JOURNAL OF CELLULAR PHYSIOLOGY 145~303-309(1990)

Vitamin D Regulates Transferrin Receptor Expression by Bone Marrow Macrophage Precursors HlROYUKl TANAKA AND STEVEN 1. TEITELBAUM* Department of Pathology and laboratory Medicine, lewish Hospital at Washington University Medical Center, St. Louis, Missouri 63 I 10 1,25-Dihydroxyvitamin D, [l ,25(OH),D,] is known to prompt monocytic differentiation of a variety of leukemic lines. We previously extended these observations to non-transformed bone marrow macrophage precursors by demonstrating that the steroid enhances plasma membrane expression of the macrophagespecific mannose-fucosereceptor (Clohisy et al., ) Biol Chern 262:15922-I 5929, 1987). Because this membrane protein is involved in non-opsonin mediated endocytosis, these observations raised the possibility that 1,25(OH),D, globally upregulates endocytic receptors. The present study, aimed at addressing this issue, turns to the transferrin receptor as a paradigm for endocytic receptors and explores the impact of 1,25(0H),D, on its expression. We found that in contrast to the mannose-fucose receptor, plasma membrane transferrin receptor expression by bone marrow-derived macrophage precursors declines by at least 30% in a dose-dependent fashion with exposure to 1 ,25(OH),D3. The effect reflects diminished receptor capacity with no change in Kd, and is independent of cell cycle. Moreover, while, ,V , of receptor-ligand internalization mirrors plasma membrane occupancy, Kuptakeremains unaltered in the presence of vitamin D, indicating that the down-regulating event does not reflect on enhanced rate of endocytosis. Further, pulse chase experiments show parallel cell surface, intracellular, and medium redistribution of radioligand with time steroid-treated and control cells. In a similar vein, while total cell-associated radioligand falls in the presence of vitamin D, the percentage of intracellular and surface bound counts at equilibrium are constant in both groups. Finally, immunoprecipitation studies reveal that the down-regulating effects of 1,25(OH),D, cannot be explained by inhibition of transferrin receptor synthesis. Thus, the decrease in total cellular transferrin binding sites is likely to represent either enhanced degradation or synthesis of "cryptic" receptors which fail to recognize '251-transferrin.

The active metabolite of vitamin D, namely, 1,25dihydroxycholecalciferol (1,25(0H),D3), is classically viewed as a calcium regulating hormone. Studies performed within the past decade indicate, however, that the biological repertoire of this steroid is more global than previously appreciated and, in fact, 1,25(OH)2D3 appears to promote differentiation of native (Clohisy et al., 1987) and transformed cells (Abe et al., 1981; Morel et al., 1986; Bar-Shavit et al., 1983). Our laboratory has focused on the maturational effects of vitamin D, on hematopoietic cells and initially reported that the steroid prompts a bipotential leukemic line, HL-60, to differentiate along a monocytic pathway (Bar-Shavit et al., 1983; Reitsma et al., 1983; Bar-Shavit et al., 1986), findings which we recently extrapolated to authentic, bone marrow-derived macrophage precursors (Clohisy et al., 1987). This latter study utilized the lineage-specific membrane protein, the mannose-fucose receptor as a marker of macrophage differentiation, which we found is progressively up-regulated in cells exposed to 1,25(OH), D,. This protein is of particular interest because it is an 0 1990 WILEY-LISS, INC

endocytic receptor involved in the macrophage-typical event of non-opsonin-mediated particle uptake (SunSang et al., 1983). The present study extends these observations to the effects of 1,25(OH)2D3 on the kinetics of receptormediated endocytosis. In this circumstance, we chose to explore the steroid's impact on expression of the transferrin receptor as it is an endocytic protein whose distribution within the cell is easily followed. We found that in contrast to its enhancing effects on the mannose receptor, 1,25(OH)2D3down-regulates surface expression of the transferrin receptor. Received May 8, 1990; accepted July 30, 1990. *To whom reprint requestdcorrespondence should be addressed. Abbreviations used CSF-1, colony stimulating factor-1; BMDMs, bone marrow-derived mononuclear phagocytes; a-MEM, a-modification of Eagle's Minimal Essential Medium; PBS, phosphate buffered saline; HHBG, Hanks' Balanced Salt Solution + 10 mM Hepes, + 10 m M Tris t 0.1% glucose + 10 mglml bovine serum albumin (pH 7.1); 1,25(OH),D,, 1,25-dihydroxycholecalciferol; TCA, trichloroacetic acid.

