- Email: [email protected]
The differentiation of normal and transformed human fibroblasts in vitro is influenced by electromagnetic fields
Experimental Cell Research 182 (1989) 610-621
The Differentiation of Normal and Transformed Human Fibroblasts in Vitro Is Influenced by Electromagnetic Fields H. PETER RODEMANN,**’ KLAUS BAYREUTHER,*‘* and GERHARD PFLEIDERERt *Institut fiir Genetik, Universitht Hohenheim, D 7tMOStuttgart 70, and tlnstitut fir Biochemie, Universitiit Stuttgart, D 7000 Stuttgart 80, Federal Republic of Germany
We studied the effect of symmetric, biphasic sinusoidal electromagnetic fields (EMF) (20 Hz, 6 mT) on the differentiation of normal human skin tibroblasts (HH-8), normal human lung fibroblasts (WI38), and SV40-transformed human lung fibroblasts (WI38SV40) in in vitro cultures. Cells were exposed up to 21 days for 2x6 h per day to EMF. Normal mitotic human skin and lung fibroblasts could be induced to differentiate into postmitotic cells upon exposure to EMF. Concomitantly, the synthesis of total collagen as well as total cellular protein increased significantly by a factor of 5-l 3 in EMF-induced postmitotic cells. As analyzed by two-dimensional gel electrophoresis of [3SS]methionine-labeled polypeptides, EMF-induced postmitotic cells express the same differentiation-dependent and cell type-specific marker proteins as their spontaneously arising counterparts. In SV40transformed human lung fibroblasts (cell line WI38SV40) the exposure to EMF induced the differentiation of mitotic WI38SV40 cells into postmitotic and degenerating cells in subpopulations of WI38SV40 cell cultures. Other subpopulations of WI38SV40 cells did not show any effect of EMF on cell proliferation and differentiation. These results indicate- that long-term EMF exposure of fibroblasts in vitro induces the differentiation of mitotic to postmitotic cells that are characterized by differentiation-specific proteins and diierentiation-dependent enhanced metabolic activities. @I 1989 Academic RCSS, IW.
Various studies have shown that electromagnetic fields (EMF)3 can modify the behavior of cells in culture, e.g., macrophages [l], neural crest cells [2], and fibroblasts [3]. Clinical studies have demonstrated significant improvements in, for instance, wound healing and bone fracture reunion [4-71 by sinusoidal as well as pulsed electromagnetic fields. Studies into the mechanisms of the effects of EMF on cultured cells have been restricted to short-term studies (2-7 days of exposure) measuring various biochemical parameters like DNA and collagen synthesis and glycosylation of proteins in cultured chicken, rabbit, and human fibroblasts [8-121. Human skin fibroblasts in uiuo and in vitro have recently been demonstrated to be differentiating cells with distinct morphological and biochemical properties ’ Present address: Lehrstuhl fur Entwicklungsbiologie, Universitiit Bielefeld, D 4800 Bielefeld 1, FXG. * To whom reprint requests should be addressed. ’ Abbreviations used: MF, mitotic flbroblast; PMF, postmitotic fibroblast; CPDL, cumulative population doublings level; DMEM, Dulbecco’s modified Eagle’s medium; EMF, electromagnetic fields. Copyright @ 1989 by Academic Press, Inc AU rights of reproduction in any form reserved 0014-4827/89 $03.00
610
Fibroblast digerentiation in EMF
611
[ 11, 121.Three mitotic (MF I, MF II, MF III) and four postmitotic (PMF IV, PMF V, PMF VI, PMF VII) flbroblast types can be characterized by morphological and biochemical markers [ll] differentiating along the terminal cell lineage MF I-MF II-MF III-PMF IV-PMF V-PMF VI-PMF VII in uivo and in vitro [ 13, 141. In the present study we used the well-defined human skin and lung fibroblast cell system to analyze the effects of biphasic, sinusoidal EMF on the differentiation behavior of normal and transformed human fibroblasts. Athermal, biphasic, sinusoidal electromagnetic fields have first been described by Kraus et al. to be very effective for the treatment of traumatic and pathological lesions of bone 14, 5, G-181. We will provide morphological and biochemical evidence that sinusoidal EMF induce the differentiation of mitotic to postmitotic normal human fibroblasts and the transition of dividing to nondividing and subsequently degenerating transformed human lung fibroblasts. MATERIAL
AND METHODS
Electromagnetic fields. A Magnetodyn function generator was used to drive six solenoid coils, each of which could take up six petri dishes. The experimental set-up and the spatial distribution of the magnetic field in the coils are shown in Fig. 1. Two signal shapes were used to generate the electromagnetic fields: (A) a continuously sinusoidal signal with a frequency of 20 Hz and a maximum magnetic induction of 8.4 mT=6 mT& (signal shape 1 in Fig. 2); (B) an asymmetric pulsed monophase signal with a pulse frequency of 8O/min and a maximum magnetic induction of 8.4 mT (signal shape 2 in Fig 2). The distortion factor was less than 1% for signal shape 1, as measured with a gaussmeter Bell 646 and a spectrum analyzer HP 3582 A. These signal shapes correspond to those used clinically to improve fracture and wound healing [5]. The coils were installed in a hum&fed 95 % air : 5 % COrincubator at 37°C and were prevented from increasing the incubation temperature by a Haake cooling system with control of the temperature in the nutrient solution of the cultures (37f0.3”C). Cell culrures. Normal human skin fibroblasts of the cell line HH-8 were cultured as previously described [13]. Normal human lung fibroblasts of the cell line WI38 and their SV40-transformed counterparts, WI38SV40, were cultured according to published procedures [19]. For each experiment normal human skin and lung fibroblasts were used at a cumulative population doublings level (CPDL) of 30-34. At this CPDL fibroblast populations consist predominantly of the mitotic fibroblast cell type MF II [13]. Bansformed human lung fibroblasts were studied at CPDL 130-140 (passage level 40-50). Cells were seeded at densities of 5x1@ or 5xld cells per 50 cm’ and incubated in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal calf serum (fcs) and antibiotics [13]. Cells were incubated in humiditied 95% air: 5% CO*. Medium was changed once each week. Cells were exposed for 2x6 h treatments per day to EMF (signal shape I, Fig. 2). To ensure an equal average intensity of EMF (6 mT) in all six culture dishes per coil the dishes were numbered and position of the culture dishes was changed every day routinely. Parallel control cell cultures of the same CPD were incubated under identical culture conditions in a separate CO*-incubator of the same type (Heraeus COr-autozero) without exposure to EMF. After 7, 14, or 21 days of incubation the cell number, morphology and various biochemical parameters, i.e., total protein synthesis, total collagen synthesis and [3’S]methionine-labeled polypeptide pattern, were studied. Cell ?ype frequency and cell number. Cell cultures were washed with phosphate-buffered saline (PBS) and were subsequently fixed with 3.5% paraformaldehyde and 70% ethanol (10 min each) and stained for 30 s with 0.1% Coomassie blue and 30 min with Giemsa solution as described elsewhere 1131.For normal human skin and lung flbroblasts the frequencies of the mitotic cell types MF I, MF II, and MF III and for the postmitotic fibroblast cell types PMF IV, PMF V, and PMF VI were determined by morphological criteria [13] counting at least 2000 cells in six paralIe1 experimental series. In Table 1 only the changes of the predominant cell types MF II and PMF VI are listed. For transformed human lung fibroblasts the frequency of mitotic flbroblasts (MF) and postmitotic fibroblasts (PMF) were determined by morphological criteria. For all cell cultures the number of cells-per 50 cm’ was determined by trypsinizing the cells and counting them in a Fuchs-Rosenthal hemocytometer 1131.
