifolds. Sci Am July:94–106. 17. Macieira-Coelho A (1967) Relationship between DNA synthesis and cell density in normal and virus transformed cells. Int J Cancer 2:296–303 18. Carter SB (1965) Principles of cell mobility: the direction of cell movement and cancer invasion. Nature 208:1183–1187 19. Macieira-Coelho A (1967) Dissociation between inhibition of movement and inhibition of division in RSV transformed human fibroblasts. Exp Cell Res 47:193–200 20. Macieira-Coelho A, Avrameas S (1972) Modulation of cell behavior in vitro by the substratum in fibroblastic and leukemic mouse cell lines. Proc Natl Acad Sci USA 69:2469–2473 21. Macieira-Coelho A, Berumen L, Avrameas S (1974) Properties of protein polymers as substratum for cell growth in vitro. J Cell Phys 83:379–388 22. Avrameas S, Ternynck T, Macieira-Coelho A (1976) Loss of mitogenic activity by immobilized lectins. Biochem Biophys Res Commun 72:790–795 23. Macieira-Coelho A (1983) Changes in membrane properties associated with cellular aging. Int Rev Cytol 83:183–220 24. Wahrmann JP, Delain D, Bournoutian C, Macieira-Coelho A (1981) Modulation of differentiation in vitro. Influence of the attachment surface on myogenesis. In Vitro 17:752–762 25. Sénéchal H, Wahrmann JP, Delain D, Macieira-Coelho A (1984) Modulation of differentiation in vitro. II. Influence of cell spreading and surface events on myogenesis. In Vitro 20:692– 698 26. Ivanov VI, Minchenkova LE, Minyat EE, Schyolkina AK (1983) Cooperative transitions in DNA with no separation of strands. Cold Spring Harbor Symp 47:243–250
References 1. Lima L, Macieira-Coelho A (1972) Parameters of aging in chicken embryo fibroblasts cultivated in vitro. Exp Cell Res 70:279–284 2. Macieira-Coelho A, Azzarone B (1990) Correlation between contractility and proliferation in human fibroblasts. J Cell Phys 142:610–614 3. Macieira-Coelho A (1991) Chromatin reorganization during senescence of proliferating cells. Mutat Res 256:81–104 4. BeMiller PM, Miller JE (1979) Cytological changes in senescing WI-38. A statistical analysis. Mech Ageing Dev 10:1–15
171