WES m MGWQ WC ENZYME

Thasfs far the Beams 0? 3321 a.
WWW STAYS. Efi‘éi‘vmgfi‘
EMES 50% 33mm

“we“

L5:

 
  

39-112.}!

    

This is to certify that the

thesis entitled

Studies On Angiotensin Converting Enzyme

presented by

James John Summary

has been accepted towards fulfillment
of the requirements for

 

Ph . D a degree in Biochemi St Ty

 

 

 

 

11I131213 r f’
MlChlgan L we
University

ABSTRACT

STUDIES ON ANGIOTENSIN CONVERTING ENZYME

BY
James John Summary

Since angiotensin I converting enzyme acts as a dipep-
tidyl carboxypeptidase on the decapeptide angiotensin I by
liberating histidyl-leucine and the hypertensive octapeptide
angiotensin II, we sought a relatively inexpensive
peptide substrate analog to routinely assay and characterize
this enzyme. An enriched preparation of human angiotensin
converting enzyme was isolated from the middle portion of three
fractions obtained by gel filtration of serum over Sephadex
G-200 in 0.1 M tris-hydrochloric acid buffer at pH 8.0, with
subsequent dialysis and lyophilization. Commercially avail-
able porcine angiotensin converting enzyme was included for
comparison. In addition to converting angiotensin I to
angiotensin II, both preparations catalyzed the hydrolysis
of the peptides L—phenylalanyl-glycyl—glycine, L-phenylalanyl-
glycyl-L-phenylalanyl-glycine, L-phenylalanyl-glycyl-glycyl-
L-phenylalanine, glycyl-L-histidyl-glycine and hippuryl-
glycyl-glycine. This was demonstrated by incubating various
concentrations of converting enzyme for one to three hours
with the substrate analog in 0.05 M phosphate buffer at pH

7.25 containing 0.1 M sodium chloride. The dipeptides

James John Summary

liberated were identified by thin-layer chromatography and
quantitively determined using colorimetric and fluorometric
techniques.

Two methods were developed for the direct chemical
assay of human plasma angiotensin converting enzyme. They
were based on the spectrOphotometric and spectrofluoro-
metric determination of histidyl-glycine, which was a product
of the hydrolysis of the substrate analog glycyl-L-histidyl-
glycine. o-Phthalaldehyde reacted with the imidazole moiety
of histidine to produce a fluorophore and a chromophore under
acid and alkaline conditions respectively; the activity of
the converting enzyme was proportional to the fluorophore
and chromophore concentration.

Glycyl-L-histidyl—glycine was hydrolyzed at a rate
about twice that of the natural substrate angiotensin I,
with a Km of 6.0 x 10-4 M as compared to 4.5 x 10"5 M for
angiotensin I. The hydrolysis of angiotensin I and glycyl-
L-histidyl-glycine were both inhibited by EDTA, pyrophos-

phate, p-chloromercuribenzoate, 8-hydroxyquinoline, Bothrops

 

Jararaca venom extract and two new natural inhibitors, human
urinary factor and plasma inhibiting factor.

The reaction of o-phthalaldehyde and B-mercapto-
ethanol with amino acids, peptides, and proteins yielded a
fluorescent product. Neither the mechanism of the reaction nor
the structure of the fluorophore had heretofore been character-
ized. For the amino acid derivatives, this was elucidated by

the synthesis of possible intermediates and characterization

James John Summary

of a solid reaction product by absorption and resonance
Spectroscopy. Evidence was found that the hemimercaptal was
produced with one aldehyde group of o-phthalaldehyde and
that a Schiff base with the amino acid was formed at the
other aldehyde group. The reaction was further extended to
peptides and proteins and characterized. The fluorescence
of the protein-o-phthalaldehyde~B-mercaptoethanol adduct was
found to be prOportional to protein concentration and mole-
cular weights of the proteins investigated. The reaction
with proteins showed essentially the same characteristics as

the reaction with amino acids. Energy transfer was observed

from the trytoPhan residues in bovine sorum albumin to the
O-phthalaldehyde—B-mercaptoethanol adduct. The fluorescence
from the o-phthalaldehyde-B-mercaptoethanol reaction was used
for the estimation of the molecular weight of four purified
proteins, for the quantitative determination of proteins,
peptides, and amino acids. It can also be used for the

assay of proteases and peptidases, since cleavage of peptide
bonds results in a net increase in the number of free amino

groups and thus an increase in derivable fluorescence.

STUDIES ON ANGIOTENSIN CONVERTING ENZYME

BY

James John Summary

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Biochemistry

1972

ACKNOWLEDGMENTS

To my fellow men:

"A hundred times every day I remind myself that my
inner and outer life depend on the labors of other
men, living and dead, and that I must exert myself
in order to give in the same measure that I have
received."

—-- A. Einstein

To the Department of Army, Office of the Surgeon General:

I sincerely appreciate the selection and total per-
sonal financial support of this endeavor.

A special thanks is given to Supply Division, PSO,
OTSG, for kindly supplying the spectrophotofluoro—
meter.

To Michigan State University:

Biochemical research is done in a laboratory adorned
with equipment and men of intellect. I apprecia—
tively acknowledge that MSU has responded in full
measure to the following exhortation:

"Take interest, I implore you, in those sacred
dwellings which one designates by the expressive
term: laboratories. Demand that they be multiplied,
that they be adorned. These are the temples of the
future. . . temples of well-being and happiness.
There it is that humanity grows greater, stronger,
better."

—-- L. Pasteur
To Dr. Hans A. Lillevik:
Your daily advice and literature references, eXperi-
mental suggestions, historical reviews, informational

sessions, and laboratory conviviality contributed
immeasurably to this biochemical treatise.

ii

To Mr. Jeffery L. Browning:

An especial thanks is extended for your camaraderie
and friendship, zestful technical assistance, and
youthful infusion of ideas.

To Dad and Janie:

"To one fixed trust my spirit clings."
--— J. Whittier

To the Profession of Chemistry:
"The chemists are a strange class of mortals,
impelled by an almost insane impulse to seek their
pleasure among smoke and vapor, soot and flame,
poisons and poverty. Yet, among all these evils, I
seem to live so sweetly that may I die if I should

change places with a Persian king!"

--- J. Becker
17th Century Chemist

iii

LIST OF TABLES

TABLE OF CONTENTS

LIST OF FIGURES . . . . . . . . . . . .

LIST

I.

II.

III.

OF ABBREVIATIONS . . . . . . . . . .

INTRODUCTION 0 O O O O O O O C O 0

LITERATURE REVIEW . . . . . . . . .

EXPERIMENTAL PROCEDURE . . . . . . . .

A.

C.

MATERIALS AND EQUIPMENT . . . . . .

l.
2.

Materials A . . . . . . . . .
Major Equipment Items . . . . .

ANALYTICAL METHODS . . . . . . . .

1.

Protein Determinations and

Analyses . . . . . . . . . .

a. The Folin-Ciocalteau-Lowry
Procedure . . . . . . . .

b. Direct Spectrophotometry . . .

c. Electrophoresis in Polyacryla-
mide Gels . . . . . . . .

d. Fluorometry . . . . . . .
Amino Acid and Peptide Determination
and/or Analyses . . . . . . .

a. Spectrophotofluorometric Analysis
b. Thin-layer Chromatography . . .
c. Absorption Spectroscopy .
d. Infrared Spectroscopy . .
e. NMR Spectroscopy . . . .

PREPARATIVE METHODS . . . . . . .

Desalting of Human Plasma on Sephadex
G-25 o o o o o o o o a o 0
Human Serum Angiotensin Converting
Enzyme by Gel Filtration . . . .
Human Serum Angiotensin Converting
Enzyme by Ion Exchange Chromatography

iv

Page
viii
ix

xii

10
10

10
ll

11

11

12
12

13
15

15
15
16
17
18
18

18

l8
18

19

E.

Human Blood Plasma Protein Fractions
by Cohn's Procedure . . . . . . .
Human Urinary Factor and Plasma
Inhibiting Factor . . . . . . .
Bothrops Jararaca Pit Viper Venom
Extract . . . . . . . . . . .
Canine Adrenal Gland Extract . . . .
Synthesis of Intermediates Involved

in the Fluorometric Assays . . . .

a.

b.

c.
d.

o-Phthalaldehyde and Glycine
Reaction Product . . . . .
o—Phthalaldehyde, Glycine and
EFMercaptoethanol Reaction
Product . . . . . . .
o-Hydroxymethylbenza1dehyde . .

Synthesis of Modified Substrates .

a.
b.

C.

o—Dihydroxymethylbenzene . . . .

N-Acetyl Glycyl-Histidyl-Glycine
Glycyl-Histidyl-Glycine Ethyl
Ester . . . . . . . .
Angiotensin I Ethyl Ester . . .

ASSAY METHODS FOR DETERMINATION OF ANGIO-
TENSIN CONVERTING ENZYME ACTIVITY . . .

l.

3.

Current Methodology for Detecting
Peptide Bond Cleavage . . . . . .

a.
b.

Ninhydrin Method . . . . .
Fluorometric Assay Using o-Phthal-
aldehyde . . . . . .

Methodology Developed in This Study .

a.

d.

Fluorometric Assay Using Sub-
strate Analog Glycyl—Histidyl-
Glycine and o-Phthalaldehyde,
Macro Method . . . . . . . .
Fluorometric and Colorimetric
Assays Using Substrate Analog
Glycyl-Histidyl-Glycine and
o-Phthalaldehyde, Micro Methods. .
Fluorometric Assay Using Substrate
Analogs, o-Phthalaldehyde, and
B—Mercaptoethanol . . . . . .
Fluorometric Assay of Angiotensin
Converting Enzyme in Human Plasma .

Converting Enzyme Activity in Cohn
Human Plasma Fractions . . . . . .

REACTION MECHANISMS . . . . . . . .

1.

Factors Influencing Angiotensin Con-
verting Enzyme Activity on Substrate

Analogs O O O C I O O O O O O

V

Page

20
21

22
22

23
23
23
24
25
25
25
26
26
26
26
27

29
30

30

31

33
33
34

35

35

IV.

RESULTS

A.

a. Effect of Chelating Agents . . .
b. Influence of Anions . . . . .
c. Influence of Inhibitors . . . .
d. Effect of Esterified and
N-terminal Modified Substrates . .
Small Peptide Hydrolysis by Human
ACD Plasma . . . . . . .
Studies on the Mechanism of the Fluoro—
genic Reaction of o-Phthalaldehyde
with E-Mercaptoethanol . . . . . .
a. Reaction of Proposed Inter-
mediates with Asparagine . . . .
b. Effect of Oxygen . . . .
c. Influence of Amino and Carboxyl
Group Proximity . . . . . . .
d. Titration with p-Mercuribenzoate .
e. Effect of Reducing Agents . . .
f. Fluorescence as a Function of
Protein Concentration and Mole-
cular Weight . . . . . . . .

9. Effect of pH . . . . . .
h. Binding of o-Phthalaldehyde to
Albumin O O O O O O O O

1. Binding of B-Mercaptoethanol to
o-Phthalaldehyde and Albumin . .

ISOLATION, PURIFICATION AND
ASSAY OF ANGIOTENSIN CONVERTING
ENZYME . C O O O C O O I O O O

l.

Angiotensin Converting Enzyme Prepara-
tions and Their Assay . . . . . .
a. Desalted Human Plasma Activity . .
b. Isolation by Gel Filtration . . .
c. Purification by Ion Exchange
Chromatography . . . . . .
d. Activity in Cohn Fractions of
Human Plasma . . . . . .
Cofactor Requirements for Activity
a. Divalent Metal Cation . . .
b. Monovalent Anion . . . .
Substrate Analogs . . .
Fluorometric Assay Using o-phthal-
aldehyde and EFMercaptoethanol . . .
Photometric and Fluorometric Assays
Using Glycyl—Histidyl-Glycine as
Substrate . . . . . . . . . .
a. Converting Enzyme Assay in
Human Serum Preparations . . . .
b. Angiotensin Converting Enzyme
Assay in Plasma . . . . . . .

vi

Page
35
36
36
37

37

38

38
39

39

4O
4O

4O
41
41

41

43

43
43
43
48
57
57
57
62
62

69

74
74

89

MECHANISM OF THE FLUOROGENIC REACTION .
l. Ultraviolet Spectral Changes . . .
2. Infrared SpectroscoPy . . . . . .
3. Nuclear Magnetic Resonance Spectro-
sc0py . . . . . . . . . . .
4. Effect of Carboxyl Group Proximity
on Fluorescence . . . . . . .
5. Influence of Oxygen, Reducing Agents
. and Alkylating Agents on Fluorescence
6. Postulated Intermediates . . . . .
7. Fluorescence of Peptides . . . . .
8. Fluorescence of Proteins . . . . .

V. DISCUSSION

A.

B.

C.

REFERENCES

ACTION OF ANGIOTENSIN CONVERTING ENZYME ON
SUBSTRATE ANALOGS O O O O O O C 0

SIDE CHAIN SPECIFICITY OF ANGIOTENSIN
CONVERTING ENZYME . . . . . . . .

THE FLUORESCENCE REACTION OF AMINO ACIDS,
PEPTIDES, AND PROTEINS WITH O-PHTHAL-
ALDEHYDE O O O C O O C C C O 0

vii

Page
97

97
101

101
106
106
108

108
108

127

131

138

152

Table

II.

II.

III.

IV.

VI.

VII.

VIII.

IX.

Scheme

II.

LIST OF TABLES

Purification of Angiotensin Converting
Enzyme o o o o o o o o o o o o o

A. Substrates Tested for Dipeptidyl Carboxy—
peptidase Activity with Angiotensin
Converting Enzyme . . . . . . . .

B. Peptides UnhydrolyzedknrAngiotensin Con-
verting Enzyme . . . . . . . . .

Hydrolysis of Glycyl-L-Histidyl—Glycine by
Polyacrylamide Gel Electrophoresis Fractions

Inhibition of Human Plasma Angiotension Con-
verting Enzyme . . . . . . . . . .

Effect of Angiotensin Converting Preparations
and ACD Plasma on Small Peptides . . . .

Effect of Compounds and Extracts on Angio—
tensin Converting Enzyme Activity in Human
Plasma and Serum Preparation . . . . .

Equimolar Fluorescence Intensities of Products

Derived From Selected Amino Acids After
Reaction with o-Phthalaldehyde, B-Mercapto-

ethanol, or Analogs- . . . . . . . .

N-Terminal Modified Substrate Analogs Tested
for Converting Enzyme Activity . . . . .

Amino Compounds and Conditions Necessary for

Their Fluorophore Formation and Fluorescence
Upon Reaction with o-Phthalaldehyde . . .

The Renin-Angiotensin System and Role of
Angiotensin Converting Enzyme . . . . .

Isolation of Angiotensin Converting Enzyme .

viii

Page

55

70

71

87

88

96

98

107

133

139

56

Figure

l.

10.

11.

12.

LIST OF FIGURES

Elution pattern obtained by chromatography
of human plasma on a column of Sephadex
G-25 o o o o o o o o o o o o 0

Gel filtration of human serum on Sephadex
6-200 0 O O O O O O O O O O O O

Effluent diagram of Fraction II from DEAE—
Sephadex A-50 . . . . . . . . . .

Standard curve for histidyl-leucine after
reaction with o-phthalaldehyde . . . . .

Polyacrylamide gel electr0phoresis patterns

Effect of EDTA on the hydrolysis of L-phenyl-
alanyl-glycyl-glycine and hippuryl-glycyl-
glycine . . . . . . . . . . . .

Effect of EDTA on the hydrolysis of L-phenyl-
alanyl-glycyl-L-phenylalanyl-glycine and L-
phenylalanyl-glycyl-glycyl-L-phenylalanine .

Effect of chloride ion on the activity of
converting enzyme using substrate analog
L-phenylalanyl-glycyl-glycyl-L-phenylalanine

Effect of chloride ion on the activity of
converting enzyme using substrate analogs
L-phenylalanyl—glycyl-L-phenylalanyl—glycine
and hippuryl-glycyl-glycine . . . . . .

Glycyl-glycine standard curve using the nin-
hydrin procedure for colorimetric assay . .

Fluorometric determination of peptide hydroly-

sis using o-phthalaldehyde and Bdmercapto-
ethanol

Activity of human converting enzyme as a
function of protein concentration using L-
phenylalanyl-glycyl-L-phenylalanyl-glycine
as substrate and o-phthalaldehyde and EFmer-
captoethanol fluorescence assay . . . .

ix

Page

45

47

50

52

53

59

61

64

66

68

73

76

Figure Page

13. Protein and histidyl—glycine spectral
absorbance curves after reaction with
o-phthalaldehyde . . . . . . . . . . 79

14. Hydrolysis of glycyl-L-histidyl-glycine by
plasma angiotensin converting enzyme . . . 81

15. Effect of substrate concentration on the
activity of plasma angiotensin converting
enzyme, measured by the rate of glycyl-L-
histidyl-glycine hydrolysis; Michaelis
Constant . . . . . . . . . . . . . 83

16. Comparison of spectrophotometric and spectro-
fluorometric assays for angiotensin converting
enzyme as a function of enzyme concentration . 85

17. ACD Plasma catalyzed hydrolysis of glycyl-L-
histidyl-glycine as a function of time . . . 92

18. Activity of human angiotensin converting
enzyme as a function of plasma concentration . 94

19. Ultraviolet spectra of the reaction involving
o-phthalaldehyde, ELmercaptoethanol and
gIYCine O O O O O O O O O O O O O 100

20. Ultraviolet Spectra of the crystalline
o-phthalaldehyde, Ekmercaptoethanol and
glycine adduct . . . . . . . . . . . 103

21. Diagramatic nuclear magnetic resonance
spectra of o-phthalaldehyde, ELmercapto-
ethanol, and adducts . . . . . . . . . 105

22. pH profile of the fluorescence reaction with
albumin . . . . . . . , , , , , , 110

23. Corrected fluorescence excitation and emis-
sion spectra of albumin: o—phthalaldehyde
and albumin; and o-phthalaldehyde, B-mercapto—
ethanol and albumin . . . . . . . . . 112

24. Corrected fluorescence excitation and emis-
sion spectra for o-phthalaldehyde plus
B-mercaptoethanol and o-phthalaldehyde,
B—mercaptoethanol, and glycine . . . . . 114

25. Partial quantum efficiency of fluorescence
as a function of wavelength for o-phthal-
aldehyde, B-mercaptoethanol and glycine;
o-phthalaldehyde and albumin: and o-phthal-
aldehyde, B-mercaptoethanol, and albumin . . 117

X

Figure

26.

27.

28.

29.

30.

31.

Page

Fluorescence as a function of concentration
for aldolase, albumin, chymotrypsinogen A,
and ovalbumin . . . . . . . . . . . 119

Fluorescence of proteins as a function of
molecular weight . . . . . . . . . . 121

Binding of o-phthalaldehyde to albumin in the

presence of excess B-mercaptoethanol and the

binding of B-mercaptoethanol to o-phthalal-

dehyde and albumin . . . . . . . . . . 123

Scatchard plots of the binding of o-phthal-
aldehyde to albumin and of B-mercaptoethanol
to o-phthalaldehyde . . . . . . . . . 125

Proposed active site of angiotensin converting
enzyme based upon substrate analog and inhibi-
tor experiments . . . . . . . . . . . 137

Proposed reaction of o-phthalaldehyde with
B-mercaptoethanol and amino acids . . . . . 142

xi

LI ST OF ABBREVIATIONS

Acid Citrate-Dextrose
Benzyloxycarbonyl

Bovine Serum Albumin
t-Butyloxycarbonyl
p-Chloromercuribenzoate
o-Dihydroxymethylbenzene
Diisopropylfluorophosphate
Dithiothreitol

Distilled and Deionized
Ethylenediaminetetraacetate
Glutathione

Human Urinary Factor
‘o-Hydroxymethylbenzaldehyde
B-Mercaptoethanol
o-Phthalaldehyde

Plasma Inhibiting Factor

xii

ACD

BSA
BOC
PCMB
DMB
DFP
DTT
DDI
EDTA
GSH
HUF
HMB
BME
OPT

PIF

 

I. INTRODUCTION

Hypertension is regarded as a physiological state of
elevated blood pressure and, in humans, may increase with
age or certain diseases. Essential hypertension in man is
of ill defined etiology, whereas renal hypertension may be
explained by a series of biochemical or molecular events.
These have been described as comprising the renin-angiotensin
system.

The operation of the renin-angiotensin biochemical
apparatus can be initiated when reduced renal perfusion pres-
sure develops. The resulting dimished renal blood flow and
arterial pressure stimulates the renal cortex to secrete
into the blood stream the proteolytic enzYme called renin.

A plasma glycoProtein of the a-Z globulin fraction has been
identified as the substrate (also called angiotensinogen)
that undergoes limited proteolysis by renin, liberating from
the N-terminal end the biologically inactive decapeptide
angiotensin.I.

The second phase of the renin-angiotensin system
comes into play when angiotensin I encounters plasma or
tissue converting enzyme, a dipeptidyl carboxypeptidase
that catalyzes the hydrolysis of angiotensin I into the
vasoactive octapeptide angiotensin II and the C-terminal

derived dipeptide histidyl-leucine. Angiotensin

II has a prime role in the regulation of blood pressure by
its ability to effect arteriole constriction and to stimu-
late the production of aldosterone by the adrenal glands.
Aldosterone causes sodium retention by the kidney and
increases blood volume with the concomitant elevation in
pressure within the renal circulation.

