Articles |
From the Medizinische Poliklinik (H.S., M.M., M. van der G., M.T., J.B., I.G., C.S., W.Z.), Institut für Medizinische Physik (E.N., M.K.), and Institut für Biochemie (H.W.), Universität Münster (Germany).
Correspondence to Prof W. Zidek, Medizinische UniversitätPoliklinik, Albert-Schweitzer-Str 33, 48129 Münster, Germany.
| Abstract |
|---|
|
|
|---|
Key Words: coenzyme A glutathione disulfide adrenal gland vasoactive agents
| Introduction |
|---|
|
|
|---|
Among the potential sites of production, the adrenal gland has been recognized as a source of several important vasoactive agents, such as neuropeptide Y, the catecholamines, and, recently, endogenous ouabain.4 Furthermore, the adrenal gland may produce other endogenous inhibitors of the Na+,K+-ATPase.5 Therefore, in the present study extracts from bovine adrenal glands were systematically assayed for vasoactivity, and in one vasoactive fraction the underlying substance was identified as coenzyme A glutathione disulfide (CoASSG).
| Materials and Methods |
|---|
|
|
|---|
Chromatography
In step 1, the delipidated methanol extract of adrenal glands
was dissolved in 5 mL eluent and 1 mol/L acetic acid and subjected to a
size-exclusion chromatography (flow, 1 mL/min; column dimension,
26x950 mm; Sephacryl 100 HR gel, Pharmacia). Then the combined active
fractions from size-exclusion chromatographies were dissolved in 5 mL
of 20 mmol/L triethylammonium acetate (TEAA) in water (eluent A) and
chromatographed (flow, 1.5 mL/min) with a C4 reversed-phase
column (22x250 mm, Protein-Plus, Zorbax) in the displacement mode
(displacer, 200 mmol/L n-butanol in eluent A)9
(step 2). In step 3, the active fraction was run (flow, 0.5 mL/min) on
a reversed-phase HPLC column (Lichrosorb RP-C18, 250x4 mm, Merck) with
the following gradient: 0 to 10 minutes, 100% eluent A; 10 to 50
minutes, 0% to 20% eluent B (acetonitrile). Thereafter, the active
fraction was rechromatographed (flow, 0.5 mL/min) with a reversed-phase
HPLC column (Superspher RP-C18 endcapped, 250x4 mm, Merck)
with the following gradient: 0 to 10 minutes, 100% to 96% eluent C
(0.1% trifluoroacetic acid in water); 10 to 60 minutes, 4% to 12%
eluent B (step 4). The final purification step (step 5) was a
chromatography (flow, 0.5 mL/min) on a reversed-phase HPLC column
(Lichrospher 60 RP-select B, Merck) with the following gradient: 0 to 4
minutes, 100% to 94% eluent A; 4 to 64 minutes, 4% to 10% eluent
B.
The extract obtained from the chromaffin granules isolated from bovine adrenal medulla was chromatographed with a reversed-phase column with the conditions given above for step 3. The supernatant of adrenal medulla slices incubated in the presence or absence of carbachol was subjected to precipitation with 0.6 mol/L perchloric acid. After neutralization with KOH, the supernatant was concentrated with a reversed-phase column (Lichroprep, Merck). The lyophilized eluate (40% acetonitrile) was chromatographed with an anion exchange column (MonoQ, Pharmacia; eluent A, 10 mmol/L K2HPO4 [pH 7]; eluent B, 50 mmol/L K2HPO4 [pH 7] plus 1 mol/L NaCl; gradient: 0 minutes, 0% eluent B; 10 minutes, 15% eluent B; 60 minutes, 40% eluent B; and 65 minutes, 100% eluent B). Then the fraction containing CoASSG, as determined by comparison of retention times with authentic CoASSG, was further chromatographed on a reversed-phase column (Superspher C18 endcapped, 4x250 mm; eluent A, 40 mmol/L TEAA; eluent B, 80% acetonitrile; gradient: 0 minutes, 0% eluent B; 5 minutes, 0% eluent B; 35 minutes, 30% eluent B; 45 minutes, 60% eluent B; and 48 minutes, 80% eluent B).
Analytical Procedures
UV spectroscopy of the vasopressor fraction in step 5 was
performed at pH 6.5 (water) and at pH 2 (0.01 mol/L HCl).
