Interactions between inositol phosphates and cytosolic free calcium following bradykinin stimulation in cultured human skin fibroblasts

Biochimica et Biophysica Acta, 1091 (1991) 409-416

409

© 1991 Elsevier Science Publishers B.V. 0167-4889/91/$03.50 ADONIS 0167488991001117 BBAMCR 12857

Interactions between inositol phosphates and cytosolic free calcium following bradyldnin stimulation in cultured human skin fibroblasts H s u e h - M e e i H u a n g , L o u r d e s T o r a l - B a r z a a n d G a r y E. G i b s o n Cornell University Medical College, Burke Rehabilitation Center, White Plains, NY (U.S.A.)

(Received 7 May 1990) (Revised manuscript received 5 September 1990)

Key words: Bradykinin; Br-A23187; Inositol phosphate; Cytosolic free calcium; Fibroblast

The inositol triphosphate (IP3) that results from hydrolysis of phosphatidylinositol 4,5-bisphosphate (PlP2) is generally accepted to be responsible for the mobilization of intracellular calcium. However, some studies suggest that low concentrations of agonists elevate cytosolic free calcium concentration ([Caz+ ]i) without IP3 formation. Thus, in the present studies, a comparison of the temporal response of inositol phosphates OP3, IPz and liD) and [CaZ+| i to a wide range of bradykinin concentrations was used to examine the relation of these two signal transduction events in cultured human skin fibroblasts (GM3652). In addition, the effects of alterations in internal or external calcium on the response of these second messengers to bradykinin were determined. Bradykinin stimulated accumulation of inositol phosphates and a rise of [Caz+ li in a time- and dose-dependent manner. Decreasing the bradykinin concentration from 1/LM to 0.1 /zM increased the time until the lP3 peak, and when the bradyHnin concentration was reduceed to 0.01/zM IP3 was not detected. ICaZ+ii was examined under parallel conditions. As the bradykinin concentration was reduced frod[ 1/~M to 0.01/zM, the time to reach the peak of [CaZ+li increased progressively, but the magnitude of the peak was unaltered. These two second messengers were variably dependent on external calcium. Although the bradykinin-stimulated initial spike of [Caz+ l i did not depend on extracdlular calcium, the subsequent sustained levels of ICaZ+l~ were abolished in calcium free medium. The bradykinin-stimulated inositol phosphate formation was not dependent on the extracellular calcium nor on the elevation of [CaZ+li that was produced with Br-A23187. These results demonstrate that bradykinininduced lP3 formation can be independent of [Ca2+l~ and of external calcium, whereas changes in ICaZ+li are partially dependent on external calcium. Introduction

The hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP2) and an increase of cytosolic free calcium, ([Ca2+]i) are among the earliest responses of cells to

Abbreviations: ANOVA, analysis of variance; [Ca2+ ]i, concentration of cytosolic free calcium; DMEM, Dulbecco's modified Eagle's medium; EGTA, ethyleneglycol-bis(fl-aminoethylether) N,N,N',N'tetraacetic acid; fura-2/AM, fura-2 acetoxymethyl ester; Hepes, N2-hydroxyethylpiperazine-N'-2 ethanesulfonic acid; IP3, inositol trisphosphate; IP2, inositol bisphosphate; IP, inositol monophosphate; PCA, percholoric acid; PIP2, phosphatidylinositol 4,5-bisphosphate; PiP, phosphatidylinositol 4-phosphate; Pl, phosphatidylinositol; SNK, Student-Newman-Keuls. Correspondence: G.E. Gibson, Comell University Medical College, Burke Rehabilitation Center, 785 Mamaroneck Average, White Plains, NY 10605, U.S.A.

many hormones, neurotransmitters or growth factors. The receptor-mediated hydrolysis of PIP2 leads to the formation of IP3 and diacylglycerol. Diacylglycerol activates protein kinase C [1], whereas IP3 is generally regarded to trigger the mobilization of calcium from internal stores [2-8]. However, several investigators [912] report that low concentrations of agonists elevate [Ca2+]i witohut detectably altering IP3. In some systems external calcium is required for receptor-stimulated breakdown of PIP2 or IP3 formation [13-17], which suggests that a rise of [Ca2+]i due to an increased influx and not IP3 might be the initial event of the receptormediated transduction in these cells. Although increased inositol phosphates formation [18-21], Ca 2+ influx [22] and mobilization of intracellular calcium [23] have been observed in response to mitogens (EGF, insulin, bradykinin and vasopressin) in fibroblasts, neither the temporal nor the dose responses of IP3 formation and elevation of [Ca2+]i after

410

bradykinin stimulation have been compared under analogous conditions. Bradykinin-induced IP3 formation is independent of external calcium [19]; whether this is secondary to the elevation of [Ca2+]i is not clear. Thus, the present studies were designed to characterize the relationship between the phosphatidylinositol cascade and [Ca2+]i after bradykinin stimulation by examining the temporal response, by varying bradykinin concentrations and by manipulating external and internal calcium concentrations. These two signal transduction events have not been rigorously compared in fibroblasts in the same laboratory, so that the relation of the temporal response of these two receptor mediated second messengers is not well established. We present a refined analysis of inositol polyphosphates and [Ca2+]i after bradykinin stimulation which will advance our knowledge concerning pattern of changes in these cells.