304

TANAKA AND TEITELBAUM

MATERIALS AND METHODS Materials 1,25(OH)2D3was provided by Dr. Milan Uskokovic (Hoffman-LaRoche, Inc., Nutley, NJ). Unless otherwise specified, all chemicals were obtained from Sigma (St. Louis, MO). Murine L929 cells were a gift of Dr. H.S. Lin (Washington University Medical School, St. Louis, MO). Preparation of Stage I CSF-1 Stage I colony stimulating factor (CSF-1) was prepared from serum-free conditioned media from L929 cells by a batch calcium phosphate method as previously described (Clohisy et al., 1987). Final specific activity was typically lo6 units/mg protein. Culture medium Complete culture medium was a-modification of Eagle's Minimal Essential Medium (a-MEM) with 500 unitdm1 Stage I CSF-1,15% fetal calf serum, penicillin (100 unitdml), and streptomycin (100 pglml). Cells were treated with 1,25(OH)2D3or comparable amounts of carrier ethanol at the start of culture. Total ethanol concentrations was less than 0.01%. Preparation of adherent bone marrow-derived mononuclear phagocytes (BMDMs) Cells were obtained from the bone marrow of 9-12week-old male A/J mice (Jackson Labs, Benton Harbor, ME) as described previously (Clohisy et al., 1987).Bone marrow plugs were pre ared by flushing femurs with ice-cold a-MEM throug a 25 gauge needle, dispersed by several passages through an 18 gauge needle, and the cells cultured overnight in the presence of complete medium containing 900 U/ml Stage I CSF-1. The non-adherent population was recovered and the adherent cells discarded. The non-adherent cells were centrifuged (10,000xg/lO minutes/26"C), resuspended in 1 ml of Pronase solution (0.02%w/v B grade, Calbiochem, 1.5 mM EDTA in PBS per lo7 nucleated cells), and incubated for 15 minutes at 37°C. Pronase activity was stopped by the addition of horse serum (0.2 mg/lO ml Pronase solution), and the suspension layered on 15 ml ice-cold horse serum. After 15 minutes' incubation on ice, the upper fraction was once again placed on ice-cold horse serum and after 15 minutes, centrifuged (1,200xg for 7 minutes at 4°C). The cells were resuspended in complete medium, and then cultured in 24-well plates at the density of 0.2 x 106/m1/2ml per well. As previously documented (Clohisy et al., 19871, withdrawal of the macrophage-specific growth factor CSF-1 for 24-48 hours leads to death of all cells and, after 7 days in culture, each expresses the macrophageassociated enzyme, a-naphthyl acetate esterase. Adherent BMDM cell count Parallel cultures were washed with cold PBS, and the cells detached from the well by 5 minutes' incubation on ice in 0.005%Zwittergent (Behring Diagnostics) and counted in a hemocytometer. Transferrin iodination Human transferrin, established to be iron saturated by a A,,,/Azso ratio in excess of .04, was iodinated by