Examination of past investigations into the renin~
angiotensin system indicates that the second phase involving
the reaction catalyzed by angiotensin converting enzyme to
produce the vasopressive product has been least thoroughly
researched. Boucher gt_§1. (60) put it this way: "Although
converting enzyme is a cornerstone of the renin-angiotensin
system, little is known about its physiological significance
and importance." Since angiotensin converting enzyme occu-
pies a central position in the sequence of hydrolytic reactions
to produce angiotensin II, examination of its enzymic proper-
ties may provide additional clues to the mechanism of renovas-
cular hypertension and aldosteronism. This research effort,
therefore, was directed toward the isolation and biochemical
characterization of this enzyme in order to develop a more
detailed knowledge of its physical and chemical properties
as a protein; to examine the structural requirements for its
catalytic activity and control; and to provide information
about its role at the cellular level. Implied in the latter
was the development of new methods of direct chemical assay

in order to implement these studies. Thus experiments were

undertaken to structurally modify substrates and the enzyme,
to introduce new natural inhibitors and substrate analogs,
and to study hormonal influences on the hydrolysis of sub-
strates. .

Throughout this research endeavor, we have sought
for a relatively inexpensive substrate analog to routinely
assay and characterize this enzyme in order that medical and
clinical laboratories could incorporate the assay for the
evaluation of human hypertensive diseases. To this end, new
substrate analogs for the assay of human angiotensin I con-
verting enzyme, using spectrophotometric and spectrofluoro-
metric techniques, were investigated in detail.

During the development of a fluorescence technique
for the assay of converting enzyme, a sensitive method for
quantitation of proteins and peptides was explored. Using
purified proteins and o-phthalaldehyde, a rapid method for
the confirmation of the molecular weights of four purified
proteins was investigated.

A thorough study was undertaken to elucidate the
mechanism of the reaction of o-phthalaldehyde and B-mercapto-
ethanol with amino acids, to characterize its reaction with
peptides and proteins, and to determine the chemical structure

of the fluorophore that is produced.

II . LITERATURE REVIEW

Thirty-six years after Tigersted and Bergman (1)
found that kidney cortical extracts contained renin (a sub-
stance which increased arterial blood pressure), Goldblatt
§E_gl. (2) in 1934 demonstrated that experimental hyper-
tension could be sustained in the dog by renal artery
stenosis. Six years later Braun-Menendez §t_gl, (3) and
Page and Helmer (4) independently found that renin liberated
a heat stable and dialyzable vasoconstrictor, angiotensin
(hypertensin, angiotonin), from a plasma protein. The
plasma protein, angiotensinogen, was subsequently shown to
be a glycoprotein (5), and occurred with the alpha-2
globulin fraction of blood plasma (6,7). Skeggs gt_gl. (8)
incubated angiotensinogen with trypsin, and subsequently
isolated a tetradecapeptide, which also served as a poly-
peptide renin substrate. Renin by now was identified as a
proteolytic enzyme, present in the renal cortex and cleaved
the leucyl-leucyl bond of angiotensinogen and liberated the
physiologically much less active decapeptide angiotensin I
(9-11). Angiotensin I was subsequently found to be hydro-
lyzed by a plasma and/or tissue converting enzyme at the
phenylalanyl-histidyl bond with the simultaneous production
of the vasoactive octapeptide angiotensin II and the dipep-
tide histidyl-leucine (12). The amino acid sequence of

4

 

angiotensin I and II isolated from horse blood (13), but
not bovine (14,15), has been shown to be identical to human
angiotensin (16,17). To recapitulate, the chemistry
involved is shown in Scheme I.

Skeggs gt_al. isolated three major (A, C1, C2) and
two minor (Bl, B2) forms of the angiotensinogen glyc0protein
(18). The three major forms after purification were found to
have the same amino acid content and molecular weight of
58,000, but their carbohydrate moieties differed in the
amount of sialic acid, glucosamine or hexose contents. Any
other function of these five forms of angiotensinogen is
unknown since they all yield angiotensin I at approximately
the same rates (10).

Angiotensin I converting enzyme was first partially
purified from horse plasma in 1956 by Skeggs and coworkers
(19) and they suggested that the enzyme was a metalloprotein.
Recently Dorer et_§l, (20) from the same laboratory again
partially purified the same enzyme using different method-
ology but could not unequivocally establish the metal
requirement of the native enzyme. Bahkle (21) partially
purified a particulate enzyme from dog lung and found it to
be qualitatively similar to horse plasma converting enzyme,
a finding which was recently reaffirmed (22). Oparil et_gl.
(23) reported lung angiotensin converting enzyme was an
exopeptidase capable ofiydrolyzing the decapeptide, but not

the nonapeptide. Partially purified preparations of plasma

 

mmpflumwm m>fluomsH
mommcflmcwuowmcd a fitfiuomv HH camcmuowmcd

smqlmflm + mzmuoumlmfimusmaHummenam>nmu¢|mm¢

ca m m H

memncm mcfluuo>cou Am>fiuom waxmmsv H cameDOflwsm
samuOHQOOSHOIHmmIHMBIHM>Ismq + smalmfimlmnmnoumnmfimlsmaHuumanam>nmn¢|mm¢

oa m m H

cflcmm cmmocflmcmuoflmcm

aflwuoumoomelummIhmh1am>lsqusmqlmwmnmnmloumlmflmlsmaHIHhBIHm>Imudlmm4

ea OH m H

 

meancm mcfluum>soo swmcmDOflmc< mo
mflom can smumsm camcmuOAScmucflcmm mneuu.H mzmmom

angiotensin converting enzyme was reported to inactivate
bradykinin and hydrolyze various peptides (24). Thus a
definitive characterization of converting enzyme using
direct chemical methods was needed to aid in the under-
standing of.its cellular role in the mechanism of hyper-
tension, since it had been found in heart, liver, aorta,
ileum (25) as well as blood and lung (19,26).

The vasopressor response to injections of angio-
tensin I or II is of short duration since many tissue
extracts of the body contain hydrolytic (peptidase) enzymes
which destroy the active angiotensin II (27). Bumpus g£_al.
(28) and Regoli gt_al. (29) characterized the activities of
these enzymes in rat plasma and found that the major com-
ponent was the aminopeptidase angiotensinase A (60%) and an
endopeptidase that cleaved angiotensin II into two tetrapep-
tides (30). The action of plasma angiotensinases on human
angiotensin I as substrate has not been studied in detail,
since angiotensin II has been used primarily for degradative
studies. Skeggs g£_al. (31) reported that the trypsin
derived tetradecapeptide substrate but not angiotensinogen
was hydrolyzedkurangiotensinases. Ng and Vane (26) reported,
however, that a 105,000 x g supernatant of dog lung homo-
genate contained "destroying enzymes" which inactivate bOth
angiotensin I and II.

For many years the assay of the components of the

renin-angiotensin system solely depended upon the pressor

response of angiotensin II in animals (32-40). Of late
radioimmunoassay techniques (41-49,58,59), though sensitive
are exceedingly involved, expensive and lengthy. Simpler,
sensitive and more direct methods which utilize gas-liquid
chromatography (50), fluorometry (51-54), and spectrophoto-
metry (20,24,55) have been introduced for the assay of some
isolated components of the renin-angiotensin system. Direct
chemical methods were introduced in order to assay for the
converting enzyme during the purification process and its
biochemical characterization. -

Roth (77) empirically develOped a sensitive fluoro-
metric method for the detection of amino acids using o-phtha-
laldehyde (OPT) and B-mercaptoethanol (BME). Although some
basic parameters of the reaction were advanced, no reaction
mechanism or further application to peptides and proteins
was given.

OPT, a weakly fluorescent aromatic dicarbonyl com—
pound, reacts in strong acid or alkali with many amino com-
pounds to produce nonfluorescent and fluorescent products.
For example, glycine can be determined colorimetrically with
OPT (78-84), but usually sensitive fluorometric methods were
used for the determination of an imidazole moiety (54,89,90),
polyamines (87-90), and glutathione (91,92). In contrast,
OPT in the presence of SME, produces highly fluorescent
compounds with amino acids under relatively mild alkaline

conditions. Variations of Roth's procedure for amino acids

were employed in several analytical procedures, and the
ease, sensitivity, and rapidity of the method encouraged

further investigation and extrapolation to peptides and pro-

teins.

 

III. EXPERIMENTAL PROCEDURE
A. MATERIALS AND EQUIPMENT

1. Materials. Unless otherwise stated, all commercial chemi-
cals were analytical reagent grade and used without further

purification.

a. Substrates; Glycyl-L-histidyl—glycine, L—phenyl—
alanyl-glycyl-glycyl-L-phenylalanine, L—phenylalanyl-glycyl—L-
phenylalanyl-glycine and L-phenylalanyl-glycyl-glycine were
purchased from Sigma Chemical Corporation; angiotensin I was
purchased from Schwarz-Mann and hippuryl-glycyl-glycine from

Nutritional Biochemicals. Only the L isomers of peptides and

amino acids were used in this study.

b. Proteins; Ovalbumin, aldolase and chymotrypsinogen
A were procured from Pharmacia Fine Chemicals: Versatol A, a
ly0philized human serum residue, was purchased from General

Diagnostics: bovine serum albumin and porcine angiotensin

converting enzyme was supplied by Pentex Corporation.

c. Chemicals and Biologicals. o-Phthalaldehyde,
B-mercaptoethanol, aldosterone, Bothrops Jararaca pit viper
venom, L-histidyl-leucine, L-histidyl-glycine and chromato-
graphically pure amino acids were ordered from Sigma Chemi-

cal Corporation. Phenol Reagent, for the total protein

10

 

 

11

determination, was purchased from Fisher Scientific Co.

2. Major quipment Items

 

a. Spectrgphotpmetegs; .Absorption spectra were recorded
with a Beckman DB spectrOphotometer and a Sargent SRL
recorder or obtained manually using a Beckman DU spectro-
photometer. Some absorbance measurements were accomplished

with a Coleman Jr. II spectrophotometer.

b. FleogogeEeESL Fluorescence was measured with a

Coleman Model 12 C filter fluorometer and an Aminco-Bowman

Spectrophotofluorometer.

c. Spectrometers. Beckman IR-S and Unicam SP 1000
Infrared spectrometers were used for infrared analysis, and
the Varian T-60 NMR was used for recording the resonance

spectra.

d. thomafiogrep§y_Celemgs;‘ Gel filtration of human
serum proteins was accomplished using 2.5 x 100 cm columns,
connected to a 3-way stopcock and Mariotte flask, supplied

by Pharmacia Fine Chemicals.
B. ANALYTICAL METHODS

1. Protein Determinations and Analyses. All water used
throughout this study was distilled and deionized (DDI)
water. All solution concentrations indicated in % were pre-

pared by weight to volume (w/v) unless otherwise noted.

 

 

12

& EglyggeEygyfiyevjngyfiygysy
Eogry Erecedere (62,63). For the determination, the follow-
ing reagents were prepared: 1 volume of Phenol Reagent was
diluted with 2 volumes of DDI water: anhydrous sodium car-
bonate (2 g), potassium sodium tartrate tetrahydrate (2 g)
and copper sulfate pentahydrate (l g) were separately diluted
to a deciliter (d1) with water. For the protein control,
1y0philized human serum (Versatol A, Lot 2275049) was diluted
with water to yield a concentration of 4.5 g/dl. After 20
minutes, the solution was quantitatively transferred to a
volumetric flask and diluted to 200 ml with 0.9% saline.

The final concentration of proteins was 1.13 mg/ml. To pre-
pare the Folin working solution, 0.5 m1 of the 1% copper
sulfate and 0.5 ml of 2% sodium potassium tartrate was added
to 49 ml of 2% sodium carbonate. The solution was used
immediately.

For the protein determination, 0.20 ml of sample,
standard and water designated as "test," "standard" and
"blank" was added to 4.0 ml of Folin working solution. The
tubes were mixed, allowed to stand 10 minutes, and then 0.4
ml of phenol reagent was added to each. The absorbances of
the solutions were measured at 580 nm between 30 and 60 min-

utes after the addition of the phenol.

m yelygeessmyyyayyyssssswyyz
ESEFX.§FEQESEPE.EP_WEF§P£9_§ES_FEFESEiEP (64). Protein

fractions eluting from the chromatography columns were

13

monitored at 280 nm on a Beckman DU spectrophotometer, using

1.0 cm quartz cells and a slit width of 0.5 mm.

c. ElecErepEogeeie ef_P£oEeinebin_Pelyaerylemide_§els
(67). The separating gel was prepared by dissolving 700 mg
of acrylamide and 21 mg of bisacrylamide in 10 ml of 0.3 M
tris, pH 8.8, and adding 0.060 ml of 5% ammonium persulfate
and 0.020 ml of N,N,N',N'—tetramethylenediamine. After
mixing, it was poured into glass tubes, carefully covered
with a little water, and allowed to set 20 to 30 minutes.

A multicompartmental disc electrophoresis apparatus
employed is described elsewhere (68a & 68b). Briefly, each
tube had its own upper buffer reservoir and current supply
which allowed for independent removal of a gel tube without
disturbing the others. The lower common buffer tank held
1400 ml, the same total volume of the eight upper reservoirs.
The electrodes were composed of 20-guage platinum wire. Cur-
rent was supplied by a Spinco Duostat Model RD constant cur-
rent regulated DC power supply.

Following polymerization, 0.025 ml of protein sample
(10 mg/ml in 20% sucrose containing bromphenol blue as the
tracking front) was carefully layered on the gel, and the
tube inserted into the empty upper buffer reservoir. The
0.05 M borate buffer, pH 9.2, was gently poured into each
upper tank and into the lower buffer reservoir. As an alter-
nate method for sample application, the tubes containing the

gel were inserted into the empty upper buffer reservoirs,

14

then filled with buffer, and using a pasteur pipet, the
sample was layered onto the gel surface. A 3 mA current was
applied to each tube until the bromphenol blue tracking front
migrated to the bottom of the gel tube (ca 35-45 minutes).

The gels were carefully extruded by inserting between
the gel and the glass tube wall a 22-guage needle attached to
a hypodermic syringe filled with water to provide lubrication.
The gels were electrolytically destained for 2 hours. The
destaining unit, powered by a 12 volt, 1 1/2 amp battery
charger, contained in a discarded battery case, used 7% acetic
acid as the electrolyte and two stainless steel plates as
electrodes.

Polyacrylamide gel electr0phoretic analysis was done
on DEAE subfractionated plasma and serum protein preparations
as described in Part C. Following electrophoresis, the gels
were cross-sectioned with a razor blade every 5 mm in order
to locate angiotensin converting enzyme activity. Two bands
were slightly stained as a result of the addition of the
tracking dye bromphenol blue. This provided a convenient
spatial correlation between gels, since converting enzyme
activity could not be detected within a single gel section.
Ten gels were sectioned at a time and the slices soaked over-
night in 2 ml of 0.05 M phosphate knffer, pH 7.25, containing
0.1 M sodium chloride, and the substrate (angiotensin I, 38.5
nmoles: Gly-His-Gly, 373 nmoles) at 4°. The mixtures were

incubated for 2 hours at 370 and assayed fluorometrically.

15

d. greteie_§nelysis_by’Flecgomefiry (116). Usually 10.0
ml of a buffered o-phthalaldehyde (OPT) containing B-mercap-
toethanol (8MB), prepared by diluting 2.5 m1 of OPT (10 mg
OPT diluted to 10 ml with B-mercaptoethanol in methanol) to
a deciliter with 0.05 M borate buffer, pH 9.5, was mixed with
a 0.20 ml aliquot of 1.0 mg/ml protein solution. After
standing at room temperature from 10-60 minutes, fluorescence
was measured in a fluorometer. Protein standards were run
simultaneously and the protein concentration obtained from
the standard curve. Spectrofluorometric scans were measured

under conditions described in the next paragraph.

2. Amino Acid and Peptide Determination and/or Analysis

 

a. §pec£rep§o£o£leo£omeEric_Aeelyeie,* Fluorescence of
solutions containing proteins, peptides, or amino acids with
fluorescence generating reagent was measured with several
instruments. For routine use the Coleman fluorometer (exci-
tation 365 nm; emission greater than 495 nm) was sufficient.
These readings were checked for continuity on Aminco-Bowman
spectrophotofluorometer at various intervals (excitation at
340 nm and emission at 455 nm for solutions using BME;
excitation at 365 nm and emission at 495 nm for solutions

without BME). Corrected fluorescence and partial quantum

 

*I wish to thank Dr. J. F. Holland for his aid and
eXpertise in obtaining and interpreting fluorescence
spectra.

16

efficiency scans were obtained using a unique computer cen-
tered fluorometer that was capable of simultaneous absorbance-
fluorescence measurements (101). The corrected fluorescence
curves were acquired in the form of quanta fluoresced per
unit wavelength and were automatically corrected for the
intensity of the excitation beam and absorption in the

sample cell (102). The resulting quantum efficiency curve
produced a quantity which was linearly related to quantum
efficiency at each point along the wavelength axis of the
excitation scan.

In a typical procedure for the quantitation of amino
acids and peptides, usually 10 m1 of buffered OPT-BME, pre-
pared as described in section B.l.d., was mixed with an
aliquot of amino acid or peptide that contained 50 to 100
nmoles. After standing at room temperature from 10 to 60
minutes, the fluorescence was measured in either the Coleman

or Aminco-Bowman fluorometer.

b. §9232515_9£,ESEF£SE$_SESEAE129_AESS§_PX.EPEPZPEYEF
ghgomangrepEy (65). Filter paper strips, saturated with
migrating solution composed of n-butanol, acetic acid, and
water (4:1:1) were placed along four sides of a glass covered
Brinkman chromatography chamber. Solution was added to a
depth of 8 mm, and the closed system allowed to equilibrate
overnight at 24° before the silica gel G plates (Analtech)
or silica gel F254 plates (Merck) were migrated to a height

of 10 to 15 cm.

17

For visual detection of the separated amino acids or
peptides, the plates were dried at 1100 for 10 minutes and
sprayed with a modified ninhydrin reagent (66) prepared in
the following manner:

Solution I: To 50 m1 of 0.2% anhydrous ethanolic
ninhydrin solution, was added 10 ml of glacial acetic acid
and 2 m1 of 2,4,6-collidine. Solution II: To 1 g of copper
nitrate trihydrate was added 100 m1 of anhydrous ethanol.
Just before use as a Spray, 50 parts of solution I were
mixed with 3 parts of solution II.

The plate was sprayed and held over a hot plate
until even color development occurred; then the spots were
observed by transmitted light and outlined with pencil.

The enzymatic 1ytic products of Gly-His-Gly were
separated using the migrating solution composed of acetone,
water, acetic and formic acids (70:20:8:2) as utilized by

Detterbeck and Lillevik (115).

m Dgyeyéyyyyktyygygfiygyyyr
The electronic absorption spectra of amino acids, peptides
and proteins in the visible and ultraviolet were recorded
at 240 with a Beckman DB spectrophotometer and Sargent SRL
recorder. Quantitative absorbance measurements and some
spectral scans were obtained manually using either the
Beckman DU or Coleman Jr. II spectrOphotometers. Ultra-
violet absorbance measurements were made in aqueous solu-
tions using quartz cuvets of 1.0 cm light path except

where noted.

18

d. Infrered_Speet£oecepy.* Infrared spectra of com-
pounds were obtained with a Beckman IR-5 or a Unicam SP 1000
infrared spectrometers. The solid compounds were scanned in

Nujol mulls.

e. NueleaE Magnetic_Resenenee_Speet£oecepy.* NMR spectra
were obtained with a Varian T-60 NMR spectrometer. The OPT :
reaction preparations were dissolved in D6—acetone, OPT in
carbon tetrachloride, glycine in D20 and B-mercaptoethanol Li

was run neat.
C. PREPARATIVE METHODS:

1. Desalting of Human Plasma on Sephadex G-25. About 2 g of

 

Sephadex G-25 were swelled overnight in 0.9% saline and
poured into a 10 ml pipet having a glass wool plug covered
with glass beads at the lower end. With a Pasteur pipet,
1.0 m1 of fresh human venous plasma was applied directly to
the drained surface of the bed and eluted with saline. The
0.6 m1 effluent fractions were monitored on the Beckman DU
spectrophotometer at 280 nm and by the turbidity produced

when a drop was layered on a 20% trichloroacetic acid solution.

2. Preparation of Human Serum Angiotensin Converting Enzyme

 

by Gel Filtration with Sephadex G-200 (69). About 17 g of

 

*I wish to thank Ms. Elizabeth I. Pupko for her
skilled technical assistance in obtaining NMR spectra and
Professor A. Timnick for helpiJlobtaining IR spectra.

19

Sephadex G-200 was heated on a boiling water bath for 5 hours
in 650 m1 of 0.1 M tris—hydrochloric acid buffer, pH 8.0,
containing 0.2 M sodium chloride. The swollen gel was
allowed to cool to room temperature and a slurry was care-
fully poured down the 2.5 x 100 cm column wall. The flow was
immediately started after filling the column, and the gel was
washed with 1500 m1 of buffer in order to thoroughly equili—

brate the gel. In most eXperiments, upward flow elution was

used in order to decrease the tendency of the bed to pack
due to gravity. The column preparation was modified in later
eXperiments when it was reported (114) that improved flow
rates could be achieved by using an internal support of sili-
conized glass beads 6 mm in diameter.

Using a three-way valve, 2.0 ml of freshly drawn
human serum was applied and immediately followed with 2 ml
of 10% sucrose in buffer to ensure sharp sample application.
The flow rate was adjusted to 15 ml per hour and 5 ml frac-
tions were collected. The entire operation was conducted at
24°. Proteins in the effluent were monitored at 280 nm on a

Beckman DU spectrOphotometer.

