This dissertation has beenmicrofilmed exactly as received 69-16,648

BENSON, Ann Marie, 1936-THE AMINO ACID SEQUENCE OF LEUCAENAGLAUCA FERREDOXIN.

University of Hawaii, Ph.D., 1969Biochemistry

University Microfilms, Inc., Ann Arbor, Michigan

THE AMINO ACID SEQUENCE OF

LEUCAENA GLAUCA FERREDOXIN

A DISSERTATION SUBMITTED TO THE GRADUATE DIVISION OF THE

UNIVERSITY OF HAWAII IN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

IN BIOCHEMISTRY

JANUARY 1969

by

Ann Marie Benson

Dissertation Committee:

Kerry T. Yasunobu, ChairmanLawrence H. PietteMorton MandelJohn B. HallJohn A. Hunt

DEDICATION

To my mother

ACKNOWLEDGMENTS

To the National Institutes of Health

for support, 1967 to 1969.

To Mrs. Kerry T. Yasunobu and Chrissie and Steven Yasunobu

for gathering a great many pounds of Leucaena glauca

leaves.

To Dr. Hans Georg Kloepfer

for felling numerous Leucaena glauca trees.

TABLE OF CONTENTS

LIST OF TABLES •

LIST OF FIGURES

ABBREVIATIONS

ABSTRACT . . • • .

v

vi

.. viii

x

1. INTRODUCTION

A. Discovery of Ferredoxins .

· . . .· ......

1

1

B. Classification, Properties, and PhysiologicalFunctions of Ferredoxins . . . . . 3

C. Primary Structure of Ferredoxins • 15

D. Statement of the Problem .

II. MATERIALS AND METHODS

A. Materials . . . . · · . . · . · · . ./.

l. Materials and Reagents ObtainedCommercially . · · . . · . · ·

2. Materials Obtained as Gifts · ·3. Materials Obtained by Preparative Methods

B. Methods

1. Isolation of Ferredoxin .

25

33

33

33

34

34

35

35

2. Preparation of the CarboxymethylcysteineDerivative .• . . . . • . . . . . . . . 41

3. Methods of Hydrolysis by Endopeptidases . 42

c. Bacillus subtilis Neutral Protease.

· . . . . . . . . . . .

a. Trypsin

b. Chymotrypsin.

d. Thermolysin

· . . . . . . . . . .42

42

42

43

4. Methods of Peptide Purification. . . . . . .ii

43

a. Chromatographic Separation ofTryptic Peptides . • • • . . . . . . . 43

b. Chromatographic Separation ofChymotryptic Peptides . . . . . 44

c. Further Purification of Trypticand Chymotryptic Peptides • . . . . . . . 44

d. Chromatographic Separation ofB. subtilis Neutral ProteaseFragments of Tryptic Peptide T-7a 45

e. Partition Chromatographic Separationof B. subtilis Neutral ProteaseFragments of Chymotryptic Peptide C-lO 46

f. Chromatographic Separation of ThermolysinFragments of Chymotryptic Peptide C-lO •. 46

5. Determination of Amino Acid Composition • •• 47

6. Methods of Amino Acid Sequence Determination 48

a. Methods of NH2-terminal Analysis . . • •. 48

1. Edman Degradation • • . • • • 48

. . .

50

50

50

50

51

52

. . .. .

. . .2. Carboxypeptidase A

3. Carboxypeptidase B

2. Dinitrophenylation

3. Dansylation • . • .

b. Methods of COOH-terminal Analysis

1. Hydrazinolysis

7. Studies on Ferredoxin from Individual Trees. 52

. . . .A. Isolation of Ferredoxin

III. RESULTS • . . • . . . . . . . . . .. . . .

. . 54

54

B. Preparation of the CarboxymethylcysteineDerivative . . • . • . • • • • • • • •

C. Amino Acid Composition of Ferredoxin • •

. . . 54

54

111

D. Amino and Carboxyl Terminal Residues · · · · · · 62

E. Tryptic Peptides • · · · · · · • · · · 62

1- Initial Chromatography · · · · · · 62

2. Further Purification of Tryptic Peptides 68

3. Amino Acid Composition · · · · · · · 68

4. Amino Acid Sequence · · · · · 75

F. Chymotryptic Peptides · • · · · · · 83

1- Initial Chromatography · · · · · · · 83

2. Further Purification of ChymotrypticPeptides . . · · · · · · · · · · · · · · 83

3. Amino Acid Composition · · · · · · · · · 86

4. Amino Acid Sequence · · · · · · · · · · · · · 86

G. Complete Amino Acid Sequence of Leucaenaglauca Ferredoxin • · · · · · · · · · · · · • · 113

H. Distribution of Sequence Heterogeneityamong the ~. glauca Population · · · • · · · · · 113

IV. DISCUSSION AND CONCLUSIONS · · · • · · · · · · · · 120

A. Characteristics of the Amino Acid Sequenceof L. glauca Ferredoxin · · · · · · · · · • · · 120

B. Comparison of Plant and Algal Ferredoxinsfrom Four Species · · · · · · · · · · · · · · · 121

C. Acidic Residues in Plant and AlgalFerredoxins. . . · · · · • · · • · · · · · · · · 125

D. Genetic and Evolutionary Aspects ofVariations in Ferredoxin Sequences · · · · · 127

1- Plant and Algal Ferredoxins · · · · · · · 127

2. Heterogeneity in ~. glauca Ferredoxin · · 132

E. Techniques in Ferredoxin Purification · · · 140

F. Specificities of Proteolytic EnzymesEmployed in these Stud1.es · · · · · · · · • · · 143

iv

G. Determination of the Complete Structureof L. glauca Ferredoxin • • • • . • • • • • • .• 144

. . . . . . . . . . . . . . . . . . . .V.

VI.

SUMMARY

BIBLIOGRAPHY • • . . . . . ... . . . . . . . . . . .146

148

LIST OF TABLES

I. Amino Acid Compositions of Ferredoxins fromL. gl~uca and Other Plants • • • • • • • • •

II. Tryptophan Content of ~. glauca Ferredoxin • · . .

v

61

63

III. Amino Acid Composition and Properties ofTryptic Peptides . . . . . . . . . . . · · · . 71

IV. Amino Acid Composition and Properties ofChymotryptic Peptides . . . . . . . . · · · . 97

V. Distribution of Amino Acid Residues at ThreePoints of Heterogeneity in L. glauca Ferredoxin 119

VI. Acidic Residue Content of Proteins of VariousType s • • • • •• ••••.••••••• · . 126

VII. Mutation Distances and Values of MinimumBase Difference per Codon • . . . . • • .

VIII. Types of Mutations in L. glauca FerredoxinHeterogeneity•.•••••••••••••

· . . .

· . . .

130

141

vi

LIST OF FIGURES

1. Spectra of Type I ferredoxins •• . . . . . . . . . 6

2. Spectra of typical Type II and Type IIIferredoxins • • • • . • • • . • . . . . . . . . 8

3 a. Summary of the functions of ferredoxin ingreen plants •• • • • • • • • • . • • • . . . . . 11

3 b. Summary of the functions of ferredoxin inphotosynthetic bacteria • • • • . • • . • . 13

20

3 c. Summary of the metabolic reactions in whichType III ferredoxin is known to function • • . •. 17

4 a. Amino acid sequences of three Type IIIferredoxins • • • • . • • • • . . • • •

4 b. The structure of a possible precursor ofType III ferredoxins • • • • • • • • . • · . . 22

5. Comparison of the amino acid sequences offerredoxins from C. butyricum and Chromatium

6. Comparison of spinach and Clostridialferredoxins . • • . • • • • • • . • • • • • • · . .

24

27

60· . .

7. Amino acid sequences of ferredoxins from alfalfa,spinach, and Scenedesmus, a green alga • • • • •• 29

8. Photograph of a specimen of Leucaena glauca • • 32

9. Procedure for isolation of ~. glauca ferredoxin 38

10. Photomicrograph of ~. glauca ferredoxin crystals 56

11. Absorption spectrum of ~. glauca ferredoxin. • 58

12. Purification of CMFd by gel filtration througha column of Sephadex G-25 • • • • • . • • • •

13. Action of carboxypeptidase A on CMFd . . . . 65

14. The elution pattern of the tryptic digest ofCMFd from AG l-X2 • • . • • • • 67

15. Purification of tryptic peptide T-2 on AG 50-X2 70

16. Summary of evidence establishing the sequenceof tryptic peptide T-7a • • • • • • • • • • • • 81

· . .

17. Elution pattern of the chymotryptic digestof CMFd from a column of AG 50-X2 • • • • •

18 a. Purification of peptide c-4 by gel filtration .

18 b. Purification of peptide C-5 by gel filtration.

19. Resolution of peptides C-7a, C-7b, C-12, andc-14 on a column of AG l-X2 • . • • • • • • •

· .· .· .

vii

85

88

90

92

20 a. Pattern of elution of a mixture containingpeptide C-9 from Sephadex G-15 •• • . . • · . . . 94

. . . . .

20 b. Distribution of peptide C-9 and a contaminantin the effluent fractions from Sephadex G-15,as determined by paper chromatography . • • .

21. Summary of the evidence establishing thesequence of peptide C-lO • • . • • • . •

· . . 96

107

22. Elution pattern of peptides from thermolysindigest of chymotryptic peptide C-lO • • • •

23. Tryptic and chymotryptic peptides arrangedin order in L. glauca ferredoxin • • • • •

· . . .· . . .

110

115

. . . . . . .

24. Elution pattern of tryptic peptides of theferredoxin of Tree 1 from AG l-X2 • • • • • •

25. The amino acid sequences of ferredoxins fromL. glauca, alfalfa, spinach, and Scenedesmus

26. Phylogenetic tree of the ferredoxins

· . .· . .

118

123

134

27. Inheritance patterns of allelic and non-allelicgenes . . . . . . . . . . . . . . . . . . . . . 137

A

Ala

Arg

Asp

Asn

Asx

BSNP

C

CMCys

CMFd

CoA

Cys

DEAE

DNP

DNS

EDTA

G

Glu

GIn

Glx

Gly

His

lIe

Leu

Lys

LIST OF ABBREVIATIONS

Adenine

Alanine

Arginine

Aspartic acid

Asparagine

Aspartic acid or asparagine

Bacillus subtilis neutral protease

Cytosine

S-carboxymethylcysteine

Carboxymethylated ferredoxin

Coenzyme A

Cysteine or half-cystine

Diethylaminoethyl

2,4-Dinitrophenyl

5-dimethylamino-I-napthalene sulfonyl

Ethylenediaminetetraacetic acid

Guanine

Glutamic acid

Glutamine

Glutamic acid or glutamine

Glycine

Histidine

Isoleucine

Leucine

Lysine

viii

Met

PCAW

Phe

Pi

PPNR

Pro

PTH

Rf

Ser

Th

Thr

TLCK

TPCK

TPN

Tris

Trp

Tyr

U

UV

Val

ix

Methionine

Pyridine-collidine-acetic acid buffer

Phenylalanine

Inorganic phosphate

Photosynthetic pyridine nucleotide reductase

Proline

Phenylthiohydantoin

Partition coefficient

Serine

Thermolysin

Threonine

L-(1-tosylamido-2-lysyl)ethyl chloromethyl ketone

L-(1-tosylamido-2-phenyl)ethyl chloromethyl ketone

Triphosphopyridine nucleotide

Trishydroxymethylamino methane

Tryptophan

Tyrosine

Uracil

Ultraviolet

Valine

x

ABSTRACT

Ferredoxin, a protein of unusually low oxidation­

reduction potential, and also the first protein found to con­

tain both iron and labile sulfide, plays a very fundamental

role in the photosynthetic process, that of acceptor of

electrons from light-activated chlorophyll.

Ferredoxin was isolated from the leaves of LeUcaena

glauca, a species of small leguminous tree. The carboxy­

methylcysteine derivative of the combined ferredoxin from

many plants was used for the amino acid sequence studies.

The products of tryptic and chymotryptic hydrolyses were

purified by ion exchange chromatography, gel filtration, and

paper chromatography. Thermolysin and Baci'llus subtilis

neutral protease were uRed to further degrade large peptides.

Most of the sequence was determined by the subtractive

method of Edman degradation (J. BioI. Chern., 237, 2547, 1962).

The amino acid sequence, deduced primarily from the

structures of the tryptic and chymotryptic peptides, is as

follows: Ala-Phe-Lys-Val-Lys-Leua/Valb-Leu-Thr-Pro-Asp-Gly­

Proa/Alab-Lys-Glu-Phe-Glu-Cys-Pro-Asp-Asp-Val-Tyr-Ile-Leu-Asp­

Gln-Ala-Glu-Glu-Leu-Gly-Ile-Glux/Aspx-Leu-Pro-Tyr-Ser-Cys-Arg­

Ala-Gly-Ser-Cys-Ser-Ser-Cys-Ala-Gly-Lys-Leu-Val-Glu-GIy-Asp­

Leu-Asp-Gln-Ser-Asp-Gln-Ser-Phe-Leu-Asp-Asp-Glu-Gln-Ile-Glu­

Glu-Gly-Trp-Val-Leu-Thr-Cys-Ala-Ala-Tyr-Pro-Arg-Ser-Asp-Val­

Val-Ile-Glu-Thr-His-Lys-Glu-Glu-Glu-Leu-Thr-GlYx/Alax-COOH.

The sequence of this ferredoxin resembles those of

xi

spinach and alfalfa ferredoxins, showing a slightly greater

similarity to alfalfa, as would be expected from taxonomic

considerations. Over 70% of the sequence, including the five

cysteine residues, is invariant in these three species. Many

of the changes between species are conservative, but radical

changes involving charged residues and proline residues occur

also.

Heterogeneity was found in positions 6, 12, 33, and 96

within the L. gl"auca ferredoxin sequence, each of which was

occupied by two different amino acid residues, as shown.

However, the tryptic peptide containing residues 6 and 12

occurred in only two of the four possible forms, one of which

contained leucine and proline in positions 6 and 12, respec­

tively, whereas the other contained valine and alanine. Thus

the presence of different genes is indicated, rather than

ambiguity in translation of the genetic code.

A study of the distribution of the different forms of

the protein among the ~. glauca population was undertaken.

In a normal distribution of allelic genes, 50% of the popula­

tion would be expected to be homozygous. Ferredoxin was

isolated from ten different plants and three of the four

points of heterogeneity were investigated (residues 6, 12,

and 96). Heterogeneity was found in each of these positions

in each of the ten individual trees. Thus the probability

that the differing allelic nuclear genes are involved is less

than 0.001. The two most probable explanations for the

xii

observed sequence heterogeneity are (1) that differing non­

allelic genes for ferredoxin are present, that is, that gene

duplication has occurred in the evolution of this species,

followed by point mutations in the individual genes, or (2)

that the ferredoxin genes are located not in the nucleus of

the call but rather in the chloroplasts, and their distribu­

tion therefore follows a pattern quite different from that of

allelic nuclear genes.

I. INTRODUCTION

A. Discovery of Ferredoxins

Ferredoxins, a group of non-heme, non-flavin proteins

containing iron and labile sulfide and possessing unusually

low oxidation-reduction potentials, function as electron

carriers in cellular redox reactions driven by light energy

or by the hydrogenase system. Proteins of this type have been

found in anaerobic bacteria such asC16~t~idia (1), in photo­

synthetic bacteria such as Chromatium (2), in algae such as

Nostoc (3) and in chloroplasts of spinach (4), alfalfa (5)

and other species of green plants.

In 1952 a soluble factor, then called "methaemoglobin

reducing factor," was isolated from parsley chloroplasts by

Davenport, Hill, and Whatley (6). This sUbstance, believed

to be a protein, catalyzed the reduction of methemoglobin in

the presence of chloroplast fragments and light. Four years

later, San Pietro and Lang (4) isolated a similar factor from

spinach chloroplasts, found it to be involved in the photo­

reduction of NADP, and named it "photosynthetic pyridine

nucleotide reductase," or "PPNR." In 1957, Arnon, Whatley

and Allen (7) independently isolated the same factor from

spinach chloroplasts and named it "TPN-reducing factor."

When Davenport (8) reported in 1960 that the partially purified

methemoglobin reducing factor, which had since been found to

catalyze the photoreduction of other heme proteins as well (9)

and had been renamed "haem-reducing factor," was also active

2

in catalyzing the photoreduction of NADP by chloroplasts, it

became evident that the three separately isolated and named

factors, haem-reducing factor, PPNR, and TPN-reducing factor,

were identical.

In the following two years factors which were active in

the photoreduction of NADP, and which were similar in some

but not all respects to the plant factors previously found,

were isolated from bacteria. Losada, Whatley, and Arnon (2)

isolated a pyridine nucleotide reductase from Chromatium, a

photosynthetic bacterium, in 1961 and in the following year

Mortenson, Valentine, and Carnahan (10) reported isolation of

a non-heme iron protein, which they named "ferredoxin," from

Clostridium pasteurianum, a species of anaerobic non-photo­

synthetic bacterium in which this protein served as an electron

carrier. Q. pasteurianum ferredoxin was obtained in crystalline

form by Tagawa and Arnon and was shown to be capable of cata­

lyzing the photoreduction of NADP by illuminated spinach

chloroplasts (11). These investigators also found that the

chloroplast protein contained iron, as did the C. pasteurianum

ferredoxin (10), that it was reversibly oxidized and reduced,

with accompanying spectral changes, and that both of these

proteins had unusually low oxidation-reduction potentials,

E'o at pH 7.55 being -417 mv for C. pasteurianum ferredoxin

and -432 mv for the spinach protein (11). These values are

close to the potential of hydrogen gas and approximately 100

mv more negative than that of the pyridine nucleotides. The

3

interchangeability of the native spinach protein and C.

pasteurianum ferredoxin in the chloroplast reaction and the

fact that both of these proteins contain iron and have unusu­

ally low oxidation-reduction potentials led to the proposal

by Tagawa and Arnon (11) that the name "ferredoxin" be extended

to include the chloroplast protein as well. Thus the class of

proteins known as "ferredoxins" has been defined as those non­

heme, non-flavin, iron containing proteins which have an

oxidation-reduction potential close to that of hydrogen gas,

and which are capable of catalyzing the photoreduction of NADP

by washed chloroplasts.

B. Classification, Properties, and Physiological Functions of

Ferredoxins

Within the past few years ferredoxins have been isolated

from numerous other species, including plants, algae, and

photosynthetic and non-photosynthetic anaerobic bacteria.

Since the properties of these proteins was found to var~ with

their source, the ferredoxins were classified into three types

(12) according to their origins:

(I) Ferredoxin of green plants and algae.

(II) Ferredoxin of photosynthetic bacteria.

(III) Ferredoxin of non-photosynthetic anaerobic bacteria.

These three types of ferredoxin differ in molecular

weight, iron and labile sulfur content, and spectral character­

istics.

Type 1 ferredoxins have the highest molecular weight,

4

approximately 11,500 (5,13,14) and contain two atoms of iron

and two atoms of labile sulfide (5) per molecule. The

characteristic absorption spectra of these plant and algal

ferredoxins have maxima at approximately 465, 420, 330, and

276 m~. The visible and UV spectra of ferredoxins from spinach

(11) and Scenedesmus (15), a green alga, are shown in Figure 1.

