Chromatin Binding of Epidermal Growth Factor, Nerve Growth Factor

Home , Bar 2 (TV Poland), Epidermal growth factor

Proc. Natl. Acad. Sci. USA Vol. 83, pp. 3728-3732, June 1986 Biochemistry Chromatin binding of epidermal growth factor, nerve growth factor, and platelet-derived growth factor in cells bearing the appropriate surface receptors (intracellular distribution/DNase IU-sensitive sites) EWA M. RAKOWICZ-SZULCZYNSKA*, ULRICH RODECK, MEENHARD HERLYN, AND HILARY KOPROWSKIt The Wistar Institute of Anatomy and Biology, 36th Street at Spruce, Philadelphia, PA 19104 Contributed by Hilary Koprowski, February 5, 1986

ABSTRACT We analyzed the uptake and intracellular Table 1. Growth factor receptors on cell lines distribution of 12-5-labeled epidermal growth factor, nerve growth factor, and platelet-derived growth factor in different Growth Receptor cell lines that express or do not express the respective surface factor Present Not detected receptors for these factors. After 1 hr of incubation, all three PDGF WI-38*, AG 1523* WM 266-4t, 1810t, SK-MEL-21t growth factors were detected in the cytoplasmic fraction and in EGF SW948§ SW707§ the nucleus, tightly bound to chromatin. The amount of NGF Hs 294t, SW707§ SW948§, SW1116§ chromatin-bound growth factors continued to increase during *Normal human fibroblasts (6, 7). the incubation, and analysis at 48 hr revealed each chromatin- tHuman melanoma (15, 16). bound labeled growth factor in a nondegraded form. After tLung carcinoma (17). limited digestion ofchromatin with DNase 1 (10-20% digested §Human colorectal carcinoma (18, 19). sequences), specific release of all three growth factors was detected only after 1 hr of incubation but not after 24 and 48 lodination of Growth Factors. Mouse ,3-NGF was iodinated hr, suggesting that the DNA regions involved in growth factor by the lactoperoxidase method (21). Human recombinant binding became nuclease-resistant. Binding of labeled epider- EGF (purchased from AMGEN Biologicals, Thousand Oaks, mal growth factor and nerve growth factor to isolated chro- CA) was iodinated using the Iodo-Gen method. Glass tubes matin was inhibited by monoclonal antibodies specific for the (10 x 75 mm) were coated with Iodo-Gen (Pierce). Human respective growth factor receptor. The data suggest that EGF [5 ,ug in 50 ,u1 of phosphate-buffered saline (PBS: 0.15 chromatin binding may represent an important step in the M NaCl/10 mM phosphate, pH 7.25) was incubated with 200 pathway of growth factor action. ,pCi of 1"¶I (Amersham; 13-17 kCi'jug; 1 Ci = 37 GBq) for 10 min at room temperature. Bound1 I was separated from free The precise mechanism by which different growth factors 2I on a column of Sephadex G-25 medium (Pharmacia). exert their effects is unclear, although interaction with the Incubation of Cells with 1251-Labeled Growth Factors. Cell appropriate surface receptor and internalization (1-9) are lines were grown as monolayers and replated in fresh medi- common to all of these factors. The fate of internalized um. 125I-EGF (10 ng/ml, 30 cpm/pg), 125I-NGF (10 ng/ml, growth factors is controversial. Nerve growth factor (NGF) 22.5 cpm/pg), or 125I-PDGF (1 or 5 ng/ml, 60 cpm/pg) were (10, 11), epidermal growth factor (EGF) (12, 13), and insulin added 24 hr later. After 1, 24, and 48 hr of incubation, cells (14) have been reported to accumulate in the nucleus. Binding were washed 3 times with PBS and fractionated as described of NGF to the chromatin of embryonic dorsal root neurons below. In competitive inhibition studies, Hs 294 cells were has been suggested by Andres and Bradshaw (11). In con- preincubated for 1 hr with unlabeled NGF (500 ng/ml), and trast, Rohver et