| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cell Growth & Differentiation |
Department of Human Physiology, University of Malaga, Campus Teatinos, 29080 Malaga, Spain [M. G. F.]; Department of Biomedical Sciences, Section of General Pathology, University of Modena and Reggio Emilia, 41100 Modena, Italy [M. G. F., L. T., L. M., M. N., M. P., A. C.]; Department of Experimental Pathology, Section of Microbiology, University of Bologna, 40126 Bologna, Italy [S. S.]; and Laboratory of Confocal Microscopy and Image Analysis, Department of Biophysics, Jagiellonian University, 30-387 Krakow, Poland [J. D.]
| Abstract |
|---|
|
|
|---|
-treated U937 cells), we found that in apoptotic cells CL moves to the outer leaflet of mitochondrial inner membrane in a time-dependent manner. This occurs before the appearance of apoptosis markers such as plasma-membrane exposure of phosphatidylserine, changes in mitochondrial membrane potential, DNA fragmentation, but after the production of reactive oxygen species. The exposure of a phospholipid on the outer surface during apoptosis thus occurs not only at the plasma membrane level but also in mitochondria, reinforcing the hypothesis of mitoptosis as a crucial regulating system for programmed cell death, also occurring in cancer cells after treatment with antineoplastic agents. | Introduction |
|---|
|
|
|---|
Evidence shows that mitochondria, the main intracellular source of energy, play a crucial role in several processes linked to apoptosis by disrupting electron transport and energy metabolism, by releasing soluble proteins such as cyt c4 , certain procaspases (2 and 9), and apoptosis-inducing factor from the intermembrane space into the cytoplasm, or by altering cellular redox potential (5 , 6) . Thus, mitochondria operate not only as power stations required for aerobic life but also as amplifiers or sometimes even as sources of signals for apoptosis. These two alternative functions are interrelated because electron transfer from respiratory substrates to oxygen, involved in energy conservation, is inevitably accompanied by formation of ROS that act as inducers or secondary messengers of apoptosis (7) . Recently, an interesting theory has been put forward, according to which mitochondria possess an autonomic system that allows them to degrade when irreversibly damaged, thus committing a sort of suicide, defined as mitoptosis by analogy with apoptosis (1 , 8) . Massive mitoptosis attributable to a massive stimulus such as, e.g., that provided by large amounts of ROS, can result in apoptosis because of the release of cyt c or apoptosis-inducing factor normally contained in the organelle.
Mitochondria contain CL, a unique phospholipid with dimeric structure, containing four unsaturated fatty acids and two negative charges (9) . The CL is located exclusively in the inner membrane of these organelles and particularly in the intermembrane contact sites (10) . CL has special properties, which include the organization into surface head group microdomains (11) , asymmetry in bilayer membranes (12 , 13) , and potential to form nonbilayer HII-phase (14 , 15) . Moreover, CL is essential as the boundary lipid for the normal functions of various mitochondrial proteins such as NADH:ubiquinone oxidoreductase, cyt c oxidase, F0F1 ATPase, adenine nucleotide transporter, and cyt c (10 , 16, 17, 18) , as well as for the function of cytosolic proteins such as tBid (19) .
Recently, we have observed a significant difference between HL-60 cells and their apoptosis-resistant clone HCW-2 (20)
, with respect to CL distribution in the mitochondrial inner membrane and have hypothesized that such distribution was related to a different metabolism and pathway of producing ATP. Furthermore, in these cell lines we analyzed mitochondrial sensitivity to a variety of damages and the capacity to undergo apoptosis under different stimuli and found that the metabolic changes occurring in HCW-2 could be, at least in part, responsible for either resistance to drug-induced apoptosis or cross-resistance to immunological attacks.5
Thus, changes in the metabolism of these type of cancer cells, highly resistant to a variety of antineoplastic agents, could explain their malignity. Using this model, together with the classical one of U937 cells treated with TNF-
(21)
, we have investigated the role of CL distribution in the earliest phases of apoptosis and its relationship to the production of oxygen-free radicals.
