Cell,
Vol. 15, 813-822,
Schwann
November
1978, Copyright
0 1978 by MIT
Cell Growth
Factors
Martin C. Raff, Erika Abney, Jeremy P. Brockes* and Ann Hornby-Smith MRC Neuroimmunology Project Zoology Department University College London London WC1 E 6BT, England
Summary Purified rat Schwann cells were found to proliferate very slowly in normal growth medium containing 10% fetal calf serum (FCS). Crude extracts of hovine pituitary or brain markedly enhanced Schwann cell growth, while similar extracts of nerve roots, liver and kidney did not. Pituitary extracts were more potent than brain extracts, and extracts from both anterlor and posterior pituitary were active. The mitogenic activity of pituitary extracts was reduced by treatment with trypsin, and abolished by pronase and by boiling. A variety of known anterior and posterior pituitary hormones, as well as fibroblast, epidermal and nerve growth factors, were not mitogenfc. FCS (>l%) was required for Schwann cell proliferation, but even high concentrations of FCS did not substitute for pituitary or brain extracts, and serum from various other species did not support Schwann cell growth. Although various agents that increase cyclic AMP levels (such as cholera toxin) had been shown to be Schwann cell mitogens, extracts of pituitary or brain did not increase cyclic AMP levels. Extracts of various bovine tissues, including pituitary, brain, liver and kidney, acted synergistically with cholera toxin in stimulating Schwann cell proliferation, although the increase in cyclic AMP induced by the mixture was not greater than that seen with cholera toxin alone. We conclude that there are at least two separate pathways for stimulating Schwann cell division, only one of which involves an increase in intracellular cyclic AMP. Introduction Schwann cells are the major glial elements in the peripheral nervous system. They surround all peripheral nerve fibers and in many cases wrap themselves concentrically around the axons to form a multilayered myelin sheath. Although Schwann cell proliferation has an important role in the development of peripheral nerves (Peters and Muir, 1959), in adults these cells normally do not divide (Asbury, * Present Technology,
address: Division of Pasadena, California
Biology, 91125.
California
Institute
of
1967) unless they are stimulated to do so by nerve injury (Abercrombie and Johnson, 1946; Bradley and Asbury, 1970). We have previously described a surface antigenic marker for rat Schwann cells, rat neural antigen 1 (Ran-l), which can be used to identify these cells unambiguously in dissociated cell cultures of newborn rat sciatic nerve (Brockes, Fields and Raff, 1977). By using antiserum against the Thy-l surface antigen in conjunction with anti-Ran-l serum in immunofluorescence assays, we showed that rat sciatic nerve cultures consisted of only two major cell types: Ran-l+/Thy-l- Schwann cells and Thy l+/Ran-l- fibroblasts (Brockes et al., 1977). By exposing these cultures to 3H-thymidine and combining autoradiography with immunofluorescence, we found that most of the fibroblasts were dividing rapidly in medium containing 10% fetal calf serum (FCS), while most of the Schwann cells were not (Brockes et al., 1977; Brockes, Fields and Raff, 1978; Raff, Hornby-Smith and Brockes, 1978). We were thus able to kill most of the fibroblasts by exposing the primary cultures to the mitotic inhibitor cytosine arabinoside (AraC); the remainder of the fibroblasts could be killed by treating the cells in suspension with anti-Thy-l serum and complement prior to their first passage (Brockes et al., 1978). Using this strategy, we have consistently been able to obtain cultures of Schwann cells that are >99.5% pure (Brockes et al., 1978). To obtain large numbers of Schwann cells for cell biological and biochemical studies, it was necessary to find ways of stimulating the purified cells to divide more rapidly, since their doubling time in regular growth medium containing 10% FCS was approximately 8 days. Preliminary experiments showed that crude extracts of bovine pituitary glands were mitogenic for Schwann cells in culture (Brockes et al., 1978). Moreover, a variety of agents that are known to raise intracellular concentrations of cyclic AMP, such as cholera toxin and dibutyryl cyclic AMP, were found to be Schwann cell mitogens (Raff et al., 1978). This paper reports that the Schwann cell mitogen(s) in pituitary extracts appears not to be any of the previously characterized growth factors or classical pituitary hormones; that extracts of bovine brain are also mitogenic for Schwann cells; that crude extracts from a variety of bovine tissues, including pituitary and brain, act synergistically with cholera toxin in stimulating Schwann cell proliferation; and that the mitogenic and synergistic factors in these extracts apparently do not act by raising intracellular concentrations of cyclic AMP. Our results suggest that there are at least two distinct pathways for stimulating Schwann cell division.
Cell 014
Results Effects of Tissue Extracts on Schwann Cell Proliferation When purified secondary Schwann cells growing in 10% FCS were exposed to 3H-thymidine for 24 hr and then studied by autoradiography, ~20% of the nuclei were usually labeled. If pituitary extract (100-500 pg/ml) was added to the growth medium for 48 hr, the proportion of labeled nuclei was greatly increased (Table 1). By studying Schwann cell growth, we showed that this effect reflected an increase in cell proliferation rather than an increase in thymidine transport or DNA repair: while cells cultured in Dulbecco’s modified Eagle’s medium (DMEM) plus 10% FCS (DMEM-FCS) doubled their number about every 8 days (even when the medium was changed every 3 days), their doubling time was 500 pg/ml; Figure 2 and Table 2). Extracts of both anterior lobes and posterior (plus intermediate) lobes of pituitary were mitogenic, although extracts of anterior pituitary were less so than extracts of posterior or whole pituitary (Figure 2). An extract of rat pituitary was also mitogenic. The mitogenic activity of bovine pituitary extracts was unaffected by heating at 60°C for 5 min, partially inactivated by heating at 65°C for 10 min and abolished by boiling for 20 min (Figure 3). It was reduced by treatment with trypsin (Worthington, 214 U/ml, 0.5% (w/w) for 60 min at 37°C) and abolished by treatment with pronase [ex Streptomyces griseus, Koch Light, 1% (w/w) for 60 min at 37”C] (Figure 3). At these concentrations, neither trypsin nor pronase was mitogenic on its own; nor did either one or the digested extracts inhibit the mitogenic activity of pituitary extract when added to the growth medium. While pituitary extracts were always mitogenic, similar extracts of bovine liver, kidney and nerve roots (cauda equina) were not over a wide concentration range (l-lo3 pg/ml) (Table 2). Extracts of nerve roots were in fact inhibitory both on their own (Table 2) and when added to stimulatory concentrations of pituitary extract (data not
-105
by
Cells Labeled
16 + 2 60 r5
Pituitary extract was added (at a concentration shown to be optimal for ‘V-Udr incorporation) for a total of 72 hr, and 5Hthymidine was present for the final 24 hr of the culture. The results in this and subsequent tables are expressed as means t SD of triplicate cultures.