al. (9) have reported that nuclear accumu- WM 266-4 cells, with unlabeled PDGF at the same concen- lation of 125I-labeled NGF (125I-NGF) does not occur under tration; 125I-NGF (10 ng/ml) and 1251-PDGF (5 ng/ml) were physiological conditions. added, respectively, and incubated further for 1 hr. In one In this report, we describe the binding of NGF, EGF, and experiment, AG 1523 cells were incubated with 1251-PDGF in platelet-derived growth factor (PDGF) to chromatin in cell the presence of anti-PDGF antibodies (22). lines expressing surface receptors for these factors. This Cell Fractionation. Cell fractions were obtained using a binding is accompanied by a change in DNase II sensitivity modification (23) ofthe procedure of Augenlicht and Baserga at the growth factor-binding region of the chromatin. (24). Cells were washed 3-5 times with PBS, disrupted in 0.35 M sucrose/10 mM KCl/1.5 mM MgCl2/10 mM Tris Cl, pH 7.6/0.12% Triton X-100/12 mM 2-mercaptoethanol and cen- MATERIALS AND METHODS trifuged at 600 x g for 10 min. The supernatant (crude Cell Lines. Cells were grown in Eagle's minimal essential cytoplasmic fraction) was centrifuged further for 1 hr at medium or Leibovitz's L-15 medium supplemented with 10% 105,000 x g to obtain the microsomal and pure cytoplasmic fetal bovine serum. Table 1 summarizes receptor expression fractions. The nuclear pellet was washed first with 0.2 M in the cell lines used. sucrose/3 mM CaCl2/50 mM Tris Cl, pH 7.6, and then with Monoclonal Antibodies (mAbs). Anti-NGF receptor mAb 0.14 M NaCl/10 mM Tris Cl, pH 8.3, and centrifuged at 700 ME20.4 has been described (20). Detailed characterization of x g for 10 min. Nucleoplasmic proteins extracted with 0.14 anti-EGF receptor mAb 425, derived by immunization of M NaCl were defined as the "sap-protein" fraction. The mice with epidermoid carcinoma A431 cells, will be present- ed elsewhere. Abbreviations: NGF, nerve growth factor; PDGF, platelet-derived growth factor; EGF, epidermal growth factor; mAb, monoclonal antibody; 125I_, 125I-labeled. The publication costs of this article were defrayed in part by page charge *On leave ofabsence from: Institute of Human Genetics ofthe Polish payment. This article must therefore be hereby marked "advertisement" Academy of Sciences, Poznan, Poland. in accordance with 18 U.S.C. §1734 solely to indicate this fact. tTo whom reprint requests should be addressed. 3728 Downloaded by guest on September 30, 2021 Biochemistry: Rakowicz-Szulczynska et A Proc. Natl. Acad. Sci. USA 83 (1986) 3729

pellet was swelled in a small amount of 1 mM Tris Cl (pH 7.9). iments, anti-NGF receptor mAb ME20.4 or anti-EGF receptor Chromatin was pelleted at the bottom of the tube, nucleo- mAb 425 was used instead of unlabeled growth factors. plasm (residual fraction after extraction with 0.14 M NaCl) Electrophoresis. Radiolabeled unbound and chromatin- was recovered from the top, and nuclear membranes were bound 12 I-labeled growth factors were analyzed by 0.1% taken at the interface. Chromatin was further fractionated by NaDodSO4/15% PAGE (28). Gels were run at 100 V for 4 hr, extraction with 0.35 M and 2 M NaCl as described (25). stained with Coomassie blue, dried, and autoradiographed. Radioactivity of 125I-labeled growth factor bound to the particular fractions was measured in a gamma counter. The RESULTS number of growth factor molecules bound was calculated using Avogadro's number and the specific activity of each Uptake and Intracellular Distribution of 1251-EGF and 1251_ growth factor. NGF. To determine whether intracellular uptake and chro- Digestion of Chromatin with DNase II. Chromatin was matin binding of 125I-EGF and 125I-NGF is mediated by cell digested