612 Rodemann, Bayreuther, and Pfleiderer
in nutrlont mdutfonz 3t’rD.$c cl+-‘ncutIotor 6ohdd cdl P6 Tqcontrd
Potrl6lmlmc
&mldd
toll
PO
Fig. 1. Experimental set-up and spatial distribution of the magnetic field in solenoid coils P6. Six solenoid coils P6 (designed to take up six IO-cm tissue culture dishes per coil) were installed in a Heraeus AutoZero CO,-incubator. The incubation system was prevented from increasing the incubation temperature by a water driven Haake cooling system (C.S.). Temperature of the nutrient solution was controlled over the 21-day incubation period. Electromagnetic fields were generated by a function generator M70S (F.G.M70S).
Total prorein synthesis. Cells wer@@eled for 6 h with 1 @i/ml [‘Hlleucine (New England Nuclear, 5 mCi/mmol) in DMEM supplement&vith W% fcs. Radioactivity incorporated into cellular proteins was determined as described elsewhere [ZO]. Toral collagen synthesis. Cells were labeled for 18 h with 4 @ml [‘Hlproline (New England
Fibroblast differentiation in EMF 613
Signal
shape 1
A
Signal shape
2
Fig. 2. Signal shapes and physical properties of the electromagnetic fields. -Signal- shape I: ._ _ _ -.. . .. _ . symmetric, lslpham sinusoidal electromagnetic tieids; maximum maguetic induction 8.4 mT=6 mT&.
Signal shape 2: asymmetric pulsed, monophasic electromagnetic fields; maximum magnetic induction 8.5 mT.
614 Rodemann, Bayreuther, and Pfleiderer TABLE 1 EMF-induced changes in the cell type frequency of mitoticfibroblasts MF ZZand postmitoticfibroblasts PMF VI in normal human skin and lungfibroblast populations Incubation time (days)
starting %MFII
7 Days
14 Days
21 Days
%MFII
%PMFVI
%MFII
%PMFVI
%MFII
%PMFVI
Human skin fibroblasts CP 82.7~k7.9 EMF-P 82.9k7.8
83.6k7.1 53.2k6.3
O.lfO.O1 25.6k4.7
81.6k5.9 7.2fl.l
O.lkO.03 59.7f6.2
81.9k7.2 0
0.2kO.03 85.4f5.9
Human lung tibroblasts CP 86.2k8.1 EMF-P 86.4k8.2
85.3k7.2 73.2k4.7
0.2f0.03 19.7f3.9
86.126.6 9.8k2.4
0.3f0.02 48.3k5.9
84.7k5.6 0
0.3f0.01 84X3+8.4
Note. Normal human skin fibroblast cell line HH-8, CPDL 3&32. Normal human lung fibroblast cell line W138, CPDL 32-34. Numbers shown represent the mean values + SD of six parallel experiments. At each time point 2000 cells were analyzed and the percentage of MF II and PMF VI was listed. CP, control populations; EMF-P, EMF-exposed populations.
Nuclear, 40 Ci/mmol) in serum-free DMEM supplemented with 0.1 mg/ml sodium-ascorbate and 0.1 mgknl @ninoproprionitrile. Incorporation of [3H]proline into total extra- and intracellular pepsinresistant material consisting at least of 80% collagen was analyzed according to standard procedures 121,221. [3’S]Methionine-labeledpolypeptide pattern. Cells were labeled for 18 h with 200 @i/ml [“Slmethionine (New England Nuclear, 400 Ci/mmol) in DMEM supplemented with 10% fcs as described [23, 241. After labeling the cells were lysed [23, 241 and radioactivity was determined. Aliquots of the various samples containing 500,000 cpm were analyzed by two-dimensional polyacryiamide gel electrophoresis [23].
RESULTS Normal Fibroblasts Cell type frequency in normal human skin and lung jibroblasts. The frequencies of mitotic fibroblasts MF II (Figs. 3A and B) and postmitotic fibroblasts
PMF VI (Figs. 3 C and D) in HH-8 and WI38 cell populations were analyzed at Days 7, 14, and 21 of incubation with and without exposure to biphasic, sinusoidal EMF (signal shape I, Fig. 2). Table 1 shows that normal human skin fibroblast populations of the cell line HH-8 which initially comprised 82.9% MF II type cells differentiate upon exposure to EMF to populations comprising 85.4% PMF VI at Day 21. Likewise normal human lung fibroblast populations of the cell line WI38 show similar changesin the cell type composition of populations exposed to EMF (Table 1). Control populations show no change in the frequencies of MF II and PMF VI during the 21 days of incubation (Table 1). Total protein and collagen synthesis. Measurements of the total protein and collagen synthesis in control populations of HH-8 and WI38 tibroblasts showed a
Fibroblast dijerentiation
in EMF
615
Fig. 3. Mitotic fibroblasts MF II of normal human skin libroblasts, cell line HH-8 (A), and of normal human hmg fiboblasts, cell line WI38 (B) at Day 21 of control populations. EMF-induced postmitotic flbroblasts PMF VI of the cell line HH-8 (C) and WI38 (0) after 21 days of exposure to EMF. Cells were seeded at an identical density of 5x10 cells per 50 cm2 before EMF treament or incubation under control conditions. During the 21day incubation period, untreated control cell populations underwent 2.1-2.9 CPD before reaching confluency. EMF-exposed cell populations underwent during the first 2 weeks of EMPtreatment a total of 0.243 CPD before all cells were arrested in the irreversible nondividing, postmitotic differentiation state. Magnification for Fig. A, B, C, D: x 100.
significant decrease over the incubation period of 21 days reflecting the downregulation of these processes as a function of increasing contact regulation. Skin and lung fibroblasts populations exposed to sinusoidal EMF, however, showed a significant increase in the synthesis of total protein and total collagen. This reflects the differentiation-dependent increase in the capacity for the synthesis of macromolecules in PMF VI type cells (Table 2) [25, 261. [35S]Methionine polypeptide pattern. Normal human skin tibroblasts exposed to EMF for 21 days show significant changes in the two-dimensional polypeptide pattern of total soluble cellular proteins (Fig. 4). Cell populations exposed to EMF express protein spots (Fig. 4B) that are not present in control populations (Fig. 4A), but that have been demonstrated to be specific for postmitotic fibroblasts PMF VI (Fig. 4B) [13]. Similar changes in the protein pattern between mitotic and EMF-induced postmitotic cells could be observed using the human lung tibroblasts cell line WI38 (data not shown). 40-898336
616 Rodermann, Bayreuther, and Pfleiderer
a
Fig. 4. [%]Methionine poIypeptide pattern of normal human skin fibroblasts cell line HIi-8. (A) Control population at Day 21 of the experiment; (B) EMF-exposed population at Day 21 of the experiment. Protein spots indicated by arrowheads are cell type-specific proteins of the fibroblast type PMF VI. IEF, isoelectric focusing, acidic end to the right. M,, molecular weight.
Transformed Fibroblasts Frequency of mitotic and postmitotic transformed human lungfibroblasts.
