Proc. Natl. Acad. Sci. USA Vol. 74, No. 11, pp. 5026-5030, November 1977
Cell Biology
Complexes of inorganic pyrophosphate, orthophosphate, and calcium as stimulants of 3T3 cell multiplication (growth stimulation/divalent cations)
H. RUBIN AND H. SANUI Department of Molecular Biology and Virus Laboratory, University of California, Berkeley, California 94720
Communicated by Robley C. Williams, August 15, 1977
insoluble complexes of Ca2+, HP042, and/or PP1 interact with the cell membrane to produce responses similar to those induced by serum or hormones. The possibility is considered that pathological alterations in the intracellular availability of these normal cell constituents might be a source of abnormal growth behavior in the intact animal.
Addition of 0.1-0.5 mM sodium PPi for 17 hr ABSTRACT to confluent cultures of BALB/c 3T3 cells in low serum concentrations stimulated the incorporation of [3Hlthymidine into DNA to an extent equal to that produced by high serum concentration. PP; prevented much but not all of the cell detachment that accompanies decreasing the serum concentration of confluent cultures and it increased the saturation density of cultures in high serum concentrations. The stimulation had a sharp concentration dependence and was associated with the appearance in the medium of a flocculent precipitate. Stimulation and precipitate formation were dependent on Ca2+ and inorganic orthophos hate (HP042-) and were inhibited by Mg2+. M~ore than halfrthe Ca2rrequirement could be met with Sr2+. In the absence of PPi, supranormal concentrations of either Ca2+ or HP042- caused graded increases in [3H]thymidine incorporation and total cell yield. The effect of supranormal [Ca2+J depended on [HP042- and vice versa, and the Ca2+ requirement could be partially met by Sr2+. The stimulation was associated with increasing turbidity of the medium. Various other complexing agents of Ca2 , including the divalent cation ionophore A 23187, failed to produce stimulation of 3T3 cells. We conclude that water insoluble complexes of PP1, HP042-, and Ca2+ or, at much higher concentrations, the latter two together, stimulate 3T3 cells and we speculate that this is brought about by the association of these complexes with the cell membrane.
METHODS BALB/c 3T3 cells obtained from J. Bartholomew were cultured in Vogt and Dulbecco's modification of Eagle's medium (DME) with 10% calf serum (5). They were seeded at 5 X 104 or 105 cells per 60-mm Falcon plastic petri dish and grown to confluency before use in experiments except when a growth curve of the cells was determined. During the experiments, the medium was changed as indicated. Chicken embryo fibroblasts were obtained from 10-day-old embryos and cultured as described (6). They were grown in medium 199 with 2% tryptose phosphate broth and 1% chicken serum. The use of DME with calf serum in the culture of the chicken embryo fibroblasts did not alter the results of experiments. Cultures were labeled with [3H]thymidine and processed for scintillation counting or autoradiography as described (6). DME was prepared without Ca2 , Mg2+, HP042, or HCO3- for experiments in which any of these constituents were to be varied. Stock 100 mM aqueous solutions of the watersoluble chemicals used were made and diluted into the medium to achieve the desired concentration. The divalent cation ionophore A23187 was kept as a 2 mM solution in ethanol and diluted more than 1:100 in medium for experimental use.
We have proposed that the coordinate response of chicken embryo fibroblasts to external effectors is regulated by the availability of Mg2+ within the cell (1, 2). In extending this concept to mouse 3T3 cells we found these cells to be inhibited by Mg2+ deprivation in a manner similar to that of the chicken fibroblasts (unpublished data). This deprivation can be brought about by limiting the Mg2+ in the medium in the presence of decreased [Ca2+] or by complexing the Mg2+ with large amounts of inorganic pyrophosphate (PPi). However, we also found that PP1 in concentrations much smaller than those of Mg2+ markedly stimulated DNA synthesis in the 3T3 cells, particularly when they were in the confluent, quiescent state. This stimulation was dependent on Ca2+ and inorganic orthophosphate (HP042-)* and it was correlated with the appearance of a flocculent precipitate in the medium. This led us to investigate the stimulation of 3T3 cells by large amounts of Ca2+ (in the absence of PPj) previously reported by Dulbecco and Elkington (4). We found that this stimulation is also dependent on the presence of HP042- and is correlated with the formation of a precipitate. A similar stimulation, with precipitate formation, could be induced with the conventional concentrations of Ca2+ in tissue culture medium (-1.7 mM) by simply increasing the concentration of HP042 . In both the stimulation by PP1 and by supranormal [Ca2+], most of the Ca2+ requirement could be met by Sr2+. The results suggest that
RESULTS Varying amounts of PPi were added to cultures containing varying concentrations of dialyzed calf serum for 17 hr, and the rate of [3H]thymidine incorporation and the total protein content were determined. In the absence of serum and PP1, there was extensive cell detachment (Fig. 1). In 0 or 0.05 mM PP1 the rate of [3H]thymidine incorporation increased with increasing serum concentration. Increasing [PPj] to 0.1 mM caused a marked stimulation in the incorporation of [3H]thymidine and in cell survival with serum concentrations from 0 to 5% but much less stimulation with 10% serum. Indeed, the rate of incorporation of [3H]thymidine in 1% serum plus 0.1 mM PP1 far exceeded that in 10% serum in either 0 or 0.1 mM PP1. There was no further stimulation of [3H]thymidine incorporation, except in 10% serum, when PP1 was increased up Abbreviation: DME, Dulbecco-Vogt modification of Eagle's medi-
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
um.
* At pH 7.4, about 80% of the orthophosphate exists as HP042- and
most of the rest as H2PO41- (3). We shall refer to it here as HP042, keeping in mind that other forms exist.
5026
so 50
f
e
F-
A
----:---::------------A..
,
300
20-,
I~~~~~~~~~~~~~~~~
E0 300
5027
Proc. Natl. Acad. Sci. USA 74 (1977)
Cell Biology: Rubin and Sanui
0
250
C
300+ 200 -m 2.0
-
j;.'
50
o Co2M
30 _,' 1
++
0.