Materials and Methods Materials Chemical and isotopes were from the indicated companies: Dulbecco's modified Eagle's medium (DMEM) and fetal bovine serum (GIBCO; Grand Island, NY); D-myo-[2-3H]Inositol (20 Ci/mmol), [2-3H]IP3, [2-3H] IP2 (Amersham International; Arlington, IL); bradykinin, ammonium formate, disodium tetraborate, sodium formate, phosphoinositides, Hepes, EGTA and LiCI (Sigma, St Louis, MO); anion-exchange resin AG1-X8 (formate, 200-400 mesh; Bio-Rad, Richmond, CA); liquid scintillation fluid Aquasol-2 (New England Nuclear, Boston, MA); Fura-2, Fura-2/AM and 4-BrA23187 (Molecular Probes; Eugene, OR); Merck silicagel 60 high-performance thin layer chromatography (HPTLC) plates, modular incubator (VWR Scientific, Piscataway, N J). Cells Human skin fibroblasts (cell line GM 3652 from an apparently normal 24-year-old male) were obtained from the Human Genetic Mutant Cell Repository at the Coriell Institute for Medical Research in Camden, NJ. Cell culture The cells were grown to confluence as a monolayer in DMEM supplemented with 16.6~ heat-inactivated fetal bovine serum and 1~ glutamine in T75 flasks. The cells were utilized between passages 9 and 15. 7 days before the experiments the cells were subcultured in 35 mm plastic petri dishes at a seeding density of 1.104 cells/cm2. Cells were grown until confluent (3 days) in 2 ml of DMEM with 20~; fetal bovine serum at 37 ° C in a humidified modular incubation that was aerated with 5~$ CO2 in 95~ air.

Experiments on inositol phosphates On the 3rd day after subculturing, the medium was replaced with serum free DMEM. 1 day later the fluid was replaced with 2 ml of DMEM with [2-3H]inositol (5 /~Ci/ml). After 3 days with the [2-3H]inositol, the medium was removed and fresh DMEM with 25 mM Hepes and 10 mM LiCI (pH 7.4) was added. The cells were incubated for an additional h at 37 °C in a modular incubator. The studies on the temporal and dose responses were initiated by removal of the medium and addition of fresh Hepes-LiCI-DMEM with appropriate treatments. Incubations were terminated by aspirating the assay medium and then adding 2 ml 0.2 M ice-cold perchloric acid (PCA). After 1 h on ice the acidified tissues were scraped with a rubber policeman and the contents of the dish were transferred to a 15 ml capped centrifuge tube. The dish was rinsed twice with 1 ml of PCA and the washes were combined with the initial 2 ml extract. After centrifugation at 1500 × g for 10 rain, the PCA supernatant was removed and neutralized with 1.5 M KOH/75 mM Hepes in the presence of a pH indicator as described elsewhere [24-26]. The same amount of the neutralization media was added to the samples to give a pH of between 7.2-7.6. Separation of inositol phosphates lnositol phosphates were separated by the procedure of Berridge et al. [27] with the modifications that were described previously [28]. In brief, the water-soluble products were eluted as follows: 15 ml of H20 to remove inositol, 15 ml of 5 mM disodium tetraborate in 60 mM sodium formate to elute glycerol phosphoinositol, 15 ml of 0.2 M ammonium formate in 0.1 M formate to elute IP, three 5 ml rinses of 0.4 M ammonium formate in 0.1 M formate to elute IP2 and three 5 ml rinses of 1 M ammonium formate in 0.1 M formate to elute IP3. The procedures for extraction and neutralization have been validated by the use of the individual radiolabelled inositol polyphosphate standards. The addition of the neutralization step to our previously published methods did not alter the recovery. Peaks of [2-3H]IP3 and [2-3H]IP2 were verified by comparison with radiolabelled standards. A 1 ml aliquot was taken from each fraction and mixed with 10 ml of Aquasol-2 for measurement of radioactivity with a Beckman LS 5801 liquid scintillation counter. The lipids were extracted from the pellets with a modified method [29] of Eichberg and Hauser [30]. Briefly, 2 ml of chloroform/methanol (2 : 1) and 0.5 ml of water were added and vigorously mixed for 20 s. After the lower organic phase was removed, the aqueous phase was extracted further with 1 ml of chloroform/ methanol/12 M HCI (4:1:0.0125, v/v). The neutral and acidic lipid extracts were combined and evaporated to dryness under N 2 and applied to silica gel HPTLC

411 plates together with authentic phosphoinositide standards. The plates were developed with chloroform/ methanol/15 M ammonium hydroxide/water (18 : 14 : 1:3, v/v). Lipids were visualized with 12 vapor. Each band that corresponded to phosphoinositide standards was removed and radioactivity was determined.