L 4
10
16
DAYS Figure
1. Effect
of Pituitary
Schwann cells were with added pituitary (C-W
cultured extract
Extract
on Schwann
Cell Proliferation
in DMEM-FCS alone (+O) (200 pg of dry weight per
or ml)
Schwann
Cell Growth
Factors
815
Table 2. Effect Schwann Cells
of Tissue
Extract
Concentration (pg of Protein
-
-
Pituitary
Extracts
per ml)
Figure ‘Y-Udr
4 Pituitary 2. Effect of Extracts Incorporation
zp
Extractbg/ml)
of Anterior
HP
.
and Posterior
on
Schwann cells were cultured in the presence of varying concentrations (fig of protein per ml) of extracts of whole (A-A), anterior (O--O) or posterior (Lm) pituitary. The ‘251-Udr incorporation is plotted so that the stimulation index is equal to cpm with extracbcpm with DMEM-FCS alone.
shown). Although extracts of liver and kidney were not inhibitory on their own, except at very high concentration (>l mg of protein per ml) (Table 2), they were slightly but consistently inhibitory at 100 and 500 pg protein per ml when added to stimulatory concentrations (250 or 500 pg protein per ml) of pituitary extract. Extracts of bovine brain were mitogenic but usually much less so than pituitary (Table 2); brain extracts did not show a prozone at high concentrations (Table 2) and enhanced the mitogenic effect of pituitary extract, even when both were used at 500 Erg of protein per ml (data not shown). Pituitary extract did not stimulate 1251-Udr incorporation into Schwann cells in ~1% FCS, and the stimulation induced by the extract in the presence of 10% FCS was not appreciably different from that induced in the presence of 20 or 30% FCS. Thus although FCS was required for Schwann cell proliferation, even high concentrations of FCS did not substitute for pituitary extract.
Failure of Known Growth Factors and Pituitary Hormones to Stimulate Schwann Cell Proliferation Nerve growth factor (NGF), fibroblast growth factors (FGF) purified from bovine pituitary and bovine brain (Gospodarowicz, 1975, 1976), and epidermal growth factor (EGF) failed to stimulate 1251-Udr incorporation into Schwann cells when added at 10 ng, 100 ng, 1 pg and 10 kg/ml (Table 3). FGF and EGF did stimulate lz51-Udr incorporation into serum-starved BALB/c3T3 cells with plateau stimulation at 10 rig/ml (data not shown). Six different anterior pituitary hormones-luteinizing hormone (LH), luteotrophic hormone (LTH), follicle-stimulat-
56
1006 r 110 (3.5) 1485 k 249 (5.2)
500
3330 r
38 (11.7)
372 r 199 (1.3)
20
413 +
89 (1.5)
449 5
54 (1.6)
500
352 2
12 (1.2)
1000
264 +
63 (0.9)
20
470 f
166 (1.6)
Kidney
100
508k
500
458 + 132 (1.6)
1000 Nerve
82(1.8)
208 f
137 (0.7)
20
218 2
20 (0.8)
100
215 -c
21 (0.8)
500
220 +
31 (0.8)
1000
162 r
63 (0.6)
20
Brain
by
Incorporation
100
1 .w....!t’ Pituitary
Incorporation
100
1000
1
lZSI-Udr (cpm) 284%
20
Liver
on ?-Udr
416 5
95 (1.5)
100
547 I
74 (1.9)
500
1017 2
65 (3.5)
loo0
1142 4 151 (4.0)
The numbers in parentheses stimulation indices (= cpm
in this and subsequent tables with stimulanbcpm in controls).
are the
ing hormone (FSH), thyroid-stimulating hormone (TSH), growth hormone (GH) and adrenocorticotropic hormone (ACTH)-and two posterior pituitary hormones-oxytocin and vasopressin - (all obtained from Sigma) were unable to stimulate 1251Udr incorporation in Schwann cells when added at 50 ng, 500 ng, 5 pg and 50 pg/ml (Table 4). The same was true for prolactin, GH, LH and FSH obtained from the Pituitary Hormone Distribution Program of the National Institute of Arthritis, Metabolism and Digestive Diseases, although their preparation of LH stimulated weakly (stimulation index ~2.0) when used at 5 and 50 pg/ml (Table 4). The high concentration required suggested that the stimulation may have been related to a contaminant rather than to the LH itself. The divalent cation carrier A23187 did not stimulate Schwann cell proliferation when used at 1O-g-1O-5 M (data not shown).
Synergistic Effects of Tissue Extracts and Cholera Toxin in Stimulating Schwann Cell Proliferation When
either
cholera
toxin
or pituitary
extracts
were
Cell 818
Table 3. Failure Cell Growth
of Known
Growth
Concentration (fig of Protein per ml)
Factor 0
to Stimulate
Extract
Growth
20
Factor
Schwann
‘*%Udr Incorporation (cpm) 1255e
Pituitary
Nerve
Factors
127
2620 2 176 (2.0)
100
4196
500
6543 2 163 (5.0)
+ 241 (3.2)
0.01
1213 2
0.1
1294 + 174 (1.0)
91 (1.0)
1290 -c 49 (1 .O) Figure 3. Effect of Boiling and Proteolytic genie Activity of Pituitary Extract
Digestion
on the Mito-
Schwann cells were cultured with untreated pituitary extract (LO) or pituitary extract that had been boiled for 20 min (V-V), or treated with trypsin (e--e) or pronase (m-m). The ‘?-Udr incorporation is plotted as the stimulation index, and the concentrations of the extracts are Hg of protein per ml.