Cell
populations at two initial seeding densities (5 x 104or 5x 10scells per 50 cm’) were used to analyze the effects of biphasic, sinusoidal EMF on the differentiation of the transformed human lung fibroblast cell line WI38SV40. As shown in Table 3 control populations at either cell density showed no significant change in the frequencies of mitotic or postmitotic cells during the 21 days of incubation. Cell populations with an initial density of 5x 10’ cells could be analyzed only up to 14 days of culture. Longer incubation periods without passagingthe cells resulted in spontaneousdegeneration of the cell populations. ‘Ikansformedlung fibroblasts (initial density 5 x lo4 cells) exposed to sinusoidal EMF for up to 21 days showed significant changes in the frequency of mitotic and postmitotic cells. After 21 days of exposure to EMF, WI38SV40 cell populations were composed of 71% postmitotic nondividing cells (Table 3). However, in contrast to their nontransformed counterparts (WI38 cells, Table 1) transformed lung fibroblasts cannot be induced completely to postmitotic, nondividing cells by EMF exposure. After 21 days 29% of the living cells were still of the mitotic cell type and were actively dividing. In cell populations of an initial density of 5 x 10’ cells per 50 cm* this effect was
Fibroblast diflerentiation
in EMF
617
TABLE 2 EMF-induced changes in total protein and collagen synthesis in normal human skin and lung fibroblasts Incubation time Initial cell density/SO cm2
7 Days
Protein synthesis (cpm 13H]Leu incorporation/id Normal human skin fibroblasts CP 5x10’ EMF-P 5x10’
CP EMF-P
21 Days
cells] 23.2k6.3 91.4k12.9’ (x3.93)
19.1f3.6 176.7+20.8* (x9.25)
34.1k2.7 90.2f6.8* (x2.64) 5x16 24.4f5.1 13.7f1.7 26.7f5.3 5Xld 51.8+8.4* (x3.78) Collagen synthesis (cpm TH]Pro incorporation/l@ cells)
37.8f4.1 190.1+10.4* (X5.03) 13.8k1.8 132.6+ 14.2* (x9.61)
Normal human lung fibroblasts CP 5x1@ EMF-P 5x10’
Normal human skin fibroblasts CP 5x104 EMF-P 5x10’ Normal human lung fibroblasts CP 5x10’ EMF-P 5x 10’
28.5k5.7 33.7f8.9 (x1.18)
14 Days
34.1k7.1 33.2f5.8
42.7k6.3 71.9f8.9* (X1.68)
14.6k1.6 103.5+9.7* (x7.09)
8.5kO.9 118.9+11.8* (x13.98)
39.7k6.3 87.4+9.6* (X2.2)
15.9f1.8 143.4*12.6* (X9.01)
13.6k1.9 180.4+15.9* (x 13.26)
* ZWO.05, n=lB. Note. Normal human skin iibroblast cell line HH-8, CPDL 30-32. Normal human lung fibroblast cell line W138, CPDL 32-34. Numbers shown are the means + SD of six parallel experiments with two subsamples in each group. For statistical analysis, data were compared by a generalized block design combined with the Dunnett test [31]. CP, control populations; EMF-P,EMF-exposed populations.
much more pronounced. As shown in Table 3 after 14 days of EMF exposure 15% of the living cells were of postmitotic cell type, as compared with 83 % mitotic and actively dividing cells. As tested by the fluorescein diacetate/ethidium bromide assay [27], the exposure of transformed lung fibroblasts to EMF results not only in a differentiation of subpopulations to postmitotic cells but also in a degeneration of other subpopulations of these cells. After 14 to 21 days the percentage of dead cells was between 20 and 48%, depending on the initial cell density of the cultures (data not shown). Total protein and collagen synthesis. As demonstrated in Table 4 control cell populations of transformed lung fibroblasts (initial cell density 5X lo4 cells) showed comparable rates of protein and collagen synthesis as their non-transformed counterparts (Table 2). Cells exposed to EMF for 21 days showed parallel to the increasing number of postmitotic cell types a significant increase in total
618 Rodemann, Bayreuther, and Pfleiderer TABLE 3 EMF-induced changes in the frequency of mitotic and postmitotic cell types of transformed human lung fibroblast populations Incubation time 7 Days
14 Days
21 Days
Initial cell density150 cm’
%MF
% PMF
%MF
% PMF
%MF
% PMF
CP EMF-P
5x104 5x104
100 91.1f3.4
0 8.9f3.4
99.7kO.l 52.Ok3.1
0.3fO.l 48.Ok3.1
99.1f0.2 29.1k3.2
0.9f0.2 70.9k3.2
CP EMF-P
5x10’ 5XlOJ
100 95.2k2.6
0 4.8k2.6
99.3+0.4 85.Ok5.1
0.7kO.4 15.Ok5.1
-
-
Note. SV&transformed human lung fibroblast cell line WI38SV40. Numbers shown represent the mean values f SD of six parallel experiments. At each time point cells were analyzed as de&bed under Methods and the percentage of mitotic cells (MF) and postmitotic cells (PMF) is listed. Two thousand cells were analyzed in each experimental series. CP, control populations; EMF-P, EMFexposed populations.
protein and collagen synthesis by a factor of 8-10 (Table 4). Likewise, transformed lung fibroblasts at an initial density of 5 x ld cklls upon exposure to EMF for 14days exhibited a significant increase in total protein synthesis by a factor of 6 (Table 4). TABLE 4 EMF-induced changes in total protein synthesis and collagen synthesis of transformed human lung fibroblast populations Incubation time _..-
Initial ceil density150 cm2
7 Days
Protein synthesis (cpm [ti]Leu incorporation/lO’ cells) 39.1f9.7 CP 5x104 EMF-P 5x104 31.3f16.0 CP EMF
5x105 5Xld
31.7k6.1 35.6f16.1
Collagen synthesis (cpm [‘HIPro incorporation/id cells) 5x104 53.7f6.9 CP EMF-P 5x 10’ 99.4k8.3” (X1.85)
14 Days
21 Days
24.3zb5.2 I98.lf61.4* k(X8.15)
34.1f2.1 257.5+72.8* (X7.55)
34.2+ 16.4 206.4+82.9* (x6.03) 32.1k4.8 125.6+ 10.6* (X3.91)
16.9f2.1 133.5+13.6* (X7.89)
* PcO.05, n=18. Note. SV40-transformed human lung fibroblast cell line WI38SV40. Numbers shown are the means + SD of six parallel experiments with three subsamples in each group. For statistical analysis, data were compared by a generalized block design combined with the Dunnett test [31]. CP, control populations; EMF-P, EMF-exposed populations.
Fibroblast differentiation in EMF 619 [35S]Methioninepolypeptide pattern. A comparison of the polypeptide pattern of control WI38SV40 cells and cells which had been exposed to EMF for 21 days revealed sign&ant changes in a variety of proteins. After 21 days of exposure to EMF these cells expressed at least 11 proteins that are not present in control populations after 21 days of incubation (data not shown). DISCUSSION The results of this study suggest that in vitro exposure of normal and transformed human fibroblast population to biphasic, sinusoidal EMF (20 Hz, 6 mT) stimulates the differentiation of mitotic fibroblast cell types to postmitotic tibroblast cell types that are characterized by significantly enhanced metabolic activities. Normal fibroblasts from various organs of vertebrate and mammalian species including man have recently been demonstrated to be cells which differentiate along a terminal cell lineage representing the morphologically recognizable compartments of the fibroblast stem cell system [ 13, 14,21,25,28]. Independently of the species or organs analyzed fibroblast populations are composed of three mitotic cell types (MF I, MF II, MF III) and four postmitotic cell types (PMF IV, PMF V, PMF VI, PMF VII) 113, 14, 291. In routinely used secondary fibroblast populations the frequencies of the cell types MF I, MF II, and MF III change as a function of the CPDL [13, 141. Similarly in spontaneously arising postmitotic fibroblast cultures the frequencies of PMF IV, PMF V, PMF VI, and PMF VII are a function of the duration of the stationary postmitotic culture [131.Morphological and biochemical data revealed that the mitotic cell type MF II is the predominant cell type in secondary cultures of early and middle passage(CPDL 2040) cell populations [13]. PMF VI represents the predominant and biosynthetitally most active postmitotic and terminally differentiated cell type of the fibroblast stem cell systemJl3, 261. As demonstrated by morphological and biochemical criteria sinusoidal EMF (20 Hz, 6 mT) induce the differentiation of both normal human skin and lung fibroblast MF II populations to PMF VI populations within 21 days. Concomitantly the physiological activity, i.e., total protein and collagen synthesis, increased significantly as a function of the differentiation state of the populations analyzed. The terminal differentiation state of the fibroblasts after 21 days of exposure to EMF is also demonstrated by the expression of six new proteins which have recently been described to be specific for postmitotic tibroblasts [ 13, 301. Recently Goodman and Henderson have also described qualitative and quantitative changes in patterns of protein expression of dipteran salivary gland cells in culture upon exposure to EMF [32]. Similar results of a stimulation of collagen synthesis in bone marrow fibroblasts after exposure to short-term pulsed electromagnetic fields (PEMF) have been described by Famdale and Murray [9]. From these studies 191,however, it is unclear whether the stimulation of collagen synthesis was due to a stimulation of the synthetic capacity of mitotic cells or was due to the PEMF-induced postmitotic differentiation or modulation of the cell populations used.