3. Purification of Human Serum Aggiotensin Converting Enzyme

 

by DEAE-Sephadex Ion Exchange Chromatography (70). The DEAE-

 

Sephadex A—50 was allowed to swell in a large excess of 0.02
M phosphate buffer, pH 6.8, containing 0.075 M sodium chlo-
ride, for 24 hours. Before use, the ion exchanger was shaken

and was allowed to settle for 7 minutes, and the supernatant

20

buffer discarded. The process of removing fine particles
was repeated five times.

About 40 mg of enriched human converting enzyme
preparation from the middle cut of three fractions obtained
by gel filtration of serum over Sephadex G-200 was dissolved
in 0.5 m1 of 0.05 M phosphate buffer, pH 7.25, containing 0.1
M sodium chloride and centrifuged for five minutes. The
clear supernatant was applied to a 1 x 32 cm column con-
taining ee. 1.5 g of the ion exchanger and eluted with a
stepwise ionic strength gradient of sodium chloride (0.075,
0.15, 0.20, and 0.30 M sodium chloride in 0.02 M phosphate
buffer, pH 6.8). The 2.5 m1 effluent fractions were monitored

on the Beckman DU Spectrophotometer at 280 nm.

4. Human Blood Plasma Protein Fractions by Cohn's Ethanol

Procedure (76). Blood plasma is obtained after removing blood

 

cells by centrifugation. Fractionation is accomplished with
ethanol by controlling such variables as pH, alcohol, ionic
strength, protein concentration and temperature. The protein
pastes* used in this study were designated as Fraction I,
Fraction II + III, Fraction IV-l, Fraction IV-4 and Fraction

V.

 

*The Cohn fractions were kindly.provided by Dr. K. B.
McCall, Assistant Director, Human Plasma Products Section,
Division of Laboratories, Michigan Department of Health,
Lansing, Michigan.

21

5. Partial Purification of Human Urinary Factor (HUF) and

Plasma Inhibitinngactor (PIF). Since angiotensin con-

 

verting enzyme had low demonstrable activity in human serum
(104), blood and urine were examined for possible angio-
tensin converting enzyme inhibitors. Freshly voided human
urine was dialyzed overnight against an equal volume of
water and the dialysate was acidified with glacial acetic
acid to pH 3.5. A method for isolating amino acids and
peptides was then applied (115). A chromatography column
(0.6 x 15 cm) was filled with moist cation exchange resin
(Dowex 50W-X8, 100-200 mesh, H+ form, 300 mg/2 ml of urine
dialysate), and the acidified dialysate was poured gently
onto the column. The acidic effluent was discarded, and

the resin was eluted with 4 m1 of 2 N triethylamine in 20%
aqueous acetone. The eluate was evaporated overnight 13
geeee over concentrated sulfuric acid. The residue (7 mg/2
m1 of urine) was redissolved in 4.0 m1 of water. The plasma
inhibiting factor was prepared similarly.

Further purification of HUF was accomplished by
adding to a column of Dowex 1-X8 (OH-form) 0.2 m1 of HUF, the
aqueous residue obtained by Dowex 50W—X8 fractionation, and
eluting with water. The wash was collected and 1yophilized.

Extraction of an active component was attempted using
either chloroform/methanol (3:2), benzene or petroleum ether.
The Dowex 50W-X8 purified HUF and PIF preparations were sub-

jected to acid hydrolysis by mixing equal volumes of PIF or

 

22

HUF solution with concentrated hydrochloric acid, sealing
them in a capillary tube, and heating at 90° for ee. 24

hours.

6. Extraction of Bothrops Jararaca Pit Viper Venom (110).

g; lgnagaga venom (5 mg) was extracted by the addition of 1.0
ml of absolute ethanol, triturated to a fine suspension, and
allowed to stand at 37° for 48 hours, and centrifuged to
remove turbidity. The crude extract was preincubated with

1 mg of the angiotensin converting enzyme preparation for

15 minutes, incubated with the substrate for 90 minutes, and
assayed according to the microfluoremetric procedure

(Section D.2.b.).

7. Preparation of Canine Adrenal Gland Extract (117).
Freshly excised adrenal glands* (2 g) from a mongrel dog were
extracted with 99% acetone overnight with Stirring at 4°.
Further extraction was done similarly with 80% aqueous ace-
tone and the two extracts combined. The acetone solution

was filtered and concentrated overnight in vacuo, the remain-
ing residue extracted twice with petroleum ether, and the
aqueous mixture incubated at 37° to remove traces of petro-
leum ether. Aliquots were incorporated directly into the

assay procedure.

 

*We are indebted to C. C. Chou, Professor of Physio-
logy and Medicine, for performing the adrenalectomy.

 

 

23

8. synthesis of Intermediates Involved in the Fluorometric

Assays

a. ReecEien_P£oeuet_o£ e-ghEheleleeEyee_aed_Glyeiee;
One gram (0.15 moles) of OPT was dissolved in 3 ml of
methanol and added to 0.55 g (0.15 moles) of glycine in 4 ml
of 50% aqueous methanol. After 4 minutes the solution turned
a deep green and the product crystallized upon slow evapora-
tion of solvent at 40°. The colored product melted between
105 and 110° and migrated only as one component upon Silica
gel thin-layer chromatography, when either methanol/water
(1:1) or butanol/acetic/water (4:1:1) was used as a migrating
solution. The TLC plates showed no new component when
sprayed with ninhydrin. Ultraviolet, infrared, and nuclear
magnetic resonance spectra were obtained on the solid prepara-

tion.

b. ReecEien_P£oguet_of QP'I_‘,_G_1_yei_r_1_e_agd_azMereapteeEhenele
Synthesis was effected as described immediately above, except
that 0.5 ml of pure s-mercaptoethanol was incorporated into
the OPT solution, and upon its addition to glycine the solu-
tion turned a deep red. Slow evaporation of the solvent at
400 yielded a red crystalline product that melted in the
range of 105 to 110° and gave one spot upon TLC when methanol/
water (1:1) was the migrating solution. However, when
butanol/acetic acid/water (4:1:1) was used, four colored prod-

ucts plus one ninhydrin positive component appeared. When a

10 pl aliquot of the reaction mixture was taken before

24

evaporation of the methanol, and diluted to 10.0 ml with 0.05
M borate buffer at pH 9.5, an intense blue fluorescence was
seen. Ultraviolet, infrared and nuclear magnetic resonance

spectra were obtained on the solid preparation.
c- yetheeie 2f_°:de£°§meEhxleeezelseeysa(ENE) _-.

(l) OPT-Ethylene Glycol Monoacetal Synthesis. One
gram (0.07 moles of OPT)and 2 g (0.08 moles) of aluminum
trichloride were mixed with 10 m1 of ethylene glycol and the
mixture was warmed for 3e. 5 minutes to dissolve the com-
ponents. The resulting dark-orange solution was extracted
with 30 ml of benzene, and the extract was washed succes-
sively with 1% sodium hydroxide and 2% sodium carbonate.

The washed benzene extract was dried over calcium sulfate
and then distilled free of benzene to yield an orange syrupy
residue of the acetal. IR spectra of the product indicated
that steric hindrance apparently prevented ethylene glycol

addition to both aldehyde groups of OPT.

(2) Borohydride Reduction. One ml of the acetal
mixture was dissolved in 5 ml of n-propyl alcohol, and to
this was added 0.3 g of sodium borohydride, and the mixture
gently refluxed for 10 minutes. Gentle heating was continued
until bubbling ceased and the colored solution cleared. Ten
m1 of water was added and the mixture allowed to stand over-
night at 37°. When concentrated sulfuric acid was added

until pH 3 resulted, a viscous layer of HMB separated from

 

25

solution. These same properties of the product were described
by Rodd (103). The pale yellow liquid was washed with 2%
sodium carbonate and dried over calcium sulfate. The struc-
ture of HMB and OPT ethylene glycol monoacetal was verified
by IR spectroscopy, reaction with phenylhydrazine and a

positive Tollen's test.

d- eyeteeeie 2f_O:Dihxd£0me£hxleeezene 110E131 2r
e:§y1yleee_Glyeol. One gram of OPT was dissolved in 5 ml of
n-propanol and to this was added 0.3 g of sodium borohydride,
and the mixture was gently refluxed for 10 minutes. Sodium
borohydride reduction was effected as previously described
with the acetal. After addition of 10 m1 of water, the mix-
ture was concentrated by distillation to ee. 8 m1, whereupon
DMB separated as a white solid, (MP 60—63°; Rodd (103) reported

64°). The structure was verified by IR spectroscopy.

9. Synthesis of Modified Substrates for Angiotensin Con-

 

verting Enzyme (118).

 

m wygflfiygyygyfljygyjpggyeToaLO
ml reaction vial containing 2.0 mg Gly-His-Gly dissolved in
50 ul of 0.05 M phosphate buffer, pH 7.25, was added 50 U1 of
saturated sodium acetate. The solution was cooled to 0° and
4 pl of ice-cold acetic anhydride was added in 1 ul aliquots
every 15 minutes. A reaction was evidenced by immediate
clearing of the solution. The resulting mixture was then

diluted to 3.6 ml with 0.05 M phosphate buffer, pH 7.25,

26

thus yielding a 2 mM solution of N-acetyl—Gly-His-Gly. Thin-
layer chromatography of the preparation, using acetone/water/
acetic acid/formic acid (70:20:8:2) as the migrating solvent,
showed one ninhydrin negative component. The N-acetyl-Gly-
His-Gly was visualized by placing the TLC plate in a closed
chamber containing an iodine vapor saturated atmOSphere

(Rf, 0.58).

b- slxcxlflistieyl-Elzcine Ethyl Ester Eyeteeeie- Gly-
His-Gly (1.0 mg) was suspended in 50 ul of absolute ethanol
containing 0.1 M hydrochloric acid and allowed to stand at 4°
for five days. Thin-layer chromatography of the reaction
mixture, using acetone/water/acetic acid/formic acid (70:20:
8:2) as the migrating solution, demonstrated one primary
ninhydrin positive component, the ethyl ester of Gly-His-Gly

(Rf, 0.40) and a trace of Gly—His-Gly (Rf, 0.16).

c. engieteneip I EtEYi Espe£.§yptpeeie. The ethyl ester
of angiotensin I was prepared as described immediately above,
except 0.2 mg were dissolved in 20 ul of the acidic ethanol
solution. Thin-layer chromatography of the mixture, using
acetone/water/acetic acid/formic acid as the migrating solu-
tion, showed one spot (Rf, 0.84) corresponding to ethyl ester

of angiotensin I.
D. ASSAY METHODS FOR DETERMINATION OF ENZYMATIC ACTIVITY.

1. Current Methodology for Detecting Peptide Bond Cleavage

 

27

a. EiEhYSEiE eseay 9f4329£°E°251P_°9829£t192 EPEYES
(71,72). A 4 M lithium acetate buffer was prepared as fol-
lows: 24 g of anhydrous lithium hydroxide was added to 60
3 m1 of water and stirred until about half was dissolved:
cautiously and in small portions were added 64 m1 of glacial
acetic acid. Warming and vigorous bubbling occurred. The
solution was diluted almost to 250 ml with water, and the
pH measured after a 1:3 dilution with water. If the pH was
not 5.2 1 0.05, it was adjusted to this value with 1 M lithium
hydroxide or glacial acetic acid. After cooling to 20°, the

solution was diluted to 250 m1.

(1) Ninhydrin reagent. To 1.0 g of ninhydrin
(Pierce Chemical Co.) was added 37.5 ml of dimethylsul-
foxide and 12.5 ml of 4 M lithium acetate buffer. The
reagent was used immediately, since the cherry red solution
faded to a light yellow when exposed to air. Potency was
maintained for 1-2 weeks if stored at 4° in small brown

bottles filled to the brim.

(2) Standards. A stock standard solution of 50 mM
glycyl-glycine was prepared by diluting 660.6 mg Gly-Gly to
a deciliter with water. To prepare a dilute 100 nmoles/ml
Gly-Gly standard, 1.0 ml of the 50 mM stock standard was
diluted to 500 ml with water. This standard was further
diluted with water to give working standards of suitable

concentration.

‘9“_‘“
a

 

28

(3) Substrates. To prepare a 0.05 M sodium phos—
phate buffer, 3.55 g of anhydrous dibasic sodium phosphate
and 3.0 g sodium chloride was dissolved in 2e: 250 m1 of
water and titrated to pH 7.25 with 0.1 M hydrochloric acid
and diluted to 500 ml with water. To prepare 1.0 mM (1000
nmoleS/ml substrates) 8.6 mg of Phe-Gly-Gly-Phe, 8.6 mg. of
Phe—Gly-Phe-Gly and 5.9 mg of Hippuryl—Gly-Gly were diluted
to 20 ml and 7.0 mg of Phe-Gly-Gly were diluted to 25 ml

with the 0.05 M phosphate buffer.

(4) Angiotensin Converting Enzyme. Enriched prepara-
tions by Sephadex G-200 fractionation of human serum ("Frac-
tion II"), or a subfraction ("Fraction III") of Fraction II
obtained by column chromatography with DEAE-Sephadex A-50
and commercially available porcine plasma converting enzyme

were used for assay.

(5) Procedure. To 0.5 ml of substrate was added
0.5 m1 of enzyme preparation and the solution was incubated
for 3 hours at 37°, using 0.5 m1 of substrate and 0.5 m1 of
phosphate buffer as the reagent blank. All tubes were
placed in a boiling water bath for 5 minutes to terminate
the enzymatic reaction and centrifuged if turbid. A 0.5 m1
aliquot was removed from the Hippuryl-Gly-Gly reaction mix—
ture and 0.1 ml aliquot was taken from the Phe-Gly-Gly, Phe-
Gly-Gly-Phe, Phe—Gly-Phe-Gly incubated mixtures as well as
the corresponding blanks, and each was diluted to 1.0 ml

with water. To each was added 0.5 ml of ninhydrin reagent.

29

Likewise, 0.5 m1 of ninhydrin was added to 1.0 m1 of working
standard and 1.0 m1 of water for the ninhydrin blank. The
tubes were mixed on a vortex mixer for 30 seconds and
covered with aluminum caps. The tubes were placed in vigor-
ously boiling water for exactly 15 minutes, then cooled for
10 minutes in tap water. To each was added 2.5 m1 of water,
mixed on a vortex mixer and centrifuged if necessary. The Y
Coleman Jr. II spectrophotometer was set at 580 nm and the

absorbance read versus water in 1/2 in. cuvets.

b. Flpopogepric_Aesey_pf_engieteneip’Eopverpipg_Epzyme:
$999§P£gJPPQEEFFEPQR?UPPF@FS

Reagent, Macro-Method. (54).

(1) Reagents. For the diluent, a 0.05 M sodium
phosphate buffer, pH 7.25, containing 0.1 M sodium chloride
was prepared. The substrate, angiotensin I (1 mg or 770
nmoles/ml) was then diluted to 1.0 ml with phosphate buffer.
For the standards, a stock histidyl-leucine solution was
prepared by dissolving 1.0 mg or 3720 nmoles/ml in water.
Aliquots were removed to give working standards of suitable
concentration. In addition 2 N sodium hydroxide, 6 N hydro-
chloric acid, and 10 mg/ml of o—phthalaldehyde in methanol
were other reagents required for the determination. Enriched
preparations of converting enzyme from Sephadex G-200 frac-
tionation of human serum ("Fraction II"), from the subfrac-

tionation of Fraction II on DEAE-Sephadex A-50 ("Fraction

30

III"), and commercially available porcine plasma converting

enzyme were used for assay.

(2) Procedure. The assay procedure was initiated by
incubating 0.1 m1 of the enzyme preparation with 0.1 m1 of
angiotensin I at 37°. Meanwhile, a substrate and enzyme
blank were prepared by adding 0.05 ml of each to 7.55 ml of 4
phosphate buffer. At the end of the one to three hour incu- fi
bation period, 0.1 m1 of the incubation mixture was added to
7.5 m1 of buffer. Into all tubes was pipeted 1.0 ml of
sodium hydroxide and 0.25 ml of o-phthalaldehyde. The
tubes were mixed on a vortex mixer and allowed to stand for
four minutes. Then 0.5 m1 of 6 N hydrochloric acid was
added to each. For standards, 2.5, 5.0, 10.0 and 20.0 ul of
stock His-Leu was pipeted into 7.6 m1 of buffer, and sodium
hydroxide, OPT, and hydrochloric acid added as above. A
reagent blank without His-Leu was also prepared. The fluores-

cence was measured either in the Aminco-Bowman Spectrophoto-

fluorometer or the Coleman Model 12 C filter fluorometer.

2. Methodology Developed in This Study

 

a. Flpopopepric_Aesey_of_engieteneip Eopverpipg_Epzyme
@ygyyyyeM§gfiygyyng§ygygygy
gfifl§§§§®j§fifijg@9mygyfiyflfygyL
M°EF97§°Eh2°L. Angiotensin converting enzyme was serially

diluted with 0.05 M phosphate buffer at pH 7.25, containing

0.1 M sodium chloride, to yield 3.25, 1.12, and 0.56 mg

31

protein per 0.5 ml. For the substrate 2.7 mg of Gly-His-

 

Gly were dissolved in 10.0 ml of the same phosphate buffer
to yield a 1.0 mM solution. Aqueous His-Leu (1 mg/ml) was
used as a stock standard.

For the hydrolysis of substrate as a function of
time, 0.5 m1 of 1.0 mM Gly-His-Gly was incubated with 0.5
m1 of enzyme (3.25 mg) at 37°. A 0.1 m1 aliquot of the
reaction mixture was removed at 0, 15, 30, 45, 75, 90, 105,
and 120 minutes after initiation of the reaction and quickly
frozen in a methanol-dry ice bath.

For the catalyzed hydrolysis of substrate as a func-
tion of enzyme concentration, 0.50 ml of substrate was incu-
bated at 37° for 2 hours with 0.50 ml of enzyme containing
3.25, 1.12, and 0.56 mg of protein. Zero time controls were
frozen as above. To each of the 0.10 ml aliquots was added
7.50 ml of 0.05 M phOSphate buffer, 1.0 ml‘2 N sodium hydrox-
ide, 0.25 ml OPT (10 mg/ml in methanol). After 5 minutes,
0.50 ml 6 N hydrochloric acid was added and the fluorescence
measured in the Aminco—Bowman spectrOphotofluorometer. For
standards, 2.5, 5.0, 10.0, 15.0, and 20.0 ul of stock His-
Leu was added to a 0.60 ml buffer, followed by sodium hydrox—

ide OPT, and then hydrochloric acid as above.

b. EPSSEF9PE°E°£1E°£°E¢EFEF_32°_SESEtEPEhEFEWEFEiE.55531
251A29192925EP_°SPY?£FEPE;EPEYT¢_9§32915EPEFE?E?.AE§£PE
Elzszlzflistiéylr§%xsine 299.E92.2'EhEhilElQ?EY§S.RE§2¢2F_F2

ESESEWEPE.EifitidllZQEYESESL.E32F27HSEPESEI

32

(1) Fluorometric Assay. Partially purified con-
verting enzyme preparation (0.05 - 1.0 mg) was dissolved in
0.10 ml of 0.05 M phosphate buffer, pH 7.25, containing 0.1
M sodium chloride, and incubated at 37° for 90 minutes with
0.10 ml of 1.0 mM Gly-His-Gly or angiotensin I (1.0 mg/ml).
The enzymatic reaction was initiated with the addition of Gly-
His-Gly or angiotensin I substrate solution and terminated
with the addition of 0.1 ml of 2 N sodium hydroxide.

The fluorogenic reaction was initiated by the addition of
0.05 ml of OPT (10 mg/ml in methanol). This solution was
allowed to stand for 5 minutes: then 0.1 m1 of 6 N hydro-
chhloric acid was added, and the final volume brought to 1.0
ml (or 2.5 ml if less sensitivity was desired). The addi-
tion of acid precipitated the proteins which were removed by
centrifugation at ca. 2000 rpm for 5 minutes in an Inter-
national Model H clinical centrifuge. After 10 to 30 minutes

following the addition of hydrochloric acid, the fluorescence

was measured in the Aminco-Bowman spectrOphotofluorometer.

(2) Spectrophotometric Assay. The procedure is iden-
tical to the one described immediately above, except the
aborbance of the solution was measured at 420 nm, in 1.0 cm
cuvets using a Beckman DU spectrophotometer, after the addi-
tion of sodium hydroxide. Addition of hydrochloric acid was
omitted and the final volume was brought to 1.0 ml with
water. The absorbance was best measured 20 to 30 minutes

after the addition of the OPT reagent.