The vasopressor fraction in step 5 was further examined with matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS).10 Briefly, a reflector-type time-of-flight mass spectrometer, equipped with a nitrogen laser (337 nm; pulse length, 4 nanoseconds) was used for ion generation and mass analysis. Speed-Vacdried samples were dissolved in 10 µL water. Then 0.5 and 1.5 µL of the 3-hydroxy-picolinic acid matrix solution (0.25 mol/L) in H2O were mixed on a flat metallic sample support and dried in a stream of cold air. Desorption of analyte ions was performed by laser shots of irradiances in the range of 106 to 107 W/cm2 focused to spot sizes of typically 50 to 100 µm in diameter. The ions generated were accelerated to an energy of 12 keV. For detection, a conversion dynode in combination with a secondary electron multiplier (EMI 9643) was used. The spectra were registered by a LeCroy 9400 transient recorder and typically accumulated from 10 single laser shots.
Enzymatic Cleavage
The vasoactive fraction from the final purification step 5 was
incubated at 37°C with alkaline phosphatase (EC 3.1.3.1, Boehringer
Mannheim; 1 mU/mL; incubation time, 2 hours), phosphodiesterase
(5'-nucleotidase, EC 3.1.15.1, Boehringer Mannheim; purified according
to Sulkowski and Laskowski11 ; 5 mU/mL; incubation time, 30
minutes), and glutathione reductase (EC 1.6.4.2, type VII, Sigma; 10
mU/mL; incubation time, 30 minutes). The split products of alkaline
phosphatase and of phosphodiesterase underwent reversed-phase
chromatography (conditions as described above for step 5). The split
products were identified by their molecular masses with the use of
MALDI-MS and by their retention times (AMP, 30.9 minutes;
dephospho-CoASSG, 59.1 minutes). The split products of glutathione
reductase were chromatographed as described above for step 3.
Then, after derivatization with ortho-phthaldialdehyde, fractions
eluting from 4 to 15 minutes were chromatographed with a reversed-phase
column (0.5 mL/min; Nucleosil 120-5 C18,
Macherey-Nagel). Eluent D consisted of 0.2 mol/L potassium acetate (pH
6) in water, and eluent E consisted of methanol. A gradient with 0% to
21% eluent D in 0 to 10 minutes, 21% to 21% eluent D in 10 to 33
minutes, and 21% to 36% eluent D in 33 to 40 minutes was developed.
Fluorescence was monitored with 330-nm excitation and 445-nm
emission.
To test for the presence of disulfide bonds, 1 µL mercaptoethanol with 5 µL of the active fraction was boiled for 20 minutes.
To identify the retention time of dephospho-CoASSG, this compound was prepared by incubation of authentic CoASSG with alkaline phosphatase as described above. Dephospho-CoASSG was purified by using chromatography as described above for step 5.
The vasopressor compound was derivatized with ortho-phthaldialdehyde as follows: The dried sample was dissolved in 60 µL water, and 40 µL of the ortho-phthaldialdehyde reagent (500 µL ethanol, 20 mg ortho-phthaldialdehyde, 9 mL of 0.4 mol/L potassium borate buffer [pH 10.4], and 500 µL mercaptoethanol) was added.
Measurement of [Ca2+]i in Vascular
Smooth Muscle Cells
Measurements of [Ca2+]i were
performed with the calcium-sensitive dye fura 2, as described by
Grynkiewicz et al12 and Williams et al,13 by
using monolayers of vascular smooth muscle cells (VSMCs) grown on round
coverslips with a diameter of 13 mm according to the method of Capponi
et al14 and Okada et al.15 Briefly, VSMCs
were washed twice in PSS containing (mmol/L) NaCl 135, KCl 5,
CaCl2 1, MgCl 1, D-glucose 5.5, and HEPES 10,
buffered with NaOH to pH 7.4, and then incubated for 60 minutes at
37°C with 0.5 µmol/L fura 2-AM (Sigma). At the end of the loading
period, the coverslips were washed twice in PSS and inserted into
quartz glass cuvettes with 2 mL PSS. After loading of the cells with
fura 2, the experiments were continued only when a viability of >95%
was observed by trypan blue exclusion. The fluorescence intensity of
fura 2loaded VSMCs was measured at 37°C by using a
spectrofluorophotometer (RF-5001 PC, Shimadzu) equipped with a
thermostatically controlled cuvette holder and with software designed
for the measurement of intracellular calcium (Shimadzu). The complete
intracellular hydrolysis of fura 2-AM to fura 2 was judged by changes
in the excitation and emission spectra. The fluorescence of fura 2 was
measured by using a data-sampling interval of 0.5 seconds with
alternate excitation wavelengths of 340 and 380 nm (bandwidth, 5 nm),
and emission was collected at 510 nm (bandwidth, 5 nm).