| Results |
|---|
|
|
|---|
640 nm (red, detectable in FL3 channel; Ref. 22
). The curve of red emission shows a peculiar phenomenon that is related not only to the dose of dye but also to its distribution in the mitochondrial inner membrane (12)
. Indeed, increasing NAO concentrations saturate all of the CL present on the outer leaflet of the mitochondrial inner membrane, and a plateau of fluorescence intensity is soon observed. Because NAO forms dimers with CL, the increase in FL3 is accompanied by a decrease in the fluorescence of monomers (detected in FL1). Then, NAO can pass through the lipid bilayer and bind all of the CL residues present on the internal surface of the mitochondrial inner membrane. The saturation of CL residues gives a second fluorescence plateau that represents the maximum fluorescence and that is considered as 100% fluorescence intensity (i.e., the sum of NAO bound to all of the CL present on either the outer or internal leaflet of the inner mitochondrial membrane). Clearly, the intensity of the first plateau corresponds to the percentage of CL present on the cytoplasmic (outer) face of the mitochondrial inner membrane (20)
.
The NAO-CL binding curves, obtained by analyzing FL3 histograms, showed the typical biphasic pattern characteristics of CL asymmetric distribution in the mitochondrial inner membrane leaflets (20)
. This phenomenon was observed in all of the cell types studied and under all of the experimental conditions tested. Fig. 1
refers to a representative experiment on HL-60 (top panels) and HCW-2 (bottom panels) cells and shows that a different percentage of CL is present on the outer leaflet of the mitochondrial inner membrane, as already reported by our group.
|
We then used a different model of apoptosis, that of TNF-treated U937 cells. As we have previously shown that an antioxidant is capable of protecting such cells from mitochondrial damage and apoptosis (21 , 25) , we treated parallel samples with the antioxidant BHT. This molecule was able to prevent changes in CL distribution during incubation with TNF (data not shown) and to protect cells from apoptosis (see below).
Fig. 2
shows the kinetics of change in CL distribution in the two experimental models used. In both models, these modifications were an early phenomenon, already visible after a 1.5-h incubation with the apoptotic stimulus. By contrast, at this point in time very few cells were positively stained by annexin V (i.e., early apoptotic), and no cells were frankly apoptotic, i.e., showed the typical hypodiploic peak after staining with propidium iodide (Fig. 3)
. Moreover, at this time no cells had changes in 
(data not shown).
|
|
|
, U937 cells showed a small increase in peroxide production; this increase could be prevented by pretreatment with BHT. | Discussion |
|---|
|
|
|---|
We found that the phenomenon of programmed cell death in STS-treated HL-60 cells was characterized by a consistent modification of CL distribution on the mitochondrial inner membrane that was not influenced by a pan-caspase inhibitor. Changes in CL distribution were also observed during apoptosis of TNF-treated U937 cells. Such a change was observable in a very early phase of apoptosis, was time dependent, and was absent in apoptosis-resistant HCW-2 cells as well as in U937 cells protected with the antioxidant BHT. Thus, changes in CL distribution and, in particular, increased exposure of CL on the outer leaflet of the mitochondrial inner membrane can be considered early markers of apoptosis. Interestingly, the latter phenomenon closely resembles the exposure of another phospholipid, phosphatidylserine, on the outer surface of another membrane, the plasma membrane, which is one of the earliest modifications occurring during cell death of apoptotic type.