for various times at 37°C with DNase II [1 enzyme surface-bound receptors, we exposed cells expressing or not unit plus 500 ,g ofchromatin per ml in 25 mM sodium acetate expressing receptors (Table 1) to these growth factors. In (pH 6.6)] (26). The reaction was stopped by cooling in ice and SW948 cells, which express EGF receptor but not NGF raising the pH to 7.5 with 0.1 M Tris (pH 11). Undigested receptor, uptake of 125I-EGF was 40- to 100-fold higher than chromatin was removed by centrifugation. The percent of uptake of 1251-NGF and 150-fold higher than 125I-EGF uptake digested sequences (supernatant) was determined spectro- in EGF receptor-negative SW707 rectal carcinoma cells photometrically. The supernatant was then precipitated with (Table 2). Uptake 1251-NGF by SW707 cells, which express 2 mM MgCl2 to obtain Mg2+-soluble and -insoluble fractions. NGF receptor, was 5- to 12-fold higher than uptake of According to Gottesfeld et al. (26), the Mg2+-soluble fraction '251-EGF by these cells. However, 125I-NGF uptake by Hs contains DNase II-sensitive, actively transcribed sequences. 294 melanoma cells, which overexpress NGF receptor, was Solid-Phase Radioimmunoassay. Chromatin was resus- 30-fold higher than that by SW707 cells. pended in 1 mM Tris Cl/1 mM EDTA at pH 7.4 and adsorbed More than 90% of 125I-EGF in SW948 cells was found in the to the wells of polyvinyl chloride microwell plates at 4°C, as cytoplasmic fraction throughout the incubation period (Table described (27), but without chromatin sonication. After the 2). After 1 hr of incubation, a small amount (1.6%) of the plates were washed with water and incubated with 1% bovine 125I-EGF was chromatin-bound, which increased to 4.2% by serum albumin in PBS to block nonspecific binding, they were 24 hr and 4.8% by 48 hr. Sap-protein and nucleoplasmic incubated with mAb tissue culture supernatant (100 ,ul per well), fractions also contained trace amounts ofthe 125I-EGF, as did washed, and incubated with '25I-labeled rabbit anti-mouse IgG, the nuclear membranes. and bound radioactivity was measured. The intracellular distribution of 125I-NGF differed in Binding of '251-Labeled Growth Factors to Isolated Chro- SW707 and Hs 294 cells (Table 2). In SW707 cells, the matin. Chromatin was adsorbed to the wells of microwell greatest amount of 125I-NGF was detected in the cytoplasm plates as described above. 1251I-EGF, -NGF, or -PDGF at after 1 hr of incubation, and this decreased after 24 hr and concentrations from 0.1 to 40 ng/ml was added to each well after 48 hr. The chromatin fraction, however, contained and incubated for 2 hr at room temperature. After washing 20.5% ofthe 125I-NGF after 1 hr, which increased by 24 hr and with PBS, bound radioactivity was measured. In competitive 48 hr to -50%. The sap-protein and nucleoplasmic fractions binding assays, chromatin adsorbed to the wells was incu- contained a small amount of the 125I-NGF, and the least bated with unlabeled EGF (700 ng/ml), NGF (700 ng/ml), or amount was in the nuclear membrane fraction. In Hs 294 PDGF (500 ng/ml), washed with PBS, and then incubated melanoma cells, the cytoplasm also contained the highest with the respective 111-labeled growth factor. In some exper- amount of 125I-NGF during the first 24 hr of incubation, but