10 Pituitary Growth
Fibroblast Factor
921 k 102 (0.7)
0.001
1259 * 159 (1 .O)
0.01
1475 2 210 (1 .l)
0.1
1708k
58(1.3)
1781 + 161 (1.4)
used at their optimal stimulatory concentrations, they induced Schwann cells to double their numbers in 2-3 days, compared with 8 days for control cultures; when both were added together, the doubling time was further reduced to under 2 days (Figure 4). The same synergistic effect was seen when Schwann cell proliferation was assayed by 3H-thymidine incorporation determined by autoradiography (Table 5) or by 1251-Udr incorporation (Table 6 and Figure 5). Remarkably, extracts of liver, kidney and brain were able to act synergistically with cholera toxin as well as extracts of pituitary, while extracts of nerve roots did so only weakly (Table 6). Choleragenoid, the biologically inactive subunit of cholera toxin which binds equally well to the same ganglioside receptor, G,,, , (Finkelstein, 1973; Bennet and Cuatrecasas, 1977), did not stimulate on its own or act synergistically with the tissue extracts (data not shown). While dibutyryl cyclic AMP did act synergistically with tissue extracts in some experiments, its synergistic activity was less striking and less consistent than that of cholera toxin. It is interesting that when cholera toxin (10 pg/ml) and dibutyryl cyclic AMP (5 x lo+ M) were used together, the stimulatory effect was consistently less than with either used alone (data not shown). Heating the pituitary extract at 65°C for 10 min reduced its synergistic activity much more than its direct stimulatory activity (Figure 5). The poor synergistic activity of heated pituitary extract was not due to the presence of a heat-induced inhibitor, since mixing it with unheated pituitary extract did not affect the synergistic activity of the latter (Figure 5). Boiling the pituitary extract for 20 min completely destroyed its ability to act synergistically with cholera toxin, while treatment with trypsin or pronase re-
Brain Fibroblast Growth Factor
0.001
1067 r
0.01
1283 + 102 (1 .O)
0.1
1253 k 148 (1 .O) 1380 t
Epidermal Factor
94 (0.8)
97 (1 .I)
Growth 0.001
1233 + 206 (1 .O)
0.01
1289 k 124 (1.0)
0.1
1447 + 103 (1.2)
1
1376 -t 146 (1 .l)
duced but did not abolish its synergistic activity. The synergistic activity of the tissue extracts was probably not related to the extracts’ preventing the breakdown of cholera toxin or increasing the binding of cholera toxin to the Schwann cell surface, since the effect was similar whether cholera toxin was used at 0.01 or 10 pg/ml, and the tissue extract could not be replaced by pretreating the cells with exogenous G,,, (Bennet and Cuatrecasas, 1977). Moreover, synergy was also seen when Schwann cells were exposed to cholera toxin for 2 hr, washed and then incubated with tissue extract (data not shown). We tested a variety of other substances for their ability to act synergistically with cholera toxin in stimulating ‘251-Udr incorporation into Schwann cells, and none were active. These included all the pituitary hormones and all the growth factors discussed above. Intracellular Cyclic AMP Levels Cyclic AMP levels were measured in Schwann cells after 2, 6 and 20 hr of exposure to control medium (DMEM-FCS) or to medium with added pituitary
Schwann 017
Table
Cell Growth
4. Failure
Factors
of Pituitary
Hormones
to Stimulate ‘*%Udr
Hormone Experiment
Pituitary
Extract
Luteotrophic Sigma)
Hormone
incorporation
(cpm)
0
0.05 pg/ml
310+6rl
-
-
-
-
-
411 f 65 (1.3)
-
343?53(1.1)
-
at:
0.5 pg/ml
5 rglml
50 pglml
100 pg/ml
-
-
-
-
2391 + 915 (7.7)
363 + 26 (1.2)
360 2 97 (1.2)
282 r 37 (0.9)
360 + 42 (1 .I)
393 A 36 (1.3)
280 + 47 (0.9)
259 2 25 (0.8)
371 2 26 (1.2)
362 -t 117(1.2)
202 -c 11 (0.7)
313 + 44 (1 .O)
330 2 29 (1 .O)
262 2 38 (0.8)
291 ? 95 (0.9)
296 e 51 (1 .O)
314 c 25 (1 .O)
359+
343 2 56 (1 .l)
366 f 47 (1.2)
353 I 61 (1 .l)
381 k 37 (1.2)
(LTH;
Follicle-Stimulating (FSH; Sigma)
Hormone
Thyroid-Stimulating (TSH; Sigma)
Hormone
Hormone
(GH; Sigma)
Adrenocopticotropic (ACTH; Sigma)
Hormone
Vasopressin Experiment
(Sigma)
le(l.1)
416 2 41 (1.3)
2
0
646~
Pituitary
Extract
Prolactin Growth
Cell Proliferation
1
0
Growth
Schwann
-
-
(NIAMDD) Hormone
72
(GH; NIAMDD)
Follicle-Stimulating (FSH; NIAMDD)
-
-
-
-
506 t 82 (0.8)
479 2 40 (0.7)
543 f 59 (0.8)
596 f 166 (0.9)
-
466 ? 77 (0.7)
465 zt 95 (0.7)
547 t 34 (0.8)
580 k 26 (0.9)
-
695?82(1.1)
626 r 99 (1 .O)
683 + 66 (1 .l)
610 r 81 (1.0)
656 2 57 (1 .O)
746 + 37 (1.2)
119k
456 f 51 (0.7)
631 ‘- 140 (1 .O)
610 -t 26 (0.9)
2056 -t 50 (3.2)
Hormone
Luteinizing Hormone (LH; NIAMDD) Luteinizing Hormone (LH; Sigma)
-
-
extract (100-400 pg/ml), cholera toxin (10 pglml) or both. In eleven separate experiments, cholera toxin consistently increased cyclic AMP levels 4-30 fold at the three time points (the increase usually being greatest at 6 hr), while pituitary (or brain) extract did not have a significant effect even in the presence of the potent phosphodiesterase inhibitor 3,isobutyL1-methyl-xanthine (IBMX; 3 x 10m5 M) (Matsuzawa and Nirenberg, 1975). When Schwann cells were exposed to cholera toxin and pituitary extract together, the rise in cyclic AMP was not significantly different from that seen with cholera toxin alone. The results of a typical experiment are given in Table 7; since we had previously shown that prostaglandin E, and isoproterenol were not mitogenic for cultured Schwann cells (Raff et al., 1978), it was interesting to observe that these reagents did not increase cyclic AMP levels in these cells when assayed at 5 min or at 2,6 and 20 hr (Table 7). Discussion As we (Raff et al., 1978) and other investigators (Wood and Bunge, 1975) have previously observed, purified Schwann cells normally grow very slowly
124(1.9)
1239 -c 117 (1.9) 526 -t 71 (0.8)
in cell culture in conventional serum-containing medium. Since their growth rate under these conditions is essentially linear when plotted on semilog coordinates (Figures 1 and 4; Raff et al., 1978), it is probable that these cells constitute a homogeneous population of infrequently dividing cells rather than a mixture of rapidly dividing and nondividing cells. We found that crude extracts of bovine pituitary and agents which would be expected to increase intracellular levels of cyclic AMP (such as cholera toxin and dibutyryl cyclic AMP) (Raff et al., 1978) were potent Schwann cell mitogens. Since their mitogenic effects were demonstrated on highly purified populations of Schwann cells (>99.5% pure) (Brockes et al., 1978; Raff et al., 1978), it is probable that they acted directly on Schwann cells rather than by inducing another cell type to secrete substances that are mitogenic for Schwann cells. In this study, we have further investigated the actions and interactions of these Schwann cell growth factors on purified rat Schwann cells in culture. While extracts of bovine (and rat) pituitary were always mitogenic for purified rat Schwann cells, similar extracts of bovine liver, kidney or nerve roots were not. Extracts of bovine brain (cerebral
Cell 818
Table 5. Synergy between Stimulating “H-Thymidine Pituitary Extract (Kg of Dry/Weight per ml)
4
DAYS
Figure 4. Synergy between Pituitary Extract Stimulating Schwann Cell Proliferation Schwann cells were cultured cholera toxin (10 rig/ml) (m-D), weight per ml) (A-A) or extract (*-*).