We investigated whether changes in CL distribution could be the consequence of overproduction of ROS and found that increased ROS production preceded changes in CL distribution. This was further demonstrated by the protective effect of BHT on TNF-treated U937 cells. ROS and other free radicals generated by normal metabolic processes are potentially very toxic to cells, and because of their highly reactive nature, can damage proteins, lipids, and nucleic acids, inactivating or inhibiting their normal function (8 , 26) . One of the destructive effects of ROS is the initiation of lipid peroxidation, resulting in runaway chain reactions leading to destruction of the cell membrane, breakdown of compartmentalization, and release of protein-catabolic enzymes, with subsequent cell death (26) . Evidence shows that increased ROS generation induces the peroxidation of CL because CL contains significant quantities of highly unsaturated fatty acids (27, 28, 29) . Moreover, it has been recently suggested that oxidation of CL is an early phenomenon of apoptosis (30 , 31) that could facilitate the binding of Bax and tBid, which occurs only in the presence of decreased stability of the planar phospholipid bilayer (32 , 33) . Nonbilayer hexagonal structures are normally transient, but their formation can be stabilized by drugs, antibiotics, nonpolar peptides, changes in temperature, changes in pH, and modifications in the concentration of divalent cations (34) . It is well known that events such as alterations in Ca2+ concentration, pH, and a shift toward the glycolytic pathway (4 , 35 , 36) , events that occur during apoptosis, could be responsible for changes in CL distribution and structure 10. Moreover, peroxidation of CL alters the molecular conformation and packing, leading to the formation of a nonbilayer hexagonal structure (34) . This modification could expose neo-nanodomains or neo-epitopes that can be immunogenic (34 , 37 , 38) . It is likely that the formation of these structures could provide specificity for targeting of tBid to mitochondria during apoptosis (19) . In addition, when peroxidized, CL fails to bind cyt c by disrupting molecular interactions between the phospholipid and the protein (39) ; the conformation of cyt c can also be modified by ROS (29) . The sum of these modifications could provoke the release of cyt c from mitochondria into the cytosol. In fact, if CL peroxidation is inhibited, cyt c release and apoptosis are inhibited (29) . Thus, it can be hypothesized that change in CL distribution represents a sort of signal that, according to the theory of mitoptosis (7 , 8) , could inform the rest of the cell of the presence of organelles damaged by ROS.
In our previous work, we demonstrated that HCW-2 cells are characterized by a higher amount of CL on the outer leaflet of the mitochondrial internal membrane, which was interpreted as a skewed metabolism toward the production of ATP in the cytoplasm rather than in mitochondria. Such cells also have mitochondria with a lower 
value than that of the HL-60 parental line, reinforcing the idea that these organelles are less functional as far as energy production is concerned (20)
. It can be speculated that a minor production of ROS and resistance to CL oxidation, very likely along with an overexpression of endogenous antioxidants, could represent a mechanism of resistance first to mitoptosis, then to apoptosis.
In conclusion, the data reported in this study, together with our previous results, suggest that: (a) increased production of ROS during apoptosis could modify intramitochondrial CL distribution; (b) this phenomenon precedes the release of cyt c and 
impairment, events that certainly destabilize mitochondrial homeostasis; and (c) HCW-2 cells possess mitochondria that are less efficient in phosphorylative oxidation but more resistant to changes in CL distribution. As a result, mitoptosis and then apoptosis are inhibited. However, additional investigations are required to understand at the molecular level how changes in CL distribution are the cause or the effect of the formation of nonbilayer hexagonal structures and whether oxidation of CL precedes or follows changes in its intramitochondrial distribution. In any case, our observations on the mechanism by which apoptosis-resistant cells survive could have a particular interest in oncology because drug-resistant clones clearly represent a major problem in the treatment of tumors. Understanding their metabolism can thus help in developing new strategies.
| Materials and Methods |
|---|
|
|
|---|
, CHX, BHT, and other common chemicals were from Sigma (St. Louis, MO) and were of analytical grade. Z-VAD.fmk was from Alexis Biochemicals (San Diego, CA). Stock solutions were prepared according to manufacturers instructions.
Cell Cultures.
Mycoplasma-free human leukemia cell lines U937 (monocytic), HL-60 (promyelocytic), and its apoptosis-resistant clone HCW-2 [the last provided by Drs. James H. Wyche and Zhihua Han (Brown University, Providence, RI)] were grown in suspension in complete culture medium (RPMI 1640 supplemented with 10% heat-inactivated FCS, 2 mM L-glutamine, 100 IU/ml penicillin, and 100 µg/ml streptomycin) and kept at 37°C in a humidified atmosphere (5% CO2 in air). Cells were collected during the log phase of growth, washed in PBS, counted, and adjusted to a density of 0.5 x 106 cells/ml.
Cell Treatment.