0100 -+ . . -~ ?~~4.0mM/
C
E' A;
~ ~~~~~~~~~~~~~~~~-
'
K
E/ / ~50 .2
~~~1.0mM1
a) 30
10
~ ~
/0.6mM
±
5+ 0
0.1
0.2
0.3
0.4 PP., mnM
0.5
0
1.0
FIG. 1. Effects of PPi on [3H]thymidine incorporation and total protein of 3T3 cells in varying concentrations of dialyzed calf serum. Cultures of 3T3 cells were grown to confluency in DME with 10% calf serum. The medium was then changed to DME with the different concentrations of dialyzed calf serum shown on the curves in the Lower panel and varying amounts of PPi. The cultures were incubated at 370 for 17 hr, and the medium was replaced with DME containing [3H]thymidine at 1.0 gCi/ml. After 1 hr of further incubation, the cultures were washed with saline, treated with cold 5% trichloroacetic acid for 10 min, and extracted with 0.1 M NaOH for scintillation counting and Lowry protein determinations. In addition, two cultures that had had no medium change were labeled with [3H]thymidine in DME and extracted at the start of the 17-hr incubation period (* 4-). A flocculent precipitate was seen floating in the medium, with some settling on the cells, at PPi concentrations of 0.1 mM or greater, but none was visible at 0 or 0.05 mM PP1.
to 0.5 mM. In 1.0 mM PP1, the rate of incorporation decreased from the maximum at all serum concentrations. Although PP1 in low serum concentrations was more effective than 10% serum in stimulating [3H]thymidine incorporation, it did not fully substitute for 10% serum in preventing cell detachment. At all [PP ] > 0.1 mM, a flocculent precipitate appeared in the medium, thus coinciding with stimulation of [3H]thymidine incorporation. The minimal concentration of PPi that produced both effects was independent of serum concentration from 0 to 5% serum, indicating that the effects were not mediated by complexes of PP1 with some serum constituent. The ionic requirements were determined for the stimulation of [3H]thymidine incorporation by PP1. The concentration of PP1 required for stimulation was inversely proportional to [Ca2+] (Fig. 2). There was no stimulation by PP1 up to 1.0 mM when [Ca2+] was 0.6 mM or less (data not shown). Again, a coincidence was found between the capacity of PP1 to stimulate and the appearance of a flocculent precipitate. At the high PP1 concentration of 1.0 mM, however, at which most of the stimulatory activity was lost, the precipitate persisted. As in the previous experiment, the stimulation went from nil to maximal with a small change in [PPj]. In contrast to the requirement of Ca2+ for stimulation by PPi, Mg2+ interfered with the effect (Fig. 3). Indeed, 2.5 mM Mg2+ completely blocked the stimulation, although there was some precipitate formed with the two highest [PP1]. Decreasing the pH of the medium from 7.4 to 6.8 caused an increase in the
0.1
0.2
0.4
0.3
1.0
0.5
FIG. 2. Effects of varying [Ca2+] on stimulation of 3T3 cells and precipitate formation by PPi. The growth medium of 3T3 cells was replaced with DME containing the concentrations of Ca2+ indicated on the curves, with 2% undialyzed calf serum. The cultures were incubated for 17 hr and processed as in Fig. 1. With each point on each curve is a sign indicating the presence (+), absence (-), or barely detectable presence (i) of a flocculent precipitate in the medium and on the cells.
minimal [PPj] required to produce both stimulation and precipitation (data not shown). The efficacy of PP1 stimulation, and precipitation, increased with increasing [HP042-] (conventional concentration -1.0 mM) (Fig. 4). When dialyzed serum was used and no HP042- was added to the medium (final [HP042-] = 0.004 mM), there was only a 2.5-fold instead of >15-fold maximum stimulation by PP1 although full stimulation could be obtained at such low [HP042-1 by adding 20% dialyzed
cm
E a)0._ C
._
E .A,
1I
0
0.1
0.2
0.3
0.4
0.5
1.0
PP;, TiMi FIG. 3. Effects of Mg2+ on the stimulation of 3T3 cells by PP1. The procedure was the same as that of Fig. 2 except that [Mg2+] was varied from 0.1 to 2.5 mM as shown. [Ca2+I and [HP042-] were constant at 1.7 and 1.0 mM, respectively. Precipitate formation was seen wherever there was stimulation. However, precipitate also appeared at all [Mg2+] in the presence of 1.0 mM PP1, and these combinations caused little or no stimulation.
Proc. Natl. Acad. Sci. USA 74 (1977)
Cell Biology: Rubin and Sanui
5028
100 300 c
200k am 0.
./
0)
3.0mM
E 100
30F
i
-
- Pptt / 30 -No 4
A
._.
0
0. 0)
1.0mM015M/ /0.145mM /0.045mM
j
50
/
HP042
0.
C
IO
/p
50
Aso
201
-
Y
E0. C
ppt.,
10
E
,_-I-_
20 I-
5
\ -.% n -+A. _A
/
3
.0
0.15 0.20 0.25 0.30 PPi, mM FIG. 4. Effects of HP042- on the stimulation of 3T3 cells by PP1. The procedure was the same as that of Figs. 2 and 3 except that [HP042-] was varied as shown and [Ca2+] and [Mg2+] were constant at 1.7 and 0.8 mM, respectively. Precipitate was seen in all markedly stimulated cultures but in none of the unstimulated cultures, as indicated by the horizontal broken line. The rate of [3H]thymidine incorporation was unusually high in this experiment because multiply 0.05
0.10
passaged cells were used.
(data not shown). These results show that the formation of insoluble complexes of PPi, HP042-, and Ca2+ correlated with stimulation but that, if these complexes formed in the presence of high [Mg2+], the stimulation was blocked. Sr2+ supplemented but did not completely replace Ca2+ in stimulation of the cells by PPi (Fig. 5). Thus, a minimum [Ca2+] of 1.2 mM was required for cell stimulation by 0.2 mM PP when Ca2+ alone was used, but only one-third as much Ca2+ was required for substantial stimulation when supplemented with Sr2+. The capacity of PPi to stimulate with Sr2+ in the absence of Ca2+ was difficult to ascertain because Sr2+ stimulated the cells slightly in the absence of both PPi and Ca2 When either PPi or 10% serum was added to confluent cultures maintained overnight in low (2%) serum concentration, there was a lag period of about 10 hr before the rate of [3H]thymidine incorporation increased. It then increased sharply in both groups but declined more rapidly with 10% serum than with PPi. There was no effect of PP1 on multiplication rate of cells growing exponentially, but PP1 increased the saturation density 2-fold. When the rate of incorporation of [3H]thymidine was measured by autoradiographic counting of labeled nuclei, the relative increases caused by PP1 or by high serum concentrations were the same as those found by scintillation counting of alkali-extracted material. Scintillation counting gave more precise and reproducible results, however, because the labeled nuclei in quiescent cultures were unevenly distributed, with a heavy concentration in the le.ss-crowded periphery of the dish. The requirement for Ca2+ and HP042- in stimulation by PP1 led us to examine stimulation by these medium constituents in the absence of PPi. Because of the frequent and erratic toxicity of precipitates formed by high concentrations of Ca2+ and HP042- in low serum concentration, it was necessary to use 10% serum, which kept most of the precipitate from settling on the cells. Because 10% serum itself gave substantial stimulation after overnight incubation, it was necessary to wait about 40 hr, when serum stimulation had subsided, to measure the effects of supranormal concentrations of Ca2+ and HP042-. Increasing
serum
.