620 Rodemann, Bayreuther, and Pjleiderer
Long-term treatment with asymmetric pulsed monophasic EMF (signal shape 2, Fig. 2) (21 days, 2 x 6 h per day) induced in subpopuh+tionsof normal human skin and normal human lung fibroblast cell lines (HHS, WI38) and in transformed human lung fibroblasts (WI38SV40) a modulation into reversibly nondividing cells. The modulated cells increase in size, show an enhanced protein synthesis, but do not undergo endopolyploidization. Upon termination of the EMF treatment the modulated cells revert to normal size and function and regain the capacity for cell divisions (Rodemann et al., in preparation). As recently demonstrated by the use of two-dimensional gel electrophoresis 1291spontaneously transformed BN-rat fibroblasts can differentiate spontaneously at low frequencies into nondividing, irreversible postmitotic cells. As demonstrated by morphological and biochemical criteria WI38SV40 cells can be induced by biphasic sinusoidal EMF to differentiate into irreversibly postmitotic and subsequently degenerating cells. However, as compared to normal lung fibroblasts the frequency of the EMF-induced postmitotic cells was significantly lower in transformed lung fibroblast populations and was furthermore a function of the initial cell density. After 21 days of exposure to EMF actively dividing mitotic fibroblast cell types were present in different numbers in cultures of both initial cell densities tested. These data indicate that mitotic mass cultures of WI38SV40 cells are composed of subpopulations that are either sensitive or completely insensitive to the EMF-treatment regime used. These results therefore make the discussion of the relevance of EMF treatment for the developmentof strategies in cancer therapy impossible at the present time. Taken as a whole, the data presented in this study suggestthat the exposure of normal human skin and lung fibroblasts in vitro to sinusoidal EMF induces the differentiation of mitotic to irreversibly postmitotic fibroblasts. The arising postmitotic fibroblasts are significantly more active with regard to total protein and collagen synthesis than the corresponding mitotic fibroblast pd&ilations. This finding is of particular interest for determining the cellular mechanismsby which biphasic sinusoidal EMF induce changes in the behavior and function of secondary fibroblast populations in vitro. Since primary and secondary skin fibroblasts of man differentiate along the same terminal cell lineage [14] it can be expected that fibroblasts in uiuo show similar or identical EMF-induced changes of cellular function and behavior, e.g., during wound healing and bone fracture reunion. We are indebted to the help and advice of W. Kraus, Institut fib Medizinische Physik, Munchen, F.R.G., and for installing the experimental set-up. We are grateful to Drs. G. MoIineux and T. M. Dexter, Paterson Institute for Cancer Research, Christie Hospital & Holdt Radium Institute, Manchester, UK, for critically reading the manuscript. This work was supported in part by the Deutsche Forschungsgemeinschaft (Ba 526/5 and SFB 223), the Fritz-Thyssen-Stiftung, and the Breuninger Stiftung. H. P. R. is a recipient of a Heisenberg grant from the Deutsche Forschungsgemeinschaft (Ro 52712-l).
REFERENCES 1. Grida, N., and Feldman, J. D. (1982) Cell Motil. 8, 243. 2. Nuccitelli, R., and Erickson, C. A. (1983) Exp. Cell Res. 147, 195. 3. Stump, R. F., and Robinson, K. R. (1983) Cell Biol. 97, 1226.
Fibroblast differentiation in EMF
621
4. Kraus, W., and Lechner, F. (1972)Miinch. Med. Wochenschr. 114, 42, 1814. 5. Kraus, W. (1984)Orthopiide 13,78. 6. Bassett,C. A. (1978)in Metabolic Surgery(Buchwald, H., and Varco, R. L., Eds.), p. 255,Grune
& strattoll, NY. 7. Watson,J., and Dowries, E. M. (1978)Japan J. Appl. Phys. 17, 215. a. Takahashi,K., Kaneko, I., Date, M., and Fukada, E. (1986)Experientiu 42, 185. 9. Famdale, R. W., and Murray, J. C. (1985)Cal& Tissue Inf. 37, 178. 10. L&off, A. R., Williams, T., Strong, D. M., and Wistar, R. (1984)Science 223, 1984. 11. Fisher, S. J., and Dulling, J. (1986)J. Bioelectr. 5, 253. 12. Lee, R. C., Rich, J. B., Kelley, K. M., Weismann,D. S., and Mathews,M. B. (1983)Amer. Surg. 48,567.
13. Bayreuther, K., Rodemann,H. P., Hommel, R., Dittmann, K., Albiez, M., and Francz, P. I. (1988)Proc. Natl. Acad. Sci. USA 84, 5112. 14. Bayreuther, K., Rodemann,H. P., Francz, P. I., and Maier, K. (1988)J. Cell Sci., in press. 15. Lechner, F., and Ascherl, R. (1978)Acto Orthop. Be/g. 44, 699. 16. Ascherl, R., Bltlmel, G., Lechner, F., and Kraus, W. (1985)Bioelectrochemistryand Bioenergetits 14, 161. 17. Lechner, F. (1972)Ifefle 2. U&Uheifkd. 114, 325. 18. Kraus, W. (1979)2nd Meeting of the European Society of Biomechanics.Strasbourg.[Abstract] 19. Hayflick, L.. and Moorehead, P. S. (l%l) Exp. Cell Res. 28, 585. 20. Rodemann,H. P., Bayreuther, K., and Schumacher,E. (1987)Biochem. Biophys. Res. Commun. 14&l. 21. Rodemann,H. P., and Bayreuther, K. (1984) Proc. Nutl. Acad. Sci. USA 81, 5130. 22. Mollenhauer, J., and Bayreuther, K. (1986)D&%rentiution 32, 165. 23. Bravo, R. (1984)in Two Dimensional Gel Electrophoresisof Proteins (Celis, J., and Bravo, R., Eds.), p. 4, Academic Press,New York. 24. Rodemann,H. P., and Bayreuther, K. (1986)Proc. Natf. Acad. Sci. USA 83, 2086. 25. Bayreuther, K., Rodemann,H. P., Kontermann, K., Mollenhauer, J., Francz, P. I., Dittmann, K., Mtlller, C., and Mtlller, G. (1987)Eur. J. Cell Biol. 43, 4. 26. Francz, P. I., Bayreuther, K., Rodemann,H. P., Hommel, R., and Dittmann, R. (1989)Different&ion, in press. 27. Rotman, B., and Papermaster,B. W. (1966)Proc. Natl. Acad. Sci. USA 55, 134. -~ 28. Kontermann, K., and Bayreuther, K. (1979)Gerontology 25, 261. 29. Francz, P. I., Kontermann, K., Mollenhauer, J., and Bayreuther, K. (1987)Anticancer Res. 7, 5A; Abstr. 168,921. 30. Rodemann,H. P., Bayreuther, K., Francz, P. I., Dittmann, &and Albiez, M. (1989)Exp. Cell Res. 1(10,84.