16
lo and Cholera
Toxin
in
in DMEM-FCS alone (u), pituitary extract (400 pg of dry both cholera toxin and pituitary
hemispheres) were mitogenic, but less so than pituitary extracts. Since mixing experiments showed that extracts of liver, kidney and nerve roots partially suppressed the mitogenic effect of pituitary extracts, it is difficult to determine whether their lack of mitogenic activity was related to the presence of inhibitors or to the absence of stimulatory factor(s). This is particularly so in the case of extracts from nerve roots, which were markedly inhibitory. Since fibroblast growth factor (FGF), which is isolated from bovine pituitary and brain (Gospodarowicz, 1975), has been reported to be mitogenic for glial cells (Westermark and Wasteson, 1975), we were interested in determining whether the Schwann cell stimulatory activity in our pituitary and brain extracts was due to FGF. This seemed improbable since our extracts from pituitary were more potent than those from brain, while the concentration of FGF in bovine brain has been reported to be 5-10 times greater than that in pituitary (Gospodarowicz, 1974). The possibility was formally excluded by showing that purified FGF from brain or pituitary (provided by Dr. D. Gospodarowicz) did not stimulate Schwann cell proliferation. Epidermal growth factor (EGF) and nerve growth factor (NGF) were also inactive in this respect.
Cholera MO-w
Toxin
Schwann
Cells
Toxin
in
Labeled
W
0
0
15*9
500
0
56 -c 5
0
10
55 2 7
500
IO
91 e 3
Pituitary extract and 3H-thymidine
1
Pituitary Extract and Cholera Uptake by Schwann Cells
and cholera toxin were added for a total of 48 hr, was present for the final 24 hr of the culture.
Since extracts from both anterior and posterior (including intermediate) lobes of pituitary were both mitogenic for Schwann cells, it seemed improbable that any of the classical pituitary hormones were responsible for this activity. We tested most of the anterior and posterior pituitary hormones, including LH, LTH, FSH, TSH, GH, ACTH, prolactin, oxytocin and vasopressin, and found that they do not stimulate Schwann cell growth. One preparation (but not another) of LH was weakly mitogenic when used at high concentration (25 pg/ml), suggesting that it might be contaminated with Schwann cell growth factor(s). It is clear that the mitogenic activity of our pituitary extracts could not be due solely to the presence of LH. Two small peptide hormones, neurotensin (Carraway and Leeman, 1976) and gastrin (Rehfeld, 1978), have recently been demonstrated in both anterior and posterior lobes of pituitary. It is improbable that these or other similar small peptides could account for the Schwann cell stimulatory activity in our pituitary extracts since the activity of these peptides is usually resistant to boiling, while the activity of our extracts was abolished by this treatment. The fact that the mitogenic activity was reduced by heating at 65°C for 10 min and by proteolytic digestion suggests that the active principle(s) in the pituitary extracts is a protein; further biochemical characterization and purification will be required, however, to determine its nature and relationship to known macromolecules. Although Schwann cells appeared to survive in 1% fetal calf serum (FCS), they did not proliferate in response to pituitary extract or cholera toxin in ~1% FCS, suggesting that FCS provided a factor(s) which was necessary for Schwann cell proliferation but which was not specific for the mitogenic stimulus. Although serum was required for Schwann cell growth, even high concentrations of FCS (up to 30%) could not substitute for pituitary or brain extract in stimulating this growth, suggesting that the stimulating factor(s) in these extracts was not present in mitogenic concentrations in FCS. It is surprising that sera from other species, including
Schwann 619
Cell Growth
Factors
Table 6. Synergy between Tissue Extracts and Cholera Stimulating ‘V-Udr Uptake by Schwann Cells
Tissue Extract (pg of Protein per ml) Pituitary
Brain
Figure Activity
5. Effect of Heating of Pituitary Extract
on
the
Mitogenic
and
Synergistic
Schwann cells were incubated in pituitary extract (V-V), pituitary extract that had been heated for 10 min at 65°C (+---+), cholera toxin (O-O), pituitary extract plus 10 rig/ml cholera toxin (LO), heated pituitary extract plus cholera toxin (m-m), cholera toxin plus 60 kg/ml each of pituitary extract and heated pituitary extract (O), or cholera toxin plus 250 Kg/ml each of pituitary extract and heated pituitary extract (*). The ‘YUdr incorporation is plotted as the stimulation, index and the concentrations of extracts are calculated as pg of protein per ml.