1000.0

30.0 25.0 20.0

500.0

15.0

!

lO.O o

Cytosolic free calcium measurement Fibroblasts were cultured exactly as for the experiments on inositol phosphates, except that the petri dishes in which the cells were seeded contained glass cover slips (25 mm) and [2-3H]inositol was not included in the medium. On the day of the calcium measurements, the medium was replaced with fresh serum free DMEM containing 1% bovine serum albumin, 25 mM Hepes (pH 7.4, 37°C) and 2 /~M fura-2AM and the cells were then incubated at 37 ° C for 1 h. The medium was then aspirated and the cells were rinsed three times with fresh incubation medium to remove any unincorporated fura-2/AM. The coverslip was placed in the chamber of a temperature-controlled microincubator (Medical System, Greenvale, NY) which was positioned on an inverted Olympus fluorescence microscope (model IMTe). Cytosolic free calcium was monitored by alternating the excitation wavelength betwee 355 and 378 nm 20-times for 1 s with a delta scan (Photon Technology International, Princeton, N J). The fluorescence of the cells was monitored with a photomultiplier tube after passing through a 40 × Nikon ultraviolet objective and 40 nm band pass filter with a peak at 509 rim. The ratio of the emitted fluorescence signal after excitation at 355 and 378 nm provided a measure of [cae+]i by the method of Grynkiewicz [31]. Measurements were done on groups of cells (at least five) that were within a 4 × 4 pm grid. Resting values were monitored for 60 s. The medium was then replaced with the treatment medium. Calcium free DMEM was made by adding 2.5 mM EGTA to the DMEM. The resulting calcium concentration was less than 10/~M as measured with a calciumselective electrodes (WPI, New Haven, CT).

Statistics Statistical significance (P < 0.05) was determined by analysis of variance (ANOVA) followed by StudentNewman-Keul's test (SNK) (based on SPSS/PC software program; SPSS, Chicago, IL). Results

Standardization of incubation conditions Cultured fibroblasts were incubated with [2-3H]in ositol and the distribution of radioactivity into various phosphoinositides (PIPe, PIP and PI) was examined at 24, 48 and 72 h. The labelling of PIPe, PIP and PI

5.O 0.0

o .o

PZP~

PIP

PI

Fig. 1. Time-course of incorporation of [2-3Hlinositol into phosphoinositides. Fibroblasts were incubated with [2-3H]inositol in DMEM for 24, 48 or 72 h. Phosphoinositides were separated by HPTLC as described in Materials and Methods. Results are expressed as the mean+S.E, of five to seven observations at each time point. ~,, denotes significant difference from the 24 h values; "[',denotes significant difference from the 48 h values by ANOVA followed by the SNK test (PIPe: F(2, 17) = 13.3, P < 0.05; PiP: F(2, 17) ffi21, P < 0.05; Pl: (2, 17) = 30, P < 0.05).1:3,24 h; ~, 48 h; and@, 72 h.

increased with the time of incubation up to 72 h (Fig. 1). Incorporation of [2-3H]inositol into PIP and PIP2 increased significantly at 72 h as compared to 24 and 48 h. Incorporation into PI was significantly elevated at 48 h as compared to 24 h and increased significantly more by 72 h. About 70-80% of the radioactivity in the aqueous phase remained as free inositol after 72 h and approx. 20-30% of the radioactivity was incorporated into phosphoinositides. About 95% of the radioactivity in the phosphoinositides was in PI, while only 2% was in PIP and PIP e. All of the subsequent experiments were done by prelabelling the fibroblast phosphatidylinositol with [2-3H]inositol for 72 h, because the incorporation was significantly increased by this time in all fractions that were monitored.

Time-course of bradykinin stimulated release of inositol phosphates The time-course of the accumulation of inositol phosphates in response to 1 ltM bradykinin was examined (Fig. 2). Bradykinin induced a rapid and transient increase of [2-aH]IP3 formation. By 5 s, the first time point, IP3 increased to 190% above control (3431 + 171 vs. 1802 + 108 dpm/dish), then declined rapidly, but sust',dned a significantly elevated level (135% of the basal) for 30-60 s and finally returned to the basal level by 5 min. Bradykinin-stimulated generation of [2-3H]IP2 was also rapid, but appeared to peak later than IP3. [2-3H]IP2 was significantly elevated by 5 s, continued to increase to 442% above the control level by 10 s (6234 + 633 vs 1412 + 88 dpm/dish), stabilized between 10 s and 1 min and then declined to only 37% above the basal within 5 min. [2-3H]IP was significantly above the basal by 15 s and remained 170-200% above the basal through the whole 5 min period.

412 5.0

4.0 "

IPa

IP3

15.0

IPa

4.0

,t ..i

11i

3.0

*

10.0

_,

3.0

w ,°'°'~°"°

/.'°

/" •

5.0!

'-'

0.0

I

-10

II

i

-9

Log [gK] M

~

,

,

.

.

,

.

.

.

.

.

.

.

.

.

.

,

. . . .

!

. . . . .

,

9.0 8.0 7.O

t

." I t / ~

1.0 0.0 _ioil '

,

!