33

c. ElpogomeprlceAesey_o£ engieteneip_gopverpipg_Epzyme;
¥9my§92g}3939@9®5§@§@¥99399%
EEJM29NPFESEEEygtyiwffiyyeg
pfgepceppoetpapoi. In a typical assay, 0.5 m1 of 1.0 mM Phe-
Gly-Phe-Gly was mixed with 0.5 ml of converting enzyme solu-
tion and incubated for 3 hours at 37°. For blanks, 0.05 M

borate buffer, pH 9.5, and enzyme preparation were combined

A

as above and boiled for 5 minutes. A 0.5 m1 aliquot of both
digest and blank was added to 9.5 m1 of borate buffered OPT.
Borate buffered OPT was prepared by diluting 2.5 ml of OPT
(10 mg OPT diluted to 10 ml with 5 ul/ml of B-mercaptoethanol
in methanol) to a deciliter with 0.05 M borate buffer, pH
9.5. The fluorescence was measured between 5 and 25 minutes

after the addition of the borate buffered OPT reagent.

d. SpectrOphotometric Assay for Angiotensin Converting

H°§§E¥E.EPS.EYESE.2iE°Eti°ESL.M32?97§°Eh9°;. The substrates,
1.0 mM Gly-His-Gly and OJT7mM angiotensin I, were dissolved
in 0.05 M phosphate buffer, pH 7.25, containing 0.1 M sodium
chloride. To 0.10 ml of acid citrate-dextrose (ACD) plasma
and 0.10 ml of substrate was added 0.10 ml of phosphate buf-
fer and the mixture incubated at 37° for 10 minutes. The
reaction was terminated by the addition of 0.20 ml of 6N

sulfuric acid and 0.20 ml of 10% sodium tungstate. The pre—

cipitated proteins were removed by centrifugation at 2000 rpm

34

 

for 5 minutes in an International Model H clinical centri-
fuge. To 0.40 ml of the clear supernatant was added 0.10

ml of 2 N sodium hydroxide followed by 0.050 ml of OPT (10
mg/ml in methanol). After 10 minutes, the absorbance of the
solution was-measured in 1.0 cm quartz cuvets at 420 nm with a
Beckman DU Spectrophotometer. The absorbance of a protein A
blank (or blank containing the inhibitor in enzyme inhibitor 4
studies) was subtracted and the net absorbance compared to a

His-Gly (Gly-His-Gly as substrate) or His-Leu (angiotensin I

as substrate) standard curve to ascertain the amount of dipep—

tide liberated during the incubation period.

3. Converting Enzyme Activity in Cohn Human Plasma Protein

 

Fractions. Each of the Cohn human plasma fractions (Fraction

 

I, Fraction II + III, Fraction IV-l, Fraction IV-4 and Frac-
tion V), previously lyOphilized from ethanol pastes, were
dissolved in 0.9% saline and dialyzed overnight against water
at 4°. The dialyzed fractions were shell frozen and 1yo-
philized. These preparations were dissolved in 0.05 M sodium
phosphate buffer, pH 7.25, containing 0.1 M sodium chloride,
centrifuged, and the supernatant was serially diluted to give
concentrations from 0.5 to 2.5 mg/ml. One ml of each was
incubated for 3 hours at 370 with 1.0 ml of 1.0 mM Hippuryl-
Gly-Gly: a 0.20 ml aliquot from the incubation mixture was
removed and the liberated Gly-Gly determined by the ninhydrin
procedure.

A fluorometric assay was also performed on each of

35

the Cohn fractions using the substrate analog Gly-His-Gly.

To 0.10 ml (1.0 mg) of the dialyzed and lyOphilized Cohn
fraction was added 0.10 ml of substrate and 0.30 ml of phos-
phate buffer. The mixture was incubated at 370 for 60
minutes and assayed by the microspectrophotofluorometric pro-
cedure.

To demonstrate inhibition by inorganic perphosphate,
8-hydroxyquinoline, and diisopropylfluorophosphate on angio-
tensin coverting enzyme detected in the Cohn fractions, 0.10
ml of 10 mM pyrophosphate, 1.0 ul of 200 ng/ml diisopropyl—
fluorophosphate, and 50 ul of 20 mM 8-hydroxyquinoline were
incorporated into the above described 0.50 ml incubation mix-

tures.
E. REACTION MECHANISMS

1. Factors Influencing Angiotensin Converting Enzyme Acti-

 

vity on Substrate Analogs

 

a. §f£e2t_o£>Eheleting.§genpse Porcine plasma con-
verting enzyme was dissolved in 0.05 M phosphate buffer, pH
7.25, containing 0.1 M sodium chloride, and serially diluted
with the same buffer to protein concentrations of 1.5, 0.75
and 0.375 mg/ml. To 5.0 m1 of the 1.5 mg/ml enzyme prepara-
tion was added 2.0 mg of disodium ethylenediaminetetraacetate
(EDTA) and left overnight at 4°. The protein concentration

was serially diluted as above.

36

b. £9£12°E°E efeApiene. Chloride-free phosphate buffer
was prepared by dissolving 7.1 g of anhydrous dibasic sodium
phosphate in double deionized water, titrating to pH 7.25
with phosphoric acid, and diluting to a liter with water.
Both porcine plasma and human converting enzyme (Fraction II)
preparations were dissolved in chloride-free buffer. Sub-
strates were prepared as described in section D.1.a. (Nin-
hydrin Assay); however, sodium chloride was deleted from the
substrate preparations.

Equal volumes of substrate and enzyme were combined,

incubated for 3 hours at 37°, and the liberated products

quantitated by the ninhydrin assay.

c. Influence of Inhibitors.

(1) Inhibition Studies Using Fraction II. Gly-His-
Gly (100 nmoles) was incubated for 90 minutes with 0.50 mg
of converting enzyme preparation (Fraction II from Sephadex
G-200 preparation) and the indicated amount of inhibitor
(Table IV) in total volume of 0.30 ml. The liberated His-
Gly was determined spectrophotometrically as detailed in

section D.2.b.

(2) Inhibition Studies Using Fraction II and ACD
Plasma. Gly-His-Gly (100 nmoles) was incubated for 90
minutes with 0.50 mg of converting enzyme preparation (Frac-
tion II from Sephadex G-200 preparation) and the indicated

amount of compound or biological extract (Table VI) in total

37

volume of 0.30 ml. The liberated His-Gly was determined
spectrOphotometrically as detailed in section D.2.b.
Similarly, 0.10 ml of ACD plasma was incubated for 10
minutes with the indicated amount of compound or biological
extract, and the liberated His-Gly determined spectrophoto-
metrically as detailed in section D.2.d.

In all inhibition studies, the concentration of the
compound or biological extract was made equimolar'in every
tube in order to correct for any effect that the compounds or

extracts may have had on the assay procedure.

d. Effeet_o£ ESE°£i£i§°_§E°EN2tEFT;Eal.M°§i£ifi°_52bl
etpapee. Angiotensin I (77.0 nmoles), ethyl ester of angio-
tensin I (77.0 nmoles), Gly—His-Gly (100 nmoles), ethyl ester
of Gly—His-Gly (100 nmoles) and n-acetyl Gly-His-Gly (100
nmoles) were incubated for 90 minutes with 0.50 mg of angio-
tensin converting enzyme preparation, in 0.05 M phosPhate
buffer, pH 7.25, containing 0.1 M sodium chloride and a final
volume of 0.30 ml. Similarly, 0.10 ml of ACD plasma was sub—
stituted for the Fraction II preparation and incubated for 10
minutes with the substrates. The liberated His-Gly and His-

Leu were determined spectrophotometrically as described in

section D.2.d.
2. Small Peptide Hydrolysis pypHuman ACD Plasma

ACD plasma (0.10 ml) was incubated for 10 minutes at

370 with 0.10 ml of 1.0 mM peptide (Table V) in a total volume

38

of 0.30 ml. The enzymatic reaction was terminated and the
proteins precipitated by the addition of 0.20 ml of 0.6 N
sulfuric acid and 0.20 ml of 10% sodium tungstate. After
centrifugation, 2.0 m1 of buffered OPT, prepared as described
in section B.l.d., was added to 0.40 ml of the protein free
supernatant. After 10 minutes the fluorescence was measured
in the Aminco—Bowman spectrophotofluorometer with the excita—
tion monochromator set at 340 nm and the emission monochroma-
tor set at 455 nm. The fluorescence of an unincubated blank,
containing all components of the reaction mixture, was
measured simultaneously and the resulting value subtracted
from the fluorescence value obtained from the incubated
sample.

Peptide hydrolysis, catalyzed by a porcine plasma
angiotensin converting enzyme preparation, was measured in an
identical manner, except 0.10 ml of a 5.0 mg/ml solution of
the enzyme was substituted for the plasma and incubated for
140 minutes. The buffered OPT reagent (2.0 ml) was added
directly to the incubation mixture and the fluorescence

measured. Appropriate blanks were prepared as above.

3. Studies on the Mechanism of the Fluorogenic Reaction of
o-Phthalaldehyde and B-Mercaptoethanol with Amino Acids, Pep-

tides,‘and'Proteins

 

a. Relative Fluorescence Produced from the Reaction of

39

reagent was made by addition of 5 ul of HMB (10 mg/ml in
methanol) to 0.70 ml of BME solution (5 ul/ml in 50%
methanol) and diluting to 10.0 ml with 0.05 M borate buffer,
pH 9.5. The buffered DMB-BME reagent was composed of 10.0
mg of DMB preparation and 0.10 ml of BME diluted to 10.0 ml
with the same borate buffer.

Similarly buffered solutions of OPT, HMB, and DMB
were prepared without 8MB. To produce the fluorescent
derivatives, 0.20 ml of a 2.0 mM solution of asparagine was
diluted to 10.0 ml with each of the aforementioned reagents

and the developed fluorescence was measured fluorometrically.

b. §f£¢st_9£.232929_92 Fleopeeceneee_ A 0.20 ml aliquot
of 2.0 mM glycine was diluted to 10.0 ml with buffered OPT-
HME reagent and the resulting fluorescence was measured.

With another identical sample, oxygen was bubbled through the
solution for 10 minutes, and then its fluorescence was
measured and compared with the unongenated sample. A similar
‘ sample without BME was prepared and compressed nitrogen bub-
bled through the solution for 10 minutes and the fluorescence

noted.

a EsyflwyeEJNWENE§WSIEQEBBEFX
en_Fluereseepce. Aqueous solutions of 2.0 mM glycine,
B—alanine, Yeaminobutyric and-e-aminocaproic acids, as well
as the ethyl ester of glycine, were prepared, and from each

a 0.20 ml aliquot was taken and diluted to 10.0 ml with

40

buffered OPT-BME reagent. The resulting fluorescence from

each was measured.

d. Titration of the Fluorescent Derivative with

taining 200 umoles of BME was added 0.20 ml of 2.0 mM gly-
cine, and this solution was titrated with 20.0 mM p-mercuri-
benzoate. The fluorescence was noted after each 0.10 ml

addition.

e. The Effect of Reducing Agents on the Production of

dithiothreitol were substituted for BME in the preparation
of the buffered OPT reagent and reacted with glycine as
before. The intensity of fluorescence resulting from each

reagent was measured.

f. Fluorescence as a Function of Protein Concentration

(1.95 mg/ml), aldolase (0.55 mg/ml), chymotrypsinogen A

(2.85 mg/ml) and ovalbumin (2.50 mg/ml) were prepared in 0.05
M phosphate buffer, pH 7.25, containing 0.1 M sodium chloride,
which produced protein concentrations of 28.9, 3.50, 114.0

and 55.7 nmoles/ml respectively. Aliquots of each protein
solution (0.050 or 0.025 ml) were diluted to 10.0 ml with
buffered OPT-BME reagent and the resulting fluorescence was

measured.

41

m yeyfigytwyeSeflwyyegyf92¥§§e
eriyapigee. Four buffered OPT-BME reagent solutions were
prepared as follows. OPT (0.050 ml, 10 mg/ml in methanol)
plus 0.50 ml of BME (5.0 ul/ml in methanol) were diluted to
a final volume of 20.0 ml with either 0.05 M borate buffer,
pH 11.0; 0.05 M borate buffer, pH 9.5: 0.05 M phOSphate
buffer, pH 7.25; 0.05 M acetate buffer, pH 5.2. For testing
the effect of each on fluorescence, 0.10 ml of bovine serum
albumin (2.5 mg/ml) was diluted to 10.0 ml with the pre-
viously described buffered OPT—6MB reagents at their speci-
fied pH. To observe the fluorescence of the bovine serum
albumin adduct as a function of pH after the fluorophore had
been formed at pH 9.5, bovine serum albumin was reacted with.

buffered OPT-8MB reagent and, after 10 minutes, the Solution

was titrated to the desired pH.

h. Eipdeng’ef_OPT_te’Eozipe_Serem_Albemineb A constant
amount of albumin (0.10 ml of a 15 nmoles/m1 solution) was
reacted with increasing aliquots of OPT solution (74.5 nmoles
per ml of 0.05 M borate buffer of pH 9.5) ranging from 0.05
to 2.0 ml and making the final volume to 10.0 ml with borate
buffer containing 1450 nmoles/ml of BME. Fluorescence was

measured after 10 minutes.

i- Eiedine ef_Bé4_E_te QPI-eoxiee_Serem_Albemz—.nr Aliquots
ranging from 0.65 to 0.20 ml of BME (5 ul/ml in methanol)

were reacted with constant amounts of OPT and albumin (0.10

42

ml of 10.0 mg/ml and 0.10 ml of 15.0 nmoles/ml respectively)
and diluted to 10.0 ml with 0.05 M borate buffer, pH 9.5.

After 10 minutes the fluorescence was measured.

IV. RESULTS

A. ISOLATION, PURIFICATION AND ASSAY OF ANGIOTENSIN CON-

VERTING ENZYME PREPARATIONS

1. Angiotensin Converting Enzyme Preparations and Their Assay

m MyggsagygymfiymeyyflyjeEyyg
EPE§2.2125291 In order to increase the sensitivity of the
ninhydrin procedure for converting enzyme assay, the back-
ground of ninhydrin positive substances was partially reduced
by removal of lower molecular weight compounds from the
plasma proteins using Sephadex G-25. A typical elution curve
is seen in Figure 1. The fractions containing the proteins,
as detected by trichloroacetic acid turbidity and absorbance
at 280 nm, were incubated with Hippuryl-Gly-Gly for 14 hours

and assayed by the ninhydrin procedure. No converting enzyme

activity could be demonstrated in the protein fraction.

b. Human Serum Angiotensin Converting Enzyme by Gel

Filtration. A typical protein elution pattern from Sephadex
G-200 column is shown in Figure 2. Three peaks successively
appeared, the fractions corresponding to the second peak

("Fraction II"), containing most of the activity, were

pooled, dialyzed overnight against water and 1y0phi1ized.

43

44

.Ec omm um magmaomummm mcHQHOmnm mmocmquSm msomom
Icflmuoumlcoc mmumoHUCA m xmmm .GOHuomum ucmSHmmm one CH
pflom oeumomouoanoflup mom LUHB wuwpfinusw mo mocmmnm may
moumoflpcfl Bouum map paw mcflmpoum mEmmam UNSHMDGOO 4 xmmm
.mpfluoHso EDHUOm mm.o :uHB pmpdam paw Hmm mnu mo con m m
m on pmflammm mmz mammHm mo HE wco .mpHHoHno ESHGOmemd
SH mmlw xwpmnmmm mo CESHOU m :0 mammam amass mo wnmmum
IoumEouno an Umcflmuno cumuumm SOHDSHM|I.H musmflm

45

BZHDAhhfl hO A!
0.mH 0.NH 0.0

0.0 0.m

 

_ _ _

 

 

[\7

 

 

0.0

«.0

v.0

\D
O
O

 

'EDHVEHOSEV

mu 083

46

Figure 2.--Gel filtration of human serum
on Sephadex G-200. Elution pattern was obtained
by chromatography of human serum on 2.5 x 100
cm column of Sephadex G-200 (17 g) in 0.1 M tris-
hydrochloric acid buffer, pH 8.0, containing 0.2
M sodium chloride. The middle peak, Fraction II,
contained angiotensin converting enzyme. The
fractions between the arrows were pooled, dialyzed
and 1y0philized.

 

47

 

1.4 )-

1.2-
”Fraction II“

280 nm
.2"
Re
1

5

.°
a:
I

 

“semen,
.°
0
l

 

 

 

0.4r-
0.2—
) l
0.0 l 1 J 1 1 I

 

 

O 5 10 15 20 25 3O 35 4O
FRACTION NUMBER

48

m @§§§9@31¥¥§?@@£¥£¥@@$5¥@¥$
grephy. The lyophilized preparation, Fraction II from
Sephadex G-200, was applied on a DEAE-Sephadex A—50 column
and eluted using a stepwise ionic strength gradient. The
effluent plot is shown in Figure 3, indicating four major
protein fractions. The peak that was eluted with 0.2 M
sodium chloride contained converting enzyme which hydrolyzed
angiotensin I, Hippuryl-Gly-Gly, Phe-Gly-Gly, Phe-Gly-Gly-
Phe, Phe-Gly—Phe-Gly and Gly-His-Gly. The natural sub-
strate, angiotensin I, was used in the assay of the 1yophil-
ized Fraction III preparation and the amount of His-Leu
enzymatically released was quantitated fluorometrically. As
seen in Figure 4, as little as 4.0 nmoles of His-Len could
be readily detected. Greater sensitivity (100-1000 pmoles/
m1) could be easily attained using the microspectrophoto-
fluorometric procedure.

Diagramatic representations of polyacrylamide gel
electrophoresis patterns of Fraction II, the four DEAE-
Sephadex fractions, and porcine plasma preparation (Figure
5) indicate that additional purification techniques are
necessary if purer preparations are desired.

Table I and Scheme II outline the partial purifica-
tion of human angiotensin converting enzyme from 2 m1 of
serum, using angiotensin I as substrate to monitor the
progress of purification by fluorometric assay. About a

12.5 fold purification was achieved after gel filtration and

49

.0 0mm .eeaaoo 50
mm x H .m.e mm .umeesn mumeemoem 2 No.0 ea meeuoaeo
ESH©Om mo ucmflpmum numcmuum Decca mmfl3mmum m mcwms
om|4 xmpmnmmmum<ma so cmnmmumoumeouno cam .mpfluoHno
Esflpom z H.o mcflcflmucoo .mm.h mm .uowmsn mumsmmonm
z mo.o 00 HS m.o cw pm>a0mmflo mums Geopoum m0 m5 munom
.Ednmm amass .Aoomuu xwpmnmmm coauomum mappflzv

HH coauomum mo EMHmMHU ucmsammmal.m musmfim

50

 

00H 05H N

HUM! I m.0

____

gnu—MM .m0 5-

on." ONH 0.: 00H R§ 0m 0w on on 0H

0

 

J::____j1wq_:a1_j_

cHHH ceauouuha

UM
H062 S «.0 w* H.0=—-

____r—___

_

_

 

__~L__

_

c

 

HUM—a. 2 .mrc .0

_

1

 

 

0.0

«.0

¢.e

0.0

0..n

Ntn

«tn

0...”

0...”

m 08: *aomaosav

51

Figure 4.--Standard curve for the
fluorescent determination of histidyl-leucine
after reaction with o-phthalaldehyde. Activa-
tion and fluorescence monochromators were set
at 365 nm and 495 nm, reSpectively.

RELATIVE FLUORESCENCE INTENSITY

52

 

70'-
60-
50'-

40— _

30

20

  

 

 

 

l l l l l l l l l
10 20 30 40.50 60 7O 80 90

was HISTI DYL-LEUCIN'E

 

53

.Hommm> mpouuomam

as ~.m ma .Hmemsn mumuon z mo.o “Hem ca m.m mm .med»

2 hm.o .SOHumummmum mahncm mafiuum>coo mEmMHm mcflouom 0cm

.oomlw xmpmnmom Eonw HH cofluomum .A>H I H mcofluomumv omld

xopmnmmmlmdmm mo mcumuumm mammuonmoupooam How mpaemeuom
twaom mo coaumucmmouamn vapmamumMHaul.m musmflm

54

 

 

 

J: EB

 

 

H_ _:_ __

 

 

ME; _IHD

 

 

 

HEM:

 

::

2005.5
than MIHUKOH

HH .30Hnbgh .

>H ZOHHDGKN

H HH IOHBUgh

HH 855m

H ZOHsgh

55

$-

.meeuoHeo eaeeom z H.o meecemueoo .mm.e

mm .Hmmman mumnmmonm z mo.o .OOSM w SHE\H chqmuonsm Eoum
pmmmmHoH wcHosmHlepHumHn mo mHOE: H uuHss wahucm mcHuum>coo

 

 

mm oo.a m.m m.m m.~H memnuxmemedmm

we mm.o oa He H.m oomnw xmememmm

OOH mo.o NH omH o.H mammHm
mum>oomm mE\muH:D wuH>Huo¢ Skuoum coHymoHMHHsm

w .os .mm Hmuos me pHom

 

meancm mcHuHo>coo chcmuoncd mo SOHumoHMHnsm .H mHnt

56
SCHEME II.-—Isolation of Angiotensin Converting Enzyme (ACE)

 

FRESH VENOUS BLOOD

1. Let coagulate.
.2. Centrifuge.

 

‘9 1
Serum Clot (REC , fibrin ,
' etc.)

1. Apply 2 ml to Sephadex G-200 equili-
brated with 0.1 M Tris-HCl, pH 8.0, 0.2 M NaCl.
2. Elute with same buffer.

 

    
  

 

1. Combine fractions of second peak,
dialyze e 4'C vs. H 0: 1yophilize.
ACE 2. Apply to DEAE-Sephagex 8 elute by
' stepwise ionic strength gradient .
0.075 l0.ld0.20'0.30 with NaCl in 0.02 M phosphate
buffer, pH 6.8 ACE eluted with

0.2 M NaCl.

AbsorbanCe Q 280- gm:

 

 

F

Peak III fractions are combined,
dialyzed, and lyophilized. 12.5
ml of effluent x purification.

 

 

 

57

ion exchange chromatography.

m Mygs§£9we9m3¥we939592E99£
Frecpiene ef_Hemen_Pleemef By the ninhydrin procedure, Cohn
Fractions I and IV-l catalyzed the hydrolysis of Hippuryl-
Gly-Gly to give a rate of 0.55 and 0.45 nmoles Gly—Gly/min/
mg of protein. Fraction V had weak activity but the repro-
ducibility was poor. No activity could be detected in
Fractions II + III and IV—4.