adult bovine, neonatal and adult rat, adult horse and adult human sera, did not support pituitary extract (or cholera toxin)-induced Schwann cell growth. Moreover, when FCS was mixed with any of these sera, it no longer supported Schwann cell proliferation, suggesting that these sera were actively inhibitory rather than deficient in an essential nutrient. Wood and Bunge (1975) have reported that
Vol. 15, 813-822,
Schwann
November
1978, Copyright
0 1978 by MIT
Cell Growth
Factors
Martin C. Raff, Erika Abney, Jeremy P. Brockes* and Ann Hornby-Smith MRC Neuroimmunology Project Zoology Department University College London London WC1 E 6BT, England
Summary Purified rat Schwann cells were found to proliferate very slowly in normal growth medium containing 10% fetal calf serum (FCS). Crude extracts of hovine pituitary or brain markedly enhanced Schwann cell growth, while similar extracts of nerve roots, liver and kidney did not. Pituitary extracts were more potent than brain extracts, and extracts from both anterlor and posterior pituitary were active. The mitogenic activity of pituitary extracts was reduced by treatment with trypsin, and abolished by pronase and by boiling. A variety of known anterior and posterior pituitary hormones, as well as fibroblast, epidermal and nerve growth factors, were not mitogenfc. FCS (>l%) was required for Schwann cell proliferation, but even high concentrations of FCS did not substitute for pituitary or brain extracts, and serum from various other species did not support Schwann cell growth. Although various agents that increase cyclic AMP levels (such as cholera toxin) had been shown to be Schwann cell mitogens, extracts of pituitary or brain did not increase cyclic AMP levels. Extracts of various bovine tissues, including pituitary, brain, liver and kidney, acted synergistically with cholera toxin in stimulating Schwann cell proliferation, although the increase in cyclic AMP induced by the mixture was not greater than that seen with cholera toxin alone. We conclude that there are at least two separate pathways for stimulating Schwann cell division, only one of which involves an increase in intracellular cyclic AMP. Introduction Schwann cells are the major glial elements in the peripheral nervous system. They surround all peripheral nerve fibers and in many cases wrap themselves concentrically around the axons to form a multilayered myelin sheath. Although Schwann cell proliferation has an important role in the development of peripheral nerves (Peters and Muir, 1959), in adults these cells normally do not divide (Asbury, * Present Technology,
address: Division of Pasadena, California
Biology, 91125.
California
Institute
of
1967) unless they are stimulated to do so by nerve injury (Abercrombie and Johnson, 1946; Bradley and Asbury, 1970). We have previously described a surface antigenic marker for rat Schwann cells, rat neural antigen 1 (Ran-l), which can be used to identify these cells unambiguously in dissociated cell cultures of newborn rat sciatic nerve (Brockes, Fields and Raff, 1977). By using antiserum against the Thy-l surface antigen in conjunction with anti-Ran-l serum in immunofluorescence assays, we showed that rat sciatic nerve cultures consisted of only two major cell types: Ran-l+/Thy-l- Schwann cells and Thy l+/Ran-l- fibroblasts (Brockes et al., 1977). By exposing these cultures to 3H-thymidine and combining autoradiography with immunofluorescence, we found that most of the fibroblasts were dividing rapidly in medium containing 10% fetal calf serum (FCS), while most of the Schwann cells were not (Brockes et al., 1977; Brockes, Fields and Raff, 1978; Raff, Hornby-Smith and Brockes, 1978). We were thus able to kill most of the fibroblasts by exposing the primary cultures to the mitotic inhibitor cytosine arabinoside (AraC); the remainder of the fibroblasts could be killed by treating the cells in suspension with anti-Thy-l serum and complement prior to their first passage (Brockes et al., 1978). Using this strategy, we have consistently been able to obtain cultures of Schwann cells that are >99.5% pure (Brockes et al., 1978). To obtain large numbers of Schwann cells for cell biological and biochemical studies, it was necessary to find ways of stimulating the purified cells to divide more rapidly, since their doubling time in regular growth medium containing 10% FCS was approximately 8 days. Preliminary experiments showed that crude extracts of bovine pituitary glands were mitogenic for Schwann cells in culture (Brockes et al., 1978). Moreover, a variety of agents that are known to raise intracellular concentrations of cyclic AMP, such as cholera toxin and dibutyryl cyclic AMP, were found to be Schwann cell mitogens (Raff et al., 1978). This paper reports that the Schwann cell mitogen(s) in pituitary extracts appears not to be any of the previously characterized growth factors or classical pituitary hormones; that extracts of bovine brain are also mitogenic for Schwann cells; that crude extracts from a variety of bovine tissues, including pituitary and brain, act synergistically with cholera toxin in stimulating Schwann cell proliferation; and that the mitogenic and synergistic factors in these extracts apparently do not act by raising intracellular concentrations of cyclic AMP. Our results suggest that there are at least two distinct pathways for stimulating Schwann cell division.
Cell 014
Results Effects of Tissue Extracts on Schwann Cell Proliferation When purified secondary Schwann cells growing in 10% FCS were exposed to 3H-thymidine for 24 hr and then studied by autoradiography, ~20% of the nuclei were usually labeled. If pituitary extract (100-500 pg/ml) was added to the growth medium for 48 hr, the proportion of labeled nuclei was greatly increased (Table 1). By studying Schwann cell growth, we showed that this effect reflected an increase in cell proliferation rather than an increase in thymidine transport or DNA repair: while cells cultured in Dulbecco’s modified Eagle’s medium (DMEM) plus 10% FCS (DMEM-FCS) doubled their number about every 8 days (even when the medium was changed every 3 days), their doubling time was 500 pg/ml; Figure 2 and Table 2). Extracts of both anterior lobes and posterior (plus intermediate) lobes of pituitary were mitogenic, although extracts of anterior pituitary were less so than extracts of posterior or whole pituitary (Figure 2). An extract of rat pituitary was also mitogenic. The mitogenic activity of bovine pituitary extracts was unaffected by heating at 60°C for 5 min, partially inactivated by heating at 65°C for 10 min and abolished by boiling for 20 min (Figure 3). It was reduced by treatment with trypsin (Worthington, 214 U/ml, 0.5% (w/w) for 60 min at 37°C) and abolished by treatment with pronase [ex Streptomyces griseus, Koch Light, 1% (w/w) for 60 min at 37”C] (Figure 3). At these concentrations, neither trypsin nor pronase was mitogenic on its own; nor did either one or the digested extracts inhibit the mitogenic activity of pituitary extract when added to the growth medium. While pituitary extracts were always mitogenic, similar extracts of bovine liver, kidney and nerve roots (cauda equina) were not over a wide concentration range (l-lo3 pg/ml) (Table 2). Extracts of nerve roots were in fact inhibitory both on their own (Table 2) and when added to stimulatory concentrations of pituitary extract (data not
-105
by
Cells Labeled
16 + 2 60 r5
Pituitary extract was added (at a concentration shown to be optimal for ‘V-Udr incorporation) for a total of 72 hr, and 5Hthymidine was present for the final 24 hr of the culture. The results in this and subsequent tables are expressed as means t SD of triplicate cultures.