,

I

i

-8 -7 -6 Log [BK] M

I

-5

-4

Fig. 3. Formation of [2-3H]lP3, [2-3H]IP2 and [2-3H]IP after stimulation by various concentrations of bradykinin. Fibroblasts were prelabeled with [2-3H]inositol as described in Materials and Methods and incubated with the indicated concentrations of bradykinin for 5 or 15 s. Values are means+S.E, of six observations, t, denotes a significant effect of bradykinin at 5 s, w h i l e , denotes a significant effect of bradykinin at 15 s as determined by ANOVA followed by the SNK test (IP3: F(7, 47) = 47, P < 0.05; IP2: F(7, 47) -- 88, P < 0.05; IP: F(7, 47) = 17, P < 0.05). A, 5 s and ®, 15 s.

x 3,0 2,0

suggests that increasing bradykinin concentrations from 0.1 pM to 1 pM shifted the time of the IP3 peak from 15 s to 5 s. Bradykinin-induced [2-3H]IP3 formation was not detectable at concentrations of bradykinin below 0.01 IgM at either time point. Bradykinin-induced [23H]IP2 formation was also dependent on the concentrations of bradykinin. At 5 s maximal stimulation was between 1 and 10 pM and half-maximal stimulation was at 0.1 /~M. [2-3H]IP2 formation at 15 s was ap-

l.O 40.0

30.0

~

20.0

X

~= lo.o 0,0

'''' 0

...... 60

' ..... 120 Time

' ............ 180 240

2.0

300

,

(s)

Fig. 2. Time-course of the accumulation of [2-3H]inositol phosphates in fibroblasts after stimulation with bradykinin. Cells were prelabeled with [2.~H]inositol as described in Materials and Methods. On the day of the measurements cells were incubated with (@) or without ( o ) bradykinin (! ~tM) in the Hepes-LiCI-DMEM for 0, 5, 10, 15, 30, 60 or 300 s. [2-3H]lP3, [2-3H]IPI and [2-3H]IP were monitored. Values are means:t: [i.E, of five observations each from two separate experiments, • , denotes a significant effect of bradykinin compared to the non-stimulated value at the same time point as determined by ANOVA followed by SNK test (IP3: F(12,134) = 20, P < 0.05; IP2: F(12,140)

l.O0 uN

1.5 Patio

of Channel I/2

t0

= 28, P < 0,05).

Effects of various bradykinin concentration on the formation of inositol phosphates and [Ca z +]~ The effects of bradykinin on the formation of inositol phosphates were examined at concentrations that ranged between 0.1 nM and 10/~M after 5 s (peak of IP3) and 15 s (peak of IP2) of stimulation (Fig. 3). Although measurements at discrete points do not permit precise definition of the peak, the results suggested that lowering the concentrations of bradykinin delayed the appearance of the peak. Bradykinin at I pM stimulated maximal [2-3H]IP3 formation, whereas maximal values with 0.1 /LM did not occur until 15 s, which

0.5

4()

t

|

60

eb

,

'

Time (seconds) Fig. 4. The effects of various bradykinin concentrations on [Ca 2+ ]i. The cells were preloaded with fura-2/AM as described in Materials and Methods. Channel 1 is the fluorescence after excitation at 355 and channel 2 is the fluorescence after excitation at 378 nm. The vertical axis is the ratio of fluorescence after excitation at 355 to that at 378 nm and estimates [Ca 2+ ]i. The arrow indicates when bradykinm was added. The values are means of two experiments (n ffi 6). The delta scan allows us to average every point of the six runs. Peak times vary significantly from each other as determined by ANOVA followed by the SNK test (F(2, 15) = 64.5, P < 0.05).

413 parent with concentrations as low as 0.01/tM and did not appear to have peaked at 10/~M. The effects of bradykinin concentrations that ranged between 0.01 and 1/LM o n [Ca2+]i were examined (Fig. 4). Bradykinin (1/~M) instantaneously increased [Ca 2+ ]i from a basal of 213 4- 13 nM to a peak of 999 4- 86 nM, which was followed by a gradual decline over several rain. In a manner similar to the IP3 response, lowering the concentrations of bradykinin lengthened the time to peak but not the magnitude of peak. The time to reach the peak of [Ca2+]i lengthened significantly at lower bradykinin concentrations. The time f o r [ C a 2 + ] i t o peak in response to bradykinin at concentrations of 1, 0.1 and 0.01/~M were 0.8 4- 0.11, 4 + 0.53 and 15 4- 1.18 s, respectively.

Effects of external calcium on the formation of inositol phosphates and [Ca" +]i To evaluate the role of external Ca 2+ in the formation of the IP3, a series of experiments were performed to determine whether or not IP3 formation and tt:e rise of [Ca 2+]i could be dissociated. Fibroblasts were stimulated with bradykinin (1/~M) with three different treatment paradigms in which external calcium was manipulated and the effects on [Ca2+]i and the phosphatidylinositol cascade were examined. In the 'control group', the cells were stimulated with bradykinin in normal calcium medium (1.3 mM); in the 'Ca2+-free

F~~

Control Ca2+free

1.4 Patio

Caz+free-pretreated

Channel

t12 0.9

aK

0'40

.