Autofluorescence was measured in similar cells that had not been loaded
with fura 2-AM and was <5% of the total fluorescence of fura
2loaded VSMCs. After the subtraction of autofluorescence for each
wavelength, the ratio (R) of the measured fluorescence values at 340-
and 380-nm excitation was calculated.12 13 As reported by
Cobbold and Rink16 for fura 2 (which exists in only two
forms, free and calcium bound), the signal from these two wavelength
pairs is uniquely determined by the ratio of free and bound dye and
therefore by the free cytosolic calcium. The ratio method eliminates
the variation due to instrumental fluctuations and changes in the dye
content of the cells, eg, changes due to bleaching or
leakage.16 The excitation ratio of resting VSMCs (F340
nm/F380 nm) remained constant during the whole experiment, indicating a
stable resting [Ca2+]i in VSMCs.
Calibration of the fluorescence signal in terms of
[Ca2+]i was performed with digitonin
and EGTA at the end of each measurement. Digitonin (1 mmol/L) and EGTA
(5 mmol/L) were sequentially added to determine the maximum (Rmax) and
the minimum (Rmin) of the F340 nm/F380 nm excitation ratio,
respectively. Control experiments confirmed that further increase of
the digitonin or EGTA concentration had no effect on Rmax or Rmin,
respectively. [Ca2+]i was calculated
according to the equation of Grynkiewicz et al12 :
[Ca2+]i=K · (R-Rmin)/(Rmax-R),
where K is Kd · Fmin 380/Fmax 380 (Fmin
380/Fmax 380 represents the ratio of the fluorescence at 380-nm
excitation measured in EGTA plus digitonin to that measured in 1 mmol/L
external Ca2+ plus digitonin, and
Kd represents the dissociation constant
of fura 2 for Ca2+, which was set to be 224
nmol/L, according to Grynkiewicz et al). Leakage during the measurement
was <4% of the total fluorescence, as observed by quenching external
fluorescence with 100 µmol/L MnCl2 as described
elsewhere.17
Hemodynamic Measurements
To assess the effect of vasopressor fractions on vascular
resistance, perfusion pressure of an isolated rat kidney perfused with
constant flow was monitored; the fractions to be tested were injected
in a volume of 100 µL.18 To test the effects of CoASSG
in a further vascular preparation, isolated rat mesenteric vascular
beds were prepared according to Kawasaki et
al.19 Pressure monitoring and perfusion with PSS were
similar to the procedures used for the isolated perfused rat kidney,
except a flow rate of 5 mL/min was used. Mean arterial pressure
was measured in the anesthetized rat (1.4 g/kg urethane) after
intra-aortic injection of the substances to be tested as described
previously.20 The experimental procedure was in accordance
with the institutional guidelines for animal research.
To test whether vasopressor effects were caused by
noradrenaline or other catecholamines activating
-receptors,
fractions obtained from adrenal gland extracts were also tested in the
isolated perfused kidney in the presence of 10-6 mol/L
phentolamine.
| Results |
|---|
|
|
|---|
-receptor blocker
phentolamine, suggesting that the second vasoactivity eluted most
likely contained catecholamines stimulating
-receptors. The first
vasoactivity remained essentially unchanged, indicating that substances
other than catecholamines were effective. This vasopressor fraction
then underwent reversed-phase displacement chromatography (step 2, Fig 1C
|
|
The active substance in the vasoactive fraction in step 5 was first
analyzed by UV spectroscopy. As shown in Fig 2A
, the UV
spectrum of adenine was obtained, including the characteristic shift
obtained by acidification to pH 2. MALDI-MS revealed a molecular mass
of 1072 D (Fig 2B
). To assess the number of charges in the vasopressor
compound, MALDI-MS was performed with 10 mmol/L KCl in the analyte
solution. Under these conditions, not only protons but also
K+ ions can be attached to the compound to produce
single-charged molecules, generating additional signals according to
the number of negative charges. Therefore, in K+-containing
medium, the number of signals is in accordance with the number of
possible combinations of attached H+ and K+
ions, the total number of attached cations exceeding the number of
negative charges by one, since only positively charged molecules are
detected. After the addition of K+, MALDI-MS
revealed seven additional peaks, indicating the presence of six
negative charges (Fig 2C
). Thereafter, attempts were made to cleave the
molecule enzymatically. Calf spleen phosphodiesterase (3'-nucleotidase)
specific for 3' phosphoester linkages and snake venom phosphodiesterase
(5'-nucleotidase) were ineffective, but alkaline phosphatase cleaved
one phosphate group from the active substance, as revealed by MALDI-MS
analysis (data not shown) and retention times of the cleavage
products (Fig 3A
). When snake venom phosphodiesterase
(5'-nucleotidase) was added subsequently, MALDI-MS (data not shown) and
the retention time of the split product indicated that AMP was cleaved
from the molecule (Fig 3A
). Both mercaptoethanol (data not shown) and
glutathione reductase (Fig 3B
and 3C
) yielded coenzyme A and
glutathione, as revealed by retention times and MALDI-MS (data not
shown). Furthermore, the retention time of the active fraction in step
5 was identical to that of authentic CoASSG, and MALDI-MS of
commercially available CoASSG with and without a K+ salt
revealed a pattern identical to that in the fraction of step 5.
|
|
CoASSG was also found in chromaffin granules isolated from bovine
adrenal medulla, as shown by the identical retention time (Fig 4A
) and MALDI-MS analysis (data not shown).