L 4
10
16
DAYS Figure
1. Effect
of Pituitary
Schwann cells were with added pituitary (C-W
cultured extract
Extract
on Schwann
Cell Proliferation
in DMEM-FCS alone (+O) (200 pg of dry weight per
or ml)
Schwann
Cell Growth
Factors
815
Table 2. Effect Schwann Cells
of Tissue
Extract
Concentration (pg of Protein
-
-
Pituitary
Extracts
per ml)
Figure ‘Y-Udr
4 Pituitary 2. Effect of Extracts Incorporation
zp
Extractbg/ml)
of Anterior
HP
.
and Posterior
on
Schwann cells were cultured in the presence of varying concentrations (fig of protein per ml) of extracts of whole (A-A), anterior (O--O) or posterior (Lm) pituitary. The ‘251-Udr incorporation is plotted so that the stimulation index is equal to cpm with extracbcpm with DMEM-FCS alone.
shown). Although extracts of liver and kidney were not inhibitory on their own, except at very high concentration (>l mg of protein per ml) (Table 2), they were slightly but consistently inhibitory at 100 and 500 pg protein per ml when added to stimulatory concentrations (250 or 500 pg protein per ml) of pituitary extract. Extracts of bovine brain were mitogenic but usually much less so than pituitary (Table 2); brain extracts did not show a prozone at high concentrations (Table 2) and enhanced the mitogenic effect of pituitary extract, even when both were used at 500 Erg of protein per ml (data not shown). Pituitary extract did not stimulate 1251-Udr incorporation into Schwann cells in ~1% FCS, and the stimulation induced by the extract in the presence of 10% FCS was not appreciably different from that induced in the presence of 20 or 30% FCS. Thus although FCS was required for Schwann cell proliferation, even high concentrations of FCS did not substitute for pituitary extract.
Failure of Known Growth Factors and Pituitary Hormones to Stimulate Schwann Cell Proliferation Nerve growth factor (NGF), fibroblast growth factors (FGF) purified from bovine pituitary and bovine brain (Gospodarowicz, 1975, 1976), and epidermal growth factor (EGF) failed to stimulate 1251-Udr incorporation into Schwann cells when added at 10 ng, 100 ng, 1 pg and 10 kg/ml (Table 3). FGF and EGF did stimulate lz51-Udr incorporation into serum-starved BALB/c3T3 cells with plateau stimulation at 10 rig/ml (data not shown). Six different anterior pituitary hormones-luteinizing hormone (LH), luteotrophic hormone (LTH), follicle-stimulat-
56
1006 r 110 (3.5) 1485 k 249 (5.2)
500
3330 r
38 (11.7)
372 r 199 (1.3)
20
413 +
89 (1.5)
449 5
54 (1.6)
500
352 2
12 (1.2)
1000
264 +
63 (0.9)
20
470 f
166 (1.6)
Kidney
100
508k
500
458 + 132 (1.6)
1000 Nerve
82(1.8)
208 f
137 (0.7)
20
218 2
20 (0.8)
100
215 -c
21 (0.8)
500
220 +
31 (0.8)
1000
162 r
63 (0.6)
20
Brain
by
Incorporation
100
1 .w....!t’ Pituitary
Incorporation
100
1000
1
lZSI-Udr (cpm) 284%
20
Liver
on ?-Udr
416 5
95 (1.5)
100
547 I
74 (1.9)
500
1017 2
65 (3.5)
loo0
1142 4 151 (4.0)
The numbers in parentheses stimulation indices (= cpm
in this and subsequent tables with stimulanbcpm in controls).
are the
ing hormone (FSH), thyroid-stimulating hormone (TSH), growth hormone (GH) and adrenocorticotropic hormone (ACTH)-and two posterior pituitary hormones-oxytocin and vasopressin - (all obtained from Sigma) were unable to stimulate 1251Udr incorporation in Schwann cells when added at 50 ng, 500 ng, 5 pg and 50 pg/ml (Table 4). The same was true for prolactin, GH, LH and FSH obtained from the Pituitary Hormone Distribution Program of the National Institute of Arthritis, Metabolism and Digestive Diseases, although their preparation of LH stimulated weakly (stimulation index ~2.0) when used at 5 and 50 pg/ml (Table 4). The high concentration required suggested that the stimulation may have been related to a contaminant rather than to the LH itself. The divalent cation carrier A23187 did not stimulate Schwann cell proliferation when used at 1O-g-1O-5 M (data not shown).
Synergistic Effects of Tissue Extracts and Cholera Toxin in Stimulating Schwann Cell Proliferation When
either
cholera
toxin
or pituitary
extracts
were
Cell 818
Table 3. Failure Cell Growth
of Known
Growth
Concentration (fig of Protein per ml)
Factor 0
to Stimulate
Extract
Growth
20
Factor
Schwann
‘*%Udr Incorporation (cpm) 1255e
Pituitary
Nerve
Factors
127
2620 2 176 (2.0)
100
4196
500
6543 2 163 (5.0)
+ 241 (3.2)
0.01
1213 2
0.1
1294 + 174 (1.0)
91 (1.0)
1290 -c 49 (1 .O) Figure 3. Effect of Boiling and Proteolytic genie Activity of Pituitary Extract
Digestion
on the Mito-
Schwann cells were cultured with untreated pituitary extract (LO) or pituitary extract that had been boiled for 20 min (V-V), or treated with trypsin (e--e) or pronase (m-m). The ‘?-Udr incorporation is plotted as the stimulation index, and the concentrations of the extracts are Hg of protein per ml.
