freshly isolated monocytes, TGF-,8 mRNA was readily de- tected in both control (lanes 1-3) and PMA-treated (lanes. 4-6) U937 cells. We did detect a 3- to 5-fold ...
Proc. Natl. Acad. Sci. USA Vol. 84, pp. 6020-6024, September 1987 Biochemistry
Expression and secretion of type 13 transforming growth factor by activated human macrophages (monocytes/differentiation/wound repair/atherogenesis)
RICHARD K. ASSOIAN*t, BARBARA E. FLEURDELYS*, HENRY C. STEVENSONt, PAUL J. MILLER:, DAVID K. MADTES§, ELAINE W. RAINES§, RUSSELL Ross§, AND MICHAEL B. SPORN* *Laboratory of Chemoprevention, National Cancer Institute, Bethesda, MD 20892; tBiological Therapeutics Branch, National Cancer Institute-Frederick Cancer Research Facility, Frederick, MD 21701, and §Department of Pathology, University of Washington School of Medicine, Seattle, WA 98195
Communicated by Arno G. Motulsky, March 16, 1986
onset of a fibrotic response (19, 20), is a paracrine source of growth factors. This paracrine role of macrophages during wound repair in vivo can be demonstrated in studies that show that fibrosis is suppressed when monocyte infiltration is blocked (21). Since activated macrophages are known to secrete PDGF (22-24), the present studies were designed to see if macrophages might also be a paracrine source of TGF-P. Both PDGF and TGF-,3 may have important roles in stimulating smooth muscle cell proliferation in atherogenesis (25) as well as in fibroblastic proliferation in connective tissue repair (13, 16).
ABSTRACT Alveolar macrophages activated with concanavalin A and peripheral blood monocytes activated with lipopolysaccharide secrete type fi transforming growth factor (TGF-j6). There is minimal TGF-/3 secretion in unactivated monocytes, even though TGF-4 mRNA is expressed in these cells at a level similar to that in activated, lipopolysaccharidetreated cultures. U937 lymphoma cells, which have monocytic characteristics, also express mRNA for TGF-j3. Freshly isolated monocytes, both control and lipopolysaccharide-treated, secrete an acid-labile binding protein that inhibits TGF-,B action. We conclude the following: (i) that expression of TGF-1B mRNA is unrelated to monocyte activation, (ii) that secretion of TGF-fJ is induced by monocyte activation, and (iii) that cosecretion of TGF-13 and its monocyte/macrophage-derived binding protein may modulate growth factor action. In contrast, monocytic expression of other growth factor genes, such as the B chain of platelet-derived growth factor, is not constitutive and requires activation.
MATERIALS AND METHODS Isolation and Activation of Human Alveolar Macrophages. Alveolar macrophages were obtained by bronchoalveolar lavage of normal nonsmoking human volunteers; samples typically contained >90% alveolar macrophages, -8% lymphocytes, and 95% of adherent cells. Conditioned medium was collected every 24 hr for 4 days and pooled. Isolation and Activation of Human Peripheral Blood Monocytes. Monocytes were isolated from normal donors, following informed consent, by standard methods (27, 28); purity was >92% as determined by nonspecific esterase staining, phagocytosis of latex beads, and overall morphology after Wright's staining; viability was >99% (trypan blue). They were cultured in serum-free medium in Teflon-coated vessels (108 cells per 100 ml) as described (28). Parallel cultures were incubated in the absence and presence of lipopolysaccharide (LPS, 10 ug/ml) for 24 hr. After incubation, cells and conditioned medium were separated by gentle centrifugation and used for RNA isolation and analysis of TGF-,3 secretion, respectively (see below). Culture and Phorbol Ester Treatment of U937 Cells. Human U937 cells (29) were cultured in RPMI 1640 medium containing 10% (vol/vol) heat-inactivated fetal bovine serum.