Table 2. Intracellular distribution of 125I-EGF and 125I-NGF in SW948 colon carcinoma and SW707 rectal carcinoma cells and of 125I-NGF in Hs 294 melanoma cells Time of % in the fraction Cell Growth incubation, Total Sap- Nuclear line factor hr incorporation* Cytoplasmt protein Nucleoplasm membranes Chromatin SW948 '25I-EGF 1 1122.4 ± 50.2 95.6 2.2 0.6 0.05 1.6 24 1158.5 ± 54.0 94.1 0.8 0.3 0.6 4.2 48 1453.1 ± 25.5 98.3 1.2 0.7 0.05 4.8 125I-NGF 48 36.7 ± 2.1 55.0 3.4 0.7 7.0 25.5 SW707 125I-NGF 1 15.2 ± 1.0 73.2 4.6 0.6 1.1 20.5 24 69.3 ± 3.3 49.0 2.6 0.8 1.5 46.1 48 70.5 ± 2.9 43.1 2.3 0.8 1.3 52.5 125I-EGF 48 6.7 ± 0.5 58.8 1.3 1.5 2.4 35.5 Hs 294 125I-NGF 1 1321.1 ± 150.2 97.7 0.7 0.3 0.07 1.2 Pt 330.0 ± 25 98.1 0.6 0.2 0.07 1.0 24 1999.9 ± 196.8 85.4 0.7 0.5 0.8 8.6 48 2481.7 ± 227.1 13.8 10.8 1.0 0.8 73.6 125I-NGF§ 1 990.0 ± 12 97.3 0.8 0.03 0.03 1.2 24 985.0 ± 20 77.4 3.8 1.10 0.10 17.6 *Values (pg per 106 cells) calculated from specific activities of radiolabeled EGF and NGF are the means ± SEM from two experiments (SW948 and SW707) or three experiments (Hs 294). tIncludes microsomal fraction, which contains up to 2% of the radioactivity. tUnlabeled NGF (500 ng/ml) was added 1 hr before 125I-NGF. §Cells were incubated for 1 hr with 125I-NGF (5 ng/ml), washed, and incubated for 24 hr in 125I-NGF-free medium. Downloaded by guest on September 30, 2021 3730 Biochemistry: Rakowicz-Szulczynska et al. Proc. Natl. Acad Sci. USA 83 (1986)

1050 - determine whether intracellular uptake and chromatin binding of PDGF is mediated by the presence of cell surface- 1000- / bound PDGF receptors, we incubated the following cells with o 200_ _ -v '25I-PDGF: WI-38 and AG 1523 normal fibroblasts, which express surface PDGF receptor but do not synthesize PDGF; 150- WM 266-4 melanoma cells and 1810 lung carcinoma cells, c L. o2 u which produce PDGF but no detectable PDGF surface :L 100- receptor (B. Westermark, personal communication); and cu _ i " 0 SK-MEL-21 melanoma cells, which neither synthesize 5. I0 ':f PDGF nor express the surface receptor (Table 1). WI-38 and A _ E AG 1523 normal fibroblasts incorporated the largest amount ofthe growth factor. Although WM 266-4 melanoma cells and A B C D E F G H 1810 lung carcinoma cells incorporated lesser amounts of FIG. 1. Total incorporation of 125I-PDGF. Bars represent incor- '25I-PDGF, these were still significant as compared with sn-.. in poration after 1 hr (light stippling), 24 hr (heavy stippling), and 48 hr SK-MEL-21, which incorporation was at background (black) of incubation. (A and B) WI-38 cells with 125I-PDGF at 1 and levels. During 1 hr of incubation, 125I-PDGF incorporation in 5 ng/ml, respectively. (C) AG 1523 cells with I251-PDGF at 5 ng/ml AG 1523 cells was 18-fold higher than in WI-38 cells and (results for 1 hr shown). (D and E) WM 266-4 cells with I25I-PDGF 50-fold that in WM 266-4 melanoma cells and 1810 lung at 1 and 5 ng/ml, respectively. (F and G) Line 1810 cells with carcinoma cells (Fig. 1). After 1 hr of incubation, 17-62% of 1I-PDGF at 1 and 5 ng/ml, respectively. (H) SK-MEL-21 with 125_PDGF taken up by WI-38 and 1810 cells was detected in 123I-PDGF at 1 ng/ml. (A-C) Cells express PDGF receptor. (D-G) Malignant cells produce PDGF but no detectable receptor. (H) chromatin, in marked contrast with WM 266-4 melanoma Malignant cells produce no PDGF and no detectable receptor. cells where 65-66% of the 1251-PDGF taken up was found in chromatin (Table 3). The levels of chromatin-bound PDGF remained relatively constant in all three cell lines, except that this decreased rapidly by 48 hr. Concomitantly, only 1.2% of after exposure ofWM 266-4 cells to 125I-PDGF at 5 ng/ml, the the 125I-NGF was chromatin-bound after 1 hr, but by 48 hr, levels ofchromatin-bound PDGF decreased between 1 and 48 74% of the 125I-NGF was bound to chromatin. hr after exposure. The 1251-PDGF bound to chromatin after 48 In another experiment, Hs 294 cells were incubated with hr of incubation comigrated in electrophoresis with free 125I-NGF for 1 hr, washed, trypsinized, and incubated for 24 IPDGF (results not shown). hr