10 Pituitary Growth
Fibroblast Factor
921 k 102 (0.7)
0.001
1259 * 159 (1 .O)
0.01
1475 2 210 (1 .l)
0.1
1708k
58(1.3)
1781 + 161 (1.4)
used at their optimal stimulatory concentrations, they induced Schwann cells to double their numbers in 2-3 days, compared with 8 days for control cultures; when both were added together, the doubling time was further reduced to under 2 days (Figure 4). The same synergistic effect was seen when Schwann cell proliferation was assayed by 3H-thymidine incorporation determined by autoradiography (Table 5) or by 1251-Udr incorporation (Table 6 and Figure 5). Remarkably, extracts of liver, kidney and brain were able to act synergistically with cholera toxin as well as extracts of pituitary, while extracts of nerve roots did so only weakly (Table 6). Choleragenoid, the biologically inactive subunit of cholera toxin which binds equally well to the same ganglioside receptor, G,,, , (Finkelstein, 1973; Bennet and Cuatrecasas, 1977), did not stimulate on its own or act synergistically with the tissue extracts (data not shown). While dibutyryl cyclic AMP did act synergistically with tissue extracts in some experiments, its synergistic activity was less striking and less consistent than that of cholera toxin. It is interesting that when cholera toxin (10 pg/ml) and dibutyryl cyclic AMP (5 x lo+ M) were used together, the stimulatory effect was consistently less than with either used alone (data not shown). Heating the pituitary extract at 65°C for 10 min reduced its synergistic activity much more than its direct stimulatory activity (Figure 5). The poor synergistic activity of heated pituitary extract was not due to the presence of a heat-induced inhibitor, since mixing it with unheated pituitary extract did not affect the synergistic activity of the latter (Figure 5). Boiling the pituitary extract for 20 min completely destroyed its ability to act synergistically with cholera toxin, while treatment with trypsin or pronase re-
Brain Fibroblast Growth Factor
0.001
1067 r
0.01
1283 + 102 (1 .O)
0.1
1253 k 148 (1 .O) 1380 t
Epidermal Factor
94 (0.8)
97 (1 .I)
Growth 0.001
1233 + 206 (1 .O)
0.01
1289 k 124 (1.0)
0.1
1447 + 103 (1.2)
1
1376 -t 146 (1 .l)
duced but did not abolish its synergistic activity. The synergistic activity of the tissue extracts was probably not related to the extracts’ preventing the breakdown of cholera toxin or increasing the binding of cholera toxin to the Schwann cell surface, since the effect was similar whether cholera toxin was used at 0.01 or 10 pg/ml, and the tissue extract could not be replaced by pretreating the cells with exogenous G,,, (Bennet and Cuatrecasas, 1977). Moreover, synergy was also seen when Schwann cells were exposed to cholera toxin for 2 hr, washed and then incubated with tissue extract (data not shown). We tested a variety of other substances for their ability to act synergistically with cholera toxin in stimulating ‘251-Udr incorporation into Schwann cells, and none were active. These included all the pituitary hormones and all the growth factors discussed above. Intracellular Cyclic AMP Levels Cyclic AMP levels were measured in Schwann cells after 2, 6 and 20 hr of exposure to control medium (DMEM-FCS) or to medium with added pituitary
Schwann 017
Table
Cell Growth
4. Failure
Factors
of Pituitary
Hormones
to Stimulate ‘*%Udr
Hormone Experiment
Pituitary
Extract
Luteotrophic Sigma)
Hormone
incorporation
(cpm)
0
0.05 pg/ml
310+6rl
-
-
-
-
-
411 f 65 (1.3)
-
343?53(1.1)
-
at:
0.5 pg/ml
5 rglml
50 pglml
100 pg/ml
-
-
-
-
2391 + 915 (7.7)
363 + 26 (1.2)
360 2 97 (1.2)
282 r 37 (0.9)
360 + 42 (1 .I)
393 A 36 (1.3)
280 + 47 (0.9)
259 2 25 (0.8)
371 2 26 (1.2)
362 -t 117(1.2)
202 -c 11 (0.7)
313 + 44 (1 .O)
330 2 29 (1 .O)
262 2 38 (0.8)
291 ? 95 (0.9)
296 e 51 (1 .O)
314 c 25 (1 .O)
359+
343 2 56 (1 .l)
366 f 47 (1.2)
353 I 61 (1 .l)
381 k 37 (1.2)
(LTH;
Follicle-Stimulating (FSH; Sigma)
Hormone
Thyroid-Stimulating (TSH; Sigma)
Hormone
Hormone
(GH; Sigma)
Adrenocopticotropic (ACTH; Sigma)
Hormone
Vasopressin Experiment
(Sigma)
le(l.1)
416 2 41 (1.3)
2
0
646~
Pituitary
Extract
Prolactin Growth
Cell Proliferation
1
0
Growth
Schwann
-
-
(NIAMDD) Hormone
72
(GH; NIAMDD)
Follicle-Stimulating (FSH; NIAMDD)
-
-
-
-
506 t 82 (0.8)
479 2 40 (0.7)
543 f 59 (0.8)
596 f 166 (0.9)
-
466 ? 77 (0.7)
465 zt 95 (0.7)
547 t 34 (0.8)
580 k 26 (0.9)
-
695?82(1.1)
626 r 99 (1 .O)
683 + 66 (1 .l)
610 r 81 (1.0)
656 2 57 (1 .O)
746 + 37 (1.2)
119k
456 f 51 (0.7)
631 ‘- 140 (1 .O)
610 -t 26 (0.9)
2056 -t 50 (3.2)
Hormone
Luteinizing Hormone (LH; NIAMDD) Luteinizing Hormone (LH; Sigma)
-
-
extract (100-400 pg/ml), cholera toxin (10 pglml) or both. In eleven separate experiments, cholera toxin consistently increased cyclic AMP levels 4-30 fold at the three time points (the increase usually being greatest at 6 hr), while pituitary (or brain) extract did not have a significant effect even in the presence of the potent phosphodiesterase inhibitor 3,isobutyL1-methyl-xanthine (IBMX; 3 x 10m5 M) (Matsuzawa and Nirenberg, 1975). When Schwann cells were exposed to cholera toxin and pituitary extract together, the rise in cyclic AMP was not significantly different from that seen with cholera toxin alone. The results of a typical experiment are given in Table 7; since we had previously shown that prostaglandin E, and isoproterenol were not mitogenic for cultured Schwann cells (Raff et al., 1978), it was interesting to observe that these reagents did not increase cyclic AMP levels in these cells when assayed at 5 min or at 2,6 and 20 hr (Table 7). Discussion As we (Raff et al., 1978) and other investigators (Wood and Bunge, 1975) have previously observed, purified Schwann cells normally grow very slowly
124(1.9)
1239 -c 117 (1.9) 526 -t 71 (0.8)
in cell culture in conventional serum-containing medium. Since their growth rate under these conditions is essentially linear when plotted on semilog coordinates (Figures 1 and 4; Raff et al., 1978), it is probable that these cells constitute a homogeneous population of infrequently dividing cells rather than a mixture of rapidly dividing and nondividing cells. We found that crude extracts of bovine pituitary and agents which would be expected to increase intracellular levels of cyclic AMP (such as cholera toxin and dibutyryl cyclic AMP) (Raff et al., 1978) were potent Schwann cell mitogens. Since their mitogenic effects were demonstrated on highly purified populations of Schwann cells (>99.5% pure) (Brockes et al., 1978; Raff et al., 1978), it is probable that they acted directly on Schwann cells rather than by inducing another cell type to secrete substances that are mitogenic for Schwann cells. In this study, we have further investigated the actions and interactions of these Schwann cell growth factors on purified rat Schwann cells in culture. While extracts of bovine (and rat) pituitary were always mitogenic for purified rat Schwann cells, similar extracts of bovine liver, kidney or nerve roots were not. Extracts of bovine brain (cerebral
Cell 818
Table 5. Synergy between Stimulating “H-Thymidine Pituitary Extract (Kg of Dry/Weight per ml)
4
DAYS
Figure 4. Synergy between Pituitary Extract Stimulating Schwann Cell Proliferation Schwann cells were cultured cholera toxin (10 rig/ml) (m-D), weight per ml) (A-A) or extract (*-*).