The fluorometric assay using Gly-His-Gly as substrate
confirmed that converting enzyme activity occurred primarily
in Fraction IV—l. Further proof that this activity was attri-
butable to angiotensin converting enzyme was obtained from
experiments using known inhibitors of the enzyme. Thus the
activity in Fraction IV-l was inhibited by inorganic pyro-

phosphate and 8-hydroxyquinoline and unaffected by diisoprop-

ylfluorophosphate.

2. Cofactor Requirements for Angiotensin Converting Enzyme
Activity

a. QizaiepteMete1_Cetiop. Since angiotensin converting
enzyme required calcium or a divalent metal ion for activity,
it was found also that incubation with 1.0 mM ethylenediamine-
tetraacetate produced inhibition with the new substrate
analogs used in this study. Figures 6 and 7 graphically,
depict that the activity was prOportional to the quantity

of enzyme and that the hydrolysis of four different Substrates

58

Figure 6.--Effect of EDTA (lower curves)
on the hydrolysis of Phe-Gly-Gly (Figure 6A) and
Hippuryl-Gly-Gly (Figure 6B). The enzyme in 0.5 M
phosphate buffer, pH 7.25 containing 0.1 M sodium
chloride, was preincubated overnight at 4°C with 1.0
mM EDTA. The experiment was initiated by the addi-.
tion of substrate. The controls (upper curves)
contained no inhibitor.

59

 

 

 

 

 

 

 

 

 

 

MG CONVERTING ENZYME

A
5 0.5r- l I I l
o 0.4— _
p
U) P66
0
a 0.2" o '-
O
_ O/ .—
g 0.1
a X? 1 +513“ 1
0.0
0 0.2 0.4' 0.6 0.8
MG CONVERTING ENZYME
B
a 0.5— -—1
2 004'— '1
‘3
‘ 0.3 - -
50.2 — —l
a 0.1 F- -‘
0.0 A l 1 1
0.2 0.4 0.6 0.8

60

Figure 7.--Effect of EDTA (lower curves)
on the hydrolysis of Phe—Gly-Phe-Gly (Figure 7A)
and Phe-Gly-Gly-Phe (Figure 7B). The enzyme, in
0.05 M phOSphate buffer, pH 7.25, containing 0.1
_M sodium chloride was preincubated overnight at
4°C with 1.0 mM EDTA. Theexperiment was initi-
ated by the addition of substrate. The controls
(upper curves) contained no inhibitor.

61

 

 

 

_ _

O 0
a: cum .8382

PGPG
’9'?

0.2

0.8

0.6

0.4

 

_ _ o
.4 .3 .4 1. .u

0 0 O

0 0 0

MG CONVERTING ENZYME

 

 

 

0.4 0.6 0.8

0.2

 

 

O

a _ fl. _
_A
,1
i
o In
D.
m
P
p _ _ r
.A .J .2 1. O
000.0.0.

5: Ohm .mofldflmomfld

MG CONVERTING ENZYME

62

was inhibited by EDTA. Gly-His-Gly (not shown in Figures 6

and 7) produced similar results.

b. MopoyalepteApien; Studies have shown that chloride
is also necessary for maximum converting enzyme activity and
as seen in Figures 8 and 9 deletion of chloride ions dimin-
ished enzyme activity but not to the extent that had been
reported for angiotensin I. The enzymatic effect of angio-
tensin converting enzyme on the hydrolysis of Phe-Gly-Gly

and Gly-His-Gly (not shown) was independent of chloride ion

concentration.

3. Substrate Analogs for.Angiotensin Converting_Enzyme. The

 

ninhydrin assay was used to follow the hydrolysis of some of
the substrate analogs: the sensitivity and linearity of the
method using Gly-Gly standards, may be observed in Figure 10.
With the aid of thin-layer chromatography, Gly-Gly
(Rf, 0.11) and Phe (Rf, 0.49) from Phe-Gly-Gly (Rf, 0.30);
Phe-Gly (Rf, 0.37) from Phe-Gly-Phe-Gly (Rf, 0.49); Phe-Gly
and Gly-Phe (Rf, 0.35: isomers not resolved) from Phe-Gly-
Gly-Phe: Gly-Gly (Rf, 0.11) from Hippuryl-Gly-Gly: His-Leu
(Rf, 0.52) from angiotensin I (Rf, 0.09); and His-Gly (Rf,
0.21)and glycine (Rf, 0.38) from Gly-His-Gly were identified
as products arising from the incubation of substrates with
human and porcine plasma converting enzyme preparations. The
detection of a terminal amino histidine residue by fluores-

cence also indicated cleavage at the position indicated.

63

Figure 8.--Effect of chloride ion on the
activity of converting enzyme using substrate
analog Phe-Gly-Gly-Phe. Varying concentrations
of enzyme in 0.05 M phosphate buffer, pH 7.25,
containing 0.1 M sodium chloride (upper curve),
and without sodium chloride (lower curve), were
incubated with the substrate for 3 hours at 37°C
and the liberated products determined using the
ninhydrin procedure.

ABSOREANCE, 570 nm

64

 

1.0L.

 

0.0

PGGP

l l l
0.8 1.6 2.4

MG CONVERTING ENZYME

 

 

65

Figure 9.-—Effect of chloride ion on
the activity of converting enzyme using sub-
strate analogs Phe-Gly-Phe-Gly and Hippuryl-
Gly-Gly. Varying concentrations of enzyme in
0.5 M phosphate buffer, pH 7.25, containing
0.1 M sodium chloride (upper two curves,
+c1‘, and without sodium chloride (lower two
curves), were incubated with substrate for 3
hours at 37°C, and the liberated products
determined using the ninhydrin procedure.

66

 

 

+c1'

PGPG

1.0"-

_
3
.0

a:

_ P,
,6 .4

0 O
o5 .uogomma

+c1'

0.2-'

EGG

 

1.6 2.4 3.2

MG CONVERTING ENZYME

0.8

67

Figure 10.--G1ycy1-g1ycine standard
curve using the ninhydrin procedure for the
colorimetric assay. To varying concentrations
.of glycly-glycine contained in 1.0 ml was
added 0.5 ml of ninhydrin, and the solutions
vigorously boiled for 15 minutes and then
cooled. After adding 2.5 m1 of water to each
the absorbance was read at 580 nm.

AESOREANCE, 570 nm

68

 

 

 

T I I I d
0.6»—
0.5-
0.4F-
0.3F' C)
002?-
0.1——
0.01 I 1 1 l 1
0 20 40 60 80 100
NINOMOLES GLYCYL-GLYCINE

 

69

In search for possible new substrate analogs for
angiotensin converting enzyme, a large number of small pep-
tides was tested, and Table II A and II B Show a compilation
of all those investigated. Initial investigations involved
the ninhydrin assay, as developed by Dorer ep_el. (20) for
the determination of peptidase activity. Later studies used
the newly developed converting enzyme assays. The DEAE
preparation of plasma angiotensin converting enzyme was incu-
bated with the substrate tested and the amount of hydrolysis
was determined using spectrophotometric and fluorometric
methods. Table II B lists peptides which were not hydrolyzed

by converting enzyme.

4. Fluorometric Assay of Angiotensin Converting Enzyme Using

 

OPT and SME. o-Phthalaldehyde generally reacts with amino
acids and peptides in alkaline medium in the presence of
B-mercaptoethanol to giVe strongly fluorescing compounds
(77,116). It occurred to us that this reaction might be
suitable for the assay of converting enzyme provided the
1ytic products were fluorescent.

To determine peptides in aqueous solution, borate
buffered OPT was added to the peptide in borate buffer,
allowed to stand 5 to 25 minutes, and the fluorescence
measured. Figure 11 compares the fluorescence obtained for
Hippuryl-Gly-Gly, Phe-Gly-Phe-Gly, and their 1ytic dipep-
tides. In a typical assay procedure, 1.0 mM Phe-Gly-Phe-Gly

was mixed with converting enzyme of various concentrations

70

Table II-A.-_Substrates Tested for Dipeptidyl Carboxypep-
tidase Activity with Angiotensin Converting

 

 

 

Enzyme
Estimated* TLC
Relative for
Substrate Rate Dipeptide Investigators
Z-Phe-His-Leu' '7 1000 His-Leu Piquilloud ep_e£.
Gly-His Gly 200 His—Gly Summary e£_el,
Z-Pro-Phe-His-Leu 100 His-Leu Piquilloud e£_e£.
Angiotensin I 100 His—Leu Summary ep_e1.
Phe-Gly-Gly-Phe 40 Phe-Gly/ Summary eE_el,
Gly-Phe
Phe-Gly-Phe-Gly 20 Phe-Gly Summary e£_e£.
Phe-Gly-Gly 20 Gly-Gly Summary e£_e£.
Hippuryl—Gly-Gly 20 Gly-Gly Yang e£_el,
Hippuryl-His-Leu 9 His-Leu' Cushman and Cheung
Boo-Phe (N02)-Phe-G1y 8 Phe-Gly Yang ep_el.
Z-Pro-Leu-Gly 5 Leu-Gly Yang ep_e£.
Z-Gly-Gly-Gly 5 Gly-Gly Summary e£_e$.
Ac-Ser-Pro-Phe-Arg 3 Phe-Arg Yang ep;e£.l
Bradykinin (Not reported)Phe-Arg Yang EE_El'

 

*Angiotensin I as 100

71

Table II-B.--Peptides Unhydrolyzed by Angiotensin Con-

verting Enzyme

 

 

 

 

 

Peptide Investigators
Hippuryl-His Cushman and Cheung
Hippuryl-Leu Cushman and Cheung
Hippuryl-Gly Cushman and Cheung
t-Boc-His-Leu Cushman and Cheung
Z-Gly-Gly Yang et a1.
Hippuryl-Phe-NH2 Yang'epeel.
Z-Pro-Phe-NH2 Yang‘et a1.
t-Boc-Phe(N02)-Phe Yang et a1.
Ala-Gly-Gly Summary 22.21-
Phe-Phe-Phe-Phe Summary ep_el.
Z-Gly-Gly-Phe Summary ep_el,
Z-Gly-Gly-Leu Summary ep_e£.
t-Boc-(Pro)4 Summary ep_el,

 

72

Figure ll.--Fluorometric determination
of peptide hydrolysis using o-phthalaldehyde and
B-mercaptoethanol. Standard solutions of Phe-
Gly-Phe-Gly, Hippuryl-Gly-Gly, and their 1ytic
dipeptides were prepared, and their relative
fluorescence intensities determined after
reaction with o-phthalaldehyde and B-mercapto-
ethanolin 0.05 M borate buffer, pH 9.5.

 

RELATIVE FLUORESCENCE INTENSITY

73

 

 

 

 

l T T l l l I T l
70 __
PG C)
G
60-— C)
/A
5° " ° reps
,a””"13
40'—
30 — 0/0
20'-
S; HGG
__.o...7
0 OT 1 l l l J l l

 

 

O 10 20 3O 4O 50 60 7O 80 90 100
NANOMOLES OF PEPTIDE

74

and incubated for 3 hours. An aliquot of each was added to
buffered OPT and the fluorescence intensities were measured
in a spectrophotofluorometer. Figure 12 shows the change in
fluorescence as a function of enzyme concentration using

DEAE-Sephadex "Fraction III" as the source.

5. §pectrophotometric and Spectrophotofluorometric Assays

 

of Angiotensin Converting Enzyme Using Glycyl-Histidyl-

 

Glycine As Substrate

 

a. gopverpipg_Epzyme eseay{in_Perpie1ly_Pprified‘Prepere-
Eiene‘ef_Hpmen_Serme The spectrophotometric and fluoro-
metric methods developed for the assay of converting enzyme
were all quite suitable since the substrates were commer-
cially available. The substrate Gly-His-Gly was particularly
attractive since its hydrolysis was catalyzed at a rate equal
to twice that of the natural substrate angiotensin I (Table
II A), thus shortening the incubation times. In addition,
one had the option of selecting either a spectrophotometric
or Spectrofluorometric procedure, or if desired, they could
be run sequentially.

The production of a yellow color resulting from the
reaction under alkaline conditions of OPT with the imidazole
group of histidyl peptides allowed spectrOphotometric and
hence quantitative determination of His-Gly.

Proteins in the enzyme preparation also reacted to pro-

duce a yellow chromophore, but the blank values were low.

’5.

75

Figure 12.--Activity of human converting
enzyme as a function of protein concentration. The
enzyme, partially purified on Sephadex G-200 and
subfractionated on DEAE-Sephadex A—50, was incu-
bated at 37°C in phosphate buffer, pH 7.25, con-
taining 0.1 M sodium chloride, with 500 nmoles of
Phe-Gly-Phe-Gly in total volume of 1.0 m1.
Phenylalanyl-glycine was measured by the fluoro-
genic reaction of the dipeptide with o-phthal-
aldehyde in the presence of Ekmercaptoethanol.

 

 

 

 

*4 ‘ fiL“—.__Afi

76

 

 

r
,O
,6
MB

_ r _ C r
0 0 0 0 0
5 4 3 2 1
Hmzmezn mozmomumopqm m>Ha¢qmm

 

0.20

0.15

MG ENZYME

0.05

77

Figure 13 shows the absorption spectra of the various com-
ponents of the reaction mixture. At 420 nm, the absorbance
of His-Gly was quite satisfactory, and the absorbance from
the protein chromophore was minimized. The color due to the
His-Gly-OPT adduct reached maximum intensity from 5 to 10
minutes after addition of OPT; however, the reaction with
proteins was not complete until 30 to 40 minutes after the
addition of OPT reagent. Readings were made in this time
frame, since the color slowly began to decay after 40 minutes.
The fluorometric method also measured the amount of His-Gly
liberated by angiotensin converting enzyme; however, for
analysis OPT was reacted with the imidazole moiety in the
presence of 2 N sodium hydroxide, and then the fluorescence
was measured under acid conditions. Since proteins did not
form a fluorophore under these conditions, there was no back-
ground fluorescence.

The hydrolysis of Gly-His-Gly was found to be pro-
portional to both the time of incubation at fixed enzyme and
substrate concentrations and to the enzyme concentration
(Figures 14 A and 14 B). Figure 15 depicts the hydrolysis
of Gly-His-Gly as a function of substrate concentration.

The Lineweaver-Burk plot inset indicated a Michaelis con-
stant of 6.0 x 10-4pm There was a close correlation between
the percent hydrolysis as shown by spectrOphotometric and
Spectrofluorometric methods (Figure 16).

Incubation of the enzyme with Gly-His-Gly at various

-

_ '4'“ 4

h;

78

Figure l3.--Protein (curve B) and histidyl-
glycine (curve A) spectral absorbance curves after
reaction with OPT. o-Phthalaldehyde was reacted
with a preparation of angiotensin converting enzyme
(0.5 mg/ml, Fraction II) and a mixture of hydrolytic
products of Gly-His-Gly, 0.10 mM His-Gly and 0.10 mM
Gly. The absorbance of the resulting chromophores
were determined as a function of wavelength.

 

 

 

0.8

0.6

ABSOREANCE

0.2

 

 

 

l

 

380

400

420

440

WAVELENGTH (run)

460

 

1..
. 1 PIN.

80

Figure l4.--Hydrolysis of glycyl-histidyl-
glycine by plasma angiotensin converting enzyme.
For the hydrolysis of Gly-His-Gly as a function
of time (Figure 14A) the substrate was incubated
with varying concentrations of enzyme at 37°C.
Aliquots were removed at the times indicated and
the histidyl-glycine quantitated spectrophoto-
metrically. For the hydrolysis of substrate as
a function of enzyme concentration((Figure 14B)
the substrate was incubated at 37°C for 90
minutes with varying concentrations of enzyme and
the liberated dipeptide quantitated, as above,
using o-phthalaldehyde.

81

 

 

120

60

 

HHUImHN mchzz

MINUTES

 

 

_ _
n

MRUImHm mmHQSZ

_

4.

40

20

 

MG ENZYME

82

Figure 15.--The effect of substrate con-
centration on the activity of angiotensin con-
verting enzyme of plasma, measured by the rate
of hydrolysis of glycyl-histidyl-glycine. G1 -
His-Gly, 25 - 200 nmoles, was incubated at 37 C
with 0.10 mg of DEAE Fraction III enzyme prepara-
tion in 1.0 ml of 0.05 M phosphate buffer, pH
7.25, containing 0.1 M sodium chloride. The
liberated histidyl-glycine was determined fluoro-
metrically using o-phthalaldehyde. The inset
shows a Lineweaver-Burk plot of the same data.

 

83

 

 

 

 

 

l l I l
C)
40 -—
w
E? 30 -
0)
a /0 1.0“-
U)
o .1. -- _.
g 20» //, v
o 005')”
10 /// -- -
C)
l 1 1 1 2;
-20 -10 0 10 20 s
l l l l
50 100 150 200

NMOLES GLY-HIS-GLY/ML

 

84

Il'l' i
Iv! II")
11.1. II!!!

 

.mpmsmp

IHmnuzmIo mchs wHHMOHuumEouosHm paw >HHmoHuumEHHoHoo

UmumuHucmsv mumz wcHomeIHmpHumHE pmumquHH on“

NO muongHm paw .wHOImHmIMHO SE m.o SHHS UmumnsoSH

mp3 mpHHoHno ESHUOm z H.o mchHmwcoo .mm.h mm

.ummmsn mpmnmmonm z mo.o mo HE o.H cH AMEmMHm amass

.HH coHuomHm mo 08 o.m1mm.Hv coHumumdmnm mewucm mnh

.SOHumuucmocoo mewucm mo GOHUUSSM H mm msmucm mcHuuo>

Icoo chSODOHmcm mom mmmmmm OHHumEouosHmouuoomm paw
OHHDmEouozmouuommm mo SOmHHmmEOOII.wH mHDmHm

85

SEN!” g
m.N

 

 

 

 

0m

00H

S ISA'IONGLH %

86

temperatures indicated that the temperature of maximum hydro-
lysis was 43°. Similarly, incubation in 0.05 M phOSphate
buffer, with the pH adjusted between 6 and 9, showed an
optimum at pH 7.25. This agrees with the value of pH 7.25
reported by Huggins ep_el. (56) for plasma angiotensin con-
verting enzyme.

Human plasma converting enzyme, as revealed by amido
black staining, was a rapidly moving protein component during
polyacrylamide gel electrophoresis. Table III shows con-
verting enzyme activity as a function of the distance of
migration along the gel. Both Gly-His-Gly and angiotensin I
were hydrolyzed by the same regions. The activity at the
origin is associated with protein which failed to enter the
gel. The faintness of the bands associated with the active
regions confirms the low levels of plasma converting enzyme
as was found by H-J Lee e£_ei. (105).

Since angiotensin converting enzyme required a
divalent metal ion for activity, incubation with 0.20 mM EDTA
produced 100% inhibition. As previously reported (56), the
sulfhydryl agent, p-chloromercuribenzoate, also was a potent
inhibitor of the enzyme (Table IV).

Pyrophosphate was found to be comparable to 8-hydroxy-
quinoline as an inhibitor. Aldosterone, an important
mineralcorticoid regulator of kidney function, was found to
be a good inhibitor (more than 50%) at relatively low con-

centrations.

87

Table III.--Hydrolysis of Gly—His—Gly and Angiotensin I by
Polyacrylamide Gel Electrophoresis Fractions

 

 

 

Converting
Enzyme
Activitya
.Substrate
Section . Intensity of
(cm from origin) AI GHG Bands Present
0.0 - 0.5 +++ +++ _
0.5 - 1.0 — + +
1.0 - 1.5 — _ +++
1.5 - 2.0 - _ 1+
2.0 - 2.5 _ - -
2.5 - 3.0 - _ +1
4.0 - 4.5 +++ ++++ ++

.?fi;_

 

aComparative activity of sections:

4+ = good hydrolysis

3+ = moderate hydrolysis
2+ = weak hydrolysis

1+ = slight hydrolysis

no hydrolysis

bSee Figure 5, p. 53

 

88

Table IV.--Inhibition of Human Plasma Angiotensin
Converting Enzyme

 

 

 

 

 

 

 

Pre- Percent Inhibition
Compound incubation 0.5 1.0 1.5 2.0 2.5 (mM)
EDTA * - 100 100 100 100 100
8-Hydroxyquinoline - 20 43 61 80 96
p-Chloromercuri- 40 70 100 100 100
benzoate -
N-Ethylmaleimide 15 minutes, 9 19 28 38 47
' 37°C
Iodoacetic acid 15 minutes, 0 0 O 1 3
‘ 37°C
Pyrophosphate* - 30 55 75 91 100
Angiotensin II* - 23 49 67 90 100
5.0 10.0 15.0 20.0 (p1)
Adrenal glands 15 minutes, l9 l4 - -
extract 37°C
.2; Jararaca venom 15 minutes, 21 21 - -
extract* 37°C
Human urinary - 70 84 87 90
factor* ,
Plasma inhibiting - 55 79 100 100
factor*
0.075 0.15 (mM)
Aldosterone 15 minutes, 52 64

37°C

 

+ 100 nmoles of Gly-His-Gly were incubated for 90 minutes with
0.5 mg of Sephadex G-200 fractionated human plasma angiotensin
converting enzyme preparation and the indicated amount of
inhibitor in a total volume of 0.3 ml. The liberated His-Gly
was determined spectrophotometrically.

* Confirmed using angiotensin I as the substrate. Angioten—
sin I (35 nmoles) was incubated as described above and the
liberated His-Leu was determined spectrophotometrically.

 

89

A crude extract of Bothrops Jararaca venom, a speci-
fic inhibitor of angiotensin converting enzyme (108,110),
was found to inhibit the enzyme when either angiotensin I or
Gly-His-Gly was the substrate. The acetone extract of dog
adrenal glands, presumably containing adrenal steroids, gave
comparable inhibition of Gly-His-Gly hydrolysis.

Angiotensin II, the natural octapeptide product of
angiotensin I hydrolysis, caused 49% inhibition of Gly-His-
Gly hydrolysis at 1.0 mM concentration. Product inhibition
was confirmed using the natural substrate angiotensin I.

The low activity in partially purified preparations
of angiotensin converting enzyme prompted a look for effectors
in human plasma and urine. Following dialysis of both urine
and of plasma and passage of the dialysate over a cation
exchange column (Dowex 50W-X8), the components bound to the
resin were found to inhibit plasma converting enzyme (e.g.
100% inhibition by 15 pl of Plasma Inhibiting Factor, Table
IV). The active component could not be extracted into ether
or benzene and was only slightly soluble in chloroform/
methanol (3/2). Human Urinary Factor was also bound by an
anion exchange resin (Dowex 1-X8). Acid hydrolysis of either
the urine or plasma preparations did not affect their inhi-

bitory ability.

0 yfigyyegygymfiymeyyxyfigfiyws
The spectrophotometric assay developed in this study was a

rapid, specific and sensitive method for the determination of

 

90

angiotensin converting enzyme activity in small amounts of
plasma or serum. As shown in Figure 17, the percent hydro-
lysis was a linear function of time at fixed plasma (0.10 ml)
and substrate concentrations (100 nmoles), provided the time
of incubation did not exceed 30 minutes. As is characteris-_
tic of enzyme catalyzed reactions, the percent hydrolysis of
Gly-His-Gly was proportional to the plasma (angiOtensin con-
verting enzyme) concentration at a fixed substrate concentra-
tion (Figure 18). A potent peptidase inhibitor, diisopropyl-
fluorophOSphate, incorporated into the incubation mixture had
no effect on the hydrolytic reaction.

Since His-Gly, the product of Gly-His-Gly hydrolysis,
as well as the plasma proteins, reacted with OPT, the assay
procedure required the preparation of a protein free filtrate
prior to addition of the OPT reagent. Further, the amber
color of serum or plasma, similar in color to the OPT-His-Gly
chromophore, absorbed blue light (420 nm) strongly. Protein
precipitation with tungstic acid reduced the background color
to negligible levels, permitting the measurement of nanomole
quantities of His-Gly.

Inhibition studies (Table IV) and polyacrylamide gel
electrophoresis experiments (Table III) using purified human
serum preparations demonstrated that Gly-His-Gly was a suit-
able substrate analog for the assay of angiotensin converting
enzyme. Thus substrate modification, small peptide assays,

and inhibition eXperiments were undertaken to determine

 

91

 

.mmEHu topmoHch mnu
pm 800 nqu mHHmoHuumEODOEQouuommm pmcHEHmump mp3
Seonmam oooaaoofle one one .ooapoaeo aaaooa z H.o
maeoeooaoo .mm.e ma .uoeeon mooeomoem z mo.o aH Haw
Imemusao Lo moeoea ooa the; oooaoooaa was lee oa.ov
HEmMHm 004 .oEHp mo coHuoS9m m mm mHOImHmImHO mo
mHm>Houphn pmNmHMDmo mammHm 00411.>H musmHm

92

 

0mg”!

 

 

0v

SI SX'IONCILH %

 

93

 

.Bmo nuHs MHHMOHHumEouonmoueowmm
pmcHEHmump HHOImHm on» can .mpHHoHno EsHpom z OH.O
oaHaHauaoo .m~.e ma .uoemso mooaamoem 2 mo.o oH ohm
um mmuscHE OH wow >HOImHmI>HO EDHB pmumnsocw mmS
.HmHo>Huommmmu mmHouHo paw mmHOSMHuuv who usocuH3
paw EDHB .mEmmHm mo ucsoam pmumoHch 0:9 .COHumnu
Iaoocoo mEmmHm mo SOHuocsm H mm memmcm OCHuHm>coo
GHmcmuonsm amass mo muH>Huo<II.OH mHSmHm

94

 

 

 

 

 

<10! 1 l
L.
. J. L J.

S ISA'IONGLH %

0.15

0.10

0.050

ML PLASMA

95

whether the catalyzed hydrolysis of Gly-His—Gly by plasma
was due to angiotensin converting enzyme or plasma peptidases.

Since it was reported that Hippuryl-His-Leu (106)
and Hippuryl—Gly-Gly (107) were catalytically hydrolyzed by
converting enzyme preparations, we acetylated the amino
group of Gly-His-Gly in order to preclude aminopeptidase
catalyzed hydrolysis of the substrate. N—acetyl-Gly-His-Gly
was hydrolyzed at an appreciably reduced rate in both the
purified angiotensin converting enzyme preparation and in
plasma (Table V).

The effect of esterification of the natural sub-
strate, angiotensin I, was not reported, although the enzyme
was termed a "dipeptidyl carboxypeptidase" (107). The ethyl
esters of both Gly-His-Gly and angiotensin I were found to
be hydrolyzed at rates approximately equal to those of the
unesterified substrates (Table V).

If the hydrolysis of Gly-His-Gly by the partially
purified preparation of human serum comverting enzyme and
by plasma was catalyzed by the same enzyme, then both sources
should produce comparable results when incubated with small
peptides and substrate analogs. Table V lists the small
peptides that were tested and semiquantitatively tabulates
the amount of hydrolysis observed with plasma and the par-
tially purified converting enzyme preparation. The partially
purified preparation of converting enzyme and plasma cat-

alyzed the hydrolysis of the same substrate analogs.

96

Table V.--Effect of Angiotensin Converting Enzyme Prepara-

tions and ACD Plasma on Small Peptides

 

Porcine Plasma

 

Converting

Enzyme Human

Peptide Preparation Plasma
Gly-His-Gly +++++ +++++
Gly-His-Gly, ethyl ester +++++ +++++
Phe-Gly—Phe-Gly ++++ ++++
Phe-Gly—Gly-Phe ++++ +++++
Hippuryl-Gly-Gly ++ 0
N-acetyl Gly-His-Gly + +
Z-Gly-Gly-Gly + ++
Phe-Phe-Phe-Phe 0 +
Z-Gly-Gly-Phe 0 0
Z-Gly-Gly-Leu 0 0
t-Boc-Pro-Pro-Pro-Pro 0 0

Angiotensin I

Angiotensin I, ethyl ester

Gly-His-Gly

Gly-His-Gly, ethyl ester

N-acetyl Gly-His-Gly

Nanomoles of peptide hydrolyzed,

per 90 min*

 

19
18
40
37

5

per 10 min

 

20
40
38

4

 

*Fraction II of Human Serum,

5+
4+
2+
1+
0

good hydrolysis
moderate hydrolysis
weak hydrolysis
Slight hydrolysis
no hydrolysis

Sephadex G-200 preparation

97

Inhibition eXperiments provided additional proof
that the Gly-His-Gly assay was measuring angiotensin con-
verting enzyme activity in plasma. As is shown in Table VI,
fluoride, citrate, oxalate and diisopropylfluorophosphate,
compounds known to inhibit peptidases, had little effect on
the enzyme catalyzed hydrolysis of Gly-His-Gly, while inhi-
bitors of converting enzyme (Table IV, VI) produced identi-
cal effects in the purified angiotensin converting enzyme

preparation and in plasma.

B. FINDINGS ON THE MECHANISM OF THE FLUOROGENIC REACTION OF
AMINO ACIDS, PEPTIDES AND PROTEINS WITH O-PHTHALALDEHYDE AND

B-MERCAPTOETHANOL

1. Ulpreviolep_epecprel_Cpapgese. As noted from Figure 19,
o-phthalaldehyde and B-mercaptoethanol must initially inter—
act for a strongly fluorescing product to develop upon the
addition of amino acid, peptide or protein. Thus the very
prominent transition at 200 nm, attributable to the carbonyl
moiety of OPT, is reduced and a new weak transition at 340
nm appears when BME reacts with OPT. Upon the addition of
glycine to form the fluorescent adduct, an intense transi-
tion is evidenced in the 225 nm region due to the formation
of an imine group or a charge transfer band. It is accom-
panied by enhancement of the transition in the 340 nm region
upon binding of the amino acid to OPT.

The OPT-BME adduct showed intense fluorescence

98

Table VI.--Effect of Compounds and Extracts on Angiotensin
Converting Enzyme Activity in Human Plasma and

Serum Preparation

 

 

 

 

 

 

Compounds Fraction II* ACD
or Preparation Plasma
ExtractS- 1.0 2.0 1.0 2.0 (mM)
EDTA +++++ +++++ +++ ++++
p-Chloromercuribenzoate ++++ +++++ +++++ +++++
8-Hydroxyquinoline +++ ++++ +++ ++++
Angiotensin II +++ ++++ +++ +++
Fluoride 0 0 0 0
Citrate 0 0 0 0
Oxalate 0 0 0 O
5.0 10.0 5.0 10.0 (ulrt
Diisopropylfluorophosphate O 0 + +
Bothrops Jararaca Venom Extract + ++ + ++
Adrenal Gland Extract - ++ - +++
*Fraction II of Human Serum, Sephadex G-200 preparation
+ Aliquots were preincubated with Fraction II or plasma prior

to initiation of the enzymatic reaction.
5+ = almost total inhibition

4+ = strong inhibition

3+ = moderate inhibition

1+ = slight inhibition

0 = no inhibition

99

.mcHomHm + Hocmsumoummoumalm + mpwnmp

-HoHoeueauo loo aoaeosHo + mosaooHaHaeoamIo luv

“Hocmnumoummoumfilm + m©>£opHmHmnu£mIo Amy “mpmzmp

IHMHmnunmlo Adv .mHm>HuoQOmu S IOH x O.m paw

. IOH x mv.H .mIOH x OOH mumB maHome OSMHocmspm

cummonEIm .mpmnmprHmnusmIo OSH>H0>SH coHuommH
mnu mo muuommm umH0H>muuHDII.mH musmHm

100

 

 

 

 

SONVBNOSEV

 

 

 

101

before removal of the methanol solvent; however, slow evapora-
tion and subsequent redissolving in borate buffer of pH 11.0
caused a reduction in fluorescence. This was in agreement
with other observations that fluorescence in dilute solution
decays after several hours. The adducts that were analyzed
were chromatographically pure by thin-layer chromatography,
thus indicating little side product formation or degradation
during synthesis. If the OPT—BME glycine adduct was chroma-
tographed by thin-layer chromatography in acidic medium,
several degradation products were visible including one nin-
hydrin positive component. The ultraviolet spectrum of the

crystalline OPT-BME-glycine adduct is shown in Figure 20.

2. Infrared Spectroscopy. Infrared spectroscopy of the

 

adducts Showed that one of the carbonyl groups of OPT dis-
appeared upon the addition of glycine, since there was a
reduction in intensity of the band attributable to the C-H
stretch on the aldehyde group. Further, the OPT-glycine
adduct showed no free amino group. The OPT-BME-glycine
adduct showed additional change in the carbonyl region and
further reduction of the C-H stretch. No S-H bands could
be detected. The imine region was obscured by the Nujol

preparation.

3. Nuclear Magnetic Resonance Spectroscopy. The NMR
Spectra of the adducts are shown in Figure 21. The OPT-

glycine adduct showed four separate proton environments. The

102

.uoSppm OSHUHHO 0cm
HocmsumoummoumEIm .mpmnmpHmHmsunmlo mcHHHmumwuo
030 m0 muuommm umH0H>muuHDII.Om musmHm

103

 

 

300

 

 

 

0.6

0.2 ~—

340

260

220

WAVELENGTH (mu)

104

 

 

.mcHomHm + Hocmcumoummouwfilm + moazmpHMHmsunmIo ADO

.mcHomHm + mO>SOUHMHmaunmuo HOV .Hocmsumoummoumslm

Amy .mpwnmpHmHmnunmIo A40 mo muuommm monocommu
OHDoSmmE ummHosc UHDMEMHOMHQII.HN musmHm

CL Emma guano

 

 

 

 

 

 

 

 

 

m m e o a v. m a ..
_ _ . _ _ _ . . H.
.3 5: .2 :ll
_ a
a 5 ____| ___ .
l
.————— —__..8:._.o ._ _.I._— —:1 = U
__II..|II._.:.._:.:. :2:
a

Il.._____l____l ______...._..__________

106

one aldehyde proton and the four phenylic protons were seen
at t less than 6. The broad band at t = 7 was characteristic
of the N-H bond. The doublet which overlayed this broad sig-
nal may be interpreted as belonging to the proton on the
alpha carbon of glycine. The integrated area attributed to
the phenylic protons was set at four, and the rest of the
areas evaluated relative to the phenylic region. The number
of protons indicated that one molecule of OPT was bound to
one molecule of glycine; further, one molecule of BME was
bound to this adduct when SME was a co-reagent. The OPT-BME,
glycine adduct showed seven proton environments, and loss of
the Signal due to the aldehyde moiety. The broad band at

t = 5 was due to two alcohol groups.

4. Effect of Carboxyl Group Proximity on Fluorescence.

Table VII shows fluorescence observed under a variety of con-
ditions. Thus it may be seen that the carboxyl group has
little effect on the fluorescence of the product, as the
number of methylene groups which separated it from the
aromatic ring were increased. In addition, glycine ethyl

ester fluoresced intensely.

5. Influence of Oxygepl Reducing Agents and Alkylating
AgentS'on F1UoresCenCe. Since BME can reduce dissolved
oxygen in solution, the effects of dissolved oxygen on
fluorescence were tested. As can be observed from Table VII,

the presence or absence of dissolved oxygen had no effect on

107

Table VII.--Equimolar Fluorescence Intensities of Products
Derived From Selected Amino Compounds After
Reaction With OPT, 3MB, or Analogs

 

 

Reactantsb FluorescenceC
Glycine + OPT + 3MB 100
Ethyl Glycinate + OPT + BME 100
Glycine + OPT + BME + PCMB - 87
Glycine + OPT + DTT 50
Glycine + OPT + NaBH4 35
Glycine + OPT + NaHSO3 6
Glycine + OPT + BME + 02 100
Glycine + OPT + N2 6
Glycine + OPT 6
OPT + BME (blank) 2
Asparagine + HMB + 8MB 0.3
Asparagine + HMB + NaBH4 0.2
Asparagine + HMB 0
HMB + 8MB (blank) 0
Asparagine + DMB + 8MB 0
Asparagine + DMB 0
DMB + BME (blank) 0
B-Alanine + OPT + 8MB 98
Phenylalanine + OPT + BME 150
(Phe) + OPT + BME 14
Gly-61y + OPT + BME 60
Hipp-Gly-Gly + OPT + BME 3
Phe-Gly-Gly + OPT + 6MB 18
Cysteine + OPT 15
Homocysteine + OPT 25
GSH + OPT 1,180
AMP + OPT + BME 3
BSA + OPT 810
BSA + OPT + BME 10,700

 

aExcitation wavelength 340 nm; fluorescence at
455 nm.

bConcentrations Of reactants: amino acids, peptides,
nucleotide or proteins were 20 11M; OPT, HMB, or DMB were 0.187
mM; and SME, NaHSO , NaHBH4, DTT or PCMB were 1.45 mM in 0.05
M borate buffer of pH 9.5.

c
Fluorescence spectrophotometer readings were rela-
tive to the glycine fluorOphor 100 fluorescent units.

108

fluorescence. Sodium-borohydride substituted effectively for
8MB, while sodium bisulfitewas ineffective in producing
fluorescence. p-Mercuribenzoate did not inhibit production
of fluorescence until after at least two moles of p-mercuri-

benzoate were added per mole of SHE.

6. Postulated Intermediates. A postulated intermediate,

 

o-hydroxymethylbenzaldehyde, gave very poor fluorescence
when substituted for OPT and reacted with glycine. NO
fluorescence occurred at all with HMB when ENE was deleted.
Dihydroxymethylbenzene produced no fluorescence with glycine

when reacted with or without reducing agents.

7. Fluorescence of Peptides. Di-, tri—, and tetra peptides

 

indicated in Table VII reacted with buffered OPT-fiME reagent
to give fluorescent products. The relative fluorescence of
a few peptides are compared in Table VII and shows that the

fluorescence decreased with an increase Of peptide bonds.

8. Fluorescence of Proteins. The binding of OPT and ENE to

 

proteins was accompanied by a bright blue flourescence which
wasquite sensitive to pH changes. The pH profile of the
bovine serum albumin fluorophore is plotted in Figure 22 and
indicates a maximum fluorescence at pH greater than 9.0.
This is similar to the relationship Observed with amino
acids (77). Highly alkaline pH did not favor the formation

of the fluorescent adduct involving serum albumin.

109

.OOxHE OHO3 mucmuommu wnu EOHQ3 um mm may mo
cOHuocSm m up SHESQHM Esumm OGH>OD mo mocmommuosHm
IImOHOHHO .m.m H mm um UOEHOM coma SOMOMHO Own
THOSQOHOSHM on» menus mm mo cOHuocsm m up SHESQHO
Esumm mcH>On mo mocmommnosHMIImmHmcmHHB .cOHuomou
mocmommnosHm mo OHHmOHm mm11.~m wusmHm

110

 

 

_ _ _

 

 

80-——

_
0 0
6 4

NOZflUmMNODHh

20 L-

12

pH

111

The excitation and emission spectra for the reaction
of buffered OPT-3MB with bovine serum albumin and glycine may
be seeniJIFigures 23 and 24. Note (Figure 25) that the par-
tial quantum efficiency of buffered OPT-albumin adduct was
negligible when compared to the OPT-BME-albumin adduct in the
280 nm region.

A sensitive fluorometric method which required only
a single reagent, namely buffered OPT—8MB, was developed for
the quantitative determination of proteins. The fluores-
cence was directly proportional to the protein concentration
for several proteins as Shown in Figure 26.

A very interesting parameter of protein fluorescence
was the proportionality of the fluorescence to the molecular.
weight. Shown in Figure 27 is the fluorescence of aldolase
(Mol. wt. 158,000), bovine serum albumin (67,000), ovalbumin
(45,000), and chymotrypsinogen (25,000) as a function of
molecular weight.

The fluorescence Of the reaction was proportional to
the concentration of bound ligand in a system in which the
concentration of ligand was not in excess, thus providing a
simple and direct method for the examination of the binding
process. When bovine serum was titrated with OPT or BME,
and the fluorescence plotted as a function of the OPT or BME
concentration, rectangular hyperbolic curves were Obtained
(Figure 28). Scatchard plots (Figure 29) revealed that

55 molecules of OPT bind per molecule of albumin, and 14

112

.pcmummme mum
mcmom mSOHum> on» How pmonmEO muouomm OSHHmom paw
mcoHumuucwocoo one .Amnuommm cOHmmHEO Op mummmu
meumummv Ea ovm u me .Ec mmv u HMH .GHEDQHO Esnmm
OSH>OQ + HocmnumODQOOHOEIm + OpmnmpHmHmnunmIO HOV
«ES ovm u xOH .Ec OHv u HMH .SHEDQHO Eduwm OSH>OQ
+ mosaooHoHoaoeauo Amv Nea new u xox. .ea mmm u Hue
.cHEdnHm Edumm OSH>OQ Adv mo muuommm GOHmmHEm paw
COHumpHoxm oocmomOHOSHw OODOOHHOOII.mN OHSOHm

113

 

 

 

I

450

 

400

1
350
WAVELENGTH (nm)

 

1
300

 

 

O
In

NONEDSEEODTJ

75—-

25—-

114

Figure 24.--Fluorescence excitation and emission
spectra for (A) o-phthalaldehyde + Ekmercaptoethanol,
Afl = 450 nm, xex = 340 nm; (B) o-phthalaldehyde +
Ekmercaptoethanol + glycine, Afl = 450 nm, Xex = 340
nm (asterisk refers to emission Spectra). The concen—
trations and scaling factors employed for the various
scans are different.

 

 

FLUORESCENCE

115

80

7!

t
1' A

O, 1 l J l 4!’1-.““‘£;-.i

 

290 330 370 410 450
WAVELENGTH (nm)

490

 

116

Figure 25.--Partial quantum efficiency
of fluorescence as a function of wavelength for
(A) o-phthalaldehyde + B-mercaptoethanol + gly—
cine, (B) o-phthalaldehyde + bovine serum albumin,
(C) o-phthalaldehyde + B-mercaptoethanol + bovine
serum albumin.

117

 

z
a:
H
L)
E

ency a

1 for 0.50

gly-.

albumin,

bovine
O
.4
.c
H
64
gr
4
a.

0.25

 

 

l l l

275 300 325
VflVWELENGflEI bun)

 

 

 

118

Figure 26.-~Fluorescence intensity as a
function of concentration for several proteins
following reaction with o-phthalaldehyde and
B-mercaptoethanol. (A) aldolase, (B) BSA, (C)
chymotrypsinogen A, (D) ovalbumin.

FLUORESCENCE

119

 

 

 

 

 

it J 1
2 w 3 4

pnoTEIN CONCENTRATION (pH)

 

120

Figure 27.—-Relative fluorescence inten-
sity of equimolar solutions of several proteins
following reaction with o-phthalaldehyde and
B-mercaptoethanol. The upper curve contained
twice the amount of protein.

FLUORESCENCE

121

 

 

  
 

(0.025 ml)

llllll
6‘10 14 18

MOLECULAR WEIGHT (X 104)

 

122

.mcflocfib mpmcmonHmcucmlo no mmaoufio ou mumwmu
ommflomnm Hmzoq .mcwocfln Hocmcuwoummonmfium no mmawcm
Iflup on mnowmu Mmmflombm momma .Amm u coaumuucmocoo
«mm\cowumupcmocoo Hmov Ammaoec m.av GHEDQHM Edumm
mcfl>on paw mowcmoamamcpnmlo ou Hocmnumoummoumaam
mo mcflocfln may pom .Hocmnumoummonmelm mmmoxm mo
mocmmmum may ca AmmHoEc m.av GHESQHM Edumm w:fl>0b ou
mumgmeamamnpgmuo mo maaccfimuu.m~ musmflm

I ll. ill I." llru

123

9&0 mflAQIOZ¢Z

 

 

 

 

 

 

00H ONH om ov
_ _ e _ _ _ _
\q K1
\\\\\&Q
l|l\\5Q|\|\\\\\\\\\\MH\\ )1Q\\\\\\\
)10 II
_. _ _ _ _ _ _
com com oov 00m

88 mmqozozaz

ow
5
cc M
m
S
D
3
N
8m
om

124

.ocoo 9m0\.ocoo
u AMEmv\/ ..ocoo <mm\.ocoo BmO n Aamovxw .ucflom soap
amusumm map um xmameoo mcu ou ocmmwa mcflocwn mo oHuwH
mzu ma Kw .cHESQHm Esnmm wcfl>on ou mowchHmHmcucmlo
mo mcflocfln mcu no mmaouflo ou momma mumcwbuo ucmfln cam
mmmfiomnm Hosoq .mownmoamamcunmlo ou mcflocflnaocmnum
.Oummoumaum no mamcmauu 0» mommy mumcwcuo puma cam
mmmflomnm Amman .mbwnmoamamnunmlo on HocmcpmoummoumEIm
mo pcm GHESQHM Edumm mcH>OQ on momcwoamHmcucmuo

no mcaocfln map co uon numgoumomuu.mm magmam

125

 

 

 

 

cm on cm
a _
or
.01
O]
_ _ _
NH m w

 

Ea

 

126 '

molecules ofBME were bound per molecule of OPT. Since the
plots were linear, the sites were non-interacting (78). A
direct plot of V versus concentration of OPT (not shown)

was non-sigmoidal, also indicating non-interacting sites.

 

V. DISCUSSION

A. ACTION OF ANGIOTENSIN CONVERTING ENZYME ON SUBSTRATE

ANALOGS

The first synthetic substrates for converting enzyme,
tripeptides benzoyl-glycyl-glycyl-glycine (79) and benzoyl-
glycyl-histidyl-leucine (55), were cleaved rather slowly
compared to angiotensin I. The site of enzymic action was
the peptide bond to which glycine or histidine contributed
the NH group. A decisive advance was made by Piquilloud gt
' El. (52) who showed that shortening the peptide chain of
angiotensin I from the amino end to form Z-phenylalanyl-
histidyl-leucine does not abolish its susceptibility to
hydrolytic cleavage by converting enzyme. Piquilloud's work
was the first indication that the action of converting
enzyme on small substrate analogs is favored by the presence
of an aromatic side chain on the carbonyl side of the sensi-
tive peptide bond.

Since 1969 studies have shown that the action of
angiotensin converting enzyme on small peptides could involve
the cleavage of many kinds of peptide bonds and, with our
present work, converting enzyme can be considered an enzyme
of rather broad side chain specificity (52,55,7S,79,106,107).

The lack of inexpensive and readily obtainable sub-
strate analogs, as well as the uncertain state of the problem

127

 

128

of the enzymic action of converting enzyme, prompted us to
study the hydrolytic action of this enzyme of synthetic sub-
strates. Our first objective was to secure peptides whose
structure could be compared to those known to be hydrolyzed
by converting enzyme. Table II A incorporates the new sub-
strate analogs and shows the relative rate of hydrolysis of
various peptides, using the natural substrate, angiotensin I,
as one hundred.

Although the converting enzyme preparations used in
our substrate analog studies were not homogenous as demon-
strated by polyacrylamide gel electrOphoresis, the following
evidence strongly supports the conclusion that the conversion
of angiotensin I to angiotensin II and the hydrolysis of
Phe-Gly-Gly, Phe-Gly-Gly-Phe, Phe-Gly-Phe-Gly, hippuryl-Gly-
Gly, and Gly-His-Gly are catalyzed by the same enzyme.

First, the enzyme we have purified on its ability to cleave
substrate analogs i1; also found to have increased specific
activity with angiotensin I as substrate. Certainly Table II
demonstrates that angiotensin converting enzyme cleaves a
variety of peptides. Secondly, in agreement with Yang (107),
converting enzyme liberates glycyl-glycine from hippuryl-Gly-
Gly. The dipeptides liberated from the substrate analogs
were identified by thin-layer chromatography and quantita-
tively determined by the ninhydrin reaction, except for his-
tidyl dipeptides liberated from angiotensin I and Gly-His-
Gly. These dipeptides were reacted with o-phthalaldehyde in

alkaline solution and measured spectrophotometrically or,

If]!!!
‘ll'll‘ll

129

upon acidification, the fluorescence was measured in a
spectrofluorometer. Thirdly, since converting enzyme

requires a divalent metal ion for activity, perhaps calcium,
incubation of the new substrate analogs with EDTA and
8-hydroxyquinoline produced inhibition. Thus Figures 6 and 7
graphically depict that the activity is proportional to the
quantity of enzyme, and the hydrolysis of four different sub-
strates is inhibited by EDTA. In addition, all other studies
have shown that Gly-His-Gly hydrolysis is inhibited by the
same agents which affect the hydrolysis of angiotensin I,
especially the ethanoic extract of Bothrops Jararaca venom
(Tables IV and VI), which is a specific inhibitor of converting
enzyme (108-111). The potent inhibition by p-chloromercuri-
benzoate of angiotensin I and Gly-His-Gly hydrolysis indicates
the presence of a sulfhydryl group at the active site.

Studies have shown that chloride ions are necessary for maxi-
mum activity, and assays performed in chloride—free medium
diminished converting enzyme activity; that is, chloride ion
stimulated the hydrolysis of Phe-Gly-Gly-Phe, Phe-Gly-Phe-Gly
and hippuryl-Gly-Gly, while Gly-His-Gly and Phe-Gly-Gly hydro—
lysis was independent of the chloride ion concentration.
Finally, hydrolysis of angiotensin I and Gly-His-Gly by the
same polyacrylamide gel slices after electrophoretic separa-
tion and the absence of peptidases (104) in the DEAE-
Sephadex preparation leads to the conclusion that Gly-His-Gly
and the other substrate analogs presented here are hydrolyzed

by the same enzyme which catalyzes the cleavage of angiotensin I.

130

The plasma and urine inhibiting factors exhibited simi-
lar properties. The cationic and anionic nature of this substance
as evidenced by binding to Dowex 50 and Dowex l, and the lack
of acid hydrolysis under conditions which would destroy any
peptide, implies that the inhibitor may be an amino acid.
Lipid soluble components were eliminated since the inhibitor
is not chloroform or ether extractable. Histidine, a major
amino acid component of blood and urine, was suspected; how—
ever, inhibition of converting enzyme by 2.5 mM histidine
could not be demonstrated.

The inhibition of converting enzyme by very low con-
centrations of aldosterone (Table IV) leads one to postulate
that a feed-back control mechanism is operative between the
mineralcorticoid/renin angiotensin system. Since angiotensin
II directly stimulated the production of aldosterone by the
adrenal cortex (112,113) the subsequent increased levels of
aldosterone could inhibit converting enzyme and thus lower
angiotensin II levels. Inhibition of converting enzyme by
the acetone extract of adrenal glands lends additional support
to this observation.

The micro-colorimetric assay using Gly-His-Gly is the
best method employed thus far in this laboratory. Quantita-
tion of 20 to 100 nanomoles of histidyl-glycine liberated in
90 minutes by a crude preparation of plasma converting enzyme
is easily accomplished. The plasma assay for angiotensin

converting enzyme is just as sensitive and more rapid. The

131

fluorometric analysis is preferable to the spectrophotometric
method when great sensitivity is desired, since 100 to 1000
picomoles ofhistidyl-glycine per m1 can be detected. A dis-
advantage is the turbidity produced upon acidification, thus
requiring centrifugation prior to analysis.

Care must be exercised in the selection of compounds
to be incorporated into the assay solution since they may
absorb the excitation beam and diminish fluorescence. Appro-
priate reagent blanks are necessary to correct fluorescence
values should this occur.

Chromophore formation often accompanies acidification
of the assay mixture, with the concurrent absorption of
fluorescence. This problem, as well as the one described
above, can be partially avoided by bringing the concentration
of all components (enzyme and effectors) to the same level
following the termination of the ‘enzyme reaction with sodium
hydroxide. The concentration of the cleavage products, how-
ever, cannot be made equimolar. Since these components can
form chromOphores, some quenching will occur which can be

diminished by using low substrate concentrations.
B. SIDE CHAIN SPECIFICITY OF ANGIOTENSIN CONVERTING ENZYME

Converting enzyme catalyzes the hydrolysis of pep-
tides at the second peptide bond from the carboxyl end (i.e.,
a dipeptidyl carboxypeptidase), although the ionized carboxyl

group is not required for the hydrolysis of angiotensin I or

 

 

Jll’llll'llll

132

Gly-His-Gly. Thus the ethyl ester of angiotensin I and Gly-
His-Gly are cleaved at the same rate as the unesterified sub-
strate. (Table V) Substrates are most efficiently hydrolyzed
if R' is aromatic as in the amino acid phenylalanine, and R"

is an imidazole moiety of histidine:

R-NH-CH—CO-NH-CH-CO-NH-CH-COOH + HOH )
A. flu 1;...
R-NH-CH-COOH + HN-CH-CO-NH-CH-COOH
.1. “in 1;...

The systematic modification of the nature of the
N-terminal amino acid residues of the type X-glycylglycine
are shown in Table VIII. In all cases where cleavage was
found, the site of enzymatic action was restricted to the
X—glycyl peptide bond. It will be noted that the first two
dipeptides are not cleaved while the latter three tripeptides
are, indicating that a minimal chain length equivalent to a
tripeptide is required for hydrolysis. A tripeptide sub-
strate containing only aliphatic amino acids, alanyl-glycyl-
glycine, is not hydrolyzed, but upon the introduction of an
aromatic amino acid, as in phenylalanyl-glycyl-glycine,
peptide bond cleavage occurs. Note also that the first eight
peptides unhydrolyzed by converting enzyme (Table II B) are
dipeptides and, as indicated above, Ala-Gly-Gly is completely
aliphatic. The completely aromatic peptide, tetra-L-phenyl-

alanine, is not hydrolyzed, indicating that the enzymic site

 

133

TABLE VIII.--N-terminal Modified Substrate Analogs Tested
for Converting Enzyme Activity

 

 

‘ Estimated*
Substrate Analog Relative
X-Gly-Gly Rate Investigators
Benzoyl—Gly-Gly O Yang et al.
Benzyloxycarbonyl-Gly-Gly 0 Yang et al.
Alanyl-Gly-Gly 0 Summary et al.
Phenylalanyl-Gly-Gly 20 Summary et al.
Benzoylglycyl-Gly-Gly 20 Yang et al.
Benzyloxycarbonylglycyl-
Gly-Gly 5 Summary et al.

 

* Angiotensin I as 100

134

cannot bind four planar aromatic groups, perhaps due to
steric hindrance. In the substrates tested for activity
with converting enzyme (Table II A), the four that are
cleaved most rapidly, except for Gly-His—Gly, have an aro-
matic residue at the R' position and an imidazole moiety at
the R" position. The favorable effect of the aromatic and
planar substitutent at the beta carbon of the third amino
acid from the carboxyl end, is emphasized by the fact that
when R' is a hydrogen the relative rate of hydrolysis is

one fifth that for the phenylalanyl residue. The data in
Table II A also indicate that the prolyl residue in Z-Pro-
Phe-His-Leu and angiotensin I disrupts the interaction of
that portion of the enzymic region that binds planar aromatic
groups. Deletion of proline from Z-Pro-Phe-His—Leu increases
the rate of hydrolysis by a factor of ten.

In comparing Z—Phe-His-Leu and Phe-Gly-Gly, observe
that the introduction of the benzyloxycarbonyl("Z") group
causes a large (SO-fold) increase in the rate of cleavage of
the substrate, assuming, of course, the effect of the
terminal amino acids (histidyl-leucine vs. glycyl-glycine) to
be minimal. (Note only the slight two-fold increase in hydrol-
ysis of hippuryl-Gly-Gly vs. hippuryl-His-Leu.) These
results indicate that the introduction of the hydrophobic
group may cause significant increases in the susceptibility
of the phenylalanyl-histidyl bond to converting enzyme action

and suggests that the "Z" unit may play a special role in the

135

catalytic process. If one speculates that the primary binding
energy comes from the interaction of the phenylalanyl-histidyl
unit with the enzyme, similar to the phenylalanyl-phenylalanyl
unit with pepsin (80), then the hydrophobic group may also
interact with a locus on the enzyme and increase catalytic
efficiency, perhaps by exerting conformational changes at the
active site. These presumed secondary hydrophobic inter-
actions are not necessarily reflected in the binding energy
for the interaction of the entire substrate molecule with the
enzyme, nor should it be construed that hydrOphobic groups are
necessary for increased catalytic efficiency. For if one com-
pares the relative rate of hydrolysis of Gly-His-Gly with
hippuryl-His-Leu, one may conclude that the apparently weaker
binding of Gly-His-Gly enhanced the catalytic efficiency.

Thus it may be concluded for the synthetic substrates
studies so far, that enzymic cleavage by converting enzyme
should not only be viewed in terms of the amino acid residues
that flank the sensitive peptide bond, but also the secondary
hydrOphobic interactions to aid in binding, as well as the
disruptive influence of the prolyl residue, that are removed
from the sensitive peptide bond. Figure 30 depicts the active
site of angiotensin converting enzyme and illustrates the

binding of Z-Phe-His-Leu prior to hydrolytic cleavage.

 

136

Figure 30.--Pr0posed active site of angio-
tensin converting enzyme based upon substrate
analog and inhibitor experiments.

“u.“

13'7

 

 

 

H3 H3.
C/u......\C U
o E
_ ..L
O OIN S
\\ \ mm
H / E
GIN C R H
\ \\ / \ P
O O .
Z

\\\\\\\\\\\\\\

138

C. THE FLUORESCENCE REACTION OF AMINO ACIDS, PEPTIDES, AND

PROTEINS WITH O-PHTHALADEHYDE

IR, NMR, and UV data indicate the formation of a
Schiff base between OPT and glycine. To eXplain the dis—
appearance of an intense glycine proton resonance peak at
tau = 4.4 and the appearance of an intense peak at tau = 2.0,
it was necessary to assume that the double bond shifted to
the alpha-carbon side of the amino acid nitrogen atom. This
rearrangement occurs readily in alkaline solution (100).

The structural requirements for the formation of a
fluorescent complex between OPT in the absence of 8MB and
indoles, imidazoles, polyamines, guanidino amines, gluta-
thione, and amino acids are contingent upon the temperature
and pH of the fluorogenic reaction and the pH during fluores-
cence measurement. Thus the fluorogenic reaction of S-hydrox—
ytryptamine requires 60 minute boiling period at pH less than
1, while histidine, histamine, agmatine, Spermidine, and
a, e-diaminopimelic acid are dependent on strongly alkaline
conditions. In the alkaline group (Table IX), only agmatine
requires heat (88°) for the fluorogenic reaction to occur.
The fluorophore is stable in acid solution for OPT deriva-
tives of indolakylamines, imidazoles, and polyamines, but
guanidinoamines and glutathione require an alkaline pH for
fluorescence. The fluorescent product from OPT with 8MB and
amino acids is labile under acidic conditions and shows a

pH dependence similar to the guanidinoamines and glutathione.

139

Table IX.--Amino Compounds and Conditions Necessary for
Their Fluorophore Formation and Fluorescence
Upon Reaction with O-Phthalaldehydea

 

 

A. Fluorophore Formation

 

Heat and Acid pH Neutral pH Alkaline pH
Indolalkylamines: Proteinsb Imidazoles:
Serotonin Histidine
S-Methoxytryptamine Histamine
S-Hydroxytryptamine Guanidinoamines:
Arginine
Agmatine
Citrulline
Polyamines:
Spermidine

IminobisprOpylamine
Diethylenetriamine
Amino Acidsb

Glutathione

 

B. Fluorescence of Fluorophore Produced

 

Acid_pH Alkaline pH
Indolalkyamines Guanidino Amines
Imidazoles Amino Acids
Polyamines Glutathione

Proteins

 

aLiterature references to all of the compounds listed here
are given in the introductory part of this paper.

bRequire the presence of excess BME.

 

140

Shore gt_al. (90) postulated that the condensation reaction
of histamine and OPT formed a three-membered ring system,
because alkylation of the amino group prevented fluorophore
formation, and m-phthalaldehyde could not replace OPT. Cohn
and Shore (91) assumed that the reaction of OPT with the
guanidinoamine, agmatine, terminated with the Schiff base,
since acid was not necessary for fluorOphore formation. Cohn
and Lyle (95) suggested that the hemimercaptal was formed
with glutathione since the pK of the sulfhydryl group is at
the point of optimal fluorescence.

Similarly, the addition of 8MB to OPT allows the base
catalyzed formation of the hemimercaptal, and a Schiff base
is formed upon the reaction of the remaining aldehyde group
with the amino acid. The prOposed reaction is shown in
Figure 31. Hemimercaptal formation is evidenced from spectra
by the disappearance of the carbonyl and sulfhydryl groups
and the appearance of one or more alcohol groups. The nucleo-
philic addition of BME to the imine bond was not indicated as
evidenced by the appearance of the new strong transition in
the UV spectra. If addition to the imine bond occurred, then
the appearance of a new chromophore would not be seen. The
lack of change in the tau = 8.0 and 6.6 signals upon the
addition of BME is indicative that no further change occurs
in the imine bond region. The proximity of the sulfur and
nitrogen atoms suggests an interaction which may lead to an

enhancement of the native fluorescence of the imine bond.

141

.moflom ocHEm ocm Hocmnumoummoumfilm nufiz mpmzmo
lamamzucmlo mo coflwommu Ummomoumll.am mnsmflm

llllll

142

 

N/omz 02.24

 

A)
Io .NIo -NIo m.I I
mmmoxm

 

JV.

143

Movement of the imine bond relative to the sulfur atom
should be minimized in order for strong fluorescence to be
observed. A molecule of water, forming strong hydrogen
bonds with the electron rich nitrogen and sulfur atoms will
tie these two side chains together. This would account for
the decrease in fluorescence upon addition of ethanol to the
buffered OPT-8MB amino acid mixture.

As shown in Table VIII glutathione reacts with OPT
alone to produce a strongly fluorescent derivative, while
cystine does not. The increase in fluorescence upon
insertion of a carbon in the side chain (homocysteine)
suggests that the chain is not flexible enough to allow hemi-
mercaptal formation with cysteine. In the case of gluta-
thione, there is sufficient flexibility to allow hemimercaptal
formation and hence, fluorescence.

Small shifts in the fluorescence spectra maxima
indicate that the addition of amino acid or protein to
buffered OPT-8MB slightly alters the resulting fluorophore.
Since the absence of carbonyl groups adjacent to the aro-
matic ring appears to be necessary for fluorescence, the
dihemimercaptal is probably the OPT-6MB fluorophore. The
molar ratio of OPT to 3MB (8:1) in buffered OPT-8MB reagent
is favorable for dihemicaptal formation. The slight differ-
ence in the excitation and emission maxima between the OPT-
BME fluorophore and the OPT-BME amino acid fluorOphore

suggests that a 11' to 11' * transition is involved.

 

 

 

4|. Vltllr‘ I I'll-II II I .I

 

144

This was concluded due to the lack of a possible n t01r*
transition in the dihemimercaptal adduct. The excitation
maximum (325 nm) and emission maximum (435 nm) for the OPT
(no 8MB) BSA adduct indicates a considerably different
fluorOphore.

The reaction of malonaldehyde with amines was
studied by Chio and Tappel (98b and they attributed the
fluorescence to formation of a Schiff base. Crowell and
Varsel (97) investigated twenty-six substituted aromatic
aldehydes and found that they generally did not fluoresce
even when other ring substitutents were capable of promoting.
fluorescence. This indicates that the aldehyde group exerts
a strong quenching effect on the fluorescence property of the
aromatic system. When methyl hemiacetals were formed, most
of them exhibited appreciable fluorescence. The acetals
were shown to have the same excitation and fluorescence maxima
as the corresponding alkyl derivatives. Likewise, the effects
of the carboxylic acid group on the fluorescence of benzoic
acid was investigated by Tournon and El-Bayoumi (99), and
they found that the group reduced the fluorescence when com-
pared to that of toluene. A similar effect is noted with OPT.
The carbonyl group must be removed by either reduction by
NaBH4or hemimercaptal formation with BME.

Varying the number of methylene groups between the
amino and carboxyl ends of the amino acid suggests that the

carboxyl group does not participate in the reaction itself,

 

ll ill)" 11‘

 

 

Ill :1] .l. .. fill .‘|.Iu .111 I’IFIi III III III I II In I

145

but does have an integral role in the fluorescence. The
reduction of fluorescence with histamine, as compared to his-
tidine, in the presence of the buffered OPT-8MB reagent
supports this conclusion (77). With the glycine-OPT-BME
adduct in alkaline solution, the dissociated form of the
carboxylic acid if fluorescent, but upon protonation with
hydrochloric acid there is a marked depression in fluores-
cence. Since the glycine ethyl ester adduct also showed
fluorescence equal to glycine, it appears unlikely that the
carboxylate ion is essential for fluorescence. Under acid
conditions the fluorophore is also labile, degrading into
colored products and a ninhydrin positive component as demon-
strated by TLC. As emphasized by Roth (77), alkaline pH was
essential for fluorOphore formation but the fluorophore was
stable from pH 6.0 to 11.5. There is a much greater decrease
in fluorescence when the glycyl-OPT—BME adduct is acidified
as compared with the ethyl ester. The undissociated car-
boxylic acid group can quench the fluorescence of the adduct.
Possibly the quenching of fluorescence may be due to proton
transfer between the acid group and either the aromatic ring
or the imine bond. The pH effect observed may also be due to
protonation of the nitrogen in the imine bond.