in I251-NGF-free medium. After the incubation tirne, the Inhibition of 125I-NGF and '251-PDGF Uptake. Unlabeled amount of chromatin-bound '25I-NGF had increased from NGF, when added at a 50-fold excess to the cell culture 1 hr 1.2% to 17.6% (Table 2). before addition of '25I-NGF, inhibited uptake of 125I-NGF by 125I-EGF and 125I-NGF associated with chromatin after 48 Hs 294 cells by 75% (Table 2). Similarly, unlabeled PDGF hr ofincubation showed the same electrophoretic mobility as added at a 100-fold excess to WM 266-4 cells, inhibited unbound growth factors (data not shown). uptake of 125I-PDGF by these cells by 80% (Table 3). In the Uptake and Intracellular Distribution of '251-PDGF. To presence of anti-PDGF antibodies, added to the incubation

Table 3. Intracellular distribution of 12I1-PDGF in WI-38 and AG 1523 fibroblasts, WM 266-4 melanoma cells, and 1810 lung carcinoma cells Time of % in the fraction 125I-PDGF, incubation, Total Sap- Nucleo- Nuclear Cell line ng/ml hr incorporation* Cytoplasmt protein plasm membranes Chromatin WI-38 1 1 17.5 ± 1.2 60.2 2.9 5.7 4.3 26.9 24 24.4 ± 2.5 60.5 1.8 9.0 3.9 24.8 48 31.0 ± 2.8 74.1 2.4 1.0 2.1 20.4 5 1 74.6 ± 4.2 79.0 0.4 2.5 0.7 17.4 24 158.1 ± 16.3 53.1 2.4 1.7 2.2 40.6 48 199.4 ± 18.5 62.6 1.8 1.1 2.4 32.1 AG 1523 S 1 1029.9 ± 25.0 69.5 18.7 0.7 1.7 9.4 S 1* 187.8 ± 15.5 59.3 13.9 2.8 4.1 19.9 WM 266-4 1 1 14.4 ± 1.20 17.6 4.5 5.9 5.6 66.4 24 14.1 ± 1.1 18.6 2.1 0.8 1.0 77.4 48 32.8 ± 2.0 10.3 4.0 1.2 1.4 83.1 1 21.3 ± 1.8 32.4 1.5 0.2 0.2 65.7 24 40.5 ± 3.9 57.1 4.4 0.25 0.25 38.0 48 74.0 ± 6.5 50.9 8.9 2.0 0.8 37.4 5 1§ 4.3 ± 0.5 42.6 0.4 3.9 5.7 47.4 1810 1 1 9.8 ± 1.0 40.8 6.1 1.0 9.2 42.9 24 19.0 ± 1.6 53.7 3.2 3.1 7.4 32.6 48 20.5 ± 1.4 54.2 5.0 3.0 3.1 34.0 5 1 15.4 ± 1.1 56.5 8.4 0.02 2.6 32.5 24 24.8 ± 1.8 57.7 6.5 0.08 6.0 29.0 48 68.4 ± 3.2 74.2 2.8 0.7 2.0 20.3 *Values (pg per 106 cells) calculated from specific activity of 125I-PDGF are the means ± SEM from two experiments. tIncludes microsomal fraction, which contains up to 3% Qf the radioactivity. *Incubation in the presence of anti-PDGF antibodies. §Unlabeled PDGF (500 ng/ml) was added 1 hr before 1251-PDGF. Downloaded by guest on September 30, 2021 Biochemistry: Rakowicz-Szulczynska et al. Proc. Natl. Acad. Sci. USA 83 (1986) 3731

Table 4. Distribution of 125I-EGF, '25I-NGF, and 125I-PDGF in microwell plates was studied (Fig. 2). mAb ME20.4 bound to chromatin fractions after 48 hr of incubation the microsomal fraction (used as control) and chromatin from Percent of the growth factor Hs 294 melanoma cells, which express surface-bound NGF receptor, but not to chromatin from SW1116 colorectal Chromatin fraction '25I-EGF 125I-NGF 1251-PDGF carcinoma, which does not express NGF receptor. To ex- 0.35 M NaCl-soluble 18.5 ± 1.8 23.1 ± 1.5 24.9 ± 1.5 clude the possibility that mAb binding to the chromatin of Hs 2 M NaCl-soluble 1.5 ± 0.2 1.6 ± 0.4 10.1 ± 1.5 294 cells reflects contamination with the NGF receptor- Residual fraction 80.0 ± 2.0 75.3 ± 2.1 65.0 ± 3.0 containing microsomal fraction ofthese cells, the microsomal '25I-EGF incorporation was studied in SW948 cells; '251-NGF, in fraction was mixed with the nuclear pellet of SW1116 cells Hs 294 cells; and 1251-PDGF, in WM 266-4 cells. Data are given as and chromatin was prepared as described in Materials and means ± SEM from three separate experiments. Methods. As shown in Fig. 2, there was little difference in the binding of mAb to this preparation as compared to chromatin of SW1116 cells alone. medium together with 1251I-PDGF, intracellular uptake by AG Binding of 125I-Labeled Growth Factors to Isolated Chro- 1523 cells was inhibited 82% (Table 3). matin. Binding of '25I-EGF, -NGF, and -PDGF to chromatin Presence of 1251-Labeled Growth Factors in Chromatin of the appropriate cell line was inhibited 80-90% in the Fractions Obtained by Salt Extraction. To show that growth presence of the homologous unlabeled growth factor (Fig. 3). factors are tightly bound to chromatin and are not adsorbed Binding of 125I-EGF to chromatin from SW948 cells was to chromatin from the cytoplasm during the fractionation inhibited 90% in the presence of anti-EGF receptor mAb 425. procedure, chromatin isolated from different cell lines (Table Binding of 125I-NGF to chromatin from Hs 294 cells was 4) 48 hr after incubation with the growth factor was extracted inhibited 90% in the presence of mAb ME20.4 (Fig. 3 A and with 0.35 M and 2 M NaCl. Up to 80% of chromatin-bound 1251I-EGF, 75% of 125I-NGF, and 65% of 125I-PDGF were present in the residual chromatin fraction (Table 4), contain- 8r ing only approximately 25% of chromatin protein and 30% of the DNA (data not shown). 6 Presence of 1251-Labeled Growth Factors in DNase II- Sensitive and -Insensitive Chromatin. After 1, 24, and 48 hr of 4 incubation of SW948 cells with 125I-EGF, of Hs 294 cells with 125I-NGF, and of WM 266-4 cells with 125I-PDGF, chromatin 2 was isolated and treated with DNase II until a limit digest of 12-20% DNase II-sensitive sequences was obtained. Ap- proximately 11-13% of the digested DNA was Mg2+-soluble o0 (data not shown), which suggests that the digest obtained az -25I-EGF, ng/ml contains DNase II-sensitive, Mg2+-soluble, actively transcribed sequences that are preferentially digested (26). Re- I- lease of 125I-labeled growth factors during chromatin diges- 0 tion was measured. After 1 hr of incubation with each of the three growth o factors, digestion of =13% of DNA sequences - released 70% of 125I-EGF (nonspecific release before digestion was -18%), 64% of 125I-NGF (nonspecific release 15%), - and 50% of 125I-PDGF (nonspecific release 19%). After 24 hr ~0 ofincubation, DNA digestion did not result in specific release 3 of 125I-NGF and 125I-PDGF and released only 10% of 1251I EGF. After 48 hr, no specific release of any of the growth L- factors was detected (data not shown). Binding of Anti-Growth Factor Receptor mAb to Isolated .C 1251-NGF. ng/ml Chromatin. To determine whether binding of the growth factor to chromatin involves receptor-like molecules, binding 80 -c of anti-NGF receptor mAb ME20.4 to chromatin adsorbed to 60

4 40

3- 20

0 0 10 20 30 40 1251-PDGF, ng/ml FIG. 3. Binding of 251I-labeled growth factors to isolated chromatin. (A) SW948 cells: 251I-EGF alone (0-.); preincubated with unlabeled EGF (700 ng/ml, o-o); preincubated with anti-EGF 1 2 3 4 receptor mAb 425 (_-_). (B) Hs 294 cells: 1251I-NGF alone (o*-); preincubated with unlabeled NGF (700 ng/ml, A-A); preincubated FIG. 2. Binding of anti-NGF receptor mAb ME20.4 to with anti-NGF receptor mAb ME20.4 (o---o). SW707 cells: 125I1 microsomes (bar 1) and chromatin (bar 2) of melanoma cell line Hs NGF alone (o-o), preincubated with unlabeled NGF (A-A); 294. Neither chromatin ofcolorectal carcinoma cell line SW1116 (bar preincubated with anti-NGF receptor mAb ME20.4 (o--). (C) 3) nor microsomes of SW1116 (not shown) are bound by anti-NGF- WI-38 cells; 1251-PDGF alone (-*); preincubated with unlabeled receptor mAb. Preincubation of SW1116 chromatin with Hs 294 PDGF (500 ng/ml, *---). SW 266-4 cells: 1251-PDGF alone (o0o); microsomes did not increase mAb binding levels (bar 4). preincubated with unlabeled PDGF (o---a). Downloaded by guest on September 30, 2021 3732 Biochemistry: Rakowicz-Szulczynska et al. Proc. Natl. Acad. Sci. USA 83 (1986) B) and to SW707 chromatin, 50% in the presence