I

11

i

16o

*

lg0

i

,

2s0

Time (seconcls) Fig. 5. The effects of external calcium on bradykinin-induced changes in [Ca2+ ]i. Fibroblasts were preloaded with fura-2/AM as described in Materials and Methods. Three different treatment paradigms were utilized: (Ca 2+ free-pretreated group), the cells were pre-incubated with calcium free DMEM for 10 rain and were then treated with calcium free DMEM with 1 ~M bradykinin; (Ca 2+ free group), the cells were treated with calcium-free DMEM with bradykinin; (control), the cells were treated with normal calcium, DMEM with hradykjnin, Channel I is the fluorescence after excitation at 355 and channel 2 is the fluorescence after exci:ation at 378 nm. The vertical axis is the ratio of fluorescence after excitation at 355 to that at 378 nm and estimates [Ca 2+ ]i- The values are means of two experiments in triplicate (n -- 6). The delta scan allows us to average every point of the six runs.

5.0 IP,

4.0 "g3

3.0

? o

2.0

x

c~ 1.0

WA

0.0 Sisal

8X

Prelnctmat san . i t f l r~;~*

Basal

Premcu~ation IIltfl 0 Ca2.

Sisal

[

~elnc~ation with Ca;~*

Basal

8g

PreiN:Ut~lLion uRfl 0 Ca a°

Fig. 6. Effects of external calcium on bradykinin-stimulated formatmn of inositol phosphates. Fibroblasts were prelabeled with [2-3H]inositol for 3 days as described in Materials and Methods. On the day of the measurements, the media were replaced with fresh Hepes-LiCI-DMEM (pH 7.4) for I h. The media were then changed and the cells were preincubated with or without calcium for 10 min. The media were then replaced with media with or without bradykinin (1/~M) in the presence or absence of calcium for 5 s. The values are means :t: S.E. of six observations from two separate experiments (n = 12). ~', denotes a significant bradykinin effect as determined by ANOVA followed by the SNK test (IP3: F(7,95)--15, P < 0 . 0 5 ; IP2: F(7,95)=16, P < 0.05). El, absence of and I~, presence of calcium during stimulation.

group', the cells were stimulated with bradykinin in calcium-free DMEM; in the 'Ca 2+ free-pretreated group', the cells were pretreated with calcium free DMEM for 10 min, then stimulated with bradykinin in the calcium free medium. When [Ca2+]i was measured under these conditioBs, the two components of the peak responded differently. In the 'Ca2+-free ' condition, the initial [Ca2+]i spike in response to bradykinin was maintained, but the sustained elevated level was abolished (Fig. 5). When the cells were pretreated with calcium free DMEM for 10 rain (Ca2+-free-pretreated group), the initial spike of cytosolic free calcium in response to bradykinin was reduceed from 836 4. 124 nM to 433 4- 40 nM and the sustained elevated level was abolished. The phosphatidylinositol cascade was evaluated under similar conditions (Fig. 6). The cells were preincubated with calcium or calcium free DMEM for 10 min, and then stimulated with bradykinin (1 ~M) for 5 s in the presence or absence of calcium. The basal levels of labelled IP3 and IP2 were not significantly altered by the presence of calcium. The levels of labelled IP3 and IP2 after bradykinin stimulation were similar whether or not calcium was included in the preincubation medium or during stimulation. Thus, neither the bradykinin-induced formation of IP3 nor that of IF2 depends on the extracellular Ca 2+ in the media under either'Ca2+-free' o r ' C a 2+-free-pretreated' conditions.

Effects of ionophore Br-A23187 on [Ca2+]~ and its relation to bradykinin-induced formation of inositol phosphates Cytosolic free calcium can be elevated from intracellular calcium stores as well as from external calcium.

414

The experiments that are presented in Figs. 7 and 8 are to test whether bradykinin-stimulated IP3 formation can occur under normal resting [Ca 2+ ]i, but in the absence of an elevation of [Ca2+]t. To determine whether the phosphatidylinositol cascade could be stimulated at resting [Ca2+]t, the ability of bradykinin to increase calcium was impaired by prior treatment with the ionophore Br-A23187. Cells were treated with Br-A23187 in calcium-free DMEM and then stimulated with bradykinin (1 /~M) in calcium-free DMEM. In these experiments, [Ca2+]t increased rapidly to a peak of 1626 4-338 nM after ionophore treatment and then returned to the basal level of 174.3 4-12.2 nM after 4 rain. A second addition of ionophore after 4 min caused only a small, nonsignificant increase of [Ca2+]t (data not shown) and no further increase occurred upon the addition of bradykinin (Fig. 7). The effects of ionophore addition on the formation of inositol phosphates were examined under conditions similar to those described in the calcium experiments. The cells were treated with Br-A23187 for 4 min in calcium-free DMEM, and then the effects of stimulat-

.1.0

0.9

nitro O, 8 ol Ogll~ll

|la

0,7

0,6 t~K

~-t23tB?