31. Neter, J., Wassermann,W., and Kutner, M. H. (1985)Applie&inea Statistical Models, 2nd ed., Richard D. Irwin Inc. 32. Goodman,R., and Henderson,A. S. (1988)Proc. Natl. Acaa. Sci. USA 85, 3928. ReceivedJuly 27, 1988 Revisedversion received December28, 1988
Printedin Sweden
The Differentiation of Normal and Transformed Human Fibroblasts in Vitro Is Influenced by Electromagnetic Fields H. PETER RODEMANN,**’ KLAUS BAYREUTHER,*‘* and GERHARD PFLEIDERERt *Institut fiir Genetik, Universitht Hohenheim, D 7tMOStuttgart 70, and tlnstitut fir Biochemie, Universitiit Stuttgart, D 7000 Stuttgart 80, Federal Republic of Germany
We studied the effect of symmetric, biphasic sinusoidal electromagnetic fields (EMF) (20 Hz, 6 mT) on the differentiation of normal human skin tibroblasts (HH-8), normal human lung fibroblasts (WI38), and SV40-transformed human lung fibroblasts (WI38SV40) in in vitro cultures. Cells were exposed up to 21 days for 2x6 h per day to EMF. Normal mitotic human skin and lung fibroblasts could be induced to differentiate into postmitotic cells upon exposure to EMF. Concomitantly, the synthesis of total collagen as well as total cellular protein increased significantly by a factor of 5-l 3 in EMF-induced postmitotic cells. As analyzed by two-dimensional gel electrophoresis of [3SS]methionine-labeled polypeptides, EMF-induced postmitotic cells express the same differentiation-dependent and cell type-specific marker proteins as their spontaneously arising counterparts. In SV40transformed human lung fibroblasts (cell line WI38SV40) the exposure to EMF induced the differentiation of mitotic WI38SV40 cells into postmitotic and degenerating cells in subpopulations of WI38SV40 cell cultures. Other subpopulations of WI38SV40 cells did not show any effect of EMF on cell proliferation and differentiation. These results indicate- that long-term EMF exposure of fibroblasts in vitro induces the differentiation of mitotic to postmitotic cells that are characterized by differentiation-specific proteins and diierentiation-dependent enhanced metabolic activities. @I 1989 Academic RCSS, IW.
Various studies have shown that electromagnetic fields (EMF)3 can modify the behavior of cells in culture, e.g., macrophages [l], neural crest cells [2], and fibroblasts [3]. Clinical studies have demonstrated significant improvements in, for instance, wound healing and bone fracture reunion [4-71 by sinusoidal as well as pulsed electromagnetic fields. Studies into the mechanisms of the effects of EMF on cultured cells have been restricted to short-term studies (2-7 days of exposure) measuring various biochemical parameters like DNA and collagen synthesis and glycosylation of proteins in cultured chicken, rabbit, and human fibroblasts [8-121. Human skin fibroblasts in uiuo and in vitro have recently been demonstrated to be differentiating cells with distinct morphological and biochemical properties ’ Present address: Lehrstuhl fur Entwicklungsbiologie, Universitiit Bielefeld, D 4800 Bielefeld 1, FXG. * To whom reprint requests should be addressed. ’ Abbreviations used: MF, mitotic flbroblast; PMF, postmitotic fibroblast; CPDL, cumulative population doublings level; DMEM, Dulbecco’s modified Eagle’s medium; EMF, electromagnetic fields. Copyright @ 1989 by Academic Press, Inc AU rights of reproduction in any form reserved 0014-4827/89 $03.00
610
Fibroblast digerentiation in EMF
611
[ 11, 121.Three mitotic (MF I, MF II, MF III) and four postmitotic (PMF IV, PMF V, PMF VI, PMF VII) flbroblast types can be characterized by morphological and biochemical markers [ll] differentiating along the terminal cell lineage MF I-MF II-MF III-PMF IV-PMF V-PMF VI-PMF VII in uivo and in vitro [ 13, 141. In the present study we used the well-defined human skin and lung fibroblast cell system to analyze the effects of biphasic, sinusoidal EMF on the differentiation behavior of normal and transformed human fibroblasts. Athermal, biphasic, sinusoidal electromagnetic fields have first been described by Kraus et al. to be very effective for the treatment of traumatic and pathological lesions of bone 14, 5, G-181. We will provide morphological and biochemical evidence that sinusoidal EMF induce the differentiation of mitotic to postmitotic normal human fibroblasts and the transition of dividing to nondividing and subsequently degenerating transformed human lung fibroblasts. MATERIAL
AND METHODS
Electromagnetic fields. A Magnetodyn function generator was used to drive six solenoid coils, each of which could take up six petri dishes. The experimental set-up and the spatial distribution of the magnetic field in the coils are shown in Fig. 1. Two signal shapes were used to generate the electromagnetic fields: (A) a continuously sinusoidal signal with a frequency of 20 Hz and a maximum magnetic induction of 8.4 mT=6 mT& (signal shape 1 in Fig. 2); (B) an asymmetric pulsed monophase signal with a pulse frequency of 8O/min and a maximum magnetic induction of 8.4 mT (signal shape 2 in Fig 2). The distortion factor was less than 1% for signal shape 1, as measured with a gaussmeter Bell 646 and a spectrum analyzer HP 3582 A. These signal shapes correspond to those used clinically to improve fracture and wound healing [5]. The coils were installed in a hum&fed 95 % air : 5 % COrincubator at 37°C and were prevented from increasing the incubation temperature by a Haake cooling system with control of the temperature in the nutrient solution of the cultures (37f0.3”C). Cell culrures. Normal human skin fibroblasts of the cell line HH-8 were cultured as previously described [13]. Normal human lung fibroblasts of the cell line WI38 and their SV40-transformed counterparts, WI38SV40, were cultured according to published procedures [19]. For each experiment normal human skin and lung fibroblasts were used at a cumulative population doublings level (CPDL) of 30-34. At this CPDL fibroblast populations consist predominantly of the mitotic fibroblast cell type MF II [13]. Bansformed human lung fibroblasts were studied at CPDL 130-140 (passage level 40-50). Cells were seeded at densities of 5x1@ or 5xld cells per 50 cm’ and incubated in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal calf serum (fcs) and antibiotics [13]. Cells were incubated in humiditied 95% air: 5% CO*. Medium was changed once each week. Cells were exposed for 2x6 h treatments per day to EMF (signal shape I, Fig. 2). To ensure an equal average intensity of EMF (6 mT) in all six culture dishes per coil the dishes were numbered and position of the culture dishes was changed every day routinely. Parallel control cell cultures of the same CPD were incubated under identical culture conditions in a separate CO*-incubator of the same type (Heraeus COr-autozero) without exposure to EMF. After 7, 14, or 21 days of incubation the cell number, morphology and various biochemical parameters, i.e., total protein synthesis, total collagen synthesis and [3’S]methionine-labeled polypeptide pattern, were studied. Cell ?ype frequency and cell number. Cell cultures were washed with phosphate-buffered saline (PBS) and were subsequently fixed with 3.5% paraformaldehyde and 70% ethanol (10 min each) and stained for 30 s with 0.1% Coomassie blue and 30 min with Giemsa solution as described elsewhere 1131.For normal human skin and lung flbroblasts the frequencies of the mitotic cell types MF I, MF II, and MF III and for the postmitotic fibroblast cell types PMF IV, PMF V, and PMF VI were determined by morphological criteria [13] counting at least 2000 cells in six paralIe1 experimental series. In Table 1 only the changes of the predominant cell types MF II and PMF VI are listed. For transformed human lung fibroblasts the frequency of mitotic flbroblasts (MF) and postmitotic fibroblasts (PMF) were determined by morphological criteria. For all cell cultures the number of cells-per 50 cm’ was determined by trypsinizing the cells and counting them in a Fuchs-Rosenthal hemocytometer 1131.