Type II ferredoxins, those of the photosynthetic bacteria,

have a molecular weight of approximately 10,000 (16), inter­

mediate between Type I and Type III ferredoxins. They contain

five to six gram-atoms each of iron and labile sulfide per

mole (12). The typical Type II ferredoxin spectrum has maxima

at 390, 310, and 280 m~, as can be seen in Figure 2, which

shows the spectrum of Chromatium (16) ferredoxin. This figure

also shows the spectrum of ferredoxin from Clostridium

pasteurianum (1), a typical Type III ferredoxin, which has the

same visible maxima as ferredoxins of Type II. However, the

ferredoxins from non-photosynthetic anaerobic bacteria differ

from those of the photosynthetic bacteria in other properties,

having a lower molecular weight, approximately 6,000 (1), and

containing seven to eight atoms each of iron and labile sulfide

per molecule (12).

Type I ferredoxin plays a very fundamental role in the

process of photosynthesis, being the primary acceptor of light­

activated electrons from chlorophyll, and thus the first stable

reductant formed as a result of the radiant energy trapped (11).

In 1962 Whatley, Tagawa, and Arnon (17) separated the light and

5

Figure 1. Spectra of Type I ferredoxins. Curve A is

the spectrum of spinach ferredoxin in 0.033 M Tris buffer of

pH 7.80. Curve B is the spectrum of Scenedesrnus ferredoxin

in 0.01 M Tris buffer of pH 8.0. These spectra, which

represent different concentrations of protein, both exhibit

maxima at 325, 420, and 463 m~. The ratios among the

absorbancies at these three maxima are identical in the

two spectra. In the region below 300 m~, the slight shift

in maxima and the differences in absorbancy ratios between

these maxima and the other maxima in the corresponding

spectra are due to differences in content of aromatic amino

acids, especially tryptophan. The spectra shown in the

figure are reproduced from pUblished spectra (11, 15).

I&JUZCma::o(I)IIIC

I&J>-t-C....I&J0:

250 350 450 550

-_......

WAVELENGTH (mfJ)

7

Figure 2. Spectra of typical Type II and Type III

ferredoxins. Curve A is the spectrum of ferredoxin from

Chromatium, a photosynthetic bacterium. The protein was

dissolved in 0.005 M phosphate bUffer, pH 7.6. Curve B is

the spectrum of Clostridium pasteurianum ferredoxin, in 0.07 M

Tris buffer of pH 7.3. These are reproductions of pUblished

spectra (16, 1). These spectra, which represent different

protein concentrations, are very similar. Each exhibits

maxima at 280 and 390 rn~ and a shoulder at 310 m~.

I&JUZea:aa:ofnmeIIJ>-l-e.J1&.1a:

300 350 400 450 500

WAVELENGTH (mfl)

9

dark reactions in electron transfer during photosynthesis,

showing that photoreduction of NADP occurs in two steps. The

first step is the reduction of ferredoxin by light-activated

chlorophyll. The reduced ferredoxin then reduces NADP in a

dark reaction which is mediated by a flavoprotein enzyme, NADP

reductase. Reduced ferredoxin was shown by these investigators

to be the physiological electron carrier in NADP reduction and

in cyclic and noncyclic photo-phosphorylation. Another func­

tion of ferredoxin in plants involves nitrogen fixation.

Nitrate is converted to ammonia by illuminated spinach grana,

by a two-step reaction (18):

FMNN03- ------------------>

nitrate reductase

FerredoxinN02- ------------------> NH3nitrite reductase

In the second step of this reaction, reduced ferredoxin

supplies the electrons, via the reductase enzyme, for the

reduction of nitrate to ammonia. The functions of ferredoxin

in green plants are summarized in Figure 3 a.

Figure 3 b shows some of the functions of ferredoxin in

photosynthetic bacteria. Although these bacteria do not con­

tain chloroplasts, the process of photosynthesis is similar

to that in plants. However, unlike plants, the photosynthetic

bacteria require a reductant other than water, the oxidation

of which supplies electrons to ferredoxin via photoactivated

chlorophyll. Since water cannot serve as the reductant,

oxygen is not evolved. In contrast to chloroplasts, which

have no hydrogenase, photosynthetic bacteria can photoproduce

Figure 3 a. Summary of the functions of ferredoxin in green plants.

I-'o

1'0:I:Z

f

.1'0oz

0..­+

..J «

..J>-g: > ON

oa::g"""l::<_~O:I: ~U

a.. :I:o a..« _----:>....... 0Z «z

a..I­«

0..­+a..o«

t

II

Figure 3 b. Summary of the functions of ferredoxin in photosynthetic bacteria.

I-'I\)

H2-CO? :> PYRUVATE

CO2

Ir:==--~_~_N_A_DP--;;;~NADPH

ACETYL

CO2 . 01 -KETO-

SUCCINYL coA4 GLUTARATE

LIGHT > CHLOROPHYLL :> I FERREDOXIN

1REDUCTANT

14

hydrogen gas, which isa product of the reduction of hydrogen

ions by reduced ferredoxin.

In 1966 Evans, Buchanan, and Arnon (19) discovered a new

ferredoxin-dependent carbon reduction cycle in a green sulfur

photosynthetic bacterium, Chlorobium thiosulfatophilum. Each

turn of the cycle fixes four molecules of C02' yielding one

molecule of oxaloacetate. Compounds ranging in size from C2

to C6' including pyruvate and ot-ketoglutarate, are also

synthesized from C02 in this cycle. Although evidence for this

cycle has been obtained only with this species, these reactions

are believed to occur in other species of photosynthetic

bacteria also.

In the non-photosynthetic anaerobic bacteria ferredoxin

acts as an electron carrier in numerous oxidation-reduction

reactions. Ferredoxin can be reduced either by hydrogen in

the presence of bacterial hydrogenase (10), or by substrate,

as in the oxidation of pyruvate to CO 2 and acetyl phosphate

in C. acidi-urici, a species in which the hydrogenase enzyme

is not present (20). The reduced ferredoxin can then function

as an electron donor in the reduction of pyridine nucleotides

(11), the conversion of nitrite and hydroxylamine to ammonia

(21), the production of pyruvate from acetyl coenzyme A and

CO2 (22,23,24,25), the reduction of sulfite to sulfide (26),

and the reduction of urate to xanthine (27). Each of these

reactions is mediated by a specific enzyme (e.g. nitrite

15

reductase, xanthine oxidase). Oxidized ferredoxin has been

found to serve as the electron acceptor in the oxidation of

various sUbstrates, such as pyruvate, hypoxanthine (27), 0(­

ketoglutarate (28), formate (29), and in the case of Methano­

bacillus, acetaldehyde (30). Specific dehydrogenase enzymes

catalyze these reactions. The role of ferredoxin in

anaerobic non-photosynthetic bacteria is summarized in

Figure 3 c.

C. Primary Structure of Ferredoxins

Since the ferredoxins play such a fundamental role in the

processes of photosynthesis and nitrogen fixation, and since

they appeared to be such unique proteins with their extremely

low oxidation-reduction potentials, and, furthermore, were

the first proteins reported to contain both iron and labile

sulfide, there has been much interest in determining the

structure-function relationships eXisting in this class of

proteins. Also, since three types of ferredoxin have been

isolated, all of which are to some extent functionally inter­

changeable, it was suspected that the primary structures of

these proteins might show evidence of evolutionary relation­

ship among the three types. Thus, in 1962 the first steps

were taken toward elucidating the primary structure of a

ferredoxin, that of Clostridium pasteurianum. Since that

time, amino acid sequence determinations have been performed

on ferredoxins of various species, including representatives

of all three of the ferredoxin types previously described.

I-'0\

Figure 3 c. Summary of the metabolic reactions in which Type III ferredoxin is

known to function. The compounds at the left of the diagram are oXidized by ferredoxin,

which then acts as an electron donor in the reactions at the right.

ACETALDEHYDE

"'~>---'t::H2 -----------1FORMATE

HYPOXANTHINE

a-KETOGLUTARATE

S0'3 > S=

-~> NADH

N02

Y\NH 3

Y

ACETYLC~ PYRUVATE

CO2

,;

18

The sequence of C. pasteU:rianum (31) ferredoxin was completed

in 1964, followed by C. bu:tyrlcum (32) in 1966, and Micrococcus

aerogenes" (33) in 1968. These Type III ferredoxins, shown in

Figure 4 a, were each found to consist of a single polypeptide

chain, 54 to 55 residues in length. Each has eight residues

of cysteine, all of which are involved in iron binding, and

all of which occupy homologous positions in the sequence in

all three species. A striking similarity between the first

half and the second half of the ferredoxin molecule was noted

in the case of C. pasteuriartum (31), and was found in the

other two species as well. It was this similarity between the

two "half-molecules" which led to the postulation of a pre­

cursor molecule (34), approximately half the size of these

ferredoxin molecules, which gave rise to the Type III

ferredoxins by gene duplication. Figures 4 a and 4 b show

possible precursor structures.

In 1968 the amino acid sequence of a Type II ferredoxin,

that of Chromatium (12), was completed. As shown in Figure 5,

a large amount of homology was found to exist between the

Type III ferredoxins and Chromatium ferredoxin, indicating a

close evolutionary relationship between these two groups of

proteins (12). Chromatium ferredoxin was found to contain an

insertion of nine amino acid residues and an extra 17 residues

at the carboxyl terminus, as compared with the' Clo'stridial

ferredoxins. Eight of the nine residues of half-cystine in

Chromatium ferredoxin were found to be homologous to the eight

19

Figure 4 a. Amino acid sequences of three Type III

ferredoxins. The six homologous half-molecules are aligned

for comparison. Lines (1), (2), and (3) show the amino acid

sequences of residues 1 through 28, and lines (4), (5), and

(6) show the sequences of residues 29 through 56, of the

ferredoxins of Clostridium pasteurianum, c. butyricum, and

Micrococcus aerogenes, respectively. The possible structures

of a precursor are shown on line (7) and include all the amino

acid residues found at each position in the six half-molecules.

-Ala-Asp-Ser-Cys-Val-Ser-Cys-Gly-Ala-

1

(1) Ala-Tyr-Lys-Ile-

(2) Phe Val

(3) Tyr Val

29

5

Asn

Asn

34

10

Val Ser

Ile Ala

39

20

Ile-Phe-Val-Ile-Asp-Ala-Asp-Thr-Cys-Ile-Asp-Cys-Gly-Asn-(4)

(5)

(6)

GIn Phe Val

Ile Tyr Ala

1 5

Thr

Ser

10

Asn

Ser

(7) Ala-Tyr-Lys-Ile-Asp-Ala-Asp-Ser-Cys-Val-Ser-Cys-Gly-Ala-Ile Phe Val Asn Thr Ile Ala AsnGIn Ala Asp Ser

Cys-Ala-Ser-Glu-Cys-Pro-Val-Asn-Ala-Ile-Ser-Gln-Gly-Asp-Ser

Cys-Ala-Asn-Val-Cys-Pro-Val-Gly-Ala-Pro-Val-Gln-Glu

(1)

(2)

(3)

(4)

(5)

(6)

14

43

14

Ala Gly

Lys Pro

Asn

Ser

20

49

20

Ser Ala

Asn

Thr

GIn

Asn GIn

Asn Pro

55

28

Asp Thr

Ser

Asp.

28

(7) Cys-Ala-Ser-Glu-Cys-Pro-Val-Asn-Ala-Ile-Ser-Gln-Gly-Asp-SerLys Gly Val Ser Pro Thr Pro Glu Thr

Pro Gly GInAsn Val

Asn

21

Figure 4 b. The structure of a possible precursor of

Type III ferredoxins. In this structure, each position is

occupied by the predominant residue in the corresponding

positions in the six half-molecules of Figure q a. Three

positions are each occupied by two amino acids since neither

dominates in the half-molecules. The placement of

phenylalanine in position 2 is based on the belief that

c. butyricum is the most conservative of the three species

compared.

1 7 14

22

Ala-Phe-Val-lle-Asp-Ala-Asp-Ser-Cys-lle-Asp-Cys-Gly-Ala-

15 21 28Glu lIe Gly

Ala-Asn-Val-Cys-Pro-Val-Gly-Ala-Pro-Asn-Gln-Glu-Asp-Ser

23

Figure 5. Comparison of the amino acid sequences of

ferredoxins from ~. butyricum and Chromatium. The numbers

below the two rows of sequences refer to the minimum number

of base changes for each codon which would be required to

convert one sequence to the other.

(A)

(B)

(A)

(B)

(A)

(B)

(A)

(B)

(A)

(B)

(B)

24

1£. butyricum A1a-Phe-Va1-I1e-Asn-Asp-Ser-Cys-Va1-Ser-Cys-

Chromatium A1a-Leu-Met-I1e-Thr-Asp-G1n-Cys-I1e-Asn-Cys­o 1 1 0 1 0 2 0 110

G1y-A1a-Cys-A1a-G1y-G1u-Cys-Pro-Va1-Ser-A1a-I1e-Thr-G1n­

Asn(G1x,Cys,Asx,Pro,G1x,Cys,Pro,Va1)G1y-A1a-I1e-Ser-G1n­2 2 0 2 2 1 0 0 0 1 001 0

G1y-Asp-Thr-G1n-Phe-Va1-I1e-Asp-A1a-Asp-Thr-Cys-I1e-Asp­

G1y-Asp-G1u-Thr-Tyr-Va1-I1e-G1u-Pro-Ser-Leu-Cys-Thr-G1u­o 0 2 2 1 0 0 1 1 2 2 011

41 42Cys-G1y- -Asn-Cys-A1a-

41 51Cys-Va1-G1y-His-Tyr-G1u-Thr-Ser-G1n-Cys-Va1-Asp-Cys-Va1-o 1 101

Asn-Va1-Cys-Pro-Va1-G1y-A1a-Pro-Asn-G1n-G1u

G1u-Va1-Cys-Pro-I1e-Lys-Asp-Pro-Ser-His-G1u-G1u-Thr-G1u­20001 210 110

81Asp-G1u-Leu-Arg-A1a-Lys-Tyr-G1u-Arg-I1e-Thr-G1y-G1u-G1y

25

half-cystine residues of the Type III ferredoxins.

Before amino acid sequence determinations had been per­

formed on plant ferredoxins, it was thought that they might

have evolved by gene duplication from the Type III ferredoxins,

just as the latter appeared to have evolved from their postu­

lated precursor. When the sequence of spinach ferredoxin (14)

was published, it could be seen that the evolution of this

protein had not followed such a simple course. However,

comparison of the spinach ferredoxin sequence with those of

Clostridial ferredoxins showed some regions of homology, as

shown in Figure 6, indicating possible evolution of these two

types of ferredoxin from a common precursor (12,14). The

amino acid sequence of alfalfa ferredoxin (35) has also been

determined and, as would be expected, is similar to that of

spinach ferredoxin. Recently, an algal ferredoxin, that of

Scenedesmus (36) was also sequenced and was found to be very

similar to the ferredoxins of the higher plants. Figure 7

shows the comparison of Scenedesmus ferredoxin with the two

plant ferredoxins.

D. Statement of the Problem

The amount of homology between spinach ferredoxin and

the clostridial ferredoxins is small enough to leave serious

doubts as to whether the plant and bacterial ferredoxins are

truly related in the evolutionary process. The possibility

of convergent evolution, in which similar structural patterns

evolve independently due to their advantage to the function

C\)

0\

Figure 6. Comparison of spinach and Clostridial ferredoxins. The amino acid

sequence of spinach ferredoxin, on line (2), is aligned with portions of the sequence of

Clostridium butyricum ferredoxin, on line (1), to show the regions of homology between

the structures of these two proteins. Gaps have been introduced into both sequences to

strengthen homology. The carboxyl terminal portion of spinach ferredoxin was not

homologous to any portion of the C. butyricum ferredoxin, and is therefore compared with

residues 27 through 34 of the spinach ferredoxin sequence. The numbers below the two

lines of sequences show the minimum base difference per codon. No value was assigned by

Matsubara et ale (12) in those positions where gaps are postulated in the sequences.--The average value obtained for the minimum base difference per codon for these alignments

was 1.07, which is interpreted (12) as an indication of evolutionary relationship between

spinach ferredoxin and C. butyricum ferredoxin.

1 20A1a-Phe-Va1-I1e-Asn-Asp-Ser-Cys-Va1-Ser-Cys-G1y-A1a-Cys-A1a-G1y-Glu-Cys-Pro-Va1­

20-Leu-Val-Thr-Pro-Thr-Gly-Asn-Va1-Glu-Phe-G1n-Cys-Pro-Asp

1 2 212 0 2 212 1 0 0 1

1A1a-Ala-Tyr-Lys-Val-Thr­

o 1 2 1 1

(1)

(2)

(1)

(2)

21 34 4 8Ser-A1a-Ile-Thr-Gln-Gly-Asp-Thr-Gln-Phe-Val-Ile-Asp-Ala-Ile-Asn-Asp-Ser-Cys-21 34 35 39 41Asp-Va1-Tyr-Ile-Leu-Asp-Ala-Ala-Glu-Glu-Glu-Gly-Ile-Asp-Ile-Pro-Tyr-Ser-Cys-Arg-A1a-212 1 1 1 1 1 1 3 1 2 2 1 1 210 0

(1)

(2)

9 29Va1-Ser-Cys-Gly-A1a-Cys-Ala-Gly-G1u-Cys-Pro-Va1-Ser-Ala-Ile-Thr-Gln-Gly-Asp-Thr-G1n-42 62Gly-Ser-Cys-Ser-Ser-Cys-A1a-G1y-Lys-L~u-Lys-Thr-Gly-Ser-Leu-Asn-Gln-Asp-Asp-Gln-Ser-

1 001 1 0 0 012 2 211 1 1 0 1 022

49-Asn-Cys-A1a-Asn-Val-Cys-Pro-Val­

83-Tyr-Pro-Va1­100

30Phe-Val-Ile-Asp-Ala-Asp-Thr-Cys-I1e-Asp-Cys-G1y­63Phe-Leu-Asp-Asp-Asp-G1n-I1e-Asp-G1u-G1y-Trp-Va1-Leu-Thr-Cys-A1a-A1a-0120121 2 2 1 1 1 1 0 0 2

(1)

(2)

(1)

(2)

50 55G1y-Ala- -Pro-Asn-Gln-Glu84 89Ser-Asp-Val-Thr-Ile-Glu-Thr-

1 1 111 2

27 34(2) Ala-A1a-G1u-Glu-Glu-Gly-Ile-Asp

90 97His-Lys-Glu-G1u-G1u-Leu-Thr-Ala

2 2 0 0 0 2 1 1

N-.:I

28

Figure 7. Amino acid sequences of ferredoxins from

alfalfa, spinach, and Scenedesmus, a green alga. The resi­

dues below the spinach ferredoxin sequence at positions 31

and 33 are believed to occupy these positions in a minor

form of the spinach protein. Amino acid residues which are

identical in all three species are underlined.