The cells were induced to die by apoptosis after different stimuli. U937 cells were incubated in presence of 50 IU/ml TNF for 14 h after a 2-h preincubation with 4 µM CHX 21. HL-60 and HCW-2 cells were treated with 5 µM STS for 14 h (40)
. For studies of ROS production, U937 were collected 1 and 2 h after addition of TNF-
and HL-60 and HCW-2 1 h after addition of STS. In some experiments, U937 cells were pretreated for 30 min with the antioxidant BHT (50 µM) before adding TNF-
. In other assays, before adding STS, HL-60 were preincubated for 30 min at 37°C with 100 µM Z-VAD.fmk.
Analysis of CL Distribution.
To analyze CL intramitochondrial distribution, we have used a new assay, recently set up by us (20
, 21)
. This method uses the lipophilic cation probe NAO, and, in particular, its capacity to form dimers when it interacts with diacidic phospholipids, along with the fact that its fluorescence emission shifts from 525 nM (monomeric form of the dye) to
640 nm (under dimeric conditions when NAO and CL are present in a stoichiometric ratio of 2:1; Refs. 21
, 41
). Cells were labeled with different concentrations of NAO to take advantage of the capacity of this fluorescent dye to gradually bind CL with high affinity (41)
. At low doses, NAO binds CL on the outer leaflet of mitochondrial inner membrane in a monomeric form; this phenomenon gives a green fluorescence emission after excitation at 490 nm. Increasing NAO concentrations results in a qualitative change in fluorescence emission because of the capacity of NAO to form dimers (i.e., two NAO molecules bind one CL molecule; Refs. 12
, 41
), and cells emit in the red channel. Simultaneously, while increasing the red fluorescence, they lose the capacity to emit in the green channel (as the number of NAO monomers decreases). The curve of red emission shows a peculiar phenomenon related to the dose of dye and its distribution into the mitochondrial inner membrane. Indeed, when NAO dimers occupy all of the possible residues of CL that are present on the outer leaflet of mitochondrial inner membrane, a plateau in fluorescence emission is reached. Then, the interactions between NAO and CL induce modifications of the inner membrane permeability so that the dye may cross the membrane and may have free access to the CL present in the matrix-side, with the subsequent formation of other dimers, and an increase in red fluorescence. Finally, when also the inner leaflet is saturated, i.e., when all of the phosphate residues of CL have bound NAO, a new and final plateau is reached. The first plateau indicates the saturation of the CL residues exposed on the outer leaflet, the second the saturation of all of the intramitochondrial CL, i.e., the maximal fluorescence. The ratio between the fluorescence intensity of the first plateau and that of the second corresponds to the percentage of CL present on the cytoplasmic (outer) face of mitochondrial inner membrane.
For NAO staining, cells were fixed in 1% formaldehyde (in PBS) for 15 min at room temperature, then washed twice in PBS and adjusted to a concentration of 1 x 106 cells/ml in PBS. Increasing amounts of NAO (0.135 µM) were added, and cells were kept at room temperature for 15 min, washed twice with cold PBS, and resuspended in PBS and analyzed. The red and green fluorescence emission intensity was plotted as a function of the amount of NAO present in the incubation mixture; however, only the red fluorescence was used for the analysis of CL distribution. Samples stained with scalar amounts of NAO were acquired in linear scale, and the median fluorescence values from the resulting histograms were then used to calculate the intensity of the fluorescence emission. Median values were then plotted and analyzed using GraFit v.3.0 software; each plateau was mathematically determined by the software using the second derivative (i.e., the rate of change of the rate of change) of the fitting curve, which indicates the inflection point.
Determination of ROS.
Production of intracellular levels of peroxides was detected with the fluorescent probe CFH2, which is an uncharged, cell-permeant molecule. Inside cells, this probe is cleaved by nonspecific esterases to form carboxydichlorofluorescein that is oxidized in the presence of peroxide (42)
; the resulting compound has green fluorescence. Before the end of the incubation period, cells were labeled with 1 µM CFH2 for 1 h at 37°C, washed, and immediately analyzed.