Type , transforming growth factor (TGF-,l) (1-5), a peptide widely distributed in tissues of humans and other animals, has the unusual ability both to stimulate and inhibit the proliferation of cells in culture (6-10). Although the effect of TGF-,B on epithelial cells is consistently inhibitory (6, 7, 9), its role in mesenchymal cell proliferation is more complex (7, 8, 11). For example, TGF-,3 acts synergistically with epidermal growth factor (EGF) and platelet-derived growth factor (PDGF) to stimulate mitosis of normal rat kidney (NRK) and smooth muscle cells cultured in soft agar yet antagonizes the effect of these mitogens by inhibiting mitosis of the same cells cultured as subconfluent monolayers (7, 8, 12, 13). Although the proliferative effects of TGF-f3 on cultured mesenchymal cells are bifunctional, studies in whole animals have suggested a stimulatory role for this growth factor. For example, exogenous addition of TGF-f3 promotes wound repair in rat model systems (14, 15), and subcutaneous injection of TGF-,3 into newborn mice results in collagen deposition and a localized fibrotic response (16). The potential role of TGF-f3 as an endogenous mediator of tissue repair is further supported by in vitro studies showing that the purified growth factor stimulates collagen and fibronectin formation in cultures of primary and established fibroblasts (16, 17). TGF-P is present at sites of injury, since platelets contain relatively large amounts of this growth factor (3) and their degranulation with thrombin releases immunoreactive TGF-,3 (8). A second cell type that acts as an important source of growth factors during tissue repair in vivo is the macrophage (18). This cell, which appears in healing wounds prior to the
Abbreviations: LPS, lipopolysaccharide; TGF-,3, type / transforming growth factor; PDGF, platelet-derived growth factor; PMA, phorbol 12-myristate 13-acetate; EGF, epidermal growth factor. tPresent address: College of Physicians and Surgeons, Columbia University, New York, NY 10032.
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
6020
Proc. Natl. Acad. Sci. USA 84 (1987)
Biochemistry: Assoian et al. Cells (0.5 x 106 cells per ml) were incubated with phorbol 12-myristate 13-acetate [PMA, LC Services (Woburn, MA), 150 nM] in T-150 flasks (50 ml per flask, 3-5 flasks per experiment) for 24-48 hr to induce the macrophage-like phenotype of adherence (30). The medium containing nonadherent cells was collected, the adherent cells were washed twice with cold Dulbecco Vogt PBS, and the washed cells were scraped into fresh PBS. B5oth cell populations were collected by gentle centrifugation. Preparation of Monocyte Conditioned Medium. For bioassays, conditioned medium (50-100 ml) from control and activated monocytes/macrophages was made 1 M in acetic acid, dialyzed against two changes of 1 M acetic acid, and freeze-dried. The residue was dissolved in 0.02 vol of 4 mM HCI and dialyzed against two changes of 4 mM HCI prior to use. For receptor binding assays, some conditioned media were transiently exposed to mild acid by adding 25 pl of 5 M HCl to 1-ml portions of conditioned medium and incubating the solution for 2 hr at room temperature. The acidified medium was neutralized prior to assay with 35 Al of 0.7 M Hepes (pH 7.0) supplemented with 1.4 M NaOH. Samples of media to be assayed without acidification were supplemented with 0.33 M Hepes, pH 7.0/1.7 M HCI/1.7 M NaOH (60 p1/ml of conditioned medium) to standardize the salt composition. Isolation of RNA and RNA Gel Blot Analysis. Total RNA was isolated from cells using the guanidine hydrochloride procedure (31). Isolation of RNA from the fresh monocytes required prolonged (4-hr) precipitation of RNA at -70'C at each stage of the procedure. Yield of RNA was -50 ,ug per 108 cells. It was subjected to electrophoresis on 0.75% agarose gels containing 1.1 M formaldehyde and ethidium bromide at 1 ,4g/ml of gel, with 10 mM sodium phosphate, pH 7.2, as buffer. Fractionated RNA was transferred electrophoretically to Nytran filters (Schleicher & Schuell), and the filter was dried at 80TC. The filters were prehybridized (3 hr at 420C) in 10 ml of 0.1% polyvinylpyrollidone/0.1% Ficoll/0.1% bovine serum albumin/0.1% NaDodSO4. Hybridization to nick-translated TGF-,8 (32) or