2
I 0
0.4
0.8
1.2
1.6
Ca2+, mM FIG. 5. Substitution of Sr2+ for Ca2+ in the stimulation of 3T3 cells by PP1. The procedure was the same as that of Fig. 2, except for the following. In one group of cultures, [Ca2+1 was varied from 0.44 to 1.44 mM in the presence (0) or absence (0) of 0.2 mM PPi. In another group, [Ca2+1 was varied from 0.04 to 0.64 mM, and supplemented with Sr2+ to a combined total of 1.44 mM in the presence (,&) or absence (A) of PP1. On the PPi curves, (+) indicates presence and (-) indicates absence of precipitate. [Mg2+] was 0.8 mM and [HPO42-1 was 1.0 mM in all cultures. The PPi concentration of 0.2 mM was used because it gave the most consistent and reliable results.
[Ca2+] from 1.7 mM to 10.0 mM in the presence of 1.0 mM HP042- caused a graded increase in [3H]thymidine incorporation (Fig. 6). Increasing [HP042-] from 1.0 mM to 10 mM in the presence of 1.7 mM Ca2+ showed approximately the same effect. There was an increase in total protein per culture with increases of up to about 8 mM of either Ca2+ or HP042-, but at higher concentrations total protein decreased, indicating some cell detachment, although the rate of [3H]thymidine incorporation remained high in the remaining cells. The stimulation by Ca2+ depended on the concentration of HP042 in the medium (Fig. 7 upper). The increased stimulation achieved by increasing the concentrations of Ca2+ or HP042- was accompanied by an increase in the turbidity of the medium (Fig. 7 lower). Both stimulation and precipitate formation by either supranormal Ca2+ or HP042- had a much more gradual dependence on concentration than did the same effects produced by PP1. Replacement of half or more of the supranormal concentrations of Ca2+ by Sr2+ produced substantial stimulation of [3H]thymidine incorporation, although not fully equivalent to that produced by equimolar concentrations of Ca2+ alone (Table 1). High concentrations of Mg2+ partially interfered with the Ca2+ effect. None of the following agents that complex Ca2+ reproduced the stimulation produced by PPi within the range of 0-1.0 mM: sodium tripolyphosphate; ADP; ATP; ethylene glycol-bis(f3aminoethyl ether)-N-N'-tetraacetate (EGTA); and ethylenediaminetetraacetate (EDTA). Indeed, EDTA was highly inhibitory in the range tested, due to binding of Zn2+ (6). The divalent cation ionophore A23187 was without effect in the range 10-8-10-7 M and was toxic at concentrations >10-6 M.
Proc. Natl. Acad. Sci. USA 74 (1977)
Cell Biology: Rubin and Sanui
5029
700
600 °
400
300 100 E200
-W ~ ~~
~
~
~
~
~
~
0
I-
--7
-
CD
Co2.'
30
20/ -C E
50
PP;, FIG. 6.
Ca2
or HPO22
mM
Stimulation of 3T3 cells by supranormal concentrations
of either CaI+ or HP042-. The medium of confluent 3T3 cultures was replaced by: fresh DME with
Ca2+ mM
10%6 calf serum and varying amounts of
in 1.0 mM HP042- (0); by varying amounts of HP042- in 1.7
Ca2+ (A); or by varying amounts of PP4 in 1.0mM HP042- plus Ca26 (o). The cultures were then incubated at 370 for 42 hr
1.7 mM
before being labeled with [3H]thymidine and processed as in Fig. 1.
flocculent precipitate was present in the medium of cultures stimulated by PP:. In the cultures with increasing [Ca2r] or increasing [HP042-] there was an increasing amount of a granular precipitste,
A
which settled
on
the cells,
and of turbidity
in the medium.
PP1 produced
either no stimulation or only marginal stimulation
when added
to
quiescent chicken
centrations that were
embryo
highly stimulatory
fibroblasts
in con-
to 3T3 cells. Supra-
normal [Ca2+] produced only marginal stimulation chicken embryo fibroblasts.
of the
DISCUSSION progress of chicken embryo fibroblasts and mammalian lymphocytes through the cell cycle (7-9). In 3T3 cells, many of these materials such as proteases (10), insulin (11), and toxic metals are either ineffective or effective only under highly restrictive conditions. PP1 differs from the aforementioned effectors in that it stimulates 3T3 cells but not chicken embryo fibroblasts. In its selectivity for 3T3 cells, PP1 stimulation resembles that reported for supranormal concentrations of Ca2+ (4, 12) and confirmed above. Indeed, the stimulation by PP1 requires the presence of Ca2+, although subnormal concentrations of Ca2+ are fully adequate to produce the effect if 0.5 mM PP1 is present. The most striking features of the stimulation are: (i) its sharp dependence on the concentration of PP1, and (ii) its association with the appearance of a flocculent precipitate in the medium. All treatments that affect the concentration at which PP1 forms a precipitate with Ca2+ also affect in an identical manner the concentration that stimulates the cells. Thus, HP042- is required to form the precipitate and to stimulate the cells, and Mg2+ interferes with both effects. From the sharp delineation between concentrations of PP1 that do or do not cause both precipitation and stimulation, we conclude that the stimulation is caused by water-insoluble complexes of PP1, Ca2+, and HP042. This raises the question
A wide variety of unrelated materials accelerate
Ca2+,
mM
FIG. 7. Dependence of supranormal Ca2+ stimulation on [HP042-] in the medium. The procedure was the same as that of Fig. 6 except that a range of [Ca2+1 was used in the presence of each of the three concentrations of HP042- indicated on the curves for [3H] thymidine incorporation in the Upper panel. An aliquot of each medium combination was incubated overnight at 370 without cells to allow complexes of Ca2+ and HP042- to form, and the turbidities were measured as absorbance at 310 nm in a double-beam spectrophotometer, with the sample with the lowest [Ca2+] and [HP042-] used as reference (Lower). * indicates a heavy precipitate that settled on the cells and caused some cell death and detachment. -
of whether the stimulation is caused by providing any one or several of these constituents to act individually as internal substrates or effectors of metabolism, or whether the complex is acting in a manner similar to macromolecular stimulants in serum that are presumed to act at the cell surface. The concentration of Ca2+ in the medium can be decreased to far below the conventional levels without markedly affecting the rate of progress of 3T3 cells through the cell cycle (ref 13; Fig. 7), indicating that the external supply of Ca2+ is not a limiting factor for growth unless extreme deprivation is imposed upon the cells. It, therefore, seems unlikely that merely providing more Ca2+ Table 1. Effects of Sr2+ and Mg2+ in mimicking or inhibiting the stimulation of 3T3 cells by supranormal Ca2+ concentration Relative rate of Relative amt. [3H]thymidine Ca2+, Sr2+, Mg2+, of protein/dish mM mM incorp. mM 1.7 4.0 8.0 1.7 4.0 1.7 8.0
0 0 0 6.3 4.0 0 0
0.8 0.8 0.8 0.8 0.8 9.8 9.8
1.0 1.2 12.30 5.58 9.43 0.81 5.09
1.0 1.1 2.96
2.19 2.94 1.04 2.15
DME medium was prepared with 1.0 mM HP042- and the concentrations of Ca2+, Sr2+, and Mg2+ shown; and 10% calf serum was added. Confluent cultures of 3T3 cells were incubated in this medium at 370 for 41 hr and then labeled with [3H]thymidine and processed as in Fig. 1.
5030
Cell Biology: Rubin and Sanui
in complexes with PPi would act as a stimulant. This view is reinforced by the finding that Sr2+ can substitute for about 75% of the Ca2+ in the PPi stimulation and that the Ca2+ ionophore A23187 is totally ineffective. That PPi itself would be a limiting factor seems even less likely. It is not needed for the growth of cells and is not present in conventional medium. Because PP1 is a product of many reactions associated with nucleic acid and protein synthesis, the presence of high concentrations of PPi in the cells would tend to inhibit these reactions which are essential for cell cycle traverse. HP042- is also an unlikely candidate as a direct control because neither its rate of uptake nor its pool size is limiting in the growth of 3T3 cells (14). Substances that form soluble complexes with Ca2+, such as EGTA, EDTA, ADP, and ATP, have no stimulatory effect. Sodium tripolyphosphate causes a slight increase in turbidity within a narrow concentration range in Ca2+-containing medium (data not shown) but causes no gross precipitate or stimulation of the 3T3 cells. We think it likely, therefore, that the stimulation by PP1 is actually brought about by complexes of PPi, Ca2+, and HP042- that localize in the cell membrane, possibly because of the hydrophobicity, and stimulate the cell by altering the configuation of the membrane. The complexes of PPi, Ca2+, and HP042 appear to have some specificity because they are ineffective in stimulating chicken embryo fibroblasts. Also, when the concentration of PPi is increased to 1.0 mM, it loses much of its stimulatory power, even though large amounts of the flocculent precipitate form. This is not due to its complexing of Mg2+, because this concentration of PPi has no effect on stimulation of the 3T3 cells by high concentrations of serum. It is more likely that the ratio of the constituents of the complex or its size or both are important elements in determining interaction with cells. The major outlines of stimulation by PPi are similar to those of stimulation by supranormal concentrations of Ca2+. Both require HP042- and are accompanied by formation of insoluble complexes. Ca2+ can be replaced by Sr2+ in both cases and Mg2+ is inhibitory, although more dramatically so in PPi stimulation. Both treatments are effective in 3T3 cells and not in chicken embryo fibroblasts. Of course, the minimal amount of PPi required for a given level of stimulation is almost two orders of magnitude less than the amount of Ca2+ required for equivalent stimulation in the absence of PPi; and the minimal Ca2+ requirement for PPi stimulation is one order of magnitude less than the requirement of Ca2+ for equivalent stimulation in the absence of PPi (data not shown). These apparent differences in Ca2+ requirement for the two types of stimulation are balanced by the fact that stimulation can occur in conventional concentrations of Ca2+ (in the absence of PPi) by simply increasing [HP042-]. The common thread in all these forms of stimulation appears to be the formation of water-insoluble complexes of phosphate compounds and Ca2+ (or Sr2+) which,
Proc. Natl. Acad. Sci. USA 74 (1977)
we presume, associate with the surface membranes of the cell and act there to stimulate metabolism and growth. These complexes have been shown to be effective only in BALB/c 3T3 cells and to have little or no effect in secondary hamster (4) and chicken embryo cells and BHK cells (4). This raises once again the question of the true nature of the 3T3 cells. Although they arose from mouse embryo cultures that were fibroblastic in appearance (15), and they have been referred to repeatedly as fibroblasts (4), recent evidence indicates they are more closely related to cells involved in blood vessel formation (16). Their sensitivity to stimulation by complexes of phosphates and Ca2+ may reflect the special vasoformative nature of these cells. Finally, it should be kept in mind that PPi, Ca2+, and HP042- are normal cellular constitutents. Although Ca2+ normally occurs in cells at concentrations far below those required for precipitation with the phosphate compounds, under certain conditions, massive amounts of Ca2+ and phosphate can accumulate (17). Any defect that increases intracellular [Ca2+], such as an inward Ca2+ leak, could lead to the formation of growth-stimulatory complexes with inorganic phosphates and, if sustained, to tumor formation. We thank Berbie M. Chu for her fine technical assistance. This investigation was supported by National Institutes of Health Research Grant 15744 from the National Cancer Institute. 1. Rubin, H. (1975) Proc. Natl. Acad. Sci. USA 72,3551-3555. 2. Rubin, H. (1976) J. Cell. Physiol. 89,613-626. 3. Neuman, W. & Neuman, M. (1958) The Chemical Dynamics of Bone Mineral (University of Chicago Press, Chicago, IL). 4. Dulbecco, R. & Elkington, J. (1975) Proc. Natl. Acad. Sci. USA
72,1584-1588. 5. Vogt, M. & Dulbecco, R. (1963) Proc. Natl. Acad. Sci. USA 49, 171-179. 6. Rubin, H. (1972) Proc. Natl. Acad. Sci. USA 69,712-716. 7. Rubin, H. & Koide, T. (1975) J. Cell. Physiol. 86,47-58. 8. Novogrodsky, A. & Katchalski, E. (1971) Biochim. Biophys. Acta 228,759-783. 9. Novogrodsky, A. & Katchalski, E. (1971) FEBS Lett. 12, 297300. 10. Glynn, R., Thrash, C. & Cunningham, D. (1973) Proc. Natl. Acad. Sci. USA 70,2676-2677. 11. Jimenez de Asua, L., O'Farrell, M., Bennett, D., Clingan, D. & Rudland, P. (1977) Nature 265,151-153. 12. Boynton, A. & Whitfield, J. (1976) In Vitro 12,479-484. 13. Boynton, A., Whitfield, J., Isaacs, R. & Morton, H. (1974) In Vitro 10,12-17. 14. Barsh, G., Greenberg, D. & Cunningham, D. (1977) J. Cell. Physlol. 92, 115-128. 15. Todaro, G. & Green, H. (1963) J. Cell Biol. 17,299-313. 16. Boone, C. (1975) Science 188,68-70. 17. Lehninger, A., Carafoli, E. & Rossi, E. (1967) Adv. Enzymol. 29, 259-320.
Cell Biology
Complexes of inorganic pyrophosphate, orthophosphate, and calcium as stimulants of 3T3 cell multiplication (growth stimulation/divalent cations)
H. RUBIN AND H. SANUI Department of Molecular Biology and Virus Laboratory, University of California, Berkeley, California 94720
Communicated by Robley C. Williams, August 15, 1977
insoluble complexes of Ca2+, HP042, and/or PP1 interact with the cell membrane to produce responses similar to those induced by serum or hormones. The possibility is considered that pathological alterations in the intracellular availability of these normal cell constituents might be a source of abnormal growth behavior in the intact animal.
Addition of 0.1-0.5 mM sodium PPi for 17 hr ABSTRACT to confluent cultures of BALB/c 3T3 cells in low serum concentrations stimulated the incorporation of [3Hlthymidine into DNA to an extent equal to that produced by high serum concentration. PP; prevented much but not all of the cell detachment that accompanies decreasing the serum concentration of confluent cultures and it increased the saturation density of cultures in high serum concentrations. The stimulation had a sharp concentration dependence and was associated with the appearance in the medium of a flocculent precipitate. Stimulation and precipitate formation were dependent on Ca2+ and inorganic orthophos hate (HP042-) and were inhibited by Mg2+. M~ore than halfrthe Ca2rrequirement could be met with Sr2+. In the absence of PPi, supranormal concentrations of either Ca2+ or HP042- caused graded increases in [3H]thymidine incorporation and total cell yield. The effect of supranormal [Ca2+J depended on [HP042- and vice versa, and the Ca2+ requirement could be partially met by Sr2+. The stimulation was associated with increasing turbidity of the medium. Various other complexing agents of Ca2 , including the divalent cation ionophore A 23187, failed to produce stimulation of 3T3 cells. We conclude that water insoluble complexes of PP1, HP042-, and Ca2+ or, at much higher concentrations, the latter two together, stimulate 3T3 cells and we speculate that this is brought about by the association of these complexes with the cell membrane.
METHODS BALB/c 3T3 cells obtained from J. Bartholomew were cultured in Vogt and Dulbecco's modification of Eagle's medium (DME) with 10% calf serum (5). They were seeded at 5 X 104 or 105 cells per 60-mm Falcon plastic petri dish and grown to confluency before use in experiments except when a growth curve of the cells was determined. During the experiments, the medium was changed as indicated. Chicken embryo fibroblasts were obtained from 10-day-old embryos and cultured as described (6). They were grown in medium 199 with 2% tryptose phosphate broth and 1% chicken serum. The use of DME with calf serum in the culture of the chicken embryo fibroblasts did not alter the results of experiments. Cultures were labeled with [3H]thymidine and processed for scintillation counting or autoradiography as described (6). DME was prepared without Ca2 , Mg2+, HP042, or HCO3- for experiments in which any of these constituents were to be varied. Stock 100 mM aqueous solutions of the watersoluble chemicals used were made and diluted into the medium to achieve the desired concentration. The divalent cation ionophore A23187 was kept as a 2 mM solution in ethanol and diluted more than 1:100 in medium for experimental use.
We have proposed that the coordinate response of chicken embryo fibroblasts to external effectors is regulated by the availability of Mg2+ within the cell (1, 2). In extending this concept to mouse 3T3 cells we found these cells to be inhibited by Mg2+ deprivation in a manner similar to that of the chicken fibroblasts (unpublished data). This deprivation can be brought about by limiting the Mg2+ in the medium in the presence of decreased [Ca2+] or by complexing the Mg2+ with large amounts of inorganic pyrophosphate (PPi). However, we also found that PP1 in concentrations much smaller than those of Mg2+ markedly stimulated DNA synthesis in the 3T3 cells, particularly when they were in the confluent, quiescent state. This stimulation was dependent on Ca2+ and inorganic orthophosphate (HP042-)* and it was correlated with the appearance of a flocculent precipitate in the medium. This led us to investigate the stimulation of 3T3 cells by large amounts of Ca2+ (in the absence of PPj) previously reported by Dulbecco and Elkington (4). We found that this stimulation is also dependent on the presence of HP042- and is correlated with the formation of a precipitate. A similar stimulation, with precipitate formation, could be induced with the conventional concentrations of Ca2+ in tissue culture medium (-1.7 mM) by simply increasing the concentration of HP042 . In both the stimulation by PP1 and by supranormal [Ca2+], most of the Ca2+ requirement could be met by Sr2+. The results suggest that
RESULTS Varying amounts of PPi were added to cultures containing varying concentrations of dialyzed calf serum for 17 hr, and the rate of [3H]thymidine incorporation and the total protein content were determined. In the absence of serum and PP1, there was extensive cell detachment (Fig. 1). In 0 or 0.05 mM PP1 the rate of [3H]thymidine incorporation increased with increasing serum concentration. Increasing [PPj] to 0.1 mM caused a marked stimulation in the incorporation of [3H]thymidine and in cell survival with serum concentrations from 0 to 5% but much less stimulation with 10% serum. Indeed, the rate of incorporation of [3H]thymidine in 1% serum plus 0.1 mM PP1 far exceeded that in 10% serum in either 0 or 0.1 mM PP1. There was no further stimulation of [3H]thymidine incorporation, except in 10% serum, when PP1 was increased up Abbreviation: DME, Dulbecco-Vogt modification of Eagle's medi-
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
um.
* At pH 7.4, about 80% of the orthophosphate exists as HP042- and
most of the rest as H2PO41- (3). We shall refer to it here as HP042, keeping in mind that other forms exist.
5026
so 50
f
e
F-
A
----:---::------------A..
,
300
20-,
I~~~~~~~~~~~~~~~~
E0 300
5027
Proc. Natl. Acad. Sci. USA 74 (1977)
Cell Biology: Rubin and Sanui
0
250
C
300+ 200 -m 2.0
-
j;.'
50
o Co2M
30 _,' 1
++
0.
0100 -+ . . -~ ?~~4.0mM/
C
E' A;
~ ~~~~~~~~~~~~~~~~-
'
K
E/ / ~50 .2
~~~1.0mM1
a) 30
10
~ ~
/0.6mM
±
5+ 0
0.1
0.2
0.3
0.4 PP., mnM
0.5
0
1.0
FIG. 1. Effects of PPi on [3H]thymidine incorporation and total protein of 3T3 cells in varying concentrations of dialyzed calf serum. Cultures of 3T3 cells were grown to confluency in DME with 10% calf serum. The medium was then changed to DME with the different concentrations of dialyzed calf serum shown on the curves in the Lower panel and varying amounts of PPi. The cultures were incubated at 370 for 17 hr, and the medium was replaced with DME containing [3H]thymidine at 1.0 gCi/ml. After 1 hr of further incubation, the cultures were washed with saline, treated with cold 5% trichloroacetic acid for 10 min, and extracted with 0.1 M NaOH for scintillation counting and Lowry protein determinations. In addition, two cultures that had had no medium change were labeled with [3H]thymidine in DME and extracted at the start of the 17-hr incubation period (* 4-). A flocculent precipitate was seen floating in the medium, with some settling on the cells, at PPi concentrations of 0.1 mM or greater, but none was visible at 0 or 0.05 mM PP1.
to 0.5 mM. In 1.0 mM PP1, the rate of incorporation decreased from the maximum at all serum concentrations. Although PP1 in low serum concentrations was more effective than 10% serum in stimulating [3H]thymidine incorporation, it did not fully substitute for 10% serum in preventing cell detachment. At all [PP ] > 0.1 mM, a flocculent precipitate appeared in the medium, thus coinciding with stimulation of [3H]thymidine incorporation. The minimal concentration of PPi that produced both effects was independent of serum concentration from 0 to 5% serum, indicating that the effects were not mediated by complexes of PP1 with some serum constituent. The ionic requirements were determined for the stimulation of [3H]thymidine incorporation by PP1. The concentration of PP1 required for stimulation was inversely proportional to [Ca2+] (Fig. 2). There was no stimulation by PP1 up to 1.0 mM when [Ca2+] was 0.6 mM or less (data not shown). Again, a coincidence was found between the capacity of PP1 to stimulate and the appearance of a flocculent precipitate. At the high PP1 concentration of 1.0 mM, however, at which most of the stimulatory activity was lost, the precipitate persisted. As in the previous experiment, the stimulation went from nil to maximal with a small change in [PPj]. In contrast to the requirement of Ca2+ for stimulation by PPi, Mg2+ interfered with the effect (Fig. 3). Indeed, 2.5 mM Mg2+ completely blocked the stimulation, although there was some precipitate formed with the two highest [PP1]. Decreasing the pH of the medium from 7.4 to 6.8 caused an increase in the
0.1
0.2
0.4
0.3
1.0
0.5
FIG. 2. Effects of varying [Ca2+] on stimulation of 3T3 cells and precipitate formation by PPi. The growth medium of 3T3 cells was replaced with DME containing the concentrations of Ca2+ indicated on the curves, with 2% undialyzed calf serum. The cultures were incubated for 17 hr and processed as in Fig. 1. With each point on each curve is a sign indicating the presence (+), absence (-), or barely detectable presence (i) of a flocculent precipitate in the medium and on the cells.
minimal [PPj] required to produce both stimulation and precipitation (data not shown). The efficacy of PP1 stimulation, and precipitation, increased with increasing [HP042-] (conventional concentration -1.0 mM) (Fig. 4). When dialyzed serum was used and no HP042- was added to the medium (final [HP042-] = 0.004 mM), there was only a 2.5-fold instead of >15-fold maximum stimulation by PP1 although full stimulation could be obtained at such low [HP042-1 by adding 20% dialyzed
cm
E a)0._ C
._
E .A,
1I
0
0.1
0.2
0.3
0.4
0.5
1.0
PP;, TiMi FIG. 3. Effects of Mg2+ on the stimulation of 3T3 cells by PP1. The procedure was the same as that of Fig. 2 except that [Mg2+] was varied from 0.1 to 2.5 mM as shown. [Ca2+I and [HP042-] were constant at 1.7 and 1.0 mM, respectively. Precipitate formation was seen wherever there was stimulation. However, precipitate also appeared at all [Mg2+] in the presence of 1.0 mM PP1, and these combinations caused little or no stimulation.
Proc. Natl. Acad. Sci. USA 74 (1977)
Cell Biology: Rubin and Sanui
5028
100 300 c
200k am 0.
./
0)
3.0mM
E 100
30F
i
-
- Pptt / 30 -No 4
A
._.
0
0. 0)
1.0mM015M/ /0.145mM /0.045mM
j
50
/
HP042
0.
C
IO
/p
50
Aso
201
-
Y
E0. C
ppt.,
10
E
,_-I-_
20 I-
5
\ -.% n -+A. _A
/
3
.0
0.15 0.20 0.25 0.30 PPi, mM FIG. 4. Effects of HP042- on the stimulation of 3T3 cells by PP1. The procedure was the same as that of Figs. 2 and 3 except that [HP042-] was varied as shown and [Ca2+] and [Mg2+] were constant at 1.7 and 0.8 mM, respectively. Precipitate was seen in all markedly stimulated cultures but in none of the unstimulated cultures, as indicated by the horizontal broken line. The rate of [3H]thymidine incorporation was unusually high in this experiment because multiply 0.05
0.10
passaged cells were used.
(data not shown). These results show that the formation of insoluble complexes of PPi, HP042-, and Ca2+ correlated with stimulation but that, if these complexes formed in the presence of high [Mg2+], the stimulation was blocked. Sr2+ supplemented but did not completely replace Ca2+ in stimulation of the cells by PPi (Fig. 5). Thus, a minimum [Ca2+] of 1.2 mM was required for cell stimulation by 0.2 mM PP when Ca2+ alone was used, but only one-third as much Ca2+ was required for substantial stimulation when supplemented with Sr2+. The capacity of PPi to stimulate with Sr2+ in the absence of Ca2+ was difficult to ascertain because Sr2+ stimulated the cells slightly in the absence of both PPi and Ca2 When either PPi or 10% serum was added to confluent cultures maintained overnight in low (2%) serum concentration, there was a lag period of about 10 hr before the rate of [3H]thymidine incorporation increased. It then increased sharply in both groups but declined more rapidly with 10% serum than with PPi. There was no effect of PP1 on multiplication rate of cells growing exponentially, but PP1 increased the saturation density 2-fold. When the rate of incorporation of [3H]thymidine was measured by autoradiographic counting of labeled nuclei, the relative increases caused by PP1 or by high serum concentrations were the same as those found by scintillation counting of alkali-extracted material. Scintillation counting gave more precise and reproducible results, however, because the labeled nuclei in quiescent cultures were unevenly distributed, with a heavy concentration in the le.ss-crowded periphery of the dish. The requirement for Ca2+ and HP042- in stimulation by PP1 led us to examine stimulation by these medium constituents in the absence of PPi. Because of the frequent and erratic toxicity of precipitates formed by high concentrations of Ca2+ and HP042- in low serum concentration, it was necessary to use 10% serum, which kept most of the precipitate from settling on the cells. Because 10% serum itself gave substantial stimulation after overnight incubation, it was necessary to wait about 40 hr, when serum stimulation had subsided, to measure the effects of supranormal concentrations of Ca2+ and HP042-. Increasing
serum
.
2
I 0
0.4
0.8
1.2
1.6
Ca2+, mM FIG. 5. Substitution of Sr2+ for Ca2+ in the stimulation of 3T3 cells by PP1. The procedure was the same as that of Fig. 2, except for the following. In one group of cultures, [Ca2+1 was varied from 0.44 to 1.44 mM in the presence (0) or absence (0) of 0.2 mM PPi. In another group, [Ca2+1 was varied from 0.04 to 0.64 mM, and supplemented with Sr2+ to a combined total of 1.44 mM in the presence (,&) or absence (A) of PP1. On the PPi curves, (+) indicates presence and (-) indicates absence of precipitate. [Mg2+] was 0.8 mM and [HPO42-1 was 1.0 mM in all cultures. The PPi concentration of 0.2 mM was used because it gave the most consistent and reliable results.
[Ca2+] from 1.7 mM to 10.0 mM in the presence of 1.0 mM HP042- caused a graded increase in [3H]thymidine incorporation (Fig. 6). Increasing [HP042-] from 1.0 mM to 10 mM in the presence of 1.7 mM Ca2+ showed approximately the same effect. There was an increase in total protein per culture with increases of up to about 8 mM of either Ca2+ or HP042-, but at higher concentrations total protein decreased, indicating some cell detachment, although the rate of [3H]thymidine incorporation remained high in the remaining cells. The stimulation by Ca2+ depended on the concentration of HP042 in the medium (Fig. 7 upper). The increased stimulation achieved by increasing the concentrations of Ca2+ or HP042- was accompanied by an increase in the turbidity of the medium (Fig. 7 lower). Both stimulation and precipitate formation by either supranormal Ca2+ or HP042- had a much more gradual dependence on concentration than did the same effects produced by PP1. Replacement of half or more of the supranormal concentrations of Ca2+ by Sr2+ produced substantial stimulation of [3H]thymidine incorporation, although not fully equivalent to that produced by equimolar concentrations of Ca2+ alone (Table 1). High concentrations of Mg2+ partially interfered with the Ca2+ effect. None of the following agents that complex Ca2+ reproduced the stimulation produced by PPi within the range of 0-1.0 mM: sodium tripolyphosphate; ADP; ATP; ethylene glycol-bis(f3aminoethyl ether)-N-N'-tetraacetate (EGTA); and ethylenediaminetetraacetate (EDTA). Indeed, EDTA was highly inhibitory in the range tested, due to binding of Zn2+ (6). The divalent cation ionophore A23187 was without effect in the range 10-8-10-7 M and was toxic at concentrations >10-6 M.
Proc. Natl. Acad. Sci. USA 74 (1977)
Cell Biology: Rubin and Sanui
5029
700
600 °
400
300 100 E200
-W ~ ~~
~
~
~
~
~
~
0
I-
--7
-
CD
Co2.'
30
20/ -C E
50
PP;, FIG. 6.
Ca2
or HPO22
mM
Stimulation of 3T3 cells by supranormal concentrations
of either CaI+ or HP042-. The medium of confluent 3T3 cultures was replaced by: fresh DME with
Ca2+ mM
10%6 calf serum and varying amounts of
in 1.0 mM HP042- (0); by varying amounts of HP042- in 1.7
Ca2+ (A); or by varying amounts of PP4 in 1.0mM HP042- plus Ca26 (o). The cultures were then incubated at 370 for 42 hr
1.7 mM
before being labeled with [3H]thymidine and processed as in Fig. 1.
flocculent precipitate was present in the medium of cultures stimulated by PP:. In the cultures with increasing [Ca2r] or increasing [HP042-] there was an increasing amount of a granular precipitste,
A
which settled
on
the cells,
and of turbidity
in the medium.
PP1 produced
either no stimulation or only marginal stimulation
when added
to
quiescent chicken
centrations that were
embryo
highly stimulatory
fibroblasts
in con-
to 3T3 cells. Supra-
normal [Ca2+] produced only marginal stimulation chicken embryo fibroblasts.
of the
DISCUSSION progress of chicken embryo fibroblasts and mammalian lymphocytes through the cell cycle (7-9). In 3T3 cells, many of these materials such as proteases (10), insulin (11), and toxic metals are either ineffective or effective only under highly restrictive conditions. PP1 differs from the aforementioned effectors in that it stimulates 3T3 cells but not chicken embryo fibroblasts. In its selectivity for 3T3 cells, PP1 stimulation resembles that reported for supranormal concentrations of Ca2+ (4, 12) and confirmed above. Indeed, the stimulation by PP1 requires the presence of Ca2+, although subnormal concentrations of Ca2+ are fully adequate to produce the effect if 0.5 mM PP1 is present. The most striking features of the stimulation are: (i) its sharp dependence on the concentration of PP1, and (ii) its association with the appearance of a flocculent precipitate in the medium. All treatments that affect the concentration at which PP1 forms a precipitate with Ca2+ also affect in an identical manner the concentration that stimulates the cells. Thus, HP042- is required to form the precipitate and to stimulate the cells, and Mg2+ interferes with both effects. From the sharp delineation between concentrations of PP1 that do or do not cause both precipitation and stimulation, we conclude that the stimulation is caused by water-insoluble complexes of PP1, Ca2+, and HP042. This raises the question
A wide variety of unrelated materials accelerate
Ca2+,
mM
FIG. 7. Dependence of supranormal Ca2+ stimulation on [HP042-] in the medium. The procedure was the same as that of Fig. 6 except that a range of [Ca2+1 was used in the presence of each of the three concentrations of HP042- indicated on the curves for [3H] thymidine incorporation in the Upper panel. An aliquot of each medium combination was incubated overnight at 370 without cells to allow complexes of Ca2+ and HP042- to form, and the turbidities were measured as absorbance at 310 nm in a double-beam spectrophotometer, with the sample with the lowest [Ca2+] and [HP042-] used as reference (Lower). * indicates a heavy precipitate that settled on the cells and caused some cell death and detachment. -
of whether the stimulation is caused by providing any one or several of these constituents to act individually as internal substrates or effectors of metabolism, or whether the complex is acting in a manner similar to macromolecular stimulants in serum that are presumed to act at the cell surface. The concentration of Ca2+ in the medium can be decreased to far below the conventional levels without markedly affecting the rate of progress of 3T3 cells through the cell cycle (ref 13; Fig. 7), indicating that the external supply of Ca2+ is not a limiting factor for growth unless extreme deprivation is imposed upon the cells. It, therefore, seems unlikely that merely providing more Ca2+ Table 1. Effects of Sr2+ and Mg2+ in mimicking or inhibiting the stimulation of 3T3 cells by supranormal Ca2+ concentration Relative rate of Relative amt. [3H]thymidine Ca2+, Sr2+, Mg2+, of protein/dish mM mM incorp. mM 1.7 4.0 8.0 1.7 4.0 1.7 8.0
0 0 0 6.3 4.0 0 0
0.8 0.8 0.8 0.8 0.8 9.8 9.8
1.0 1.2 12.30 5.58 9.43 0.81 5.09
1.0 1.1 2.96
2.19 2.94 1.04 2.15
DME medium was prepared with 1.0 mM HP042- and the concentrations of Ca2+, Sr2+, and Mg2+ shown; and 10% calf serum was added. Confluent cultures of 3T3 cells were incubated in this medium at 370 for 41 hr and then labeled with [3H]thymidine and processed as in Fig. 1.
5030
Cell Biology: Rubin and Sanui
in complexes with PPi would act as a stimulant. This view is reinforced by the finding that Sr2+ can substitute for about 75% of the Ca2+ in the PPi stimulation and that the Ca2+ ionophore A23187 is totally ineffective. That PPi itself would be a limiting factor seems even less likely. It is not needed for the growth of cells and is not present in conventional medium. Because PP1 is a product of many reactions associated with nucleic acid and protein synthesis, the presence of high concentrations of PPi in the cells would tend to inhibit these reactions which are essential for cell cycle traverse. HP042- is also an unlikely candidate as a direct control because neither its rate of uptake nor its pool size is limiting in the growth of 3T3 cells (14). Substances that form soluble complexes with Ca2+, such as EGTA, EDTA, ADP, and ATP, have no stimulatory effect. Sodium tripolyphosphate causes a slight increase in turbidity within a narrow concentration range in Ca2+-containing medium (data not shown) but causes no gross precipitate or stimulation of the 3T3 cells. We think it likely, therefore, that the stimulation by PP1 is actually brought about by complexes of PPi, Ca2+, and HP042- that localize in the cell membrane, possibly because of the hydrophobicity, and stimulate the cell by altering the configuation of the membrane. The complexes of PPi, Ca2+, and HP042 appear to have some specificity because they are ineffective in stimulating chicken embryo fibroblasts. Also, when the concentration of PPi is increased to 1.0 mM, it loses much of its stimulatory power, even though large amounts of the flocculent precipitate form. This is not due to its complexing of Mg2+, because this concentration of PPi has no effect on stimulation of the 3T3 cells by high concentrations of serum. It is more likely that the ratio of the constituents of the complex or its size or both are important elements in determining interaction with cells. The major outlines of stimulation by PPi are similar to those of stimulation by supranormal concentrations of Ca2+. Both require HP042- and are accompanied by formation of insoluble complexes. Ca2+ can be replaced by Sr2+ in both cases and Mg2+ is inhibitory, although more dramatically so in PPi stimulation. Both treatments are effective in 3T3 cells and not in chicken embryo fibroblasts. Of course, the minimal amount of PPi required for a given level of stimulation is almost two orders of magnitude less than the amount of Ca2+ required for equivalent stimulation in the absence of PPi; and the minimal Ca2+ requirement for PPi stimulation is one order of magnitude less than the requirement of Ca2+ for equivalent stimulation in the absence of PPi (data not shown). These apparent differences in Ca2+ requirement for the two types of stimulation are balanced by the fact that stimulation can occur in conventional concentrations of Ca2+ (in the absence of PPi) by simply increasing [HP042-]. The common thread in all these forms of stimulation appears to be the formation of water-insoluble complexes of phosphate compounds and Ca2+ (or Sr2+) which,
Proc. Natl. Acad. Sci. USA 74 (1977)
we presume, associate with the surface membranes of the cell and act there to stimulate metabolism and growth. These complexes have been shown to be effective only in BALB/c 3T3 cells and to have little or no effect in secondary hamster (4) and chicken embryo cells and BHK cells (4). This raises once again the question of the true nature of the 3T3 cells. Although they arose from mouse embryo cultures that were fibroblastic in appearance (15), and they have been referred to repeatedly as fibroblasts (4), recent evidence indicates they are more closely related to cells involved in blood vessel formation (16). Their sensitivity to stimulation by complexes of phosphates and Ca2+ may reflect the special vasoformative nature of these cells. Finally, it should be kept in mind that PPi, Ca2+, and HP042- are normal cellular constitutents. Although Ca2+ normally occurs in cells at concentrations far below those required for precipitation with the phosphate compounds, under certain conditions, massive amounts of Ca2+ and phosphate can accumulate (17). Any defect that increases intracellular [Ca2+], such as an inward Ca2+ leak, could lead to the formation of growth-stimulatory complexes with inorganic phosphates and, if sustained, to tumor formation. We thank Berbie M. Chu for her fine technical assistance. This investigation was supported by National Institutes of Health Research Grant 15744 from the National Cancer Institute. 1. Rubin, H. (1975) Proc. Natl. Acad. Sci. USA 72,3551-3555. 2. Rubin, H. (1976) J. Cell. Physiol. 89,613-626. 3. Neuman, W. & Neuman, M. (1958) The Chemical Dynamics of Bone Mineral (University of Chicago Press, Chicago, IL). 4. Dulbecco, R. & Elkington, J. (1975) Proc. Natl. Acad. Sci. USA
72,1584-1588. 5. Vogt, M. & Dulbecco, R. (1963) Proc. Natl. Acad. Sci. USA 49, 171-179. 6. Rubin, H. (1972) Proc. Natl. Acad. Sci. USA 69,712-716. 7. Rubin, H. & Koide, T. (1975) J. Cell. Physiol. 86,47-58. 8. Novogrodsky, A. & Katchalski, E. (1971) Biochim. Biophys. Acta 228,759-783. 9. Novogrodsky, A. & Katchalski, E. (1971) FEBS Lett. 12, 297300. 10. Glynn, R., Thrash, C. & Cunningham, D. (1973) Proc. Natl. Acad. Sci. USA 70,2676-2677. 11. Jimenez de Asua, L., O'Farrell, M., Bennett, D., Clingan, D. & Rudland, P. (1977) Nature 265,151-153. 12. Boynton, A. & Whitfield, J. (1976) In Vitro 12,479-484. 13. Boynton, A., Whitfield, J., Isaacs, R. & Morton, H. (1974) In Vitro 10,12-17. 14. Barsh, G., Greenberg, D. & Cunningham, D. (1977) J. Cell. Physlol. 92, 115-128. 15. Todaro, G. & Green, H. (1963) J. Cell Biol. 17,299-313. 16. Boone, C. (1975) Science 188,68-70. 17. Lehninger, A., Carafoli, E. & Rossi, E. (1967) Adv. Enzymol. 29, 259-320.