29

1 5 10 15:1) Alfalfa Ala-Ser-Tyr-Lys-Val-Lys-Leu-Val-Thr-Pro-Glu-Gly-Thr-Gln-Glu-Phe-Glu-C:

(2) Spinach Ala-Ala-Tyr-Lys-Val-Thr-Leu-Val-Thr-Pro-Thr-Gly-Asn-Val-Glu-Phe-Gln-C;

(3) Scenedesmus Ala-Thr-Tyr-Lys-Val-Thr-Leu-Lys-Thr-Pro-Ser-Gly-Asp-Gln-Thr-Ile-Glu-C:

23 30 35 40(1) Tyr-Ile-Leu-Asp-His-Ala-Glu-Glu-Glu-Gly-Ile-Val-Leu-Pro-Tyr-Ser-Cys-Arg-Ala-Gly-Se

(2) Tyr-Ile-Leu-Asp-Ala-Ala-Glu-Glu-Glu-Gly-Ile-Asp-Leu-Pro-Tyr-Ser-Cys-Arg-Ala-Gly-SeLys Met

(3) Tyr-Ile-Leu-Asp-Ala-Ala-Glu-Glu-Ala-Gly-Leu-Asp-Leu-Pro-Tyr-Ser-Cys-Arg-Ala-Gly-Al

48 55 60 65(1) Ala-Gly-Lys-Val-Ala-Ala-Gly-Glu-Val-Asp-Gln-Ser-Asp-Gly-Ser-Phe-Leu-Asp-Asp-Asp-GJ

(2) Ala-Gly-Lys-Leu-Lys-Thr-Gly-Ser-Leu-Asn-Gln-Asp-Asp-Gln-Ser-Phe-Leu-Asp-Asp-Asp-G:

(3) Ala-Gly-Lys-Val-Glu-Ala-Gly-Thr-Val-Asp-Gln-Ser-Asp-Gln-Ser-Phe-Leu-Asp-Asp-Ser-G:

73 80 85 go(1) Trp-Val-Leu-Thr-Cys-Val-Ala-Tyr-Ala-Gly-Ser-Asp-Val-Thr-Ile-Glu-Thr-His-Lys-Glu-G

(2) Trp-Val-Leu-Thr-Cys-Ala-Ala-Tyr-Pro-Val-Ser-Asp-Val-Thr-Ile-Glu-Thr-His-Lys-Glu-G

(3) Phe-Val-Leu-Thr-Cys-Val-Ala-Tyr-Pro-Thr-Ser-Asp-Cys-Thr-Ile-Ala-Thr-His-Lys-Glu-Q

29

5 10 15 20-Tyr-Lys-Va1-Lys-Leu-Va1-Thr-Pro-G1u-G1y-Thr-G1n-G1u-Phe-Glu-Cys-Pro-Asp-Asp-Val-

-Tyr-Lys-Va1-Thr-Leu-Va1-Thr-Pro-Thr-Gly-Asn-Val-G1u-Phe-Gln-Cys-Pro-Asp-Asp-Val-

-Tyr-Lys-Va1-Thr-Leu-Lys-Thr-Pro-Ser-Gly-Asp-GIn-Thr-Ile-Glu-~-Pro-Asp-Asp-Thr-

30 35 40 45Ala-G1u-Glu-Glu-Gly-I1e-Val-Leu-Pro-Tyr-Ser-Cys-Arg-Ala-Gly-Ser-Cys-Ser-Ser-Cys-

·Ala-G1u-Glu-Glu-Gly-I1e-Asp-Leu-Pro-Tyr-Ser-Cys-Arg-Ala-G1y-Ser-Cys-Ser-Ser-Cys-Lys Met

·A1a-G1u-Glu-Ala-Gly-Leu-Asp-Leu-Pro-Tyr-Ser-Cys-Arg-Ala-Gly-Ala-Cys-Ser-Ser-~-

55 60 65 10-Ala-G1y-G1u-Va1-Asp-Gln-Ser-Asp-Gly-Ser-Phe-Leu-Asp-Asp-Asp-Gln-I1e-Glu-G1u-G1y-

-Thr-G1y-Ser-Leu-Asn-G1n-Asp-Asp-Gln-Ser-Phe-Leu-Asp-Asp-Asp-G1n-I1e-Asp-Glu-Gly-

-A1a-G1y-Thr-Va1-Asp-Gln-Ser-Asp-Gln-Ser-Phe-Leu-Asp-Asp-Ser-Gln-Met-Asp-Gly-G1y-

80 85 90 91-Va1-Ala-Tyr-A1a-G1y-Ser-Asp-Val-Thr-I1e-G1u-Thr-His-Lys-Glu-Glu-Glu-Leu-Thr-Ala

-Ala-Ala-Tyr-Pro-Va1-Ser-Asp-Val-Thr-I1e-G1u-Thr-His-Lys-Glu-Glu-Glu-Leu-Thr-A1a

~Va1-Ala-Tyr-Pro-Thr-Ser-Asp-Cys-Thr-Ile-A1a-Thr-His-Lys-Glu-G1u-Asp-Leu-Phe

I\I

!

f

30

of the molecules, must also be considered. In order to obtain

more information as to the relationships among the three types

of ferredoxins, it will be necessary to determine the amino

acid sequences of ferredoxins from additional species of

plants and bacteria. From such data it may be possible to

determine the evolutionary pattern and the types of mutations

which have occured between species.

Another important aspect of these comparative studies

involves structure-function relationships, since ferredoxin

from all of these sources can function in the photoreduction

of NADP. The functional interchangeability between bacterial

and plant ferredoxins suggests that, despite differences in

properties such as molecular weight, spectrum, and iron and

sulfur content, there is an active site common to all three

groups of ferredoxins. Thus, a comparison of the primary

structures of a greater number of bacterial and plant

ferredoxins may help in elucidation of the nature of this

active site.

Determination of the amino acid sequence of ferredoxin

from Leucaena glauca, a species of small leguminous tree

(Figure 8) abundant in Hawaii, was undertaken in hopes that

the results of this research would contribute some of the

information needed to resolve the as yet unanswered questions

involving structure-function and evolutionary relationships

among the ferredoxins.

31

Figure 8. Photograph of a specimen of Leucaena glauca.

This plant, like all legumes, has seeds in pods. The flower,

which can be seen at top center, appears similar to that of

alfalfa.

33

II. MATERIALS AND METHODS

A. MATERIALS

1. Materials and Reagents Obtained Conunercially:

a •..J. T. BakerChemic"al CO"ITipany; Ne·w ·Jersey

Hydrindantin

Sulfanilic acid

b. Biorad Laboratories, California

Cellex D

AG l-X2 (200-400 mesh)

AG 50-X2 (200-400 mesh)

c. Brown Company, New Hampshire

Solka-Floc

d. California Corporation for Biochemical Research,

California

5-Dimethylamino-l-naphthalene sulfonyl chloride

e. Cyclo Chemical Corporation, California

L-(1-tosylamido-2-phenyl)ethyl chloromethyl ketone

f. Eastman Organic Chemicals, New York

B-Dimethylaminobenzaldehyde

Phenylisothiocyanate

Trifluoroacetic acid

g. Mallinckrodt Chemical WOrks, Missouri

Ethylenediaminetetraacetic acid

h. Matheson, Coleman and Bel1 2 New Jersey

Hydrazine

Hydrazine sulfate

34

i. Nutritional Biochemicals 'Corporation, Ohio

Iodoacetic acid

Trishydroxymethylamino methane

j. Pharmacia Fine Chemicals, Incorporated, New York

Sephadex G-10, G-15 G-25F, G-75F

k. Pierce Chemical Company, Illinois

2,4-Dinitrofluorobenzene

Ninhydrin

Trichloroacetic acid

1. Carl Schleicher and SchuellCompany, New Hampshire

Diethylaminoethyl cellulose

m. Sigma Chemical Company, Missouri

Cellulose

B-Mercaptoethanol

n. Warner-Chilcott Laboratories, California

Silica gel G

o. Worthington Biochemical Corporation, New Jersey

Carboxypeptidase A

Carboxypeptidase B

Chymotrypsin, pancreatic

Trypsin, pancreatic

2. Materials Obtained as Gifts:

a. Thermolysin was a gift from Dr. H. Matsubara,

Space Sciences Laboratory, University of California,

Berkeley.

3. Materials Obtained by 'Prepa.rative Methods:

35

a. Bacillus subtilis neutral protease was isolated

by Dr. James McConn, by the method of McConn,

Tsuru and Yasunobu (37).

b. TLCK-Chymotrypsin was prepared from commercially

obtained chymotrypsin by Dr. James McConn, accord­

ing to the method of Mares-Guia and Shaw (38).

c. TPCK-trypsin was prepared from commercially

obtained trypsin according to the method of

Wang and Carpenter (39).

B. METHODS

1. Isolation of Ferredoxin:

The isolation procedure included homogenization

and acetone fractionation by a method adapted from

that of San Pietro and Lang (40). Modifications in­

cluded the addition of acetone without prior dilution,

filtration, or centrifugation of the homogenate, and

removal of particulat material from the 35% acetone

treated homogenate by filtration through Solka-Floc

rather than by centrifugation. Chromatography on

columns of DEAE-cellulose and ammonium sulfate

fractionation, carried out by procedures similar to

those of Tagawa and Arnon (41), were followed by

adsorption of the ferredoxin onto DEAE-cellulose in

ammonium sulfate solution (0.5 gram per ml) and

elution with Tris buffer, and gel filtration through

Sephadex G-75 (42). The isolation procedure is

36

summarized in Figure 9.

Six kg of leaves, which had been picked at

various locations in the vicinity of the University

of Hawaii campus and stored in the freezer, were

divided into batches of 500 grams. Each batch was

homogenized in a Waring blender at 4° for 2-10

minutes at high speed with 750 ml 12.5 roM Tris-HCl

buffer of pH 7.2 which was 44 mM in NaCl. Twelve

batches of leaves yielded 15 liters of homogenate.

To each 2 1/2 liters of homogenate were added slowly

and with vigorous stirring 810 ml of acetone of -10°.

Each batch was immediately filtered through a layer

of Solka-Floc in a Buchner funnel at 4°. All partic­

ulate matter was retained by the filter and

approximately 2.3 liters of brown filtrate were

obtained per batch of homogenate. To each 2 liters

of filtrate were added, slowly and with stirring,

3.2 liters of acetone at -10°. The flask was packed

in dry ice and the precipitate was permitted to

settle for a few minutes. Most of the supernatant

was then siphoned off and the remaining suspension

was centrifuged at -10° for 5 minutes at 5000 rpm.

All steps after this point were carried out at 4°.

When 12 batches of leaves had been processed to

this point, the precipitates thus obtained were

combined and extracted with 1.2 liters of 0.01 M

Tris buffer of pH 7.3. After removal of the residual

31

Figure 9. Procedure for isolation of Leucaena glauca

ferredoxin.

.~

12 KG. LEAVESt Warino Blender

HOMOGENATE

~ +Acetone (to35%) i FilterFILT RAT Et +Acetone(to 75%)\ DecantiCentrifuge

PRECIPI TATEExtract; DialyzeDEAE-cellulose column (5 x50cm)

Wash: 0.1 ~IO.2M Tris, pH 7.3Elute l 0.5 M Tris,pH-r.3

RED BAND (1650ml)Dilute 2, -foldDEAE 2 (3 x20cm)DEAE 3 (3 x 20 cm)DEAE 4 (3 x 6 cm)

RED BI\ND (85 mt)t + (NH4)2S04 (,O.5gm/ml),Centrifuge

SUPERNATANTDEAE 5 (I x 15 em)

Wash:O.l M Tria, pH7.3, containing0.5 gm/ml(NH4)2S04

Elute: 1M Tris, pH 7.3RED EFFLUENT (35ml)t Sephadex G-75 (4x70cm)PURE FERREDOXIN (240 mg)

t +(N~~S04(O.50m/ml)CRYSTALLINE FERREDOXIN

39

precipitate by centrifugation at 8500 rpm for 10

minutes at 0°, the supernatant was dialyzed over­

night against 4 liters of 0.01 M Tris (pH 7.3) at

4°, centrifuged again to remove precipitate which

-f-ormed during dialysis, and then applj.ed to a

column (5 x 50 cm) of DEAE-cellulose which had been

equilibrated with 0.1 M Tris. All Tris buffers

were adjusted to pH 7.3 with HC1. One liter of the

same buffer was then passed through the column to

remove any residual acetone. A dark brown band

could be observed at the top of the column, with a

red band immediately below it. The material was

left at this stage for 24 hours while another 6 kg

of leaves were processed to this point. The material

thus obtained was applied to the same DEAE-cellulose

column. The column was then washed with 2 liters of

0.1 M Tris buffer followed by 1 liter of 0.2 M Tris

buffer. The effluent, at first orange and finally

pale yellow, was discarded. Ferredoxin was eluted

from the column in 1650 ml of 0.5 M Tris buffer.

This solution was diluted 2 1/2 fold, to yield a

Tris concentration of 0.2 M, and applied to a 3 x 20

cm column of DEAE-cellulose. Development of this

and sUbsequent columns followed the procedure used

for the initial column. After chromatography on

another 3 x 20 cm column followed by a 3 x 6 cm

40

column, the ferredoxin was contained in 85 ml of

buffer. Ammonium sulfate (0.5 gram per ml) was

then added. Although the procedure of Tagawa and

Arnon (41) for isolation of spinach ferredoxin

specifies 0.6 gram ammonium sulfate per ml, it was

found that this higher concentration of ammonium

sulfate caused partial precipitation of Leucaena

glauca ferredoxin, whereas when 0.5 gram per m1 was

used no ferredoxin could be detected in the precip­

itate. After removal of the precipitate by

centrifugation, the supernatant solution was applied,

directly and without dilution, to a 1 x 15 cm

column of DEAE-ce11u1ose. Passage of 200 ml of 0.1

M Tris buffer containing 0.5 grams ammonium sulfate

per mI. through the column yielded a yellow effluent

which was discarded. Ferredoxin was eluted in 35 ml

of 1 M Tris and applied to a 4 x 70 cm column of

Sephad~x G-75 which was equilibrated and developed

with 0.1 M Tris buffer. Fractions of 16 ml were

collected. The UV and visible spectra of the

effluent fractions were determined by use of the

Beckman DK-2A ratio recording spectrophotometer.

Eight fractions, containing the majority of the

ferredoxin, were combined. Ammonium sulfate (0.5

gram per ml) was added and ferredoxin crystallized

readily. The crystals and supernatant solution were

41

applied to a 1 x 8 cm DEAE-cellulose column and the

ferredoxin was eluted in 15 ml of 1.44 M Tris

buffer of pH 8.6.

2. Preparation of the Carboxymethylcysteine Derivative:

The method of Crestfield, Moore and Stein (43)

was used, with modification of the volume of the

reaction mixture and the time of reduction. To 189

mg of native ferredoxin, dissolved in 8 ml of 1.44

M Tris buffer of pH 8.6, were added 6.4 grams of

twice recrystallized urea and 0.8 ml of a solution

containing 50 mg of EDTA per mI. The resulting

solution was flushed with nitrogen for 10 minutes,

after which 0.27 ml of mercaptoethanol was added.

After an additional 20 minutes of flushing with

nitrogen, the reaction vessel was sealed. The

reduction was permitted to proceed at room tempera­

ture for 24 hours, during which the color of the

solution changed from dark red to pale yellow.

Carboxymethylation of the cysteine residues was

accomplished by addition of 947 mg of thrice

recrystallized iodoacetic acid. The reaction

proceeded under nitrogen in the dark for 15 minutes

after which the reaction mixture was applied to a

1.5 x 83 cm column of Sephadex G-25 in 0.02 M

NH4HC03' Fractions of 3.4 ml were collected. The

carboxymethylated protein was located by its

42

absorbance at 277 mlJ. The appropriate fractions

were combined and dialyzed against 3 liters of

deionized water.

3. Methods of Hydrolysis by Endopeptidases:

a. Trypsin

TPCK-trypsin, in which chymotryptic activity

is inhibited, was used. A solution of 83 mg of

CMFd in 3 m1 of deionized water was adjusted to

pH 8.0 by addition of 1 M NaOH. TPCK-trypsin,

equivalent to 1% of the weight of sUbstrate, was

added initially and after 1 hour. The pH was

maintained between 7.9 and 8.2 by periodic addition

of 1 M NaOH. The hydrolysis proceeded at 40 0 for

4 hours and was stopped by application of the

digest to a column of AG 1-X2.

b. Chymotrypsin

A solution of 79 mg of CMFd in 3 to 4 m1 of

deionized water was adjusted to pH 8.0 by addition

of 1 M NaOH. Initially and after 3 hours, 1% (by

weight) portions of TLCK-chymotrypsin (in which

tryptic activity is inhibited) were added. After

9 hours of hydrolysis at pH 8.0 to 8.2 and 400 ,glacial acetic acid was added to lower the pH to

3.0.

c. Bacillus subti1is Neutral Protease

An aliquot of an acetone suspension of the

43

enzyme was centrifuged at 10,000 rpm for 15

minutes at -100 C. The precipitate was dissolved

in 1 ml of buffer of pH 7.1 which was 2 roM in

cacodylate and 2 roM in CaC12, and dialyzed for 4

hours against 2 liters of the same bUffer. The

enzyme concentration was calculated from the

absorbance at 280 m~. Approximately 1 ~mole of

peptide, dissolved in 0.02 M ammonium acetate

buffer of pH 7.1, was hydrolyzed by 2% (by weight)

B. subtilis neutral protease at 300 for times

ranging from 45 minutes to 5 hours. Reaction was

stopped by addition of glacial acetic acid to

lower the pH to 4.5, thus inactivating the enzyme.

d. Thermolysin

This enzyme was used to hydrolyze one of the

large chymotryptic peptides so as to produce

fragments of sizes suitable for sequence determi­

nation. The peptide (0.7 ~mole) was dissolved in

1 ml of 0.2 M ammonium acetate buffer of pH 8.2.

Initially and after two hours, portions of

thermolysin equal to 2% of the weight of peptide

were added. The hydrolysis proceeded at 400 for

7 hours after which the digest was dried in a

stream of nitrogen.

4. Methods of Peptide Purification:

a. Chromatographic Separation of Tryptic Peptides

44

The tryptic digest was applied to a column

(2.0 x 107 cm) of AG l-X2 which had been equili­

brated with starting buffer. The column was

developed with 360 ml pyridine-collidine-acetic

acid buffer of pH 7.8 (44), a linear gradient

(1800 ml) from the same buffer to 0.4 M acetic

acid, 500 ml of 0.4 M acetic acid, and finally

300 ml of 50% acetic acid. Fractions of

approximately 6 ml were collected and 0.1 ml was

taken from each fraction for ninhydrin assay (45).

b. Chromatographic Separation of Chymotryptic Peptides

The chymotryptic digest, adjusted to pH 3.0,

was applied to a 1.6 x 105 cm column of AG 50-X2,

equilibrated with starting buffer. Elution with

1360 ml of 0.2 M pyridine acetate of pH 3.0 was

followed by a linear gradient (2 liters) from the

same buffer to 2.0 M pyridine acetate of pH 5.0.

The fraction size ranged from 3.6 to 8.2 mI. A

suitable aliquot was taken from each tube for

ninhydrin assay.

c. Further Purification of Tryptic and Chymotryptic

Peptides

The purity of the tryptic and chymotryptic

peptides was determined by amino acid analysis,

paper chromatography in the butanol-acetic acid­

water (4:1:5) solvent system (46), hereafter

45

referred to as Solvent I, and thin layer

electrophoresis on cellulose layers in pyridine

acetate buffers of pH 6.5. Peptides which re­

quired desalting were passed through a 0.9 x 100

cm column of Sephadex G-15 in 0.02 M ammonium

bicarbonate, deionized water, or 0.2 M acetic

acid. Peptide mixtures remaining after the

initial chromatography were resolved by gel

filtration through Sephadex G-15, by chromatography

on columns of AG 50-X2 in pyridine acetate buffers

or AG l-X2 in pyridine collidine acetate and

acetic acid, or by paper chromatography in the

upper phase of Solvent I or, in case of two large

tryptic peptides, in pyridine-isoamyl alcohol-

0.1 M ammonia (6:3:5) (47), hereafter referred to

as Solvent II.

d. Chromatographic Separation of B. subtilis Neutral

Protease Fragments of Tryptic Peptide T-7a

The digest was lyophilized, dissolved in a

small volume of starting bUffer, and applied to

a column-(0.8 x 57 cm) of AG 50-X2. The column

was developed with 100 ml of 0.2 M pyridine

acetate, pH 3.0, followed by a 300 ml linear

gradient from the same buffer to 2.0 M pyridine

acetate, pH 5.0. Fractions of approximately 2 ml

were collected. From each tUbe, 0.05 ml was taken

46

for ninhydrin assay. Fractions contained in each

ninhydrin positive peak were pooled.

e. Partition Chromatographic Separation of B.

subtilis Neutral Protease Fragments of

Chymotryptic Peptide C-10

The digest was dissolved in a few drops of

the upper phase of Solvent I and applied to a

0.6 x 65 cm column of Sephadex G-25F equilibrated

with the lower phase of the same solvent. The

upper phase was used to develop the column.