0,5

6

,I

t6o

aao

t

8!

aoo

,

.a(

Time (seconds) Fig, 7. Effects of ionophore Br-A23187 and bradykinin on [Ca 2+ ]i in human fibroblasts in Ca 2+ free-EGTA containing DMEM. On the day of the measurements, the media were replaced with fresh HepesDMEM (pH 7,4) containing 1~ bovine serum albumin. Fibroblasts were preloaded for I h with fura-2/AM as described in Materials and Methods. Fibroblasts were then treated with calcium free DMEM with Br-A23187 (1/AM) as indicated by the first arrow for 4 rain. The media was then replaced with calcium free medium without bradykinin (1/AM) at the time indicated by the second arrow. The resting level of [Ca2+ ]i was measured in the cells were treated with Ca 2+ free-EGTA containing DMEM without either Br-A23187 or bradykinin. Channel 1 is the fluorescence after excitation at 355 and channel 2 is the fluorescence after excitation at 378 nm. The vertical axis is the ratio of fluorescence after excitation at 355 to that of 378 nm and estimates [Ca2+ ]i- The values are means of two experiments in tripficate. The delta scan allows us to average every point of the six runs. The value of the peak varies significantly from the resting value, from the value 4 rain after addition of Br-A23187 and the value after addition of bradykinin ( P < 0.05) by ANOVA followed by SNK test.

6.0 ,t

5.O J~ "lm

~

T

4.0

&0

-r-

Z.O

T

-r

"T"

,ol

0.0

O.T.

C

AC

~

J~(

I~

Fig. 8. Effects of bradykinin and ionophore Br-A23187 on the formation of inositol phosphates in fibroblasts in Ca 2+ free DMEM. The cells were prelabeled with [2-3H]inositol for 3 days as described in Materials and Methods. On the day of the experiment, the media were replaced with fresh Hepes-LiCI-DMEM (pH 7.4) for 1 h. The experiments consisted of two incubation steps except for zero time (0.T.) which was stopped without incubation. The cells were incubated with calcium-free DMEM for 4 min and then replaced with another calcium free DMEM for 5 s (C); the cells were treated with Br-A23187 (1 /tM) in calcium free DMEM for 4 min and then incubated with another calcium-free DMEM without bradykinin for 5 s (AC); the cells were treated with Br-A23187 in calcium-free DMEM for 4 min and then incubated with Br-A23187 in calcium-free DMEM for 5 s (AA); the cells were treated with Br-A23187 in calcium-free DMEM for 4 ~a~inand then treated with bradykinin in calcium-free DMEM (ABK); the cells were stimulated with bradykinin in calcium free DMEM for 5 s without preincubation (bradykinin). , , denotes a significant bradykinin effect as determined by ANOVA followed by the SNK test, while t denotes a significant difference between group ABK and BK (IP3: F(6,70)=43.6; IP2: F(6,70)--118). O, IP3 and m, Ip2.

ing with either bradykinin or BroA23187 in calcium free DMEM for 5 s were determined (Fig. 8). Formation of inositol phosphates after bradykinin stimulation without Br-A23187 pretreatment is the reference for all other conditions tested (Fig. 8; BK). [2-3H]Inositol phosphates (IP3 and IP2) did not increase significantly above either 0 time (0.T) or control (C) conditions if cells were pretreated for 4 min with Br-A23187 and then either treated for 5 s with calcium-free DMEM (Fig. 8; AC) or calcium-free DMEM that included Br-A23187 (Fig. 8; AA). However, lP3 and lP2 formation by cells pretreated with calcium free medium plus Br-A23187 for 4 min still responded to bradykinin stimulation (Fig. 8; ABK), whereas prior Br-A23187 addition reduced the bradykinin-induced formation of inositol phosphates (Fig. 8; BK). Discussion

These ~tudies demonstrate parallel comparisons of the temporal responses of bradykinin-induced changes in inositol phosphates (particularly IP3) and Ca 2+ mobilization to a wide range of bradykinin concentrations in human skin fibroblasts. The present studies