612 Rodemann, Bayreuther, and Pfleiderer
in nutrlont mdutfonz 3t’rD.$c cl+-‘ncutIotor 6ohdd cdl P6 Tqcontrd
Potrl6lmlmc
&mldd
toll
PO
Fig. 1. Experimental set-up and spatial distribution of the magnetic field in solenoid coils P6. Six solenoid coils P6 (designed to take up six IO-cm tissue culture dishes per coil) were installed in a Heraeus AutoZero CO,-incubator. The incubation system was prevented from increasing the incubation temperature by a water driven Haake cooling system (C.S.). Temperature of the nutrient solution was controlled over the 21-day incubation period. Electromagnetic fields were generated by a function generator M70S (F.G.M70S).
Total prorein synthesis. Cells wer@@eled for 6 h with 1 @i/ml [‘Hlleucine (New England Nuclear, 5 mCi/mmol) in DMEM supplement&vith W% fcs. Radioactivity incorporated into cellular proteins was determined as described elsewhere [ZO]. Toral collagen synthesis. Cells were labeled for 18 h with 4 @ml [‘Hlproline (New England
Fibroblast differentiation in EMF 613
Signal
shape 1
A
Signal shape
2
Fig. 2. Signal shapes and physical properties of the electromagnetic fields. -Signal- shape I: ._ _ _ -.. . .. _ . symmetric, lslpham sinusoidal electromagnetic tieids; maximum maguetic induction 8.4 mT=6 mT&.
Signal shape 2: asymmetric pulsed, monophasic electromagnetic fields; maximum magnetic induction 8.5 mT.
614 Rodemann, Bayreuther, and Pfleiderer TABLE 1 EMF-induced changes in the cell type frequency of mitoticfibroblasts MF ZZand postmitoticfibroblasts PMF VI in normal human skin and lungfibroblast populations Incubation time (days)
starting %MFII
7 Days
14 Days
21 Days
%MFII
%PMFVI
%MFII
%PMFVI
%MFII
%PMFVI
Human skin fibroblasts CP 82.7~k7.9 EMF-P 82.9k7.8
83.6k7.1 53.2k6.3
O.lfO.O1 25.6k4.7
81.6k5.9 7.2fl.l
O.lkO.03 59.7f6.2
81.9k7.2 0
0.2kO.03 85.4f5.9
Human lung tibroblasts CP 86.2k8.1 EMF-P 86.4k8.2
85.3k7.2 73.2k4.7
0.2f0.03 19.7f3.9
86.126.6 9.8k2.4
0.3f0.02 48.3k5.9
84.7k5.6 0
0.3f0.01 84X3+8.4
Note. Normal human skin fibroblast cell line HH-8, CPDL 3&32. Normal human lung fibroblast cell line W138, CPDL 32-34. Numbers shown represent the mean values + SD of six parallel experiments. At each time point 2000 cells were analyzed and the percentage of MF II and PMF VI was listed. CP, control populations; EMF-P, EMF-exposed populations.
Nuclear, 40 Ci/mmol) in serum-free DMEM supplemented with 0.1 mg/ml sodium-ascorbate and 0.1 mgknl @ninoproprionitrile. Incorporation of [3H]proline into total extra- and intracellular pepsinresistant material consisting at least of 80% collagen was analyzed according to standard procedures 121,221. [3’S]Methionine-labeledpolypeptide pattern. Cells were labeled for 18 h with 200 @i/ml [“Slmethionine (New England Nuclear, 400 Ci/mmol) in DMEM supplemented with 10% fcs as described [23, 241. After labeling the cells were lysed [23, 241 and radioactivity was determined. Aliquots of the various samples containing 500,000 cpm were analyzed by two-dimensional polyacryiamide gel electrophoresis [23].
RESULTS Normal Fibroblasts Cell type frequency in normal human skin and lung jibroblasts. The frequencies of mitotic fibroblasts MF II (Figs. 3A and B) and postmitotic fibroblasts
PMF VI (Figs. 3 C and D) in HH-8 and WI38 cell populations were analyzed at Days 7, 14, and 21 of incubation with and without exposure to biphasic, sinusoidal EMF (signal shape I, Fig. 2). Table 1 shows that normal human skin fibroblast populations of the cell line HH-8 which initially comprised 82.9% MF II type cells differentiate upon exposure to EMF to populations comprising 85.4% PMF VI at Day 21. Likewise normal human lung fibroblast populations of the cell line WI38 show similar changesin the cell type composition of populations exposed to EMF (Table 1). Control populations show no change in the frequencies of MF II and PMF VI during the 21 days of incubation (Table 1). Total protein and collagen synthesis. Measurements of the total protein and collagen synthesis in control populations of HH-8 and WI38 tibroblasts showed a
Fibroblast dijerentiation
in EMF
615
Fig. 3. Mitotic fibroblasts MF II of normal human skin libroblasts, cell line HH-8 (A), and of normal human hmg fiboblasts, cell line WI38 (B) at Day 21 of control populations. EMF-induced postmitotic flbroblasts PMF VI of the cell line HH-8 (C) and WI38 (0) after 21 days of exposure to EMF. Cells were seeded at an identical density of 5x10 cells per 50 cm2 before EMF treament or incubation under control conditions. During the 21day incubation period, untreated control cell populations underwent 2.1-2.9 CPD before reaching confluency. EMF-exposed cell populations underwent during the first 2 weeks of EMPtreatment a total of 0.243 CPD before all cells were arrested in the irreversible nondividing, postmitotic differentiation state. Magnification for Fig. A, B, C, D: x 100.
significant decrease over the incubation period of 21 days reflecting the downregulation of these processes as a function of increasing contact regulation. Skin and lung fibroblasts populations exposed to sinusoidal EMF, however, showed a significant increase in the synthesis of total protein and total collagen. This reflects the differentiation-dependent increase in the capacity for the synthesis of macromolecules in PMF VI type cells (Table 2) [25, 261. [35S]Methionine polypeptide pattern. Normal human skin tibroblasts exposed to EMF for 21 days show significant changes in the two-dimensional polypeptide pattern of total soluble cellular proteins (Fig. 4). Cell populations exposed to EMF express protein spots (Fig. 4B) that are not present in control populations (Fig. 4A), but that have been demonstrated to be specific for postmitotic fibroblasts PMF VI (Fig. 4B) [13]. Similar changes in the protein pattern between mitotic and EMF-induced postmitotic cells could be observed using the human lung tibroblasts cell line WI38 (data not shown). 40-898336
616 Rodermann, Bayreuther, and Pfleiderer
a
Fig. 4. [%]Methionine poIypeptide pattern of normal human skin fibroblasts cell line HIi-8. (A) Control population at Day 21 of the experiment; (B) EMF-exposed population at Day 21 of the experiment. Protein spots indicated by arrowheads are cell type-specific proteins of the fibroblast type PMF VI. IEF, isoelectric focusing, acidic end to the right. M,, molecular weight.
Transformed Fibroblasts Frequency of mitotic and postmitotic transformed human lungfibroblasts.