Fractions of 1 ml were collected. From each

fraction, one large drop was spotted on filter

paper, dried, and sprayed first with ninhydrin

solution and then with Ehrlich reagent for

tryptophan (48) to determine the pooling of

fractions.

f. Chromatographic Separation of Thermolysin

Fragments of Chymotryptic Peptide C-10

The digest was dissolved in starting buffer

and applied to a 0.6 x 90 cm column of AG l-X2.

Elution with a 120 ml linear gradient from

pyridine-collidine acetate buffer of pH 7.8 to

0.4 M acetic acid was followed by 90 ml of 0.4 M

acetic acid, an 80 ml linear gradient from 0.4 M

to 1.0 M acetic acid, 50 ml 1 M acetic acid, and

finally, 50 ml of 50% acetic acid. An aliquot

41

was taken from each tube for ninhydrin assay and

fractions were pooled accordingly.

5. Determination of Amino Acid Composition:

Samples for amino acid analysis were hydrolyzed

in sealed evacuated tubes in glass distilled constant

boiling HCl containing a small amount of phenol to

prevent oxidation of tyrosine (49). Hydrolysis was

carried out at 1050 for 18 to 96 hours, after which

the HCl was evaporated in a stream of nitrogen. The

residue was dissolved in 0.2 M sodium citrate, pH 2.2.

The Beckman/Spinco Model 120 amino acid analyzer (50)

was used for quantitative determination of all amino

acids except tryptophan, which was determined by the

method of Opienska-Blauth et ale (51). In this pro­

cedure, three samples (0.063, 0.090, and 0.135

~moles) of CMFd, were each dissolved in 1 ml deionized

water, and to each were added 2 ml of glacial acetic

acid containing 56 mg ferric iron per ml, followed by

2 ml of concentrated H2S04 (specific gravity 1.84).

The reaction mixtures were shaken, allowed to stand

for 1 hour during which the color developed, and their

absorbances at 545 m~ were read against a blank in the

Coleman Jr. spectrophotometer. The same procedure was

applied to a series of tryptophan samples of varying

concentrations. Thus a standard curve was obtained

from which the amount of tryptophan in the protein

lt8

samples could be determined.

6. Methods of Amino Acid Sequence Determination:

a. Methods of NH2-terminal Analysis

1. Edman Degradation

The phenylisothiocyanate procedure of

Edman (52), as modified by Konigsberg and

Hill (53), was used in a sequential manner

to determine most of the amino acid sequence

of the peptides. The coupling reaction was

carried out at 40° for 2 hours in

N-ethylmorpholine acetate buffer of pH 8.8.

The reaction mixture was dried in a stream

of nitrogen at 40°. After extraction of the

excess phenylisothiocyanate into benzene,

the residue was again dried thoroughly.

Anhydrous trifluoroacetic acid (0.1 to 1 ml)

was added and the cyclization reaction was

allowed to proceed at room temperature for

1 hour. After evaporation of the

trifluoroacetic acid in a stream of nitrogen

at room temperature, the residue was dissolved

in 1 ml O.lt M acetic acid. In some cases this

solution was then maintained at 40° C for 10

minutes (54). The PTH-amino acid was extracted

into two 1 ml portions of benzene and in some

cases was identified by thin layer

chromatography on layers of silica gel G in

the solvent systems of Randerath (55). In a

few instances the PTH-amino acid was hydrolyzed

in vacuo in constant boiling HCl at 1500 C

for 16 hours (56) and the regenerated amino

acid was then identified on the amino acid

analyzer. However, the method used to the

greatest extent was the subtractive procedure

of Konigsberg and Hill (53), whereby the NH2­

terminal residue is identified by its disap­

pearance from the amino acid analysis of the

peptide after Edman degradation. After

extraction of the PTH-amino acid, an aliquot

is taken from the aqueous phase which contains

the residual peptide. This aliquot is

hydrolyzed in HC1 and subjected to amino acid

analysis. In the case of acidic peptides which

were not very soluble in acetic acid, the

aqueous phase was dried at 40 0 C in a stream

of nitrogen and the residue was dissolved in

the N-ethylmorpho1ine buffer for the next

stage coupling reaction before an aliquot was

taken for analysis. This procedure was also

useful when PTH-arginine was present.

PTH-arginine is incompletely extracted from

the aqueous phase (54) and is hydrolyzed by

50

HCl to yield arginine, which thus appears in

the analysis of the residual peptide. How­

ever, when the sample is first dissolved in

the basic N-ethylmorpholine buffer, the

PTH-arginine decomposes and is eliminated as

an artifact in the analysis.

2. Dinitrophenylation

The procedure used was that of Sanger

(57) as modified by Fraenkel-Conrat (58).

The DNP-amino acids were identified by thin

layer chromatography on layers of silica gel

G as described by Randerath (59).

3. Dansylation

The reaction of dansyl chloride (5­

dimethylamino-l-naphthalene sulfonyl chloride)

with amino groups to yield a fluorescent

derivative and the hydrolysis in constant

boiling HCl, which releases the DNS-amino

acid from the NH2-terminal position in a

peptide or protein, were carried out as

described by Gray (60). The DNS-am1no acids

were identified by thin layer chromatography

on layers of silica gel G in the solvent

systems of Deyl and Rosmus (61).

b. Methods of COOH-terminal Analysis

1. Hydrazinolysis

51

The method of Bradbury (62) was used,

with minor modifications. To 0.02-0.03 pmole

peptide were added 26 mg hydrazine sulfate

and 0.2 ml 96% hydrazine. The reaction was

carried out in a sealed, evacuated tube for

16 hours at 60°. After the sample had been

dried in a stream of nitrogen, 1 ml of 0.2 M

sodium citrate buffer of pH 2.2 was added.

The pH was adjusted to 2.2 by dropwise addi­

tion of 6 M HCl, and the sample was then

analyzed directly on the amino acid analyzer.

2. Carboxypeptidase A

The method used for carboxypeptidase A

digestion was a modification of that of

Harris (63). To 1 ml of cold water were

added 10 pI of 1 M NaOH and 1 drop of an

aqueous suspension of the enzyme. The high

pH caused immediate solublization of the

enzyme. Eight pI of 1 M HCl were added to

lower the pH, followed by 0.6 ml of 0.02 M

NH4HC03 (pH 8-9). The concentration of the

enzyme was determined spectrophotometrically

(E~18 mp = 19.4). To the peptide or protein,

dissolved in 0.02 M NH4HC03

, was added an

aliquot of enzyme solution such that the

weight ratio of substrate to enzyme was 20:1.

52

The reaction mixture was incubated at 40°.

At various time intervals, aliquots were

taken and pipetted into 0.2 M sodium citrate

buffer of pH 2.2, thus inactivating the

enzyme. The amino acid analyzer was used to

quantitatively determine the amino acids

released.

3. Carboxypeptidase B

This enzyme was used to release lysine

and arginine from the carboxyl terminal

positions of tryptic peptides. The method

used was similar to that for carboxypeptidase

A, except that solublization is not required.

The weight ratio of enzyme to substrate varied

from 2 to 3%. Digests were carried out in

0.1 M sodium phosphate buffer of pH 8.0, at

40° for 2 to 8 hours.

7. Studies on Ferredoxin from Individual Trees:

Leaves were gathered from ten individual L. glauca

trees at various locations on the island of Oahu - in

Manoa Valley, on Round Top, and on Mount Tantalus.

Ferredoxin was isolated separately from each tree and

carboxymethylated as previously described. After

passage through a column of Sephadex G-25, the CMFd

was divided into three portions, for amino acid

analysis, hydrolysis by carboxypeptidase A, and

hydrolysis by trypsin. Acid hydrolysis and amino

53

acid analysis were performed as previously described.

Carboxypeptidase digests were carried out for five to

ten hours at 40° with an enzyme:substrate ratio of

1:20. Tryptic digests were performed in 0.02 M

NH4HC03

at 40° for one to seven hours with 2% (by

weight) TPCK-trypsin. After drying in a stream of

nitrogen, each tryptic digest was dissolved in a

minimal amount of starting buffer and applied to a

0.6 x 36 cm column of AG l-X2. The column was

developed with 6 ml of pyridine-collidine-acetic acid

buffer of pH 7.8, followed by a 60 ml linear gradient

from the same buffer to 0.4 M acetic aCid, then 6 ml

of 0.4 M acetic aCid, and finally 50% acetic acid.

Fractions of approximately 2 ml were collected, and

an aliquot was taken from each fraction for ninhydrin

assay. Fractions within each ninhydrin-positive peak

were pooled and aliquots were taken for amino acid

analysis.

54III. RESULTS

A. Isolation of Ferr~doxin

The yield of ferredoxin from the purification pro­

cedure described was 20 mg per kg of leaves. A

photomicrograph of ferredoxin crystals is shown in

Figure 10. Figure 11 shows the spectrum of the native

protein. The ratios of absorbances at 463, 420, and

325 m~ to that at 277 m~ are 0.43, 0.49, and 0.65,

respectively, as compared with 0.43, 0.48, and 0.65 for

the corresponding maxima in the spectrum of alfalfa

ferredoxin.

B. Preparation of the Carboxymethylcysteine Derivative

The pattern of elution of the carboxymethylation

reaction mixture from a 1.5 x 84 cm column of Sephadex

G-25F is shown in Figure 12. The yield of carboxymethylated

ferredoxin (CMFd) was 15.6 ~moles, or 100%. The molar

absorptivity index, E, at 277 In~ was 1.59 x 103 cm-l , with

the protein concentration based on amino acid analysis.

C. Amino Acid Composition of Ferredoxin

The results are summarized in Table I. Each value

is the average of those obtained from analyses of duplicate

hydrolysates of CMFd. Half-cystine was determined as

carboxymethylcysteine. The amino acid compositions of

ferredoxins from spinach (13), alfalfa (35), and

Scenedesmus (36) are included in the table for comparison.

Tryptophan determinations carried out according to the

55

Figure 10. Photomicrograph of L. glauca ferredoxin

crystals. The photograph was taken at l500X magnification.

iI

Figure 11. Absorption spectrum of ~. glauca ferredoxin. Measurements were made on

a Beckman DK-2A ratio recording spectrophotometer. The solvent was 0.1 M Tris-Hel buffer

of pH 1.3.

V1~

U)

d10 q- rt> "!d d d 0

3::>N\f8~OS8\f 31\I.l\f13~

.o

o(\Jq-

oen(\J

--

59

Figure 12. Purification of CMFd by gel filtration

through a column of Sephadex G-25. The eu1tion pattern shows

the separation of CMFd, which was detected by its absorbance

at 276 m~, from the carboxymethy1ation reagents after

passage through a 1.5 x 84 cm column of Sephadex G-25F in

0.02 M NH4HC03.

17070 100 135

......... -- ~-- ..... -.- ~ ----- -- -----

-

f-

~

-

l-

• CMFd\J

.~4 • • •

1.4

0.2

0.0

35

1.2::L

Ew \.0l"-N

r-c::x: 0.8wuz

0.6c::x:CD0::0(f)

0.4CD<t

EFFLUENT ml.

61TABLE I

Amino Acid Compositions of Ferredoxins fromL. glauca and Other Plants

AminoL. glauca

from from Alfalfa Spinach ScenedesmusAcid Analyses Sequence

residuesLysine 5.10 5 5 4 4

Histidine 0.91 1 2 1 1

Arginine 1.91 2 1 1 1

Aspartic acid 10.55 10+ 8 11 12

Asparagine 0 1 2 0

Threonine 3.94 4 6 8 10

Serine 6.72 7 8 7 8

Glutamic acid 16.5 12+ 13 9 6

Glutamine 4 3 4 4

Proline 4.50 4+ 3' 4 4

Glycine 6.87 6+ 7 6 7

Alanine 6.91 7 9 9 10

Half-cystine 4.98 5 5 5 6

Valine 6.61 6+ 9 7 5

Methionine 0.0 0 0 0 1

Isoleucine 3.87 4 4 4 3

Leucine 8.98 9+ 6 8 7

Tyrosine 2.73 3 4 4 4

Phenylalanine 3.11 3 2 2 3

Tryptophan 0.92 1 1 1 0

Total 96 97 97 96

62

method of Opienska-B1auth et a1. (51) gave the data in

Table II, which includes values obtained with tryptophan

samples of known concentration and the results from three

samples of CMFd. The value obtained for CMFd, 0.92

residue of tryptophan per molecule, was an average of the

results of the three determinations.

D. Amino and Carboxyl Terminal Residues

Dansy1ation (60) yielded only DNS-a1anine, which was

identified by thin layer chromatography on silica gel G,

in the solvent system, benzene, 16: pyridine, 4: glacial

acetic acid, 1, of Dey1 and Rosmus (61). Hydrolysis of

CMFd by 5% (by weight) carboxypeptidase A at 40° was

carried out for at total of five hours, with a1iquots

taken periodically. The results are shown in Figure 13.

E. Tryptic Peptides

1. Initial Chromatography

The elution pattern of the tryptic digest from

AG 1-X2 is shown in Figure 14. Peptide numbers were

assigned according to the order of the peptides in

the sequence (rather than according to elution

position); thus T-1 is the amino terminal tryptic

peptide. The fractions represented by each

ninhydrin-positive peak were pooled as indicated.

The results of paper chromatography, paper electro­

phoresis, and amino acid analyses indicated that

peptides T-1, T-3a, T-3b, T-5, T-7a, and T-7b did not

TABLE II

Tryptophan Content of L. glauca Ferredoxin

Sample llgrams pmolesAbsorbance at llgrams

545 mu vs. blank tryptophan(calculated)

Molestryptophanper mole

of protein

Tryptophan 0.00 -----

" 4.08 0.008

" 8.16 0.018

" 16.3 0.040

" 24.5 0.060

" 40.8 0.096

CMFd 0.0632 0.028 11.8 1.00

" 0.0900 0.035 14.8 0.88

" 0.135 0.052 22.0 0.88

0\W

Figure 13. Action of carboxypeptidase A on CMFd. The amount of each amino acid

released by the exopeptidase is plotted against time. Leucine (--0- ---- ~ ---),

threonine ( --0 0), glycine (- .. - - -), and alanine

~ -A- - - 04- - -) were released from the protein by carboxypeptidase A.

'".l=

.<1o a

\ ~ 1I 'w

\~ ~I 'z,-u

lz

w \~~

1«,.-J -<!>\

en,« Q)-~ ~~

:::s

~ ~ 0 c

~, r() E

\ \-

~\w

\ \ :iE

\ \t-

, 1,.~ LO

~\ \~~ \"

"~ "-,.., ~,

\ "4.-_---0°0 0

0 LO 0• .

00

N131.0tkJ .::10 310~ ~3d S310~

Figure 14. The elution pattern of the tr~ptic digest of CMFd from AG l-X2. CMFd~

was hydrolyzed by trypsin and the resulting peptide mixture was chromatographed on a

2.0 x 107 cm column of AG l-X2 with pyridine-collidine-acetic acid buffers. The peptides

were detected in the eluate by ninhydrin analysis of an aliquot from each fraction. The

absorbance at 570 mp represents the color obtained by ninhydrin reaction. The bars on

the abscissa indicate the fractions pooled.

0"10"1

o -fI"---~I-

.a,.------I-

EIt)

l- t!- t-Z 0 ZW W- ~C ..J« I.L0: I.LC)

LLI

~---===============-:-u +~

~ ~:I:

-oe~

o

CX)r--::I:0. _

~ t!-« '}I~====-----­o

C1- f's;:::============::;:===:;::=--_1-.s.:::......LIt)---~o::-----~It):------:(\J.----~It)~

~ a ~ d d(rfwOL9) 30N'fe~OS8'f 3/\ I.L'f13 ~

68

require further purification.

-2~ Further P~rification of Tryptic Peptides

Peptide T-2 was first desalted and then

chromatographed on a 1.0 x 58 cm column of AG 50-X2

(200-400 mesh). Elution with a 300 ml linear gradient

from 0.2 M pyridine acetate of pH 3.0 to 2.0 M

pyridine acetate of pH 5.0 yielded two ninhydrin

positive peaks, as shown in Figure 15. The major

component was peptide T-2, whereas the minor com­

ponent was found to be identical in composition to

T-l.

The material eluted from the original AG l-X2

column with 50% acetic acid, when concentrated to

dryness in vacuo and subjected to partition

chromatography on Whatman 3 MM paper in Solvent II,

gave four ninhydrin positive bands. The Rf values

were 0.15, 0.30, 0.37, and 0.45, and the visually

estimated relative intensities of ninhydrin color

were +5, +4, +5, and +3, respectively. After elution

of the peptides from the paper in 0.02 M NH4HC03

,

aliquots were taken for amino acid analysis. The

results showed that the two slower moving bands were

peptide T-4, whereas the two bands of higher Rf

corresponded to peptide T-6.

3. Amino Acid Composition

Table III shows the amino acid compositions,

69

Figure 15. Purification of tryptic peptide T-2 on

AG 50-X2. This peptide was separated from a lesser amount

of T-l by ion exchange chromatography on a 1.0 x 58 cm

column of AG 50-X2. Elution was performed with a 300 ml

linear gradient from 0.2 M pyridine acetate of pH 3.0 to

2.0 M pyridine acetate of pH 5.0. Ninhydrin assay was used

to detect the peptides.