K

the chloramine-T method. A total of 200 pg of transferrin were mixed with 1 mCi of carrier-free NA1251 (Amersham, Arlington Heights, IL), and 100 pg of chloramine T in 20 pl of 0.1M sodium hosphate buffer (pH 7.6). The reaction was terminatec r after 3 minutes on ice by addition of 150 pg of sodium metabisulfite and 600 pg of potassium iodine. The sample was then run on a Sephadex G-50 column (1 X 20 cm) buffered in 10 mM Tris-HC1 (pH 7.5). Then 400 pl fractions were collected and radioactive fractions identified by y-counting. Protein determination was performed by the Miller method (Miller, 1959). Specific activity was typically 1MCilng protein and >95% radioactivity was recovered in the 10% TCA precipitable fraction. lZ5I-transferrinbinding assay BMDMs were washed three times with HHBG (Hanks' Balanced Salt Solution, 10 mM Hepes, 10 mM Tris, 0.1% glucose, and 10 mg/ml bovine serum albumin, pH 7.0) and incubated with 200 p1 of various concentrations of lZ5I-transferrinin HHBG plus 200 p1 of HHBG 6 mglml unlabeled ion-saturated transferrin. Equilibrium binding was achieved after 1 hour at 4°C after which the cells were rapidly washed six times with ice-cold Hanks' Balanced Salt Solution, lysed in 500 pl of 0.1N NaOH and subjected to y-counting. Endogenous transferrin in this circumstance does not affect radioligand binding (data not shown) and nonspecific binding is less than 10% of total bound counts.

*

Effects of cell cycling on transferrin receptor expression As transferrin receptor expression is known to parallel DNA synthesis (Trowbridge and Omary, 1981),we sought to individualize the effects of 1,25(OH)2D3from that of cell cycle. To this end, we utilized a technique recently developed in our laboratory designed to yield PO ulations of bone marrow-derived macrophages enric ed with quiescent (GoGI)or cycling (S-phase) cells (Clohisy et al., 1989). Briefly, asynchronous BMDMs maintained in the presence or absence of vitamin D3 for two days were transferred to medium containing 50 U/ml Stage I CSF-1 (quiescent dose) which, within 24 hours, reduces 3H-thymidine incorporation and the number of cells in S hase by more than 90%. The treated and control cel s were once again divided and either placed in quiescent levels of CSF-1 or a mitogenic dose which within eight hours places more than 45% of the cells in S phase.

K

Y

Pulse-chase experiments Surface transferrin receptors were saturated with 100 nM '251-transferrin for 1 hour at 4°C. The wells were then washed three times with cold phosphate buffered saline (PBS) and HHBG containing 100 nM unlabeled iron-saturated transferrin added. The cells were placed at 37°C and at designated intervals 1) medium containing, 2) surface residing (acid strippable) and 3) internalized (residual after acid stripping) counts were determined. Acid stripping After removing binding medium and/or washing in PBS, 400 ~1 of acid stripping solution (0.5 M NaCl,

305

VITAMIN D REGULATES TRANSFERRIN RECEPTORS

0.2M acetic acid) was added to the wells. After 8 minutes' incubation at 4"C, the medium was collected and cells were rinsed briefly with an additional 400 pl of acid strip ing solution. These two aliquots were combined an subjected to y-counting. Total protein synthesis BMDMs maintained in the presence or absence of 5 x 10-8M, 1,25(OH)& for 3 days were pulsed for 6 hours with 2 p.Ci [3]H-leucine(Dupont). The cells were then washed 3 times with PBS and incubated for 1hour on ice with 10% trichloroacetic acid (TCA). They were then re-rinsed with ethano1:ether (3:1) and extracted for counting in 0.1N NaOH. Immunoprecipitation of transferrin receptor Bone marrow macrophage precursors were plated in 100 mm tissue culture dishes at an initial seeding density of 0.2 X 106/mlin 50 ml of medium f 5 x lo-' M 1,25(OH)& Three days later, at which time the vitamin D-treated cell number was 2.45 2 0.05 x lo6/ plate and control, 2.95 0.05 x 106/plate, the cells were labeled with 100 piWm135S-methionine(Tran35SLabell, ICN, Irvine, CA) for 2 hours at 37°C. The dishes were then rinsed and the cells incubated another 30 minutes in standard medium containing 2 mM L-methionine. They were then washed with PBS and lysed with 500 pl of lysis buffer (1%Triton-X, 1 mM PMSF 0.2 TIU/ml aprotinine in PBS). The lysate was spun in a microfuge (Beckmann) for 10 minutes at 20,OOOg. The pellet was resuspended in lysis buffer in which the concentration was adjusted to the equivalent of lo6 cells/250 p1. Monoclonal antibody raised against the murine transferrin receptor was purified on a protein A affinity column from medium conditioned for 7 days by hybridoma R17.217.1.3 (ATCC TIB 219). The cell lysate and 1 pg antibody were incubated overnight at 4°C in immunoprecipitation buffer (1%Triton-X, 1%SDS, 0.5% deoxycholate, 0.5% bovine serum albumin in PBS) in a final reaction mixture volume of 500 pg of rat IgG?,. The immunocomplex was then adsorbed to protein A-sepharose b continuous mixing for 2 hours at 4°C. The protein 1-sepharose was then washed 3 times sequentially with immunoprecipitation buffer and PBS. The final protein A-sepharose pellet was boiled with 40 pl sample buffer (10% glycerol, 1%SDS, 1mM DTT, 0.001% bromophenol blue in 0.5% M Tris buffer, H 6.8) for 5 minutes and loaded onto an 8% SDS!AGE gel which, after overnight and 3 days' exposure, was subjected to fluorography. RESULTS 1251 fe transferrin binding to BMDM's 1,25(OH),D3 decreases surface expression of transferrin binding sites. As seen in Figure 1, down-regulation (P < .001, -30%) occurs by day 3 of steroid treatment, and although of less magnitude, is still resent on day 4 (P< .Ol), by which time radioligand bpinding by control cells has diminished. As documented by Scatchard analysis, the reduction in 1251-transferrin binding reflects decreased capacity with no change in receptor-ligand affinity (Fig. 2). The event is dose dependent with significant (I' < .01) inhibition of bind-

B

*

T

Day 2

Day 3

Day 4

Fig. 1. Time course of transferrin receptor down-regulation by 1,25(OH),D,. Maximum binding site number was determined using a saturating amount of lZ5I-transferrin (50 nM, see Fig. 2). Each bar represents specific binding (displaceable) t S.D. Statistical analysis was performed by Students t test.

6.25 12.5

25

50

ADDED 1251-TRANSFERRIN (nM) Fig. 2. Saturation analysis and Scatchard plot of transferrin receptor on day 3 BMDMs. lZ5I-transferrinbinding assay was performed at 4°C on cells maintained for 3 days with (o...~) or without (M) 1,25(OH),D, (50 nM). From the Scatchard plot (inset), Nmax of vitamin D-treated cells was calculated as 3.05 fmoles/104cells (3.75 fmoles/104cells in control) and Kd was 4.02 x W 9 M in 1,25(OH),D,treated cells (4.18 X 10-9M in control).

306

TANAKA AND TEITELBAUM

ing detectable at 10-lOM1,25(OH)2D3,a physiological concentration, particularly in the presence of 15% serum-containing medium (Fig. 3). Cell cycling a n d transferrin receptor expression Transferrin receptors appear coordinately with DNA synthesis (Trowbridge and Omary, 1981) and thus, changes in surface expression may reflect altered replication by agents such as 1,25(OH)2D,known to induce cells into GoGI phase (Bar-Shavit et al., 1983). To explore this possibility, we synchronized BMDM’s and then exposed them to concentrations of CSF-1 which induce them to either undergo rapid DNA synthesis or remain in a quiescent state. Thus, BMDM’s were treated with 5 x 10-sM 1,25(OH)zD, for 48 hours and then placed in fresh, vitamin D3-bearing medium containing either a “mitogenic” (500 U/ml) or “quiescent” (50 Ulml) concentration of CSF-1 for an additional 8 hours after which time 1251-transferrin binding was determined. As illustrated in Figure 4, inhibition of DNA synthesis leads to a 25% reduction in transferrin receptor expression independent of the presence or absence of vitamin D. Most importantly, however, 1251-transferrinbinding by either cycling or non-cycling cells exposed to the steroid is approximately 75% of that of their untreated counterparts. Similar results are obtained when cells are exposed to 1,25(OH)2D3for 72 instead of 48 hours prior to mitogenic manipulation (data not shown). Intracellular traffickin of receptor-bound ‘251-transerrin Experiments described thus far establish that 1,25(OH)?D3 down-regulates transferrin receptor expression independent of cell cycle and ambient CSF-1 levels but offer no insight into the mechanism by which this may occur. One likely explanation would, of course, be differences in the rates of receptor cycling. We began to explore this possibility by studying the effects of the steroid on an initial step in receptor trafficking, namely, endocytosis. To this end we added increasing amounts of lZ5I-transferrin to cells at 4°C and determined uptake at 37°C over 15 minutes, by which time cell-associated radioligand is in steady state (see below). Figure 5 shows that consistent with receptor down-regulation, less radioligand is lost from vitamin D-treated cells’ surface, leading to a proportional decrease in V,,,. In contrast, however, no differences in Kuptake(K,) exist between steroid-exposed and control cells. These data suggest that differences in cell-associated lZ5I-transferrinin vitamin D and control cells at 37°C reflect receptor capacity and not the rate at which the ligand is endocytosed. To further explore this possibility, we saturated surface transferrin receptors of day 3 cells by exposing them to 1251-transferrinfor 1hour at 4°C. The plates were washed and the cells incubated for various periods at 37°C in HHBG containing an equal amount of unlabeled transferrin. At specific intervals, the wells were quickly chilled and surface bound 1251transferrin removed by “acid-stripping” was measured. Medium-residing (i.e., exocytosed or displaced) counts were also assessed as were those cell-associated after acid exposure (ie., internalized).

Regardless of the presence of steroid, surface bound counts decline by 50-60% within the 5 minutes of warming, after which they continue to slowly decrease (Fig. 6). Thus, the calculated half life of 1251-transferrin is 4.5 and 4.2 minutes, respectively, in control and 1,25(OH)2D3treated cells again indicating that vitamin D does not alter the steady state rate of receptorligand internalization. Consistent with the fall in surface-bound counts, intracellular 1251-transferrinrises rapidly in both circumstances, plateauing at 15% of total within the first five minutes of warming. Medi-

I

-T n z m

2.0

z

a a

$

1.5

z

§

f

0

10-11 10-10 10-910-8

10-7

lS25(OH)2D3(M) Fig. 3. Dependency of transferrin receptor down-regulation on 1,25(OH),D, concentrations. BMDM’s were cultured 3 days with various levels of 1,25(OH),D, and binding assessed at 4°Cwith 50 nM radioligand. Each point represents the mean of quadruplicate assays +- S.D. Statistical analysis was performed by ANOVA. *P < 0.01 compared to control, **P< 0.01 compared to M.

3t t-

I 0

1

I

12.5 25

I

50

I

100

ADDED 1251-TRANSFERRIN (nM) Fig. 4. Effect of cell cycle on transferrin receptor expression. Cells were enriched for those in G,G, or S phase by ambient levels of CSF-1, and the l”1-transferrin binding assay was performed at 4°C.

307

VITAMIN D REGULATES TRANSFERRIN RECEPTORS

um-containing counts continue to increase in a linear fashion. Assessment of total cellular transferrin receptor While our previous data indicate that vitamin D does not alter the steady state rate of transferrin receptor trafficking, they do not exclude the possibilit that the steroid changes relative surface and intrace lular distributions of the binding protein. We examined this hypothesis by incubating BMDM’s with a saturatin concentration of 100 nM lZ5I-transferrin at 37” wherein cell-associated counts rapidly plateau reaching a dynamic steady state by 15 minutes (Fig. 7). One hour after warming, surface ligand was measured after “acid-stripping” as were residual, cell-associated counts. While as expected, 1,25(OH)2D3treatment decreases cell associated radioligand, there is no change in the percentage expressed on the plasma membrane. These data confirm that vitamin D,-induced transferrin receptor down-regulation reflects a decreased number of receptor copies and not its redistribution. Total protein and transferrin receptor synthesis Finally, we explored the effects of 1,25(OH)2D3on total protein and transferrin receptor synthesis. Thus, BMDM’s were metabolically labeled with ,H-leucine and incorporation into TCA-precipitable counts found to be similar in vitamin D,-treated (234.7 5 9.7 d p d lo4 cells) and control (235.9 2 9.7 dpm/104 cells) cells. Next 35S-methioninelabeled cells were lysed and the cell extract immuno recipitated with anti-transferrin receptor antibody. T e resultant fluorograms contain numerous bands, the most prominent of which has an apparent MW of 90,000 D,, approximating that of the reduced murine receptor (Seligman and Allen, 1987). When densitometrically analyzed, however, fhorograms derived from vitamin D3-treated cells are similar to control (Fig. 8). DISCUSSION Delivery of intracellular iron by transferrin constitutes the first step in incorporation of this essential cofactor into a variety of polypeptides, most notably, hemoglobin and several respiratory proteins (Vogt et al., 1969).Internalization of bound transferrin occurs by receptor-mediated endocytosis wherein the complex is transported to vesicles where, in an acidic milieu, iron is released (Octave et al., 1983). Upon its subsequent return to the cell surface, the apoprotein disassociates from the receptor in preparation for diferric reoccupancy and reinitiation of endocytosis. Thus, transferrin receptor cycling is relatively well understood and has become a paradigm of receptor-mediated endocytosis. Our interest in the transferrin receptor was prompted by studies of the effects of 1,25(OH)2D3on binding of mannosylated albumin by developing bone marrow macrophages (Clohisy et al,, 1987). We found in those experiments that expression of the macrophage-specific mannose receptor is upregulated on BMDM’s with time in culture and, furthermore, that 1,25(OH)&,enhances its expression. As endocytosis is a property characteristic of mature phagocytes, upregulation of the mannose receptor indicated vitamin D3

P

8

K

may enhance the endocytic capacity of developing macrophages. We had, in our previous study, raised the issue of specificity of mannose receptor upregulation by 1,25(OH)zD3and showed, in fact, that the steroid does not effect surface binding of 1251-athrombin (Clohisy et al., 1987). Bound a-thrombin does not, however, undergo classical receptor-mediated endocytosis and, hence, the issue of whether 1,25(OH),D3 globally enhances surface ex ression of endocytic receptors, remained unresolveb: The present series of experiments, therefore, utilized the transferrin receptor as an example of membrane endocytic binding proteins and focused on the effects of 1,25(OH)zD, on its expression by developing macrophages. We found that in contrast to the mannose receptor, 1,25(OH)2D, inhibits surface expression of the transferrin receptor by about 30% on a per cell basis, an event entirely attributable to decreased capacity with no change in receptor-ligand affinity. It should be appreciated that BMDM’s exposed for 72 hours to vitamin D, contain approximately 25% more protein per cell than do the wild type (unpublished data) and thus, we have presented the degree of 1,25(OH)2D3induced down-regulation in the most conservative manner. Most importantly, the event is dose-dependent

25

-

c . Control 0.-0

I

0

I

I

12.5 25

1,25(OH)&

1

I

50

100

ADDED 1251-TRANSFERRIN(nM)

-

0.6

,06

V ~ n 4 . 9 ) i 0 . 0 6 h n ~ ~ l ~ c e l b x 1 5 m i n0. 0

I(IO.94fOYxl(P

I

0

I

I

I

I

0.040.08

0.16 0.32 1 /ADDED 1251-TRANSFERRIN(nM)

Fig. 5. Kinetics of transferrin receptor complex internalization. Cells were incubated with various concentrations of radioligand at 4°C and uptake determined during the first 15 minutes of warming to 37°C. V,,, and K, are expressed ? S.D.

308

TANAKA AND TEITELBAUM Surface

lntracellular 100 r

100,0-0

Medium

lOOr

CONTROL

0-0 CONTROL 0 - 0 1,25(OH)zD3

0-0 1,25(OH)2D3

0-0

CONTROL

0-0 1,25(0fo2D3

0 0

m

0

c L

/-

a

u)

0

0

5 10 15 20 Time (min)

5 10 15 20 Time (min)

0

5 10 15 20 Time (min)

Fig, 6. Trafficking of transferrin receptor. Cells were incubated with 50 nM ‘”I-transferrin at 4°C and chased with an equal amount of unlabeled ligand. Surface bound counts were determined by acid stripping. Internalized radioligand was resistant to acid.

-

-I

0

65.2f 0.9

o 63.3 f 3.8

€ I

0

15

30

60

90

P 1

4°C binding

TIME (min.) Fig. 7. Time course of 1251-transferrin-ce11association. BMDMs maintained at 37°C for 3 days in the presence (-0) or absence 1-( of 5 x lo-@M 1,25(OH)2D, were incubated with a saturating amount of Y-transferrin (100 nM) with or without excess unlabeled ligand. Specific cell-associated radioligand was determined with time. At 60 minutes, parallel cells were acid stripped and the percentage of cell-associated counts which were surface bound calculated.

with picomolar concentrations of the steroid capable of eliciting detectable inhibition of expression. 1,25(OH)2D, is known to induce differentiation of native and transformed cells and in so doing, universally reduces its rate of DNA synthesis (Clohisy et al., 1987; Abe et al., 1981; Morel et al., 1986; Bar-Shavit et al., 1983). Beause transferrin receptor expression occurs parri passu with cell replication (Trowbridge and Omary, 19811, it was essential that we determine if the down-regulating effects of vitamin D3 reflect only its impact on cell-cycling. To this end, we utilized an approach recently developed in our laboratory in which BMDMs may be synchronized in GoGl by culture in low levels of CSF-1 and either maintained in a quiescent state or induced to undergo rapid mitosis by

markedly increasing the ambient levels of the macrophage-s ecific growth factor (Clohisy et al., 1989). We founcf as exzected, that replicating cells bind -25-30% more I-transferrin than do their quiescent counterparts. Most importantly, however, 1,25(OH)2 D,, regardless of the state of cell-cycling and ambient CSF-1, down-regulates surface transferrin receptor expression by an additional 25-30%. Having established that vitamin D3 “specifically” reduces plasma membrane transferrin receptor capacity, we turned to the mechanisms by which this may occur. Because DMSO-induced maturation of erythroleukemia cells enhances the rate of transferrin receptor internalization (Mulford and Lodish, 1988), we explored the effects of 1,25(OH)2D3on trafficking of the rece tor-ligand com lex. We found that while vitamin D,-bifferentiated ce 1s endocytose less radioligand per unit time, the event reflects surface receptor capacity (Vmax).In fact, the rates of occupied receptor internalization (Kuptake)of both treated and control cells are similar. Furthermore, in each circumstance, intracellular radioligand accumulates rapidly and plateaus in like pro ortions within 15 minutes of endocytosis initiation. Faken with the continual and parallel accumulation of medium counts, these experiments support the conclusions that 1,25(0H),D,-induced down-regulation of the transferrin receptor does not represent alterations in its rate of trafficking. Alternatively, decreased surface receptor capacity might reflect its redistribution into an intracellular compartment. We examined this possibility by saturating total cell 1251-transferrinbinding sites in a kinetic steady state and independently measured membrane and intracellular counts. While total cell-associated radioligand is reduced in the steroid-exposed cells, the distribution of counts is identical t o control. We conclude, therefore, that 1,25(0H),D,-induced receptor down-regulation reflects decreased total cell transferrin receptor content. On the other hand, we found that given the sensitivities of the techniques, both total protein and transferrin receptor synthesis are unal-

P

VITAMIN D REGULATES TRANSFERRIN RECEPTORS

309

LITERATURE CITED 67 kDa

1

94 kDa

1

RE L AT I V E D ISTANCE Fig. 8. Densitometrogram of fluorographs of cellular material immunoprecipitated with anti-transferrin receptor antibody. ---,immunoprecipitate from 1,25(0H),D,-treated cells (5 x 10-8M for 3 days); -, from untreated cells; and . . . ., from negative control using IgGz,.