of the same physiological conditions in culture medium. The mechanisms antibody (Fig. 3B). ofnuclear uptake and binding to chromatin acceptor sites and the sequences possibly activated by different growth factors DISCUSSION remain to be investigated. The results show that growth factors are taken up by cells into We thank Dr. B. Westermark for kindly supplying '25I-PDGF, the cytoplasm and that some of the growth factor molecules PDGF, anti-PDGF antibodies, and the AG 1523 and 1810 cell lines become bound to the chromatin. In the case of EGF and and for helpful discussion; Dr. A. Ross, for anti-NGF antibody; Dr. NGF, uptake and subsequent binding to chromatin are Bothwell, for NGF; Dr. P. Phillips, for WI-38 cells; Dr. G. J. Todaro, for Hs 294 cells; and Dr. K. 0. Lloyd, for SK-MEL-21 cells. This associated with the presence of the appropriate receptors on study was supported in part by a grant from the Sandoz Foundation, the cell surface. For NGF, 84,000 molecules per nucleus a grant from Deutsche Forschungsgemeinschaft (Ro 619), and by (based on total incorporation, Table 2) were bound to the grants CA25874, CA21124, CA10815, and RR05540 from the Nation- chromatin of Hs 294 cells, which expresses the highest al Institutes of Health. number of cell surface NGF receptors (15), as compared with 1800 molecules bound to chromatin of SW707 cells, which 1. Landreth, G. E. & Shooter, E. M. (1980) Proc. Natl. Acad. have low-level receptor expression (data not shown); NGF Sci. USA 77, 4751-4755. binding to chromatin of SW948 cells, which do not express 2. Schechter, A. L. & Bothwell, M. A. (1981) Cell 24, 867-874. receptors, was at background levels. For EGF, 7000 mole- 3. Bernd, P. & Greene, L. A. (1984) J. Biol. Chem. 259, cules were bound to the chromatin of receptor-expressing 15509-15516. 4. Carpenter, G. & Cohen, S. (1976) J. Cell Biol. 71, 159-171. SW948 cells, whereas EGF binding to chromatin of SW707 5. Vlodavsky, I., Brown, K. D. & Gospodarowicz, D. (1978) J. cells, which do not express the receptor, was at background Biol. Chem. 253, 3744-3750. levels. PDGF, on the other hand, was bound to chromatin of 6. Heldin, C. H., Westermark, B. & Wasteson, A. (1981) Proc. both WM 266-4 and 1810 cells, neither of which expresses Natl. Acad. Sci. USA 78, 3664-3668. PDGF receptors (Table 3). However, after a 1-hr exposure to 7. Heldin, C. H., Wasteson, A. & Westermark, B. (1982) J. Biol. 125I-PDGF at 5 ng/ml, two human fibroblast lines that express Chem. 257, 4216-4221. PDGF receptors took up more PDGF than did WM 266-4 and 8. Bowen-Pope, D. F. & Ross, R. (1982) J. Biol. Chem. 257, 1810 cells. Moreover, within 48 hr ofexposure to PDGF, 1300 5161-5171. 9. Rohver, H., Schafer, T., Korsching, S. & Thoenen, M. (1982) molecules were bound to chromatin per nucleus (based on J. Neurosci. 2, 687-697. total incorporation, Table 3) of WI-38 human fibroblasts, as 10. Yanker, B. A. & Shooter, E. M. (1979) Proc. Natl. Acad. Sci. compared with only 550 and 400 molecules per nucleus of USA 76, 1269-1273. WM 266-4 and 1810 cells, respectively. Unlike the fibroblast 11. Andres, R. Y. & Bradshaw, R. A. (1977) Proc. Natl. Acad. lines, both WM 266-4 and 1810 cells secrete PDGF into the Sci. USA 74, 2785-2789. medium, and the lower uptake of PDGF in these cells might 12. Johnson, L. K., Vlodavsky, I., Baxter, J. D. & Gospo- result from the competition between labeled PDGF and darowicz, D. (1980) Nature (London) 287, 343-348. endogenous PDGF for binding sites on the surface receptors. 13. Savion, N., Vlodavsky, I. & Gospodarowicz, D. (1981) J. Biol. The lower density of125I-PDGF in chromatin ofthese two cell Chem. 256, 1149-1154. 