primer-extended H2 (33) cDNA probes (107 and 106 cpm, respectively) proceeded overnight at 42TC in 10 ml of a fresh solution of the above, also containing 10% (wt/vol) dextran sulfate; probes were 32P-labeled (34, 35). All filters were washed (four times) at 420C. Filters incubated with the TGF-.8 probe were additionally washed twice at 680C. All washes were in 0.2x SSC/0.1% NaDodSO4. (lx SSC = 0.15 M NaCl/0.015 M sodium citrate, pH 7.0.) Assays for Growth Factors. Colony formation (1, 3) and mitogen assays (36) were performed in NRK 49F cells. The radioreceptor assay for TGF-,8 was performed under both "simultaneous" and "sequential" protocols. In both protocols A549 cells (2 x 105 cells per well, 24-well plate) were incubated overnight at 370C in Dulbecco's modified Eagle's medium (DMEM) containing 5% (vol/vol) fetal bovine serum and washed twice with DMEM, human serum albumin at 1 mg/ml prior to use. In the simultaneous protocol, the washed cells were incubated (2 hr at 230C in an atmosphere of 5% C02/95% air) with 0.2 ml of RPMI 1640 with human serum albumin at 1 mg/ml, containing selected volumes of conditioned medium and 105 cpm of 1251-labeled TGF-,8 (37). In the sequential protocol, cells were incubated (2 hr as described above) with 0.2 ml RPMI 1640 with human serum albumin at 1 mg/ml containing selected volumes ofconditioned medium, the monolayers were washed twice with ice-cold PBS containing bovine serum albumin at 1 mg/ml, and the washed cells were incubated for an additional 2 hr as described above with '251-labeled TGF-p (105 cpm in 0.2 ml of RPMI 1640 with human serum albumin at 1 mg/ml). After the incubation with radiolabel, cells in both protocols were washed four times with ice-cold PBS containing bovine serum albumin at 1
6021
mg/ml and extracted (30 min at 370C) with 0.5 ml of 10 mM
Hepes, 10% (vol/vol) glycerol, and 1% Triton X-100 (pH 7.0).
TGFU-j
levels were measured in these extracts with a known standard (3, 38).
RESULTS TGF-f Biological Activity and Receptor Binding Activity in
Activated Alveolar Macrophages and Blood Monocytes. Initial studies focused on secretion of TGF-,B by freshly isolated macrophages. Alveolar macrophages and peripheral blood monocytes were activated in vitro by culture with concanavalin A and LPS, respectively. Conditioned medium from these cell cultures was analyzed for TGF-,8, using stimulation of NRK cell colony formation as a bioassay. The results show that activated alveolar macrophages (Fig. 1A) and activated peripheral blood monocytes (Fig. 1B) secrete significant amounts of TGF-13. As expected from the well-documented synergism of TGF-,3 and EGF in this NRK cell assay (12), the conditioned media from these cultures required EGF for efficient colony formation (data not shown). The ability to isolate unactivated peripheral blood monocytes and then to activate these cells in vitro by exposure to LPS allowed us to compare the secretion of TGF-f3 by these two cell populations. Fig. 1B shows that peripheral blood monocytes secrete almost no TGF-,3 in culture and that exposure of these cells to LPS results in a 10-fold increase in measurable TGF-,8 in the conditioned medium within a 24-hr period. Based on a standard curve for TGF-B in the colony formation assay, we estimate that the conditioned medium (100 ml) from LPS-treated monocytes (108 cells) contains 20 pM TGF-,8 (determined in two separate experiments). The inducible secretion of TGF-f3-like biological activity during monocyte activation with LPS was paralleled by a similar increase in the level of mitogenic activity as tested on NRK fibroblasts (Fig. 1C). However, TGF-f3 is not mitogenic for NRK cells under the assay conditions used (36), suggesting that the observed increase in mitogenic activity is due to secretion of other growth factors by the activated macrophages. In fact, this interpretation agrees well with studies (22-24) showing a large increase in secretion of a PDGF-like peptide upon activation of alveolar and peritoneal macrophages. To confirm that TGF-,8 is secreted during monocyte activation with LPS, we examined the ability of conditioned medium from control and LPS-treated cells to compete for radioiodinated TGF-,B binding in a specific radioreceptor assay (Table 1, experiment 1). The results show that acidified conditioned medium from control monocytes minimally inhibits the binding of 1251-labeled TGF-P to its cell surface receptor, while acidified medium from LPS-activated monocytes is a more effective competitor, containing 15-40 pM TGF-,B (determined in four separate experiments). LPS itself does not affect the binding of TGF-p to its receptor (Table 1, experiment 2). The bioassay (Fig. 1) and receptor binding assay (Table 1) give similar values for the concentration of TGF-,B in conditioned medium from activated cells. These results demonstrate that exposure of monocytes to LPS in vitro causes inducible and coordinated secretion of TGF-/3 and other mesenchymal mitogens. TGF-fi Binding Protein Secreted by Monocytes. In our initial experiments on growth factor secretion from activated monocytes and macrophages, we noticed that TGF-p biological activity was not detected in conditioned medium unless they were briefly exposed to mild acid (pH 2-3, data not shown). This result suggested that the conditioned medium contained an acid-labile protein that was binding radioiodinated TGF-,8 and, thereby, inhibiting its biological activity. Such an interpretation is consistent with the results of others (39, 40) show-
Biochemistry: Assoian et al.
6022
Proc. Natl. Acad. Sci. USA 84 (1987) Table 1. Analysis of conditioned medium from control and LPS-treated monocytes by radioreceptor assay for TGF-f3 Medium cpm bound % competition TGF-, pM Exp.
C,) w
z
0 -j 0
I
U-
w m z
CU) w
z
0 -O 0 -~ Ofd
m z0
~0 2
-c
8Z;;-
C
Z x
5, X
0
0
conditioned medium was incubated with cells in the absence of radioiodinated TGF-,8, the cells were washed to remove unbound material, labeled TGF-f3 was incubated- with the washed cells, and finally, the amount of bound radiolabel was determined. Assayed under these conditions, a binding protein would not be exposed to labeled TGF-13 and should not inhibit ligand-receptor interactions. Fig. 2A shows that unacidified conditioned medium was an effective inhibitor of I251-labeled TGF-f3 binding when assayed in the simultaneous protocol and that this inhibitory activity could be destroyed by transient exposure of the conditioned medium to mild acid. These results could be expected if an acid-labile binding protein were present in the conditioned medium. Unacidified conditioned medium did not effectively inhibit the binding of I251-labeled TGF-f3 to cells when tested in the sequential protocol (Fig. 2B). Under these conditions, unacidified and transiently acidified condi-
'x
'o
3832
2.8 3132 18 19 1919 50 2 4992 0 0 Fresh 5043 0 0 Fresh/LPS Conditioned medium (0.2 ml) from control (CM) and LPS-treated (CM/LPS) monocytes was transiently acidified and tested for TGF-3, in a radioreceptor assay performed under the sequential protocol. The results are presented as specific (total minus nonspecific) binding. Nonspecific binding was determined in the presence of 10 nM unlabeled TGF-43 (2000 cpm for experiment 1 and 1100 cpm for experiment 2). Maximal specific binding of l25l-labeled TGF-,/ was determined with fresh medium in the absence of LPS (Fresh). LPS was used at 10 p&g/ml.
ox c
Fresh Cm CM/LPS
1
sb
--_____
I 100
101
102
103
100
VOLUME OF CONDITIONED MEDIUM (gAl) FIG. 1. Release of TGF-/3 by activated monocytes and macrophages in culture. Conditioned medium from cultures of alveolar macrophages (A) and peripheral blood monocytes (B and C) were collected, acidified, and concentrated for bioassay. Aliquots of the concentrated media were assayed for their ability to stimulate colony formation (A and B, diameter >62 gm) and [3H]thymidine incorporation (C) in NRK fibroblasts. Solid and open symbols show results obtained with conditioned medium from activated and control cell cultures, respectively. Colony formation was always measured in a saturating concentration of exogenous EGF (0.5 nM). This concentration of EGF and an optimal concentration (20 pM) of purified TGF-,8 (36) induced formation of 2000-2500 colonies (diameter >62 ,um). In the mitogen assay, [3H]thymidine incorporation was 0.5 x 105 and 2 x 105 cpm for the negative (medium alone) and positive [5% (vol/vol) fetal calf serum] controls, respectively. Results are shown as volume of 1x conditioned medium that would be required to induce the observed effects.
z c
z
50
-J
x LL
0
25. 100
I-
I
i
z
wL
0=wL
o
B-
o
-_Q
I-
751 50C 1) r
4b~1
ing that many cells secrete TGF-f3 in a latent form, which be activated by exposure to mild acid. To demonstrate that monocytes secrete a binding protein for TGF-,8, we compared the ability of neutral and transiently acidified conditioned medium from unactivated monocytes to inhibit the binding of TGF-,3 in receptor binding assays performed under simultaneous and sequential protocols. In the simultaneous protocol, conditioned medium and radiolabeled TGF-f3 were added concomitantly to cells; under these conditions a binding protein could form a complex with labeled TGF-,3 and inhibit binding of the labeled growth factor to its cell-surface receptors. In the sequential protocol,
A
N
75
2
5
10 20
50 100
VOLUME OF CONDITIONED MEDIUM
200
(MI)
can
FIG. 2. Detection of a binding protein for TGF-,3 in monocyte conditioned medium. Conditioned medium from cultures of peripheral blood-derived monocytes was assayed for its ability to inhibit the binding of I251-labeled TGF-,B to its plasma membrane receptor on intact A549 cells. The medium was assayed without concentration at pH 7.4 either directly (open circles) or after treatment with mild acid (solid circles). Results are from assays performed under simultaneous (A) and sequential (B) protocols. Maximal specific binding of 1251-labeled TGF-,f in this experiment was 5000 cpm. All data are corrected for nonspecific binding (1090 cpm, determined in the presence of 10 nM unlabeled TGF-/3).
Biochemistry: Assoian et al.
Proc. Natl. Acad. Sci. USA 84 (1987)
tioned media were equipotent in their ability to inhibit receptor binding of radioiodinated TGF-,B. The small amount of inhibitory activity in the conditioned medium from unactivated monocytes likely represents the level of active TGF-,3 released from these cells. The results of Fig. 2 show that monocytes secrete an acid-labile binding protein for TGF-,3 and that the complex of this protein with the growth factor effectively prevents TGF-,8 binding to its cell-surface receptor. Similar experiments performed on the conditioned medium from LPS-treated cultures showed that activated monocytes also produce this binding protein. Constitutive Expression of TGF-8 mRNA in Monocytes and U937 Cells. We continued our studies of TGF-p secretion by examining expression of TGF-/3 mRNA in cultures of control and LPS-treated peripheral blood monocytes. Total RNA from control and LPS-treated cells was fractionated on agarose gels containing formaldehyde, transferred to Nytran filters, and hybridized to a nick-translated cDNA for TGF-3. Fig. 3 (Upper) shows that the expected 2.5-kilobase mRNA for TGF-f3 was present in the control (lanes 1-3) as well as LPS-treated (lanes 5-7) cell cultures. Moreover, the level of TGF-f3 mRNA in the two cell populations was similar and within a factor of 2; this result was confirmed by dot-blot analysis (data not shown). The cells used for this RNA gel/blot and the conditioned media used to examine TGF-f3 secretion (Fig. 1) were derived from the same experiment. Although unactivated monocytes expressed TGF-p mRNA, they did not secrete the peptide itself. The results preclude the possibility that expression of TGF-p mRNA in the control monocyte culture was due to partial activation of the cells during isolation and suggest that expression of TGF-p mRNA is constitutive, whereas secretion of the growth factor itself is inducible. The somewhat unusual combination of inducible protein secretion and constitutive mRNA expression for TGF-f3 in monocyte activation prompted us to examine TGF-p mRNA levels in U937 cells (29, 30), an established monocyte-like cell line that can be induced to acquire a limited macrophage phenotype by exposure to agents such as PMA. U937 cells -28S 'p -
18S
-28S
-
1
2
3
4
5
6
18S
7
FIG. 3. Expression of TGF-,8 mRNA in monocytes and macrophages. Total RNA, isolated from control and LPS-treated monocytes, was fractionated on formaldehyde/agarose gels, transferred to Nytran filters, and hybridized to a cDNA probe for TGF-,8 (Upper). Washed filters were exposed to film with intensifying screens for 3-10 days at -70°C. Lanes contain 1, 3, and 10 jig of total RNA from control (lanes 1-3) and from LPS-treated (lanes 5-7) monocytes. Lane 4 shows nonspecific binding of the labeled probes; it contains 10 ug of autologous lymphocyte RNA that had been depleted of mRNA by repeated passage over a column of oligo(dT)-cellulose. (Lower) Hybridization of the RNA to an H2 probe, demonstrating that similar levels of RNA were fractionated in corresponding lanes. The positions of 18 and 28S rRNA were determined by ethidium bromide staining.