To measure intracellular superoxide anion levels, another fluorescent probe, DH2, was used. DH2 freely enters cells, where it can be directly oxidized by superoxide anion to ethidium bromide. The trapped, nonpermeable ethidium bromide can then intercalate into DNA to become fluorescent (orange/red; Ref. 43 ). Before the end of the incubation period, cells were labeled with 2 µM DH2 for 15 min at 37°C, washed, and immediately analyzed.
Detection of Apoptosis and 
.
Apoptosis was detected by two diverse methods able to monitor the phenomenon at different stages: (a) the exposure of phosphatidylserine on the plasma membrane, an early event of apoptosis, by FITC-conjugated annexin V labeling; and (b) the condensation and loss of fragmented chromatin, a late event, by PI labeling. Concerning the first method, cells were resuspended in 200 µl of binding buffer [10 mM HEPES/NaOH (pH 7.4), 140 M NaCl, and 2.5 mM CaCl2], incubated in the dark with 1 µl of annexin V for 10 min at room temperature, washed, and analyzed in the same binding buffer (44)
. Briefly, for PI labeling, the cell pellet was gently resuspended in 1 ml of hypotonic fluorochrome solution (0.1% sodium citrate, 0.1% Triton X-100, and 50 µg/ml PI) and kept 30 min at 4°C before analysis (45)
.

was analyzed using the lipophilic cation JC-1. Briefly, cells were treated as described above and, at the end of the incubation period, stained with the 
-sensitive probe JC-1 (2.5 µg/ml) in RPMI 1640 containing 10% FCS for 10 min at room temperature in the dark as described previously (46
, 47)
.
Flow Cytometry and Data Analysis.
Cytofluorimetric analyses were performed using a FACScan cytometer (Becton Dickinson, San José, CA) equipped with an argon-ion laser tuned to 488 nm, as reported previously (48)
. Green fluorescence from NAO, CFH2, JC-1 monomers, and annexin V was detected through the standard band-pass filter centered at 520 ± 10 nm, orange fluorescence from PI and JC-1 aggregates at 575 ± 10 nm, and red fluorescence from NAO and DH2 through the long pass filter (630 ± 15 nm). A standard cytogram based on the measurement of right-angle scatter versus forward-angle scatter was designed to eliminate cellular debris and aggregates.
Statistical Analysis.
Data are given as mean ± SE and represent a minimum of four experiments. To analyze the homogeneity among groups, the Kruskall-Wallis test was used, followed by multiple post hoc comparisons using the Mann-Whitney t test with Bonferroni adjustment. Any P < 0.05 was considered statistically significant. Calculations were performed with SPSS version 10.0 (SPSS, Chicago, IL) under Windows ME.
| Acknowledgments |
|---|
| Footnotes |
|---|
1 This work has been supported by grants from Istituto Superiore di SanitàProgramma Nazionale di Ricerca sullAIDS 2001 (Progetto Patologia, Clinica e Terapia dellAIDS), from Murst Cofin 2000 (to A. C.), the Foundation for Polish-German Cooperation (Warsaw), Wellcome Trust (London), and The Polish State Committee for Science (Warsaw) (to J. D.). L. M. was supported by a fellowship from Associazione Angela Serra, Modena (Italy). ![]()
2 M. G. F. and L. T. have equally contributed to the paper and are considered first authors. ![]()
3 To whom requests for reprints should be addressed, at Chair of Immunology, Department of Biomedical Sciences, University of Modena and Reggio Emilia, via Campi 287, 41100 Modena, Italy. Phone: 39-059-205-5415; Fax: 39-059-205-5426; E-mail: cossariz{at}unimo.it ![]()
4 The abbreviations used are: cyt c, cytochrome c; ROS, reactive oxygen species; CL, cardiolipin; TNF, tumor necrosis factor; NAO, nonyl acridine orange; STS, staurosporine; Z-VAD.fmk, Z-Val-Asp-fluoromethylketone; BHT, butyl hydroxytoluene; 
, mitochondrial membrane potential; CFH2, 6-carboxy-2',7'-dichlorodihydro-fluorescein diacetate,di(acetoxymethyl ester); DH2, hydroethidine; JC-1, 5,5',6,6'-tetrachloro-1,1',3,3'-tetraethylbenzimidazolylcarbocyanine iodide; CHX, cycloheximide; PI, propidium iodide. ![]()
5 S. Salvioli, G. Storci, M. Pinti, D. Quaglino, L. Moretti, M. Merlo-Pich, G. Lenaz, S. Filona, A. Fico, M. Bonafé, D. Monti, L. Troiano, M. Nasi, A. Cossarizza, and C. Franceschi. Apoptosis-resistant phenotype in HL-60-derived cells HCW-2 is related to changes in expression of stress-induced proteins that impact on redux status and mitochondrial metabolism, in press. ![]()
Received for publication 2/12/02. Revision received 6/17/02. Accepted for publication 6/24/02.
| References |
|---|
|
|
|---|
-induced apoptosis in U937 cells: the role of mitochondria. Exp. Cell Res., 220: 232-240, 1995.[Medline]
This article has been cited by other articles:
![]() |
K. Gillick and M. Crompton Evaluating cytochrome c diffusion in the intermembrane spaces of mitochondria during cytochrome c release J. Cell Sci., March 1, 2008; 121(5): 618 - 626. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. A. Parone, D. I. James, S. Da Cruz, Y. Mattenberger, O. Donze, F. Barja, and J.-C. Martinou Inhibiting the mitochondrial fission machinery does not prevent bax/bak-dependent apoptosis. Mol. Cell. Biol., October 1, 2006; 26(20): 7397 - 7408. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. Y. Tyurina, V. Kini, V. A. Tyurin, I. I. Vlasova, J. Jiang, A. A. Kapralov, N. A. Belikova, J. C. Yalowich, I. V. Kurnikov, and V. E. Kagan Mechanisms of Cardiolipin Oxidation by Cytochrome c: Relevance to Pro- and Antiapoptotic Functions of Etoposide Mol. Pharmacol., August 1, 2006; 70(2): 706 - 717. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Karawajew, P. Rhein, G. Czerwony, and W.-D. Ludwig Stress-induced activation of the p53 tumor suppressor in leukemia cells and normal lymphocytes requires mitochondrial activity and reactive oxygen species Blood, June 15, 2005; 105(12): 4767 - 4775. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. C. Sparagna, C. A. Johnson, S. A. McCune, R. L. Moore, and R. C. Murphy Quantitation of cardiolipin molecular species in spontaneously hypertensive heart failure rats using electrospray ionization mass spectrometry J. Lipid Res., June 1, 2005; 46(6): 1196 - 1204. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Lucken-Ardjomande and J.-C. Martinou Newcomers in the process of mitochondrial permeabilization J. Cell Sci., February 1, 2005; 118(3): 473 - 483. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Petrosillo, F. M. Ruggiero, M. Pistolese, and G. Paradies Ca2+-induced Reactive Oxygen Species Production Promotes Cytochrome c Release from Rat Liver Mitochondria via Mitochondrial Permeability Transition (MPT)-dependent and MPT-independent Mechanisms: ROLE OF CARDIOLIPIN J. Biol. Chem., December 17, 2004; 279(51): 53103 - 53108. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Sugioka, S. Shimizu, and Y. Tsujimoto Fzo1, a Protein Involved in Mitochondrial Fusion, Inhibits Apoptosis J. Biol. Chem., December 10, 2004; 279(50): 52726 - 52734. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. PETROSILLO, F. M. RUGGIERO, and G. PARADIES Role of reactive oxygen species and cardiolipin in the release of cytochrome c from mitochondria FASEB J, December 1, 2003; 17(15): 2202 - 2208. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Liu, Q. Dai, J. Chen, D. Durrant, A. Freeman, T. Liu, D. Grossman, and R. M. Lee Phospholipid Scramblase 3 Controls Mitochondrial Structure, Function, and Apoptotic Response Mol. Cancer Res., October 1, 2003; 1(12): 892 - 902. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cell Growth & Differentiation |