tered by vitamin D3 treatment. Taken with the clear fall in lZ5I-transferrinbinding sites induced by the steroid, these findings suggest that vitamin D3 accelerates transferrin receptor degradation or prompts a shift to “cryptic” receptors which, although expressing the same epitopes as the native molecule, fail to recognize the radioligand. Thus, 1,25(OH),D3 treatment of macrophages is associated with decreased expression of functional transferrin receptors independent of cell cycle and kinetics of endocytosis. While the significance of this event as relates t o macrophage differentiation are unknown, our findings indicate that 1,25(OH)2D3exerts distinct effects on various endocytic receptors. ACKNOWLEDGMENTS This work was supported by NIH grant DE05413.

Abe, E., Miyaura, C., Sakagami, H., Takeda, M., Konno, K., Yamazaki, T., Yoshiki, S., and Suda, T. (1981) Differentiation of mouse myeloid leukemia cells induced by la,25-dihydroxyvitamin D,. Proc. Natl. Acad. Sci. USA, 78:49904994. Bar-Shavit, Z., Teitelbaum, S.L., Reitsma, P.H., Hall, A,, Pegg, L.E., Trial, J., and Kahn, A.J. (1983) Induction of monocytic differentiation and bone resorption by 1,25-dihydroxyvitamin D,. Proc. Natl. Acad. Sci. USA, 805907-5911. Bar-Shavit, Z., Kahn, A.J., Stone, K.R., Trial, J., Reitsma, P.H., and Teitelbaum, S.L. (1986) Reversibility of vitamin D-induced human leukemia cell-line maturation. Endocrinology, 118:679-686. Clohisy, D.R., Bar-Shavit, Z., Chappel, J., and Teitelbaum, S.L. (1987) 1,25-dihydroxyvitamin D, modulates bone marrow macrophage precursor proliferation and differentiation: Upregulation of the mannose receptor. J . Biol. Chem., 262:15922-15929. Clohisy, D.R., Chappel, J.C., and Teitelbaum, S.L. (1989) Bone marrow-derived mononuclear phagocytes autoregulate mannose receptor expression. J. Biol. Chem., 2645370-5377. Miller, G.L. (1959) Protein determination for large numbers of samples. Anal. Chem., 31:964. Morel, P.A., Manolagas, S.C., Provvedini, D.M., Wegman, D.R., and Chiller, J.M. (1986) Interferon-y-induced IA expression in Wehi-3 cells is enhanced by the presence of 1,25-dihydroxyvitamin D,. J. Immunol., 136:2181-2186. Mulford, C.A., and Lodish, H.F. (1988) Endocytosis of the transferrin receptor is altered during differentiation of murine erythroleukemic cells. J . Biol. Chem., 2625455-5461. Octave, J.-N., Schneider, W.J., Trouet, A., and Crichton, R.R. (1983) Iron uptake and utilization by mammalian cells: I Cellular uptake of transferrin and iron. Trends Biochem. Sci., 8:217-220. Reitsma, P.H., Rothberg, P.G., Astrin, S.M., Trial, J., Bar-Shavit, Z., Hall, A., Teitelbaum, S.L., and Kahn, A.J. (1983) Regulation of myc gene expression in HL-60 leukaemia cells by a vitamin D metabolite. Nature, 306:492494. Seligman, P.A., and Allen, R.H. (1987) Isolation of transferrin receptor from human placenta. Methods Enzymol., 147;239-247. Sun-Sang, J.S., Nelson, R.S., and Silverstein, S.C. (1983) Yeast mannans inhibit binding and hagocytosis of zymosan by mouse peritoneal macrophages. J . Celf Biol., 96:160-166. Trowbridge, I.S., and Omary, M.B. (1981) Human cell surface glycoprotein related to cell proliferation is the receptor for transferrin. Proc. Natl. Acad. Sci. USA, 78:3039-3043. Vogt, A,, Mishell, R.I., and Dutton, R.W. (1969) Stimulation of DNA synthesis in cultures of mouse spleen cell suspensions by bovine transferrin. Exp. Cell Res., 54:195-200.