Fluorescence spectra of the amino acids, which had
varying distances between the amino and carboxyl ends, showed
no changes in the partial quantum efficiency of fluorescence.

The UV spectral scans ofthe amino acids with OPT and BME were

'1

146

all the same (except for the aromatics) indicating that
despite a variance in the relative fluorescence among amino
acids as indicated by Roth (77) they all react in the same
manner.

The various amino acid alkyl and aryl groups may
extend quenching and enhancement contributions to the basic
fluorophore. An examination of the data allows the following
generalizations to be made: 1) the presence of a conjugated
system (tyrosine) or a double bond, such as a carbonyl group
in glutamine, tends to enhance the fluorescence over the
glycine standard; 2) the presence of a lone pair of electrons,
such as on an amine (ornithine), tends to quench the fluores-
cence; and 3) increasing the hydrophobic content of the group
also decreases the fluorescence. Structural models of the
product show that the amino acid side chains are free to
extend close to the imine bond region of the molecule.

Another factor affecting the relative fluorescence
of the various amino acids is competition between the separate
groups on one amino acid for OPT. In the case of lysine and
arginine, the epsilon amino and guanidino groups may also
form Schiff bases. This would tend to limit the number of
possible fluorOphores.

Fluorescence spectra obtained with BSA, OPT, and SME,
show the same excitation and emission maxima as is seen for
the amino acid derivatives, and the conclusion is that basi-

cally the same reaction is occurring with proteins.

 

 

147

Fluorescence is much more intense after reaction with OPT

in the presence of 8MB than in the absence of 8MB, so the
reaction involving histidine without BME was disregarded as
the source of fluorescence. The pH profile for the BSA
fluorescent derivative (Figure 20) shows that highly alkaline
conditions prevent fluorophore formation, but the fluoro—
phore, once formed, is stable at high pH as with amino acids.

The binding of 14 molecules of BME to one of OPT is
rather high, but this may be the amount necessary to shift
the equilibria in favor of the hemimercaptal derivative.

This number for BME is not corrected for reduction of any
disulfide linkages in BSA.

The degree of protein crosslinking by the OPT-8MB
reagent is dependent on concentration. In high concentra-
tions of reagent, BSA precipitates and the centrifuged pro-
duct is a rubbery pellet. If the concentrations of OPT and
8MB are kept slightly higher than saturation levels (55 OPT/l
BSA), no such polymerization occurs and it is assumed that
little crosslinking characteristic of dialdehydes is occurring.

Studies with oligOpeptides indicated that the pep-
tide bond quenches fluorescence when compared with glycine;
yet, on an equimolar basis, bovine serum albumin produces
a fluorescence that is a hundred fold greater than glycine.
Quenching by the peptide bond may be the result of fewer car-
boxylic acid groups to enhance fluorescence. The enhance-

ment of fluorescence seen with the bovine serum albumin deri-

 

148

vative may be accounted for by the fact that the amino acids
in bovine serum albumin, which react with OPT, may have a
greater relative fluorescence compared to glycine. The
fluorescence spectra (Figures 23 and 24) give evidence for a
secondary mechanism of fluorescence, that is energy transfer
from tyrosine and tryptOphan to the OPT-8MB adduct. The
primary mechanism is direct excitation of the fluorophore at
340 nm and flourescence emission at 454 nm. Excitation at
270-280 nm will allow energy absorption by the typtophan and
tyrosine residues in bovine serum albumin and fluorescence
emission at 340 nm. This emission band overlaps the excita-
tion band for the OPT-8MB fluorOphore. Energy transfer from
tyrosine to the OPT-8MB fluorophore does not occur since the.
partial quantum efficiency for excitation at 280 nm is about
20% of that observed at 340 nm when emission is observed at
450 nm, thus indicating a possible transfer of energy to the
OPT adduct. The partial quantum efficiency of the glycine
adduct, under conditions specified above, shows a baseline
value.

The OPT-8MB reaction has been used as an analytical
method for fluorescent determination of amino acids (77).
This reaction is used extensively in this laboratory as a
quantitative method for the determination of amino acids, pep-
tides and proteins in place of the ninhydrin reaction. The
simplicity of the method allows tests to be made within 10
to 15 minutes, and the sensitivity allows one nanomole of
amino acid or one hundredth of a nanomole of protein to be

detected.

 

 

149

The reagent may be used for the determination of
total protein, especially where low concentrations and small
amounts of sample are available, as in cerebrospinal fluid.
OPT in conjunction with NaBH4 or BME may be used as a fluor-
escent probe for detection of protein conformation changes.
Measurements would be limited to a few hours; however, the
label can be reacted with dilute protein solutions in sev-
eral minutes. Several of the current probes require consid-
erable time and dialysis to prepare the specimen.

Bifunctional reagents such as o-phthalaldehyde under-
go complex reactions with proteins apparently involving many
different side-chains. For example, glutaraldehyde was
used to confer structural rigidity on carboxypeptidase (119).
Since a primary amino group is required for Schiff base form-
ation, o-phthalaldehyde probably reacts with lysine, argi-
nine, and terminal amino groups.

The molecular weight relationship of the four purified
proteins to fluorescence intensity permitted quick estimation
of their individual molecular weights, which were determined
by generally more lengthy methods. Thus using the inde-
pendently determined molecular weight of a protein, a solution
equimolar to that of other standard protein solutions may be
prepared. Then the fluorescence of its OPT-8MB derivative
may be compared to other protein standards as depicted on
the standard curve in Figure 27. The reason for the linear
relationship between fluorescence intensity and molecular
weight of the four proteins is unknown, and the dependence

must be established for other proteins.

 

150

Although the fluoreecence obtained with adenosine
monophosphate was only about a hundredth that of glycine on
an equimolar basis, the reagent shows the possibility of its
use as a fluorescent probe in nucleic acids.

The OPT-8MB reagent can be used for the assay of any
peptidase or protease activity, since cleavage of peptide
bonds results in a net increase in the number of free amino

groups and thus an increase in derivable fluorescence.

 

 

I 0". ‘i III I]! I54- in! il II. I |.I..I|.Ij I'll

REFERENCES

151

 

'r

10.

11.

12.

13.

14.

15.

REFERENCES

Tigerstedt, R. and P. G. Bergmann. Scand. Arch. Physiol.
8, 223 (1898).

 

Goldblatt, J. Lynch, R. F. Hanzal, and W. W. Summerville.
J. Exptl. Med. 119, 389 (1934).

 

Braun—Menendez, J. C. Fasciolo, L. F. Leloir, and J. M.
Munoz. J. Physiol. London, 98, 283 (1940).

 

Page, I. H. and O. M. Helmer. J. Exptl. Med. 71, 29

 

Skeggs, L. T., K. E. Lentz, H. Hochstrasser, and J. R.
Kahn. J. Exptl. Med. 118, 73 (1963).

 

Plentl, A. A., I. H. Page, and W. W. Davis. J. Biol.
Chem. 147, 143 (1943).

 

Green, A. A. and F. M. Bumpus. J. Biol. Chem. 210, 281
(1954).

 

Skeggs, L. T., J. R. Kahn, K. E. Lentz and N. P. Shumway.
J. Exptl. Med. 106, 439 (1957).

 

Skeggs, L. T., K. E. Lentz, J. R. Kahn, and N. P. Shumway.
J. Exptl. Med. 108, 283 (1958).

 

Skeggs, L. T., K. E. Lentz, H. Hochstrasser, and J. R.
Kahn. Can. Med. Assoc. J. 90, 185 (1964).

 

Bumpus, F. M., P. A. Khairallah, K. Arakawa, I. H. Page,
and R. R. Smeby. Biochim. BiOphys. Acta 46, 38 (1961).

 

Skeggs, L. T., K. E. Lentz, N. P. Shumway, and K. R.
Woods. J. EXptl. Med. 104, 193 (1956).

 

Plentl, K. E., L. T. Skeggs, K. R. Woods, J. R. Kahn,
and N. P. Shumway. J. Exptl. Med. 104, 183 (1956).

 

Peart, W. S. Biochem. J. 62, 520 (1956).

 

Elliot, D. F. and W. S. Peart. Biochem. J. 65, 246
(1957).

 

152

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

153

Aoyagi, H., K. Arakawa and N. Izumiya. Bull. Chem. Soc.
Japan 41, 433 (1968).

 

Arakawa, K., M. Nakatani, A. Mino ara, and M. Nakamura.
Biochem. J. 104, 900 (1964).

 

Skeggs, L. T., K. E. Lentz, H. Hochstrasser, and J. R.
Kahn. J. EXptl. Med. 118, 73 (1963).

 

Skeggs, L. T., J. R. Kahn, and N. P. Shumway. J. EXptl.
Med. 103, 295 (1956).

 

Dorer, F. E., L. T. Skeggs, J. R. Kahn, K. E. Lentz, and
M. Levine. Anal. Biochem. 33, 102 (1970).

 

Bahkle, Y. S. Nature. 220, 919 (1968).

Corcoran, R. J., H. W. Henry, G. W. Sander, and C. G.
Huggins. Fed. Proc. 29, 281 (1970).

 

S. Oparil. Fed. Proc. 29, 282 (1970).

 

Yang, H. Y. T., E. G. Erdos, T. A. Jenssen, and Y. Levin.
Fed. Proc. 29, 281 (1970).

 

Page, I. H. and F. M. Bumpus. Physiol. Rev. 41, 331
(1961).

 

Ng, K. K. K., and J. R. Vane. Nature. 216, 762 (1967).

Braun-Menendez, E., J. C. Fasciolo, L. F. Leloir, J. M.
Munoz and A. C. Taquini. Renal Hypertension. Spring-
field, Ill., Thomas (1946).

 

Bumpus, F. M., R. R. Smeby, I. H. Page, and P. A.
‘Khairallah, Can. Med. Assoc. J. 90, 190 (1964).

 

Regoli, D., B. Riniker, and H. Brunner.- Biochem. Phar—
macol. 12, 637 (1963).

 

chubu T., H. Akutsu,S. Fujimoto, E. Jeda, K. Hiwada,
and Y. Yamamura. Biochim. Biophys. Acta 191,558 (1969)

 

Skeggs, L. T., K. E. Lentz, A. B. Gould, H. Hochstrasser,
and J. R. Kahn. Fed. Proc. 26, 42 (1967).

 

Skeggs, L. T., J. R. Kahn, and N. P. Shumway. J. Ethl.
Med. 95, 241 (1952).

 

Peart, W. S. Biochem. J. 59, 300 (1955).

 

Pickens, P. T., F. M. Bumpus, A. M. Lloyd, R. R. Smeby
and I. H. Page. Circulation Res. 17, 438 (1965).

 

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

154

Helmer, O. M. and W. E. Judson. Circulation. 27, 1050
(1963).

 

Skinner, S. L. Circulation Res. 20, 391 (1967).

 

Gould, A. B., L. T. Skeggs, and J. R. Kahn. Lab.
Invest. 15, 1802 (1966).

Boucher R., R. Yeyrat, J. de Champlain, and Genest.
Can. Med. AsSoc. J. 90, 194 (1964).

Scornik, O. A. and A. C. Plaldini. Am. J. Physiol. 201,
526 (1961).

 

Brown, J. J., D. L. Davies, A. F. Lever, J. J. S.
Robertson, and M. Tree. Biochem. J. 93, 594 (1964).

 

Yallotton, M. B., L. B. Page, and E. Haber. Nature.
215, 714 (1967).

Goodfriend, T. L., D. L. Ball and D. B. Farley. J. Lab.
and Clin. Med. 72, 648 (1968).

 

Gocke, D. J., L. M. Sherwood, I. Oppenhoff, J. Garten,
and J. H. Laragh. J. Clin. Endocr. 28, 1675 (1968).

 

Hollemans, H. J. G., van der Meer, and J. L. Touber.
Nature. 217, 277 (1968).

Boyd, G. W. and Peart, W. S. Lancet. July 20,129
(1968).

Goodfriend, T. L., L. Levine and G. D. Fasman. Science.
144, 1344 (1964).

Harber, E., L. B. Page, and G. A. Jacoby. Biochem.
4, 693 (1965).

 

Boyd, G. W., J. Landon, and W. S. Peart. Lancet. 1002
(Nov., 1967).

Catt, K. J., M. C. Cain, and J. P. Coghlan. Lancet. 1005
(Nov., 1967).

Landowne, R. Chim. Analyt. 47, 589 (1965).

 

Roth, M. and A. Reinhartz. Helv. Chim. 'Acta 49, 1903
(1966). j

Piquilloud, Y., A. Reinharz,and M. Roth. Helv. Chim. Acta
49,1903 (1966).

Reinhartz, A. and M. Roth. Eur0pean J. Biochem. 7, 334
(1969).

 

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68a.

68b.

69.

155

Roth, M. in Meth. of Biochem. Analysis. 18, 261 (1968).

 

Cushman, D. W. and H. S. Cheung. Fed.‘Proc. 28, 280
(1969).

 

Huggins, C. G., R. J. Corcoran, J. S. Gordon, H. W.
Henry, and J. P. John. Circulation Res. 27, Suppl.
1, 93 (1970).

 

Flodin, P. and J. Killander. Biochim. Biophys. Acta
63, 403 (1962).

 

Gocke, D. M., J. Gerten, L. M . Sherwood, and J. H. Laragh.
Hypertension. 17, 131 (1969).

 

Haber, E., T. Loerner, L. Page, B. Kliman, and A.
Purnode. J. Clin. Endo. and Metab. 29, 1349 (1969).

 

Boucher, R., H. Kurihara, C. Grise, and J. Genest. Cir-
culation Res. 27, Suppl. 1, 83 (1970).

 

 

Piquilloud, Y., A. Reinharz and M. Roth. Biochim.
Biophys. Acta 206, 136 (1970).

 

Folin, O. and V. Ciocalteau. J. Biol. Chim. 73, 627
(1927).

 

Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J.
Randall. J. Biol. Chem. 193, 265 (1951).

 

Warburg, O. and W. Christian. Biochem. Z. 310, 384
(1941)

 

Brenner, M. and Pataki, G. Helv. Chim. Acta 44, 1420
(1961).

 

Moffat, E. D. and R. I. Lytle. Anal. Chem. 31, 926
(1959).

 

Gordon, A. H. and L. N. Louis. Anal. Biochem. 21, 190
(1967).

 

Dalby, P. L., and H. A. Lillevik. "Taxonomic Analysis
of ElectrOphoretic Blood Serum Patterns in the Cotton
Rat, Sigmodon.” Publicn. Museum Mich. State Univ.
4, 75 (1969).

Besaw, L. G., P. L. Dalby and H. A. Lillevik. Anal.
Biochem. accepted for publication (1972).

 

Fischer, L. An Introduction to Gel Chromatography. Ameri—
can Elsevier Publishing Co., N. Y. (1969).

 

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.
82.

83.

84.

85.

86.

87.

88.

156

"Sephadex Ion Exchangers, An Outstanding Aid in Bio-
chemistry" by Pharmacia Fine Chemicals.

Moore, 8. J. Biol. Chem. 243, 6281 (1968).

 

Block, J. R. "Amino Acid Analysis of Protein Hydrolysis."
In AnalytiCal Methods in Protein Chemistry_(P. Alexander
and R. J. BIock, eds.), Vol. II, p. 34. Permagon Press,
Oxford (1960).

 

"Sephadex-Gel Filtration in Theory and Practice" by
Pharmacia Fine Chemicals.

Yang, H. Y. T. and E. G. Erdos. Nature. 215, 1402
(1967).

Yang, H. Y. T., E. G. Erdos, and Y. Levin. Biochim. and
Biophys. Acta 214, 374 (1970).

Strong, L. E. "Blood Fractionation." In Enc clo edia
of Chemical Technology. Vol. III, p. 567 1196;).

Roth, M. Anal. Chem. 43, 880 (1971).

 

 

 

 

Scatchard, G., J. S. Coleman and A. L. Shen. J. Am.
Chem. Soc. 56, 658 (1934).

 

Yang, H. Y. T., E. G. Erdos, T. A. Jenssen, and Y. Levin.
Fed. Proc. 29, 281 (1970).

 

Inouge, K. and J. S. Fruton. Biochemistry. 7, 1711
(1968).

 

Zimmerman, W. Z. Physiol. Chem. 189, 4 (1930).

 

Patton, A. R. J. Biol. Chem. 108, 267 (1935).

 

Klein, G., and H. Linder. Z. Physiol. Chem. 20, 251
(1932).

 

Sandstrom, w. M. and H. A. Lillevik. Ind. Eng. Chem.
Anal. Ed. 13, 781 (1941).

 

 

Patton, A. R. and E. M. Foreman. Science. 109, 339
(1949).

Curzon, G. and Giltrow, J. Nature. 173, 314 (1954).

Maickel, R. P. and F. P. Miller. ”Anal. Chem. 38,
1937 (1966).

 

Bogdanski, D. R., A. Pletscher, B. B. Brodie and S.
Ydenfriend. J.'Pharmacol. Exptl.'Therap, 117, 82
(1956).

 

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

107.

157

Gerber, D. Anal. Biochem. 34, 500 (1970).

 

Shore, P. A., Alan Burkhalter, and Victor H. Cohn, Jr.
J. Pharmacol. and EXptl. Therap. 127, 182 (1959).

 

Cohn, V. H. and P. A. Shore. Anal. Biochem. 2, 237
(1961).

 

 

Kremzner, L. T. Anal. Biochem. 15, 270 (1966).

Rogers, C. J., C. W. Chambers, and N. A. Clarke. Anal.
Biochem. 20, 321 (1967).

 

Pisano, J. J., J. D. Wilson, L. Cohen, D. Abraham, and
S. Udenfriend. J. Biol. Chem. 236, 499 (1961).

 

Cohn, V. H. and J. Lyle. Anal. Biochem. 14, 434 (1966).

 

McNeil, T. L. and L. V. Beck. Anal. Biochem. 22, 431
(1968).

 

Crowell, E. P. and C. J. Varsel. Anal. Chem. 35, 189
(1963).

 

Chio, K. S. and A. L. Tappel. Biochemistry. 8, 2821
(1969).

 

Tournon, J. and M. A. El-Bayoumi. J. Am. Chem. Soc. 93,
6386 (1971).

 

Bernhard, S. A. The Structure and Function of Enzymesz
W. A. Benjamin Inc., p. 215 (1968).

 

Holland, J. F. Ph.D. Thesis, Michigan State University
(1971).

Holland, J. F. R.E. Teets, and Timnick. Anal. Chem.
accepted for publication (1972).

 

Rodd, E. H., ed. Chemistry of Carbon Compounds. Elsevier,
Vol. III, p. 284 (1956).

 

Lee, H. J., J. Larne, and I. Wilson. Biochem. Biophys.
Acta 235, 521 (1971).

 

 

Boucher, R., H. KuriHara, C. Grise, and J. Genest.
Circulation Res. 26 and 27, Suppl. 1, 83 (1970).

 

Cushman, D., H. Cheung. Biochem. Pharmacol. 20, 163
(1971).

 

Yang, H. Y. T., E. Erdos, and Y. Levin. Biochem. Biophys.
Acta 214, 3, 4 (1970).

 

108.
109.

110.

111.

112.

113.

114.

115.

116.

117.

118.

119.

158

Bakhle, Y. S. Nature. 220, 919 (1968).
Ng, K. K. F. and J. R. Vane. Nature. 225, 1142 (1970).

Ondetti, M. A., H. Williams, E. Sabe, J. Pluscec, E.
Weaver, and O. Kecy. BiochemiStry. 10, 4033 (1971).

 

Stewart, J. M., S. H. Ferreira and L. J. Greene. Bio—
chem. Pharmacol. 20, 1557 (1971).

 

Gross, F. Klin. WochSChr. 36, 693 (1958).

 

Davis, J. 0., "The Renin-Angiotensin System in the Con-
trol of Aldosterone Secretion." pp. 173—200. Kidngy
Hormones. (J. W. Fisher, ed.) Academic Press (1971).

 

Sachs, D. H. and E. Painter. Science. 175, 781 (1972).

Detterbeck, F. J. and H. A. Lillevik in Progress in
Thin-layer Chromatography and Related Methods (A.
NiederWIeser and G. Pataki, eds.) pp. 235-247. Ann
Arbor Science Publishers, Inc. (1971).

 

Summary, J. J., J. L. Browning, and H. A. Lillevik.
Submitted for publication to Biochemistry (1972).

Hawk, P. B., B. L. Oser, and W. H. Summerson.
Practical Physiolo ical Chemistry. McGraw-Hill Book
Co., 13th ed., p. 64 (I954).

Fraenkel-Conrat, H. "Methods for Investigating Essen-
tial Groups for Enzyme Activity." In Methods in
Enzymology. (S. P. Colowick and N. O. Kapean, eds.),
Vo1. IV. 251, 254. Academic Press (1957).

 

 

Quiocho, F. A. and F. M. Richards. Biochemistry.
5,4062 (1966).

 

   

”7(111117111’11111((1)1111)1:71))”

46 1944