Cell
populations at two initial seeding densities (5 x 104or 5x 10scells per 50 cm’) were used to analyze the effects of biphasic, sinusoidal EMF on the differentiation of the transformed human lung fibroblast cell line WI38SV40. As shown in Table 3 control populations at either cell density showed no significant change in the frequencies of mitotic or postmitotic cells during the 21 days of incubation. Cell populations with an initial density of 5x 10’ cells could be analyzed only up to 14 days of culture. Longer incubation periods without passagingthe cells resulted in spontaneousdegeneration of the cell populations. ‘Ikansformedlung fibroblasts (initial density 5 x lo4 cells) exposed to sinusoidal EMF for up to 21 days showed significant changes in the frequency of mitotic and postmitotic cells. After 21 days of exposure to EMF, WI38SV40 cell populations were composed of 71% postmitotic nondividing cells (Table 3). However, in contrast to their nontransformed counterparts (WI38 cells, Table 1) transformed lung fibroblasts cannot be induced completely to postmitotic, nondividing cells by EMF exposure. After 21 days 29% of the living cells were still of the mitotic cell type and were actively dividing. In cell populations of an initial density of 5 x 10’ cells per 50 cm* this effect was
Fibroblast diflerentiation
in EMF
617
TABLE 2 EMF-induced changes in total protein and collagen synthesis in normal human skin and lung fibroblasts Incubation time Initial cell density/SO cm2
7 Days
Protein synthesis (cpm 13H]Leu incorporation/id Normal human skin fibroblasts CP 5x10’ EMF-P 5x10’
CP EMF-P
21 Days
cells] 23.2k6.3 91.4k12.9’ (x3.93)
19.1f3.6 176.7+20.8* (x9.25)
34.1k2.7 90.2f6.8* (x2.64) 5x16 24.4f5.1 13.7f1.7 26.7f5.3 5Xld 51.8+8.4* (x3.78) Collagen synthesis (cpm TH]Pro incorporation/l@ cells)
37.8f4.1 190.1+10.4* (X5.03) 13.8k1.8 132.6+ 14.2* (x9.61)
Normal human lung fibroblasts CP 5x1@ EMF-P 5x10’
Normal human skin fibroblasts CP 5x104 EMF-P 5x10’ Normal human lung fibroblasts CP 5x10’ EMF-P 5x 10’
28.5k5.7 33.7f8.9 (x1.18)
14 Days
34.1k7.1 33.2f5.8
42.7k6.3 71.9f8.9* (X1.68)
14.6k1.6 103.5+9.7* (x7.09)
8.5kO.9 118.9+11.8* (x13.98)
39.7k6.3 87.4+9.6* (X2.2)
15.9f1.8 143.4*12.6* (X9.01)
13.6k1.9 180.4+15.9* (x 13.26)
* ZWO.05, n=lB. Note. Normal human skin iibroblast cell line HH-8, CPDL 30-32. Normal human lung fibroblast cell line W138, CPDL 32-34. Numbers shown are the means + SD of six parallel experiments with two subsamples in each group. For statistical analysis, data were compared by a generalized block design combined with the Dunnett test [31]. CP, control populations; EMF-P,EMF-exposed populations.
much more pronounced. As shown in Table 3 after 14 days of EMF exposure 15% of the living cells were of postmitotic cell type, as compared with 83 % mitotic and actively dividing cells. As tested by the fluorescein diacetate/ethidium bromide assay [27], the exposure of transformed lung fibroblasts to EMF results not only in a differentiation of subpopulations to postmitotic cells but also in a degeneration of other subpopulations of these cells. After 14 to 21 days the percentage of dead cells was between 20 and 48%, depending on the initial cell density of the cultures (data not shown). Total protein and collagen synthesis. As demonstrated in Table 4 control cell populations of transformed lung fibroblasts (initial cell density 5X lo4 cells) showed comparable rates of protein and collagen synthesis as their non-transformed counterparts (Table 2). Cells exposed to EMF for 21 days showed parallel to the increasing number of postmitotic cell types a significant increase in total
618 Rodemann, Bayreuther, and Pfleiderer TABLE 3 EMF-induced changes in the frequency of mitotic and postmitotic cell types of transformed human lung fibroblast populations Incubation time 7 Days
14 Days
21 Days
Initial cell density150 cm’
%MF
% PMF
%MF
% PMF
%MF
% PMF
CP EMF-P
5x104 5x104
100 91.1f3.4
0 8.9f3.4
99.7kO.l 52.Ok3.1
0.3fO.l 48.Ok3.1
99.1f0.2 29.1k3.2
0.9f0.2 70.9k3.2
CP EMF-P
5x10’ 5XlOJ
100 95.2k2.6
0 4.8k2.6
99.3+0.4 85.Ok5.1
0.7kO.4 15.Ok5.1
-
-
Note. SV&transformed human lung fibroblast cell line WI38SV40. Numbers shown represent the mean values f SD of six parallel experiments. At each time point cells were analyzed as de&bed under Methods and the percentage of mitotic cells (MF) and postmitotic cells (PMF) is listed. Two thousand cells were analyzed in each experimental series. CP, control populations; EMF-P, EMFexposed populations.
protein and collagen synthesis by a factor of 8-10 (Table 4). Likewise, transformed lung fibroblasts at an initial density of 5 x ld cklls upon exposure to EMF for 14days exhibited a significant increase in total protein synthesis by a factor of 6 (Table 4). TABLE 4 EMF-induced changes in total protein synthesis and collagen synthesis of transformed human lung fibroblast populations Incubation time _..-
Initial ceil density150 cm2
7 Days
Protein synthesis (cpm [ti]Leu incorporation/lO’ cells) 39.1f9.7 CP 5x104 EMF-P 5x104 31.3f16.0 CP EMF
5x105 5Xld
31.7k6.1 35.6f16.1
Collagen synthesis (cpm [‘HIPro incorporation/id cells) 5x104 53.7f6.9 CP EMF-P 5x 10’ 99.4k8.3” (X1.85)
14 Days
21 Days
24.3zb5.2 I98.lf61.4* k(X8.15)
34.1f2.1 257.5+72.8* (X7.55)
34.2+ 16.4 206.4+82.9* (x6.03) 32.1k4.8 125.6+ 10.6* (X3.91)
16.9f2.1 133.5+13.6* (X7.89)
* PcO.05, n=18. Note. SV40-transformed human lung fibroblast cell line WI38SV40. Numbers shown are the means + SD of six parallel experiments with three subsamples in each group. For statistical analysis, data were compared by a generalized block design combined with the Dunnett test [31]. CP, control populations; EMF-P, EMF-exposed populations.
Fibroblast differentiation in EMF 619 [35S]Methioninepolypeptide pattern. A comparison of the polypeptide pattern of control WI38SV40 cells and cells which had been exposed to EMF for 21 days revealed sign&ant changes in a variety of proteins. After 21 days of exposure to EMF these cells expressed at least 11 proteins that are not present in control populations after 21 days of incubation (data not shown). DISCUSSION The results of this study suggest that in vitro exposure of normal and transformed human fibroblast population to biphasic, sinusoidal EMF (20 Hz, 6 mT) stimulates the differentiation of mitotic fibroblast cell types to postmitotic tibroblast cell types that are characterized by significantly enhanced metabolic activities. Normal fibroblasts from various organs of vertebrate and mammalian species including man have recently been demonstrated to be cells which differentiate along a terminal cell lineage representing the morphologically recognizable compartments of the fibroblast stem cell system [ 13, 14,21,25,28]. Independently of the species or organs analyzed fibroblast populations are composed of three mitotic cell types (MF I, MF II, MF III) and four postmitotic cell types (PMF IV, PMF V, PMF VI, PMF VII) 113, 14, 291. In routinely used secondary fibroblast populations the frequencies of the cell types MF I, MF II, and MF III change as a function of the CPDL [13, 141. Similarly in spontaneously arising postmitotic fibroblast cultures the frequencies of PMF IV, PMF V, PMF VI, and PMF VII are a function of the duration of the stationary postmitotic culture [131.Morphological and biochemical data revealed that the mitotic cell type MF II is the predominant cell type in secondary cultures of early and middle passage(CPDL 2040) cell populations [13]. PMF VI represents the predominant and biosynthetitally most active postmitotic and terminally differentiated cell type of the fibroblast stem cell systemJl3, 261. As demonstrated by morphological and biochemical criteria sinusoidal EMF (20 Hz, 6 mT) induce the differentiation of both normal human skin and lung fibroblast MF II populations to PMF VI populations within 21 days. Concomitantly the physiological activity, i.e., total protein and collagen synthesis, increased significantly as a function of the differentiation state of the populations analyzed. The terminal differentiation state of the fibroblasts after 21 days of exposure to EMF is also demonstrated by the expression of six new proteins which have recently been described to be specific for postmitotic tibroblasts [ 13, 301. Recently Goodman and Henderson have also described qualitative and quantitative changes in patterns of protein expression of dipteran salivary gland cells in culture upon exposure to EMF [32]. Similar results of a stimulation of collagen synthesis in bone marrow fibroblasts after exposure to short-term pulsed electromagnetic fields (PEMF) have been described by Famdale and Murray [9]. From these studies 191,however, it is unclear whether the stimulation of collagen synthesis was due to a stimulation of the synthetic capacity of mitotic cells or was due to the PEMF-induced postmitotic differentiation or modulation of the cell populations used.
620 Rodemann, Bayreuther, and Pjleiderer
Long-term treatment with asymmetric pulsed monophasic EMF (signal shape 2, Fig. 2) (21 days, 2 x 6 h per day) induced in subpopuh+tionsof normal human skin and normal human lung fibroblast cell lines (HHS, WI38) and in transformed human lung fibroblasts (WI38SV40) a modulation into reversibly nondividing cells. The modulated cells increase in size, show an enhanced protein synthesis, but do not undergo endopolyploidization. Upon termination of the EMF treatment the modulated cells revert to normal size and function and regain the capacity for cell divisions (Rodemann et al., in preparation). As recently demonstrated by the use of two-dimensional gel electrophoresis 1291spontaneously transformed BN-rat fibroblasts can differentiate spontaneously at low frequencies into nondividing, irreversible postmitotic cells. As demonstrated by morphological and biochemical criteria WI38SV40 cells can be induced by biphasic sinusoidal EMF to differentiate into irreversibly postmitotic and subsequently degenerating cells. However, as compared to normal lung fibroblasts the frequency of the EMF-induced postmitotic cells was significantly lower in transformed lung fibroblast populations and was furthermore a function of the initial cell density. After 21 days of exposure to EMF actively dividing mitotic fibroblast cell types were present in different numbers in cultures of both initial cell densities tested. These data indicate that mitotic mass cultures of WI38SV40 cells are composed of subpopulations that are either sensitive or completely insensitive to the EMF-treatment regime used. These results therefore make the discussion of the relevance of EMF treatment for the developmentof strategies in cancer therapy impossible at the present time. Taken as a whole, the data presented in this study suggestthat the exposure of normal human skin and lung fibroblasts in vitro to sinusoidal EMF induces the differentiation of mitotic to irreversibly postmitotic fibroblasts. The arising postmitotic fibroblasts are significantly more active with regard to total protein and collagen synthesis than the corresponding mitotic fibroblast pd&ilations. This finding is of particular interest for determining the cellular mechanismsby which biphasic sinusoidal EMF induce changes in the behavior and function of secondary fibroblast populations in vitro. Since primary and secondary skin fibroblasts of man differentiate along the same terminal cell lineage [14] it can be expected that fibroblasts in uiuo show similar or identical EMF-induced changes of cellular function and behavior, e.g., during wound healing and bone fracture reunion. We are indebted to the help and advice of W. Kraus, Institut fib Medizinische Physik, Munchen, F.R.G., and for installing the experimental set-up. We are grateful to Drs. G. MoIineux and T. M. Dexter, Paterson Institute for Cancer Research, Christie Hospital & Holdt Radium Institute, Manchester, UK, for critically reading the manuscript. This work was supported in part by the Deutsche Forschungsgemeinschaft (Ba 526/5 and SFB 223), the Fritz-Thyssen-Stiftung, and the Breuninger Stiftung. H. P. R. is a recipient of a Heisenberg grant from the Deutsche Forschungsgemeinschaft (Ro 52712-l).
REFERENCES 1. Grida, N., and Feldman, J. D. (1982) Cell Motil. 8, 243. 2. Nuccitelli, R., and Erickson, C. A. (1983) Exp. Cell Res. 147, 195. 3. Stump, R. F., and Robinson, K. R. (1983) Cell Biol. 97, 1226.
Fibroblast differentiation in EMF
621
4. Kraus, W., and Lechner, F. (1972)Miinch. Med. Wochenschr. 114, 42, 1814. 5. Kraus, W. (1984)Orthopiide 13,78. 6. Bassett,C. A. (1978)in Metabolic Surgery(Buchwald, H., and Varco, R. L., Eds.), p. 255,Grune
& strattoll, NY. 7. Watson,J., and Dowries, E. M. (1978)Japan J. Appl. Phys. 17, 215. a. Takahashi,K., Kaneko, I., Date, M., and Fukada, E. (1986)Experientiu 42, 185. 9. Famdale, R. W., and Murray, J. C. (1985)Cal& Tissue Inf. 37, 178. 10. L&off, A. R., Williams, T., Strong, D. M., and Wistar, R. (1984)Science 223, 1984. 11. Fisher, S. J., and Dulling, J. (1986)J. Bioelectr. 5, 253. 12. Lee, R. C., Rich, J. B., Kelley, K. M., Weismann,D. S., and Mathews,M. B. (1983)Amer. Surg. 48,567.
13. Bayreuther, K., Rodemann,H. P., Hommel, R., Dittmann, K., Albiez, M., and Francz, P. I. (1988)Proc. Natl. Acad. Sci. USA 84, 5112. 14. Bayreuther, K., Rodemann,H. P., Francz, P. I., and Maier, K. (1988)J. Cell Sci., in press. 15. Lechner, F., and Ascherl, R. (1978)Acto Orthop. Be/g. 44, 699. 16. Ascherl, R., Bltlmel, G., Lechner, F., and Kraus, W. (1985)Bioelectrochemistryand Bioenergetits 14, 161. 17. Lechner, F. (1972)Ifefle 2. U&Uheifkd. 114, 325. 18. Kraus, W. (1979)2nd Meeting of the European Society of Biomechanics.Strasbourg.[Abstract] 19. Hayflick, L.. and Moorehead, P. S. (l%l) Exp. Cell Res. 28, 585. 20. Rodemann,H. P., Bayreuther, K., and Schumacher,E. (1987)Biochem. Biophys. Res. Commun. 14&l. 21. Rodemann,H. P., and Bayreuther, K. (1984) Proc. Nutl. Acad. Sci. USA 81, 5130. 22. Mollenhauer, J., and Bayreuther, K. (1986)D&%rentiution 32, 165. 23. Bravo, R. (1984)in Two Dimensional Gel Electrophoresisof Proteins (Celis, J., and Bravo, R., Eds.), p. 4, Academic Press,New York. 24. Rodemann,H. P., and Bayreuther, K. (1986)Proc. Natf. Acad. Sci. USA 83, 2086. 25. Bayreuther, K., Rodemann,H. P., Kontermann, K., Mollenhauer, J., Francz, P. I., Dittmann, K., Mtlller, C., and Mtlller, G. (1987)Eur. J. Cell Biol. 43, 4. 26. Francz, P. I., Bayreuther, K., Rodemann,H. P., Hommel, R., and Dittmann, R. (1989)Different&ion, in press. 27. Rotman, B., and Papermaster,B. W. (1966)Proc. Natl. Acad. Sci. USA 55, 134. -~ 28. Kontermann, K., and Bayreuther, K. (1979)Gerontology 25, 261. 29. Francz, P. I., Kontermann, K., Mollenhauer, J., and Bayreuther, K. (1987)Anticancer Res. 7, 5A; Abstr. 168,921. 30. Rodemann,H. P., Bayreuther, K., Francz, P. I., Dittmann, &and Albiez, M. (1989)Exp. Cell Res. 1(10,84.
31. Neter, J., Wassermann,W., and Kutner, M. H. (1985)Applie&inea Statistical Models, 2nd ed., Richard D. Irwin Inc. 32. Goodman,R., and Henderson,A. S. (1988)Proc. Natl. Acaa. Sci. USA 85, 3928. ReceivedJuly 27, 1988 Revisedversion received December28, 1988
Printedin Sweden