16
lo and Cholera
Toxin
in
in DMEM-FCS alone (u), pituitary extract (400 pg of dry both cholera toxin and pituitary
hemispheres) were mitogenic, but less so than pituitary extracts. Since mixing experiments showed that extracts of liver, kidney and nerve roots partially suppressed the mitogenic effect of pituitary extracts, it is difficult to determine whether their lack of mitogenic activity was related to the presence of inhibitors or to the absence of stimulatory factor(s). This is particularly so in the case of extracts from nerve roots, which were markedly inhibitory. Since fibroblast growth factor (FGF), which is isolated from bovine pituitary and brain (Gospodarowicz, 1975), has been reported to be mitogenic for glial cells (Westermark and Wasteson, 1975), we were interested in determining whether the Schwann cell stimulatory activity in our pituitary and brain extracts was due to FGF. This seemed improbable since our extracts from pituitary were more potent than those from brain, while the concentration of FGF in bovine brain has been reported to be 5-10 times greater than that in pituitary (Gospodarowicz, 1974). The possibility was formally excluded by showing that purified FGF from brain or pituitary (provided by Dr. D. Gospodarowicz) did not stimulate Schwann cell proliferation. Epidermal growth factor (EGF) and nerve growth factor (NGF) were also inactive in this respect.
Cholera MO-w
Toxin
Schwann
Cells
Toxin
in
Labeled
W
0
0
15*9
500
0
56 -c 5
0
10
55 2 7
500
IO
91 e 3
Pituitary extract and 3H-thymidine
1
Pituitary Extract and Cholera Uptake by Schwann Cells
and cholera toxin were added for a total of 48 hr, was present for the final 24 hr of the culture.
Since extracts from both anterior and posterior (including intermediate) lobes of pituitary were both mitogenic for Schwann cells, it seemed improbable that any of the classical pituitary hormones were responsible for this activity. We tested most of the anterior and posterior pituitary hormones, including LH, LTH, FSH, TSH, GH, ACTH, prolactin, oxytocin and vasopressin, and found that they do not stimulate Schwann cell growth. One preparation (but not another) of LH was weakly mitogenic when used at high concentration (25 pg/ml), suggesting that it might be contaminated with Schwann cell growth factor(s). It is clear that the mitogenic activity of our pituitary extracts could not be due solely to the presence of LH. Two small peptide hormones, neurotensin (Carraway and Leeman, 1976) and gastrin (Rehfeld, 1978), have recently been demonstrated in both anterior and posterior lobes of pituitary. It is improbable that these or other similar small peptides could account for the Schwann cell stimulatory activity in our pituitary extracts since the activity of these peptides is usually resistant to boiling, while the activity of our extracts was abolished by this treatment. The fact that the mitogenic activity was reduced by heating at 65°C for 10 min and by proteolytic digestion suggests that the active principle(s) in the pituitary extracts is a protein; further biochemical characterization and purification will be required, however, to determine its nature and relationship to known macromolecules. Although Schwann cells appeared to survive in 1% fetal calf serum (FCS), they did not proliferate in response to pituitary extract or cholera toxin in ~1% FCS, suggesting that FCS provided a factor(s) which was necessary for Schwann cell proliferation but which was not specific for the mitogenic stimulus. Although serum was required for Schwann cell growth, even high concentrations of FCS (up to 30%) could not substitute for pituitary or brain extract in stimulating this growth, suggesting that the stimulating factor(s) in these extracts was not present in mitogenic concentrations in FCS. It is surprising that sera from other species, including
Schwann 619
Cell Growth
Factors
Table 6. Synergy between Tissue Extracts and Cholera Stimulating ‘V-Udr Uptake by Schwann Cells
Tissue Extract (pg of Protein per ml) Pituitary
Brain
Figure Activity
5. Effect of Heating of Pituitary Extract
on
the
Mitogenic
and
Synergistic
Schwann cells were incubated in pituitary extract (V-V), pituitary extract that had been heated for 10 min at 65°C (+---+), cholera toxin (O-O), pituitary extract plus 10 rig/ml cholera toxin (LO), heated pituitary extract plus cholera toxin (m-m), cholera toxin plus 60 kg/ml each of pituitary extract and heated pituitary extract (O), or cholera toxin plus 250 Kg/ml each of pituitary extract and heated pituitary extract (*). The ‘YUdr incorporation is plotted as the stimulation, index and the concentrations of extracts are calculated as pg of protein per ml.
adult bovine, neonatal and adult rat, adult horse and adult human sera, did not support pituitary extract (or cholera toxin)-induced Schwann cell growth. Moreover, when FCS was mixed with any of these sera, it no longer supported Schwann cell proliferation, suggesting that these sera were actively inhibitory rather than deficient in an essential nutrient. Wood and Bunge (1975) have reported that