E 0.4o~lO

~wuZ<t~ 0.3oenCD«w>-....<t-ILaJa::

180

T-2

200 220 240 260

EFFLUENT VOLUME (mU

71

TABLE III

Amino Acid Composition and Properties of Tryptic Peptides

Amino Acid

Lysine

Histidine

Arginine

CM-cysteine

Aspartic acid

Threonine

Serine

Glutamic acid

Proline

Glycine

Alanine

Valine

Isoleucine

Leucine

Tyrosine

Phenylalanine

Tryptophanb

Total residues

Yield (JlIDoles)

Purificationd

Rfe

fElectrophoresis

T-l

0.95(1)

0.22

0.20

1.00(1)

3

3.15

0.28

-0.61

T-2

residuesa0.92(1)

1.08(1)

2

3.88

AG 50

0.16

-0.73

T-3a

0.90(1)

0.97(1)

0.91(1)

2.24(2)

1.04(1)

1.94(2)

8

1.61

0.34

0.0

72TABLE III, continued

Amino Acid T-3b T-4 T-5

residuesaLysine 0.94(1) 1.10(1)

Histidine

Arginine 0.9(1)

eM-cysteine 1.5(2)c 1.97(2)

Aspartic acid 1. 08(1) 3.9(4)

Threonine 1.01(1) 0.2

Serine 1.0(1) 2.71(3)

Glutamic acid 5.3(5)

Proline 1.21(1) 2.0(2)

Glycine 1.10(1) 1.3(1) 2.17(2)

Alanine 0.77(1) 1.1(1) 2.06(2)

Valine 0.95(1) 1.1(1)

Isoleucine 2.0(2)c

Leucine 1.20(1) 3.1(3)

Tyrosine 1.9(2)

Phenylalanine 1.0(1)

Tryptophanb

Total residues 8 26 10

Yield (flmo1es) 1.54 2.41 3.75

Purificationd PC

R e 0.21 (0.30) 0.02f

E1ectrophoresis f 0.0 ----- +0.29

73

TABLE III, continued

Amino Acid T-6 T-7a T-7b

residuesaLysine 1.00(1) 0.91(1)

Histidine 0.93(1) 0.85(1)

Arginine 0.8(1)

CM-cysteine 1.1(1)

Aspartic acid 5.2(5) 1.01(1) 1.11(1)

Threonine 0.8(1) 1.77(2) 1. 79(2)

Serine 1.5(2) 0.98(1) 1.11 (1)

Glutamic acid 6.9(7) 3.76(4) 4.10(4)

Proline 1.2(1)

Glycine 1.8(2) 0.88(1)

Alanine 1.8(2) 0.98(1)

Valine 2.1(2) 1.58(2)c 1.83(2)

Isoleucine 1.3(1) 1.09(1) 0.91(1)

Leucine 3.7(4) 1.18(1) 1.02(1)

Tyrosine 1.3(1)

Phenylalanine 1.2(1)

Tryptophanb (1)

Total residues 32 15 15

Yield (J-lmo1es) 2.75 2.95 2.01

Purificationd PC

R e (0.37) 0.15 0.07fE1ectrophoresis f ----- +0.47 +0.14

TABLE III, continued

aResidue values calculated from amino acid analysis data

are given, followed by the assumed number of residues in

parentheses.

bDetermined by Ehrlich reaction.

cValues obtained after 96 hours of hydrolysis.

dpC refers to paper chromatography in Solvent II.

AG 50 refers to ion exchange chromatography on columns of

AG 50-X2.

eparentheses indicate Solvent II. All other values refer

to Solvent I.

fMovement at pH 6.5 relative to alanine (=0), lysine (=-1),

and glutamic acid (=+1).

75

yields, Rf's, electrophoretic characteristics, and

purification procedures for the tryptic peptides.

Tryptophan was determined qualitatively by Ehrlich

reaction (48) since only one residue of this amino

acid is present in the protein. All composition data

were from 22 hour hydrolysates, except for those

indicated in the table to be values obtained after 96

hours of hydrolysis. Amino acids present in analyses

in amounts less than 0.15 residue are not included in

the table.

4. Amino Acid Sequence

The data are expressed in table form, rather than

the usual text form, wherever the ease of understand­

ing the data and the interpretation thereof could be

enhanced by this form of presentation. The amino

acid sequence of each peptide is given, with the

residues of which the sequence has been determined

separated by hyphens, whereas unsequenced sections

are enclosed in parentheses with the residues therein

separated by commas.

The methods which were used are indicated in the

sequence given for each peptide. Arrows to the left

and right represent the use of carboxypeptidase A or

B and Edman degradation, respectively. A broken

arrow indicates dinitrophenylation. Dansylation and

hydrazinolysis are represented by single and double

underlinings, respectively. Aligned immediately

76

below each peptide sequence, the amino acid composition

is given for comparison with the sequence data. For

those amino acids which appear more than once in the

same peptide, the residue value is given to correspond

only with the first appearance of that amino acid.

Step 1, Step 2, etc. refer to the results of Edman

degradation. A dash in this data indicates that the

corresponding amino acid was not determined in that

step. The amino acid disappearing at each step is

indicated by underlining of the residue value. Values

less than 0.1 are not included except as required to

show the disappearance of a residue. Carboxypeptidase

and hydrazinolysis data are expressed as residues of

amino acid liberated, based on the amount of peptide

used.

T-l: ~-Phe-kYs

Composition: 1.1 1.0 1.0

Step 1: 0.3 1.1 1.0

Carboxypeptidase B (2 hours): Lys 0.56.

T-2: Val-Lys-'7

Composition: 1".1 0.9

Step 1: 0.0 1.0

T-3a: Leu-Leu-Thr-Pro-As~-Gl~Pro-Lys--,.~ ~ ---'7

Composition: 1.9 0.9 2.2 1.0 1.0 0.9

Step 1:

Step 2:

1.0

0.2

1.0 1.9 1.1 1.1

1.0 1.9 1.0 1.1

--

77

Step 3: 0.1 0.2 2.0 1.0 1.0

Step 4 : 1.0 0.9 1.1

Step 5: 1.0 0.3 1.0

Step 6: 1.0 0.4 0.6

Direct identification in 5th step: PTH-Asp.

Carboxypeptidase B (20 hours): no residues liberated.

Lysine was placed at the COOH-termina1 according

to the specificity of trypsin.

T-3b: Val-Leu-Thr-Pro-As~-G1y-A1a-kYs~ -., --""'7~ -

Composition: 1.0 1.2 1.0 1.2 1.1 1.1 0.8 0.9

Step 1: 0.0 1.0 0.9 1.1 1.0 1.2 0.8

Step 2: 0.2 0.9 1.1 1.0 1.1 0.9

Step 3: 0.2 1.0 1.0 1.2 0.8

Step 4: 0.2 0.3 1.0 1.2 0.8

Step 5: 0.4 1.1 0.9

Direct identification in 5th step: PTH-Asp.

Carboxypeptidase B (8 hours): Lys 0.54.

--Hydrazino1ysis of residual peptide: Ala 0.97.

T-4: ~-Phe-G1x-(CMCYS2,ASx4_,Ser,G1x3+,Pr02,G1Y,

A1a,Va1,I1e2,Leu3,Tyr2)-~rg

Edman degradation - direct

identification in 1st step: PTH-G1u

Dansy1ation: DNS-G1u

Dansy1ation after step 1: DNS-Phe

Dansy1ation after step 2: DNS-G1u

Carboxypeptidase B (2 hours): Arg 0.71

78

T-5: A1a-G1~-Ser-CMCys-Ser-Ser-CMCYS-A1a-G1y-kYs--'? --'7 7 -,~ 7 ----'7

Composition: 2.1 2.2 2.7 2.0 1.1

Step 1: 1.1 2.2 2.8 1.8

Step 2: 1.1 1.3 2.1 2.0

Step 3: 1.0 1.1 1.9 1.8

Step 4 : 1.0 1.1 1.1 1.2

Step 5 : 1.0 1.0 1.2 1.1

Step 6: 0.9 1.1 0.4 0.8

Step 1 : 0.9 1.1 0.2 0.2

Step 8: 0.4 1.0 0.3 0.3

Carboxypeptidase B (1 hours): Lys 0.85.

Hydrazino1ysis of residual peptide: G1y 0.72.

T-6: ~-~-G1X-(CMCys,AsP5,Thr,Ser2,G1U6,Pro,

G1Y2,A1a2,Va1,I1e,Leu3,Tyr,Phe,Trp)-Arg

Step 1: Leu) Identified as free amino acids)

Step 2: Val) after acid hydrolysis of

pheny1thiohydantoin.

Dansy1ation after step 1: DNS-Va1

Dansy1ation after step 2: DNS-G1u

Arginine was placed at the COOH-termina1 according

to the specificity of trypsin.

T-1a: Ser-As~-Va1-Va1-I1e-G1u-Thr-His-LYS-G1u--, ---'7 --;J'~

Composition: 1.0 1.0 1.6 1.1 3.8 1.8 0.9 1.0

Step 1: 0.2 1.0 1.8 1.2 4.1 1.8 0.9 0.6

Step 2: 0.3 1.9 0.9 4.0 1.8

Step 3: 1.1 1.0 4.0 1.9

79Step 1.1 : 0.1 1.0 1.1.1 1.8

Step 5: 0.4 3.9 1.8

T-7a (cont.) Glu-Glu-Leu-Thr-Gly

Composition (cont.) 1.2 0.9

Step 1: 1.0 1.1

Step 2 : 1.0 1.1

Step 3: 1.0 1.1

Step 4 : 1.0 1.1

Step 5 : 1.0 1.3

Values for valine and isoleucine in the first 3

steps are those obtained after 96 hours of hydrolysis.

Direct identification in 2nd step: PTH-Asp

Hydrazinolysis: Gly 0.1.13

The data used to establish the sequence of T-7a are

summarized in Fig. 16.

Hydrolysis of T-7a with Bacillus subtilis neutral

protease (2% by weight) for 1 hour yielded 3 peptides

which were then separated on AG 50W-X2.

Step 5 (direct analysis): 0.0

1.1 1.0

0.9 1.0

0-.9 1.0

1.0 1.0

0.0 1.0

0.2 1.1

ser-As~-Val-Val-Ile-G1U~ ---:;:>' ---::0" -.,..

0.9 1.0 1.9 1.0 1.1

0.1 1.0 1.9

0.4 1.8

T-7a-BSNP-l (59%):

Composition:

Step 1:

Step 2:

Step 3:

Step 4:

80

Figure 16. Summary of evidence establishing the se­

quence of tryptic peptide T-7a. The fragments obtained by

hydrolysis with B. subtilis neutral protease are BSNP-l,

BSNP-2, and BSNP-3. The arrows to the right above the se­

quence represent Edman degradation steps performed on the

fragments, while those below the sequence refer to degrada­

tion of T-7a. The amino terminal sequence of T-7a is the

same as that of BSNP-l; therefore BSNP-l is the amino

terminal fragment of T-7a. BSNP-3 must be the carboxyl

terminal fragment of T-7a, since hydrazinolysis released

only glycine from the latter. The placement of BSNP-2 thus

becomes unequivocal.

-d.z83

2zenm

~zenm

'1111:fa1~",__1

.a(!)

1

1<31~

1

1'§'I

1:f1.!.~

82

The values for Val and Ile in the composition and

in the first 3 steps of Edman degradation were

obtained after 96 hours of hydrolysis.

T-7a-BSNP-2 (76%): Thr-His-Ly~-Glu-Glu-Glu~ --:?' --:7' ~

Composition: 0.9 1.0 0.9 3.3

Step 1: 0.2 0.9 0.8 3.3

Step 2: 0.3 0.6 3.2

Step 3: 0.2 0.2 3.0

Direct identification in 4th step: PTH-Glu

Direct identification in 5th step: PTH-Glu

Direct analysis of residue after 5th step: Glu only

T-7a-BSNP-3 (79%):

Composition:

Step 1:

Step 2:

Leu-Thr-Gly~~

1.0 1.0 1.1

0.2 0.9 1.1

0.1 1.0

T-7b had an amino acid composition identical to that

of T-7a, except for the presence of an alanine residue

in place of the sole glycine residue of T-7a.

T-7b: Ser-(Asp-Val-Val-Ile-Glu-Thr-His-Lys-Glu-----;7

Composition: 1.1 1.1 1.8 0.9 4.1 1.8 0.9 0.9

Step 1: 0.4 1.1 2.0 0.9 4.0 2.0 0.9 0.9

T-7b (cont.): Glu-Glu)-Leu-Thr-Ala~

Composition (cont.):

Step 1:

1.0

1.1

1.0

1.0

83

The values for valine are those obtained after 96

hours of hydrolysis.

Carboxypeptidase A (4 hours, 40°): Ala 0.8S.

T-7b was hydrolyzed by B. subtilis neutral protease

under the same conditions used for T-7a and the

sequence of the alanine containing COOH-terminal

fragment was determined.

T-7b-BSNP-3 (8S%): Leu-Thr-Ala--;i" --;i"

Composition: 1.0 1.0 1.0

Step 1: 0.1 1.0 1.0

Step 2: 0.1 1.0

The sequence of the residues in parentheses in T-7b

is based on analogy with the structure of T-7a.

F. Chymotryptic Peptides

1. Initial Chromatography

The elution pattern of the chymotryptic digest

from AG SO-X2 is shown in Figure 17. Peptide contain­

ing fractions which were pooled are indicated.

Peptides C-2, C-3, C-6, and C-10 did not require

further purification, as indicated by amino acid

analyses.

2. Further Purification of Chymotryptic Peptides

Chromatography of C-l on a column of AG l-X2

yielded a single ninhydrin positive peak which proved

to be the pure peptide. Gel filtration through a

column of Sephadex G-1S in 0.2 M acetic acid separated

peptide C-4 from contaminating material which was

Figure 17. Elution pattern of the chymotryptic digest of CMFd from a column of AG

50-X2. CMFd was hydrolyzed by TLCK-chymotrypsin and the resulting peptides were

fractionated by chromatography on a 1.6 x 105 cm column of AG 50-X2 in pyridine acetate

buffers. The bars on the abscissa indicate the fractions pooled under each ninhydrin

positive peak.

00J::"

C-sC-IO0, I

~ 0t:'v " ·':1t:= ·'1' ~ •

~ 250 350 450 500 850 950 1450 1550 1650U')

l-

I C-4«f'_Q 1\

~ 0.2

~~ 0.1

V \. C-13 C-2 C-3en I C-8 Im« 0 • , • '2350 2450 II 30501950 2150 3150 3250

EFFLUENT VOLUME (m I )

86

present in minor amount. The elution pattern is

shown in Figure 18 a. Peptide C-5 was also obtained

in pure form after gel filtration. The elution pattern

is shown in Figure 18 b. C-7, C-12, and c-14 were

resolved by chromatography on AG l-X2, as shown in

Figure 19. C-7 was separated into two peaks, C-7a and

C-7b. c-8 was obtained in pure form after chromatography

on AG l-X2. C-9 yielded a single asymmetric ninhydrin

positive peak on gel filtration through Sephadex G-15,

as shown in Figure 20 a. However, paper chromatography

of aliquots taken from fractions in the region of

asymmetry showed two sets of ninhydrin positive spots

of different Rf , the relative intensities of which are

plotted in Figure 20 b. Peptide C-9 was then obtained

in pure form by pooling fractions as shown. Peptide

C-ll was separated from minor contaminants by gradient

elution from a column of AG l-X2.

3. Amino Acid Composition

Table IV shows the amino acid compositions, yields,

Rr's, electrophoretic characteristics, and purification

procedures for the chymotryptic peptides. Samples for

amino acid analysis were hydrolyzed for 22 hours, except

where 96 hours of hydrolysis is indicated in the table.

The only Ehrlich positive peptide, C-IO, contains the

sole tryptophan residue of the protein.

4. Amino Acid Sequence

87

Figure 18 a. Purification of peptide c-4 by gel

filtration. This peptide was passed through a 1 x 100 em

column of Sephadex G-15 in 0.2 M acetic acid and effluent

fractions of 1.5 ml were collected. Two ~l from each frac­

tion were spotted onto filter paper and sprayed with a 0.2%

solution of ninhydrin in acetone. The intensity of the blue

color which developed on heating gave the pattern shown when

plotted on a a to +5 scale. The UV spectrum of each fraction

was also measured. The phenylalanine containing c-4 was in

fractions 19 through 24 (indicated as pooled) and tyrosine

containing contaminants in the relative amount of 14% were in

the earlier ninhydrin positive fractions.

+5

a::0--l0 +4u

z-a::o·>- +3:cz-zw> +2- C-4~<t--lW0:

+1

10 15 20 25

FRACTION NUMBER

"~.

89

Figure 18 b. Purification of peptide C-5 by gel

filtration. This peptide was obtained in pure form in the

fractions indicated after passage through a 1 x 218 cm

column of Sephadex G-15 in deionized water. The tyrosine

containing C-5 was readily detected by its absorbance at

280 mu.

0.0 ""--&.-.. .....L-..L-__..L..-L --J

30 40 50

EFFLUENT VOLUME (mil

\DI-'

Figure 19. Resolution of peptides C-7a, C-7b, C-12, and C-14 on a column of AG 1-X2.

The elution pattern was obtained by ninhydrin assay of a1iquots from the effluent frac­

tions from a 1.0 x 55 cm column of AG 1-X2. Elution by a 200 m1 linear gradient from,

pyridine-collidine-acetic acid buffer of pH 7.8 to 0.4 M acetic acid was followed by

elution with acetic acid as indicated. Peaks which are not numbered did not contain

significant amounts of peptide material, as determined by amino acid analysis.

C-12

C-7a

C-7b

C-14

OJ2

IIlIf GRADIENT >1< 0.4 M ACETIC ACID

0.0a

0.04

0.00 I J ! ,-,,,, .£ ,....,C"""'!-, , L - '=' ," I60 100 150 200 260 320 380

E~lO

~I.LJ()Z«CD0::oCJ)

CD«I.LJ>-l-e:(...JIJJ0::

EFFLUENT VOLUME (mU

93

Figure 20 a. Pattern of elution of a mixture containing

peptide C-9 from Sephadex G-15. The effluent fractions from

a 1 x 100 cm column of Sephadex G-15 in 0.2 M acetic acid

were assayed by spotting on filter paper, spraying with

ninhydrin solution, and heating to develop the color. The

intensity of the blue color was then rated visually on a a

to +9 scale and plotted against fraction number. Incomplete

resolution of the components of the peptide mixture is

indicated by the asymmetry of the ascending portion of the

peak.

3020O-----------.l"-----~

10

+1

+9

a:: +80.-J0u +7z-a:: +6c>-::I:

+5z-Z

IJJ +4>-J- +3«.-Jwa:: +2

FRACTION NUMBER

95

Figure 20 b. Distribution of peptide C-9 and a con­

taminant in the effluent fractions from Sephadex G-15, as

determined by paper chromatography. A1iquots from fractions

in the region of asymmetry in the peak shown in Figure 20 a

were chromatographed on Whatman #1 paper in butano1-pyridine­

water (1:1:1). Ninhydrin color was developed as previously

described. Two rows of ninhydrin positive spots of different

Rf values were observed. Their intensities were plotted

separately on a 0 to +5 scale as shown. Fractions 22 through

26 were pooled, as indicated by the bar on the abscissa, to

yield peptide C-9 in pure form.

a::o<3 +5uziE +4o>-:r:6 +3z

w2:: +2I-<{.-JWa:: +1

I \I \

I \

o LL-L...L..::'~=:::J:::::L---J

15 20 25 30

FRACTION NUMBER

97TABLE IV

Amino Acid Composition and Properties of Chymotryptic Peptides

Amino Acid

Lysine

Histidine

CM-cysteine

Aspartic acid

Threonine

Serine

Glutamic acid

Proline

Glycine

Alanine

Valine

Isoleucine

Leucine

Tyrosine

PhenylalaninebTryptophan

C-l

1.04(1)

0.96(1)

C-2

aresidues1.89(2)

0.2!4

0.23

0.12

1.12(1)

0.29

C-3

1.92(2)

1.24(1)

c-!4

0.84(1)

0.98(1)

....-·0.92(1)

0.24

1.10(1)

1.!44(1.!4)

1.08(1)

0.58(0.6)

1.06(1)

Total residues 2

Yield (pmoles) 1.45

Purificationd AG 1

Rf in Solvent I 0.62

Electrophoresise 0.0

3

0.53

0.15

-0.92

5 8

0.40 3.90

G-15

0.22 0.26

-0.86 +0.22

98TABLE IV, continued

Amino Acid C-5 c-6 C-7a C-7b

residuesaLysine

Histidine

Arginine

CM-cysteine 0.92(1)

Aspartic acid 2.02(2) 1.07(1) 0.99(1)

Threonine

Serine 0.19 0.20 0.12

Glutamic acid 1.19(1) 2.92(3) 1. 04 (1)

Proline 1.12(1) 0.93(1) 1.05(1)

Glycine 0.23 0.24 0.96(1) 0.99(1)

Alanine 1.02(1)

Valine 0.90(1)

Isoleucine 0.16 0.81(1) 1.03(1) 0.99(1)

Leucine 0.24 1.86(2) 1.12(1 ) 1.02(1)

Tyrosine 0.85(1) 0.93(1) 0.96(1)

Phenylalanine

Tryptophanb

Total residues 7 8 6 6

Yield (pmo1es) 1.45 2.44 0.59 1.45

Purificationd G-15 AG 1 AG 1

Rf in Solvent I 0.37 0.64 0.64 0.66

E1ectrophoresise +1.11 +0.76 +0.26 +0.39

-- .

99

TABLE IV, continued

Amino Acid c-8 C-9 C-10 C-11

residuesaLysine 0.93(1)

Histidine

Arginine 0.94(1) 0.11

CM-cysteine 1.63(2) 1.09(1)

Aspartic acid 4.94(5)

Threonine

Serine 2.71(3) 1.00(1) 1.70(2)

Glutamic acid 0.27 0.25 6.92(7)

Proline

Glycine 1.29(1) 1.06(1) 2.04(2)

Alanine 1.13(1) 0.98(1) 0.14

Valine 1.33(1) 0.90(1)

Isoleucine 0.88(1)

Leucine 0.23 0.95(1) 2.20(2) 1.10(1)

Tyrosine

Phenylalanine 0.11 1.00(1)

Tryptophanb (1)

Total residues 8 6 22 2

Yield Cj.1mo1es) 1. 32 2.64 2.51 1.78

Purificationd AG 1 G-15 PC

Rf in Solvent I 0.07 0.25 0.13 0.62

E1ectrophoresise +0.10 0.0 +1.00 0.0

100

TABLE IV, continued

Amino Acid C-12 C-13 C-11.1

residuesaLysine 1.06(1)

Histidine 0.85(1)

Arginine 0.88(1)

CM-cysteine 0.81.1(1)

Aspartic acid 0.91(1)

Threonine 0.98(1) 1.90(2) 0.96(1)

Serine 0.20 1.16(1) 0.23

Glutamic acid 4.17(1.1)

Proline 0.98(1)

Glycine 0.15 a.al}(1)

Alanine 2.14(2) 0.51 1.04(1)

Valine 2.01(2)C

Isoleucine 0.92(1)c

Leucine 1.07(1)

Tyrosine 1.01.1(1)

Phenylalanine

Tryptophanb

Total residues 5 16 2

Yield (Jlmo1es) 0.1.10 1.45 0.66

Purificationd AG 1 G-15 AG 1

Rf in Solvent I 0.35 0.05 0.30

E1ectrophoresise +0.43 +0.21 0.0

101

TABLE IV, continued

aResidue values calculated from amino acid analysis data

are given, followed by the assumed number of residues in

parentheses.

bDetermined by Ehrlich reaction.

cValues obtained after 96 hours of hydrolysis.

dpC refers to paper chromatography in Solvent I.

G-15 refers to gel filtration through columns of

Sephadex G-15.

AG 1 refers to ion exchange chromatography on columns

of AG l-X2.

eMovement at pH 6.5 relative to alanine (=0), lysine (=-1),

and glutamic acid (=+1).

0.9 1.8 1.2

0.8 1.0 1.1

0.3 1.0 1.0

0.0 1.0-

102

The sequence data are presented in the same form

as were the data for the tryptic peptides.

C-l: Ala-Phe---;;?

Composition: 1.0 1.0

Step 1: 0.0 1.0

C-2: LY2--Val-~Ys

Composition: 1.9 1.1

Step 1: 1.0 1.0

Carboxypeptidase B (5 hours): Lys 1.04.

C-3: LY~-~LY~-~-Leu

Composition: 1.9 1.9 1.2

Step 1:

Step 2:

Step 3:

Step 4:

c-4: Thr-(Pro,Asp,Gly,Pro/Ala,Lys)-Glu-Phe~ =-==~

Composition: 0.9 1.4 1.0 1.1 0.6 0.8 1.1 1.1

Step 1: 0.1 1.4 1.0 1.1 0.6 0.7 1.1 0.9

Carboxypeptidase A (18 hours): Phe 0.62.

Hydrazinolysis of residual peptide: Glu 1.07.

Electrophoresis, pH 6.5: Acidic; therefore

Asp & Glu.

This portion of the sequence overlaps six residues

of the tryptic peptides, T-3a and T-3b, which differ

in the residue in position 12. This position is

occupied by proline in T-3a and by alanine in T-3b.

This sequence heterogeneity accounts for the non­

integral values of proline and alanine in C~4.

103

Direct identification in 1st step:

Direct identification in 4th step:

Direct identification in 5th step:

Carboxypeptidase A (4 hours):

Hydrazino1ysis of residual peptide:

~-CMCY~-~-ASg.-Ast;.-~-'tyr

1.2 0.9 1.1 2.0 0.9 0.9

C-5:

Composition:

Step 1:

Step 2 :

Step 3 :

Step 4 :

Step 5:

0.3 1.0 0.8 2.3

0.2 0.0 0.8 2.2

0.0 2.1

1.2

0.6

1.0 0.6

1.1 0.9

0.8 1.1

0.9 0.9

1.0 1.0

PTH-G1u

PTH-Asp

PTH-Asp

Tyr 1.0

Val 0.79

C-6: I1e-LeU-As~-G1n-A1a-G1u-G1u-LeU--;7~ -,~~-,

Composition: 0.8 1.9 1.1 2.9 1.0

Step 1: 0.1 1.9 1.1 3.0 1.1

Step 2: 1.1 1.0 2.9 1.0

Step 3 : 1.0 0.4 2.9 1.0

Step 4 : 1.0 0.2 2.2 1.0

Step 5 : 0.9 2.1 0.1

Step 6: 0.9 1.1

Step 7: 1.0 0.0

Direct identification in 3rd step: PTH-Asp

Direct identification in 4th step: PTH-G1n

Direct identification in 6th step: PTH-G1u

104

Direct identification in 7th step: PTH-Glu

C-7a:

Composition:

Step 1:

Step 2:

Step 3:

Step 4:

Step 5:

Gl~-Ile-Glu-Leu-Pro-Tyr-, -r ---r~1.0 1.0 1.0 1.1 0.9 0.9

0.1 1.0 1.0 1.1 1.0 0.9

0.1 1.1 1.0 1.0 0.9

0.3 1.1 1.0 0.9

0.1 0.2 1.0 1.0

0.2 1.0

Direct identification in 3rd step: PTH-Glu

C-7b:

Composition:

Step 1:

Step 2:

Step 3:

Step 4:

Gl~-~-As~-~-Pro-Tyr

1.0 1.0 1.0 1.0 1.1 1.0

0.2 1.0 0.9 1.0 1.0 1.0

0.1 0.1 1.0 1.0 1.0 1.0

0.2 1.0 0.9 1.1

0.2 1.0 1.0

Direct identification in 3rd step: PTH-Asp

Hydrazinolysis: Tyr 0.81

C-8: ser-CMCYS-Ar~-(Ala)Gly,Ser,CMCys,Ser)~ 7

Composition: 2.7 1.6 0.9 1.1 1.3

Step 1 : 1.8 2.0 0.8 1.1 1.1

Step 2: 1.9 0.9 0.7 1.0 1.2

Step 3: 2.0 0.9 0.2 0.8 1.2

Since the remaining five residues are contained in

tryptic peptide T-5, of which the sequence has been

determined, no further studies were performed on c-8.

C-9: Ser-(CMCys,Ala,Gly,Lys)-Leu~ .........

105

Composition: 1.0 1.1 1.0 1.1 0.9 1.0

Step 1: 0.3 1.0 1.0 1.1 0.7 0.9

Hydrazinolysis: Leu 0.24

Since the first five residues of this peptide are

contained in peptide T-5 and have been sequenced in

that peptide, no further studies were performed on

C-9.

C-lO: Val-Glu-Gly-Asp-Leu-Asp-Gln-Ser-Asp-Gln-Ser-Phe­---;:;-

Comp: 1.3 6.9 2.0 4.9 2.2 1.7 1.0

Step 1: 0.4 7.1 2.2 4.8 2.1 1.8 1.0

C-lO (cont): Leu-Asp-Asp-Glu-Gln-Ile-Glu-Glu-~-!fp

Composition (cont.) 0.9 (1)

Step 1: 1.1 (1)

Carboxypeptidase A (16 hours): Trp 0.54

Hydrazinolysis of residual peptide: Gly 0.84

Thus C-lO has valine as its amino-terminal residue

and -Gly-Trp as its carboxyl-terminal sequence. The

data used to establish the sequence of C-lO are sum­

marized in Fig. 21. The internal amino acid sequence

was determined by studies on fragments obtained by

enzymatic hydrolyses of C-lO. C-lO was digested with

~. subtilis neutral protease for 5 hours and the re­

sulting fragments were resolved by partition chromato­

graphy on a 0.6 x 65 cm column of Sephadex G-25

equilibrated with the lower phase of Solvent I.

Figure 21. Summary of the evidence establishing the sequence of peptide C-IO.

Peptides Th-l and BSNP-l both have valine as the amino terminal residue, as does C-IO,

and are therefore placed at the amino terminal of C-IO. BSNP-l provides an overlap,

showing that Th-l is followed by Th-2 and Th-3. Since the tryptophan containing peptide,

BSNP-2, must represent the carboxyl terminal portion of C-IO, the placement of Th-4 is

established. Edman degradation, hydrazinolysis, and the use of carboxypeptidase A,

indicated by arrows to the right, double underlining, and arrows to the left, respec­

tively, established the remainder of the sequence.

I-'o0'\

Q)~

a.~

Q)

Cf)

~re-l

I..

rt')~r (\J &11

I (\JI

II

.r:. I ~ra..

:::J-I- ,s::::.

ZC)

l-I

en I

~rCD ~r

I

I

- ~r~r

Ia.. I

z ~r

c:-en

C)I

CDI

~II~rI

I V

-~ ir

I- I

~

C),s::::.

l-I l-

I

~r irI-

I

~r ~r

108

Elution was performed with the upper phase of the

same solvent system. Two peptides, accounting for

15 of the 22 residues, were obtained in pure form.

C-lO-BSNP-l (45%):

Val-(Glu,Gly,Asp,Leu,Asp,Gln,Ser,Asp,Gln)~

Composition: 0.8 3.2 1.1 3.0 1.1 1.0

Step 1: 0.1 3.3 1.2 2.7 0.6 1.1

Thus C-lO-BSNP-l has valine at its amino-terminus

and is therefore the amino-terminal fragment of C-lO.

C-lO-BSNP-2 (13%): Ile-Glu-(Glu,Gly,Trp)~-r

Composition: 0.9 2.1 1.0 (1)

Step 1: 0.2 2.1 1.0 (1)

Direct identification in 2nd step: PTH-Glu

The presence of tryptophan.in this peptide was

shown by a positive Ehrlich reaction.

The residual peptide after step 2 was sUbjected to

thin layer electrophoresis at pH 6.5 and found to be

acidic. Thus the remaining Glx residue in this peptide

is glutamic acid and not glutamine. Since -Gly-Trp has

been shown to be the carboxyl-terminal sequence of C-lO,

the sequence of C-lO-BSNP-2 is Ile-Glu-Glu-Gly-Trp.

The remainder of the sequence of C-lO was obtained

from studies on peptides formed by the action of

thermolysin in a 7 hour digest. These peptides were

resolved by chromatography on a 0.6 x 90 cm column of

AG l-X2. The elution pattern is shown in Fig. 22.

I-'o\0

Figure 22. Elution pattern of peptides from the thermolysin digest of chymotryptic

peptide C-lO. The pattern was obtained by ninhydrin assay of portions of the effluent

from a 0.6 x 90 cm col~n of AG l-X2. Elution was· performed as follows:, 120 ml linear

gradient from pyridine collidine acetate of pH 7.8 to 0.4 M acetic acid; 140 ml of 0.4 M

acetic acid; 80 ml linear gradient from 0.4 M to 1.0 M acetic acid; 50 ml of 1.0 M acetic

acid; and 40 ml of 50% acetic acid. Th-l and Th-2 were eluted in 0.4 M acetic acid, Th-3

and Th-4 were eluted in 1.0 M acetic acid, and the residual C-lO was eluted in 50% acetic

acid. The bars on the abscissa indicate the fractions pooled under each ninhydrin positive

peak.

E-

....ZIJJ::>...JIJ­IJ­W

o~

-?--------~

I-

oT.--_-------0.--.. 0

~

o~

I _---------

~ -----_---....1 ~

-o•o

It)

od

(rfWOL9) 3~N\i8~OS8" 31\1J."13~

III

0.0 2.0 2.1 2.1 0.9

1.2 2.0 1.8 0.9

1.0 1.3 1.7 0.8

0.9 1.2 1.1 0.9

0.4 1.1 1.0 0.8

0.2 0.3 1.0 1.0

Leu-(Asp,G1n,Ser,Asp,G1n,Ser)---,1.0 1.8 2.3 1.8

0.2 2.0 2.3 1.7

LeU-AS~-G1n~Ser-As~-G1n-(Ser,Phe)---, ---, ---, -r1.0 1.9 2.1 2.0 0.9

Hydrazino1ysis:

C-10-Th-2 (37%):

Composition:

Step 1:

C-10-Th-3 (34%) :

Composition:

Step 1:

Step 2:

Step 3:

Step 4 :

Step 5:

Step 6:

C-10-Th-1 (71%): Va1-G1u-G1y-Asp-r --'7

Composition: 1.0 1.0 1.1 1.0

Step 1: 0.1 1.0 1.1 1.0

Step 2: 0.3 1.0 1.0

Direct identification in 2nd step: PTH-G1u

Asp 0.49

Direct identification in 2nd step: PTH-Asp

Direct identification in 3rd step: PTH-G1n

Direct identification in 5th step: PTH-Asp

Direct identification in 6th step: PTH-G1n

C-10-Th-4 (29%): Leu-As~-As~-G1u-G1n~ ----::r'

Composition: 0.9 2.0 2.1

Step 1: 0.1 2.0 2.1

Step 2: 1.2 2.0

Step 3: 0.3 2.0

112

Step 4 : Direct analysis of the residue yielded GIn,

0.80 residue.

Direct identification in step 2: PTH-Asp

Direct identification in step 3: PTH-Asp

Direct identification in step 4 : PTH-Glu

C-ll: Val-Leu---,Composition: 0.9 1.1

Dinitrophenylation: DNP-Val

C-12:

Composition:

Step 1:

Step 2:

Step 3:

Step 4:

Thr-CMCys-Ala-Ala-Tyr---'7 7 ---"7 ---,

1.0 0.8 2.1 1.0

0.0 1.0 2.0 1.0

0.2 1.9 1.1

Ll 0.9

0.0 1.0

C-13: ~-Ar~-(Ser,Asp,val,val,lle,GlU,Thr,His,

Composition: 1.0 0.9 1.2 0.9 2.0 0.9 4.2 1.9 0.9

Step 1:

Step 2:

0.3 0.8 1.1 1.0 1.9

0.3 1.0 1.0 1.7

LO 4.3 1.8 0.9

1.0 4.3 1.9 0.8

C-13 (cont.): Lys,Glu,Glu,Glu,Leu,Thr,Gly/Ala)

Composition (cont.):

Step 1 (cont.):

Step 2 (cont.):

1.1

0.9

0.8

1.1

1.2

1.2

0.8 0.5

0.8 0.5

0.8 0.5

The values for Val and lIe are those obtained after

96 hours of hydrolysis.

Position 96 is one of the points of sequence

heterogeneity, as shown in the studies on tryptic

113

peptides T-7a and T-7b. This accounts for the

fractional residue values for glycine and alanine in

C-13. The remaining portion of the sequence was

determined in studies on T-7a and T-7b.

C-14: Thr-Ala~

Composition: 1.0 1.0

Step 1: 0.0 1.0

G. Complete Amino Acid Sequence of Leucaena glauca Ferredoxin

The complete sequence was deduced primarily from the

structures of the tryptic and chymotryptic peptides and

the overlaps among them. Analogy with the sequences of

spinach (14) and alfalfa (35) ferredoxins was used to

establish the position of c-6 with respect to C-7a and

C-7b, and of C-11 with respect to C-12. Figure 23 shows

the complete sequence and the positions of the tryptic and

chymotryptic peptides. The four positions of heterogeneity,

positions 6,12,33, and 96, are enclosed in rectangles. The

correlation between residues at positions 6 and 12 is known,

since peptide T-3a contains leucine and proline, whereas

peptide T-3b contains valine and alanine, in these two

positions, respectively.

H. Distribution of Sequence Heterogeneity among the L. glauca

population

The individual trees yielded from 1 to 5.6 kg of

leaves. The recovery of ferredoxin ranged from 30 to 100%

of the previous value of 20 mg per kg of leaves. Trees

with a large proportion of new leaf growth yielded higher

Figure 23. Tryptic and chymotryptic peptides arranged in order in L. glauca

ferredoxin. The numbers of the tryptic peptides are given above, and those of the

. chymotryptic peptides, below, the amino acid sequence. The residues at the points of

heterogeneity are enclosed in rectangles.

I-'I-'~

:r-I~T.2-t T-3g

Ia-Phe-Lys-Va1-LY~LeUl-Thr- Pro-Asp-Glywrol-l..¥sLGlu-Phe-Glu-Cys-Pro-Asp-Asp-Va1-TYfrIle-Leu-1 ~ C- 2 ~1YgJJ 10 ~ 1 20

C-I-t-C-3 C-4 C-5----- ~--

T- 4 t T-5 1Asp-G1n-Alo-GIU-GIU-~~1GIY-Ilet~LeU-Pro-TYlSer- Cys-Arg-:~-G,y.ser.Cys-SelSer-Cys- AIa-GIy- Lys-

c-s t c- 7g t c-a t-- c-g

t T- 6

;~ ivaI-Glu-GIY-Asp-Leu-ASP-Gln-Ser-Asp-~~-ser-Phe-LeU-Asp-ASP-GIU-Gln-I le-GIu-~t'IY-Trp,Val- L.eUT

-4 c -10 +-C-II-1

r T-7 g •IThr.cYS-AIO-Ala-TYrt-~~- Arg-Ser-Asp-VoI-VaI-Ile-Glu-Thr-His-':~GIU-GIU-G Iu-LeUIThr-~i~I

~C-14C-12 C-13 JI

116

recoveries of ferredoxin. Carboxypeptidase A released

alanine, glycine, threonine, and leucine from the

carboxymethylated protein in each case. The chromatography

of the tryptic digests on AG l-X2 yielded five ninhydrin

positive peaks. A typical elution pattern is shown in

Figure 24. The results of amino acid analysis showed

that the first peak contained peptides T-l and T-2; the

second peak was a mixture of T-3a and T-3b; the third peak

contained only peptide T-5; the two carboxyl terminal

peptides, T-7a and T-7b, were eluted together as the fourth

peak; and finally, peptides T-4 and T-6 were eluted in 50%

acetic acid. No evidence for heterogeneity other than

that found in the sequence studies was evident in the

amino acid composition data obtained from the tryptic

peptides of the ferredoxin of the individual trees. The

amino acid composition of peak T-3 (peptides T-3a and T-3b)

showed fractional residue values for leucine, valine,

alanine, and proline in each of the ten trees studied.

Thus, heterogeneity exists in positions 6 and 12 in each

case. Also, T-7 (peptides T-7a and T-7b) invariably

contained both glycine and alanine, in fractional residue

amounts, confirming that heterogeneity exists in position

96 in the ferredoxin of each of the ten individual trees,

as was indicated by the data obtained in the

carboxypeptidase studies. The results of the amino acid

analyses of T-3 and T-7 from the individual trees are

summarized in Table V.

Figure 24. Elution pattern of tryptic peptides of the ferredoxin of Tree 1 from

AG l-X2. The 0.6 x 36 cm column was developed with 6 ml of pyridine collidine acetate

of pH 7.8, a 60 ml linear gradient from this buffer to 0.4 M acetic acid, 6 ml of 0.4 M

acetic acid, and finally 20 ml of 50% acetic acid. The elution pattern was obtained by

ninhydrin assay of an aliquot from each fraction. The pooling of fractions is indicated

by the bars on the abscissa.

......

......~

•-E...Z

011I

·3~

­•°0

CD

N•o.C • •

..-J~<D0%

~~il ~

010

2"-~ 00 ~

•I-

~C. ~

U ~C0%:

:i¢ .0 •-

t It)III: .....~ ~c

(J0-..r-

.~z ~ rt)L\J- c:: t!-oc""')a::

(!)

4~ --N

~- t tLt-L

TABLE V

Distribution of Amino Acid Residues at Three Points

of Heterogeneity in L. glauca Ferredoxin

119

T - 3 T - 7Tree Residue 6 Residue 12 Residue 96

Leucine Valine Proline Alanine Glycine Alanine

1 0.45 0.40 0.69 0.44 0.62 0.38

2 0.57 0.53 0.41 0.51 0.66 0.37

3 0.50 0.53 0.50 0.53 0.69 0.34

4 0.52 0.39 0.43 0.36 0.62 0.45

5 0.47 0.44 0.50 0.37 0.67 0.42

6 0.60 0.50 0.47 0.53 0.68 0.36

7 0.50 0.39 0.38 0.56 0.63 0.45

8 0.45 0.38 0.40 0.38 0.6-3 0.34

9 0.26 0.57 0.31 0.60 0.66 0.37

10 0.52 0.49 0.51 0.46 0.64 0.35

Mean 0.48 0.46 0.46 0.47 0.65

Standarddeviation 0.03 0.02 0.03 0.03 0.01 0.01

120

IV. DISCUSSION AND CONCLUSIONS

A. Characteristics of the Amino Acid Sequence of L. glauca

Ferredoxin:

In this study, L. glauca ferredoxin was found to

contain 96 amino acids in a single chain. Methionine

and asparagine are absent from this protein, and

tryptophan and histidine are present only in the amount

of one residue each. L. glauca ferredoxin contains two

residues of arginine, in contrast to other plant and-"...,

algal ferredoxins in which only one residue of this amino

acid has been found. The number of cysteine residues is

of special interest because cysteine has been implicated

in the chelate structure through which the two atoms of

iron are bound to the plant ferredoxin molecule (5).

L. glauca was found to contain five residues of cysteine,

as do spinach (13) and alfalfa (35) ferredoxins.

Heterogeneity was found to exist at four positions in the

~. glauca ferredoxin sequence, at residues 6, 12, 33, and

96. The finding of both glycine and alanine in position:

96, the carboxyl terminal position, permitted interpreta­

tion of the data previously obtained in the carboxypeptidase

A studies on the carboxymethy1ated protein. In each of

three of the four positions of heterogeneity, two very

similar amino acid residues are present. These include

position 6 (leucine and valine), position 33 (aspartic

acid and glutamic acid), and position 96 (glycine and

121

alanine). However, in position 12, alanine is present

in one form of the protein, whereas proline, an amino

acid which can have a drastic effect on the secondary

structure of protein molecules (64), is present in the

other form.

B. Comparison of Plant and Algal Ferredoxinsfrom Four Species:

In Figure 25, the amino acid sequence of L. glauca

ferredoxin is given in full, with the residues which

differ in alfalfa, spinach, and Scenedesmus ferredoxins

on the lines below the L. glauca sequence. The residues

are numbered according to the system used in the published

sequence of spinach ferredoxin (14). Thus, in ~. glauca

ferredoxin, which has one less residue at the amino

terminus when the homologous residues of the four

sequences are aligned, the amino terminal residue is

numbered "2". This difference at the amino terminal

accounts for the fact that only 96 amino acid residues

are found in L. glauca ferredoxin, whereas spinach and

alfalfa ferredoxins each have 97. Scenedesmus ferredoxin,

which also has only 96 residues, is lacking one residue

at the carboxyl terminal, as compared with the ferredoxins

of the higher plants~

Each of the four above-mentioned species has a

cysteine in each of positions 18, 39, 44, 47, and 77.

Scenedesmus ferredoxin has one additional cysteine

resi~ue, at position 85. It~has been found that all of

Figure 25. The amino acid sequences of ferredoxins from ~. glauca, alfalfa, spinach,

and Scenedesmus. The L. glauca ferredoxin sequence is given in fUll, with the additional

residues present at points of sequence heterogeneity shown on the line above. Residues

in alfalfa, spinach, and Scenedesmus ferredoxins which differ from the corresponding

residues in L. glauca are given on the lines below the L. glauca ferredoxin sequence, as

indicated. A minor form of spinach ferredoxin, having lysine in position 31 and methionine

in position 33, is not shown.

~NN

Val 10 Ala 20La. -Ala-Phe-Lys-Val-Lys-Leu-Leu-Thr-Pro-Asp-Gly-Pro-Lys-Glu-Phe-Glu-Cys-Pro-Asp-Alf. ~Ser Tyr Leu Val Glu Thr GinSp. Ala Tyr Thr Leu Val Thr Asn Val GinSc. AlaThr Tyr Thr Leu Lys Ser Asp Gin ThrIle

21 30 Asp 40L.g. Asp-Val- Tyr-lle-Leu-Asp-GIn-Ala-Glu~Glu-Leu-Gly-Ile-Glu-Leu-Pro-1Yr- Ser-Cys-Arg-

. Alt. His Glu ValSp. Ala Glu AspSc. Thr Ala Ala Leu Asp

41 50 60L.g. Aia-Gly-Ser-Cys-Ser-Ser-Cys-Aia-Gly-Lys-Leu-Val-Glu-Gly-Asp-Leu-Asp-Gln-Ser-Asp-Alt. Vol Ala Ala Glu ValSp. Lys Thr SIr AspSc. Ala Val Glu Ala Thr Val

61 70 80L.g. Gln-Ser-Phe-Leu-Asp-Asp-Glu-Gln-I le-Glu-Glu -GIy-Trp-Val-Leu-Thr -Cys-Ala-Ala-Tyr-Alf. Gly Asp Val~ ~p A~Sc. Ser Met Asp Gly Phe VaI

81 90 AlaL.g. Pro-Arg-Ser-Asp-Val-Val-Ile-Glu-Thr-His-Lys-Glu-Glu-Glu-Leu-Thr-GlyAlt. Ala Lys Thr AlaSp. Val Thr AlaSc. Thr Cys Thr Ala Asp Phe -

124

the cysteine residues in spinach ferredoxin are titrable

by heavy metals (5) and Matsubara et ale (14) have

suggested a model in which the folding of the polypeptide

chain brings together the five cysteine residues, which

are proposed to be chelated to the iron atoms. The three

residues of cysteine at positions 39, 44, and 47 are

located in a portion of the molecule in which the sequence

varies very little in these four species. This section,

from residue 35 through residue 50, includes only one

site, residue 43, at which variation is observed.

The similarities in the amino acid sequences of

these ferredoxins are more striking than are the differences

among them. Fifty-eight per cent of the sequence is

identical in these four species. Residues which do not

vary are: 4, 5, 9, 10, 12, 18-21, 23-26, 28-30, 32, 35-42,

44-50, 54, 58, 60, 62-66, 68,72, 74-77, 79, 80, 83, 84,

87, 89-93, and 95. If the ferredoxins of only the three

species of higher plants are compared the similarity is

even greater, with over 70% of the sequence invariant.

Many of the changes between species involve amino acid

residues which are structurally similar. Among these are

Tyr ---> Phe, Val ---> Leu, Ala ---> Gly, and Glu ---> Asp.

Such changes are considered conservative, since they would

not be expected to exert any significant effect on activity.

However, radical changes, such as those involving residues

which differ in charge and those involving proline

125

residues, occur also. It can be concluded that those

residues which vary with species are not essential to the

function of the protein. Thus, tryptophan and methionine,

and the histidine residue in position 27, which are found

in some but not all species, must not be directly involved

in catalysis or in the binding of iron. Many other

residues may also be considered non-essential on these

grounds. However, it may be noted that the requirements

for activity, even when not specifying that a certain

amino acid must occupy a given position in the molecule,

might nevertheless control the type (acidic, basic,

hydrophobic) of residue which may be present.

c. Acidic Residues in Plant and Algal Ferredoxins:

The amino acid sequences of the four plant and algal

ferredoxins shown in Figure 25 reveal an unusually high

content of acidic residues (aspartic and glutamic acids).

Table VI lists a number of proteins of various types and

the per cent of acidic residues in each, as established

in each case by amino acid sequence determinations. Most

of these proteins, including the bacterial ferredoxins,

have a content of acidic residues less than 15%. In

plant and algal ferredoxins these values range from 19%

in Scenedesmus to 24% in L. glauca. Of the proteins

listed, only rubredoxins, another class of non-heme iron

proteins, equal plant ferredoxins in their content of

acidic amino acids.

126TABLE VI

Acidic Residue Content of Proteins of Various Types

ProteinAcidic Total %Acidic Ref.Residues Residues Residues

Bacillus subtilisBPN' subtilisin 11 275 4 (65)

Bovine trypsinogen 14 229 6 (66)

Bovine ribonuclease 10 124 8 (67)

Rattlesnakecytochrome c 9 104 9 (68 )

Human hemoglobin 27 287 9 (69)

Human cytochrome c 11 104 11 (70)

E. coli tryptophan28 267 (71)synthetase Alpha 11

Clostridium butyricumferredoxin 7 55 13 (32)

Clostridium pasteurianumferredoxin 7 55 13 (31)

Human growth hormone 35 188 19 (72 )

Scenedesmus ferredoxin 18 96 19 (36)

Spinach ferredoxin 20 97 21 (14)

Alfalfa ferredoxin 21 97 22 (35)

....... Peptostreptococcuselsdenii rubredoxin 12 52 23 (73)

Leucaena glaucaferredoxin 23 96 24

Micrococcus aerogenesrubredoxin 13 53 25 (74 )

127

Of the 23 acidic residues in ~. glauca, 12 occur in

clusters in the sequence (residues 20-21, 29-30, 65-67,

70-71, and 92-94, as numbered in Figure 24). Such group­

ings of acidic residues have also been noted (14) in the

sequence of spinach ferredoxin and can be seen in the

alfalfa and Scenedesmus ferredoxin sequences as well. In

comparing the four sequences shown in Figure 25, only

four clusters, including nine acidic residues (residues

20, 21, 29, 30, 65, 66, 92, 93, and 94), are found to be

common to all four species. It is not known whether any

of these groupings are required for the function of the

protein. It may be noted that no clustering of acidic

residues is found in the amino acid sequences of the

clostridial ferredoxins, which are also capable of

catalyzing the photoreduction of NADP.

D. Genetic and Evolutionary Aspects of Variations in

Ferredoxin Sequences:

1. Plant and Algal Ferredoxins

The differences among the ferredoxins of the

four species shown in Figure 25 clearly indicate three

types of mutations which may be responsible: addi­

tion, deletion, and point mutations. The differences

in chain length, which involve the presence of one

less residue at the amino terminus of L. glauca

ferredoxin and at the carboxyl terminus of

Scenedesmus ferredoxin, are probably examples of

128

addition to or deletion from the corresponding genes

of one codon in each case. In the case of ~. glauca,

deletion rather than addition seems to be more

probable, since the other three species, including

the phylogenetically more primitive Scenedesmus, all

have an amino acid residue in this position. The

"missing residue" at the carboxyl terminal of

Scenedesmus, however, cannot be assumed to imply that

addition of a codon has taken place in the evolution

of the higher species, since a deletion in the

ferredoxin gene of Scenedesmus or of one of its

evolutionary precursors after divergence from the

precursor of the higher plants would be an equally

probable explanation. It may also be noted that it

is possible in the case of terminal residues that a

point mutation in the corresponding codon could

result in loss of one residue without affecting the

rest of the gene, thus giving the appearance of a

deletion, if the point mutation resulted in a chain

termination codon. However, in the two cases under

consideration, the amino acid present in the three

other species is alanine, which would require a

minimum of two point mutations within the codon for

such a change to occur. The codons considered are

those listed by Brimacombe et al. (75).

All of the variations within the interiors of

129

these sequences appear to have been caused by point

mutations in the ferredoxin genes. No addition or

deletion of interior residues is evident in the four

sequences, and although variation is found to exist

in as many as five consecutive positions in the

sequences, the codons corresponding to the amino

acids involved are such as to provide no evidence for

the occurrence of a frame shift. The differences

among the ferredoxin sequences in all four species

could be accounted for by a total of 45 single and

14 double mutations, with transversions occurring at

more than twice the frequency of transitions.

A method for relating phylogenetic differences

to the mutation distances between species has been

described by Fitch and Margoliash (76), who define

mutation distance as the minimal number of nucleotides

to be altered in order for the gene for one protein

to code for the other. A low value for mutation

distance between homologous proteins from different

species indicates close phylogenetic relationship

between the species. The mutation distances between

the ferredoxins of the four species are given in

Table VII. Also included in the table are the values

of the minimum base difference per codon which is

equal to the mutation distance divided by the number

of codon pairs compared (77).

130TABLE VII

Mutation Distances and Values of Minimum

*Base Difference per Codon

Ferredoxin LeucaenaAlfalfa Spinach ScenedesmusSource glauca

Leucaena glauca 24 29 44

Alfalfa (0.25) 27 40

Spinach (0.30) (0.28) 39

Scenedesmus (0.46) (0.42) (0.41)

*Mutation distances are given in the upper right half of thetable. Values of minimum base difference per codon are inthe lower left half of the table and are enclosed inparentheses.

131

The amino acid sequence of L. glauca ferredoxin

differs in 18, 21, and 25 positions from those of

spinach, alfalfa, and Scenedesmus, respectively, and

the corresponding mutation distances are 29, 24, and

44. The mutation distances indicate that L. glauca

is closest to alfalfa in the evolutionary scheme.

This is qualitatively in agreement with phylogenetic

considerations, since L. glauca and alfalfa are both

of the family Legl~inosae of the order Rosales of the

group Calyciflorae (cup flowers) whereas spinach is

of the group Thalamiflorae (receptacle flowers) and

Scenedesmus is a green alga (Chlorophyta). However,

from the closeness of the phylogenetic relationship

between alfalfa and L. glauca, it is surprising that

the calculated mutation distance is not much smaller.

Assuming that Scenedesmus is more primitive than

the higher plants, it can be seen that spinach and

alfalfa ferredoxins, with mutation distances from

Scenedesmus ferredoxin of 39 and 40, respectively,

have evolved equally since their divergence from a

common evolutionary precursor. L. glauca ferredoxin

must have diverged from the alfalfa line of evolution

within a relatively short time after the divergence

of the spinach and alfalfa precursors. These

evolutionary relationships, which are based on the

assumption of uniform rates of change in all groups,

132

are shown in Figure 26, in the form of a phylogenetic

tree.

2. Heterogeneity in ~. glauca ferredoxin

Heterogeneity has been found in the amino acid

sequences of ferredoxins from spinach and alfalfa.

In each of these species one major form of the protein

was found to be present in far greater amount than the

minor components. In~. glauca ferredoxin, however,

each of the four positions of heterogeneity is

occupied by two amino acid residues in ratios ranging

from 1:1 to 2.5:1, as determined from peptide yields.

Thus it is apparent that the heterogeneity in L.

glauca ferredoxin is not due to recent mutations,

since in that case the mutant forms would be present

in much lower amount than the original forms.

The tryptic peptide including residues 6 and 12

occurred in only two of the four possible forms, one

of which contained leucine and proline in positions

6 and 12, respectively, whereas the other contained

valine and alanine. Therefore the heterogeneity in

these two positions must be due to the presence of

different genes rather than to ambiguity in transla­

tion of the genetic code, as may be the case in

positions 33 and 96. Since the four points of

heterogeneity thus represent a maximum of three

variables, there may be as many as eight forms of the

133

Figure 26. Phylogenetic tree of the ferredoxins. The

evolutionary relationships indicated by the mutation distances

among amino acid sequences of the ferredoxins are shown diagram­

matically. All of these ferredoxins have regions of homology

among their primary structures, indicating possible common

evolutionary origin. The ferredoxins of the anaerobic

nonphotosynthetic bacteria have the most primitive structure,

having evolved little since their apparent origin by the

doubling of a precursor gene. The ferredoxin of Chromatium is

quite similar in structur~ and is therefore placed on the same

branch of the tree. The ferredoxins of the higher plants and

that of Scenedesmus have evolved much further. The branching

off of Scenedesmus ferredoxin and the inter-relationships among

the ferredoxins of the higher plants as shown here are based on

the degree of homology existing among the amino acid sequences

and are discussed in the text.

LEUCAENAGLAUCA ALFALFA SPINACH

ANAEROBICNONPHOTOSYNTHETIC

BACTERIA

135

protein in this species.

The population distribution of the different

amino acids at three points of heterogeneity was

studied in order to obtain more information as to

the genetic causes for the presence of multiple forms

of ferredoxin in ~. glauca. If the heterogeneity at

positions 6 and 12 were due to differing allelic

nuclear genes of equal frequencies, the distribution

of the two forms of ferredoxin among the L. glauca

population would be expected to follow a 1:2:1 ratio,

as shown in Figure 27. Thus 50% of the trees would

be heterozygous and contain both forms of the protein

(with respect to residues 6 and 12). However, of the

other 50% of the trees, half would be expected to be

homozygous for each form of ferredoxin. Thus, in

selecting trees at ~a~dom from the L. g!~ popula­

tion, the probability that any given tree would be

homozygous and contain only one of the two forms of

the protein is 50%, or 1/2.

On the other hand, if the differing genes were

non-allelic, as would be the case if duplication of

the ferredoxin gene had· occurred in the evolution of

this species, each individual tree would be expected

to contain both forms of the protein, probably (but

not necessarily) in approximately equal amounts.

Such a pattern of distribution of the different forms

136

Figure 27. Inheritance patterns of allelic and non­

allelic genes. The distribution among progeny of allelic

genes of equal frequencies, in this case genes for two

forms of ferredoxin as shown, according to classical

genetic principles, would follow a 1:2:1 pattern, with

equal numbers of homozygotes and heterozygotes resulting.

The distribution of non-allelic genes, however, would yield

only progeny which are homozygous for both forms of the

protein, as shown at the right.

•

..C\JX--t---...-....

(.)-.-.JW-.J-.J«

(.)

:Jw.-.J.-.J«I

zoz

• •• •

I · I

I II II II II £

« 0~ a::L~__-;;:;-I « c..

It] I I-"iii""""rtJ- : 1~ ::::>~ ~I II II III II

LLt.