415 provide a refinement and advance in our knowledge of the pattern of these two signal transduction systems in a single cell line under analogous conditions. Bradykinininduced formation of !p3 and the elevation of [cae*]i were transient and rapid. An exact comparison of the temporal response of IP3 to that for [Ca2+]i is difficult, because the calcium measurements were continuc,us an5 IP3 determinations were only at discrete time points. Furthermore, the calcium peak occurred by the earliest time that was practical to examine IP3 (5 s) in this study. In spite of these limitations, the use of multiple concentrations of agonist provide a practical approach to make comparisons between these two signal transduction mechanisms. The generation of inositol phosphates and elevation of [Ca2+]i were both dependent on the concentrations of bradykinin. The time to the IP3 peak, as well as that of [Ca2+]i, lengthened as the concentrations of bradykinin decreased. At 0.01 /tM bradykinin an increase of [Ca2+]i but not IP3 formation was detected. Furthermore, at 0.1 #M bradykinin, the increase of [Ca2+]i clearly preceded the formation of the IP3. Similarly, in hepatoeytes the calcium peak precedes the IP3 peak and low concentrations of agonists promote calcium entry without detectable changes in inositol phosphates [10]. These results suggest that at low bradykinin concentrations mechanisms other than IP3 are responsible for calcium mobilization. Low concentrations of bradykinin may depolarize the membrane potential and induced Ca 2+ influx or release Ca z+ from an IP3-insensitive calcium pool. However, if low concentrations of bradykinin caused only small increases in IP3 they would be very difficult to detect. By the current methods, an increase in IP3 as little as 10% (estimated from the S.D. of the experiments) could be detected and no elevation was obtained. Thus, at concentrations of bradykinin at which the magnitude of the calcium response was unaltered, the IP3 response was only 10% of its maximal response. If IP3 is iw,olved in calcium mobilization under these conditions, then IP3 at levels less than 10% of maximal response is capable of producing a full calcium response. Although labelled IP2 could serve as an index for IP3 formation at lower bradykinin concentrations, it may be derived from sources other than IP3. At 1 #M bradykinin concentrations, bradykinin induced rapid increases in IP3 and IP2 as early as 5 s. The bradykinin-induced rapid formation of IPz may be derived from not only dephosphorylation of IP3 but also from the hydrolysis of PIP directly. In several cell systems, agonist-stimulated phospholipase C directly hydrolyzes PIP as well as PiP2 [32,331. The increase of the [Ca2+]i in response to bradykinin may be from intracellular and extracellular sources. The experiments with varying calcium demonstrate that the initial spike of [Ca2+]i upon bradykinin stimulation is from intracellular stores, since it occurs in the absence

of external calcium. However, the subsequent sustained elevated level of [Ca2+]i is due to Ca 2+ influx, since it is abolished in the calcium free medium. A large literature indicates that the transient IP3 formation triggers the release of Ca 2+ from endoplasmic reticulum in a number of systems [2,5]. The finding in the present study that the change of the phosphatidylinositol cascade did not depend on the rise of [Ca2+]i further supports the role of this pathway in initiation of calcium mobilization. The mechanism of regulation of Ca 2+ entry that gives rise to the sustained level at the peak is poorly understood. Some of the inositol polyphosphates or G proteins may modulate the entry of calcium from the extracellular sources. Inositol 1,4,5-trisphosphate [34], or the combination of inositol 1,4,5-Msphosphate with inositol 1,3,4,5-tetrakisphosphate [35,36], has been proposed to play a role in regulation of Ca 2÷ entry. The present studies demonstrate that the phosphatidylinositol cascade can be dizsociated from changes in calcium homeostasis. Bradykinin-induced IP3 formation was not dependent upon extracellular calcium nor on the elevation of [Ca 2 + ]i- Bradykinin stimulated IP 3 formation similarly in both Ca 2+- and EGTA-containing medium, which suggests that the response does not depend on extracellular Ca 2+. Furthermore, when the cells were preincubated with EGTA containing medium for 10 min, the generation of IP3 following bradykinin stimulation was unaffected. Ca 2+ influx is abolished and internal calcium stores are depleted after incubation in calcium-free and subsequent treatment with ionophore Br-A23187 as in this study or in other cells [37-39]. Under such conditions, [cae+]i remained at resting level after bradykinin or Br-A23187 stimulation. However, IP3 formation was still markedly increased after bradykinin application. These studies demonstrate that receptor-mediated IP3 formation can occur at the resting level of [Cae+]t. The initial addition of Bro A23187 did not increase IP3 production, while it elevated [Ca2+]i, which indicates that in the absence of receptor activation a large increase of [Ca z+]i due to ionophore is insufficient to initiate the hydrolysis of phosphoinositides. A second addition of BroA23187 in Ca 2+ free DMEM altered neither [Ca2+]i nor IP3 formation. This approach allowed us to dissociate the bradykinin-induced accumulation of IP3 from changes in [cae+]i and to demonstrate that the transient formation of IP3 is not secondary to calcium entry or to Ca z+ mobilization. Although bradykinin stimulation still increased the phosphatidylinositol cascade without elevation of [Ca2+]i, the response was diminished. The reduction may be related to changes in calcium. Although an increase in calcium is not required for the activation of phospholipase C, elevated calcium can amplify the activation of phospholipase C and lead to enhanced IP3 formation [40]. Therefore, depletion of the internal calcium stores by ionophore could impair I ~ formation

416

by this mechanism. Another possiblity is that A23187 disrupts protein-fipid interactions [41] and alteras the environment of the bradykinin receptor.

Acknowledgements We thank Dr. Joseph Appleby for his excellent assistance and Miss Angela Smith for maintaining the cell cultures.

References 1 2 3 4

Nishizuka, Y. (1986) Science 233, 305-312. Berridge, M.J. (1987) Annu. Rev. Biochem. 56, 159-193. Berridge, M.J. and h-vine, R.F. (1989) Nature 341, 197-205. Burgess, G.M., Godfrey, P.P., McKinney, J.S., Berridge, M.J., h'vine, R,F. and Putney, J.W., Jr. (1984) Nature 309, 63-68. 5 Streb, H., h'vine, R.F., Berridge, M.J. and Schulz, !. (1983) Nature 306, 67-69, 6 Joseph, S~K., Thomas, A.P., Wilfiams, R.J., h'vine, R.F. and Willb~nson, J,R. (1984) J. Biol. Chem. 259, 3077-3081. 7 Prentki, M., Biden, T.J., Janjic, D., h'vine, R.F., Berridge, M.J. and Woilheim, C.B. (1984) Nature 309, 562-564. 8 Hirata, M., Suematsu, E., Hashimoto, T., Hamachi, T. and Koga, T, (1984) Biochem. J. 223, 229-236. 9 Ambler, S.K., Thompson, B., Solski, P.A., Brown, J.H. and Taylor, P. (1987) Mol. Pharmacol. 32, 376-383. 10 Extort, J.H. (1988) FASEB J. 2, 2670-2676. 11 Lynch, C.J., Blackmore, P.F., Charest, R. and Exton, J.H. (1985) Mol. Pharmacol. 28, 93-99. 12 Merritt, .I.E., Taylor, C.W., Rubin, R.P. and Putney, J.W. (1986) Biochem. J. 238, 825-829. 13 Akhtar, R.A. and Abdel-Latif, A.A. (1978) J. Pharmacol. Exp. Ther. 204, 655-688. 14 Cockeroft, S., Baldwin, J.M. and Allan, D. (1984) Biochem. J. 221, 477-482.

15 Cockroft, S., Bennett, J.P, and Gomperts, B.D. (1981) Biochem. J. 200, 501-508. 16 Fisher, S.K. and Agranoff, B.W. (1981) J. Neurochem. 37, 968-977. 17 Fisher, S.K., Hob., R.W. and Agranoff, B.W. (1981) J. Neurochem. 37, 491-497. 18 Vicentini, L.M. and Villereal, M.L. (1984) Biochem. Biophys. Res. Common. 123, 663-670.

19 Jamieson, G.A., Jr. and Villereal, M.L. (1987) Arch. Biochem. Biophys. 252, 478-486. 20 Fu, T., Okano, Y. and Nozawa, Y. (1988) Biochem. Biophys. Res. Commun. 157, 1429-1435. 21 Olsen, R., Santone, K., Melder, D., Oakes, S.G., Abraham, R. and Powis, G. (1988) J. Biol. Chem. 263, 18030-18035. 22 Mix, L.L., Dinerstein, R.J. and Villereal, M.L. (1984) Biochem. Biophys. Res. Commun. 119, 69-75. 23 Berridge, M.J., Heslop, J.P., Irvine, R.F. and Brown, K.D. (1984) Biochem. J. 222, 195-201. 24 Bone, E.A., Fretten, P., Palmer, S., Kirk, C.J. and Michell, R.H. (1984) Biochem. J. 221, 803-811. 25 Audigier, S.M.P., Wangs, J. and Greengard, P. (1988) Proc. Natl. Acad. Sci. USA 85, 2859-2863. 26 Shulda, S.D. (1985) Arch. Biochem. Biophys. 240, 674-681. 27 Berridge, M.J., Dawson, R.M.C., Downes, C.P., Heslop, J.P. and Irvine, R..F. (1983) Biochem. J. 212, 473-482. 28 Huang, H.-M. and Sun, G.Y. (1988) J. Neurochem. 50, 366-374. 29 Sun, G.Y., Huang, H.-M. and Chandrasekhar, R. (1988) Neurochem. Int. 13, 63-68. 30 Eichberg, J. and Hauser, G. (1967) Biochim. Biophys. Acta 146, 415-422. 31 Grynkiewicz, G., Poenie, M. and Tsien, R.Y. (1985) J. Biol. Chem. 260. 3440-3450. 32 Wilson, D.B., Neufeld ~.J. and Majeurus, P.W. (1986) J. Biol. Chem. 260, 1046-1051. 33 Claro, E., Wallace, M.A., Lee, H.-M. and Fain, J.N. (1989) J. Biol. Chem. 264, 18288-18295. 34 Kuno, M. and Gardner, P. (1987) Nature 326, 301-303. 35 h'vine, R.F. and Moore, R.M. (1986) Biochem. J. 240, 917-920. 36 Morris, A.P., Gallacher, D.V., Irvine, R.F. and Petersen, O.H. (1987) Nature 330, 653-655. 37 Vicentini, L.M. and Villereal, M.L. (1984) Biochem. Biophys. Res. Commun. 123, 663-670. 38 Lambert, T.L., Kent, R.S. and Whorton, A.R. (1986) J. Biol. Chem. 261, 15288-15293. 39 Yano, K., Higashida, H., Hattori, H. and Nozawa, Y. (1985) FEBS Lett. 181,403-406. 40 Eberhald, D.A. and Holz, R.W. (1988) Trends Nuerosci. 11, 517-520. 41 Klausner, R.D., Fishman, M.C. and Karnovsky, M.J. (197,°) Nature 281, 82-83.