6023
-28S
_q
- 18S
-28S p
w*.
1 2 3 4 5 6 7
18S
-
8 9 10
FIG. 4. Expression of TGF-P mRNA in U937 cells. Total cellular RNA from control and PMA-treated U937 cells was subjected to RNA gel blot analysis as described in the legend to Fig. 3. Filters were hybridized to 32P-labeled cDNA probes for TGF-,f (Upper) and H2 (Lower). Lanes contain 1, 3, and 10 ,g of total cellular RNA from control U937 cells (lanes 1-3), from PMA-treated, adherent U937 cells (lanes 4-6), and from PMA-treated cells that had yet to adhere (lanes 8-10). Lane 7 (nonspecific binding control) contains 10 ,g of fractionated control U937 RNA that had been depleted of mRNA by repeated passage over oligo(dT)-cellulose. (Lower) Hybridization of the RNA to an H2 probe, demonstrating that similar levels of RNA were loaded in corresponding lanes. The positions of 18 and 28S rRNA were determined by ethidium bromide staining.
were incubated in the absence (control) and presence of PMA until most of the PMA-treated culture became adherent (40-48 hr). Total RNA was prepared from these two cell populations and expression of TGF-,3 mRNA was examined by RNA gel/blots (Fig. 4). Consistent with our results in freshly isolated monocytes, TGF-,8 mRNA was readily detected in both control (lanes 1-3) and PMA-treated (lanes 4-6) U937 cells. We did detect a 3- to 5-fold increase in the level of TGF-f3 mRNA in the PMA-treated U937 cells (which was not observed after LPS-treatment of fresh monocytes), but control studies showed that PMA-treated U937 cells that did not attach still expressed the higher level of TGF-,/ mRNA (lanes 8-10). Thus, the PMA-induced increase in TGF-f3 mRNA expression in this system is not directly associated with activation of U937 cells into an adherent, macrophage-like population.
DISCUSSION The studies reported here demonstrate that TGF-,rmRNA is expressed at similar levels in unstimulated monocytes and in monocytes activated to become macrophages. However, TGF-,8 itself is only secreted by activated macrophages. Thus, secretion of this growth factor is similar to that observed for the PDGF-like protein produced by these cells (22-24), despite the fact that the mechanisms controlling secretion of TGF-,3 and PDGF-like protein during monocyte activation differ. Other studies (22-24) have demonstrated that c-sis mRNA is present in macrophages but not in unstimulated monocytes. Secretion of PDGF-like protein by macrophages thus appears to be controlled, at least in part, by induction of its mRNA. The control mechanism responsible for induction of TGF-/3 secretion during monocyte activation does not appear to act at the level of RNA since mRNA levels for TGF-,f are similar in unstimulated and activated monocytes. These results indicate that different mechanisms regulate gene expression and secretion of macrophage growth factors. Our mRNA expression studies with U937 cells support the results we obtained with freshly isolated monocytes. TGF-13 mRNA is present in control as well as PMA-treated U937 cells, and the increase in the level of TGF-p mRNA observed after PMA-treatment was not
6024
Biochemistry: Assoian et al.