138

of ferredoxin might also be observed if the ferredoxin

genes were located in the chloroplast DNA rather than

in the nuclear chromosomes, since an individual plant,

while inheriting a total of only one diploid set of

nuclear genes, may receive numerous proplastids which

develop into chloroplasts.

The present studies show that, of the ten trees

examined, all contained both forms of ferredoxin with

respect to residues 6 and 12. The probability that

allelic nuclear genes are responsible is thus (1/2)10,

or less than 0.001. However, it is not possible to

distinguish between the other two possibilities, non­

allelic nuclear genes and allelic genes of the

chloroplasts, from the data obtained in these studies.

The fact that the ratio of glycine to alanine in

T-7 is 1.7:1, which is quite different from the 1:1

ratio found to exist between the residues present in

positions 6 and 12, indicates that the genetic causes

of the heterogeneity in this protein are more complex

than the presence of two non-allelic genes, which

would yield the same ratio between amino acids at all

points of heterogeneity. The fact that both glycine

and alanine are found at position 96 in the ferredoxin

of each of the ten trees studied shows that allelic

genes cannot be responsible for this heterogeneity.

If multiple non-allelic genes are responsible, there

139

must be more than two such genes to account for the

differences in ratios of amino acids present at the

different points of heterogeneity. Three non-allelic

genes could not account for these ratios unless they

had different activities. Four non-allelic genes of

the same activity could at best give a 3:1 ratio of

glycine to alanine, in contrast to the 1.7:1 ratio

actually found. If the ferredoxin genes were contained

in the chloroplast DNA rather than in the nucleus of

the cell it would be possible for three o~ more differ­

ing ferredoxin genes, present in different amounts in

the cell, to account for the observed ratios. However,

it seems probable that more variation in ratios of

amino acids at each point of heterogeneity would be

found in different trees if this were the case.

Another, more probable explanation for the

different ratios of residues present at the points of

heterogeneity would be ambiguity in translation of the

terminal codon. Ambiguous translation of a special

codon has been suspected to occur in the genes for the

c(-chain of horse hemoglobin (78), with either lysine

or glutamine being inserted in position 60. Another

example of possible ambiguity in coding, involving

serine and threonine, has been found recently in

studies on mouse hemoglobin (79). The possibility

that the finding of more glycine than alanine in

140

position 96 might be an artifact of the methods

used for tryptic hydrolysis or peptide separation is

counteracted by the results of hydrolysis of the

carboxymethylated protein by carboxypeptidase, which

indicate the presence of more glycine than alanine at

the carboxyl terminus. The most probable explanations

for heterogeneity in L. glauca ferredoxin therefore

appear to be:

1. Either two non-allelic nuclear genes or

differing chloroplast genes, in the case of

residues 6 and 12.

2. Ambiguity of translation at position 96.

The cause of heterogeneity at position 33 has not been

investigated.

The amino acid differences and types of mutations

involved in the observed heterogeneity are summarized,

in Table VIII. In each case a single point mutation

could be responsible. All appear to be transversions,

and all involve the guanine-cytosine base pair, with

the possible exception of the aspartic acid-glutamic

acid change in which the nucleotides involved can not

be ascertained.

E. Techniques in Ferredoxin Purification:

The procedure used for purification of ferredoxin

from ~. glauca leaves included three modifications which

may be of use in the isolation of this protein from other

141

TABLE VIII

Types of Mutations in L. glauca Ferredoxin Heterogeneity

ResidueNumber Amino Acid Changes Codon Changes

RequiredMutation

Type

6 Leu <-----> Val UuA or CUX <--> GUX TransversionG

12 Pro <-----> Ala CCX <---------> GCX Transversion

33 Asp <-----> G1u GAU <---------> GAA TransversionC G

96 G1y <-----> Ala GGX <---------> GCX Transversion

142

species as well. The first of these was the addition of

acetone directly to the unfiltered homogenate. This was

necessary in the case of L. glauca due to the fact that

the leaf homogenate, in contrast to that from some other

species, was very viscous and could not have been

filtered without the prior addition of a large amount of

buffer, thus greatly increasing the volume to be processed

in sUbsequent steps. The second modification, the use of

a bed of Solka-floc for filtration of the acetone-treated

homogenate, had the advantage of being much faster than

centrifugation, and also succeeded in removing all

particulate homogenate material from the extract in one

step, along with the material precipitated by 35% acetone.

The third modification was introduced at a later stage

after the ferredoxin had been treated with 0.5 gm per ml

ammonium sulfate to precipitate much of the remaining

contaminating protein material. The ferredoxin, in

solution with this high concentration of ammonium SUlfate,

was ~pplied directly and without prior dilution to a

column of DEAE-cellulose, which was then washed with 0.1 M

Tris buffer of pH 7.3 to which had been added 0.5 gm

amm0nium sulfate per mI. The ferredoxin was absorbed as

a dark red band at the top of the column, while a large

amount of yellow material passed through. This property

of fe~redoxin, that of being strongly bound by DEAE­

cellulose in very high salt concentration is highly

unusual and therefore results in a very large amount of

purification with little or no loss of ferredoxin. The

reason for this phenomenon is not known, but logically

would appear to involve a decreased solubility of the

protein in the ammonium sulfate solution in the presence

of DEAE-cellulose, rather than simple ionic interaction

between the ferredoxin and the DEAE-cellulose, such as

occurs in solutions of low salt concentration.

F. Specificities of Proteolytic Enzymes Employed in these

Studies:

Some observations may be made regarding the

specificities of the proteolytic enzymes used in the

sequence studies. TPCK-trypsin hydrolyzed both of the

Arg-X bonds but only four of the five Lys-X bonds in the

carboxymethylated protein. The unhydrolyzed bond was

Lys-Glu, in positions 90-91. However, another Lys-Glu

bond, in position 13-14, was hydrolyzed in the tryptic

digest. Thus it may be assumed that the presence of an

acidic residue following lysine is not in itself

inhibitory to the splitting of the bond, but that other

factors, involving additional portions of the sequence,

are involved. In the chymotryptic digest, one Lys-X

bond was partially cleaved, where X is either leucine or

valine, in position 5-6, even though the enzyme had been

pretreated with TLCK to inhibit tryptic activity. Four

of the nine Leu-X bonds were not cleaved; the four X

144

groups were three residues of aspartic acid and one of

proline. The Phe-Leu bond (positions 62-63) was not

hydrolyzed by the TLCK-chymotrypsin. In studies on

spinach ferredoxin (14) this bond was found to be

hydrolyzed to only a minor extent in the chymotryptic

digest. One unusual cleavage by the TLCK-chymotrypsin

was the bond between two serine residues, at positions

44 and 45. Apparently the action of the enzyme was

responsible, since this bond is stable in tryptic

peptide T-5.

In the hydrolysis of peptides T-7a and C-IO by

B. subtilis neutral protease, two of the bonds that were

cleaved, GIn-Leu (residues 67-68), and Glu-Leu (residues

93-94), were consistent with the known specificity of

the enzyme for linkages in which a hydrophobic residue

contributes the amino group (80, 81, 82). However, two

other bonds, Gln-Ser (residues 60-61) and Glu-Thr (residues

87-88) were also hydrolyzed. Thermolysin, which has a

specificity (83) like that of the B. subtilis neutral

protease, hydrolyzed only those bonds in peptide C-IO

which were in keeping with its known specificity, Asp-Leu

(residues 54-55), Ser-Phe (residues 61-62), Phe-Leu

(residues 62-63), and GIn-lIe (residues 67-68).

G. Determination of the Complete Structure of L. glauca

Ferredoxin:

The present studies have elucidated only the primary

145

structure of this protein. The secondary, tertiary, and

chelate structures remain to be determined. Possibly the

most definitive method of determining the complete

structure of L. glauca ferredoxin will be by X-ray

diffraction studies of the crystalline protein. To this

end samples of the protein have been sent to Dr. Lyle

Jensen at the University of Washington and, hopefully,

the determination of the complete structure will be

possible.

146

v. SUMMARY

Ferredoxin, a non-heme iron protein involved in

photosynthesis and nitrogen fixation in plants, was isolated

from the leaves of Leucaena glauca, a species of small

leguminous tree. The protein was carboxymethylated and

hydrolyzed by trypsin and chymotrypsin to yield two sets of

peptides which were then isolated by ion exchange and

partition chromatography. Chemical and enzymatic methods

were used to determine the amino acid sequences of the

peptides, from which the sequence of the protein was

reconstructed.

L. glauca ferredoxin was found to contain 96 amino acid

residues in a single chain. Comparison of its sequence with

those of spinach (14), alfalfa (35), and Scenedesmus (36)

ferredo~ins showed that 58% of the residues are identical in

these four species, including five cysteine residues which

are believed to be involved in the binding of the two iron

atoms to the polypeptide chain. The types of mutations which

appear to be responsible for the differences among the

ferredoxins of these four species are addition, deletion, and

point mutations. The mutation distances among these

ferredoxins indicate that L. glauca is more closely related

in evolution to alfalfa than to the other species, and that

it has evolved further from Scenedesmus than have spinach and

alfalfa.

Heterogeneity was found in the amino acid sequence of

147

~. glauca ferredoxin. A study of the distribution of the

observed heterogeneity within the ~. glauca population showed

that the genes responsible for the different forms of the

protein are not allelic nuclear genes, but may be either

chloroplast genes or non-allelic genes of the nucleus.

148VI. BIBLIOGRAPHY

1. Lovenberg, W., Buchanan, B. B., and Rabinowitz, J. C.,

J. BioI. Chern., 238, 3899 (1963).

2. Losada, M., Whatley, F. R., and Arnon, D. I., Nature,

,190, 606 (1961).

3. Arnon, D. I., Science, 149, 1460 (1965).

4. San Pietro, A., and Lang, H. M., Science, 124, 118 (1956).

5. Keresztes-Nagy, S., and Margoliash, E., J. BioI. Chern.,

241, 5955 (1966).

6. Davenport, H. E., Hill, R., and Whatley, F. R., Proc.

Roy. Soc., ~., 139, 346 (1952).

7. Arnon, D. I., Whatley, F. R., and Allen, M. B., Nature,

180, 182 (1957).

8. Davenport, H. E., Biochern. ~., 77, 471 (1960).

9. Davenport, H. E., and Hill, R., Biochern. J., ~, 493 (1960).

10. Mortenson, L. E., Valentine, R. C., and Carnahan, J. E.,

Biochern. Biophys. Res. Commun., 1, 448 (1962).

11. ~agawa, K., and Arnon, D. I., Nature, 195, 537 (1962).

12. Matsubara, H., Jukes, T. H., and Cantor, C. R., Brookhaven

Symposia in Biology, Structure, Function, and Evolution

in Proteins, No. 21 (1968), in press.

13. Matsubara, H., Sasaki, R. M., and Chain, R. K., J. BioI.

Chern., 243, 1725 (1968).

14. Matsubara, H., and Sasaki, R. M., J. BioI. Chern., 243,

1732 (1968).

15. Matsubara, H., J. BioI. Chern., 243, 370 (1968).

149

16. Sasaki, R. M., and Matsubara, H., Biochern. Biophys. Res.

Commun., 28, 467 (1967).

17. Whatley, F. R., Tagawa, K., and Arnon, D. I., Proc. Nat1.

Acad. Sci. (U. S.), 7, 266 (1963).---- -- -18. Losada, M., Paneque, A., Ramirez, J. M., and del Campo,

F. F., in A. San Pietro (Editor), Non-heme iron proteins:

role in energy conversion, Antioch Press, Yellow Springs,

Ohio, 1965, p. 211.

19. Evans, M. C. W., Buchanan, B. B., and Arnon, D. I., Proc.

Nat1. Acad. Sci. (U. S.), 53, 1420 (1965).

20. Korkes, S., J. Bio1. Chern., 216, 737 (1955).

21. Valentine, R. C., Mortenson, L. E., Mower, H. F.,

Jackson, R. L., and Wolfe, R. S., J. Bio1. Chern., 238,

856 (1963).

22. Raeburn, S., and Rabinowitz, J. C., in A. San Pietro

(Editor), Non-heme iron proteins: role in energy

conversion, Antioch Press, Yellow Springs, Ohio, 1965,

p. 189.

23. Stern, J. R., in A. San Pietro (Editor), Non-heme iron

proteins: role in energy conversion, Antioch Press,

Yellow Springs, Ohio, 1965, p. 199.

24. Bachofen, R., Buchanan, B. B., and Arnon, D. I., Proc.

Nat1. Acad. Sci. (U. ~.), 51, 690 (1964).

25. Andrews, I. G., and Morris, J. G., Biochem. Biophys.

Acta, 97, 176 (1965).

- .....

150

26. Akagi, J. 00., Biochem. Biophys.oRes. COrnniun., 21,72

(1965 ).

27. Valentine, R. C., Jackson, R. L., and Wolfe, R. S.,

Biochem. Biophys. Res. Commun., 7, 453 (1962).

28. Valentine, R. C., and Wolfe, R. S., J. Bacterio1., 85,

1114 (1963).

29. Bradshaw, W. H., and Reeder, D. J., Bacterio1. Proc.,

110 (1964).

30. Brill, W. J., and Wolfe, R. S., Fed. Proceed., 24, 233

(1965).

31. Tanaka, M. T., Nakashima, T., Benson, A., Mower, H. F.,

and Yasunobu, K. T., Biochem. Biophys. Res. Commun. ~ 16,

422 (1964).

32. Benson, A.M., Mower, H. F., and Yasunobu, K. T., Proc.

Nat1. Acad. Sci., (U. S.), 55, 1532 (1966).

33. Tsunoda, J., Whiteley, H. R., and Yasunobu, K. T., J.

BioI. Chern., in press.

34. Eck, R. v., and Dayhoff, M. 0., Science, 152, 363 (1966).

35. Keresztes-Nagy, S., Perini, F., and Margo1iash, E., to

be pUblished, (1968).

36. Sugeno, K., and MatSUbara, H., Biochem. Biophys. Res.

Commun., 32, 951 (1968).

37. McConn, J. D., Tsuru, D., and Yasunobu, K. T., J. BioI.

Chern., 239, 3706 (1964).

38. Mares-Guia, 00., and Shaw, E., Fed. Proc., 22, 528 (1963).

151

39. Wang, S. S., and Carpenter, F. H., J. Bio1. Chern., 240,

1619 (1965).

40. San Pietro, A., and Lang, H. M., J. Bio1. Chern., 231,

211 (1958).

41. Losada, M., and Arnon, D. I., in Modern rnethods~f plant

analysis, Vol. VII, H. F. Linskens, B. D. Sanwa11, and

M. V. Tracey, (Editors), Springen, Vienna, 1964, p. 594.

42. Bendall, D. S., Gregory, R. P. F., and Hill, R., Biochern.

~., 88, 29P (1963).

43. Crestfie1d, A. M., Moore, S., and Stein, W. H., J. Bio1.

Chern., 238, 622 (1963).

44. Schroeder, W. A., Jones, R. T., Cormick, J., and McCalla,

K., Anal. Chem., 34, 1570 (1962).

45. Moore, S., and Stein, W. H., J. Bio1. Chern., f76, 367

(1948).

46. Katz, A. M., Dreyer, W. J., and Anfinsen, C. B., J.

Bio1. Chern., 234, 2897 (1959).

47. Harris, J. I., and Hindley, J., J. Mol. Bio1., 13, 894

(1965).

48. Smith, I., Nature, 171, 43 (1953).

49. Sanger, F., and Thompson, E. O. P., Biochern. Biophys.

Acta, 71, 468 (1963).

50. Spackman, D. H., Stein, W. H., and Moore, S., Anal. Chern.,

30, 1190 (1958).

51 •. Opienska-B1auth, J., Charezinski, M., and Berbec, H.,

Anal. Biochem.,~, 69 (1963).

152

52. Edman, P., and Sjoquist, J., Acta Chern. Scand., 10,

1507 (1957).

53. Konigsberg, W., and Hill, R. J., J. BioI. Chern., 237,

2547 (1962).

54. Konigsberg, W., in C. H. W. Hirs (Editor), Methods in

enzymology, Vol. XI, Academic Press, New York, 1967,

p. 464.

55. Randerath, K., Thin Layer Chromatography, Academic

Press, New York, 1966, p. 122.

56. Fraenkel-Conrat, H., Harris, J. I., and Levy, A. L.,

Methods Biochem. Anal., £, 391 (1955).

57. Sanger, F., Biochem. J., 39, 507 (1945).

58. Fraenkel-Conrat, H., Harris, J. I., and Levy, A. L.,

Methods Biochem. Anal., £, 388 (1955).

59. Randerath, K., Thin Layer Chromatography, Academic Press,

New York, 1966, p. 116.

60. Gray, W. R., in C. H. W. Hirs (Editor), Methods in enzy-

mology, Vol. XI, Academic Press, New York, 1967, p. 139.

61. Deyl, Z., and Rosmus, J., J. Chromatog., 20, 514 (1965).

62. Bradbury, J. H., Biochem. ~., 68, 475 (1958).

63. Fraenkel-Conrat, H., Harris, J. I., and Levy, A. L.,

Methods Biochem. Anal., £, 397 (1955).

64. Szent-Gyorgyi, A. G., and Cohen, C., Science, 126, 697

(1957).

65. Smith, E. L., Markland, F. S., Kasper, C. B., Delange,

R. J., Landon, M., and Evans, W. H., J. BioI. Chern., 241,

153

5974 (1966).

66. Walsh, K., and Neurath, H., Proc. Natl. Acad. Sci. U. ~.,

52, 884 (1964).

69. Braunitzer, G., Gehring-Muller, R., Hilschmann, N.,

Hilse, K., Hobom, G., Rudloff, V., and Wittrnann-Liebold,

B., ~. Physiol. Chernie, 325, 283 (1961).

70. Matsubara, H., and Smith, E. L., J. BioI. Chern., 237,

3575 (1962).

71. Yanofsky, C., Drapeau, G. R., Guest, J. R., and Carlton,

B. C., Proc. Natl. Acad. Sci. U. S., 57, 296 (1967).

72. Li, C. H., Liu, W-K., and Dixon, J. S., J. Am. Chern.

Soc., 88, 2050 (1966).

73. Bachrnayer, H., Yasunobu, K. T., Peel, J. L., and Mayhew,

S., J. BioI. Chern., 243, 1022 (1968).

74. Bachmayer, H., Yasunobu, K. T., and Whiteley, H. R.,

Biochern. Biophys. Res. Commun., 26, 435 (1967).

75. Brirnacornbe, R., Trupin, J., Nirenberg, M., Leder, P.,

Bernfield, M., and Jaouni, T., Proc. Natl. Acad. Sci.

U. ~., 54, 954 (1965).

76. Fitch, W. M., and Margoliash, E., Science, 155, 279

(1967).

77. Cantor, C. R., and Jukes, T. H., Proe. Natl. Acad. Sci.

154

Q. s., 56, 177 (1966).

78. Kilmartin, J. V., and Clegg, J. B., Nature, 213, 269

(1967).

79. Rifkin, D. B., Hirsh, D. I., Rifkin, M. R., and

Konigsberg, W. H., Cold Spring Harbor Symp. Quant.

BioI., 31, 715 (1966).

80. Feder, J., Biochemistry ~, 2088 (1967).

81. Morihara, K., Biochem. Biophys. Res. Commun., 26, 656

(1967).

82. Benson, A. M., and Yasunobu, K. T., Arch. Biochem.

Biophys., 126, 653 (1968).

83. Matsubara, H., and Sasaki, R. M., Fed. Proc., 26,723

(1967).