14. Goldfine, I. D., Smith, G. J., Wong, K. Y. & Jones, A. L. lines might reflect binding of endogenous intracytoplasmic (1977) Proc. Natl. Acad. Sci. USA 74, 1368-1372. PDGF to chromatin, leaving only a restricted number offree 15. Fabricant, R. N., De Larco, J. E. & Todaro, G. J. (1977) Proc. binding sites in the chromatin ofthese cells. The 80% inhibition Natl. Acad. Sci. USA 74, 565-569. of 125I-PDGF incorporation into WM 266-4 cells by unlabeled 16. Herlyn, M., Balaban, G., Bennicelli, J., Guerry, D., IV, PDGF and the 80o inhibition of125I-PDGF uptake into AG 1523 Halaban, R., Herlyn, D., Elder, D. E., Maul, G. G., Step- fibroblasts by anti-PDGF antibodies suggest that entry of lewski, Z., Nowell, P. C., Clark, W. H. & Koprowski, H. "I-PDGF into the cell is mediated by surface receptors. (1985) J. Natl. Cancer Inst. 74, 283-289. At present, it is not known whether the growth factors exist 17. Bergh, J., Nilsson, K., Dahl, D., Andersson, L., Virtounen, in a free state or bound to the respective receptors. However, M. & Lehto, P. (1984) Lab. Invest. 51, 307-316. 18. Herlyn, D., Steplewski, Z., Herlyn, M. & Koprowski, H. one suspects the presence of receptor-like molecules in the (1980) Cancer Res. 40, 717-721. chromatin, since mAbs that react with either NGF or EGF 19. Herlyn, M., Guerry, D. & Koprowski, H. (1985) J. Immunol. receptors prevent binding of the corresponding growth fac- 134, 4226-4230. tors to the chromatin. All three growth factors were recov- 20. Ross, A. H., Grob, P., Bothwell, M., Elder, D. E., Ernst, ered from the chromatin in a nondegraded state 48 hr after C. S., Marano, N., Ghrist, B. F. D., Slemp, C. C., Herlyn, incubation. That the growth factors are tightly bound to M., Atkinson, B. & Koprowski, H. (1984) Proc. Natl. Acad. chromatin and do not originate from the cytoplasmic fraction Sci. USA 81, 6681-6685. is indicated by the fact that up to 80% of the total chromatin- 21. Sutter, A., Rippelle, R. J., Harris-Warrich, R. M. & Shooter, bound growth factor was recovered from the residual chro- E. M. (1979) J. Biol. Chem. 254, 5972-5982. 22. Bowen-Pope, D. F., Vogel, A. & Ross, R. (1984) Proc. Nail. matin fraction after extraction with 2 M NaCl. In addition, Acad. Sci. USA 81, 2396-2400. binding ofgrowth factors to specific DNase II-sensitive regions 23. Horst, A., Rakowicz-Szulczynska, E. M. & Wiland, E. (1981) ofchromatin eliminates the possibility that the binding sites are Mol. Cell. Biochem. 37, 3-7. randomly distributed in the chromatin. These growth factor- 24. Augenlicht, L. H. & Baserga, R. (1973) Arch. Biochem. binding regions were DNase II-sensitive after 1 hr ofincubation Biophys. 158, 89-96. but DNase II-resistant 24 hrlater. A protective effect on isolated 25. Rakowicz-Szulczynska, E. M. & Horst, A. (1981) Biochim. genes has been described (29, 30). Biophys. Acta 653, 69-82. Nuclear uptake of NGF (10, 11) and binding to the nuclear 26. Gottesfeld, J. M., Bagi, G., Berg, B. & Bonner, J. (1976) membrane (10) have been reported. Rohver et al. (9) reported Biochemistry 15, 2472-2483. 27. Vanderbilt, J. N. & Anderson, J. N. (1983) J. Biol. Chem. 258, that nuclear accumulation of125I-NGF in PC12 cells occurred 7751-7756. only when cells were maintained in 0.2% glucose/0.1% 28. Laemmli, U. K. (1971) Nature (London) 227, 680-685. bovine serum albumin instead of culture medium; accumu- 29. Scheidereit, C., Geisse, S., Westphal, H. M. & Beato, M. lation was observed only in plasma membranes and cyto- (1983) Nature (London) 304, 749-752. plasm. In our hands, all three growth factors were directly 30. Emerson, B. M., Lewis, C. D. & Felsenfeld, G. (1985) Cell 41, bound to chromatin of several cell types grown under 21-30. Downloaded by guest on September 30, 2021