associated with the macrophage phenotype of adherence. These data identify a monocyte-like cell line displaying a phenotype very similar to that of fresh monocytes with regard to expression of TGF-13 mRNA. The results reported here also describe a mechanism controlling TGF-,B action that operates subsequent to growth factor secretion. As observed in many other cell types, the monocyte/macrophage produces a binding protein for TGF,3 as well as the growth factor itself. The binding protein effectively inhibits the interaction between TGF-,8 and its cell-surface receptor; this result agrees with studies by others showing that the TGF-,8-binding protein complex is biologically inactive (39, 40). Although the initial studies with transforming growth factors were based on the premise of their being cancerspecific peptides, the identification of TGF-03 in a large number of normal tissues and cells (1-5) strongly suggests that this protein has a fundamental regulatory role in human physiology. It is now clear that effector cells involved in inflammation and repair (platelets, macrophages, and lymphocytes) are all sources of TGF-,B (refs. 2-4 and this report). With the stimulatory effects of TGF-/3 on formation of collagen and fibronectin (16, 17), these data strengthen the suggestion that TGF-pB may play an important role as an endogenous mediator of tissue repair. Moreover, the effects of TGF-p8 on proliferation of aortic smooth muscle cells in culture (8) and the likely involvement of platelets, macrophages, and T cells in atherogenesis (25) and several other proliferative diseases (41) all suggest that TGF-,3 may be of importance in many diseases other than cancer. We thank Gilbert Jay for the H2 cDNA probe 12A and Sue Perdue for help with the manuscript. D.K.M. is a fellow of the Parker B. Francis Foundation. This work was supported in part by grants from the Public Health Service (HL-18645) and from R. J. R. Nabisco. 1. Roberts, A. B., Anzano, M. A., Lamb, L. C., Smith, J. M. & Sporn, M. B. (1981) Proc. Natl. Acad. Sci. USA 78,
5339-5343. 2. Childs, C. B., Proper, J. A., Tucker, R. F. & Moses, H. L. (1982) Proc. Natl. Acad. Sci. USA 79, 5312-5316. 3. Assoian, R. K., Komoriya, A., Meyers, C. A., Miller, D. M. & Sporn, M. B. (1983) J. Biol. Chem. 258, 7155-7160. 4. Kehrl, J. H., Wakefield, L. M., Roberts, A. B., Jakowlew, S., Alvarez-Mon, M., Derynck, R., Sporn, M. B. & Fauci, A. S. (1986) J. Exp. Med. 163, 1037-1050. 5. Seyedin, S. M., Thompson, T. C., Bentz, H., Rosen, D. M., McPherson, J. M., Conti, A., Siegel, N. R., Galluppi, G. R. & Piez, K. A. (1986) J. Biol. Chem. 261, 5693-5695. 6. Tucker, R. F., Shipley, G. D., Moses, H. L. & Holley, R. W. (1984) Science 226, 705-707. 7. Roberts, A. B., Anzano, M. A., Wakefield, L. M., Roche, N. S., Stem, D. F. & Sporn, M. B. (1985) Proc. Natl. Acad. Sci. USA 82, 119-123. 8. Assoian, R. K. & Sporn, M. B. (1986) J. Cell Biol. 102, 1217-1223. 9. Masui, T., Wakefield, L. M., Lechner, J. F., LaVeck, M. A., Sporn, M. B. & Harris, C. C. (1986) Proc. Natl. Acad. Sci. USA 83, 2438-2442. 10. Leof, E. B., Proper, J. A., Goustin, A. S., Shipley, G. D.,
Proc. Natl. Acad. Sci. USA 84
11. 12. 13. 14.
15. 16.
17.
18. 19.
20. 21. 22. 23. 24. 25. 26. 27. 28.
29. 30. 31.
32. 33. 34.
35. 36. 37.
38. 39. 40. 41.
(1987)
DiCorleto, P. E. & Moses, H. L. (1986) Proc. Natl. Acad. Sci. USA 83, 2453-2457. Massagud, J., Kelly, B. & Mottola, C. (1985) J. Biol. Chem. 260, 4551-4554. Anzano, M. A., Roberts, A. B., Meyers, C. A., Komoriya, A., Lamb, L. C., Smith, J. M. & Sporn, M. B. (1982) Cancer Res. 42, 4776-4778. Assoian, R. K., Grotendorst, G. R., Miller, D. M. & Sporn, M. B. (1984) Nature (London) 309, 804-806. Sporn, M. B., Roberts, A. B., Shull, J. H., Smith, J. M., Ward, J. M. & Sodek, J. (1983) Science 219, 1329-1331. Lawrence, W. T., Norton, J. A., Sporn, M. B., Gorschboth, C. & Grotendorst, G. R. (1986) Ann. Surg. 203, 142-147. Roberts, A. B., Sporn, M. 13., Assoian, R. K., Smith, J. M., Roche, N. S., Wakefield, L. M., Heine, U. I., Liotta, L. A., Falanga, V., Kehrl, J. H. & Fauci, A. S. (1986) Proc. Natl. Acad. Sci. USA 83, 4167-4171. Ignotz, R. A. & Massagud, J. (1986) J. Riol. Chem. 261, 4337-4345. Nathan, C. F., Murray, H. W. & Cohn, Z. A. (1980) New Engl. J. Med. 303, 622-626. Ross, R. & Benditt, E. P. (1961) J. Biophys. Biochem. Cytol. 11, 677-700. Ross, R. (1986) Biol. Rev. 43, 51-96. Leibovich, S. J. & Ross, R. (1975) Am. J. Pathol. 78, 71-91. Shimokado, K., Raines, E. W., Madtes, D. K., Barrett, T. B., Benditt, E. P. & Ross, R. (1985) Cell 43, 277-286. Martinet, Y., Bitterman, P. B., Mornex, J.-F., Grotendorst, G. R., Martin, G. R. & Crystal, R. G. (1986) Nature (London) 319, 158-160. Mornex, J.-F., Martinet, Y., Yamauchi, K., Bitterman, P. B., Grotendorst, G. R., Chytil-Weir, A., Martin, G. R. & Crystal, R. G. (1986) J. Clin. Invest. 78, 61-66. Ross, R. (1986) N. Engl. J. Med. 314, 488-500. Martin, T. R., Altman, L. C., Albert, R. K. & Henderson, W. R. (1984) Annu. Rev. Respir. Dis. 129, 106-111. Stevenson, H. C. (1984) Methods Enzymol. 108, 242-249. Stevenson, H. C., Dekaban, G. A., Miller, P. J., Benyajati, C. & Pearson, M. L. (1985) J. Exp. Med. 161, 503-513. Sundstrom, C. & Nilsson, L. (1976) Int. J. Cancer 17, 565-577. Ralph, P., Williams, N., Moore, M. A. S. & Litcofsky, P. B. (1982) Cell. Immunol. 71, 215-223. Deeley, R. G., Gordon, J. I., Bums, A. T. H., Mullinix, K. R., Bina-Stein, M. & Goldberger, R. F. (1977) J. Biol. Chem. 252, 8310-8319. Derynck, R., Jarrett, J. A., Chen, E. Y., Eaton, D. H., Bell, J. R., Assoian, R. K., Roberts, A. B., Sporn, M. B. & Goeddel, D. V. (1985) Nature (London) 316, 701-705. Cosman, D., Khoury, G. & Jay, G. (1982) Nature (London) 295, 73-76. Ley, T. J., Anagnou, N. P., Pepe, G. & Nienhuis, A. W. (1982) Proc. Natl. Acad. Sci. USA 79, 4775-4779. Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY). Assoian, R. K. (1985) J. Biol. Chem. 260, 9613-9617. Frolik, C. A., Wakefield, L. M., Smith, D. M. & Sporn, M. B. (1984) J. Biol. Chem. 259, 10995-11000. Assoian, R. K. (1987) Methods Enzymol., in press. Lawrence, K. A., Pircher, R., Kryceve-Martinerie, C. & Jullien, P. (1984) J. Cell. Physiol. 121, 184-188. Pircher, R., Lawrence, D. A. & Jullien, P. (1984) Cancer Res. 44, 5538-5543. Sporn, M. B. & Harris, E. D., Jr. (1981) Am. J. Med. 70, 1231-1236.