Similarly, when adrenal medulla slices were stimulated with
10-4 mol/L carbachol, CoASSG was identified in the
supernatant by its retention time and by MALDI-MS analysis (Fig 4B
). From the UV absorption by CoASSG, it was estimated that from 15 g
of adrenal tissue
20 nmol CoASSG could be released by carbachol. On
the other hand, the adrenal medulla slices incubated in the absence of
carbachol did not release detectable amounts of CoASSG.
|
Authentic CoASSG showed the following effects on vasculature (Fig 5
): (1) vasoconstriction in the vasculature of the
isolated perfused rat kidney in concentrations
10-12
mol/L (Fig 5A
and 5B
) and at isolated perfused mesenteric vascular beds
(Fig 5C
), (2) increase in blood pressure after intra-aortic injection
of 5x10-10 mol in the rat (Fig 5D
), and (3) increase in
[Ca2+]i in rat vascular smooth muscle
cells (Fig 5E
). Furthermore, CoASSG potentiates the effect of
angiotensin II on both cytosolic free Ca2+ and
vascular tone (Fig 5A
, 5B
, and 5E
). The effects of the fraction in step
5 were identical to those of authentic CoASSG.
|
| Discussion |
|---|
|
|
|---|
Ondarza21 first isolated CoASSG from rat liver. CoASSG was
found in mitochondria from rat and bovine liver22 23 24 and
in various bacteria.25 26 In Escherichia coli,
up to 90% of the total pool of coenzyme A is found as
CoASSG,27 compared with
12% in liver.28
Several biological actions of CoASSG have been described. CoASSG
inhibits RNA polymerase in E coli.29 Under
anaerobic conditions, CoASSG reversibly inhibits phosphofructokinase in
rabbit skeletal muscle.30 Furthermore, CoASSG inactivates
the hydroxymethylglutaryl coenzyme A reductase, probably by
inactivating reactive sulfhydryl groups.31
In the literature, the vasoactive properties of CoASSG have as yet not
been described. The angiotensin IIpotentiating effect of CoASSG
suggests that besides its direct vasoactive actions, CoASSG may act by
modulating the angiotensin II effects. In earlier studies by Michelakis
et al32 and Mizukoshi and Michelakis,33 a
circulating agent potentiating the vasopressor effects of angiotensin
II has been described in renovascular hypertension. Since these authors
reported that this substance had a molecular mass of
1000 D, it
could well be identical to CoASSG.
A vasopressor effect of CoASSG was demonstrated in both renal and mesenteric vasculature. The actions of CoASSG on isolated VSMCs suggest that the vasopressor effects are due to an increase in [Ca2+]i in VSMCs. Higher concentrations of CoASSG are required in isolated VSMCs than in intact vessels. This may be due to the changes in VSMCs occurring after repeated culturing. For example, it is known that Ca2+ handling in subcultured VSMCs is partially altered.34 It is yet to be determined which type of receptor is required for the action of CoASSG and which signal transduction pathways mediate the action.
CoASSG was also found in the chromaffin granules of the adrenal
medulla, and after stimulation with carbachol, CoASSG is secreted by
adrenal medulla slices in concentrations sufficient to induce systemic
effects. Therefore, systemic actions of CoASSG can be assumed. CoASSG
is a potent vasopressor, as it is effective in concentrations of
10-12 mol/L, which is comparable to the efficacy of
angiotensin II or endothelin.3
| Acknowledgments |
|---|
Received September 8, 1994; accepted December 20, 1994.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. Jankowski, A. Schroter, M. Tepel, M. van der Giet, N. Stephan, J. Luo, W. Zidek, and H. Schluter Isolation and Characterization of Coenzyme A Glutathione Disulfide as a Parathyroid-Derived Vasoconstrictive Factor Circulation, November 14, 2000; 102(20): 2548 - 2552. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. Wang and N. Ballatori Endogenous Glutathione Conjugates: Occurrence and Biological Functions Pharmacol. Rev., September 1, 1998; 50(3): 335 - 356. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Research Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1995 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |