4v.

‘73'*’3;';
'-

‘
'3‘ - .
3 35.1!“ L-| }

§

(“(1
33’4“» 4:1,. :7"
. ~x :j'I
3 33- ‘
{9“ 1"“
,3.
I'if‘k‘

3'1 ‘1

a x

5:?

I' .‘ .

I
6‘. ‘.
.o 33313": 3;“, :5; I.
3, <; 3%3 ,3“ Mr
"3"? ‘. 3539:" N .
\d

,

‘1

.‘.(.

4‘. ‘
I. " 3 m“
‘3 5.4.,”

‘(“ "OW'V
q; -
«

.zz' a"
3‘
'15.).
. ’ éx‘. ~ A.
a-I'iilz;f;: , 'L l'. "'1“ J‘-
-’-- Wm i 3‘;

--f{‘“

'(z

5.3.1:.- :,

"3w .
,"' 2m:

'15:; .' 3".“ f'
_ , “~35" $211
-‘4::;1'¢ ‘-

g- , ,
._ , ‘ ,
arc-1","J «5....1..._.',, ._’_~ 1:: _J"“' "p;
, w.» .— ._ ' ~ M33"
‘3 ijmr “I. '15:?" .r
“:1;"‘z;:‘¢?:—nmw_u; .572 -..‘£':"} ‘ y“
“:4“ 1:40: o.
"._ a. V mJ' I»... 3.22::- ’ ”.J.‘
a“... ..
mix-tav-fiom

3"Amw-
m7?" ”“743

 

v u'mnmr
7‘...
:zgzgvat'rar‘? ' "E
H ' ‘uvvw ‘
3"“ ””"“’L mq... R” my... ....,,.
(4-..- M "" ‘ PL< m:
..

$3”. A’finbfim
'-

 

 

 

7) 0&4
LmRARY 736' '1
M'm'gm State llllll’lUlllHlli Illiillll M W WI lll

University 3 1293 005.74 9274

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

thesis entitled .
"A Study of the Optical Absorption Differences
Spectra Attributed to the 51—82 Transition

in the Oxygen Evolving Complex of Green Plants"

presented by

Dwight Lillie

has been accepted towards fulfillment
of the requirements for

Ph . D degree in Chemistry

 

 

Date/éajmg/ 3/???

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

.4.

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

   
 

F

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmatwe Action/Equal Opportunity Institution

A STUDY OF THE OPTICAL ABSORPTION DIFFERENCE SPECTRA
ATTRIBUTED TO THE SI --> 82 TRANSITION IN THE OXYGEN EVOLVING
COMPLEX OF GREEN PLANTS

By

Dwight Henry Lillie

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Chemistry

1 989

ABSTRACT

A STUDY OF THE OPTICAL ABSORPTION DIFFERENCE SPECTRA
ATTRIBUTED TO THE S --> S TRANSITION IN THE OXYGEN EVOLVING
CO PL OF GREEN PLANTS

By

Dwight Henry Lillie

The optical absorption dlfierence spectrum arising from the $1 -> 82
transition of the Oxygen Evolving Complex (CEO) in purified Photosystem ll
membranes from green plants has been studied under controlled conditions in
which the required anion cotactor, Cl’, has been deleted and reconstituted with a
series of monovalent anions, OH', Cl', Br', and N03'. The observed asymmetric
absorption difference band, does not undergo the substantial changes expected
if it arose from a charge transfer from the anion ligand to the Mn oi the CEO, and
it is therefore proposed that the absorption band arises from an oxygen ligand to
the Mn.

In Dedication to
Gladys Cross,

my grandmother, who passed away the 24th of January, 1988. Her life was a
testament of the love she had for her children and grandchildren. i will never.
forget the lessons she taught me, and in particular the value of hard work and the

joy there is to be found in pursuing one's ideals.

«d NJ Nu... .ux- am hi. hi. 0%.... up... haw.

ACKNOWLEDGEMENTS

There have been many people who have helped contribute to my
completion of this Ph.D. program, and I wish to thank them here.

First, and most of all, I wish to express my deepest thanks and love to
my wife, Carolynn. Her patience, understanding, and unselfish love allowed me
to accomplish this goal. Our children, Curtis, Erin, and Megan, helped me to
remember that there is indeed life after graduate school. i also want to thank my
parents, my brother Richard, and my sister Beth, for the encouragement and
support I received.

To my mentor and friend, Professor Jerry Babcock, who provided me
with an environment to discover a real joy for science and who was willing to
allow his students broad latitude in their research, i am indeed grateful.

I relied heavily upon the expertise of the staff in the Electronics, Machine,
and Glass Shops. In particular, Marty Rabb, Ron Haas, Scott, and Tom Clarke
provided far more assistance and encouragement than required. Their
dedication to teaching graduate students the fine points of electronics, "...here,
just hold on to this end for a moment while i bring up the power...," guarantees
quick studies.

The contributions of another, Dr. Tom Atkinson, who quickly won me to
the philosophy, ”if you have to do it more than once - i'd rather teach a computer
to do it..." cannot be understated. it was only been through his influence and
teaching that I was able to produce the software and hardware systems
described within as well as to appreciate the inherent advantages of taking
backups, '...well, this is trashed - you‘ve got a backup tape, don't you?”

Juan Lopez Garriaga, Chris Bender, Matt Espe, Harold Fonda, Bob
Kean, Bill Buttner, Tony Oertling, Pat Callahan, Margie, Carol, and other
members of the laboratory - thank you for the friendship, the willingness to pitch
in, the late night bull-sessions, and the beer. it was fun, wasn't it...wasn't it??

TABLE OF CONTENTS

List of Tables .......................................... viii
List of Figures .......................................... ix
List of Abbreviations ...................................... xi
CHAPTER 1 - INTRODUCTION ............................... 1
1.1 PHOTOSYNTHESIS: AN OVERVIEW ..................... 2
1.1.1 Basic Equation ................................ .2

1.1.2 Van Neils Hypothesis ............................ 2

1.1.3 Determination of the Substrate for Evolved Oxygen ......... 3

1.1.3.1 The Hill Reaction ........................... 3

1.1.3.2 Controversy Between 002 and H20 .............. 3

1.1.4 The Z-scheme ................................. 4

1.2 Current Model: Electron Transport ........................ 5
1.2.1 The Reaction Center ............................. 6

1.2.2 Reducing Side of Photosystem ll ..................... 8

1.2.2.1 Iron Mediation of Reducing Side Electron Transport ..... 9

1.2.3 The Secondary Donor, Z .......................... 9

1.2.3.1 S ctroscopic and Kinetic Characterization of Z ...... 10

1.2.3.2 E ectron Transfer Models for Z .................. 11

1.2.3.3 Problems with Chemical Identity of Z .............. 12

1.2.3.4 T rosine as Z ............................. 13

1.3 The Oxy‘gen volving Complex (OEC) ..................... 14
1.3.1 P enomological Data: The Kok Model ...... . ........... 14

1.3.2 Cofactor Requirements for Oxygen Evolution ............ 16

1.3.3 Spectroscopic Evidence of Manganese Involvement ....... 16

1.3.3.1 Multilme - EPR ............................ 16

1.3.3.2 X-Ray Studies ............................ 18

1.3.3.3 Optical Studies ............................ 19

1.4 Focus of This Thesis ................................ 19
CHAPTER 2 - SPECTROMETER DEVELOPMENT ................. 43
2.1 INT ODUCTION .................................. 43

2.2 EX ERIMENTAL REQUIREMENTS ...................... 44
2.2.1 Information to be Measured ........................ 44

2.2.2 Sample Attenuation of Exciting Pulse ................. 44

2.2.3 Signal Discrimination ............................ 45

2.2.4 Photochemistry Induced by the Probe Light ............. 46

2.2.5 Sample Dependencies on Flash Number ............... 46

2.2.6 Other Characteristics ............................ 47

2.8 IMPLEMENTATION ................................ 48
2.3.1 Monochromator and Probe Beam Source ............... 4 8

2.3.2 Sample Compartment ........................... 49

2.3.3 Excitation ................................... 50

2.3.4 Detector s .................................. 50

2.3.4.1 P Divider Chain ......................... 53

2.3.4.2 Simple Tests ............................. 55

2.3.5 Analog Electronics ............................. 55

V

2.3.5.1 Relating Measured Light Intensities to Absorption ..... 55

2.3.5.2 Detector Amplifier Circuitry .................... 57

2.3.5.3 Final Filtering ............................. 58

2.4 CONCLUSION .................................... 59
2.4.1 Susceptibility to mechanical vibrations ................. 6 0
2.4.2 Imprecise Optical Feedback Control of Xenon Arc Lamp ..... 60
2.4.3 Alternative A roaches .......................... 61
CHAPTER 3 - INSTRUME DEVELOPMENT - COMPUTER SUPPORT. .88
3.1 Introduction ...................................... 88
3.1.1 Overview .................................... 9 0
3.1.1.1 Timing Control for Events ..................... 90

3.1.1.2 Absorption Determination ..................... 91

3.1.2 Implementation Philosophy ........................ 91
3.2 HARDWARE ARCHITECTURE IMPLEMENTATION ........... 92
3.2.1 Timi S stem ................................ 93
3.2.2 Data oI ection ................................ 94
3.2.3 Transmission ................................. 94

3.3 SOFTWARE ARCHITECTURE IMPLEMENTATION ........... 95
3.3.1 Distributed Processing: An Introduction ................ 96
3.3.2 Distributed Processing: The Design & Implementation ...... 98
3.3.2.1 User Command System ...................... 98

3.3.2.2 Apparatus Control System .................... 99

3.3.2.3 Graphics Display System .................... 100

3.3.2.4 Operating System Functions .................. 101

3.3.2.5 Data Communication ....................... 102

3.4 SUMMARY ..................................... 103
3.4.1 Performance ................................ 103
3.4.2 Maintenance ................................ 104
3.4.3 Evaluation .................................. 104
CHAPTER 4 - THE PHOTOSYSTEM II ACCEPTOR SIDE ............ 120
4.1 INTRODUCTION ........................ ~ ......... 120
4.1.1 Separation of Ouinone Contributions ................. 121
4.1.2 Sim Iification of Reducing Side Chemistry ............. 122
4.1.2. Conventional Inhibition ...................... 122

4.1.2.2 Trypsin Modification of the Ob Binding Site ......... 124

4.1.2.3 Cholate Treatment ......................... 125

4.1.3 rosoopic Characterization of the Reducing Side ...... 125
4.1.3.1 Steady State Absorption S ctra ............... 126

4.1.3.2 Flash Induced Difference ra ............... 126

4.2 TERIALS AND METHODS ......................... 128
4.2.1 Biological Pre arations and Handling ................. 128
4.2.1.1 Chlorop ast preparation ...................... 128

4.2.1.2 Photosystem II membrane preparation ............ 129

4.2.1.3 Sample Preparation for Flash Absorption Spectroscopy .130

4.2.2 Instrumental Conditions ......................... 132
4.2.3 Data Handling ................................ 133
4.3 RESULTS ...................................... 134
4.3.1 Saturation of the Oa'lOa Transition .................. 134
4.3.2 The Oa'lOa Difference Spectrum ................... 135
4.3.3 Kinetics .................................... 137
4.3.4 Hydroxylamine ............................... 137

4.4 Discussion ...................................... 139
4.4.1 Reaction Center Heter eneity .................... 140
4.4.2 Alternative Secondary E ron Acceptors ............. 141

vi

CHAPTER 5 - OPTICAL STUDIES OF THE $1 —> 82 TRANSITION . . . . 168

5.1 INTRODUCTION ................................. 168

5.1.1 Polypeptide Depletion .......................... 169

' 5.1.1.1 The 23 kDa Polypeptide ........................ 169

5.1.1.2 The 17 kDa Polypeptide ........................ 170

5.1.1.3 The 33 kDa Polypeptide ........................ 171

5.1.2 Spectroscopic Probes of the Internal S-states .............. 171

5.1.2.1 EPR Signals Arising from the OXéan Evolving Complex. . 171

5.1.2.2 Optical Signals Arising from the xygen Evolv ng Complex .172

5.1.3 Goals of this Chapter ........................... 175

5.2 Materials and Methods .............................. 175

5.2.1 Biological Preparations and Handling ................. 175

5.2.1.1 Polypeptide Depletion ......................... 176

5.2.1.2 Chloride Depletion ........................... 176

5.2.1.3 Oxygen Evolution Assays ....................... 177

5.3 RESULTS ...................................... 178

5.3.1 Steady State 02 Evolution Measurements ............. 178
5.3.1.1 Anion Effects in Control and Chloride Depleted PS ll

Membranes .......................... 178

5.3.2 Absorpti‘gn Difference Spectra ..................... 180

5.3.2.1 S1 —> S Transition in Untreated Samples ...... 180

5.3.2.2 The S1 —> 82 Transition in NaCI Treated Samples. . . 180

5.4 Discussion ...................................... 182

5.4.1 Justification for Spectral Mani ulation(s) .............. 182

5.4.2 Role(s) Played by the Anion -Factors ............... 183

5.4.3 Identification of the First Flash Absorption Differences ..... 184

5.4.4 Conclusions ................................. 185

APPENDIX A - Flash Absorption Spectrometer Computer Interface ..... A213

A.1 Introduction .................................... A213

A.2 Overall descri ion ............................ A213

A.2.10pticalystem................... ........ A214

A.2.2 Electronics, analogue ....................... A217

A.2.3 Electronic, di ital .......................... A218

A.2.4 Electronics, iming ........................ A220

A.2.5 Computer, hardware ....................... A221

A.3 Miscellaneous Hardware ........................ A222

A.4 Computer software ............................ A222

A.4.1 Preliminary - some requirements before starting ..... A223

A.4.2 Starting ................................ A224

A.4.3 Execution .............................. A225

A.5 ERRORS AND SUGGESTED RECOVERY METHODS ....... A236

A.5.1 Failure to start UPLODR fully .................... A236

A.5.2 Interface at computer end not powered up ............ A237

A.6 DATA TRANSFORMATION ......................... A237

A.7 STRUCTURE OF THE SAVED DATA FILES .............. A239

A.8 GRAPHICS DEVICE 0 SELECTION .................... A241

vii

Table 1.1
Table 3.1

Table 3.1

Table 4.1

Table 5.1

Table 5.2

Table 5.3

LIST OF TABLES

Cofactors implicated in oxygen evolution

Resource comparison between AIM-65 and PDP-11
computer systems

Resource comparison between IBM-PC and PDP-11
computer systems .

Effect of NHZOH on the 320 nm Absorbance Change
Correlation between S1 -> 82 spectra taken in untreated
O]2] evolvin PS-ll membranes and in CI' depleted, X'
reconstitut PS-Il membranes

Effect of CaX _saits uponO evolution in untreated and in
CI' depleted, reconstitut PS-II membranes.

Relative absorption peak heights for CI’ depleted, X‘
reconstituted PS-Il membranes.

viii

Figure 1.1
Figure 1.2
Figure 1.3
Figure 1.4
Figure 1.5
Figure 1.6

Figure 1.7
Figure 1.8
Figure 2.1
Figure 2.2
Figure 2.3
Figure 2.4
Figure 2.5
Figure 2.6
Figure 2.7
Figure 2.8
Figure 2.9

Figure 2.10
Figure 2.11
Figure 2.12

Figure 3.1
Figure 3.2

Figure 3.3

LIST OF FIGURES
The Kok S-state Model.
Oxygen Yield as a Function of Flash Number
The Z-scheme
Membrane Topology Corresponding to the Z—scheme.
Crystal Structure of Bacterial Reaction Center

Structure and Redox Chemistry of Quinones In Photosystem
I .

Multiline Configurations

Cubane structure proposed for the OEC.

Block Diagram of Flash Absorption Spectrometer
Amplifier Circuitry for PMT.

Fourth order low-pass filter.

Roll-off characteristics of Active Low Pass Filters
Typical Dynode Chains for PMTs

Dynode Wiring Design.

Events during a Single Flash Sequence
Occurrence of Events During Multiple Sequences.
Removal of background level from signal.

PMT accuracy response from 250 nm to 350 nm.
PMT Linearity

Layout of Analog Signal Processing

Logical Schematic of Timing Logic

Communications Scheme between Computer and
Instrument

Algorithms for Setup of Timing Sequences.

ix

Figure 3.4

Figure 3.5
Figure 3.6
Figure 4.1

Figure 4.2
Figure 4.3
Figure 4.4
Figure 4.5
Figure 4.6
Figure 4.7
Figure 4.8
Figure 4.9

Figure 4.10

Figure 5.1
Figure 5.2

Figure 5.3
Figure 5.4

Figure 5.5
Figure 5.6
Figure 5.7

Figure 5.8
Figure 5.9

Figure 5.10

Algorithms for Initialization and Initiation of Timing
Sequences

Algorithms to Process Timer Completions (counter finished)
Logical Schematic of Software Control Systems

3 nchronous Relationship between Donor andReducing
des

Inhibitory Sites in the PSII Electron Transport Chain.
Saturation Profile of Oa'lOa versus excitation intensity
The Absorption Spectra of Oa’lOa.

Comparison of Two Oa'lQa Spectra

Exposure TIme Effects on the Oa'lOa change.

Effects of External Acceptors on d[Oa'/dt.

Absorption Spectrum of Tris-Washed PS—II membranes
C-550 Bandshift

Kinetic Data Processing

Effect of Replacement of CI' with NOa'.

S1 1Absotgp tion Difference Spectrum in 02 Evolving,
ninhibited ampies.

A Comparison Between the 82/81 and the Oa'lOa Spectra.

C' Depleted, No Addition (OH'), 1st Flash Difference
Spectrum.

CI' Depleted, Cl' Repleted, 1st Flash Difference Spectrum.
CI" Depleted, Br” Repleted, 1st Flash Difference Spectrum.

Cl" Depleted, N03' Repleted, 1st Flash Difference
Spectrum.

CI' Depleted, X' Repleted, fst Flash Difference Spectra.

Comparison Between Two First Flash Difference Absorption
Spectra.

Comparative Plots for the /S1 Difference Spectra Between
02 and X' substituted, X'- Cl', Br', N03", and OH’}.

ADC
AIM-65

ATP
bit
CPU

DCBO

DCMU
DEC
DMA
DPC
EGTA

EPFI
EXAFS

94.1

Hepes
HLL
IBM

ISFI

LHC

ABBREVIATIONS 8: SYNONYMS
Analogue to Digital Converter

Evaluation computer system using an 8-bit 6502
microprocessor.

Adenosine Triphosphate
Smallest piece of information in a number system of base 2
Central Processing Unit

Symbolic identig associated with a kinetic species
responsible for i nal II, slow (lis), and presumed to be of

. the same type as .

2,5 (or 2,6)-DichIoro-p«benzoquinone. Only 2,5-DCBO was
used in the work reported here.

3-(3,4 DichlorophenyI)-1,1 dimethylurea
Digital Equipment Corporation

Direct Memory Access
Diphenylcarbazide

EthyleneglycoI-bIs-(beta-aminoethyl ether)-N,N,N',N'-
tetraacetlc Acid

Electron Paramagnetic Resonance

Extended X-ray Absorption Fine Structure

:2 EPR si naI that is centefred atg - 4.1 and considered to
533;; Coengoenfrmers o the S ]2\] state of the Oxygen
N-2-Hydroxyiethyl Piperazine-N'-2-EthanosuIfonic Acid.
High Level Language(s)

International Business Machines

Interrupt Service Routine

Kilobyte, actually 211 or 1024 bytes.

Light Harvesting Complex

xi

If:
ME
MI

‘ U
51)

Mb
Mes
MIMD
MMU
MPU
NADPH
OEC
OS

PC

POP-1‘1

PMT
PPBQ
PSII
O-BUS
RC
FISX-1 1

FIT-1 1

"f

"3
"w

SIMD
task
TCP

Megabyte, actually 22°.
4-Morpholineethanesuifonic Acid

Multiple Instruction, Multiple Data

Memory Management Unit

(see CPU)

Nicotinamide Adenoine Dinucleotide Phosphate
Oxygen Evolving Complex

Operating System

Personal Computer, often used meaning IBM's interpretation
of personal computer.

Pro raCm Data Processor, 16-bit minicomputer manufactured
by E .

Photomultiplier Tube

phenyI-p-benzoquinone

Photosystem II

The backplane used in the PDP-11 computer systems.
Reaction Center

Real-time System Executive, a multiuser, multitasking

aerating system for the DEC PDP-11 systems supporting
MU.

Real Time operatin system, a single user operating system
for the DEC PDP-1 systems. There are two versions of the
system, RT-1ISJ, 'SJ' for single job, and RT-11FB, 'FB' for
foreground-background.

'Signal II, fast.‘ Fast EPR kinetic signal assigned to 2+,
detected in samples inhbited in the evolution of 02.

'Signal II, slow.‘ Slow EPR kinetic signal associated with 0"".

'Signal II, very fast.‘ Very fast EPR kinetic signal assigned to
2+, detected in samples which can evolve 02.

Single Instruction, Multiple Data
Term referring to an executable computer program.

Task Control Block

xii

ifs
VII

VIII. 5

Tris
VAX-1 1
VMS

Tris(hydroxymethyl)aminomethane
Virtual Machine Architecture, DEC's 32-bit minicomputer.

Virtual Memory System for the DEC VAX-11 computer
family.

Symbolic ident' which relates thekinetic species, D, with a
clearly identifie Tyrosine.

Symbolic identity which relates the kinetic species, 2, with an
impled chemical identity, Tyrosine.

Symbolic identity for a kinetic species implicated in electron
transfer between the Reaction Center and the Oxygen
Evolving Complex.

xiii

CHAPTER 1

1 .1 . INTRODUCTION

Photosynthetic organisms couple one electron photochemistry to two and
four electron transport, and generate intermediate species with large reductive
and oxidative potentials. The reductive equivalents are used initially for the
production of 'primary' biochemical energy in the form of ATP (Adenosine
Triphosphate) and NADPH (Nicotinamide Adenosine Dinucleotide Phosphate).
The oxidative equivalents are toxic waste products from a biochemical viewpoint,
and are duly disposed of via a water oxidation process. Water oxidation is a four
step process (see Reaction 1.9), and the biochemical catalysis site, otherwise
known as the Oxygen Evolving Complex (OEC), must be able to acquire and
stabilize four oxidative equivalents until dioxygen is stably formed and evolved.
Upon the evolution of oxygen, the storage site is reset for another four electron
process. The site implicated for catalysis is a tetranuclear cluster of manganese
atoms, the structure of which is not yet known. Several polypeptide and ionic
cofactors have been identified as being responsible for establishing and main-
taining the local environment(s) necessary for oxygen evolution. Largely thought
to be Intractable, the catalysis site has recently been proving to be amenable to
biochemical and biophysical techniques. This has been driven by the devel-
opment of biological preparations that have become increasingly resolved in

terms of their composition and purity.

12F

Adv ,Cu .,,U 9
~ ..,.. was .M . .u. a v
“n 5v. . . .2:

33":

~

3. . . -u m.“
, In I u . ..
flea aMm-Mn gum his &
T e! t lflv

at A.
Call.

2

1.2. PHOTOSYNTHESIS: AN OVERVIEW

1.2.1. Basic Eguation

Photosynthesis refers to the light to chemical energy conversion
processes found in plants (eukaryotic organisms) and also in photosynthetic
bacteria (prokaryotic organisms). The net sum of all relevant processes in higher

plants can be represented as:
GCOZ + 6H20 --> CGH‘I 206 + 602 (1 .1)

Early in this century photosynthesis was described in a similar fashion.
Centuries had been required to reach this point, and many more years remained
before the reactions underlying this equation could be properly ferreted out and
placed into context. Blackman (1905), with little in the way of experimental
evidence, first hypothesized that photosynthesis involved two setiof reactions.
One set consisted of a series of light dependent reactions that yielded photo-
chemical products for use by the second set. This second set of reactions
contained no direct dependence upon light other than the photochemical
products, and were thus referred to as the dark reactions. These suggestions
made by Blackman; however, remained largely speculation until Van Niel, in the
19208 and 1930s, began systematic investigations that firmly laid the foundations
for Equation 1.1 and the subsequent work of Hill that supported the idea of light
and dark dependent reactions.

1.2.2. Van Niel's Hypothesis

In 1941 Van Niel (1941) noted that, after comparing hydrogen donors for
bacteria and higher plants, molecular oxygen released from plants must arise

from water and not from carbon dioxide, as widely believed. In the case of

é
r\)
(A)

I“

- 4

'PA“:

3%}

bacteria, which can not oxidize water, succinate or hydrogen sulfide was used in
place of water as the reducing substrate. The underlying process in bacteria and

plants was presumed the same and could be generalized to:

HZA + C02 --—> (CHZO) + H2 + 2A (1.2)
1.2.3. Determination of the Substrate forérolved Oxygen

1.2.3.1 The Hill Reaction

Confirming the existence and independence of two separate sets of
reactions, Hill (1939) demonstrated oxygen evolution from 'osmotically shocked'
chloroplasts, subcellular components containing the photosynthetic machinery
isolated from leaves, given addition of an external oxidant. These chloroplasts
had lost the ability to assimilate carbon dioxide, and so the original purpose of
demonstrating complete photosynthesis in a subcellular component of plant
tissue was defeated; however, the retention of oxygen evolution proved more
significant. Not only did this establish the independence of two sets of reactions
responsible for oxygen evolution and carbon dioxide assimilation, but also it
strongly indicated that water, not carbon dioxide, was the substrate in the
production of oxygen. These experiments, demonstrating the production of
molecular oxygen, and the reduction of an exogenous reductant, have come to

be known as the 'Hili-Reaction.‘
1.2.3.2 Controversy Between 002 and HZO

The concept of water oxidation at this time was not novel. Ruben et al.
1941) illuminated Chlorella in oxygen labeled water and observed that the

content of the oxygen evolved was likewise labeled, and it was also shown that

the content of the labeled oxygen evolved was independent of the content of
oxygen labeled carbonate. This was objected to on the grounds of oxygen
exchange by Warburg (1964). He had found in 1960 that bicarbonate was
required for oxygen evolution (Warburg and Luttgens, 1960). This exchange is
apparently facilitated by low pH and carbonic anhydrase, conditions not well
controlled in the original experiments of Ruben et al. (1941).

It has only been recently that similar isotopic experiments (Stamler and
Radmer, 1975) using 18o-labeileci water In algae acceptably confirmed this
observation. . In these experiments the algae samples showed no carbonic anhy-
drase activity and the pH was regulated to minimize mixing. Ironically, later
experiments by Radmer (discussed later) regarding intermediates in the water

oxidation process have also fallen prey to the same type of objections.
1.2.4. The Z-scheme

It was not until the 1960s that this evidence could be interpreted properly,
which is not to say that the research carried on in the intervening years was
fruitless. Those intervening years laid the groundwork necessary for acceptance
of a new concept. Hill and Bendall (1960) proposed that the light dependent
reactions, those established as responsible for water oxidation, consisted of two
separate photochemical reaction centers operating in series driving
photosynthetic electron transport, and represented, in a fashion similar to Figure
1.3, the 'Z-scheme.‘

Duysens et al. (1961) further extended the distinguishability of the
reaction centers by using light of selective wavelengths to control the oxidation
and reduction of the cytochrome that is in the electron transport chain between
the two photosystems (see Figure 1.3). Long wavelength light selectively excited
Photosystem I and led to the oxidation of cytochrome. Short wavelength light

it‘s:

selectively excited Photosystem ll (PSII) and led to the reduction of the oxidized
cytochrome. Addition of inhibitors such as DCMU (3-(3,4 DichlorophenyI)-1,1
dimethylurea), blocked reduction by Photosystem II.

1.3. Current Model: Electron Transport

It has now been fairly well established that electron transport in higher
plants from H20 to NDAP‘i‘ involved three membrane protein complexes. These
complexes interact with each other via small carriers in a serial fashion.

Electrons are extracted from H20 by the Photosystem II complex and
transported to the b6-f complex by a plastoquinone carrier. Plastocyanin carries
electrons from the b6-f complex to the Photosystem I complex. The Photosystem
l complex in turn reduces NADP": The molecular components making up these
complexes, when they are known, plotted against their respective redox midpoint
potentials resemble a '2‘ as shown in Figure 1.3 and this diagram of the electron
transport chain has become known as the 'Z-scheme.‘ As the molecular
components transport electrons they alternately undergo reduction and oxidation.

Since the 1960s the Z-scheme has remained relatively unchanged
except for the addition of new components as they have been identified.
Association of sections of the electron transport chain with biochemical structures
has been made, and an orientation with respect to membrane topology has been
established. Figure 1.3 represents a considerable improvement in detail over the
original. Figure 1.4 displays the Z-scheme components in a model membrane
topology.

Chloroplasts possess a definite 'inside' and 'outside' orientation. This
orientation allows the free energy released during photosynthetic electron

transfer, in the form of a buildup of the free proton concentration, [H+], in the

Ime

ADI

3111 a

excl-c
'V

inner aqueous enclosure, to be partially conserved by the phosphorylation of
ADP to ATP, catalyzed by the ATP synthase.

To complete the picture we shall continue to flesh out the scheme, but
the chronological detailing will be dropped, and the emphasis will lie with
Photosystem II, the system coupled to the OEC. A more detailed historical
account can be found by Rabinowitch (1945) and Hill (1972).

L31 The Reaction Center

The reaction centers in Photosystems I and II are responsible for light
gathering and initial quasi-stable charge separation. Associated with each
reaction center is a core set of 50 chlorophylls, and somewhat more loosely is a
larger set of 200 to 250 chlorophylls in a protein pigment complex called the light
harvesting complex (LHC) or the antenna complex. These chlorophylls are
coupled electronically to each other and to the reaction center to form a potential
sink at the reaction center. Light absorption by the LHC results in a coherent
exciton packet that is funneled into the reaction center for photochemistry. A
fairly detailed physical description of such antenna complexes has been provided
by Knox (1977); also see Pearlstein (1982) and Parson and Ke (1982) for
somewhat more recent treatments.

The reaction center of higher plants is presumably a chlorophyll dimer
'special pair' that is electronically coupled to chemical groups directly, and
indirectly facilitates initial electron transfer from the excited state manifold(s) of
the special pair. In green plants and algae there are two such reaction centers
and are referred to as P680 and P700 respectively for Photosystems II and l.
The 'P' stands for pigment and the subscripted numbers reflect the absorption
maximum for the pigment (Clayton, 1980). A structural analogy between green
plants and bacteria may be drawn from crystallographic data from the bacterium

Rhodopseudomonas viridis (Deisenhofer, et al., 1984. See Figure 1.5). In the
bacterial case, the special pair, known as P370, consists of two
bacteriochlorophylls forming an electronic dimer. Nearby, on each side, forming
two branches, are two other bacteriochlorophylls. Continuing down the branches
are two more bacteriopheophytins, and terminating the branches are two
quinones. These branches are referred to as M and L branches of the reaction
center, and possess a general 02 symmetry. Despite the structural symmetry of
the two branches there is evidence that suggests the L branch is preferred over
the M branch for electron transfer by a factor greater than 5 (Breton et al., 1986).
Recent theoretical calculations have shown that the location of polar groups are
sufficient to enforce this observed unidirectionality of electron transfer (Michel-
Beyerle, 1 988).

While this structural analogy between higher plants and bacteria appears
to be generally valid, it is not without problems. The special pair in bacteria lies
perpendicular to the inside membrane surface. In green plants, however, there is
experimental support for the special pair to be positioned parallel to the
membrane surface (see Rutherford, 1985 and Breton and Vermeglio, 1982).

The initial charge separation results in the formation of an oxidized
pigment, P‘I'. Pheophytin, in the case of green plants, or bacteriopheophytin, in
the case of bacteria, is reduced in a time scale of 10'12s. In turn, it is oxidized
and the primary quinone acceptor is reduced in a time scale of 10’9s. P"‘ can
either be reduced by its normal physiological donor, 2, and the reduced quinone
acceptor will be oxidized or P"’ can undergo a back reaction with the reduced
quinone acceptor. The first scenario is preferred in untreated systems. Alteration
of one or more of the components in the electron transport chain can shift
towards a preference for the back reaction. As an example, deactivation of the

Oxygen Evolving Complex shown in Figure 1.3, results in an increased lifetime of

A
‘A’

I

m

b

Fe .
Iron I
iafisl
$5311
“wei'

96:“
HB‘

2*, the physiological donor to P680- If 2 is still oxidized when P680+ is next
oxidized then only the back reaction with the acceptor is reasonable. P680+ has
been difficult to study because of the multiplicity of kinetic phases that it exhibits.
In untreated systems it becomes reduced in tens of nanoseconds. In treated

systems it develops several micro- and milIi-second decay times.
1.3.2. Reducing Side of Photosystem II

The secondary electron acceptor of Photosystem II, the primary acceptor
being pheophytin (plants) or bacteriopheophytin (bacteria), is a plastoquinone,
referred to commonly as Oa. Figure 1.5 shows Its bacterial analogue as being
the terminal quinone of the L branch. Upon reduction, the charge separation
spans the width of the membrane, ~40A.

Oa is located in a covalently bonded site and in close proximity is
another quinone binding site, Ob. The Ob site, unlike Oa, is labile for quinones in
the plastoquinone (oxidized) and plastohydroquinone (reduced) forms, see
Figure 1.6. It is in tact lost during isolation as demonstrated in Figure 1.5. An
iron in the Fe(ll) state supports, and, as discussed below, mediates electron
transport between Oa and Ob by altering the equilibrium distribution.

The interaction between Oa and Ob serves as the interface between one
electron photochemistry and two electron transport chemistry. Two turnovers are
required as shown in Reactions 1.3 and 1.4 before the Ob site is recharged,
Reaction 1.5. Note that the second turnover is only possible when the reaction
center is 'open', Ob is reduced and Oa oxidized.

h v
OaOb --> Oa'Ob <--> OaOb' (1 .3)

hv
OaOb' --> Oa'Ob" ---> OaObz' (1.4)

2H+ PO

OaObz’ --> OaObH2 ---> OaOb + POH2 (1.5)

As will be discussed in more detail in Chapter 4, it is this binary cycle
coupled with the contribution of anionic radicals in the near UV and in the blue
absorption region that have led to disagreements over procedures that have
been used to account for and remove absorption difference spectral contributions
that arise solely from the reducing side. Anionic radicals have absorption
maxima in the near UV at 320 nm and in the blue at 440-450 nm (Bensasson and

Land,1973)
1.3.2.1 Iron Mediation of Reducing Side Electron Transport.

It was originally thought that iron removal blocked electron transport from

Oa to Ob, Parson (1978) and Blankenship and Parson (1978). It has been
demonstrated, however, that removal of the iron only appears to reduce the
measured efficiency of electron transfer between Oa and Ob by 37% (Debus, R.
J. et al., 1986). Reconstitution with divalent transition metals such as Mn2+,
002+, Ni2+, Cu2+, and Zn2+ as well as Fe2+ can restore rates to within 10% of
the control rates. An interesting aspect of the work of Debus et al. (1986) is the
stability of the Oa semiquinone with respect to the fully reduced form in the
presence and absence of the supporting divalent transition metals. The role
played by the metal appears to stabilize the charge separation indirectly by

stabilizing the semiquinone form.

1.3.3. The Secondary Donor, 2

Until recently the number of components directly involved in the electron
transfer pathway between the OEC and the oxidized reaction center has been

uncertain. Since the mid-70$ there has been reliable spectroscopic detection of

10

the redox reactions of the reaction center and its donor(s). The interpretation,
though, has been conflicting and has led to differing models. A fairly complete
summarization of the plausible models was done by Bouges-Bocquet (1980).
Experiments reported recently, time resolved EPR spectra (Hoganson
and Babcock, 1988a), in vivo deuteration (Barry and Babcock, 1987), and time
resolved optical UV spectra (Gerken et al., 1988), appear to confirm the presence
of only one intermediate involved in direct electron transport between the Oxygen

Evolving Complex and the reaction center.
1.3.3.1 Spectroscopic and Kinetic Characterization of Z.

Signal ll, an EPR signal with an isotropic g - 2.00, was observed as early
as 1956 by Commoner et al. (1956). Babcock and Sauer (1973) and Babcock et
al. (1976) associated this Signal II with the Z+ species, the donor to P680- Since
then it has been generally recognized that there are three forms of Signal II, and
two distinct molecular species giving rise to the Signal lIs. Signal IIS is a long-
Iived, slowly decaying signal, and is considered to arise from a molecular species
referred to as 0*.

Signal Ilf, ‘f‘ meaning 'fast', arising from 2*, is the direct donor to P680 in
preparations in which the capability to evolve oxygen has been inhibited. Signal
"f, first observed in thylakoids depleted of manganese (Babcock and Sauer,
1975), has been observed to closely parallel the reduction of “3680+ in the
similarly treated samples (Boska et al., 1983, and Weiss and Ranger, 1986).

Signal llvf, ‘vf' meaning 'very fast', arising from 2*, is the direct donor to
P680 in preparations retaining their oxygen evolving capacity. It was reported by
Blankenship et al. (1975) and Babcock et al. (1976). The original
characterization of "W contained a low resolution time resolved spectrum using

spinach thylakoids. In preparations that have not been inhibited in oxygen

(.

l“: the

Ias'lic

11

evolution the primary reductant phases of P680+ occur in the nanosecond time
regime (Mauzerall, 1972, and Van Best and Mathis, 1978), and they report that
the rise time of Signal Ilvf closely parallels the reduction time. Recently the time
resolved spectrum of NW has been reproduced by Hoganson and Babcock
(1988b), and their results confirm the similarity between II, and llvf.

1.3.3.2 Electron Transfer Models for Z

The most widely accepted model places only one component between

the OEC and the reaction center:

OEC ---> z ---> Peso (1.6)

Wit and co-workers (see Schlodder et al., 1984, and Brettel ef al., 1984),
have argued strongly, largely on the basis of the observed multiple kinetic phases
In their optical measurements, for multiple species arranged in serial or parallel

fashion:

OEC --> DI --> DZ ---> Z --> P680 (1.7)

D1
OEC <02> z --> P630 (1.3)

Where 'D1' and '02 are respectively for donor 1 and donor 2. Witt's group has
recently recanted these models (Gerken et al., 1988), however, in light of the
EPR evidence published by Hoganson and Babcock (1988a) and direct evidence
of the chemical identity of an intermediate (Barry and Babcock, 1987, and Debus
et al., 1988)

12

As of yet there is only spectroscopic evidence for the one donor species,
2, linking the OEC to the reaction center, whether it is the primary reductant to
P680 has yet to be shown. It has been recently reported, however, by Witt's
group (Gerken et al., 1987) that the primary donor to P680+ has been detected
optically in uninhibited oxygen evolving samples. The reported extinction
coefficients of 6000-8000 M'1cm'1 at 260 nm agree with those reported for
optically detected 2+ in inhibited samples by Dekker (1984a).

1.3.3.3 Problems with Chemical Identity of 2

Early extraction and reconstitution experiments of isotopically labelled
plastoquinone by Kohl and Wood (1969) were correlated with the subsequent
loss of Signal Ils, the EPR signal originating from Di, and led to the tentative
conclusion that D1' (and hence 2‘") was a plastoquinone cation. Other evidence
supported this as well. Wood and Bendall (1976) reported a high positive
midpoint potential for a OH2+IOH2 redox couple, O'Malley and Babcock (1984)
and O'Malley et al. (1984) analyzed experimental EPR Signal II data, EPR model
compound data, and performed computer simulations of the EPR spectra. This
all reasonably supported the plastoquinone hypothesis based on calculated ring
spin densities. Optical evidence from Diner and de Vitry (1984) and Dekker et al.
(1984a) all suggested that D and 2 was a quinone.

Unfortunately, several severe problems remained with this hypothesis
that could not easily be reconciled. (1) The experimental 9isov 2.0046, was
consistently too high for in vitro cation plastoquinones, the g-value ranges for
these are 2.0034 to 2.0038 (Sullivan and Bolton, 1968). (2) The conditions
required for model simulation of Signal II, 125 sulfuric acid, was considered by
many to be too severe for biological systems to provide. (3) The experiments of

Kohl and Wood (1969) could not be reproduced in any other laboratory, and (4)

quant
unaell
Cit/a:

was 01

(4.)
-‘

13

quantification experiments of plastoquinone levels by de Vitry et al. (1986), were
unable to find enough plastoquinone-9 in Photosystem ll particles isolated from
Chlamydomonas reinhardtii to account for D and 2 being plastoquinones. This

was confirmed by Takahashi and Katoh (1986).
1.3.3.4 Tyrosine as Z

Definitive identification of Signal IIS, 0"”, as a tyrosine on the D2
polypeptide has been reported by Barry and Babcock (1987). Two sets of in vivo
deuteration experiments were involved. First, the methyl group of quinones were
deuterated. Mass spectral data provided proof of the deuteration and the EPR
spectra showed no discernible change in D"'. The second set of experiments
deuterated tyrosine. Again, mass spectral data provided proof of deuteration,
and the EPR spectra showed a reduction in linewidth as would be expected It the
characteristic Signal ll spectra were Indeed arising from tyrosine.

Debus and Barry (Debus et al., 1988) were able to precisely locate the
position of D by reasoning from the facts that D"' is a tyrosine, spectral evidence
indicating 2+ is the same molecular species, and that 2 may be located on the
D1 polypeptide and D is located on D2 (iodination experiments by Takahashi et
al., 1986, and lkeuchl and lnoue, 1987). Trebst (1986) had published the amino
acid sequence and suggested folding patterns for the D1 and D2 polypeptides;
examination revealed two conserved tyrosines. Change of Tyr-160 on 02 to
phenylalanine by using site directed mutagenesis resulted in photosynthetic
growth but no characteristic Signal lls. Assuming the symmetry hypothesis holds
up, then a corresponding residue on D1, Tyr-161, is Z.

As a side note, this recent identification of D and (possibly) 2 as
tyrosines has prompted Hoganson and Babcock (1988a) to modify the

nomenclature conventions regarding Signal iI. The single letter abbreviation of

A
’-
8

id

In.

W

at
Vi

F. w
‘J

9

W35 '

0
l

iris

IESEE

14

tyrosine is 'Y,’ and is now used as the primary identifier. The subscripts 's', 'vf',
and 'f‘ are dropped. References to Z and 2" become Y2 and YZ+. References to
D and D+ become YD and YD+- The spectral references to "W and "f are now
folded into Y2, and depends upon the context of the discussion to determine

whether it is uninhibited or inhibited conditions.
1.4. The Oxygen Evolving Complex (OEC)

As early as 1969 phenomological evidence had been gathered by Joliot
(1968) and Kok et al. (1970) that gave rise to an elegant and abstract model of
02 evolution. This model is represented in Figure 1.1, and an example of
supporting experimental evidence is given Figure 1.2. Until 1981 the situation
was frustrating because the S-state model proposed by Kok suggests the
existence of discrete, intermediate states in the process of oxidizing water, and
. researchers were faced with problem of the apparent inability of the OEC to
withstand direct attempts to modify its behavour without complete dissolution.

At present, however, biochemical techniques have been made available
that not only provide more highly refined samples, but also allow selective
manipulation of the environment surrounding the OEC in terms of polypeptide
composition and ionic cofactors. This refinement in biochemical techniques have
led to better interpretation of information returned by the physical probes, such as
EPR, XAFS, and optical spectroscopy, in terms of the states of the OEC rather

than merely whether the system was evolving oxygen.

1.4.1. Phenomological Data: The Kok Model.

As stated before the OEC is the site for four electron chemistry with the
end result being the oxidation of two water molecules. The major function of this

complex is the removal of waste oxidative equivalents. The early experiments of

15

Joliot (1968) and Joliot et al. (1969), examined 02 yield as a function of short
saturating flashes of light with a modulated 02 electrode, and provided the first
quantitative linkage between discrete photoacts and 02 evolution. This allows
the half reaction to be written as a function of photons absorbed:

hv
2HzO --> 02 + 4H+ + 46' (1.9)

Subsequent work by Kok et al. (1970) led to the S-state model (Figure
1.3) and provided a basis for explaining the periodicity and damping observed by
both Joliot and Kok as depicted in Figure 1.4.

Two observations were key: 1) In dark adapted samples, only three
flashes of light were required for maximal 02 yield, and the next maxima
occurred four flashes later; 2) Model simulations (Forbush et al., 1971) required
inclusion of parameters allowing for double turnovers and for misses; i.e, one
flash of light could result in no S-state advance, a single S-state advance, or a
double S-state advance. Including only one or the other resulted in not only
damping but a phase shifting of the 02 yield that was not experimentally
observed. The first observation led to the idea of an equilibrium distribution
between S1 and So with the 80:81 ratio being 1:3. This has since been modified
after studies on long dark adapted material (hours) revealed that the ratio was
closer to 0:1 (Vermas et al., 1984). This dark auto-oxidation has been correlated
with the decrease in YD (Signal lls), and it has been suggested that D+ is
reduced by So in the dark over a long period of time (Zimmermann and

Rutherford, 1985).

16

1.4.2. Cofactor Reflements for Oxygen Evolution.

It is difficult to discuss the cofactors that seem to be required on an
individual basis because of the interdependency. Table 1.1 lists the known
factors that have been established as being required for 02 evolution. Those
directly involved are the extrinsic polypeptides and the ionic species. Of these,
manganese is the only one for which there is still an absolute requirement; i.e.,
manganese can neither be removed or replaced without inhibition. The accepted
stoichiometry of tour manganese per reaction center and as the site for catalysis
has been established by the work of Cheniae and Martin (1970), Ghanotakis and
Yocum (1985 and 1986), and lkeuchl et al. (1985) in thylakoids, Photosystem II
membranes, and in Photosystem lI core preparations.

The extrinsic polypeptides, 17, 23, and 33 kDa, are included only
because they are endogenous cofactors. Salt washing with high, non-
physiological concentrations of NaCI removes the 17 and 23 and inhibits 02
evolution, but higher than physiological concentrations of Ca2+ and CI' largely
removes the inhibition (see Ghanotakis and Yocum, 1985). Likewise, salt
washing with CaCIz removes all three polypeptides, and requires a six-told
increase in the Cl' concentration to sustain 02 evolution (Kuwabara et al., 1985).

The Mn cannot as yet be released without irreversible inhibition.

1.4.3. Smwscogic Evidence of Manganese Involvement,

1.4.3.1 Multiline - EPR

Perhaps the best understood spectroscopic characterization of the
manganese ensemble are the EPR detectable signals arising from the 32 state.
First observed In 1980 (Dismukes and Siderer, 1980), it was shown by Dismukes
and Siderer (1981) that a multiline EPR signal, consisting of at least 16 lines and

17

centered at g - 2.0, could be generated in oxygen competent thylakoids by
flashing a dark adapted sample, thereby generating the 82 state, and then
freezing quickly to 77 K. This suggested that this multiline signal arose from the
$2 configuration of the OEC. This observation was supported by others
(Hansson and Andreasson, 1982, and Brudvig etal., 1983a,b).

Spectral simulations of the multiline by Dismukes et al. (1982) suggested
either two dimers, Mn2(lll,lV) or a tetramer, Mn4(lll)3,IV). Hansson and co-
workers (Hansson and Andreasson, 1982, and Andreasson et al., 1983) found a
better simulation using a binuclear Mn2(ll,lll) model.

Confusing the issue has been the observation of the so-called g - 4.1
signal (referred to hereafter as the 94.1). Dark adaptation followed by treatment
with (DCMU), an inhibitor at the Ob site, and illumination between 130 - 160K
should populate the 82 state of the OEC (de Paula et al., 1985). Instead of the
expected multiline an EPR signal at g - 4.1 was found. Warming the sample to
200K resulted in the disappearance of the 94.1 and the appearance of the
multiline. Several have suggested this provided evidence for another
intermediate between the OEC and 2 (Casey and Sauer, 1984 and Zimmennann
and Rutherford, 1984).

Today, it is generally accepted that the EPR spectroscopic properties of
manganese in the 82 state are dependent upon temperature. Studies by Brudvig
and co-workers (de Paula and Brudvig, 1985, de Paula et al., 1986) and
Rutherford and co-workers (Zimmermann and Rutherford, 1986) have
demonstrated that the 94.1 and the multiline signals arose from different
configurations of the 82 state. These configurations can be expressed in an
equilibrium as depicted in Figure 1.7. This variability can be accounted for in

temperature dependent structural alterations that alter exchange couplings.

18

In terms of understanding the OEC, characterization of the 82 state
through the EPR multiline and 94.1 signals has been the singularly most
important event in the 1980s. The reader is referred to reviews by Babcock

(1987) and Brudvig (1986), and also to the Ph.D. theses of de Paula (1986) and
Beck (1 988).

1.4.3.2 X-Ra y Studies

EXAFS (Extended X-ray Absorption Fine Structure) has been used to
determine the ligand environment around the manganese in samples poised in
the S1, 82, and 53 states. In the resting, S1, state, a Mn-Mn distance of 2.7A
has been determined (Yachandra et al., 1986). A second Mn-Mn distance of
3.3A has also been reported (Guiles et al., 1987). In addition, a u-oxo bridge at
1.8A and a terminal oxygen or nitrogen ligation shell at 2.0 - 2.1A has been
identified as well; however, CI', as a ligand, has not been found.

Lack of CI' as a ligand is significant. As mentioned in the earlier section
discussing required cofactors for steady state oxygen evolution requires a
negatively charged ion such as Cl', Br', or NO3', CI’ being endogenous.
Sandusky and Yocum (1983), through competitive inhibition studies, have
suggested that Cl- serves as a bridging ligand between two manganese centers.

From the EPR studies, the model best supporting the available data is a
mixed valence tetramer (de Paula, 1986). The EXAFS data indicate the
presence of a u-oxo bridged manganese dimer. Thus, it has been proposed that
the manganese complex is in the form of a 'cubane'-like structure, Mn4O4, and

undergoes transitions as in Figure 1.8. (Brudvig and Crabtree, 1986).

19

1.4.3.3 Optical Studies

UV absorption spectroscopy has been used to detect absorption
changes associated with S-state advancement by several groups (Velthuys,
1980, Dekker et al., 1984b, Saygin and Witt, 1985 and 1987, Lavergne, 1986 and
1987, and Ranger and Weiss, 1983 and 1986). Considerable disagreement
exists regarding interpretation. Dekker et al. (1984b), after complex data
deconvolution that involves the removal of contributions from Oa, Ob (or other
acceptor quinones), and 2, reported the changes for the 30 --> S1, S1 -> 82,
and $2 -> S3 to be identical and probably arising from successive Mn(lll) ->
Mn(lV) transitions.

This finding contradicts that of Velthuys, and has since been challenged
by Lavergne (Lavergne, 1986 and 1987) and Witt's group (Saygin and Witt, 1985)
as well. These three groups claim there are no detected absorption changes on
the SD --> S1 and $2 --> 83. It appears the source of disagreements arise
because of the difficulty in deconvolving unwanted contributions from the
experimental data. Both Lavergne and Witt's group point out that deconvolving
the S-state transitions, normally dependent upon the relative S1/SO distributions
and the hit and miss parameters, does not yield unique solutions.

Dekker and Witt (Kretschmann et al., 1987) have since issued a paper
agreeing that the So ->S1 transition is a factor of three less than originally
reported by Dekker et al. (1984b) and probably corresponds to a Mn(ll) -->
Mn(lll) change. They also agree that the S1 --> 82, and $2 --> 83 changes are
identical and probably correspond to Mn(lll) -> Mn(lV) changes.

1 .5. Focus of This Thesis

It Is apparent that Cl' plays a significant role in the evolution of Oz. It is

also apparent that the existing data concerning this role are inconclusive. As a

20

result we decided to study the optical transitions in the near UV associated with
the S-states of the Oxygen Evolving Complex in green plants using Photosystem
II preparations. It was anticipated that the time resolved optical spectrometer
designed for study of the electrochromic shift and proton release patterns in
chloroplasts (Lillie, 1984) would prove to be adequate after having its wavelength
range extended into the UV and modification of the data acquisition and
computer interfaces.

Chapter Two discusses the improvements made to the spectrometer.
These changes resulted in an extension of the operation wavelength range to the
UV and near UV. Both the bandwidth and the minimum detected absorption
change ranges were extended as well.

Chapter Three discusses the improvements of the support and
functionality that were provided by the interfaced computer system. These
changes extended the type of experiments possible from repetitive single
channel collection to include repetitive multiple channel collection and conditional
event occurrences dependent upon shot and channel numbers.

Chapter Four discusses experiments performed on the reducing side of
Photosystem II. Only one other group has reported in as great of detail the
absorption changes undergone by the reducing side.

Chapter Five discusses experiments on the S-state transitions. These
experiments were conducted upon untreated and treated Photosystem ll
systems. The treated systems were those in which the Oxygen Evolving
Complex was reversibly inhibited.

Appendix A contains the user documentation for the Flash Absorption
Spectrometer and its Computer Interface (FASCI). This was a separately

prepared document for users and is included here for completeness.

21

TABLE 1.1

Cofactors lmplicated in Oxygen Evolution

Cofactors implicated in oxygen evolution. Proposed functions are
derived from Babcock (1987) and Murata and Miyao (1987), and are the ”well-
accepted” views. Alternative viewpoints exist for assigning polypeptide functions

for D1, D2, Ob, and Mn; for a review see Babcock (1987).

 

 

Cofactor Function
PSII Core Complex
2; kDa Chi-binding
34 Mn-binding, D2 (reaction
center)

33 Ob, D1 (reaction center)
9 b-559
4 Ii

Light-Harvesting Polypeptides

22-28 kDa Chi-a and b binding
Extrinsic Polypeptides

33 kDa Mnéstabilization.

23 Ca 1‘, CI'

17 Ca2+, Cl‘
Ionic Species

M +n, (n . 2,3, or 4) H20 site

Ca +

Cl'

22

FIGURE 1.1
The Kok S-state Mod_eL

The Kok model is abstract and postulates that there are five oxidative
storage states, but is unable to make any statement concerning how the
oxidative equivalents are actually stored or even what comprises the storage
center. Here, Sn, n - {0,1 ,...,4}, represents the five oxidative states of the Oxygen
Evolving Complex. '2 is the electron carrier between the OEC and the reaction

center, P680 as discussed in the text.

23

680

P
>
.
.
_
Z
>
_

V_x

\ . /
\ . /
\ _ /

\ . /
\> _ >/
\ . _ _ /
\ . _ . /

\ _ _ . /
\ . . . /
> . . . >
. . _ . .

v v v v
h h h h
0 1 2 3 4
SIIIVSIIIVSIIIVSIIIVS
> _
. .
. .
_ .
_ .
. .
. .
. .
_ .
+ IIIIIIIIIIIIIIIIIII +
\ /

\ /

. _

_ .

<

+ O

H 2

4 H

2
+

24

FIGURE 1.2
Oxygen Yiefl as a anction of Flash Number

The oxygen flash yield in isolated chloroplasts after a dark adaptation as
a function of flash number (Buttner, 1984). Kok etal. (1970) were able to fit the
damped, period four oscillations to a model by assuming a fraction of centers did
not undergo a transition and that another fraction of centers underwent two
transitions per flash. Computer modeling predicts the dark adapted distribution
of 80:81 to be 1:3.

25

cm

<1-

.anaz smo_a
_ o. m

b P b h
4 4 ‘ 4

P IF P IP
J I4 I‘I ‘ R

0.

 

 

 

 

 

 

 

LooEaz zmo_u mamav> u.om> No o>_.o.om

 

SSA/DA

26

HGURE13
The Z-scheme

The Z-scheme is a schematic representation of linear photosynthetic
electron transport in plants. The individual components are plotted along the
ordinate according to their known or estimated midpoint potentials (adapted from

Buttner, 1984).

27

03a
Ema... N\

.31.»? RD i\2mv\

 

c:
a

 

 

as}

 

ma

Q;

We

Do:

QT

WT

(A) 03 ’imuaiOd xopaa

28

FIGURE 1.4

Membrane Topology Corresmnding to‘the Z-scheme

The components involved in the Z-scheme are depicted here in their
approximate locations relative to the membrane's orientation. Unlike other
topological schemes, sizes and contours of the protein complexes have not been
articulated. Instead, an attempt has only been made to show relationships and
approximate electron transfer component locations. These components as well

as the protein complexes are discussed in the text.

Ti
Ti

91

 

29

 

 

 

{1'

er...

7.-...- -

 

b559

 

IIIIII iii A

 

 

 

I

A
\—

Oi

 

 

 

D

OJ
DC

 

 

 

 

M n Mfr

Mil Mil

 

I] 0

 

 

 

 

 

)

l

,..i
| Li

I

C

dotmm

X I

D

Water

30

FIGURE 1.5
Crystal Structure of Bacterial Reaction Center

Overview of the structure of Rhodopseudomonas viridis (reproduced
from Michel-Beyerle et al., 1987). Numbers in figure represent center to center
distances in 0.1 nm. 'P' is the reaction center formed from two
Bacteriochlorophyllb s with an interplanar distance of 0.3 nm. EM and 'BL' are
bacteriochlorophyll-b's in the M and L subunits. 'BPM' and 'H' are
bacteriopheophytin-D's in the M and L subunits. 'MO' is the acceptor, a
metaquinone in bacterial reaction centers. Each branch is terminated by a

quinone; however, the one in the M branch is lost during isolation.

31

 

32

FIGURE 1.6
Structure and Redox Chemistry of gilinones in Photosystem II.

It is important to understand this chemistry and in particular the
equilibrium states. Absorption bands in the UV and near UV of the quinone and

semiquinone disappear in the doubly reduced states (Chapter 4)

33

 

asses case
- as s 8

 

Incas

O

 

it

\/

. Io 3
+Im ———-———_—_————_
a _o_

mlgc \ OE

 

 

K\
ocmmcoguxm 33.; I 06 /\
3:309 I oo
_o_ 0.08 _o_
I I _ _
rm >£
u _.._ _.._

 

34

FIGURE 1.7
mitiline Configgrations

Proposed model for multiline and g . 4.1 signals. The proposal is that
the ligand substitution reactions influence the equilibrium between the two
configurations which give rise to the g - 4.1 and the multiline signals. (adapted

from Beck, 1988b).

35

K
31(01 ) + L <===é===> 81(L) + c1

I I
I901 I
v _ K2 v _
31(C1 ) + L <=======> 51(L) + C1

I l
IX 1 K'
I I eq

36

FIGURE 1.8
Cubane strjucture mosed for the O§_C_.

Brudvig and Crabtree (1986) have proposed this model in light of the
evidence from the EPR studies on the 82 state, EXAFS data, and optical
evidence. This particular model has been of particular interest because the
physical evidence backing it is more definitive than for other models that have
been suggested (see for example Kambara and Govindjee, 1985, or Spencer et

al., 1986).

37

/”s"\
DH .0. DH 8
+ M AI. 0
DUB/\an/ H++ 9'
/Mn\ \l Mn
' :1” H TH UH ,...fi~.\UH
S4/Mn\ (6‘ \Mn Mn:MIn//\/Iin S1

i“’/""‘i ..
[1H /l
4 Mn “RE-TI)" UIH 0 DH 8
U \( Mn\ IIJ /
e' A
+ -
0H7M"‘0H 2H3IJ M" H + e

38

REFERENCES
Babcock, G. T. (1987). ”New Com rehensive Biochemistry: Photosynthesis,”
125-158, (Amesz, J., ed.), lsevier, Amsterdam.

Babcock, G. T., Blankenship, R. E., and Sauer, K. (1976). FEBS Letters, 61,
286-289.

BabcockézG. T. and Sauer, K. (1975). Biochimica et Bigghysica Acta, 376, 315-
8.
Babcock, G. T. and Sauer, K. (1973). Biochimica et Bigghysica Acta, 325, 483.

Barry, B. A. and Babcock, G. T. (1987). Proceedings of the National Academy
of Sciences USA, 84, 7099-7103.

Beck, W. F. (1988 . "Ligand-Exchan e and Oxido-Reduction Reactions of Water
Analogs w h the Manganese etramer Complex of Photosystem II," 264,
Ph.D. Thesis, Yale University.

 

Bensass108n1, R. and Land, E. J. (1973). Biochimica et Biophysica Acta, 325, 175-

Blackman, F. F., (1905). Anng. Rev. e_o_t,, 19, 291-295.

Blankenshig, R. E. and Parson, W. W. (1978). Annual Reviews of Biochemistg,
54 , 635-653.

Blankenship, R. E., Babcock, G. T., and Sauer, K. (1975). Biochmica et
Biophysica Acta, 387, 165.

Boska, M., Sauer, K., Butter, W., and Babcock, G. T., (1983). Biochimica et
Bigghysica Acta, 722, 327-330.

Bouges-Bocquet, B., (1980). Biochimica et Biophysica Acta, 594, 85-103.

Breton, J., Martin, J.-L., Petrich, J., Migus, A., and Antonetti, 26. (1986). FEBS
Letters, 209, 37-43.

Breton, J. and Verm Iio, A. (1982). ”Energy Conversion in Plants and Bacteria",
Volume 1, 15 -194, (Govindjee ed.), Academic Press, New York.

Brettel, K., Schlodder, E., and Witt, H. T. (1984). Biochimica et Biophysica Acta,
766, 403-415.

Brudvig, G. .W., Case , J. L., and Sauer, K. (1983a). Biochimica et Biophysica
Acta, 723, 36 -371.

39

Brudvig, G. W., Casey, J. L., and Sauer, K. (1983b). "The Oxygen Evolving
System of Photosynthesis”, 159-164, (lnoue, Y., Crofts, A. R.,
Govind ee, Murata, N., Renger, G., and Satoh, K., eds), Academic
Press, okyo.

Brudvig, G. W. (1986). Journal of Bioenergetics and Biomembrane , 19, 91-103.

Brudvig, G. W. and Crabtree, R. H. (1986). Proceedin s of the National
Academy of Sciences, USA, 8_3_1, 4586-4588.

Buttner, W. J. (1984). Ph.D. Thesis, Michigan State University.

Casey, J. L. and Sauer, K. (1984). Biochimica et Biophysica Acta, 767, 21-28.

Cheniae, G. M., and Martin, I. F. (1970). Biochimica et Bigphysica Acta, 197,
219-239.

Clayton, R. K. (1980). "Photosynthesis: Physical Mechanisms and Chemical
Patterns," 88-110, Cambridge University Press, Cambridge.

Commoner, V., Heise, J. J., and Townsend, J. (1956). Proceedin of the
National Academy of Sciences, USA, 42, 710-718.

Debus, R. J., Barry, B. A., Babcock, G. T., and McIntosh, L. (1988). Proceedings
of the National Acagemy of Sciences, USA, 85, 427-430.

Debus, 2.2.81.7, Feher, G., and Okamura, M. Y. (1986). Biochemistry, 25, 2276-

Deisenhoter, J., E , 0., Miki, K., Huber, R. and Michel, H. (1984). Journal of
Molecular iolggy, 180, 385-398. .

Dekker, J. P., Van Gorkom, H. .J., Brok, M., Ouwehand, L., (1984a). Biochimica
et Biophysica Acta, 764, 301 -309.

Dekker, J. P., Van Gorkom, H. .J., Wensink, J., and Ouwehand, L., (1984b).
Biochimica et Biophysica Acta, 767, 1-9. '

de Paula, J. C. and Brudvig, G. W. (1985) Journal of the American Chemical
Sociey, 107, 2643-2648.

de Paula, J. 0., Beck, W. F., and Brudvig, G. W. (1986) Journal of the American
Chemical Soci_e_ty, 108, 4002-4009.

de Paula, J. C. (1986). Ph.D. Thesis, Yale University.
de Vitry, C., Caries, C., and Diner, B. A. (1986). FEBS Letters, 196, 203-206.
Diner, B. A. and de Vitry, C. (1984). ”Advances In Photosynthesis Research,"

Volume 1, (Junk, W., ed.) Martinus Nijhoff Publishers, Dordrecht, The
Netherlands.

Dismukes, G. C., Ferris, K., and Watnick, P. (1982). Photobiochemistgy and
Photobigphysics, 3, 243-256.

 

Jc
Jc

4O

Dismukes, G. C. and Siderer, Y. (1980). FEBS Letters, 121, 78-80.

Dismukes, G. C. and Siderer, Y. (1981). Proceedings of the National Academy
of the Sciences USA, 78, 274-278.

Duysens, L. N. M., Amesz, J., and Kamp, B. M., (1961). Nature, 190, 510-511.

 

Forbush, B., Kok, B., McGIoin, M. P. (1971). Photochemistgy and Photobiolggy,
14, 307-321.

Gerken, 6.°.1.,7B”rettel, K., Schlodder, E., and Witt, H. T., (1988). FEBS Letters, 223,
9- .

Gerken, S., Brettel, K., Schlodder, E., and Witt, H. T., (1987). FEBS Letters, 223,
376-380.

Ghanotakis, D. F. and Yocum, C. F. (1986). FEBS Letters, 1 97, 244-248

Ghanotakis, D. F. and Yocum, C. F. (1985). Photosynthesis ReseaLch, 7, 97-114.

Guiles, R. D., Yachandra, V. K., McDermott, A. E, Britt, R. l, Dexheimer, S. L.,
Sauer, K., and Klein, M. P. (1987). "P ress in Photosynthesis
Research," Volume 1, (Big ins, J., ed.), 61-564, Martinus Nijhofl Pub-
lishers, Dordrecht, The Net rlands.

Hansson, O. and Andreasson, L.-E., (1982). Biochmica et Biophysica Acta, 679,

261 -268.
AndreassonhLaE" Hansson, O., and Vanngard, T. (1983). Chemica Scrima, 21,

Hill, R. (1939). Proceedings of the Royal Socie_ty, London, 127, 192-210.

Hill, R. and Bendall, F., (1960). Nature, 186, 136-137.

Hill, R. (1972). "Proceedings of the 2nd International Congress of
Photos nthesis Research,” Vol 1, (Forti, G., Avron, M., Melandri, B.A.,
eds.), unk, The Hague.

Hoganson, C. W. and Babcock, G. T. (1988a). Biochemistgy, 27, 5845-55.

Hoganson, C. W. and Babcock, G. T. (1988b). Biochmica et Bigghysica Acta, in
press.

lkeuchl, M. and lnoue, Y. (1987). FEBS Letters, 210, 71-76.
lkeuchl, M., Yuasa, M., and lnoue, Y. (1985). FEBS Letters, 185, 316-322.
Joliot, P. (1968). Photochemistgy and Photobiology, 8, 451-463.

Joliot, P., Barbieri, G., and Chabaud, R. (1969). Photochemistgy and
Photobiolggy, 10, 309-329.

41

Knox, R. S. 1977). ”Primary Processes of Photosynthesis", Volume 2, 55-97,
Ba r, J., ed.), EIsevier/North-Holland Biomedical Press, Amsterdam,
he Netherlands.

Kohl, D. H. and Wood, P. M., (1969). Plant Physiolpgy, 44, 1439-1445.

Kok, B., Forbush, B., and McGIoin, M. (1970). Photochemistgy and Photobiolpgy,
1 1 , 457-475. ~

Kretschmann, H., Dekker, J. P., Saygin, O., and Witt, H. T. (1987). Biochimica et
Biophysica Acta, 932, 358-3 1.

Kuwabara, T., Miyao, M., Murata, T., and Murata, N. (1985). Biochimica et
Biophysica Acta, 806, 283-289.

Lavergne, J. (1986). Photochemistg and Photopiology, 43, 311-317.
Lavergne, J. (1987). Biochimica et Biophysica Acta, 894, 91-107.
Lillie, D. (1984) MS. Thesis, Michigan State University.

Mauzerall, D. (1972) Proceedin s of the National Aca_demy of the Sciences
USA, 69, 1358-1 362.

Michel-Beyerle, M. E., Plato, M., Deisenhofer, J., Michel, H., Bixon, M. and
Jortner, J. (1988). Biochimica et Biophysica Acta, 932, 52-70.

 

 

Murata, N. and Miyao, M. (1987). "Progress in Photosynthesis Research,"
Volume 1, (Biggins, J., ed. ,4 3-461, Martinus Nijhoff Publishers,
Dordrecht, The Netherlan s.

O'Malley, P. J. and Babcock, G. T. (1984). Biochimica et Biophysica Act , 765,
370-379.

O'Malley, P. J., Babcock, G. T., and Prince, R. C. (1984). Biochimica et
Biophysica Acta, 766, 283-288.

Parson, W. W. (1978). "The Photosynthetic Bacteria", 455-469, (Clayton, R. K.
and Sistrom, W. R., eds.), Plenum, New York.

Parson, W. W., Ke, B. (1982). "Energy Conversion in Plants and Bacteria",
Volume 1, 330-385, (Govindjee ed.), Academic Press, New York.

Pearlstein, R. M. (1982). "Energy Conversion in Plants and Bacteria”, Volume 1,
293-330, (Govindjee ed.), Academic Press, New York.

Rabinowitch, E. I. (1945)]. ”Photosynthesis and Related Processes,” Vol I”,
lnterscience, N. .

Renger, G. and Weiss, W. (1983). Biochimica et Biophysica Acta, 722, 1-11.
Renger, G. and Weiss, W. (1986). Biochimica et Biophysica Acta, 850, 184-196.
Rutherford, A. W. (1985). Biochmica et Biophysica Acta, 807, 189.

42

Ruben, S., Randall, M., Kamen, M., and H de, J. L., (1941). Journal of the
American Chemical Socigty, 63, 8 7-890.

Sandusky, P. O., Yocum, C. F. (1983). FEBS Letters, 162, 339-343.
Saygin, O. and Witt, H. T. (1985). Photochemistg and Photobiophysics, 10, 71-
82.

Sayin, O. and Witt, H. T. (1987). Biochimica et Biophysica Acta, 893, 452-469.

Schlodder, E., Brettel, K., and Witt, H. T., (1984). BLochimiciet Biophysica Acta,
808, 123-131.

Sullivan, P. D. and Bolton, J. R. (1968). Journal of the American Chemical
Society, 90, 5366-5370.

Takahashi, Y. and Katoh, S. (1986). Biochimica et Biophysica Acta, 848, 183-
1 92.

Takahashi, Y., Takahashi, M.-A., and Satoh, K. (1986). FEBS Letters, 208, 347-
351 .

Trebst, A. (1986). Z. Naturforsch, 41 c, 240-245.

Van Bes1t,8.é A. and Mathis, P. (1978). Biochmica et Bipphysica Acta, 505, 178-

Van Niel, C. B. (1941). Advances in Enzymolggy, 1, 263-328.
Velthuys, B. R. (1980). Annual Reviews in Plant Physiology, 31, 545-567.

Warburg, O. and Luttgens, W. (1960) Z. Naturforsch, B: Inorganic Chemis’greyJ
Or anic Chemist Biochemist io srcs io , 15 , - .
Warburg, O. (1964). Anntyal Reviews of Biochemim. 33, 1-14.

Weiss, W. and Renger, G. (1986). Biochimica et Biophysica Acta, 850, 173-183.

Wood, P. M. and Bendall, D. S. (1976). Eurogan Journal of Biochemis_t_r_y, 861,
337-344.

Yachandra, V. K., Guiles, R. D., McDermott, A., Britt, R. D., Dexheimer, S. L.,
332.9%?" and Klein, M. P. (1986). _Biochimica et Bipphysica Acta, 850,

Zimmermann, J. L. and Rutherford, A. W. (1984). .&_Biochimica et Biophysica
Acta, 767, 160-167.

Zimmermann, J. L. and Rutherford, A. W. (1985). Physiol. Veg, 23, 425-434.
Zimmermann, J. L. and Rutherford, A. W. (1986). Biochemistgy, 25, 4609-4615.

fWild ir
Observe

aSSOciat
11104 It

Ranges,

the ”Dell”

CHAPTER 2

SPECTROMETER DEVELOPMENT

2.1. INTRODUCTION

It was stated earlier that the primary instrumental technique to be used
was fast absorption spectroscopy. The spectrometer described here and in the
following chapter represents a considerable extension and reworking of the
original design (Lillie, 1984). The prototype was of a single beam design and
used for experiments examining some aspects of the electrochromic shift and of
the proton release patterns related to Photosystem II that had been observed in
chloroplasts. Additionally, It was to serve as a testbed from which to begin
experiments on the primary acceptor, Oa, and the oxygen evolving complex of
PS-ll. A review of the electrochromic shift and proton release can be found by
Junge and Jackson (1982).

The absorption changes associated with the electrochromic shift are
found in the blue green region (450 to 550 nm), and the amplitudes of the
observed changes are on the order of 103. The magnitude of the changes
associated with the OEC in the UV and near-UV (250 to 350 nm) are on the order
of 10'4 to 10's. Our original design was unable to resolve these absorption
changes, and we were forced to re-examine the design of the spectrometer.

In this chapter we will explore the experimental requirements and the
implemented design for non-computer components. Design issues arising from

the experimental requirements that affect the implementation of the computer

43

Ii fir:
tans:
0f the
case i

beam.

Chemis

#2?
A.)
(I)

the deSCI

I01 Obtain

adeSIed II

44

support such as hardware interfacing and software are discussed in the following
chapter. Specific discussion of the analogue to digital converters is also delayed

until the following chapter, though there may be some general comments here
2.2. EXPERIMENTAL REQUIREMENTS

Careful thought and attention to detail are required to bring an
instrumental setup to an operating condition such that the focus is upon the
problem at hand instead of focusing on simply being able to perform an
experiment. The following section, among other topics, highlights subtle but
important considerations we had to take into account during the building and

testing of this experimental apparatus.
Q1. Information to pe Measured.

In general, the information sought is the response of a chemical system
in time after an excitation. The response is measured by changes in light
transmission of a 'probe‘ beam at a given wavelength relative to the transmission
of the probe at the same wavelength before excitation. The excitation in this
case takes the form of a short pulse of intense light perpendicular to the probe
beam. This excitation is tuned to a wavelength optimal for inducing photo-

chemistry and minimal attenuation of the exciting pulse.

2.2.2. Sample Attenuation of Excitigg Pulse

Organic and inorganic photolytic experiments are somewhat similar to
the description just given in terms of the information sought and the method used
for obtaining it. An important difference is with respect to the chromophore's
concentration. The concentration in the photolytic experiments can usually be

adjusted to give easily detectable changes in the light intensity. Biological

45

conditions are such that the chromophore's concentration often cannot be
adjusted above the micro- to nano- molar range without aggregation and optical
density then becoming limiting factors.

The high optical density of the sample creates a problem as it attenuates
the exciting pulse of light as it passes through the sample. Consider Figure 4.8,
a spectrum of a Photosystem I I membrane preparation. The strong absorption
bands in the 650 to 690 nm range are due to chlorophylls from both the light
harvesting complex and the reaction center. Use of an exciting light in this
wavelength range would optimize excitation, but the light itself would be strongly
attenuated and would not uniformly excite across the cross-section of the sample
exposed to the probe light. Use of a wavelength region not strongly absorbed by
the sample would decrease the amount of attenuation and allow for a more

uniform sample exposure; however, the probability for excitation is lessened.

2.2.3. Signal Discrimination

The absorption change for many of the photosynthetic components in the
electron transport chain ranges from 10'5 to 103. The intensity of the detected
signal can be related to the concentration of the chromophore by Beer's Law. The
differential Beer's Law equation is:

-dI/I . aCdx (2.1)

where 'l' is the intensity of the beam before absorption, 'dl' is the fraction
transmitted over the distance 'dx' through which the beam passes, 'C' is the
concentration of molecules undergoing absorption, and 'a' is a constant of the

molecules specifying the probability for absorption. The integrated form is:

I - roe-30X (2.2)

In.‘

pl
Di'
the
an

8X0,

DilOlo
deben
qufiler

PhOIOS

46

or, rearranging,
A - -Iog(|/Io) - aCx. (2.3)

'I0' is the intensity of the measuring light before its absorption by the
sample, and 'I' is the intensity after absorption. 'A,‘ then, is the absolute
absorption of the sample. If the absolute absorption changes by 105, then this
corresponds to a change in the mo ratio by about 0.002% in the intensity of the
measuring light after absorption. It is the ability to resolve the changes in I from
the ratio, I/lo, that is important. Increased sensitivity to small changes in probe
light intensity also results in an increase in sensitivity to the exciting light. The
differences in intensity between the probe and the exciting light will be orders of

magnitude, thus complicating discrimination.

2.2.4. Photochemistgy Induced by the Probe Light.

The probe beam itself can interact with the sample and perform
photochemistry. Figure 4.6 is a plot of the time of exposure of the sample to the
probe beam before excitation versus the maximal absorption change at 320 nm,
the maxima for monitoring the change associated with Oa'IOa. This directly puts
an upper limit on the length of exposure of the sample to the probe before
excitation. This upper limit has other implications discussed in the next section.

2.2.5. Sample Dggndencies on Flash Number.

We also know that the separate electron transfer components in
photosynthetic systems can display non-synchronous damped periodic
dependencies on number and frequency of excitation. The binary and
quaternary oscillations of the reducing and oxidizing sides respectively of

Photosystem II were discussed in Chapter One. At the onset, from a resting

8)

slc
the

late

infinq
succi
SGQUE
W de
seqllel

tribe,a

47

state, the periodic reactions on both sides will be synchronous. The accumulation
of 'misses' on the oxidizing side will allow the two to shift out of phase with
respect to one another. Thus giving rise to the origin of the non-synchronous
behaviour.

Experiments aimed at elucidating kinetic or spectral information
dependent on these periodicities need to have a sample transport mechanism for
the introduction of controlled samples into the excitation chamber for
measurement. Also needed are enough data collection channels for update,

analysis, and storage to characterize the periodicities.
2.2.6. Other Characteristics

Device timing and the ability to have multiple, separate, sequential data
collection sessions is important here as well. As an example, consider an
experiment aimed at collecting absorbance changes from the oxygen evolving
complex. If we are interested in the optical changes that occur as a result of
storage of each oxidizing equivalent and their eventual simultaneous release, the
changes of at least four successive excitations will have to be recorded sepa-
rately.

Figure 2.7 depicts the events that occur per flash, these are known as
lntersequence events. Figure 2.8 depicts the events that occur between
successive flashes for a sequence group of four, these are known as intra-
sequence events. A sequence is defined as a group of pulses that go through
one data collection cycle. In the example given in Figure 2.8 the four successive
sequences will define a single iteration. Sixteen such iterations, for a total of 64
flashes, were required to provide the signal to noise resolution as shown here.

The time from opening the shutter to the beginning of recording data is
deliberately kept to less than 5 ms because of the photochemistry induced by the

minimiz

48

measuring beam (see Figure 4.6), and this implies that the detector system and
electronics must be stabilized to a relatively flat baseline within that time period.
The linearity of the baseline is important because of the lack of a reference that
could be used to correct fluctuations caused by instability in the arc, electronics,
or sample.

Once recording has begun, the probe light source must be stable, either
through feedback control or an accurate reference, as all optical changes are

recorded relative to the baseline established immediately before excitation.
2.3. IMPLEMENTATION

As a result of these deliberations, we settled on an extensive set of
modifications and enhancements to the original spectrometer and computer
system. The rest of this chapter deals with modifications of the spectrometer
(Figure 2.1) and the following chapter discusses the computer system

modifications.
2.3.1. Monochromator and Probe Beam Source

Attenuation of the probe beam was severe below 350 nm when a 100
watt tungsten-halogen lamp was used as the probe source. We were able to
resolve this by using a xenon arc lamp housing and power supply (A-1000 water-
cooled holder and LPS-200 power supply, Photon Technology, Inc.). A 75 watt
Osram xenon lamp was used for illumination. Compared to a lamp using a
filament, xenon arc lamps have a stability problem as the arc tends to wander
around the surface of the electrode tips. The Osram lamp used is of a special
design and features a very small surface area with strongly tapered electrode
tips. This tends to localize the arc into one position so that arc movement is

minimized.

49

The light output from the arc lamp passes through a beam-splitter that
diverts a fraction of the light to an optical feedback unit that regulates the current
the power supply provides to the lamp. Following the beam splitter is an electro-
mechanical shutter (Unlblitz, model 1008, Advanced Electro-mechanical
Engineering; Rochester, NY) and the entrance slit to the monochromator (H-20
monochromator with either a 600 or 1200 gr/mm ion-etched concave holographic
grating, Jobin-Yvon, Instruments S.A.; Metuchen, NJ). The shutter is isolation
mounted separately on the spectrometer platform.

The original concave holographic gratings (made commercially by
Instruments S.A.) generally have a higher spectral purity, though, the efficiency is
lower than with a more conventional Czemy-Turner type. A grating optimized for
the UV was acquired because of recent advances in the manufacturing
technology. An ion-etching process performed on the Master grating supposedly
results in a higher efficiency; though, we have not been able to quantitatively

demonstrate this.

2.3.2. Sample Compartment

The sample compartment is positioned between the monochromator and
the detector housing. Adjustable slit positions on both sides provide for a variety
of entrance pathways relative to the sample cell position for excitation. Typically
we excited perpendicular to the probe beam, but other groups have used
excitation onto the same side of the sample cell as the probe. As the pathlength
along this axis is shorter than along the other path the attenuation of the exciting
beam should be less. Examination of Figure 4.3, a plot of input excitation energy
versus detected response shows that 20% of maximum intensity has essentially

achieved 100% response.

50

Built into the wall holding the face of the detector housing is a
mechanical shutter. This allows us to position components inside of the
compartment without exposing the detector.

The sample cell holders are removable, and are designed for use with a
variety of cell-types. Currently, we have cells suitable for single sample, multiple
flash experiments as well as multiple sample flow-type experiments. One is
currently being designed for variable temperature work. A small rail inside allows

us to position focusing optics for optimal throughput.
2.3.3. Excitation

Sample excitation is provided by a pulsed dye laser (model DL-1100
Phase-R Corporation; New Durham, NH), a xenon flashlamp pumped laser rated
at a maximum power output of 250 mJ/pulse at a repetition rate of 0.25 Hz with a
pulse width of 500 ns at 640 nm (0.5 Megawatts, MW). In practice a more
realistic power output range was between 50 - 70 mJ/pulse (0.1 - 0.14 MW). A
repetition rate of 1 Hz could be sustained but only for short periods of time due to
a high degradation rate of the dye.

This type of laser is well known for emitting considerable radiation in the
radio-frequency range as well as in the visible. Therefore, it is kept in a Faraday
cage as designed by Rabb, Yerkes (1979), and Buttner (1983) to minimize
electronic interference. Electrical interference from its firing is over in less than
100 ns.

2.3.4. Detector(s)

Devices normally used for the conversion of light into electrical signals
are photomultipliers, phototubes, and photodiodes. All rely on a light sensitive

surface to produce an electric current; however, the composition of the surfaces,

eil

inl

phc
that
Phat
field
the /a

timer

dIOp b
release

M399

51

the spectral sensitivity, methods for amplifying the photoelectric current, and the
response times differ considerably. The reader is referred to Malmstadt et al.
(1981) for general information on energy transducers. Specific information on
PMTs and photodiodes can be obtained from individual manufacturers. It is
important to realize that two PMTs, each made by a different manufacturer, may
be completely unsuitable given differences in the materials used. These
differences will show up in mean and maximum anode currents, amplication per
stage, and maximum radiant intensity. An example of this is EMI-9558 (Gencom)
and R622 (Hammamatsu). The Hammamatsu tube is cited as being a re-
placement for the EMI-9558, and it does appear to have an identical S-20
response. The difference arises in the peak anode current. The EMI tube is
rated at 1mA whereas the R622 is rated only for 100 uA.

Photomultipliers (PMTs) and phototubes use a photocathodic surface to
emit photoelectrons in response to an interaction of incident radiant energy with
electrons in the photocathode material; i.e., photoemission. The photoelectrons
in phototubes are collected directly onto the anode by a potential drop between
the photocathode and anode with virtually no amplification. In PMTs the
photoelectrons are accelerated and focused by an electric field to strike a dynode
that in turn emits secondary electrons in response to the impact of the
photoelectrons. These secondary electrons are in turn accelerated by an electric
field toward the next dynode. This process is repeated until the electrons from
the last dynode are collected by the anode, and leave the PMT as an electric
current.

The electric field responsible for the acceleration is created by a potential
drop between each successive element. The number of secondary electrons
released per impact is a function of the photocathode material and the potential
voltage drop between successive electrodes. Typically the amplification

52

experienced ranges from two to four depending upon the potential drop; there is
an upper limit to the drop supplied.

Photodiodes rely on the light sensitivity of pn semiconductor junctions to
generate a reverse-biased current. They do not possess the inherent
amplification nor the spectral range of PMTs; however, photodiodes are capable
of responding in the sub-nanosecond time scale, an order of magnitude or more
faster than the response of PMTs. Often they are packaged with an internal
amplifier to obtain better amplification while keeping the fast response.

Our detection system was designed for two detectors to be used
depending upon the wavelength region of interest. Below 750 nm an EMI-
9558OB photomultiplier tube (EMI-9558OB, 11-stage venetian blind, S-20
response, EMI Gencom, lnc., Plainview, NY) was used, and above 750 nm an
FND-100H photodiode (EG&G Photon Devices, Salem, MA) was used. The
mounts for each detector were made so that the housing could accomodate
either with no other change in the setup of the apparatus. The photodiode and its
accompanying circuitry fit onto a 3 inch diameter circuit board at the head of the
detector housing. A cable transversing the distance between the circuit card and
the back of the PMT where the pro-amplifiers are attached Is required and is line
balanced to minimize signal degradation.

Photodiodes have been best applied in the near-I R in time resolving
primary donor reactions of the reaction centers. As discussed in Chapter One,
the primary donor in photosynthesis is contained in the reaction center. The re-
action center in green plants, P680+’ has a characteristic absorption band at 820
nm. One of the first diode lasers, used with a diode detector, emits a usable line
at 820 nm, and Van Best and Mathis (1979) were able to obtain extremely well

resolved kinetic decays of P680+ in the nanosecond time regime.

’9:

ex;

and F
reQUln

53

Though the use of photodiodes has been limited, recent developments
have significantly enhanced the intrinsic signal amplification properties, and have
extended the spectral responses of photodiodes further into the blue. The devel-
opment of 'diode array' spectrometers by Hewlett-Packard and Perkin-Elmer,
capable of collecting an entire optical absorption spectra at once, hold promise
for the development of instrumentation capable of time-resolving entire spectra of
fast changing multi-component systems such as those found in photosynthesis.
This would represent a considerable enhancement over the current method of

taking such a spectrum, which must now be done point-wise.
2.3.4.1. PMT Divider Chains

The potential drop, discussed in the previous section as responsible for
the acceleration of the emitted secondary photoelectrons, is usually established
by a resistive divider chain, though individual voltage sources could be used
instead. Typical divider chain configurations are shown in Figure 2.5 (Dumont,
1963 and RCA, 1985). Normally the design of the divider chain is not part of the
design consideration of the detector system. Standard configurations are
commercially available for common applications; however, in our case, the
requirements are such that the standard configurations are unsatisfactory as
explained below.

In the literature, usage of PMTs and discussions concerning their usage
falls into one of two classes: 1) low intensity light levels, relatively low time
resolution, continuous illumination as found in a more typical scanning absorption
spectrometer, or 2) high intensity, pulsed light levels, applications as found in a
Raman spectrometer. in the former case a quick response of the divider chain
and PMT to changes in the light intensity is unimportant, and the current

requirements at the anode are low. Pulse applications, however, have high

Si
us
the
a n

988

and is

54

current requirement for brief periods, and the divider chains employ capacitors to
act as current sources during heavy loads. Between pulses the capacitors
recharge, and the time required for recharge is basically the RC time constant.
This recharge problem makes their design untenable in a situation with a
continuous high current.

Our experimental conditions require that the PMT and the divider chain
have a linear and accurate response to small changes of light in high background
levels of monochromatic radiation for periods of time lasting up to several
seconds. In addition, the divider chain should have the capacity to recover
quickly from sudden pulses of light that are orders of magnitude more intense
than background incidence. Standard divider designs that accomodate low in-
tensity or pulsed applications do not suffice.

If the current measured at the anode is greater than 5% of the current
available in the divider chain then the divider chain has to be allowed time for
recharging. This recharge time is dependent upon the current at the anode and
the RC time constant. If the expected current will exceed this 5 % limit the
simplest solution is to use resistors with a lower resistance. As an example, If the
usual operating voltage is 1000 v, each resistor is valued at 200 Kohms, and
there are 10 stages; the total current available would be 0.5 mA. This would allow
a maximum current at the anode of 0.025 mA. The maximum current could
easily be doubled by using 100 Kohm resistors instead; however, the heat
radiated by the resistors, l2R, would now be 25 mW, up by a factor of 4 from 6
mW. Unless cooled, or forced air ventilation, or both is used in the housing the
additional heat production will have a deleterious effect on the PMT, and the
divider chain, and an inaccurate and noisy output signal will result.

A reasonable design for the divider chain was reached by trial and error

and is as shown in Figure 2.6. Capacitors are wired in parallel instead of in

I0 Ilne
Equai
IIIIOrm.

beam t

E

55

series as shown in Figure 2.5. Referencing the dynodes to ground localizes any
possible, sudden drop in the voltage between any two dynodes as a result of
current surges. The first two dynodes use 100 V zener diodes to force a constant
voltage drop. Dynodes #9 through #11 are open, and dynode #8 serves as the
anode. This reduces the amplification by roughly a factor of 10 per stage for a
total reduction of 1000, but the cathode can now be operated at a higher and

therefore more sensitive potential.
2.3.4.2. Simple Tests

We were able to get an accurate response model from deep in the UV
(250 nm) into the blue-green (350 nm), see Figure 2.10. Further, this
demonstrates that fractional changes in relative concentrations, dilutions by 0.95
and 0.90, are shown to be reflected correctly. Figure 2.11 shows the linearity of
the photomultiplier at 320 nm at 1/10 of the amplification used in real

experiments.

2.3.5. Analpg Electronics

The signal processing performed by the analog electronics is shown in
Figure 2.12. The reasons for two signal processing channels will be explained in

the following sections.
2. 3. 5. 1 . Relating Measured Light Intensities to Absorption

Beer's law, as described earlier in Equations 2.1, 2.2, and 2.3, allowed us
to linearly relate the changes of observed light intensity to concentration. From
Equation 2.3, this defined the absolute absorption of the sample, and required
lnfonnation about 'lo', the incident radiant intensity. Experimentally, with a single

beam spectrometer, the determination of 'o is awkward. Fortunately, we are only

56

interested in the absorption relative to the absorption before excitation. We can

thus write:

dA - Af - Ai
. -Iog(Il/l‘o) - (4090910)

. Iog(Ii/Iio) - Iog(If/Ilo) (2.4)

'Iio' and 'Iio' are the radiant intensities upon the sample before and during
excitation and it is reasonable to assume that these do not change; i.e., Ifo - lio.

Rearranging cancels out the unabsorbed radiant intensity terms and we have:
dA . iog(if/Ii) (2.5)

The quantities Ii and If can be replaced by their voltage equivalents as the current
response of either detector is linear with respect to light intensity. Substituting
V0" for Ii and rewriting If in terms of V0“ and dV:

dA . Iog(Vofl/(Vofi - dV)) (2.6)

Use of Equation 2.6 indicates that determination of relative absorption
changes requires information about the initial amount of transmitted light through
the sample, V0“, and continuous information regarding the change in Intensity
relative to the initial value, dV. This is still an awkward equation to deal with
experimentally, but only from the aspect of performing the mathematical
manipulations, the division and logarithmic operations, not from the aspect of

selecting the individual values.

 

SE
0U

ca;

the

”819
Signal.
Meier
Signal I
F'lUie :

57
2.3.5.2. Detector Amplifier Circuitry.

Obviously, the difference, dV, in Equation 2.6 is critical as it is a direct
determination of the extent of the changes of the photosynthetic samples.
Analogue to digital converters (ADCs) typically span a voltage range, 9.9., +5.0 V
to -5.0 V. If an ADC has 8 bits, that means that a 10 V span will have 256
divisions, or 39 mV per division. An absorption change of 10‘4 on a background
of 10 V corresponds to a change of 40 mV; i.e., 1 division. If the 10 V
. background level was removed from the signal a maximum amplification factor of
250 could be applied to fill the digitizing window of the ADCs (see Figure 2.9)
with kinetic information as opposed to just the background.

The separation of the analogue signal into separate baseline and
differential channels required a complete redesign of the detector pre-amplifier
and amplifier circuitry (Figure 2.2) as well as of the storage ADC itself. The sig-
nal from the detector and pre-amplifier are simultaneously input into a tracking
sample and hold amplifier (SHA) and a normal cascade amplifier chain. The
output signal from the tracking SHA is split into two. One side is connected to a
capacitor referenced to ground. This has the effect of storing a charge on the
capacitor equal to the voltage output from the tracking SHA, and this charge
changes as the output changes. The other output from the tracking SHA is
inverted and summed (resulting in a subtraction operation) with the original signal
to give a final signal that is zeroed. When triggered, the SHA enters into a 'hold'
mode, the pro-amplifier input is disconnected from the capacitor, and uses the
charge stored on the capacitor to provide its output; this then forms the reference
signal. Changes in input levels are summed with respect to the inverted original
reference, and the final output signal is one in which the background reference
signal has been removed; the effects of this on the signal are also shown in

Figure 2.7.

58

Two difficulties with this design revolve around the use of an analog

SHA. A consistent droop is induced into the output signal during the hold phase,
and this is because the discharge of the capacitor is exponential. There are also
slight phase problems resulting in the introduction of some spurious frequencies,
and these are related again to the use of a capacitor. Replacing the analogue
capacitor storage system with a digital system that uses a separate analogue to
digital and digital to analogue converters would solve both of these problems.
When in a hold mode, the reference charge would be maintained digitally and

would not droop.
2. 3.5. 3. Final Filtering.

One of the drawbacks to digital sampling of analogue signals is the
possibility for aliasing - the introduction of spurious frequencies into a sampled
digital signal due to undersampling of high frequency components (Cooper,
1978) These high frequency components are those that are present in the input
signal that are higher than allowed by the Nyquist sampling theorem. (a rather
simple explanation of this theorem is given by Swanson et al., 1975) This
sampling theorem states that the sampling rate must be greater than twice the
highest frequency component in the signal. As an example, if a signal contains a
component of of 1 MHz frequency, than the signal must be sampled at a rate no
slower than 2 MHz.

To prevent the introduction of these high frequencies as spurious low
frequencies, a fourth order filter (see Figure 2.3 for the diagram and Figure 2.4
for rolloff characteristics) was designed and used to process the differential
channel signal before digitization to prevent aliasing errors. While not proving to
be critical for the information presented in this dissertation because of the

relatively low sampling rates reported here, a distinct benefit accrued as the filter

59

eliminated high frequency ringing components that occasionally triggered the
entire system into oscillation.

As an aside, it is not quite true that sampling must be performed at twice
the desired signal frequency for accurate digitization. A practice known as
interleaving may be performed and, depending upon the degree of interleaving,
essentially the frequency of the signal to be sampled will be independent of the
frequency at which the signal can be sampled. A discussion on this interleaving
may be found by Verdun et al. (1988). Basically, assume a signal with a
frequency of 1 MHz and a maximum sampling rate of 1 MHz. The first
digitization run on the data is stored and then the second digitization run is
delayed by one half of thes sampling period. Store this second run with the first
run and the two summed together equal one digitization run sampled at 2 MHz.
In this fashion high frequency signal frequencies can be sampled with less
expensive components with only a tradeoff in terms of the total amount of time

required to collect the data.
2.4. CONCLUSION

Frequently, problem occurrences were masked by others. As an
example, one baffling problem, though simply solved, was caused by the rotation
of a bar magnet in a stir plate near the output light beam from the xenon arc
lamp. There was no effect on the low-impedance analog signal lines that were
positioned nearby, nor was was there any coupling of a mechanical vibration onto
the spectrometer platform. However, a distinct, ~60 Hz oscillation of the probe
beam intensity was observed, and this is usually indicative of AC line current
coupling into the signal lines. It finally turned out that the rotation of a magnetic
bar In a nearby stir plate at a frequency of ~60 Hz was inducing the oscillation

observed. Only through a coincidental movement of the stir plate and noticing

60

the removal of the periodic component from the probe beam on an oscillation

was this particular problem resolved.
2.4.1. Susceptibility to mechanical vibrations

The spectrometer, light source, monochromator, detector, and all
structural supports, were mounted on a heavy rigid platform. This platform was
supported by Iabjacks that were in turn isolation mounted on a regular laboratory
bench. Isolation mounting can only be done when the platform is heavy enough
to compress the isolation cushion by a significant amount. Otherwise,
mechanical vibrations continue to couple into the platform. This particular
problem is expected to be rectified when the equipment is remounted on an

isolation table designed for this.

2.4.2. Imprecise Omical Feedback Control of Xenon Arc Lamp.

A second problem, though, is much more difficult to solve and rises
largely because of the single beam design. The small area design of the
electrode tips and current feedback loop of the xenon arc lamp allow the stability
of the probe beam to be only tolerable at best. While better stability can be
achieved with a tungsten-halogen lamp, the lamp output is decreased
considerably, particularly in the UV and near-UV, and the detection problems
encountered in the prototype due to too low of a probe beam intensity resurface.

The self-correcting aspect of a dual beam spectrometer can be obtained
via an additional feed-back path. A beam-splitter could be inserted into the path
of the probe and a fraction diverted to a photodiode. The output from the pho-
todiode, properly amplified, could be used to correct fluctuations in the PMT
output due solely to fluctuations in the arc lamp. The only remaining optical

problem then remaining would be uncorrected changes in sample absorption due

61

to probe beam photochemistry, and, as demonstrated, if the exposure before

excitation is kept under 15 ms this will not be a problem.
2.4.3. Alternative Approaches

Apparently, the group led by Van Gorkom in Germany (Dekker, 1988)
has taken a somewhat different tack in their experimental setup. It was earlier
shown that the magnitude of the intensity changes in a probe beam is related to
the intensity of the probe. Therefore, the traditional approach has been to use as
intense of a probe beam as possible given other constraints such as probe-
induced photochemistry (Junge, 1976). According to Dekker, their group uses a
defocused probe of low intensity. This allows them to run their optical
experiments with the samples continuously exposed to the probe light.

From a technical point of view these differences in experimental
technique has two principal advantages: 1) Mechanical vibrations that couple into
the platform from the opening and closing of the shutter are no longer a factor as
the shutter is left open during the course of an experiment. 2) Steady state and
transient response demands of the PMT are not as severe as the constraints
imposed by our design. Our design requires that the PMT has recovered and is
tracking accurately with a few milliseconds after the shutter has opened and the
sample exposed to an intense probe beam.

From an experimental point of view, this methodology raises the question
of what is happening with the sample that is being continually exposed to low-
level Illumination. It is entirely possible that the low-level illumination modifies the
short term dark adapted S-state distribution and this in turn would affect the S-
state absorbance spectra. Model results bear this out; Lavergne (1986) and
Saygin and Witt (1985) have modeled transient absorbance changes of the OEC

that clearly show differences of the absorption change patterns merely by altering

62

the initial S-state distribution. Unfortunately the existing experimental data

(Saygin and Witt, 1985) does not select one distribution over another.

63

FIGURE 2.1

_Block Diamam of Flash Apsorption Sgctrometer

Dashed lines indicate control lines, block arrow lines are data pathways.
single lines are signal carrying lines (either analogue or digital). Components in
this Diagram are from those discussed in Chapters Two and Three. The PDP-11
computer is shown as being central. Data pathways go out to the Timing
Controller and to the Fast/Ultrafast ADC. The Timing Controller provides the
direct control over the other components shown. The Fast/Ultrafast ADC

digitizes the analogue signal output from the Amplifier, and is transferred.

64

 

 

{till it. _ L

1. Fr

 

 

.3230 van Bea

 

 

 

   

 

 

 

 

 

 

 

 

twp—41:00 Brazos—via
l. -. :Idoa Illl .I - it
0
£3.33 12:25 .
3.21:". <
a lliiil .Illr
IIIIIIIIIIII 53.6528 e25:
,. uo< _ _rriii iiiiiii I! I. III
I-.II;.II-II, 5:52:35: IIIIIII _ _n. _ali II
_I u. _
_ _ lIlr _
_ H _
IIIIIII. lllllllllllllllllllllll Li lllll _
PM a _ a
_
_ _ u
52:53 5:: 3:1on . m
_
5.622 I: -ilrli I, I IIIIIIII <
4

I, , 32:355. m_>i>:
Elana _ zuijdiiumcl— .

22:14:». lilmii. _ _ :
filthmwl. , IIDIIU

_
_
_
_
_
_
_
_
_
_
_
_
Ii. ill .Zzofi izEm “
.2522;qu .u.o I u
. 1 III: III
_
_
A\1 5:32.: .
_
. ..l _ 5:: zoiaoxxoozoa "
I I .III.. III _
. .
. . .
_

      

mzuhiu Aqua mzu;
unchao $0.: (0...."

xu<ulawmu

w
/

4(0316

L3<4loz<

202ux

65

FIGURE 2.2
Amplifier Circgitry for PMT.

This circuitry was designed by Marty Rabb and Dwight Lillie, and
provides cascade amplification and active low-pass filtering. It has a maximum
frequency response of 2 Mhz. Two output signals are provided; SIGNAL
OUTPUT, contains the kinetic signal along with the baseline, and RESTORED
SIGNAL OUTPUT, which has had the baseline subtracted.

66

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

’mo
UCDOLD so
a yawn-.Doficmhw 8a.: . .5. . x Z_ Lemmip
u m L ,.
a @ . . e 0 s onBaccm
o _ “zoz
_ _ . 0.95m IdI—I
>m_- sn_.
UcJOLo :3ch 411%
.w uJQCJD 230% r C UcJOLo

 

p.39: Gama

Eco son

doom
soon.
58m
.589
:23.
3 m8 o
32,:
an. o

 

uccwmcou ms;

 

IIIIIIII

67

FIGURE 2.3
mirth (fler low-pass filter.

This is a series of two cascaded multiple feedback 2nd order low pass
filters. The response characteristics of this filter compared to several others is in

the next Figure. C1 8 IOOPF, CZ I 22pF, R1 8 1K, R2 I 4.7K.

68

 

 

 

 

JI

 

 

 

 

 

 

 

 

 

 

 

 

69

FIGURE 2.4
Roll-off characteristics of Active Low Pass Filters

These characteristics were collected as a result of building and
evaluating a series of active low pass filters for final filtering. '.' 4th order filter
built from two cascaded 2nd 0.3dB Chebyshev filters. 'x' Modified 2nd order
0.3db Chebyshev. '*' 2nd order 0.3dB Butterworth. ’0' 2nd order 0.3db
Chebyshev. '+' Simple passive RC filter. Most advantageous about the filter
selected, the 4th order filter, is the steep roll-off characteristics at the mo - 1.0
point. Two of the 2nd order filters exhibit significant attenuation before the 0.3dB
point, and the other two have significant oscillation before (not shown due to

scale of plot) the turnover point.

dB

0.

“IO.

- 50

-40.

-60.

7O

 

 

 

(I.

l/iC (iC

f0

SOOKHZ)

71

FIGURE 2.5

 

Typical Dynode Chains for PMTs

Two designs are shown here. The first design is typical of that used in
relatively low intensity, continuously scanning spectrometers. The second design
is typical of that found in Raman spectrometers. The series connected ca-
pacitors are required only for those dynodes which may be expected to have a
peak dynode current of 1/10 of the average current through the entire network.
Series connecting the dynodes allows use of lower rated capacitors. R1 is
typically three times the resistance of R2 Rn. C1, CZ, and C3 are on the order of
0.005 uF.

72

Cathode Anode

WW

R1 Ra R3 R4 Fir-28,481 RL

Cathode R1 R2 R3 R4 neRn_1Rn Anode

QM

C, ca (:3 RL

73

FIGURE 2.6
Dynode Wirinngsign.

This figure shows the divider chain design we adopted. In place of the
first two resistors of Figure 2.5, R1 and R2, two Zener dynodes are used instead.
This insures that the potential drop between the first two dynodes does not
change. The rest of the resistors are of the same value. The last four dynodes
are left unconnected and dynode 7 is used as the anode. Dynodes 4 through 7
have parallel connected capacitors instead of series connected. This requires
capacitors of higher rating, but changes in the potential drop are localized instead

of being echoed to neighboring dynodes.

74

a
__..\m.\\.\§§r 9Q RC
QII.

 

 

 

 

 

 

 

 

a><irria><iiria><, es< +4- ‘ 42 ‘ >1-
: 5 _lm.

2.33:9 an: 2;.
L a TC 32 @355

75

FIGURE 2.7

Events during a Single Flash Seguence

This figure depicts the events that occur during a single sequence. This
particular sequence is part of an experiment which is doing repetitive single shot
excitation of a sample. Between sequences the sample chamber is continuously
flushed from a resovoir of photosynthetic material. Both of the amplifier outputs,
which are digitized by the fast and slow Analog to Digital converters, are shown
and their relative level(s) with respect to one another. The slow ADC is only
concerned with establishment of a baseline level so it handles the Output Signal
line. Note that the 'RESTORED' Output Signal only contains useful information
after the ADCs are triggered to begin conversion, Pulse 4. The ADCs only

convert for a finite period of time and will complete before the shutter is closed,
Pulse #6.

76

9:2... was s2 E2. .3

88 ESE. EM _ amaze . _

. . 83
.5 ea 5.32.: _ E
_

843$me
_
m _ n :5:

 

 

m N

\o
If)
wr
m

.3:

 

‘
-—_——_—

N
-__—————

——--—-I————_ _—

é

 

".23 see

 

 

 

———————.___

 

w
_
_
_
_
_
_
_
_
_
_

§
.5

 

 

”
_
_
_
c
_
_
_
_
_
_
_
L
_ _ as sea SEE
_

 

I

 

77

FIGURE 2.8

Occurrence of Events During Multiple Seguences.

This figure depicts some of the events that must occur during between
each flash sequence. The example chosen here is a four flash iteration and each
flash sequence is numbered in the order it occurred. The delay times between
each flash sequence is constant, except between the fourth and the start of the
next iteration. The pump for purging the sample chamber is shut down at the
beginning of a cycle and started at the end of the same cycle, the intervening
flash sequences do not trigger it. This figure coincidentally displays one of the
difficulties we had with multiple flash iterations. Each trace is the average of 16

iterations. The large droop at the the first flash is the result of a door slamming
during one flash sequence.

78

 

 

as: >32» nacho coco-5313.35 >30“.
38: coo."
can: non-3:0 39.: 52.
32.095» 5 venoun use €055.
€338: Bot-3d: 3503 :3
.333 5523'—

-- €333}
aegis; _ e g; a

 

 

 

_=
l noomuuu
l nous: N. 0.5.”. >323 coco—Gonnanucw >IHou
n. '3 >58 35.531335 >33 9.85.» cued
nus—5.» vdoH 4320955 5 Union.» van 6935
run—0&0... EM magnum Man Goa-wan _vwnnououn €026.35 wagon :3
sun-nous vovuoa = 350 .3
@ Reno: mauve—:3 @ .uuuoHo nos—£3

 

 

 

a exegesis; _

old» vuou haunt
anon—.3 vac.—

asea a gs.

 

some?» ass: essence Sea-e doom T
:1 IS» >33 coco—EonI-uufi >10“.
38:» need
3309.3» 5. concun Us. 68:5- 9
@ 60:33.... 6333: 352. '3

3:30 mono—£3

79

FIGURE 2.9
Removal of background level 'from signal.

This figure contains two plots of simulated data. The top plot shows a
kinetic signal with a maximum change of 40 mV plus a background of 10 V. This
is typical of a normal signal. Resolution of the kinetics is impossible because the
background completely fills the digitizing window. The bottom plot shows the
kinetic signal after subtraction of the background and subsequent amplification by

250. This half fills the digitizing window.

80

Ans“ «Em»
cow cap on, ov— ow. oo— on om 0v on
F? .— b h h —l r P r h b b P

P s n . P

L

AuEv oEWP
owl can our cow cv— cup cow an ac
PleplPPll-prhbbl—rP-lb

 

onw x Auc__amop i _oco_m >EovV

 

 

ow

h

r

ow

_

oc__ouoa >o— ns.; .ocm.» >sov

h

"'0

l‘

O
fi—‘fi‘-f-‘C—T-W_1—'Ff_T-m—U-T—r_f—?.PHFN

op

F—

(A) 133133 “Olidiocqy

81

Figure 2.10
PMT accuracy response from 250 nm to 350 nm.

This figure demonstrates the linearity and discrimination of the PMT and
the amplification system from 250 to 350 nm. ferricyanide at a stock
concentration of 0.625 M in water was diluted to relative concentrations of 0.95
(0.5938 M) and 0.90 (0.5625 M). The absorption spectra of each dilution was
taken with the PAS spectrometer. Data was collected by the Nicolet 1074
operating in a slow sweep mode while the grating of the monochromater was
driven mechanically from 250 nm to 350 nm. PMT voltage was 600 V, ampli-

fication at 1/10. Absolute absorption scale was greater than 2.0, scale shown

has been normalized.

 

82

"1.0
---O.95

-----o.9o

T l I I I

300.00

Wavelength (nm)

83

FIGURE 2.11
PMT Linearity .

This figure demonstrates the linearity of the EMI-955808 PMT and the
amplification system at 300 nm from 0% to 100% transmission. The amplification
factor was 0.1 and the voltage range output from the system was 10.4 V at
100% T and 0.0 V at 0% T. PMT voltage was set to 600 V. The transmission
was determined through a set of neutral density filters. Three separate sets of
experiments were conducted over the course of three weeks. One test occurred
immediately after turning on the system, the second test occured just before an
experiment, and the third set occured just after a 12 hour experiment (the PMT

was not exposed ambient light during the entire 12 hours).

84

3...:

__..:..-...0.: .._.,w.i:....-.._r o

It

.9... .E E.
p . _ _ L

o +«

.,,. .-_ 2. . _. -.._ a. ”.W.,. : a .. g m m. m

.2 2 m

 

med...

nknx.c

E. a:

85

FIGURE 2.1 2
Layout of Analog Signal Processigg

The separation of the analog signal from the PMT into two channels for
signal processing is demonstrated. The analog signal from the PMT is
comprised of both the background and the kinetic signal components. The SHA
is used to lock onto the background level, this is then inverted and subtracted
from the original analog signal. The result is a separate channel containing only
the kinetic components. The kinetic signal is amplified to fill the digitizer window,
digitized, and stored in the ADC memory buffers. The background level is
digitized and stored in the ADC memory buffers. The digitized data is retrieved at

some point under computer control.

ENS

 

0.6. SIGNAL (Bookground) Aoolog oIgnoI
o KINETIC SIGNAL ((0.013) from Ootooior

l :

 

 

 

Ioporoiion of 0.0.
or“ lIuotIo oompononto
of Anolog SIgnol

 

 

 

 

 

 

 

 

o.c. swim. xmmc sicuAL "
ONLY ONLY 1
i 0.0. Slgool Look In

Kinoiio IIgnoI Amollflooiloo
ouo fliiorln.

 

 
  

 
   
   
   

  
   
   
 

ILO'
RESOLUTION
ADO

FAST
RESOLUTION
ADO

convoroloo

ll

Olgliol Ilgnol Prooooolo.
of 0.3. ood Klnoilo Signolo
(In horoooro)

 

COMPUTER PROCESSING

87

REFERENCES
Buttner, B., (1983) Ph.D. Thesis, "Induced and Natural Delayed Luminescence of
Green Plant Photosynthesis”, M.S.U.

Cooper, J. W. (1978). 'Tranform Techniques in Chemistry," Chapter 4, 72—74,
(Griffiths, P. R., ed.), Plenum Press, N.Y.

Dekker, J. P. (1988). Private communication with Dr. Dekker after he accepted a
joint Post-Doctoral position with Professors G. T. Babcock (M.S.U) an C.
F. Yocum (Univ. of Mich.).

”Du Mont Multiplier Phototubes" (1963). Fairchild Dumont Laboratories, Clifton,
NJ.

Junge, w., (1976) ”Chemistry and Biochemistry of Plant Pigments,” Volume II,
Chapter 22, (Goodwin, T. W., ed.), Academic Press, N.Y.

Junge, W. and Jackson, J. B. (1982). ”Photosynthesis: Ener y Conversion by
Plants and Bacteria”, Volume I, Chapter 13, (Gavin ]68, ed.) Academic
Press, N.Y.

Lavergne, J. (1986). Photochemist_ry and Photobiolggy, 43, 311-317.

Lillie, D., (1984) MS. Thesis, “The Development and Characterization of a Flash
K netlc Absorption Spectrometer”, M.S.U.

Malmstadt, H. V., Enke, C. G., and Crouch, S. R. (1981). ”Electronics and
Instrumentation for Scientists”, Chapters 3 and 4, The
Benjamin/Cummings Publishing Company, Inc., CA

"RCA Photomultiplier Manual” (1985). RCA Corporation.

Sawnson, R., Thoennes, D. J., Williams, R. C., and Wilkins, C. L. (1975). Journal
of Chemical Education, 52, 530-533.

Saygin, $2. and Witt, H. T. (1985). Photochemistry and Photobioghysics, 10, 71 -

Yerkes, C., (1979) MS. Thesis, "The Kinetics of Electron Transport in
Chloroplast Photosystem ll", M.S.U.

Van Best, J., and Mathis, P. (1979). Photochemist_ry Photobiolggy, 31, 89-92.

Verdun, F. R., Ricca, T. L., and Marshall, A. G., (1988). Applied Smctroscogx,
42, 199-203.

CHAPTER 3

INSTRUMENT DEVELOPMENT - COMPUTER SUPPORT

3.1 . Introduction

Complementing the efforts in the spectrometer design have been equal
efforts in the areas of computer hardware, software, and interface design to
support the instrumental apparatus. The computer used In the prototype was an
AIM-65 microcomputer (Rockwell Industries) with 4Kb of RAM memory. The
original program development facilities consisted entirely of a BASIC interpreter
in ROM, an Assembler In ROM, a primitive line editor, manual cassette tape for
mass storage, a 20 character wide single line LED display, and a 20 character
wide thermal printer.

* The first effort added hardware to support communication with a video or
hardcopy terminal. This modification resulted in our ability to examine listings
and quickly determine errors in the code. We also connected this communication
port to a PDP-11/02, developed communications software, and allowed the PDP-
11 to handle the output from the Aim-65. At this point we could then conduct an
experiment, and transfer the results from the NICOLET 1074 through the Aim-65
to the PDP-11. The communications software on the PDP-11 would transfer this
information to a disk file. A typical experiment would collect 1023 points and the
time to transfer these points out of the NICOLET 1074 to the disk file on the PDP-
11 exceeded 90 seconds.

88

89

The second attempt upgraded the memory of the AIM-65 to 32Kb, re-
wrote the Aim-65 control Monitor, wrote a cross-assembler for the 6502
processor to compile the re-written monitor program, added code in the monitor
to support true interrupt communications, added additional hardware for the
communications, and added the FORTH language. The memory upgrade was to
allow a set of successive turnovers of the reaction center to be collected in
separate data channels. Rewriting the control monitor and improved
communications hardware decreased the time required for data transfer between
the AIM-65 and the PDP-f 1. FORTH is a language has proven to be quite
versatile in instrument control systems and it had been hoped we could bypass
the difficulties being encountered in mixing BASIC and 6502 assembly.

We were still hampered by mass-storage deficiencies, large timing
granularities (discussed later), inadequate memory, and a slow processor.
Owing to the access in our laboratory to a Digital Equipment Corporation (DEC)
PDP11/23 with the RSX-11M operating system, it was decided to forego any
further development with the Aim-65 and base future computer development and
support on the PDP-11 rather than continue with the AIM-65 or purchase a IBM-
PC type computer. Comparisons of the resources of the Aim-65 and the IBM-PC
computer relative to the PDP-11 and IBM-PC are listed in Tables 3.1 and 3.2.
Another option available would have been to use a PDP-11 with RT-11, a single-
tasking, single user operating system. However, the limitations of the PC based
solutions, mainly single task execution, would also have existed.

This chapter will establish the computer resources needed and detail the
hardware, software, and instrument interface designed to support the FAS
spectrometer. Finally, we discuss whether the path chosen was wise in light of

the cost and predominance of PC-type computers and software.

90

This chapter and the previous one examine at a detailed level the
implementation of the FASCI on a component and system level. The user
manual for use of the instrument is presented in Appendix A. This should provide

a different view of the instrument and its capabilities.
3.1.1. Overview

In the previous chapter we examined in some detail the experimental
scenarios and the instrumental requirements. While it is not necessary to go into

that detail again, we will return to Figure 2.8 and Equation 2.6.
3.1.1.1. Timing Control for Events.

Figure 2.8 reveals that a number of events occur during the course of an
experiment, and that they all occur in controlled time intervals. The timing pattern
shown in this particular case is complex and common enough to require that it be
easily performed. Time resolution ranging from seconds to microseconds is
required for lntra-sequence time delays. An intra-sequence time delay is the
delay between two consecutive pulses occuring within a sequence. A resolution
of seconds to minutes are required for the inter-sequence time delays. An inter-
sequence time delay is the delay between two consecutive sequences. Adding to
the complexity is the need for inter-sequence delays to be variable to
accomodate larger cycles that occur over several sequences, and the need to
add or delete individual pulses on a sequence dependent basis.

In the particular case of Flgure 2.8, there are four channels of data that
must be independently stored, manipulated, and displayed. Individual banks of
memory can be built and access protocols devised so that the ADC supports four
separate chanels. Simultaneous read of one channel by the computer and write

on another channel by the ADC is allowed; however, the maximum number of

91

separate channels is in the range of 16 to 32 channels, and this then incurs a
prohibitive cost.

An easier solution to the channel problem can be arrived at by using the
computer to store the channels. The need for more channels than originally
designed can be adjusted by modifying the storage space requested by the
programs and changing the algorithm limits to accomodate the new upper bound.
The difficulty of designing a hardwired timing controller with the ability for
conditional pulse generation and alternative delay cycles can also be avoided by
using a computer and a series of programmable digital timers with maskable or

optionally selectable output signals.
3.1.1.2. Absorption Determination

Equation 2.6 reveals that the information collected by the ADCs, the inital
offset voltage and a series of difference voltages, can be used to determine the
relative absorption changes. Analogue circuitry can be constructed to perform
the division and logarithmic calculations; however, they operate at a rate much
slower than the upper data collection rates. Again, the computer can be used to
receive the raw binary values output from the ADCs and to perform the

necessary calculations.

3.1.2. Imgementation Philosophy

The design of computer based instrumentation is challenging because
questions continually arise over whether a particular functionality is be
implemented in hardware or handled by a software algorithm. In general the
implementation of functions in hardware are faster in terms of execution time
than if coded in software algorithms, however, a much longer time is required in

achieving a working model.

92

The question takes on special importance in this instance as we have
chosen to base the computer support on a multiuser, multitasking computing
system in our laboratory . The potential to affect other users as well as to
degrade significantly the performance of the FAS spectrometer is large. As an
example of the data throughput rates the hardware could provide, the ADC was
originally designed to collect and temporarily store 1024 points of 8-bit data, and
then to transfer the data immediately into core storage of the NICOLET 1074 at a
rate of 200 KHz. Attempting to transfer this quantity of information at this rate
can be handled be a dedicated computer; however, a multiuser computer would
have such a performance degradation that it would appear to other users to have
become inoperable ("crash" is the term coined by computer scientists). It is for
this reason that even much more powerful computers than the PDP-f 1, such as
the VAX-11, a 32-bit minicomputer, are generally used singly, or in a 'stand

alone' mode, when involved in real-time instrument work.
3.2. HARDWARE ARCHITECTURE IMPLEMENTATION

Figure 2.1 depicts the layout of the hardware supporting the FAS
spectrometer. In the last chapter we were only concerned with the analogue
components of the spectrometer itself; i.e, optical and signal processing. Now
we'll examine some of the other components.

The digital components can be divided into roughly three systems. The
first system is responsible for timing pulse generation and distribution of the
pulses to the device. The second system is concerned with digitizing the analog
data and transfer to the computer itself. The third system, the NICOLET 1074
and display scope, is held over in its entirety from the prototype and serves as a

real-time graphics display subsystem.

93
3.2.1. Timing System

Replacing the simple timer-chip and interupt service routine concept
used in the prototype, a timing board (Codar Q-TIMER) with 16 independently
programmable channels with a 16-bit count resolution and four independently
selectable counting frequencies is used instead. The board is inserted into the
Q-BUS (or backplane; an electrical system designed for the systematic
distribution of digital signals in a computer system) of the PDP-11 computer.
Each timing channel can be programmably enabled to activate a vectored
interrupt upon completion of the count as well as triggering one or more channels
to begin counting. The logical detail of this is shown in Flgure 3.1.

The major advantage of this device in generating timing pulses is the
small load on CPU resources while maintaining programmable flexibility. An
entire timing pattern handling all devices for a sequence is loaded at once. Each
timer had a dynamic range available from 250 ns to 16 s and a granularity as fine
as 250 ns. This allowed the computer to continue normal processing completely
uninterrupted until the end of a timing sequence. The timing capability of the
prototype, however, was limited to one timer and a (pro-load) latch with a
dynamic range of 65,535 us (216) with a granularity of only 1 us for the chip
alone. For each event in a given timing sequence the CPU in the AIM-65 had to
be interrupted to load the timer latch with the next delay. Instances where a delay
exceeded the dynamic range could only be accomodated by dividing the delay
into a multiple of a calculated base count and then a residual. For example, a
delay of 136 ms is achievable by two counts of 65535 and one count of 4930
given a 1 MHz counting rate. Multiple base counts and a residual implies that the
software must tend the overhead of the loop count and to insure the timer is only

enabled on the last residual count.

94

The net result was that the minimum time between any two pulses was
140 us using the AIM-65, and the minimum time for the PDP-11 was 250 ns. The
addition of hardware enable lines on the PDP-11 also gave a conditional pulse
capability. This was needed to allow timing patterns as in Figure 2.8. The
pulse(s) required for the peristaltic pump were only enabled in the last sequence.
Figures 3.8, 3.9, and 3.10 summarize the situations for setup, initializing, and

timer completion in pseudo-code.
3.2.2. Data Collectifl '

As already noted, data is collected by a Fast/Ultra-Fast ADC with 1024
points of storage and conversion rates ranging from 33 MHz to 2.5 KHz. The fast
ADC was designed and built by Martin Rabb (electronic design engineer,
Chemistry Dept., Michigan State University) in support of the delayed
luminescence studies in plant chloroplasts by Professor Babcock and W. J.
Buttner (Buttner, 1984a and 1984b). As mentioned, this was initially designed to
transfer into a Nicolet 1074 for immediate display, and while the display system
has been left relatively untouched, the ADC has been modified to support
another 8-bit data collection channel and programmed interrupt transfer into a
computer.

The original conversion channel is used for the differential signal
containing the kinetic information, and the second channel is used to digitize the
baseline information. Under interrupt request from the ADC 1024 points of 16

bits of data (8 bits for each digitizer) is transferred into the computer.

3.2.3. Transmission

The data transmission system is the system most susceptible to errors.

It is responsible for providing electrical bi-directional transfer pathways between

95

the computer and the instrument. One of the pathways carries the timing signals,
and another pathway carries the data and control signals for the ADC. Since the
FAS spectrometer was to be based on our general computer, the computer could
not be very well removed from the laboratory if the spectrometer was positioned
elsewhere (such as the Laser Facility under construction). To allow for such an
eventuality, special attention was paid to insuring reliable data transmission. The
potential distances contemplated, ranging from 5 to 100 meters, between the
computer system and the apparatus turned this into a significant problem.

After some consideration it was realized that fast data transfer rates
could be achieved only through parallel (8 or 16 bits simultaneously) rather than
serial (1 bit at a time). The 'fastest' rate for serial transmission used in
asynchronous terminal l/O without flow control is 38.4 Kbaud (Kbits/second, or
4.8 Kbytes/second). The hardware capabilities of the parallel port used in the
PDP-11 O-BUS allows rates up to 40 kwords/second, or 80 Kbytes/second.

A dedicated transmission system was designed employing two matched
transmitter/receiver pairs (see Figure 3.2) One pair is mounted next to the host
computer system. The second pair is mounted on the same equipment rack
containing the analog to digital converters. The current medium used to support

communication are two (one for each direction) 32 bit data cables.
3.3. SOFTWARE ARCHITECTURE IMPLEMENTATION

The design of the software architecture for the FAS has uniquely applied
concepts of distributed processing to the requirements for real-time instrument
control, particularly in a multiuser, multitasking operating system environment.
The general implication of distributed processing Is that tasks can be offloaded or
'distributed' to other processing elements. For the discussion here, a 'processing

element' can be defined as a computer system running a task. This definition

96

allows two tasks running simultaneously on the same processor to be considered

as two processing elements.

3.3.1. Distributed Processing: An Introduction

The concepts of distributed processing have become an important area
of computer research in both applied and theoretical areas. In some forms, It
offers the promise of significantly increasing computing performance at all levels
without corresponding order of magnitude increases in the performance of basic
computing hardware technology.

Implementation of these concepts have taken shape in many different
forms. The vectored array processors by Cray Research Laboratories and
Floating Point Systems Inc., are examples of highly specialized implementations
designed to solve particular classes of computing problems. The concept of
pipes in various implementations of the Unix operating system (Ritchie, 1984)
represents a general solution to stream or filter oriented data processing.

A simplistic example of a computationally distributed task can be found in
the FORTRAN DO loop:

DO 100 l-1,100
A(l) - 1.0/30)
100 CONTINUE

This loop inverts 100 elements in an array B and stores the result in
array A. A single processor loops through this 100 times. In the case of a
vectored system containing 100 hundred or more computing elements one
division would be performed by each computing element reducing the time

required by as much as a factor of 100. In practice this is not quite true as there

"I. '1'
‘l "
wfl

 

 

he
int

sin

OPE
dist
nlor
the (

BnoL

97

are loading and pipelining considerations of the data that turn this 100 fold
reduction into a lower bound, but this is irrelevant here.

The example given above is representative of one of the simplest
classes of parallel processing known as 'Single Instruction, Multiple Data' (SIMD).
The same operation, division, is being performed upon multiple instances of data,
100 elements in an array. Other classes of parallel processing machines exist
such as those belonging to the class of Multiple Instruction, Multiple Data
(MIMD). In MIMD processing, the problem as a whole is considered, broken
apart, and discrete processing aspects given to separate, independent
processing elements. MIMD type of processing, while normally executing upon a
multiple processor architecture, can be simulated on a single processor using a
multitasking operating system.

In real-time instrument control several sets of relatively independent
events such as data ready notification, user interrupt, timer count down
completion, etc., may be occurring simultaneously and each event may require
significant attention. Formalization of an algorithm to handle these independent
concurrent events can be quite complex and difficult to test. Distribution of the
handling of these events to separate processing elements allows development of
individual algorithms for each processing element, and these in turn will be
simpler as the individual tasks are now far simpler.

In single processor configurations, running multitasking, multiuser
operating systems, the Digital PDP-11 or VAX-11 being specific examples,
distributed processing can refer to distributing a processing function over two or
more active tasks. Within either and across both of the Digital processor families
the operating system communication resources or networks are sophisticated

enough to allow programs that are designed around distributed processing

Pa
dirt
rlllr
COn

dUra

"lain

98

algorithms on a single processor to be distributed over multiple processors with

little if any change to the software.
3.3.2. Distributed Processing: TheQesign & lmplemegtation

Processing is distributed over four software systems as listed below and
diagrammed in Figure 3.6:
User Command System
Apparatus Control System
Graphics Display System
Operating System Functions
These systems interact with each other directly along communication
pathways and indirectly along data pathways (see Figure 3.6). These pathways

are provided as a result of the operating system inter-task communications.
3.3.2. 1. User Command System

The User Command System (UCS) serves two functions: 1.)
Coordination of the activities of all the software components responsible for
providing the integrated environment of the FASCI, and 2.) Execution of the user
commands. The UCS has multiple display screens depending on the command
being executed. A menu portion displays current values for configuration
parameters and associated mnemonic commands. Configuration parameters are
directly related to the experiment execution and comprise variables such as
number of scans to execute, number of timing sequences per scan, type of data
conversion (raw, transmission, absorption, summation, averaging), base line
duration, conditional pulse generation, etc. These configuration parameters can
be directly stored and loaded from files so that common setups can be

maintained. Commands directly control program execution such as initiating an

inte
Con
BVBl

raw

$9ch
noific
only 0|
pr Oces
Olher t;
l0 help I

sell/ICQ 1

99

experiment, saving the data, plotting data, etc. Appendix A contains a more
detailed discussion of the menu(s), contents, and interaction with user.

The most critical part of the UCS is the Asynchronous Event Handler
Subsystem (AEHS) (part of the block for the UCS in Figure 3.6). Asynchronous
events are events which occur independent of normal processing. They are
handled first by the operating system, and dispatched to appropriate handlers.
Each of the systems notify the operating system of the events they want to
handle directly; therefore, when an asynchronous event occurs it is given to the
appropriate systems' handlers. The AEHS subsystem of the UCS handles
asynchronous events in a timely fashion that are of importance to the execution
of the UCS. Message arrival notification, successful task initiation, user interrupt,

etc. are handled through here.
3. 3.2. 2. Apparatus Control System

The Apparatus Control System (ACS) is the only system that directly
interacts with hardware. Commands are received from the UCS over the
communication pathway previously described. It too has an AEHS to handle
events. It also has a data pathway it shares with the UCS for the fast transfer of
raw data from the ADC for processing.

An Interrupt Control System (ICS) contains the code for the interrupt
service routines (ISRs). ISRs, software code entered immediately after hardware
notification, are used for fast response. lSRs are generally written to respond to
only one interrupt at a time from a device and perform the barest amount of
processing because all other forms of processing by the computer, including
other tasks and system functions, are locked out. The ICS provides a framework
to help prevent potential conflicts arising between multiple interrupts requiring

service from one or more ISRs.

100

Extended periods of time in an ISR degrade system response in a drastic
fashion (PDP-11 Processor Handbook, 1981). Opposing this idea of minimal
processing while executing at such a high priority, however, is the overhead
involved in saving and restoring contexts when lSRs are entered and exited on a
rapid basis due to an almost continuous series of hardware interrupts being
generated by the same device. The net result is then little useful work being
performed by the operating system. 1024 entries into an ISR would not only be
wasteful in terms of system resources, but would also result in an unacceptably
long period of time to effect the transfer of 1024 points. The ISR, which
processes the data from the ADC, when entered, remains until all data has been
transferred or an error condition has been established; however, to allow other
interrupt processing to occur, the interrupt priority of the ISR is dropped to 0

(lowest).
3. 3.2. 3. Graphics Display System

Real-time graphics display, the updating of which can be difficult to
schedule because of other interrupt and exception sources, is handled via the
Graphics Display System (GDS). It too has a communication pathway, and
shares a data pathway for access to processed data for plotting.

Upon initiation, the GDS is given the graphics device to use and the
number of data sets (maximum of 6) it can display on a subgrid on the device.
During the execution of an experiment once a set of data for a particular scan
and a particular sequence has been collected, processed, and stored in a
memory buffer, a command is issued from the UCS to the the GDS to plot the

given set. While the UCS continues on, the GDS accesses the data and plots it.

1 01
3. 3.2.4. Operating System Functions

This is not formally a system in the manner of the previously discussed
ones; however, some of the functions of the FASCI rely on services provided by
the operating system or other support programs. Two such support programs
are a vector to raster conversion job and a data archival job.

At the completion of an experimental run the user may decide to keep
the data, plot the data, or both. If the data is to be kept, a 'save' option is
selected and a local file is generated containing all of the experimental
parameters and the collected data. This file is then queued for archival onto
either magnetic tape or other mass storage on another computer system
accessible through the DECNET network. Once the file is archived successfully
a log is updated describing the file and final location, and the local file deleted.
This archival process was handled through the operating system independent of
the subsequent execution of FASCI, minimized the storage required for the
collected data, and safeguarded it against unintentional deletion or computer
problems.

If the 'plot' Option was selected then a local file was generated containing
all of the experimental parameters and the collected data that was to be plotted.
A batch processing job was queued for execution and it ran independently of
subsequent execution of FASCI. The batch job used the local file and several
programs, BINDMP, MULPLT (Atkinson, 1982), and RASTER (Gregg, 1985), to
process and print configurational information and the plots of the data on a dot
matrix continuous feed line printer. Use of a continuous feed printer instead of a
pen plotter meant that the generation of the plots could be performed without
intervention required by either the operator or the FASCI system.

The automation of these procesSes, archival storage, and plotting,

resulted in a significant increase in the number of experiments that be performed

102

in a given amount of time. This is important when considered within the context
of sample lifetime. If samples are reliable for only 6 hours on ice, at the end of
the period a new sample must be prepared if the experiments are not completed.
Automation of a set of tasks such as plotting simply makes the FASCI available

for experiments when it would be otherwise engaged in plotting.
3.3.2. 5. Data Communication

As shown in Figure 3.6, the systems have varying access to the
Experimental and Runtime Data Bases. These are tables of data created by the
UCS for the purpose of storing experimental data. The Runtime Data Base
contains digitized data that has been uploaded by the ACS from the transient
signal averager(s) but not converted and summed with other data gathered on
previous sequences or iterations. The Experimental Data Base contains digitized
data that has been converted.

At execution initiation the UCS is started first. It then initializes and
establishes the data bases, activates the other programs, and upon their
activation are given appropriate access rights to the data bases. For example,
the ACS only needs write access to the data base containing just the data which
has just been collected. The GDS needs access to the long term converted data
for plotting.

Memory resources are conserved by making the Experimental Data
Base variably sized. If a given experimental run consists of one sequence, then
the size of the region is set to only what is required to hold one set of data. If a

run consists of 32 sequences, then the region is resized to hold 32 sets of data.

lr_\

dr
ha

USE

facft
alley
allOW

aVall;

103
3.4. SUMMARY

The net result of this effort is a sophisticated instrument interface that
allows the experimenter flexible control over the design and execution of an
experimental scenario. Designing and developing this type of instrument
interface on this powerful of an operating system allowed features to be added,
such as the above mentioned archival and plotting programs, that could not be
contemplated on other less powerful computing platforms. The type of
architecture used for the code design would not have been as modular and
independent on a system supporting only single tasking. Some areas of the
instrumentation could originally not be considered completely successful,

however.
3.4.1. Performance

The performance of the computer as a general purpose multiuser system
degraded signficantly during experiments in which the inter-sequence delay time
dropped below five seconds, and degraded to the point of being unusable and
halting periodically when the GDS was activated. This was particularly
bothersome as use of a distributed design was to have lessened the resource
usage of the processor.

Replacing the PDP-11/23 with the PDP-11/73 (resulting in roughly a .
factor of four in terms of a performance increase) removed this problem. This
allowed inter-sequence delay times to drop to one second, the smallest value
allowed for in the software; however, the GDS still consumed a large share of the
available resources because of the extensive I/O and floating point calculations.

Experimentation with a DMA (Direct Memory Access) I/O controller for
movement of data related to the graphics display indicated that most of the

processor overhead involved in I/O could be eliminated; however, the expense

int

hat
the
5770/
This

air/al.;

104

was determined to be unnecessary as the NICOLET 1074 and its display
oscilloscope still functioned adequately for this purpose.

Conversion of the data from two 8 bit words to a real numbers with
enough dynamic range to support signal averaging also proved to be a
bottleneck because of the logarithmic operations (see Equations 2.4, 2.5, and
2.6). Over 100 ms was required to convert and sum 1024 data point. Linearizing
the logarithmic term by using a Taylor series expansion and using only the first

three terms of the expansion lowered the required time to less than 10 ms.
3.4.2. Maintenance

As I am the author of this software and hardware system I have an
intimate knowledge of It and occurrence of problems can be fixed quickly and in a
fashion that Is consistent with the design of the system. New features can be
added rapidly. Even so a failure in a 7402 integrated logic chip worth about
$0.23 required one month to track down and replace. During that time the entire
instrument was inoperative. I

Those who will use this spectrometer and its computer interface after I
have departed will not have this knowledge and experience and, regardless of
the documentation and detailed schematics that are prepared, will spend
inordinate amounts of time fixing problems that arise because of infamiliarity.
This situation will be tolerated only until a product becomes commercially

available at reasonable cost.
3.4.3. Evaluation

This implementation has resulted in a methodology aptly suited for the
design of computer software and hardware for real time instrument control. A

commercial scanning spectrometer has also been Interlaced using the same

105

distributed processing philosophy and many of the same underlying software
modules. Both instruments are capable of being operated simulataneously
without interference from each other, and other users were able to continue with
word-processing, graphics, or data analysis entirely unaffected.

Use of the PDP-11 and in-house developed hardware and software as
opposed to PC type machines and off the shelf software appears to be justifiable
only if a major focus of the laboratory is the development of computer aided
instrumentation or if the needed instrumentation or computer support simply are
not commercially available.

In this instance, at the time of conception (1983 - 1984), software was
not available and equivalently equipped PC computers were much more costly
on a per user basis and on a per task basis than the PDP-11. Today (1988 -
1989), the situation has changed, and data acquisition packages (software and
hardware) are available off the shelf. The lower data acquisition ranges of the
computer, < 1.0 MHz, are certainly within range of other products. However, the
adaptability of the computer to differing experimental senarios, high speed data
acquisition, integrated approach to software and data structures (especially If
more than one commercial package from different vendors has to be used), and

overall utilization of the computer has not yet been matched.

106

TABLE 3.1

Resogce comparison between AIM-65 and PDP-11 computer systems

This table presents a comparison of resources offered by the AIM-65 and PDP-

11 computer systems. See text for details.

 

 

 

 

Equipment Aim-65 PDP-1 1
Original Equipment
Processor: 6502 PDP-1 1/23
Address Space: 4Kb 64Kb (per
task)
2Mb (total)
Operating Sys. Monitor RSX-11M/MPLUS
Number of Users: 1 >12
Number of Tasks: 1 >128
Memory: 4 Kb 256Kb
Editors: Line Editor TECO (Line
Editor , KED,
K52, DT (key-
ad Editors)
Languages: BASIC Interpreter ASIC
Assembler Macro Assember
Eortran-77
Mass Storage: Cassette Tape 492Kb Flexible
Diskette
30Mb Winchester
Display 20 Char LED (3) Tektronix
Displa 4010 Emulators
Video erminal (5) VT100
Terminals
Printer: 20 Character 132 Column
Thermal Printer Dotmatrix
Printer
Plotter: HP-7470
Additional Equipment
Processor: u1pg/r7aded to PDP
1
Added Memory: 32Kb 1280Kb

Languages:

added FORTH

107

TABLE 3.2
Resource comparison between IBM-PC and PDP-11 computer systems

Comparison of resources offered by the IBM PC and PDP-11 computer
systems. Hardware disparities between the two systems for a single program are
not great, but the IBM PC with the MS-DOS operating system is unable to
support multitasking. The IBM PC has some advantages over the PDP-11 in the
amount of memory available for a single task. The original versions compilers for
the IBM limited memory to 64 or 128 Kb, but more recently, the compilers have
come to support memory models that easily allow for the entire 640 Kb to be
used by a single program. Such an option is not available to the PDP-11.

The processor in the IBM PCs, the 8088, can physically address 1 Mb,
but the physical design of the IBM PC limited it to 640 Kb. Other hardware
vendors have adopted schemes allowing additional memory, 'expanded' or
'extended' memory to be inserted. This additional memory cannot be used to

execute code only data.

108

 

 

 

 

 

Equipment Aim-65 PDP-1 1
Original Equipment
Processor: 8088 PDP-1 1/23
Address Space: 64Kb (original) 1285b (per
tas
128 to 640Kb 2Mb (total)
(vendor
dependent)
Operating Sys. MS-DOS RSX-11M/MPLUS
Number of Users: 1 >12
Number of Tasks: 1 >128
Memory: 640 Kb 256Kb
Editors: EDLIN (Line TECO (Line
Editor)
(editor) KED, K52, EDT
(keypad
editors)
Languages: BASIC Interpreterl
Macro Assember
Igortran-77
C Mass Storage: 360Kb Flexible 492Kb Flexible
Diskette Diskette
10Mb Winchester 30Mb Winchester
Display: 80 x 25 Video (3) Tektronix
Display:1 , 4010 Emulators
5) VT100
errninals
Printer: 132 Column Dot Matrix Printer
Plotter: HP-7470 2 pen plotter
Additional
Processor: upgraded to
PDP-11/73
Memory: 'Expanded' or increased to
'extended' 1280Kb
memory has
become available

 

 

 

1oolor Graphics Adapter (CGA), Enhanced Graphics Adapter (EGA), and Hercules
(non-IBM, 3rd party standard) adapters. Emulation of Tektronix 401 x possible.

109

Hounesd’
Qgical Schematic of Timing Logic

The command control interface unit routes the commands and data to
the appropriate timing unit, TUnn, where 'nn' - {1 ,...,16}. Each timing unit can
individually programmed in terms of count, frequency of counting, and execution
when its counting is finished. As shown here all but the last timing units have its

output connected to the next timing unit. Further details can be found in the text.

Z
/Unterruot
Vector

 

\

 
  

   
 

Commands
and Data

     
 

THner(s)

Enable

110

   
    
   

'TUOI

TUO

 
  

Interrupt
Vector
Control
InterFace
Umt

   
    
  
 

TUO

   
   
   
 
   
 
   
   
   
  
   

   

TUO

TUO

TUO

TUO

TUO

TUO

TUIO

TUll

    
 
 

Command
Control

Interface
Umt

  
  

   
   
 
   
    

TUIE

TU13

TU14

TU15

To

IZMED
DEV7CES

111

FIGURE 3.2
Communications Scheme between Comp uter and Instrtument

Both ends, the computer and the instrument, each have a transmitter and
a receiver that manage the parallel data flows. The data in each direction travels
through a 32 bit communications pathway. Each and also has a front panel
display. The front panel display shows the data activity on each line and while
intended for initial debugging, proved to be continually useful for status diagnosis.

The blocks 'Location 1' and 'Location 2' indicate that each group within
are located separate from one another. The dashed connecting lines labelled '32
bit communications pathway' simply indicate an indefinite length separating the
two locations. In actuality, designing the communications between the computer
and the instrument in this fashion remove any dependencies on the actual

medium used to support the communication pathway.

1‘12

 

ES;
.25 :5:

 

 

Emma
_mccm Cacao

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

  

 

 

 

 

 

 

_ _
_ _
. .
_ _
_ c _
_ _
n m: E . n
_ cewp§m=62_ _ . s "nuns _ ae>aoea sssshssssss _
_ into as assessezsm _ _
_ _ _ _
_ _ _ _
_ _ _ _
_ _ _ _
_ _ _ _
_ _ _ _
_ _ _ _
_ .llo . . asosazou _
“ 238“ § § .5283. “ _§s§.§:§ _ aécmcgsssss E: u
_ _ _ _
_ _ _ _
_ _ _ _
_ _ _ _
A. lllllllllllllllll .a r. llllllllllllllllllll .a

m So So N \ SS Es N

113

FIGURE 3.3

Algorithms for Setup of Timing Sequences

These algorithms reflect the pocedures of preparing the AIM-65 and the
PDP-11 for performing the hardware and software for timing pulse generation.
The timer on the AIM-65, unlike the ones on the PDP-11 CODAR timing board,
does not offer a scalable base counting frequency and requires additional
counting periods, manipulated by software in the interrupt routine, to extend to
some of the timing period intervals needed. The PDP-11, on the other hand, only
requires determination of the appropriate base frequency and the number of

counts in that base. See text for further discussion.
a) Algorithm to setup, in pseudo code, the single timer for the AIM-65.

DO for all pulses (16)
IF delay count exceed 65535 ticks at 1 MHz
Calculate and store # of multiples of counts at 65000 us.
ELgEICU'ate and store the leftover residual delay count.
Store 0 multiples of counts
Store delay count.
ENDIF
Store device enable mask
ENDDO

b) Algorithm to setup, in pseudo code, the Codar O-TIMER for the
PDP-11.

DO for all pulses (16)
Calculate and store basic frequency
Calculate and store counts for basic frequency
Store device enable mask
Arm counter
ENDDO

114

FIGURE 3.4
flgorithms for Initialization of Timing Seguences

These algorithms reflect the procedures necessary for initiating a Timing

Sequence.

a) Algorithm to initialize, in pseudo code, the single timer for the AIM-
65.

Load Timer with count Decrement multiple count number
IF multiple count number <> 0
Load Timer Latch with count
ELSE
Load Timer Latch with residual count
ENDIF
Trigger Timer to begin

b) Algorithm to initialize, in pseudo code, the Codar Q-TIMER for the
PDP-11.

Trigger first timer to begin

115

FIGURE 3.5
Algorithms to Process Timer Completions (counter finishe_d_)_.

These algorithms reflect event processing that must take place when a
timer has completed. Processing occurring in hardware are preceded by
'[...hardware execution... ] and '->.' In the case of the AIM-65 the latch for the
timer is loaded by hardware into the counter and the timer is restarted. The
software prepares the latch for the following time. If the current time being
counted down is to result in a pulse, then the timer is enabled for generating a
pulse. Other than the automatic reload all of the operations occur in software
and incur a significant overhead in the duty cycle.

Contrast this with the processing required for the PDP-11 and the
CODAR timer. Handling of timer completions is performed completely in
hardware and only upon the last pulse to be sent out is the processor notified of

the event.

116

a) Algorithm to process timer completion, in pseudo code, for the AIM-

65.

[,..hardware execution... ]
--> Timer counted down
-> Interrupt enerated to CPU
--> Count in atch loaded into timer
--> Timer restarted

Decrement mulitple count number
IF multiple count number > 0
Load residual count into timer latch
ELSE
Enable Device Pulse
IF NOT last device
Load Timer latch with next device count
Reset multiple count number to new device's
ELSE
standby
ENDIF
ENDIF

b) Algorithm to process timer completition, in pseudo code, for the
Codar O-Timer for the PDP-11.

[,..hardware execution... ]
--> Any Timer but last is counted down
-> Next Timer triggered to begin counting
-> Pulse put out

[...hardware execution... ]
-> Last Timer counted down

standby...

117

FIGURE 3.6

Lpgical Schematic of Software Control Systems

The User Control System (UCS), Apparatus Control System (ACS), and
the Graphics Display System communicate directly via message pathways (dark
lines) and indirectly via data access to data bases (broad, light lines).

Access into the data bases for each system is standardized. Each data
base was designed so that the first several units of storage are addresses or
pointers into the data bases for the information. The location of these offsets are
standardized and stored in a file which is used in the compilation of all systems
accessing the data bases. In this way, modifications to either the data bases or

the programs accessing the data bases is simplified.

118

33:; $235 3;: «55235

3 3

Eh a ESE has

 
 

 

 

 

 

“Russo cassse\ .ssesag Assess
." Eaasw MEEEENLECREww “seam axgsassaasww .s
caomxm Eaamg mozqcco cagmxw_ocwcom mboachQc

 

 

 

 

 

 

 

7 as as s5 #

 

        

     

.1 lllllllllllllll J
_ insist: _

3% 5%
535.33

 

$32sz ......... L .
=38§ccg 3g—

 

$83
E: §§E$

 

 

 

 

 

 

 

51m .828 am:

i

saw:

 

 

 

:25: cssscé Same...»

 

119

REFERENCES

Atkinson, T. V. (1982). Atkinson is author of a program called 'MULPLT'
designed to perform vector plotting'of multiple data sets. This program
was developed at Michigan State niversity and has been distributed
through DECUS, Digital Equipment Corporation User's Society.

Buttner, W. J. 8a7nd Babcock, G. T. (1984a). Biochimica et Biophysica Acta, 767,
2 2-2 .

Buttner, W. J. (1984b). Ph.D. Thesis, Michigan State University.

Gregg, H. (1985). Gregg is author of a program called 'RASTER' designed to
perform generallzed vector to raster conversion. This program was
develo d at Michigan State University and has been distributed through
DECU , Digital Equipment Corporation User's Society.

Ritchie, D. M., 1984. AT&T Bell Laboratories Technical Journal, 63, 8. Reprinted
from _Ilectutre Notes onTom uter Science, 79, Language Design and
Programming Methodology, Springer-Variay, 1980.

I_3_DP-11 Processor Handbook, 1981, Digital Equipment Corporation.

Vax Hardware Handbook, 1982, Digital Equipment Corporation, pp 331-366.

CHAPTER 4

THE PHOTOSYSTEM II ACCEPTOR SIDE

4.1. INTRODUCTION

The chemistry and absorption band locations of the acceptor side
quinones that undergo one and two-electron redox reactions were briefly
Introduced in Chapter One. The uncoupling of their contributions to the transient
absorption changes observed in Photosystem Il preparations has proven to be
key for the interpretation of absorption changes attributed to the oxygen evolving
complex as alluded to earlier and detailed in the next chapter. This uncoupling
has not proven to be straightforward for reasons related both to the nature of the
systems studied and the conditions that researchers use in their studies. This
latter fact has only recently become appreciated and will be developed later in
this chapter.

At the time we set out to examine optical changes of the OEC (1984) only one
group had published a detailed kinetic difference spectrum arising from the
reducing side of PS-ll as well as a spectrum for the YZ+IYZ couple (Dekker et al.
(1984a). These results were subsequently used for removing reducing side and
YZ+IYZ contributions from the difference absorption spectra of the S-state
transitions of the OEC (Dekker et al., 1984b).

As mentioned In Chapter One and discussed in further detail in the following
chapter, the optical results reported by Dekker et al. (1984b) relating to the 8-

state transitions have been controversial because of the methods used to

120

121

deconvolvethe S-state spectra and the interpretation of the data. In their
procedure it was necessary to remove the contributions arising from Oa'lQa and
Yz‘l’, and we believed that the experimental protocol used was suspect in regards
to the exogenously added electron acceptor and dOnor systems that supported
electron transfer through the tris-washed PS-II systems. In addition, the
conditions used for determination of the Qa' lQb spectrum differed significantly
from the conditions used to determine the S-state spectra. This necessitated
repetition of some aspects of their work under conditions we felt were better
controlled. This would accomplish two things: 1) Our results would serve then as
an independent confirmation of their work. 2) Subsequent experiments we would
perform that relied upon detailed knowledge of the spectral contributions arising
from the reducing side, such as S-state transitions, would not be incorrectly
adjusted if legitimate differences existed between the spectrum we would report
for Oa'lQa versus that reported by Dekker et al. (1984a).

A detailed outline of the reactions undergone by PS" in response to
single flashes of light is shown in Figure 4.1, which reveals a synchronous
relationship between the acceptor and the donor sides. The cumulative effects of
'hits' and 'misses' of the OEC do not remain correlated with the reactions
undergone by the acceptor side, and thus the two sides dephase with respect to

one another and complicate the observed absorption spectral changes.

4.1.1. Separation of Quinone Contributions

Any optical spectroscopic investigation of the acceptor side has to
account for absorption changes by both Qa as well as by Ob. In particular, the
equilibrium condition between Oa'Qb <---> OaOb‘ must be accounted for.
Vermeglio and Clayton (1977) showed that in the near-UV and visible the
absorption spectra of OaQb' and Oa'Ob in Rhodopseudomonas sphaeroides

122

were nearly indistinguishable; however, in the red and in the near-IR, these two
states could be distinguished one from another on the basis of induced
absorption band-shifts in the bacteriochlorophyll and bacteriopheophytin.
Kleinfeld et al. (1985 and 1984) have since demonstrated that the difference of
20 % in the optical extinction coefficients between the states Oa'lOa and Oa/Qb'
at 450 nm is large enough to measure reliably and to differentiate between

different states of the acceptor side.

4.1.2. Simplification of Reducing Side Chemistgy

The chemistry of the acceptor side is the same regardless of the Integrity
of the oxygen evolving complex, and follows that given in Reactions 1.3, 1.4, and
1.5 assuming that the acceptor used is either plastoquinone or a quinone
analogue. Ideally, all modifications to the acceptor side should be reversible.
Such a condition demonstrates that the effects measured have not been
achieved through irreversible alteration of the composition and states of the Ga

and Ob sites.
4.1.2.1. Conventional Inhibition

Figure 4.2a shows known sites of selective inhibition that will affect the
yield of Qa'. Under conditions in which the OEC has been deactivated, such as
by tris-washing, which removes the manganese responsible for oxygen evolution,
the addition of ferricyanide, Fe(CN)63', as an acceptor and a herbicide, such as
3-(3,4 DIchlorophenyl)-1,1 dimethylurea, DCMU, that competes inhibitively with
endogenous quinones for access to the Ob binding site, the chemistry will
simplify to (Figure 4.2b has more detail):

hv
Qa --> Qa' (4.1)

fer
exl
so
80
0a
kin
unt
Fe
Fe:
ran

diff

the
EqL
819(

Of tj

123

Oa" + Fe3+ ---> Qa + Fe2+, 11/2 - 33 (4.2)

There are difficulties with this simplification. The oxidation of Oa‘ by
ferricyanide can be heterogenous under conditions thought to promote
exclusively the slow oxidation of Qa’. As shown in Equation 4.2, the t1/2 is 3
seconds, but the reduction of Yz+ by ferrocyanide, Fe(CN)64‘, has a t1,2 of 50 to
80 ms (Dekker et al., 1984a). This allows the backreaction between P680 and
Oa', which has a t1,2 of 120 to 200 us (Haveman and Mathis, 1976), to become
kinetically more competitive. The idea that Oa‘ could also undergo fast oxidation
under similar reaction conditions as used above was first suggested by
Petrouleas and Diner in 1986 (Petrouleas and Diner, 1986). The t1/2 for the
F93+ to Fe2+ reduction attributed to Oa' appears to be pH dependent and
ranges from 7 to 25 us (Diner and Petrouleas, 1987a and 1987b). Another
difficulty may be from an interference in the electron flow between the OEC and
YZ+.

This difficulty of heterogeneous behaviour can be generalized to include
the entire response of the Photosystem II unit. The simple model suggested by
Equations 4.1 and 4.2 assume complete reactions for at least this section of the
electron transport chain. Melis and Anderson (1983) have questioned the extent
of the Ga response in relation to the extent of charge separation and Dennenberg
et al. (1986) has provided evidence that Photosystem II reaction centers capable
of water oxidation consist of two separate populations that are characterized by
their differing response to added quinones.

Another apparent difficulty may be an interference in the electron flow
between the OEC and th. Hoganson (1988) noted in EPR studies on the

Iineshape of Y2... that ferricyanide in concentrations of greater than 1 mM

appeal

membl

changl
This 0:

states.

POpUlE

4.1.2..

leactil
1976)
hlldro
remm
3) mir
diges
the bi

preps
°deiz
Unwe
Soth

prep;

124

appeared to affect the half lives of both th and of the S-states in Photosystem II
membrane preparations.

As a final note, experiments examining multiple, sequential S-state
changes can not use ferricyanide because of its slow oxidation of Oa' to Oa.
This oxidation time is long in comparison to the deactivation of the higher S-

states. This in turn contributes to the decrease in the homogeneity of the S-state

populations.

4.1.2.2. Trypsin Modification of the Ob Binding Site

A different approach to the control of the acceptor side electron transport
reaction has been adopted by Renger (Renger, 1976a, 1976b, and Renger et al.,
1976). He found that brief incubation periods of chloroplasts with trypsin, a
hydrolytic enzyme that attacks lysine and arginine residues, had three effects: 1)
removed sensitivity to DCMU, 2) lowered the affinity for acceptor quinones, and
3) minimally altered the affinity for ferricyanide. These effects were attributed to
digestion of the intrinsic 32 kDa polypeptide (Renger et al., 1986) thought to hold
the binding sites for Ob.

Unfortunately, this technique, when extended to Photosystem II
preparations in which the membrane orientation has disappeared and both the
oxidizing as well as the reducing sides are exposed to the bulk media, has
unwelcome side-effects. Trypsin appears to attack, at a minimum, the water
soluble polypeptides of 17, 23, and 33 kDa (Volker et al., 1986). Photosystem II
preparations exhibiting the multiline, an EPR signal attributed to the 82 state of
the OEC (see Chapter 1), were subjected to tryptic digestion at pH 7.6 and 6.0.
The sample at pH 7.6 lost the multiline completely, but the sample at pH 6.0
manifested no change; however, the water soluble polypeptides all appeared to
be digested to some extent. In particular, the 33 kDa polypeptide became a 15

kDah
its dig

comp

4141

whhc
DCNl
Qafl:
plastt
as ye
react—

there

lechr
Scanl
anotr
Pefio
bagel

reflec

e"Cite

125

kDa fragment. Trypsin appears to be suitable for use in chloroplasts; however,
its digestion of polypeptides apparently associated with the oxygen evolving

complex make it unsuitable for work with Photosystem ll membranes.
4. 1.2. 3. Cholate Treatment

Bolby and Yocum (1987) investigated the treatment of the reducing side
with cholate as it held out fairly strong promise for removal and reconstitution of
DCMU sensitivity. They had demonstrated high rates of reduction of F931" from
Oa', and these rates were comparable to those achieved by endogenous
plastoquinone or exogenously added analog quinones. Unfortunately, they have
as yet been unable to prove the reversibility of the cholate treatment and
reconstitute DCMU sensitivity. Because of the uncertainty of the modifications to

the reducing side, we chose not to take advantage of this treatment.
4.1.3.§pedroscopic Characterization of the Reducing Side

The experiments on the reducing side have reflected increasing
technological sophistication. The early experiments were recorded by using
scanning spectrometers and conditions designed to pump one state over
another. Later experiments relied upon flash methods to generate for brief
periods of time the intermediate states, that could then be subtracted from
baseline samples. The spectrometer used by Joliot, the 'Joliot spectrometer,’ still
reflects the state of the art in terms of spectrometer design. This was a dual
beam, dual recording spectrometer designed explicitly for differential sample

excitation (Joliot et al., 1980, and Joliot and Joliot, 1981).

4.1.3.1

reaso
and \l
deox
term
had

nm.

SDE

pla

126

4.1.3. 1. Steady State Absorption Spectra

The so-called steady state spectra due to the Oa'lOa couple has been
reasonably well known since reported by Van Gorkom (1974) (also, see Stiehl
and Witt, 1969, and Pulles et al., 1976). The sample preparation was from
deoxycholate enriched Photosystem II membranes ('DOC 2 particles' to use the
terminology adopted by Van Gorkom) from spinach thylakoids. These samples
had lost the ability to evolve oxygen. Points were collected at a resolution of 5
nm.

Van Gorkom identified the species responsible for the difference
spectrum as plastoquinone by comparing it with the difference spectrum of the
plastosemiquinone anion obtained by Bensasson and Land ( 1973). The
spectrum collected by Van Gorkom exhibited a 15 nm red shift relative to that
obtained by Bensasson and Land; however, Van Gorkom noted that this was
appareme similar to a shift seen for ubiquinone in bacterial reaction centers

(Slooteen, 1972).
4.1.3.2. Flash Induced Difference Spectra

Flash induced absorption differences have been observed for years in
the UV and have generally been associated with acceptor side chemistry. Stiehl
and Witt (1968) reported the first definitive correlation; however, the first flash
induced spectrum of the Oa’lOa couple was not reported until 1984. Dekker et
al. (1984a) recorded the spectrum at a resolution of 2.5 nm in Photosystem ll
enriched membranes that were better characterized than those used by Van
Gorkom in the steady state experiments. This spectrum was similar to that
obtained earlier via steady state techniques in both Iineshape and relative peak

amplitudes, although, as noted, the steady state and the flash-induced spectrum

are

Ben

tris-
witl
bill
to
rel
m
Tl
tr

127

are red-shifted by 15 nm relative to the in vitro semiquinone reported by
Bensasson and Land.

Dekker et al. (1984a) found, in samples with the OEC deleted through
tris-treatment, that the absorption band for Oa'/Oa‘ is asymmetric around 325 nm
with a molar extinction of 13,000 _M’1cm'1 and a half-width of 20 to 30 nm. In the
blue, between 410 to 450 nm, the sharp, derivative shaped bands were attributed
to blue shifts in the absorption spectrum arising from the pheophytin due to Oa
reduction. They related the extent of the flash induced Oa' absorption to the
number of centers undergoing photochemistry by use of the C-550 bandshift.
This is a bandshift of the pheophytin intermediate in the Photosystem Il electron
transport chain (see Figure 1.3) caused by the reduced primary acceptor, Oa'
(Van Gorkom, 1974).

More recently, Velthuys (1988), has likewise performed a spectroscopic
characterization of the Oa’ state and is largely in agreement the results of Dekker
et al. (1984a) in the blue and in the red. He disfavours the blue shift
interpretation of the pheophytin, and suggests that a red shift of the pheophytin
and a blue shift of the chlorophyll is operative. Unfortunately, his experiments
were not carried out deep enough to the blue to also provide independent
estimations of the extinction and bandshape of the difference spectrum arising

from the Oa'lOa couple.

 

4.2. N

A
_L

4.2. 1.

128

4.2. MATERIALS AND METHODS

4.2.1. golpgical Prepgrations ant; Handling

4.2. 1. 1. Chloroplast preparation

Broken thylakoid membranes, referred to as Class II chloroplasts
(Robinson et al., 1980), were isolated as described by Robinson et al. (1981) as
summarized below. Spinach obtained commercially was washed and
depetiolated. Approximately 500 to 1000 g were homogenated for ten seconds in
a blender in a buffer solution of 0.4 M NaCl, 2 mM Mng, 1.0 mg/ml bovine
serum albumin (BSA), 1 mM ethylenediamine tetraacetic acid (EDTA), and 20
mM HEPES (pH 7.6). The homogenate was filtered through eight layers of
cheese-cloth. The filtrate was then accelerated in a centrifuge until 3500 x g was
reached and braked to a stop. This first centrifugation removed any large
particulate matter that had passed through the cheesecloth. The pellet was
discarded, and the supernatant centrifuged at 7500 x g for 10 minutes. The
supernatant from this was discarded, and the pellet resuspended in 20 mM
HEPES (pH 7.6), 150 mM NaCl, and 4 mM MgCl2 (first resuspension buffer).
This suspension was centrifuged at 7500 x g for 10 minutes and the supernatant
discarded. The pellet was resuspended and stored at -40°C in a solution of 400
mM sucrose, 50 mM HEPES (pH 7.6), and 10 mM NaCl (SHN). The oxygen
evolving rates of these chloroplasts were typically greater than 300 urnoles
Ozlmg Chi/hr.

Irreversible inhibition of oxygen evolution was accomplished by washing
the chloroplasts in a solution of Tris(hydroxymethyl)aminomethane (Tris) before
final suspension in SHN. The chloroplast pellet was suspended in 50 to 100 ml
of a solution of 0.8 M Tris-HCI and 1 mM EDTA at pH of 8.0. This suspension

was

dunn
nfinu
susp
Phou

ioHov

4.2. 1

referr
(1981
Frost
matel

descr

129

was incubated on ice at 4°C for 20 minutes and exposed to normal room lighting
during the incubation. The suspension was then re-pelleted by a 4000 x g, 10
minute centrifugation. If only chloroplasts were to be prepared, the pellet was
suspended in SHN and either stored in a -40°C freezer or immediately used. If
Photosystem II membranes were to be prepared, the procedures outlined in the

following section(s) were followed.
4.2.1.2. Photosystem II membrane preparation

Isolation of the Photosystem II complex in a preparation commonly
referred to as 'Photosystem II membranes' was carried out as in Berthold et al.
(1981), modified as in Ghanotakis et al. (1984), and described briefly below.
Freshly prepared or frozen stock Class II chloroplasts were used for the starting
material. Inhibition of oxygen evolution has already occurred as previously
described.

Chloroplasts were dark adapted and incubated on ice for five to ten
minutes. An aliquot of a solution containing 50 mM Mes, 15 mM Nacl, 5 mM
MgCIz, and 1 mM ascorbate at a pH 016.0 (Triton Suspension Buffer) was added
to the chloroplast suspension such that subsequent dropwise addition of a stock
25% Triton solution (25% by volume of Triton X-100, 50mM Hepes (pH - 7.5),
and 15mM MgClz) yielded a final chlorophyll concentration of 2 mg/ml and 7%
Triton. Once the Triton aliquot had been added, the samples were incubated on
ice in the dark for 30 minutes. Upon completion of the incubation period the
solution was centrifuged at 2000 x g for 1 minute. The supernatant was
removed, the pellet discarded, the supernatant suspended in the Triton
suspension buffer, and centrifuged at 19,000 x g for 30 minutes. The resulting
supernatant was discarded, the pellet resuspended in the Triton suspension

buffer, and centrifuged at 19,000 x g for 30 minutes. The resulting supernatant

130

was discarded and the pellet suspended in a buffer consisting of 0.4 M Sucrose,
50 mM Mes (pH - 6.0), and 10 mM NaCl (SMN buffer). Final concentration was
roughly 2 to 4 mg/ml Chl.

This final solution contained less than 1%contamination from
Photosystem I as determined by EPR (Berthold et al., 1981). If 02 evolution was
not inhibited then the average preparation was capable of supporting an 02
evolving rate of 600 to 700 umoles 02 /mg Chi/hr. If the preparations have been
washed with Tris then they exhibited no measurable 02 evolution using a Clark
oxygen electrode.

4.2.1.3. Sample Preparation for Flash Absorption Spectroscopy

Tris-washed PS-ll membranes obtained from spinach were used in the
FAS at concentrations ranging from 75 - 125 uM Chl (0.3 - 0.5 uM Reaction
Centers, assuming 250 Chi/RC). 10 uM DCMU was used to block further
electron transfer beyond Oa. Problems possibly arising from continual sample
excitation were eliminated by recirculating the sample between every laser shot.
Samples in the cuvette and tubing represented no more than 10% of the total
volume. It was ascertained via steady state oxygen measurements (data not
shown), that samples were not damaged from mechanical transport between
excitation cuvette and reservoir.

The basic reaction media was prepared by having 100 to 200 ml of buffer
consisting of 0.4 M sucrose and 0.05 M Mes at pH 6.0 stirring on ice in a round
bottom flask. An aliquot of PS-Il membranes at a stock chlorophyll concentration
between 2 to 4 mg/ml (2.3 to 4.5 mM) were added slowly in the dark for a final
concentration between 75 - 125 uM Chi and allowed to stir for five minutes. A
second solution containing only buffer (0.4 M Sucrose and 50 mM Mes at a pH of

6.0) at 4°C flowed continuously through the sample chamber and tubing. When

131

five minutes had passed the sample chamber and tubing were drained of the
buffer and the photosynthetic sample in the reaction media was allowed to flow
through.

If the reducing side reactions were to be blocked then 1 mM DCMU in
EtOH was added to the buffer before the PS-Il membrane aliquot for a final
concentration of 10 LM DCMU. After a five minute dark adaptation an equimolar
mixture of ferri- and ferrocyanide was added to achieve a final concentration of
0.25 to 1 mM Fe3+ and F9211 If the reducing side reactions were not to be
blocked 2,5 dichlorobenzene (DCBO) was added to a final concentration of 400
uM.

depe
taken
Width

opfiCa

132

The ferrous iron on the reducing side, 0400, (Petrouleas and Diner,
1986), is dependent upon the order of component addition, especially as relating
to the use of ferricyanide and DCMU (lkegami and Katoh, 1973 and Bowes et al.,
1979). The participation of 0400 is thought to lessen the extent of the absorption
change attributable to Oa' at 325 nm. It was observed that addition of DCMU
before ferricyanide slowed down the rate of oxidation of O400 and, therefore its
participation in subsequent redox chemistry, by as much as a factor of 25
(Wraight, 1985). Largely based upon the work of Bowes and co-workers (1979),
a consistent order of addition of components was followed as discussed above
and diagrammed below to avoid any problems arising from inconsistent order of
addition. A consideration of the role that 0400 might play in these studies is

found in the 'Discussion' section.

Basic Reaction Media

< ....... If inhibitin Oa’Ob --> OaOb';
add DC U.

< ........ Add Photosynthetic Material.

< ....... lf inhibitin Oa’Ob -> OaOb';

add ferri/ erro-cyanide mixture.
If NOT inhibiting Oa'Ob ->OaOb';
add 2.5 DCBO.

 

V
Prepared Sample

4.2.2. InstrMmental Conditions

The instrument time constant and scan rates used were variable and
depended on the type of experiment being performed. Kinetic spectra were
taken every 2.5 nm (slit width of 0.5 mm, optical bandpass of 3 nm) or 5 nm (slit
width of 1.0 mm, optical bandpass of 6 nm) in the range 280 to 450 nm. The

optical feedback variance limits for the xenon probe beam were set to 0.95 (i.e., If

til
G.
iUl
dis
not

103

are :

133

the average power output of the probe beam was 100 watts (W), the optical
feedback unit kept the power variation of the probe beam on the millisecond time
scale between 95 and 105 W.) Data points collected for the purpose of
constructing difference absorption spectra were taken randomly along the
wavelength range to minimize systematic errors. For each spectra between 4 to

16 flashes were averaged per point.

4.2.3. Data Handling

Absorption difference spectra were produced from a set of kinetic decay
traces collected over the selected wavelength range in a fashion analogous to a
boxcar integrator (for a brief description, see Malmstadt et al., 1981). Each trace,
one for each sampled wavelength, was stored along with its baseline data in a
computer readable format. The baseline data were used to correct the decay
trace data in the construction of the absorption difference spectra. All points
falling within a selected time window were averaged and one point resulted for
that window. Multiple windows were often specified for each trace. These points
were then plotted versus wavelength. This process is depicted in Figure 4.10.
Lines through the points were derived either by using an interpolative cubic spline
fit to the data if the point density was high enough, or by applying a Savitsky-
Golay (Savitsky and Golay, 1964) 5-point triangularly weighted, smoothing
function to the data. Coefficients in the latter case were chosen for minimal
distortion. In all cases, manipulation of data involved the actual data points, and
not the lines that were generated from the data points. The lines were generated
to serve only as an aid to the eye.

Conditions or modifications that are different from those described here

are specified in the figure legends.

I!-X

Of ‘
thu
Chic
absl
Cont
SUch

cOltlr

I10”. 5
”least

0(’"Cer

134

4.3. RESULTS

These experiments were conducted .to provide several pieces of
information: 1.) Proof of saturation of the Oa/Oa transition, 2.) Verification that we
were observing the Oa'IOa transition by observing if the lifetime of the Oa"
species decreases through the use of a quinone acceptor rather than
ferricyanide, 3.) Verification of the results obtained by Dekker and co-workers
(1984a) regarding the quantitative shape and extent of the Oa‘lOa spectrum, and
4.) Confirmation of the EPR results obtained by Ghanotakis and Babcock (1983)
that showed that the addition of hydroxylamine in low concentrations caused a
reversible interruption of electron flow between YZ and P680+- In the original
experiments, the decay of P680 under hydroxylamine treated conditions was
followed. In the experiments reported here the decay of Oa' was followed.

4.3.1. Saturation of the Oat/Oa Transition

Incomplete saturation of photochemistry leads to a situation in which all
of the PS-II centers fail to participate in charge separation and Oa reduction;
thus, the detected extent of absorption change becomes a function of both the
chlorophyll concentration and the excitation energy. This will result In an
absorption difference spectrum of Oa'/Oa that cannot be linearly related to the
concentration of the reaction centers. A difference spectrum produced under
such conditions cannot be used for the quantitative removal of Oa'lOa
contributions from other spectra.

A saturation profile was performed for chlorophyll concentrations ranging
from 50 to 200 MM [Chl] against laser intensity to determine if photochemistry, as
measured by Oa' reduction, was complete. The highest chlorophyll

concentration used, 200 MM [Chi], is shown in Figure 4.3. Saturation was

cc
tht
USI
rep
bar
Dhe
Flas

aCCe

Spec}

135

reached at 30% of the maximum laser power used for this set of experiments (85

mJ per pulse, 400 ns pulse duration, 213 kilowatts, kW).
4.3.2. The O_a£/Oa Difference Spectru_n:l_

Figure 4.4 is the spectrum we derived from a series of kinetic traces
collected point by point in wavelength range of 280 to 550 nm. In the UV and
near-UV the points were collected every 1 nm and in the visible the points were
collected every 2.5 nm. Slit widths were adjusted so that in the UV the optical
bandpass was 6 nm and in the visible the optical bandpass was 3 nm. Figure 4.5
overlays this spectrum on top of that obtained by Dekker et al. (1984a). From the
comparison in this figure it is clear that there is agreement on most of the major
spectral characteristics. However, the 325 nm absorption band we report is
broader and its peak is less well defined. Additionally, in our spectrum the
derivatively shaped bands between 390 and 450 nm are blue shifted by ~5 nm
relative to those reported by Dekker et al. (1984a).

Although we were able to reproduce the detailed Iineshape reported by
Dekker et al. (1984a), we were unable to reproduce initially the extinction
coefficients, 13 mM'1cm'1 at 325 nm, that Dekker et al. (1984a) had reported for
the Oa’lOa couple. The first effort we made to resolve this discrepancy was to
use a spectroscopically related feature known as the C-550 bandshift. First
reported by Van Gorkom (1974) and used by Dekker et al. (1984a), this is a
bandshift centered around 545 nm of the absorption spectrum of an intermediate
pheophytin induced by the reduced state of 0a“. This is a region in which our
Flash Absorption Spectrometer, the FAS, had already been shown to operate
acceptably, Lillie (1984).

Under conditions identical to those used for determining the difference

spectrum of the Oa'lOa couple, the C-550 spectrum was collected in the same

th
sa
ms
10
time
furtl
Figu
t0 ex

018%

136

fashion as described earlier (see Materials and Methods) and compared against
that reported by Van Gorkom (1974). As with the Oa'lOa spectrum the Iineshape
and the extinction coefficient, ~2.6 mM'1cm'1, we measured was similar to that
reported in the literature (Van Gorkom, 1974). There have been two other
independent measurements of the C-550 bandshape and magnitude. McCauley
and Melis (1986) reported the magnitude to be twice that reported by Van
Gorkom (1974), 5.2 mM'Icm'1; however, Jursinic and Dennenberg (1988)
support our results and those reported originally by Van Gorkom (1974).

The only reasonable explanation that occurred to us at the time for this
linear attenuation of the absorption changes in both the UV and visible regions
was that Oa was already partially reduced prior to the laser pulse. This could
happen if dark adaptation was incomplete, if other chemical reactions were
occurring in the dark, or if the probe light beam was inducing photochemistry.
The latter option was examined first by determining the extent of the absorption
change versus the sample exposure time after opening the shutter (Figure 4.6).
As can be seen in the figure, there is a direct correlation between the extent of
the flash induced absorption change and the length of the exposure time of the
sample to the probe beam. Our initial experiments used an exposure time of 50
ms and determined the extinction coefficient for 325 nm to range between 9 and
10 mM’1cm'1, which is confirmed by Figure 4.6. Subsequently, the exposure
time to the probe before excitation was shortened to less than 10 ms for all
further experiments. The affect of this on the C-550 change can be seen in
Figure 4.9 and on the Oa'lOa change in Figure 4.4. It proved to be unnecessary
to explore the other possibilities. This point is considered further in the following

discussion section.

If;

A

l-_

d9(

137

4.3.3. Kinetics

The above experiments showed that we were spectroscopically
observing similar phenomological events to those reported by Dekker et al.
(1984a); however, without further evidence, such as kinetic data, the association
cannot be convincingly made. Therefore, the behaviour of the Oa'lOa redox
couple was studied by using various combinations of the ferri- Iferrocyanide and
2,5-DCBO acceptor systems.

Figure 4.7 compares two decay curves of Oa'. The upper trace ls from a
iris-washed sample treated with DCMU to block electron transport to endogenous
quinones and a ferri- /ferrocyanide acceptor system for oxidation of Oa'. The
lower trace is from a similarly prepared sample except that DCMU was omitted
and both ferri-lferro-cyanide and 100 MM 2,5-DCBO were added as the acceptor
systems. The lower trace exhibits a faster decay time (11,2 ~ 1 - 1.5 ms)
indicating that oxidation of Oa', presumably by 2,5-DCBO. is much faster than
with the inorganic oxidant alone. The oxidation of Oa“ can be slowed down in
this sample by addition of DCMU and exposure of the sample to the probe light
for one minute (data not shown).

Fast reduction of Oa' in the presence of a quinone acceptor system and
slow reduction in the presence of an inorganic oxidant is conclusive for an
operational Ob site on the reducing side of the Photosystem ll membranes.
Oulnone acceptor systems bind in the Ob site in order to oxidize Oa'. Inorganic
systems do not use the Ob site, have a longer distance over which electron

transfer must occur, and thus exhibit slower decay constants.

4.3.4. Hydroxylamine

Hydroxylamine was observed by Ghanotakis and Babcock (1983) to
decrease the EPR signal arising from Yz‘t. A model was suggested (see Flgure

 

of

de

Up

san
CO”

81101

nUliil
and j

Iilldrc

138

4.2c) in which hydroxylamine interfered reversibly with electron flow between Y2
and P680 and, as a result of this interference, P680 is reduced by Oa". This
model relied upon the EPR data Ghanotakis and Babcock (1983) reported as
well as upon fluorescence information obtained from Chlorella by Den Hann et al.
(1976). The evidence implicated Yz+ as being the locus for inhibition as
opposed to P680 It was thought that, in repeating the same experiment, but
observing the decay of Oa' directly, we would observe a decay component of
d[Oa’]/dt with a tug of ~120 us, the (half-time for the backreaction between
P680+ and Oa'.

Tris-washed PS-ll membranes were diluted into a reaction media as
described in Section 4.2.1.3. 1 mM DCMU was added to block the reducing side
reactions and a 0.5 M ferri-lferrocyanide mixture was used as the accepter
system. At varying times before the experiment was started 0.5 mM NHZOH was
added to the reaction media. We were not able to detect the fast decay
components expected, t1,2 ~ 120 us, of Oa' back reacting with P680+- We did
observe, though, that the amplitude of the slow decay components, t1/2 ~ 50 ms,
decreased. This decrease was largely turnover dependent, and not dependent
upon dark incubation time in NHZOH. We were also able to restore the initially
observed amplitude by washing out the NHZOH. Similar experiments using the
same apparatus have been performed with the focus on the fast decay
components (Xing, 1989). The amplitude of the fast decay components was
shown to increase as a function of the number of turnovers.

The correlation between the slow decay amplitude decrease and the
number of turnovers, the inverse correlation to the fast decay amplitude increase,
and the restoration of the initial signal by washing out the NHZOH is conclusive of

hydroxylamine having only a reversible inhibition before P530"’ and supports the

re
9X
Sal
Silt
det.
redl
led:

139

original model proposed by Ghanotakis and Babcock (1983). These results are

summarized in Table 4.1.
4.4. DISCUSSION

We have ascertained that the optical transient we observe whose
Iineshape peaks around 325 nm is indeed due to the same species as reported
by Dekker et al., (1984a). They had related their observation to the species
reported by Van Gorkom under steady state conditions (1974) and ultimately to
the plastosemiquinone anion radical reported by Bensasson and Land (1973).
Our reported line shape and extinction coefficients are not signficantly different
from those reported by Dekker et al. (1984a), and the kinetic behaviour with
regards to exogenous quinone (t1/2 ~ 1.5 ms for the reaction Oa' + DCBO --> Oa
+ DCBO'., Dekker et al., 1984b) and inorganic oxidants appears to be consistent
with expectations. Further confirmation exists in the examination of a separate
spectroscopic feature, the C-550 bandshift. As a result we can be confident that
our reported absorption difference spectrum, taken at a point density of 1 to 2.5
points/nm, is reasonable.

A significant aspect to our independent confirmation of the work reported
by Dekker and co-workers (1984a) is that it has proven unnecessary for us to
employ kinetic phase deconvolution procedures for the determination and
removal of the individual component contributions. Furthermore, the
experimental conditions we used for the Oa'lOa spectrum determination are the
same we employed for examination of the S1 -> 82 transition. This is unlike the
situation reported by Dekker et al. (1984a). In particular, for the Oa‘lOa
determination, Diphenylcarbazide (DPC) had been employed by them as a
reductant of YZ+. While the interaction of DPC and its oxidizedform with other

redox active species, e.g., ferricyanide or ferrocyanide, is uncertain, Dekker et al.

ii
to

0X

blnt

aCIil

140

(1984a) did note that the stoichiometry for reaction with Yz+ was unknown
because the oxidized form of DPC they generated in vivo appeared to be
different than the chemically oxidized forrn generated in vitro reported by Vernon
and Shaw (1969).

Throughout this chapter the simple model of the reducing side suggested
by Equations 4.1 and 4.2 has been assumed to be operative; however, in Section
4.1.2.1, difficulties with the model were raised and they will be considered here.
The most difficult set of problems appear to revolve around the general issue of
heterogeneous populations of reaction centers. This has occurred because of
the observation that the apparent number of active reaction centers separately
determined from chlorophyll concentrations, extent of Oa reduction (Mells and
Anderson, 1983), and evolution of oxygen (Dennenberg et al., 1986 and Graan
and Ort, 1986) are not the same. In relationship to the work reported here,
understanding this apparent heterogeneity is crucial because of the reliance upon
the extent of the absorption change induced at 325 nm as being directly related
to the concentration of reaction centers participating in charge separation and

oxygen evolution.
4.4.1. Reaction Center Heterogeneity

Graan and on (1986), using chloroplasts, determined that the number of
binding sites for terbutryn, an inhibitor of the Ob binding site similar in mode of
action to DCMU, was higher than the number of centers reducing endogenous
plastoquinone. It was found that two populations of PS-II centers existed. One
population was effective at utilizing endogenous plastoquinone and could oxidize
water, the second population could also participate in water oxidation provided
specific exogenous acceptors were used. The effectiveness of these exogenous

acceptors was found to be (In decreasing order of promoting fast electron

S.
is.
all

evl

(lull
abSl
The

flUOn

141

transfer from Oa) halogenated quinones > methyl-substituted quinones ~

endogenous quinones >> ferricyanide.
4.4.gMXIternatlve Secongarv Electron Acceptors

It is probable that the two populations of PS-II centers they determined
were the PS-Ila and PS-Ilb centers that had already been proposed (see Melis
and Homann, 1978). These two types of centers have been shown to have
differing antenna sizes (Melis and Anderson, 1983) and different locations in the
chloroplasts (Melis and Homann, 1978); specifically, PS-llb is considered to be
located in the stroma along with PSI and has a smaller antenna size by at least
a factor of two. In the isolation of Photosystem-ll membranes, the PS-llb centers
are partitioned with the PS-l centers and removed from the preparation and their
existence is therefore of no consequence to this work as the preparations used
were almost wholly free of PS-I contamination. Alternative Secondary Electron
Acceptors.

Another concern to be addressed involves the existence of alternative
secondary electron acceptors and their effects upon the Oa'lOa difference
spectrum. This was already alluded to earlier In the 'Materials and Methods
Section' in discussing the order of addition of various chemicals; i.e., DCMU,
ferricyanide, and DCBO and their effects upon 0400. The existence of these
alternative acceptors has been widely reported and met with controversy as the
evidence has largely been indirect. As an example, Jursinic and Dennenberg
(1988) have proposed the existence of an ancillary (their terminology) or auxiliary
quinone acceptor, Aq. The basis for their proposal was the stimulation of the
absorption change at 325 nm they observed upon the addition of ferricyanide.
The absorption increase was consistent with the increase in the area over the

fluorescence curve in the same experiment. They have found that the midpoint

142

potential of Aq, 318 mV, ls independent of pH and has a 0.38:1.0 quOa ratio.
Unrelated experiments aimed at determining stoichiometric quantities of quinone
per reaction center (T akahashi and Katoh, 1986, and de Vitry et al., 1986) were
unable to support the plastoquinone radical identity for Yz“ (Ghanotakis et al.,
1983) because of the lack of quinone. These same conclusions are applicable
here and do not support an ancillary quinone hypothesis.

ltoh (1978) introduced a species, AH, to explain widely variable rates of
Oa’ oxidation under differing conditions as monitored by variable fluorescence.
This was refuted by Ghanotakis et al. (1983) by use of a simpler model that
negated the introduction of another electron transport component. However, the
existence has been confirmed in PS-Il membranes of an acceptor species
designated O400. Indirect evidence for 0400 has been reported since at least
1973 (lkegami and Katoh, 1973, also see Bowes et al., 1979). Petrouleas and
Diner (1986) reported the first direct evidence of O400 by using EPR and
Mossbauer spectroscopy to monitor the correlated behaviour of a ferric and
ferrous iron and they further demonstrated that this was the non-heme iron on the
reducing side of Photosystem-ll. This observation was confirmed by
Zimmermann and Rutherford (1986).

Under some conditions the ferrous iron becomes oxidized to the ferric
state and when Oa' is reduced, the Fe(lll) is reduced back to the Fe(ll) state with
a t1/2 ranging from 7 to 25 us (pH 6.5 and 7.5, Diner and Petrouleas, 1987a,
1987b). This fast reduction, being faster than either the fonlvard or back
reactions of Oa', can serve to reduce the detected extent of the reduction of Oa.
Zimmermann and Rutherford (1986) showed that the iron oxidation could be
caused by exogenously added quinones whose solution midpoint potentials are
greater than for F93+/Fez+, such as with phenyl-p-benzoquinone (PPBO). The
formation of Fe3+ was not observed with exogenously added 2,5-DCBO.

143

Separately, the iron oxidation was inhibited strongly by the use of DCMU
(Wraight, 1985).

Given these constraints on the formation of F93+ and our consistent use
of either 2,5-DCBO as an exogenously added quinone or of the DCMU and ferri-
/ferrocyanide acceptor system, we conclude that our results presented here are
not contaminated by 0400 oxidation of Oa' and no conclusive evidence exists for
other alternative electron transfer components that may interfere. The
experiments described in the following chapter limit themselves solely to the S1 --
> 82 transition and the experimental conditions we use are such that 0400 plays

no role.

144

TABLE 4.1
fifect of NHZQH on Lhe 320 nm Apsorbance Change

Effect of Hydroxylamine on the induced absolution change at 320 nm in
Tris-washed Photosystem II membranes. The observed extinction at 320 nm
before the addition of NHZOH was 10 mM'1cm'1. The granularity of each value
reported is 36 turnovers. The final, washed value is a control in which NI-IZOH
was washed out by two resuspensions in the reaction media. Recovery to with

10% of the original confirms the reversibility of the NHZOH treatment.

 

 

Number of Total Normalized Turnovers
Turnovers Absorption Change
in NH20H
0 (no add) 36 1.0

36 72 0.88

36 108 0.56

36 140 <0.25

[sample washed twice]

176 0.9

145

FIGURE 4.1
Synchronous Relationshippetween Donor and Reducim Sides

This figure demonstrates 1.) The synchronous relationship between the
binary reactions occurring on the reducing side of PS-Il and the quaternary
relationship on the oxidizing size, 2.) The variability in the order of reactions
depending upon the number of oxidizing equivalents that have been accumulated
by the OEC.

146

co co coco N cm mllllu_

mason: I an

/\

so

so we come N cm seen i on

so we coco +N cm.lllll.

2&3. use

mo so coco +N cm

AHHouzu

No + +14

me so once ~ om

nap—50h

__ 2mem>m0h01m 2.

25mm 6m.
nal: ocnw

 

7/ hicowv

/

 

 

”_tn. B‘QON

we so once N .+cm

3...“; I c Lou

m><31h<m o_._.mz_v_

b)

147

FIGURE 4.2

lnhibflowfifies in the PSII Electron Transmrt Chain.

Sites of selective inhibition in Photosystem II electron transport chain.

Back-reaction pathways are not shown. The numbers in parenthesis

represent sites of inhibition. See the next chapter for further details on

inhibition involving the Oxygen Evolving Complex.

1.

5.

Inhibition at the Ob Binding site. Inhibition is either competitive
(DCMU), removal of the site altogether (trypsin), or mutation.

Inhibition at the Oxygen Evolving Complex. Inhibition classically is
heat or TRIS treatment; all of which release manganese, and are
irreversib e.

Inhibition near the Oxygen Evolving Complex. This involves
removal of the peripheral polypeptides, the 33 kDa or removal of
ionic co-factors. This inhibition is distinctly different from the
previous one because the inhibition is reversible.

Inhibition interrupting the electron transport between the Oxygen
Evolving Complex and 2. Temperature variation allows other
endogenous onors such as b559 and Signal IIs to become
competitive.

Inhibition interrupting the electron transport between Z and P680;

Electron transfer in Photosystem If after TRIS-washing, inhibition with
DCMU, and addition of ferri- and ferro- cyanide.
Possible loci of inhibition for NHZOH. It was thought to either interrupt

electron transfer between 2 and P680 or to act directly upon P680 It is

now known that its mode of action is the former.

148

a)

(1)

----> Qb

680

OZIIIIIIIP

---->

---->

OEC

(5)

(4)

(3)

(2)

b)

/--< Fe3+

Fe3+ <--

c)

(1)

----> Qb

80

a 6
QTIIIVPTIIW

2
> H
. \xN
_ .\
.\\
\t
.

Z

149

FIGURE 4.3
Saturation Profile of Qa-lOa versps excitation intensity

This figure demonstrates that even with the variability in laser power
photosynthesis, as measured by the reduction of Oa at 320 nm, is complete at
even 1/3 of the original power. 100% laser power - 85 mJ, [Chi] - 200uM, tris-
washed Photosystem ll membranes. Donation to Ob blocked by 1 mM DCMU,
and Ferri-lFerro-cyanide used as acceptor system, 1 mM. The points
represented by '+' are the average of 16 turnovers and the points represented by
'*' are the average of 8 turnovers. No significant differences related to the onset

of saturation was observed. Further discussion can be found in the text.

(325 nm)

Normalized Absorbance

 

 

150

 

 

+
+
T d
+ + ‘
O
O
Q
i i
i.
+
20 40 so so 100

2 Maximum Loser Intensity

151

FIGURE 4.4
The Apsomtion Spectra of. Oa-lOa

This is a point by point absorption difference spectrum of the OaOa
transition in freshly prepared tris-washed Photosystem II membranes. The
oxygen evolution rates of the Photosystem ll membranes before Tris-washing
were greater than 600 umoles/mg Chi/hr. Sample conditions were as follows:
[Chl] - 124 uM, 1 mM DCMU, 1 meerricyanide, 1 meerrocyanide, pH
adjusted to 6.0. Each point represents the average of 16 to 49 shots. Shortly
before each shot the sample chamber was loaded from the sample reservoir and

flushed aftenrvards. Further discussion can be found in the text.

152

AEcV sam:o_o>o3
i own

 

(mo/flm/) uoltou1}x3

153

FIGURE 4.5
Comflson of Two Oa-lOa Smctra

This figure compares the Oa'lOa spectrum we determined and the
Oa’lOa spectrum reported by Dekker and co-workers (1984a). The points and
the solid lines drawn through them represent are our results and the dashed line
is is the line shape ultimately return This consists of the same spectra as in the
previous figure and also plotted over the same wavelength range is the same
spectra as shown in previous Figure except for an overlay of the spectra reported
by Dekker under similar conditions. As can be seen, the two spectra are quite
similar. Sample conditions were as described for Figure 4.4. Further discussion

can be found in the text.

Extinction (/mM/cm)

Extinction ([mM/cm)

15

10

10

 

 

 

154

T ' T ' T ‘ I ' r r r ' 1

 

290 330 370 410 450
Wavelenoth (hm)
i
\.
‘ \
, \
\.
.. / \
\ / .
. \\ .// \
. \ .—-—
/ l 'l
1'
l/ l l
! -.-u
ésof 330' ‘ ' 370' ' ‘ ito' ' ' 350

Wavelength (nm)

155

FIGURE 4.6
Exmsure Time Effects on the Oa-/Oa change.

This figure demonstrates the effect of exposure time to the probe beam
on the extent of the Oa/Oa change observed at 320 nm. The absorption extent
has been normalized to 1. tris-washed Photosystem ll membranes were used for
this experiment, [Chl] - 100 uM, [DCMU] - 1.0 mM. [Fe3+] - [Fe2+] - 1.0 mM.
Each point is the average of 16 turnovers. Further discussion can be found in the

text.

156

“new ocauoaau

,..sm. on

Pirlri. ”bib

ffi 1 TT

1

rT'

rfij

 

rfifijfi

N
O

V
0

ID
0

(n
O

(mu ozg) oounqaosqy pozijomJoN

157

FIGURE 4.7
_Efffects of External Acceptors on dIOa-[ldt

This figure compares the effects of two different acceptor systems with
Photosystem II membranes. Each trace represents the average of 16 turnovers
at 320 nm. The top trace was collected using only a ferri-lferrocyanide acceptor
system and DCMU to prevent reduction of Ob. The bottom trace was collected
using 2,5-DCBO in addition to ferri-lferrocyanide, DCMU was not added. Each
trace represents the average of 16 flashes and they were normalized before
plotting. Sample conditions were as follows: [Chi] - 120 MM, [DCBO] - 100 uM,
[Fe3+] - [Fe2+] - 0.5 mM. [DCMU] - 1 mM (top trace only), and pH adjusted to

6.0. Discussion in text.

158

 

 

4.0 '1

_.Aeq._u___
2 4. 6
3 2 1

Aoooa a; GOGMDHOmnfiw 0>RHMH0m

100

80

60

40

20

Time (ms)

159

FIGURE 4.8
_Apsorption Spectrum of Tris-Washed PS-ll mempranes

This is an absorption spectrum of tris-washed PS-Il membranes shortly
after preparation. This is illustrative of the optical thickness of the samples we
handle. The difficulty of collecting data below 300 nm can be appreciated by
observing the steady increase in background absorption as one goes further into
the UV. The sample has a higher background absorption in the green, around
450 nm, but this is compensated for as the sensitivity of the PMT in this region is
increasing as well as the output power from the Xenon lamp (data not shown).
Another complication, however, is that the background changes very rapidly and
so short exposure time to a probe beam is essential to get an accurate baseline.

For this experiment, the [Chi] = 37 mM and the data was collected by a

Perkin-Elmer scanning spectrometer.

160

 

a 20

ASE £32963

.50

n

.onm

_Ll

_ .

o Eu

a 20

.23

1|

II

 

8.0

and

end

006

on;

on.—

uongosqv

161

FIGURE 4.9
C-550 Bandshift

C-550 bandshift in tris-washed Photosystem Il membranes. This
bandshift is symmetrically centered around 545 nm and is due to the shift of the
intermediate pheophytin's absorption spectrum induced by the reduced Oa'.

Points were collected every 5 nm, slit width - 0.5 mm, wavelength
bandpass - 6 nm. [Chl] - 85 uM, [DCMU] . 10 MM, [Fe(CN)63’] - [Fe(CN)64'] .
0.5 mM. pH 6.0. Sample exposure time was 3 ms. It was necessary to use
another grating optimized for the visible wavelength region in the

monochromatoer for these experiments.

162

”Ecv swmco_o>o5
can own oem can own

Piilillb.-itrllllr._l - -21-. _.. l.ll...I p n. ill. b O O m w I

I

ooovl

T

x
oomlH

 

1 com

(mo/m/) uo;tou

T

coo—

 

{i

com.

onmlo

163

FIGURE 4.10
Kinetic Data Processigg

This figure depicts a simplified view of the processing of kinetic data
once it has been transferred from the transient signal averager into the host
computer. The 'Single Collection' is the actual collection by the transient signal
averager of 1024 points in real-time and transmission to the host computer. The
'N Collection' is the summation of the set of data collected and, while not done
real-time, is completed for each single collection cycle. The storage to a
computer readable file is done when this full set of data has been collected.
Further data transformations, such as the construction of the absorption
difference spectrum, is performed by other software once all the sets have been

collected.

164

352:...»
usualsaasann

  

      
   
 

 
 

'-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

              
       
  

238.32 . W
E50 :9: . m
3:6; 295 . . . v . 6
Emma 28 :88 com 528 c8
Eon 28m Eon. 29% use“. 295
35388“.
ceoxomv
l lllllllllllllllllllllllllllllllll l.
Emma 558 22$ _
as c: as . so a: as é c: as _
.
Amaaoowm a _
9.5.38.5 .89 _
nuns
sassy...
q s...
mm mm Vsosm
m m _ z
258.59 n
.
MN MW "seam
. .. IIIIIIII L Ml IIIIIIIIIII Mir IIIIIIIIII H I I L 95
esteem as E as so: 2 am as a new

165

References
Berthold, D. .A., Babcock, G. .T., and Yocum, C., F. (1981). FEBS Letters,
134,231 -234.

Bensasson, R., and Land, E.J., (1973), Biochimica et Bipphysica Acta, 325*, 345-
348.

Bolby, N. and Yocum, C. F. (1987). Private Discussions.

Bowes, J. M., Crafts, A. R., and Itoh, S. (1979). Biochimica et Biophysica Acta,
547, 320-335.

De Vitry, C., Caries, C., and Diner, B. A., (1986). FEBS Letters, 196, 203-206.

Den Hann, G. A., Garter de Vries, H., and Duysens, L. N. M. (1976). Biochimica
et Biophysica Acta, 430, 265-281.

Dekker, J. P., Van Gorkom, H. J., Brok, M., and Ouwehand, L., (1984a).
Biochimica et Bi0physica Acta, 764, 301-309.

Dekker, J. P., Van Gorkom, H. J., Wensink, J., and Ouwehand, L. (1984b).
Biochimica et Biomysica Acta, 767, 1-9.

Dennenberg, R. J., Jursinic, P. A., and McCarthy, S. A. (1986). Biochimica et
Biophysica Acta, 852, 222-233.

Diner, B. A. and Petrouleas, V. (1987a). Biochimica et Biophysica Acta, 895,

 

107-125.
Diner, 81 gig any8 Petrouleas, V. (1987b). flochimica et Biophysica Acta, 893,

Ghanotakis, D. F., Topper, J. N., Babcock, G. T., and Yocum, C. F. (1984).
FEBS Letters, 170, 169-173.

Ghanotakis, D. F. and Babcock, G. T. (1983). FEBS Letters, 153, 231-234.

Ghanotakis, D. F., O'Malley, P. J., and Babcock, GT (1983). "The Oxygen
Evolving System of Photosynthesis”, 91-101, (lnoue, Y., Crofts, A. R.,
Govin ee, Murata, N., Renger, G., and Satoh, K., eds), Academic
Press, okyo.

Graan, T. and Ort, D. R. (1986). Biochimica et Biophysica Acta, 852, 320-330.

l-Iavemagég. and Mathis, P. (1976). Biochimica et Biophysica Acta, 440, 346-

Hoganson, C., (1988) Private Discussions.

166

Ikegami, I. and Katoh, S. (1973). Plant Cellular Physiolpgy, 14, 829-836.
Itoh, S. (1978). Biochimica et Bipphysica Acta, 504, 324-340.
Joliot, P. and Joliot, A. (1981). FEBS Letters, 134, 155159.

Joliot, P. Beal, D., and Frilley, B. (1980). Journal de chimie physigue, 77, 209-
216. _

Jursinic, P. and Dennenberg, R. (1988). Biochimica et Biophysica Acta, 935,
225-235.

Kleinfeld, D., Okamura, M. Y., and Fehr, G., (1985). Biochimica et Biophysica
Acta, 809, 291 -31 0.

Kleinfeld, D., Okamura, M. Y., and Fehr, G., (1984). Biochmica et Biophysica
Acta, 766, 126-140.

Ikegami, I. and Katoh, S. (1973). Plant Cell Physiolpgy, 14, 829-836.

Lillie, D., (1984) MS. Thesis, "The Development and Characterization of a Flash
Klnetic Absorption Spectrometer", M.S.U.

Malmstadt, H. V., Enke, C. G., and Crouch, S. R. (1981). "Electronics and
Instrumentation of Scientists,” 422-426,The Benjamin/Cummings
Publishing Company, Inc., California.

McCauley, S. W. and Melis, A. (1986). Biochimica et Biophysica Acta, 849, 175-

 

182.

Melis, A. and Anderson, J. M. (1983). Biochimica et Biophysica Acta, 724, 473-
484. '

Melis, A. and Homann, P. (1978). Archives of Biochemistgy and Bipphyslcs, 190,
523-530.

Petroulegsév. and Diner, 8. A. (1986). Biochimica et Biophysica Acta, 849, 264-

Pulles, M. P., Van Gorkom, H. J., and Verschoor, G. A. M. (1976). Biochimica et
Biophysica Acta, 440, 98-106.

Renger,2G.4, Hagemann, R., and Fromme, R. (1986). FEBS Letters, 203, 210-

Renger, G. (1976a). FEBS Letters, 69, 225-230.
Renger, G. (1976b). Biochimica et Biophysica Acta, 440, 287-300.

Renger, G., Erixon, K., Doring, G., and Wolff, Ch. (1976). Biochimica et
Biophysica Acta, 440, 278-286.

Robinson, H. H., Sharp, R. R., and Yocum, C. F. (1981). Archives of
Biochemist_ry and Biophysics, 207, 1-8.

167

Robinson, H. H., Sharp, R. R., and Yocum, C. F. (1980). Biochemical
Biophysical Research Commgnications, 93, 755- 61.

Savitsky, A. and Golay, M. J. E. (1964). Analflical Chemi§t_ry, 36, 1627-1639.
Slooteen, L. (1972). Biochimica et Biophysica Acta, 275, 208-218.

Stiehl, H. H. and Witt, H. T. (1969). Z. Naturforsch. B. Inor anic Chemist
Organic Chemistry. Biochemistry. Eiopfiysicsl and Eiolpgy, 248, i588-

Takahashi, Y. and Katoh, S. (1986). Biochimica et Biophysica Acta, 848, 183-
192.

Van Gorkom, H. J. (1974). Biochimica et Biophysica Acta, 347, 439-442.

Velthuys, B. R. (1988). Biochimica et Biophysica Acta, 933, 249-257.

Vermegl‘i'oégA‘i 6agd Clayton, R. K. (1977). Biochimica et Biophysica Acta, 461,

Volker, M., Renger, G., Rutherford, A. W. (1986). Biochimica et Biophysica Acta,
851, 424-430.

Wraight, C. A. (1985). Biochimica et Biophysica Acta, 809, 320-330.
Xing, M. X. (1989). Private communication.

Zimmermann, J.-L. and Rutherford, A. W. (1986). Biochimica et Biophysica Acta,
851, 416-423.

CHAPTER 5
Optical Studies of the $1 —> 82 Transition
5.1 . INTRODUCTION

The oxygen evolving complex (OEC) was introduced in Chapter One. Its
distinct, spectroscopically observable, five redox state cycle and associated
nomenclature in terms of 80, S1, , Sn terminology was discussed in some
detail. In preparations of chloroplasts, Photosystem II membranes, and PS-II
'cores' (Ghanotakis and Yocum, 1986) that are active In evolving dioxygen, the
OEC will have associated with it three extrinsic water soluble polypeptides of 17,
23, and 33 kDa (these are referred to in the literature by some researchers as 18,
24, and 33 kDa), four manganese, and supporting Ca2+ and CI' ions. The exact
number of supporting ions is uncertain, Cammarata and Cheniae (1987) have
reported one Ca2+ per PS-II; however, Ono and Inoue (1987) have reported that
there are two Ca2+ per PS-Il reaction center. A recent report by Shen et al.
(1988) has placed the number back at one again. There is evidence for
involvement of the intrinsic 33 kDa polypeptide, the Intrinsic 34 kDa polypeptide,
or both In manganese binding but the available data is Inconcluslve and does not
pertain to the work here (see Metz et al., 1980, 1984, and 1986). The
requirements for the presence of most of these co-factor were listed In Table 1.1.
In this Chapter we present flash induced optical absorption spectra
corresponding to the $1 --> 32 transition in green plant PS-II systems under
conditions in which we have systematically depleted the endogenous CI' cofactor

and reconstituted with a series of monovalent anions known to restore 02

168

169

evolution to some extent. These results are currently being prepared for

publication elsewhere (Lillie and Babcock, 1989).

5.1.1. Polyppptide Depletion

5.1.1.1. The 23 kDa Polypeptide

Discriminating between the roles played by the 17 and 23 kDa
polypeptides in the OEC has been complex because of the degree of their
association. The three extrinsic polypeptides are normally not removed during
isolation of Oz evolving Photosystem II membranes. Incubation In concentrated
NaCI will specifically release the 17 and 23 kDa polypeptides and will leave the
33 kDa and the manganese functionally associated with the OEC (Akerlund et
al., 1982, Ghanotakis et al., 1984a and 1984b, and Miyao and Murata, 1984b).
This depletion inhibits 02 evolution. Oxygen evolution can be reconstituted
almost wholly in these polypeptide depleted membranes (hereafter referred to as
NaCI washed or salt washed) by rebinding the 23 kDa and, optionally, the 17 kDa
(Miyao and Murata, 1983). Reconstitution can also be achieved by use of non-
physiological concentrations of CaCl2 (Ghanotakis et al., 1984b, Miyao and
Murata, 1984a). In fact, Ghanotakis and co-workers showed that the
reconstitution is due to the divalent cation instead of the anion by showing 26%
and 42% reconstitution by using MgCIz and SrClz respectively and no
reconstitution with NaCI. An alternative viewpoint has been expressed, though,
by Blough and Sauer (1984) as they claimed it was the relative solubilities of the
polypeptides in the salt solutions that was crucial as opposed to the valency of
the salts. The NaCI washed PS-Il membranes were shown to have lost 40% of
their Ca2+ and the remaining Ca2+ apparently occupied loose binding sites on

the membranes (Ghanotakis et al., 1984b). Dialysis of these membranes against

170

EGTA, a chelating agent, removed the rest of the Ca2+. 02 evolution was not
stimulated in these NaCI, EGTA washed membranes by polypeptide
reassociation even though the polypeptides were shown to definitely rebind. 02
evolution could be stimulated, though, by the addition of Ca2+. Ghanotakis and
co-workers also showed that the degree of stimulation was dependent upon the
order of addition. If the Ca2+ was added before the polypeptides then Oz
evolution was stimulated as soon as the rebinding of the polypeptides occurred.
If the Ca2+ was added after the polypeptides were rebound then an hour or more
of incubation was required to achieve 90% reconstitution. From this it can be
concluded that an absolute requirement for Ca2+ exists for high rates of oxygen
evolution. The ability of Ca2+ to reconstitute oxygen activity in NaCI washed and
NaCl, EGTA washed membranes and the dependency of the rate of
reconstitution upon the order of addition between Ca2+ and the 23 kDa
polypeptide has led to the conclusion that Ca2+ serves in a structural and
possibly catalytic role in the OEC (Ghanotakis et al., 1984a). The 23 kDa seems
to play only a structural role by providing an environment that lessens but does

not eliminate the absolute requirement for Ca2+.
5.1.1.2. The 17 kDa Polypeptide.

A well defined role for the 17 kDa has not yet emerged. It is not essential
for 02 evolution as evidenced by the previous section. Miyao and Murata (1985)
and Akabori et al. (1984) have suggested that it reduces the requirement for CI'.
The CI' requirement is 5 to 10 mM upon rebinding of only the 23 kDa. Akabori et
al. (1984) observed that the Cl' requirement dropped to below 1 mM (also see
lmaoka et al., 1986). However, Waggoner and Yocum (1987) have shown that
they can deplete the 17 kDa by washing with 60 mM Mng (or CaCIz). Under

these conditions, they retain the 23 kDa largely on the membrane in an apparent

171

functional role. They did not observe a CI’ effect ascribed by others to the 17
kDa.

5.1.1.3. The 33 kDa Polypeptide.

The NaCl and EGTA treatments do not remove either the 33 kD
polypeptide or the four managanese atoms from the PS If membranes. An
additional incubation in either 1.0 M CaCl2 or 2.6 M urea is required to remove
the 33 kDa (Ono and lnoue, 1984, and Miyao and Murata, 1984b). Either
treatment leaves the four manganese atoms functionally associated with the
OEC; however, they appear to have become more sensitive to the bulk
constituents and their concentrations. Miyao and Murata (1984b) incubated
membranes depleted of the 17, 23, and 33 kDa polypeptides for four hours in a
series of salt buffers at differing salt concentrations. Those with the highest
concentrations of Cl' (200 mM) retained the highest 02 evolving activity, and also
showed that the cation, in the form of Na+, K1“, Mgz", or Ca2+, did not appear to
be responsible for retaining 02 rates, and, with regards to M9213 may even have
been inhibitory. The 33 kDa polypeptide and Cl' both appear to function in
structural/protectant roles. The evidence is as yet incomplete and does not allow

for discrimination between the two.

5.1.2. Sgctroscopic Probes of the Internal S-states

5.1.2.1. EPR Signals Arising from the Oxygen Evolving Complex.

All of the S-states of the model proposed by Kok (Kok et al., 1970) are
EPR silent except for the 82 state. In 1980 (Dismukes and Siderer, 1980 and
1981) determined that an EPR signal consisting of 16 or more peaks, the 'multi-
Iine,‘ was directly correlated with the 82 state. Another EPR signal that was

172

found in samples treated with DCMU and exposed to continuous illumination at
low temperatures and is centered around 9 - 4.1 (hereafter referred to as the
94.1) had also been observed. de Paula and Brudvig (1985) definitively
associated the 94.1 signal with a configuration of the 32 state different than that
which gives rise to the $2 multiline EPR signal. The formation of the 82 state in
NaCI washed and CI' depleted PS-ll membranes has been controversial. Blough
and Sauer (1984) reported that the amplitude of the multiline they observed
paralleled the percentage inhibition of oxygen‘evolution in NaCI washed and
reconstituted PS-II membrane samples. Franzen et al. (1985) are in agreement
concerning the NaCI washed membranes. However they have also observed
that Cl' depleted, but not polypeptide depleted samples, while not exhibiting
steady state oxygen evolution, do exhibit a strong multiline signal, indicating that
an Sz-like state can been formed. Ono at al. (1987) did not observe a multi-line
on a single flash. They did note however, that if Cl‘ were added the multi-Iine
would form after a single flash had been given. Beck (1988) has repeated the
experiments described by Franzen and Ono and found that the 94.1 signal is
formed and is converted to the multiline upon addition of NaCl. It was further
realized that the nature of the cryoprotectant, 0.4 M sucrose or 30% ethylene
glycol, strongly affected the conformation adapted by the 82 state. The 94.1

state appears to be the stable state when sucrose is used.
5.1.2.2. Optical Signals Arising from the Oxygen Evolving Complex.

The first report for the S1 --> 82 optical transition was a difference
spectrum recorded by Velthuys (1981) in choloroplasts. The difference spectrum
he reported was characterized by a broad, symmetrical band that peaked at 320
nm. He used NHZOH treated PS-Il membranes to obtain an optical spectrum for
which the contributions arising from the $1 --> 82 transition had been eliminated

173

(Joliot, 1966). This spectrum was then subtracted from a spectrum obtained from
untreated PS-II membrane, and the resulting difference spectrum was due, it was
presumed, solely to the OEC. Velthuys suggested that the magnitude of the
absorbance changes observed with successiveexcitations followed the pattern
{0,1,0,-1}. This is a short hand method of describing the changes: the first '0' is
the magnitude of the change for the SO --> S1 change, the '1' is for the S1 --> 52
change, the next '0' and '-1' are for the $2 --> S3 --> S4 changes respectively.
Dekker et al. (1984a) reported an absorption difference spectrum for the $1 -->
82 transition in DCMU inhibited Photosystem ll membranes by subtracting out
spectral contributions from Oa'/Oa and Z‘WZ. The spectra for these were
determined separately in tris-washed PS-Il membranes. They extended the work
(Dekker et al., 1984b) by determining the absorption difference spectrum for all of
the S-state transitions. Unlike the results reported by Velthuys, the $1 --> 82
spectrum reported by Dekker et al. (1984a) was asymmetric, peaked at 305 nm,
and the absorbance pattern for the set of S-state changes was {1,1,1 ,-3}. On the
basis of the Sn -> 8,, +1 bandshape similarity to the difference spectrum
between Mn(lll) and Mn(lV) gluconate complexes they assigned the S1 --> 82
transition to a Mn(lll) -> Mn(lV) transition. The interpretation of Dekker and co-
workers of the optical changes attributed to the OEC has been the subject of
considerable controversy in the literature in terms of the assignment of the Mn(lll)
--> Mn(lV) transitions and the similarity of the absorption spectra throughout the
S-state cycle. Lavergne (1986) favors the model proposed by Velthuys (1981) in
which the absorption changes follow the pattern {0,1,0,-1}. His analysis of the

, data reported by Dekker et al. (1984a and 1984b) indicates that the experimental
results could not distinguish definitively between either model. He presented
additional data from a mutant strain of IC. sorokiniana in which the initial So/S1

ratios had been varied. This data further supported the {0,1,0,-1} model.

174

Lavergne (1987) presented optical difference spectra for the S-state changes and
showed that the 81 --> 82 and 82 --> S3 spectra were similar in shape though
differing in magnitude, and both differed considerably from the SO --> S1
spectrum. The S1 -> 82 spectrum reported by Lavergne is of similar shape to
that reported by Dekker et al. (1984b). Vincent and Christou (1986) challenged
the identification of the manganese oxidation states on the basis of their
manganese model compound results. They have reported that the UV and
visible difference absorption spectra of Mn(ll,ill) and Mn(lll,lll) oxo-bridged
carboxylate complexes are as similar to the transitions reported by Dekker et al.
(1984a) as are the manganese gluconate complexes. Renger and Weiss (1986)
obtained spectra for the So --> S1 and $2 --> 83 transitions that were similar to
each other and peaked at 320 nm. The S1 --> 82 was shown to be different.
They used trypsin treated PS—ii membranes in conjunction with ferricyanide to
remove acceptor side contributions; however, trypsin has been shown to digest
the OEC in part (see Chapter 4, Section 4.1.2.2) and the spectra they show in the
presence of an exogenous quinone acceptor appear to have binary oscillation
components. This is suggestive that the trypsin treatment has not been
successful at removing all of _the acceptor side contributions. Saygin and Witt
(1985) initially differed with Dekker and co-workers as well. They showed that
differences between various models rely in effect upon variations in SOIS1
distributions and could not be resolved from experimental data. They went on to
use NHZOH to set the S-state population to one S-state specifically and used
silicomolybdate as an electron acceptor to bypass Qa to Qb donation (however,
see Grann, 1986, regarding this misconception). Their results were similar to
those reported by Renger and Weiss (1986) in that the changes for SO --> S1

and the $2 —> 83 were similar and those for the $1 --> 82 were different.
Recently, a joint paper was published (Kretschmann at al., 1987) in which the two

175

groups agreed that the $1 --> 82 and $2 --> S3 transitions have similar spectra
and are due to Mn(lii) --> Mn(lV) transitions. The So --> S1 spectrum is different
and they concluded that it is most likely due to a Mn(ll) -> Mn(iil) transition.

5.1.3. Goals of this Chapter

it is clear, given the above discourse, that resolution of the optical
difference spectra of the S-states is complex and lacking in external controls that
can be applied unambiguously to remove spectral contaminants such as Qa'lQa,
Qb'lQb, and 0400. Another approach, however, could be to examine relative
changes of the absorption bands as a result of Cl" depletion and X“
reconstitution, where X' - {OH‘, Cl‘, Br', and N031; note: throughout this
chapter, X', will be used and without qualifying terms it will be meant to include
the set of anions listed here. in principle, since the only factor changing in these
experiments is the anion co-factor, the other component contributions cancel out.
This Chapter, therefore, analyzes these relative changes for the 81 --> 82

transition.

5.2. MATERIALS AND METHODS

5.2.1. Biolggical Preparations and Handling

Chloroplasts and Photosystem ll membranes were prepared as
described in the previous Chapter. The following procedures presume the
availability of such materials, either from frozen stock (stored at -40°C) or freshly
prepared. The viability of the samples was determined strictly from their capacity
to support sustained oxygen evolution under continuous illumination. See the

'Oxygen Evolution Assays' section for details.

176

5.2.1.1. Polypeptide Depletion

Polypeptide depletion procedures are as described in Ghanotakis at al.
(1984b) and outlined here. The Photosystem ll membrane sample is adjusted to
a concentration of 2 mg Chi/ml in a buffer of 400 mM sucrose, 50 mM Mes (pH
6.0), 5 mM MgCiz, 15 mM NaCI (resuspension buffer). After stirring in the dark
for 5 minutes, an equal volume of 4 M NaCi and 50 my! Mes (pH 6.0) was added
dropwise, stirred for an additional 5 minutes in the dark, and allowed to incubate
on ice for 1 hour. The mixture was then centrifuged at 40,000 x g for 30 minutes.
The supernatant was discarded and the pellet suspended in the resuspension
buffer. This was then centrifuged at 40,000 x g for 30 minutes. The supernatant
was discarded and the pellet was suspended in the resuspension buffer at a
concentration of approximately 2.0 mg Chi/ml. if this were not to be used within 6
hours the sample was stored at -40°C; otherwise, it was kept on ice in the dark.
The resulting Photosystem ll membranes were free of the 17 and 23 kDa

polypetides as determined by Ghanotakis ef al. (1984a).
5.2.1.2. Chloride Depletion.

The same procedure as that described above was followed; however, the
last suspension took place in a salt-free media, 400 mM sucrose and 50 mM Mes
(pH 6.0). The suspension was stirred in the dark on ice for 30 minutes and then
centrifuged at 40,000 x g for 30 minutes. The supernatant was discarded, the
pellet resuspended in the salt free media, stirred in the dark on ice, and again
centrifuged for 30 minutes at 40,000 x 9. These samples were used within 6
hours of their preparation. We experimented with using 8042' as a replacement
for CI by adding 5 m_il_li Na2804 to the salt free buffer solution. No significant
differences were observed between the behaviour noted with chloride depleted

and reconstituted samples in either the salt free or 8042' buffers if the samples

177

were used within 18 hours of preparation. Samples stored in salt free media for
longer periods of time exhibited less of a percentage of reconstitution upon
CaCl2 addition compared to their respective controls than did the samples stored
in the SO42“ buffers. The decrease in the percentage of reconstitution over time
in the salt free media was most likely due to a leaching out of functional

manganese from the membranes.
5.2.1.3. Oxygen Evolution Assays

The intactness of the Oxygen Evolving Complex in control and treated
Photosystem ll membranes was measured by the rate of oxygen evolution in
continuous illumination by using a Clark-type polarographic electrode as
described by Berthold at al. (1981). Control 02 evolving PS-ll membranes are
defined for our purposes as the PS-ll enriched membranes that are derived as a
result of the preparative procedures described in the previous chapter.
Determination of control rates was carried out as follows: 3.5 ml of a reaction
media consisting of 50 mM Mes (pH 6.0) and 250 UM 2,5oDCBO was added to a
temperature controlled, water jacketed reaction cell and allowed to equilibrate
(the reaction media was suspended in the same resovoir that supplied the water
for the reaction cell and so equilibration required no more than two or three
minutes). An aliquot of the photosynthetic sample, quantity adjusted for a final
chlorophyll concentration of ~7.5 UM, was added and followed by an addition of
ferricyanide, final [Fe3+] - 2.5 mM. One minute of further equilibration was
allowed and then a white light source (General Electric, EJL, intensity ~8 kJ m‘2
s’1) was turned on and the evolution of oxygen with respect to time was
measured by a YSl model 53 Oxygen Probe Monitor connected to a strip chart
recorder. Freshly prepared Photosystem ll membrane preparations that exhibited

an oxygen evolution rate of less than 550 urnoles Ozlmg Chi/hr were discarded.

178

Photosystem ll membranes from frozen stock not exhibiting a rate within 10% of
their rates before freezing were also discarded. Polypeptide depleted
membranes not capable of at least a 65% restoration of control oxygen rates with
the additional infusion of 15 mM CaCl2 into the reaction media were discarded.
All untreated Photosystem ll membranes preparations were discarded within two
weeks of the date of their preparation. These criteria regarding the viability of the
samples for these experiments have arisen out of a concern for the quality of the
preparations. The oxygen evolving reaction centers that have been polypeptide
and chloride depleted become inhibited more quickly with respect to oxygen
evolution under continuous illumination than control samples that had not
undergone these treatments. The preparations that exhibited lower initial 02
evolution rates deactivate even more quickly than companion preparations that
have high (> 550 urnoles Oz/mg Chi/hr) rates (data not shown). This is
suggestive, though not conclusive, that while damaged preparations may be able
to support 02 evolution for a short time, they do deactivate and may continue to
have unwanted spectral contributions. This does not necessarily reflect a direct
causal effect; rather, it is probably indicative of the disintegration of OECs that

were already damaged during the depletion phases.

5.3. RESULTS
5.3.1. Steady State 02 Evolution Meaerements

5.3.1.1. Anion Effects in Control and Cl' Depleted PSII Membranes

The effects of various anions substituting for Cl' on oxygen evolution
rates in both control and NaCI washed, Cl' depleted Photosystem ll membranes

were examined. We were principally interested in observing if the effects of the

179

differing anions varied between the control and treated samples. The divalent
cation and the sum total of the monovalent anion concentrations were kept
constant and the oxygen evolution rate was monitored as a function of the mole

‘ fraction of the substituting anion relative to chloride. Figure 5.1 shows the effect
upon the oxygen evolution rate when N03' is the substituting anion. As can be
seen, little difference is exhibited regarding the required anion stoichiometries as
the slopes of the lines are nearly equal. The absolute magnitude of the response
differs, but this is to be expected because the exogenous polypeptides are in
place in the control samples, which is likely to limit access of the substituting
anion to the Cl' binding sites. Similar experiments were conducted for Br as well
with nearly identical results (data not shown); i.e., stoichiometry for anion effect
upon addition does not change between control and treated samples. in addition,
the reconstitution of oxygen evolution by the four anions separately, Cl', Br',
CH302’, and N03”, was examined and the results summarized in Table 5.2. This
confirms that our inhibition of 02 evolution is relatively complete and reversible.
Further, given the rates of reconstitution for Br‘, Cl', and NOa’, the behaviour of
the OEC for successive turnovers should be consistent with the expectation of
the OEC not being integrally damaged and able to participate in 'nonnal' S-state
transitions. These results are also consistent with those obtained by Damoder et
al. (1986). They examined oxygen evolution and the multiline formation with
different anions in CI' depleted PS-ll membranes, and observed the order of
reactivation to be Cl' > Br‘ ~ NO3' >> OH' for steady state Oz evolution.
However, they also found that for multiline reconstitution Sr was as effective as
Cl'.

180

5.3.2. Absogion Difference Sppctra

5.3.2. 1. The $1 --> 82 Transition in Untreated Samples

The optical absorption difference spectrum of the $1 --> 82 transition in
untreated Photosystem ll membranes was acquired at a resolution of 5 nm and is
displayed in Figure 5.2. The conditions were identical to those used to acquire
the Qa'/Oa difference spectrum (Figure 4.4). This figure displays the spectrum
with no attempt to remove possible contaminating species. This spectrum was
normalized at 325 nm, and the curve drawn is a best fit to the data (the point at
320 nm is considered an anomaly and discarded) without trying to induce
spectral features that are not warranted by the density of the data. Figure 5.3
displays the same S state transition and overlays the Oa/Oa spectrum
(normalized at 325 nm) from Figure 4.4. The peak of the absorption band for the
oxygen evolving sample is clearly shifted to 320 nm even though the data point
for 320 nm is rejected because of its extremity. As a comparison, consider the
first flash spectrum in oxygen evolving Photosystem ll particles under conditions
similar to ours that was published by Dekker (1985). When his spectrum is
normalized at 325 nm and compared to the spectrum in Figure 5.2 (see Figure
5.9), the bandshapes are largely similar; however, the spectrum reported by
Dekker peaks at 325 nm while the one we collected peaks at 320 nm.

5. 3.2.2. The $1 --> 82 Transition in Na Cl Treated Samples.

Oxygen evolving Photosystem ll membranes were treated with high salt
to remove the 17 and 23 kDa polypeptides and incubated in CI' free media to Cl’
deplete the membranes. The first flash normalized (325 nm) absorption
difference spectra for the cases, Cl“ depleted and no further addition (a condition

that corresponds to OH' repletion), Cl‘ depletion and Cl‘ repletion, Cl' depleted

181

and Br' repletion, and Cl' depleted and NO3' repletion, were collected and are
displayed successively in Figures 5.4, 5.5, 5.6, and 5.7. A reproduction of these
for comparison in one figure is given in Figure 5.8. The OH‘ and Br‘ substituted
were collected point by point every 2 to 2.5 nm,-and the Cl‘ and NO3' were
collected every 5 nm. The experimental conditions were identical to those used
for the acquisition of the 81 --> 82 transition in untreated Photosystem ll
membranes. The lines drawn are a smoothed fit without trying to induce spectral
features not warranted by the data. The OH' and the Br' substituted spectra have
shifted the peak of their absorption bands to 325 nm and the Cl' and N03'
spectra remain relatively centered around 320 nm. All of these spectra decrease
in intensity towards 290 nm. Only the OH' substituted spectra reaches the zero-
crossing point in the range 290-295 nm as also shown by the Oa'lOa spectrum, a
feature indicative of only quinone contribution in that wavelength region. A
comparison of these first flash difference spectra corresponding to the $1 -> 82
transition with varying anionic co-factors, possibly serving as ligands to the
manganese in the OEC, show that there are differences apparently related to the
differing anions; however, the underlying cause for the differences is not
immediately clear. In an effort to distinguish the relative changes, the normalized
spectrum arising from the untreated Photosystem ll membranes was subtracted
from the others and the results of each such subtraction are displayed in Figure
5.10. For each spectra that has been subtracted, the bottom curve depicts the
differences. A quantitative measure of the correlation between the untreated PS-
ll spectrum and the spectra arising from X' substitution can be determined by
summing the square of the differences. The results of this operation on each of
the difference spectra are given in Table 5.1. Throughout the range of available
data, 300 to 380 nm, the degree of spectral similarity, as determined by the

summed square differences, in descending order is, Br’ > NO3' > Cl‘ >> Ol-l’.

182

Focusing upon the broad absorbance band centered roughly at 320 - 325 nm,
through the range, 300 to 350 nm, the similarity of the spectra, in descending
order are N03' > Br' ~ Cl' >> OH'. Oualitatively, the leading slope of the 325 nm
absorption band in the OH', Br', and Cl' substituted spectra, from 305 to 325 nm,
is identical to the slope in the untreated spectrum except for differing amplitudes.
The following slope for the same band differs except for the the Cl“ substituted

case.

' 5.4. DISCUSSION

5.4.1. Justification for S ral Mani ulation s .

it is certainly possible to adjust the first flash absorption difference
spectrum obtained in untreated Photosystem ll membranes for the contributions
arising from Oa/Oa, Z172, and 0400. Since there is quite a similarity between our
spectrum and the one presented by Dekker (1985), we have, in fact, been able to
obtain a spectrum remarkably similar to that presented as the $1 —> 82 transition
through just such manipulations (data not shown). However, as discussed
earlier, without a technique independently separating the spectral contributions of
the active reaction centers from those partially and completely inactivated, this
would be a futile exercise and could not be further applied to unknown situations.
Our method of examining the $1 --> 82 transition under conditions of varying
anionic co-factors provides us with the means to compare the relative differences
in the spectra as they presumably arise because of the varying anionic co-
factors. The assumptions we are forced to make regarding reaction center and
OEC integrity can be largely supported by steady state oxygen evolution
measurements. The spectra have all been normalized on the point(s) obtained at

325 nm as this is the putative peak for the Oa'lOa transition. One of the critical

183

assumptions we make is that centers inactive with respect to oxygen evolution do
not undergo Oa -> Oa‘ transitions that will be detected during the timescale of
our experiments. This assumption is supported by the extinction coefficient at
325 nm obtained for the first flash for the anion- repletion cases of OH‘, Cl‘, Br',
and N032 as shown in Table 5.3. The absolute peak height appears to be
linearly related to the degree of reconstitution of oxygen evolution, data for these
conclusions are given in Table 5.2. Therefore, normalizing the spectra at 325 nm

allows us to standardize the Oa'lOa contribution(s).
5.4.; Role(s) Played_bv the Anion Co-Factors

Distinct spectral changes are observed that are dependent upon the
anion cofactor, and the 02 evolution data (Figure 5.1) indicates that the anions
probably operate in mechanistically similar fashions. The optical data we present
here would indicate that the large absorption band centered asymmetrically at
320 - 325 nm is shifted relative to the absorption band recorded for native 02
‘ evolving PS-ll membranes upon depletion of Cl' and X‘ repletion for the $1 ->
82 transition. Additionally, the steady state 02 evolution data reveals a nearly
identical stimulation of evolution of Oz between Cl' and NOa'. This is suggestive
of a structural role in the OEC being played by the anion. The alteration of the
band shape with reconstitution by Cl would also indicate that the structural
change of the OEC upon Cl depletion is irreversible though not inhibitory. in
apparent contradiction, however, are the results from Yachandra et al. (1986a
and 1987). For conditions forming the multiline or the 94.1 they report that no
structural changes occur between the S1 and the 32 states and that, while the
EXAFS data is unable discriminate between the presence or absence of a Cl“
ligand, high resolution studies of the multiline (Yachandra et al., 1986b) do not

show any differences between Cl' and Br' reconstituted samples. There are two

184

factors that may weigh heavily here regarding the role of halides in the OEC.
Yachandra and co-workers (1987 and 1986a) have reported Mn-Mn distances of
2.7 A, 1.75 A, and 1.98 A, and a composite Mn association number of 0.5 to 2.0
Mn atoms at an average distance of 2.7 A. A binuclear or trinuclear configuration
of the Mn could be responsible for the multiline and the 94.1 signals and the
remaining Mn could be at a distance greater than 3.0 A; CT ligated to these non-
participatory centers would not necessarily be detected. The second factor to
consider is the structure giving rise to the multiline. It has been shown that
various model systems of bi- and tri- nuclear centers can exhibit multiline-type of
signals (for example, Mabad et al., 1985, and Kessissoglou et al., 1989) in the 9
~ 2 region. The detection of a multiline signal in PS." membranes does not
necessarily imply that the configuration is identical for all cases in which the OEC

has been perturbed.
5.4.3. identification of the First Flash Absorption Differences

As discussed above Yachandra et al. (1987) have concluded that the $1
--> 82 transition reflects an oxidative state change of the manganese, probably
Mn(lll) --> Mn(lV), not requiring the participation of a halide as a ligand to the
manganese (Yachandra, 1986a). Beck (1988) has shown that in the absence of
Cl“ the 94.1 is formed and converts to a multiline upon the readdition of Cl', and
there is some evidence suggesting that a multiline will form in presence of at
least Br' (de Paula, 1988). in light of the work from Yachandra and co-workers
(1987), the formation of the 94.1 must also correspond to a manganese
oxidation. From this it can be presumed that, given our experimental conditions,
the multiline is formed on the first flash for Cl', Br“, and NO3' reconstituted PS-ll

membranes and the 941 formed for the case where only OH‘ was present.

185

5.4.4. Conclpsions

EXAFS, EPR, and the results we have shown here with UV optical
difference spectra clearly indicate that substitution of endogenous Ci‘ for
exogenous anions do not affect the spectroscopically observable $1 -> 32
transition. This could be interpreted as showing that Ci' is not involved as a
ligand to the Mn of the OEC; however, this is in direct contradiction to the 02
evolution data, obtained from both continuous and pulsed illumination turnovers
of the OEC, that anions are required for advancement beyond a functional S1
state1 and is unlikely to be the case.

A more reasonable explanation is that the asymmetric absorption band,
centered at 320 to 325 nm, simply does not arise from a charge transfer from the
Cl' ligand. This is supported by the fact that we were able to effectively
exchange the endogenous ligand for a series of exogenous ligands, OH', Br',
and N032 and not observe major changes in the absorption band. The only
other likely ligand candidate responsible for this absorption band is oxygen.
Unfortunately, this hypothesis is difficult to test directly; however, this does
provide some insight into model complexes that need to be constructed.

The adamantane models of Brudvig and Crabtree (1986) served as
useful starting points and somewhat more realistic models from a synthetic point
of view have been employed by Vincent and Christou (1986 and 1987). Using
these models as starting points, modifying terminal and bridging oxygen ligands
in a systematic fashion should result in determining the spectroscopic
phenomena that are sensitive to these changes. Alternatively, one could find

models that are dependent upon liganding anions such as Ci‘ and by

 

1As discussed in an earfier section, a modified 82 state that is of the same oxidation
state as the normal 82 can be obtained in the absence of an anion but further advancement does
not appear to be possible without addition of an anion; therefore, it is a non-functional 82 state.

186

systematically varying the anion determine the observable property sensitive to

anion substitution.

187

TABLE 5.1
Correlation between S1 --> 82 s ctra taken in untreated 02 evolvin PS-ll

membranes and in Ci- depleted. 7- reconstituted PS-ii membranes

This table presents a correlation between the spectrum of the 81 -> 82
transition in untreated PS-ll membranes and a spectrum of the same transition in
CI' depleted, X" reconstituted, where X' - {Cl', Br', N032 OH'}. The correlation
is measured by the sum of the square of the individual differences between each
set of two spectra. The first table makes use of the full range of data, and the

second table is restricted to data that contributes to the 325 nm absorbance
band.

Range: 300 - 390 nm

 

 

Substituting Anion (dA(Oz) - dA(X’))2
or 0009
Br' ‘ 0.007
NO - 0.003
Oi-P 0.012

Range: 300 - 350 nm

 

 

Substituting Anion («(02) - dA(X'))2
or 0.009
Br' 0.009
NO ' 0.007

0 0.012

188

TABLE 5.2

Effect of CaX2 salts u n 02 evolution in untreated and in CI- de leted X-
reconstituted PS-ii membranes.

Effect of CaCiZ, CaBr2, Ca(NO3)2 and Ca(OAc)2 upon oxygen evolution
in untreated and salt washed, Cl' depleted PS-ll membranes. Conditions were

as outline in Materials and Methods. See text for discussion.

 

 

Rates before Salt-Washing : 750 urnoles Ozlmg Cthr
Rates after Salt-Washing
and Cl' depletion : 70
Addition % Reconstitution
Untreated
None 100
CaCl2 5.7 mM 95
CaClz 14.3 m 97
CBBf‘z 5.7 m 95
CaBr14.3 m 81
Ca N63 2 5.7 m 98
Ca NO 2 14.3 m 90
Ca OAggz 5.7 m 45
Ca OAc 2 14.3 mM 37
Treated

None <10
CaCl2 5.7 mM) 50
CaClz 14.3 mM 92
CaBr2 5.7 mM 41

14.3 mM 84
Gags?) 32 5.7 mM 30
C8 NO 2 14.3 m 65
Ca (“312 5.7 m 16
Ca OAc 2 14.3 m <10

 

189

TABLE 5.3

Relative ibsorgjon gak heights for CI- depletedI X- reconstituted PS-ll
membranes.

Relative peak heights of the first flash absorption difference spectra
under Ci' depleted and OH', Cl', Br', and N03' repleted conditions, adjusted for

differences in Chlor0phyll concentrations.

 

 

Addition Relative Peak Height
Cl' 1 .0
Br' 0.8
N03' 0.4
OH 0.3

190

FIGURE 5.1
_E_ffect of Replacement of Cl- with NCE;

The normalized rate of 02 evolution versus the mole fraction of Cl‘ in
non-Cl' depleted (control) and Cl' depleted Photosystem ll membranes is plotted.
The maximum 02 evolving rate in the control PS-ll membranes was used for
normalizing both sets of data. The same data in somewhat different form are
also presented in tabular form below. The reaction media consisted of [Chi] - 7.5
UM, 250 MM 2,5-DCBO, and 2.5 mM ferricyanide. See text for discussion.

Rates before Salt-Washing : 695 urnoles 02 lmg Cthr
Rates after Salt-Washing : 150

[Ca2+][Ci'] [NO3'] Control
mM mM mM urnoles Ozlmg Chi/hr normalized2

 

14 29 0 695 1 .0

14 23 6 569 0.82
14 14.5 14.5 598 0.86
14 6 23 51 1 0.74
14 0 29 473 0.68

[Ca2+] [OF] [NO Cl' depleted
mM mM m_ urnoles Ozlmg Chi/hr norm3 norm4

 

14 29 0 498 1.0 0.72
14 23 6 526 1.06 0.76
14 14.5 14.5 404 0.81 0.58
14 6 23 394 0.79 0.57
14 0 29 356 0.72 0.51

 

2Normalized with respect to maximum rate in Control PS-ll membranes.
3Norrnallzed with respect to maximum rate in Ci’ depleted PS-ll membranes.

“Normalized with respect to maximum rate in Control PS-ll membranes.

nfitnoza + Ht_o.v\_t_s_

 

F m.o 9.0 v.0 N.o o
F, F -.L b L n p p P b b p p P b p P pifl O
T
_.N.o
fl
a
«.0
m...
f
_
o \\ 490
u \\ \\m
\\\ _
+ .\\\x\\\\ t _T w o
\\\ + u.
t-\- v
\\\t _
\\\ 4
r #—

lmoz vac I._U yo co_wuczu 0 mo co—wawpwmcouvm

'UOi}njOAa zo

pazjjomiou

192

FIGURE 5.2

S2181 Absorption Difference Spectrum in O2_Evolving, Uninhibited Samples.
This figure is a point by point difference spectrum of the 81/52 transition
in freshly prepared Photosystem ll membranes, 700 wholes Ozlmg Cthr. The
spectrum was normalized at 325 nm, [Chi] - 80.2 uM ([RC] ~ 0.32 uM), 1 mM
DCMU, 20 mM CaClz, 1 mM Fe3+, 1 mM Fe2+. Each point is the average of
between 8 to 16 shots. Other conditions are as described in the Materials and

Methods, see text for discussion.

02 Centers

Untreated PS-ll

193

 

 

 

O
a:
i"°
#-
'o
l‘
'I')
i.
r-
is
in
’I’)
[o
n
j-n
p
+ l"
L
O
’-
i-n
b
p
L‘O
CO
T—‘er'T‘j‘T‘ijT—r‘l‘l'Iji'r'I' N
~0- N o a no - N o
v- I- v- O O O O O

[mu gag) oouoqeosqy pezijomiou

Wavelength (nm)

194

FIGURE 5.3

A Comarison Between the SZfiii and the Qa-lOa SMra.

This figure presents a comparison between two point by point spectra for
the $1 --> 82, '+', and Qa --> Oa', transitions. Both spectra have been
normalized at 325 nm. The inset spectrum is the same data except shown over
the full wavelength range for which we have collected the Oa'lOa spectrum.
Conditions for the $1 --> 82 spectrum are as described for Figure 5.2.
Conditions for the Ga --> Oa‘ spectrum are as described in Figure 4.4. See text

for discussion.

0¢-/Oo — '.

52/51— '§':

Full Stole:

'fi

1'1
390

Vr'tv'w—r—v'rr'

 

 

 

110

'fi TV
330 350 3 o

310

5
Iaveienglh (am)

195

 

to
a:
r:

 

 

; ; 3 a 5 g ‘6 ‘6 {g .5 c
(“"1 55C) .3UDQJOIq' ’.‘!II’JIUON' i
l
O
G
\
i
O
O
+
V
i
D
'- 0
Q + _ _
N
O . . o
m . ‘—
o h'!,
+ i
p
'f' p (:g
6'!
N
riw' ' ‘1 I ‘11' V I ' ' ' i ‘ ‘I‘* i ' ‘I Y r ‘0 1,2, 1* v 1 4v I
“' ‘3‘ ‘:’ (I: “, “, C“ ‘:>
O C I . . . . .
(mu 539) ’°"DQJ°‘9V Pazjjomiou

Wavelength (nm)

196

FIGURE 5.4
CI- Depleted. No Addition (OH-L ist Flash Difference Sppctrum.

Point by point difference spectrum of the 81/82 transition in salt washed,
Cl' depleted and no added anion, corresponds to OH' substitution instead.
Spectrum was normalized at 325 nm. [Chi] - 80.2 UM ([RC] ~ 0.32 uM), 1 mM
DCMU, 20 mM CaClz, 1 mM Fe3+, 1 mM F921". Each point is the average of
between 8 to 16 shots. Other conditions are as described in the Materials and

Methods, see text for discussion.

Ci— depleted. no addition

197

+

 

V’ N 0 no 3 V' N

v- v- '- O O O O
I

(mm 53:) aouaqiosqy paz [DWJON

 

Wavelength (nm)

198

FIGURE 5.5
CI- Depleted, CI- Repletej. 1st Flash Difference Spectrppl.

ggint by point difference spectrum of the 81/82 transition in salt washed,
Cl' depleted, CI' repleted Photosystem ll membranes. Spectrum was normalized
at 325 nm. [Chi] . 100 uM ([RC] ~ 0.4 uM), 1 mM DCMU, 20 mM CaClz, 1 mM
Fe3+, 1 mM Fe2+. Each point is the average of between 8 to 16 shots. Other

conditions are as described in the Materials and Methods, see text for discussion.

Ci~ depleted. Cl— reconstituted

199

 

 

 

o
a:
l'n
t
h
o
+ ~S;
D
r-
0
_1n
Ln
)-
c
(.n
n
+ l—
P
o
P'-
n
+ i
in
o
. a:
~r N o " co co ~1- N o
.— - v- o c o o c

(mu gzg) souaqaosqy pOijomJON

Wavelength (nm)

200

FIGURE 5.6
CI- DepletedI Br- RepletedJ st Flash [_Difference Spectru_r_n_._

Point by point difference spectrum of the 81/82 transition in salt washed,
Cl' depleted, Br‘ repleted Photosystem ll membranes. Spectrum was normalized
at 325 nm. [Chi] - 100 MM ([RC] ~ 0.4 uM), 1 mM DCMU, 20 mM CaClz, 1 mM
Fe3+, 1 mM F924". Each point is the average of between 8 to 16 shots. Other

conditions are as described in the Materials and Methods, see text for discussion.

201

Aficv s~o=o.o>oa

 

 

2m . _ .2.». r .omm 0 gm . .SM .2:
.4

.4

+ 1

a.

h

vo«=*_aucoooa lam .voao.aov 1.0

 

0.9

«.c

+.o

v.0

(mu gzg) oouaqrasqy pozuauuau

202

FIGURE 5.7
CI- Depleted. N03- Repleted, 1st Flash Difference Spectru_n_i_.

Point by point difference spectrum of the 81/32 transition in salt washed,
Cl' depleted, NO3' repleted Photosystem il membranes. Spectrum was
normalized at 325 nm, [Chi] - 100 UM ([RC] ~ 0.4 UM), 1 mM DCMU, 20 mM
Ca2+, 1 mM F931: 1 mM Fe2+. Each point is the average of between 8 to 16
shots. Other conditions are as described in the Materials and Methods, see text

for discussion.

203

aficv sauce—use:
mum cam an

b b b n

r h

ses=s_s.=ess. tnoz .ees._a.e t_e

+

o.o
~.o
v.0
,2

c.o

gnu gzg) oauoqiosqy pOZ[|DmJON

204

FIGURE 5.8

CI- Depleted, X- Rflleteg, 1st Flash Difference Sacha.

Reproduction of the data in Figures 5.4, 5.5, 5.6, and 5.7 onto one page
for easier comparison. Each plot is labelled, conditions for each can be found in

the corresponding figure legend.

205

“is. ‘.’:o.o>oh “Ecv s—oco_o>eh

 

 

 

 

 

 

 

 

 

 

can osn can can o—n can can osn own » ohm, s. o—m L oaw 0.0
P r bl? p r h r h r p r b I F s p P - °.O p r F > b r P r b i f i
Y Y
N
o
s m
a
+ I
+ i
v 1
.
P
V.
q
‘
0
u
.
u
a
e
U
z
s
U V
w I
W ..F
so.=...u¢ouo. tan .vo.o_nos 1.0 v0.3...ncouo. inoz .vo.o_nos 1.0
“Es. x.uco_o>ol ”Ea. s.uco_o>ok N
0mm » own P Omn P OM” L Opp” b OGN°.° Oar” » - bchbn r - row-n h p pcnhn p b rev.“ 1 b DOD °.°
o. o. _-
v N-O m
+ + 1 W
. .6 n
z
+ . w
y 0.9 v.
s . u
+ n
+ . a.o q
+ 0
v u
s u
+ v 0.. \J
a n
S
1 a;
U
. w
1 v.—
"

 

ee_._eee o: .ee.e_aoe 1_o cu.s....seoo. ;_o .oe.e_aee 1_o

(mu gzg) OOUDQJOBQY P.2[jDWJON

(mu gag) eauaqiosqy PIZjIDWION

206

FIGURE 5.9
Comflson Between Two First FlashQiflerence Apsorption Sgctra.

Replot of the data reported by Dekker (1985) for the first flash in Oz
evolving Photosystem ll membranes, '+', and a replot of the data given in Figure
5.2, '0'. Both sets of data are normalized at 325 nm. The conditions used by
Dekker differed in that he used only 5 UM ferricyanide, no ferrocyanide, and 5
mM MgCI2. The concentration of DCMU also differed, 10 UM, but this should
have no secondary effect. The other differences noted may be significant, see

text for further discussion.

207

”Eng :uuco_o>o3

 

seasosaa ;.e_. as... .e____o .e.

b h

.zesssa

 

c.o

it)
C

O
P

(mu gag) BOUDQJOSQY pazijowioN

208

FIGURE 5.10

Comparative Plots for the S2/S1 Difference S ctra Between 9; and X-
substituted, X- dClaira N6 3-, OH}.

Each plot contains the normalized (325 nm) first flash absorption

 

 

spectrum obtained in untreated 02 evolving Photosystem ll membranes, '+', the
same spectrum obtained in X' reconstituted Photosystem ll membranes, '0', and
the difference between the two, '.'. In order, a) no substitution, OH‘
reconstitution, b) Cl' reconstitution, 0) N03' reconstitution, and d) Br'

reconstitution. See text for further discussion.

Normalized Aheerhence (325 nm)

1
O
at

Nermeileed Ibeerhenee (325 nm)

I
O
M

02 - o‘e. o“- _ 0.0. D"'(°2 - 0“-) - e.e

o
O
A

 

U
0
O

0.5 1

0.0 1

 

r ‘ fi
320

V 330'7
leveiength (an)

- Jeofi

02 - '+'. Cl- - 'e'. 0111(02 - C|-) - ‘.'

209

 

$1077 -
Ieeeieegth (no)

Iermellzed lieerheeee (323 um)

I
O
O
u-

500

 

Nermeiized Aheerhenee (325 um)

-

e

o
A

o

O

on
A

01 - '+'. NOJ- - '0', Olfl(02 - NOJ-) - '.'

 

 

 

-005 ' v 1 ' fie ' I fifi fie V—1
300 320 340 300 300
Ieeeieegth (an)
02 - '+'. Ir- - '.'. 0111(02 - .,-) - -,-
1.3 '

lermelieed Aheerbenee (325 em)

 

 

330 f ' ' 330 ' ' ' :30 ' ' sea
leeeieegth (nu)

U
0
O

02 - ON— (ee edd)

 

340 7

Ieeeiength (an)

210

REFERENCES

Akabori, K., lmaoka, A., and Toyoshima, Y. (1984). FEBS Letters, 173, 36-40.

Akerlund, H.-E., Jansson, C., and Andersson, B. (1982). Biochimica et
Biophysica Acta, 681, 1-10.

Beck, W. F. (1988). "Ligand-Exchan e and Oxido-Reduction Reactions of Water
Analogs With the Manganese etramer Complex of Photosystem ll,”
Ph.D. Thesis, Yale University.

Berthold, D. A., Babcock, G. T., and Yocum, C. F. (1981). FEBS Letters, 134,
231-234.

Blough, N. V. and Sauer, K. (1984). Biochimica et Bipphysica Acta, 767, 377-
381.

Brudvig, G. W. and Crabtree, R. H. (1986). Proceedin s of the National
Academy of Sciences. [LS5 83, 4586-4588.
Cammarata, K. V. and Cheniae, G. M. (1987). Plant Physiolpgy, 84, 587-598.

Damoder, R. Klimov, V. V., Dismukes, G. C. (1986) Biochimica et Blomysica
Acta, 848, 378-391.

Dekker, J. P. (1985 . Ph.D. Thesis, ”Electron Transfer in the Oxygen - Evolving
System of hotosynthesis,” p48.

Dekker, J. P., Van Gorkom, H. J., Brok, M., and Ouwehand, L., (1984a).
Biochimica et Biophysica Acta, 764, 301-309. '

Dekker, J. P., Van Gorkom, H. J., Wensink, J., and Ouwehand, L. (1984b).
Biochimica et Biophysica Acta, 767, 1-9.

de Paula, J. C. (1988). Personal communication r arding experiments he had
performed and were being continued in Pro essor Brudvig's laboratory.

de Paula, J. C. and Brudvig, G. W. (1985). Journal of the American Chemical
Society, 107, 2643-2648.

Dismukes, G. C. and Siderer, Y. (1981). Proceedin s of the National Academ
of the Sciences USA, 78, 274-2 8.

Dismukes, G. C. and Siderer, Y. (1980). Febs Letters, 121, 78-80.

 

Franzen, L.-G., Hansson, O., and Andreasson, L.-E. (1985). _Biochimica et
Biophysica Acta, 808, 171.

Ghanota‘kgss, D. F. and Yocum, C. F. (1986). Photosynthesis Researfi, 10, 483-

Ghanotakis, D. F., Topper, J. N., Babcock, G. T., and Yocum, C. F. (1984a).
FEBS Letters, 170, 169-173.

211

Ghanotakis, D. F., Babcock, G. T., and Yocum, C. F. (1984b). FEBS Letters,
167, 127-130.

Grann, T. (1986). FEBS Letters, 206, 9-14.

lmaoka, A., Akabori, K., Yan i, M., lzumi, K., Toyoshima, Y., Kawamori, A.,
Nakayama, H., and ato, J. (1986). Biochimica et Biophysica Acta, 848,
201-211. .

Joliot, P. (1966). Biochimica et Biophysica Acta, 126, 587-590.

Kessissoglou, D. P., Kirk, M. L., Bender, C. A., Myoung, S. L., and Pecoraro, V.
L. (1989). Chemical Communications, submitted.

Kok, B., Forbush, B., and McGIoin, M. (1970). Photochemistry and Photobioipgy,
1 1 , 457-475.

Kretschmann, H., Dekker, J. .P., Saygin, O., and Witt, H. .T. (1987). Biochimica
et Biophysica Acta, 932, 358-361.

Lavergne, J. (1987). Biochimica et Biophysica Acta, 894, 91-107.
Lavergne, J. (1986). Photochemistgy and Photobiology, 43, 311-317.

Mabad, B., Cassoux, P., Tucha ues, J.-P., Hwang, Y., and Henrickson, D. N.
(1985). Journal of the merican Chemical Society, 107, 2801-2803.

Metz, J. G., Wang, J., and Bishop, N. l. (1980). FEBS Letters, 114, 61 -64.
Metz, J. G. and Seibert, M. (1984). Plant Physiology, 76, 829-832.

Metz, J. G., Pakrasi, H. B., Seibert, M, and Amtzen, C. J. (1986). FEBS Letters.
205, 269-274.

Miyao, M. and Murata, N. (1985). FEBS Letters, 180, 303-308.

Miyao, M. and Murata, N. (1984a). FEBS Letters, 168, 118-120.

Miyao, M. and Murata, N. (1984b). FEBS Letters, 170, 350—354.

Miyao, M. and Murata, N. (1983). Biochimica Biophysica Acta, 725, 87-93.

Ono, T. and lnoue, Y. (1987). FEBS Letters, 227, 147-152.

Ono, T. and lnoue, Y. (1984). FEBS Letters, 168, 281 -286.

Ono, T., Zimmermann, J. L., lnoue, Y., and Rutherford, A. W. (1987). "Progress
in Photosynthesis Research," Volume 1, 653-656 (Biggins, J., ed.),
Martinus Nijhoff Publishers, Dordrecht, The Netherlands.

Renger, G. and Weiss, W. (1986). flpchimicg et Biophysica Acta, 850, 184-196.

Saygin, 91.3? Witt, H. T. (1985). Photobiochemistgy and Photobimhysics, 10,

212

Shen, J.-R., Satoh, K., and Katoh, S. (1988). Biochimica et Biophysica Acta,
936, 386-394.

Velthuys, B. R. (1981). "Proceedings of the 5th lntemational Photos nthesis
Congress," Vol 2, "Photosynthesis ll. Electron Transport an
Photophosphorylationf' 75-85, (Akoyunoglou, G., ed.), Balaban
International Science Services, PA.

Vincent, J. B. and Christou, G. (1987). lnorganica Chimica Ag, 136, L41 -L43.
Vincent, J. B. and Christou, G. (1986). FEBS Letters, 207, 250-252.

Waggoner, C. M. and Yocum, C. F. (1987). "Pr ress in Photosynthesis
Research" Volume1,(Biggins, J.,,enJ, 85-688, Martinus Nijhoff
Publishers, Dordrecht, he Netherla

Yachandra, V. K., Guiles, R. D., McDermott, A., Cole, J. L., Britt, R. D.,
De;hesimer, S. L., Sauer, K., and Klein, M. P. (1987). Biochemistgy, 26,
59 4- 981.

Yachandra, V. K., Guiles, R. D., McDermott, A., Britt, R. D., Dexheimer, S. L.,
8326323?" and Klein, M. P. (1986a). Biochimica et Biophysica Acta, 850,
3 - .

Yachandra, V. K., Guiles, R. D., McDermott, A., Britt, R. D., Dexheimer, S. L.,
33519321; and Klein, M. P. (1986b). Biochimica et Biophysica Acta, 850,

APPENDIX A

Flash Absorption Spectrometer Computer Interface

A.1. Introduction

Note

This Appendix is a modification of the original documentation concerning the FASCI.
As such the purpose of the document was to be instructive in the uses and capabflties of FASCI
Instead of justifying particular design considerations. Additionally, the reader will note that there
are areas of overfapplng coverage between this Appendix and Chapters 2 and 3, but in terms of

technical detail this Appendix should be more complete.

FASCI is a computer interface package which interfaces a PDP-11/xx
computer, operating under the RSX-11M/MPLUS operating system, to the
Babcock Laboratory's Flash Absorption Spectrometer. This document is written
to provide enough information to operate the spectrometer and take meaningful
data. Some of the more conventional (i.e, commonly encountered) problems will
be discussed. Hardware and/or software malfunctions are not covered, and the
user is referred to the hardware schematics and Marty Rabb.

it is assumed that users of the FASCI spectrometer have a basic
understanding of double beam and single beam absorbance spectroscopy and
familiarity with the computer operating system. it is NOT necessary to be a
programmer to use the interface; however, an understanding of the mathematical

and graphical tools available is essential.

A213

A214

The interface package was designed, and the spectrometer built by
Dwight Lillie of the Babcock Research Laboratory in partial fulfillment of the
requirements for his doctoral project in photosynthesis in the Department of
Chemistry at Michigan State University. All components built or designed at
Michigan State were done so by Lillie unless stated otherwise. This

documentation was written June, 1987 shortly before leaving M.S.U.
A.2. Overall description

The FASCI spectrometer consists of several component systems:
Optical,
Electronic, analogue,
Electronic, digital,
Electronic, timing,
Computer, hardware.

Computer, software,

A.2.1. tical S stem

The stability of the optical system ultimately determines the resolution
and reproducibility of the FAS. Current setup allows dO.D. of 0.0005 and 0.005

for simple changes and kinetic resolution respectively.

A.2. 1. 1. Probe light source.

This is a 75W Xenon Arc Lamp from Photon Technology with an optical
feedback system. The optical feedback system uses a separate photodiode that
receives its input from a fraction of the light generated by the Xenon Arc Lamp. A
beam splitter is used to divert the output fraction into the photodiode. A current is

A215

generated by the photodiode, it is amplified, and used in the feedback control of
the lamp.

A.2. 1.2. Iris.

This is a pin actuated, hand operated iris from Melles Griot and it is used

to control the quantity of light reaching the entrance slits to the monochromator.
A.2. 1.3. Electronic shutter and computer control.

This is a mechanical shutter from Uniblitz that is controlled electronically.
The power supply and computer control for the shutter were designed at
Michigan State and it can be operated either manually or automatically from a

computer.
A.2.1.4. U V- VlS monochromator.

The monochromater is an Instrument SA H-20, with fixed
interchangeable slits 010.5, 1.0, and 2.0mm giving an optical bandwidth 012.0,
4.0, and 8.0nm respectively. Two ion-etched concave holographic gratings for
the UV and near-IR, as well as a normal concave holographic grating for the UV-

VlS region.
A.2.1.5. Quartz focusing optics.

These optics are used within the sample compartment (next item) to
focus the probe light emerging from the monochromater through the sample cell

and onto the Photomultiplier Tube.

A21 6
A.2. 1. 6. Sample Compartment.

The sample compartment was built by the machine shop of the
Chemistry Department at Michigan State University. It serves a structural role in
that it holds the monochromater, the sample cell(s), and the detector at fixed
distances. It also serves to protect the sample cell from ambient light while a

sample may be equilibrating before an experiment.
A.2. 1. 7 Optical Filter holder.

A filter holder was constructed by the machine shop to protect the
detector against the light caused by the laser during sample excitation. It was
designed for space minimization between sample cell location and PMT and can
easily accomodate more than one filter to be held with thin plastic spacers
between each. It was also built with mounting slots for a mechanical shutter

(next item).
A.2.1.8. Mechanical shutter.

The mechanical shutter fits between PMT mounting and the sample
compartment for blocking light from PMT entrance while allowing the sample

compartment to be exposed to light.
AS. 1.9. Photodetector Unit.

This unit consists of a housing for a detector, a mounting platform inside
the housing for the detector, and the detector itself. The housing is bolted on the
sample compartment behind the sample cell so that its surface is flush with the
plate it is mounted onto. Two types of detectors are supported, one is an EMI
9558-OB photomultiplier tube and the other other is an EG&G FND-f 00
photodiode. The photomultiplier tube is used for work from 250 to ~700 nm and

A217

the photodiode is to be used from 700 to ~900 nm. Each detector has its own

mounting platform because the requirements differ for each.

A.2.2. Electronics, analmue

The analogue electronic system interfaces with the optical system at the
PMT, and processes the analog signal from the PMT into a form suitable for the

Analog-to-Digltai converters (ADC). The various components involved are:
A.2.2. 1. Detector Amplification Circuits.

The photodiode produces current that is converted to voltage by using a
resistor. An op-amp (Burr Brown 3551) is used to match impedances and to
send the signal through 6" of cable to reach the PMT post-amplifiers. The EMI
9558-OB requires a divider chain to take advanatage of the secondary photo-
electron multiplication effect. The divider chain determines the maximum
allowable current (up to the limit of the PMT itself) that can be extracted from the
dynodes. The current available in the divider chain must be maintained at levels
greater than 20 times that of the maximum current drawn from the dynodes via
secondary electron emission. On the other hand, too high of a current flow will
increase the temperature of the resistors and wires that make up the divider
chain. This heat increase will increase the temperature of the dynodes, thereby
lowering the potential required for secondary emission and a noiser and less

predictable output will ensue.
A.2.2.2. Photodector Post Amplifiers.

This is a two-amplifier (Burr Brown 3551, 50Mhz GBP) package we

designed that is arranged in a current-follower with an inverting amplifier

A218

configuration. A current offset control allows a balance current to adjust the

output around the zero crossing threshold.
A.2. 2. 3. Signal Amplifier.

The signal amplifier serves three functions: it provides a fixed and user
adjustable signal amplification, it provides a user adjustable time-constant, and it
provides two signal outputs. The two signal outputs contain the original signal,
possibly amplified, and the original signal with the DC component(s) subtracted

out.
A.2.2. 4. 3 Fourth-order Active filters.

A set of fourth-order active filters were constructed and used to process
the analog signal between the amplifiers and the ADC. These were used
because the frequency cutoff point is considerably sharper than with simple
passive or active first order filters. These are adjusted to provide minimal filtering
s that no digital sampling is performed on the analog signal containing

frequencies higher than the nyquist frequency.

A.2.2. 5. ADC pre-amplifiers.

The A003 each have a pre-ampiifier that adjusts the incoming analog
signal so that the digitizing window is filled. The pie-amplifier for the slow ADC
usually has a unit gain because the background signal has already been adjusted

to give ~10 V.

A.2.3. ElectronicI digital

The digital electronic component system provides an interface to the

analog electronic system via the two A008, and provides an interface to the

A219

computer hardware. It effectively includes the digital transmission, handshaking,

and other required components. The component list:
A.2. 3. 1. Analog to Digital Converter and Storage.

This system consists of the electronics necessary to convert, store, and
retrieve an analogue signal. The analogue signal is presumed to have already
been split into two signals: a DC. offset analogue signal that contains the
transient information, and the unprocessed analogue signal for baseline
information.

Two converters are used:

FAST/ULTRA-FAST Converter (FADC)

SLOW Converter (SADC)
The FADC converter is used to sample the DC. offset analog signal after having
been amplified by the ADC preamp for collection of the transient signal. The
SADC converter is used to sample the non-offset analog signal for collection of
the baseline. Each converter is connected to 1024 bytes of 8 bit memory.
Additional circuitry exists to initialize and sequentially read out of memory upon

activation of data request lines.
A.2.3.2. Signal Transmission via parallel cable.

Sixteen bits of data, 8 bits for the FADC, 8 bits for the SADC, and 2 bits
for handshaking are sent along 32 conductor parallel cable. Properly, the
number of conductors in the cable should be double the number of signal lines so
that every other conductor would be ground. This would minimize groundtalk
between the communication lines; unfortunately, not enough 40 conductor cable

was readily available so the convention of every other conductor being ground

A220

was violated for the 2 bits of handshaking. Generation of these signals are
handled by the ADC.

A.2. 3. 3. Signal Reception.

This component set comprises those electronics responsible for shaping
transmission information into a suitable form for transmission through many

meters of media, the transmission documentation, and reception of the data.
A.2. 3. 4. Computer handshaking.

This set provides signal transformation of the handshaking generated by
the ADC into what is appropriate for the parallel board used by the DEC 11/xx
computer. In the current case a DRV-11J, a simple input/output parallel port
board, is used.

A.2.4. Electronics. Timing

This system consists of 'l'i'L pulses generated by the computer and sent
out to the appropriate device. All devices under computer control are controlled
via simple pulses. Generally, while there may be one or more pulses used to
control function there is at least one pulse which forces a device to 'clear,‘ i.e.,
return to default state. The appropriate components are given below in the

following sections.

A.2.5. Codar “liming Board.

This board generates up to 16 pulses, each of which, except for the first,
is triggered to begin a delay countdown by the 'firing' of the previous pulse. Each

countdown is 16 bits, and programmable over the range of 250 ns to 16 s.

A221
A.2.6. Pulse Enable.

This is a programmable mask for the timing pulses. The timing board
has no provision for software selection of enabled output lines when the timers
are wired in a cascade configuration. The addition of a separate set of enable
lines allows a Boolean 'mask' to be created via software that fulfills this need.

The output port of a DRV-11J is used.
A.2.7. Signal Transmission.

Signal Transmission occurs via a 32 conductor cable. Unlike the earlier

situation, the convention of every other conductor being a ground is followed.
A.2.7. 1. Signal Reception and distribution.

This system handles appropriate line termination and distribution of the
signals to a panel with 15 BNC connectors. Every device to be controlled in this

manner requires a 120 ohm resistor to properly terminate the input signal line.
A.2.8. ComputerI hardware

The computer uses the hardware discussed in the following sections.
A.2. 7.2. Codar Timing Board.

See the previous section for a discussion of this.
A.2. 7.3. Parallel Port.

The output pulses emitted when a timer is finished are required to trigger
the next timer in the sequence, but the pulses need to be disabled from being
sent to the final output devices. The parallel port, a DRV-11J, is used to provide

A222

16 parallel lines for masking the timers' output. See the previous section on

'pulse enable.‘
A.3. Miscellaneous Hardware

A video display terminal is needed for user interaction. A video display
terminal, either the one used for user interaction or another separate one, is
required for graphics display. Synchronous data plotting requires either a
Hewlett- Packard plotter or a dot matrix printer capable of raster graphics. If the
plotting is to be done while unattended then a dot matrix printer iks suggested.
This situation will occur frequently when a continuous set of experiments are
being performed. At the completion of one experiment, the data plotting for the
experiment can be initiated, and the user can then begin another experiment
without having to wait for the completion of the plotting.

Storage of collected data requires some form of storage that is mounted
and accessible. The preferred device is a fast, hard disk, but also acceptable
devices include mag tape or another computer system accessible over an
ETHERNET. The problem with the last two as the principal storage device is

access time - especially in the middle of an experiment.
A.4. Computer software

Two primary tasks are responsible for working together to use all of the
other systems in a useful manner. Other support tasks also exist to provide
additional functions such as ADC calibration, data massaging, graphics, etc.

The two primary tasks are divided into a user interface and a instrument
interface, and communicate with each other through a well defined software
interface. While more detail will occur later; it is enough to mention here that the

instrument interface is not dependent upon the user interface per-se. Another

A223

task(s) could interface to it following the correct protocol and thus allow the
experiment to take on new functionalities. High level access has been provided
on the user interface and so that construction of a new user interface should not
be too onerous of an undertaking.

Original documentation and design specifications referenced a third task,
the Graphics Display Processor (GDP). it still exists and can be used but
because of its drain on system resources it is not routinely used and the current
software will require a recompilation with the support for the GDP turned on. For
the lifetime of this project the functionality provided by the Nicolet 1074 and its
display oscilloscpoe should serve adequately.

The next section will be wholly concerned with what's required to conduct
an experiment. The various screens the user is presented with during the course
of execution will be examined. The following section will then discuss some of
the software design philosophies that were employed. The last section will

discuss building the various tasks.

A.4.1. Prelimingg - some requirements before starting.

The two tasks responsible are known as 'UPLODR.TSK' and 'FAS.TSK.'
Both UPLODR.TSK and FAS.TSK are privileged. UPLODR.TSK must be
installed (default task name is assumed), and a recommended priority of no lower
than 140 (out of 255) is recommended. FAS.TSK doesn't have to be installed,
but if it's not installed then the user must be a privileged user in order to activate
it. Currently, UPLODR.TSK is installed at boot time from LB1 :[1,54] via the
GTBINS.CMD file which is run by STARTUP.CMD. A flying install is done on
FAS.TSK, located in LB1:[1,54], via CCL. The necessary command is located in
LB:[1,5]SYSCCL.CCL, and is typically activated by typing 'FASCI.‘

A224

It is a good idea for the user to be privileged because if a problem
develops and the software doesn't synchronize or falls out of synchronization
then a lockup could occur. if a lockup occurs then no other tasks can be loaded,
KILL will not be able to run (unless its priority at install time is raised above
UPLODR's), and a privileged user would have to come in and ABOrt UPLODR
and/or FAS.

A file, FASCALlB.FlL;1, MUST BE AND IS EXPECTED TO BE located in
LB:[1,1]. This file contains calibration information about the ADCs, if the file is
not there then it will have to be regenerated by the program ADCCAL.TSK.

A.4.2. Starting

If all of the above is correct, and it should be, then the interface is
activated, as mentioned above, by typing 'FASCI.‘ The screen should clear (if the
terminal is either a VT100 or a VT52 and the type is set correctly), a message
that UPLODR is running should appear, and a series of lines validating that
dynamic memory buffers were successfully created and read and write access
has been guaranteed. The screen should again be cleared off, a menu written,
and a prompt of the form 'FASCI > ' on the left side of the screen roughly 213 of

the way down.
a) Failure to Start

If only a part of this sequence has occurred and then stopped; wait for
roughly ten seconds. if nothing else has still happened, one can assume the
software de-synchronized and a lockup may have occurred. Abort the tasks in

the order: UPLODR first and then FAS.

A225
A.4.3. Execution

Continuing to describe an experiment in this serial fashion becomes not
only difficult but confusing as well. As can be seen in the next section there are
several options the user should specify to define the experiment. User can
choose them in the order and manner desired; some of the option ranges depend
on values already chosen for other selections.

Therefore, each of the option will be discussed in turn. Meaning of the
option descriptor, invocation of the option, other dependencies, etc. will be
discussed for each where appropriate. It shall be left to the user to determine

what needs to be specified and the order in which they are specified.

A.4.4. Execptjon Screen Display

 

exn execur 1.01110 WRITE SETTE nernaait
apt/[sun AVG]I[XSCALE]/[YSCALE]I[BS]/[NAS]}
neponr sAveusuu AVG] [331}

 

 

 

 

 

fast adc rea. v/b slow ado rea. vlb
_Sample Zero Otaput Device _SYO:
_Tltte
_e Timing Events data buffer
_Sequenoe Number _lteration Number
_ADc [ON OFF] _Sampie Rate q
_adc GAin _Data Transformation
_Uaar Offaet _Baokgrotmd puae I
_Faat Mode
_Mlnlma Maxima

FASCI >

 

 

 

A.4.4. 1. Execution Screen Display General Description

The above is a reasonable facsimile to the primary execution screen that
user deals with most of the time. The first three lines of words are 'executable'

commands. Each executable command is separated from another by several

A226

spaces. Some of the commands have options that can be attached as well,
these are shown as being options by the braces, {...}. Valid options are the
words within the braces. In some cases one option selection may exclude
another option selection, and these are then grouped inside of the brackets, [...].
An example above is the SAVE command. The data can be saved In an
averaged, AVG, mode, or in a summed, SUM, made. These two selections are
obviously (to me anyways) exclusive and grouped within the same bracket.
Unfortunately, not a terribly great amount of attention was paid to remaining
syntactically similar; hence, GP (Go Plot) options are delimited by a slash, and
SAVE options are delimited by spaces.

The two columns contain a field descriptor, field space, and often
something written in the field space. In cases, where the field descriptor also
doubles as a command (these commands are known as 'configuration
commands'), the command will consist of a two character tag. The two
characters are known by the letters in caps in a descriptor. Thus, the command
for '# Timing Events' is TE<cr>. The command for 'Data Transformation' is
DT<cr>.

Additional information needed will be prompted for. In some cases
(pointed out below) the command expects its argument on the line of the two

character tag.
A.4.4.2. Execution Screen Display Configuration Commands

fast adc resolution. This is not a settable option.
This is a value that reflects the resolution in terms of bits per volts (b/v)
for the fast adc. This value is modified when the ado GAin selection is modified.

slow adc resolution. This is not a settable option.

A227

This is a value that reflects the resolution in terms of bits per volts (b/v)
for the slow ado. As the slow adc has only one gain, this value is not usually
modified during execution of FAS.
SZ - Sample Zero. This is a settable option.
This allows definitions of what raw data point constitutes a ‘time zero.’
The FADC collects 1023 points. if the conversion rate is 1Mhz, 1 point per 1
micro-second (us), and the exciting flash is generated 200 us after the FADC
begins converting then the 'time zero' is 200. If the conversion rate is 1Mhz, 1
points per 2 us, the 'time zero' is still 200. The number for the Sample Zero
refers to the raw data point.
OD - Output Device. This is a settable option.
This allows definition of the physical definition the data is written to when
saved. Unfortunately, when the file is closed, the lun associated becomes
reassigned to SYO:, and this option needs to be reselected each time.
Tl - Title. This is a settable option.
This allows user to select a title to be associated with plots that are
generated, or data files that are saved.
TE - fl Timing Events. This is a settable option.
This allows definitions of the number of events that controlled by FASCI.
if the sequence of events follows something like:
Open Shutter
Trigger, FADC
Trigger, laser
Close Shutter

Then there are only 4 timing events.

Data Buffer. This is a settable option.

A228

This allows user to ‘turn off' the storage data buffer which is used to
dynamically store processed data. When a scan is being iterated, for the
particular sequence being executed, data is read out of the buffer, added to the
data just collected, and written back to the buffer. This data buffer is created
anew at the beginning of each EXECUT command, and is sized to be no larger
than the number of sequences in the current scan. Optionally turning off the data
buffer allows memory to be conserved if needed.

SN - Sequence Number. This is a settable option.

This defines number of sequences in an iteration cycle. This option must
be set BEFORE the timing pulses are configured.

A sequence is the group of pulses defined to be generated when an
experiment is running. it one is looking at steady state flash turn over
experiments then only 1 sequence is required. if one is looking at a four flash
oscillation, and wants an additional four flashes to confirm an oscillatory pattern
then a total of 8 flashes are to be generated before beginning to loop around and
continue averaging. 8 sequences are thus to be chosen far sequence number.

lN - Iteration Number. This is a settable option.

This allows definition of the number of times a sequence group will be
iterated.

SD - Sequence Delay. This is a settable option.

This defines the amount of time (in units of integer seconds) which
elapses between any two sequences when the number of sequences are greater
than one. If there is only one sequence number then the sequence delay should
be set to the iteration delay value. .

Time range is from 3 to 16000 s when Fast Mode is DISABLED. When
Fast Mode is ENABLED then the time range is from 1 to 16000 s; however, there

A229

are some qualifications. Total number of sequences cannot exceed 8; see the
description on Fast Mode for more details.

Chosing delay times between sequences is usually no problem for lower
time delays of 5 s or greater. Between 3 and 5, though, care must be taken in
the configuration of the timing pulses, consideration must be given to the number
of users logged on, etc. Remember that after every sequence 1023 points must
be converted from raw 8 bit binary values to 32 bit floating point numbers in
either transmission or absorption format. Choosing too short of a value is usually
manifested by 'TlMlNG or DATA OVERRUN' error type messages. Meaning that
not enough time is available for processing before more data becomes available.

lD - iteration Delay. This is a settable option.

This defines the amount of time (in units of integer seconds) which
elapses between any two sequences when the number of iterations are grater
than one.

Time range is from 3 to 16000 3.

AD - ADc [ON OFF]. This is a settable option.

This is an obsolete option and is disabled.

DT - Data Transformation. This is a settable option.

This allows user to define how the data is processed. More detail is
given in the next section; though, it is briefly described here. Data can be
transformed into Raw Transmission (summed or averaged Volts) or into Relative
Absorbance (summed or averaged). Raw transmission simply involves
converting the 8 bit binary value into a voltage. Relative absorbance, though,

requires baseline information.

A230

Prompts will appear for the following types of information:
Raw or Transmission
Sample and Hold Active (SHA) or Not
lf SHA is active than get baselineduration (in raw units of points)
Data zeroed or not.

The SHA is used to look a baseline for sampling. The baseline
information is usually from when the conversion is started to just before the laser
goes off. The length of time is given in raw sample points.

UO - User Offset. This is a settable option.

This allows user to define the amount (volts) the baseline has been offset
from 0.0V. This is useful to provide more of a dynamic range, and enhanced
signal-to-noise. if a background voltage (related linearly with voltage) is 10, and
dV is expected to be 0.01 ; then, if the background voltage Is 20, the dV will be
0.02. This results in a doubling of the signal fed to the FADC.

BK - Bachround pulse ft. This is a settable option.

This allows user to specify that the first sequence (implies that Sequence
Number is already set and is in the range [1 ,31]) is a 'background.‘ Prompt for
the pulse it which turns on the excitation. On the first sequence of each iteration
that pulse will be prevented, and the data recorded will be considered a
background.

Specifying this option increases the number of sequences by one, and is
indicated in the Sequence Number field by a '(+)' at the end of the field. This
cannot be set if Sequence Number was not already defined, or the 32 sequences
were already defined.

All data structures which are manipulated by the user, whether it be the

dynamic data holding the last scan or the saved file, will have one more field than

A231

originally specified. In place of the first sequence's y-axis data will be the
background.

FM - Fast Mode. This is a settable option.

This allows user to specify a shorter sequence delay than 3 s for a
limited number of sequences. Minimum time is now 1 s, and still in units of
integer seconds. Achievement of this short of a time delay is by delaying data
processing until all sequences in an iteration are completed. This means that the
Iteration Delay time has to be increased enough to handle the additional time

required for processing.
A.4.4.3. Execution Screen Display Executable Commands

These commands actually cause something to happen. This something
can be anything from running a scan, to plotting the data, to saving it, to
spawning off a command to the system, etc.

]EXIT. I would think this does the obvious - shuts down the interface.

EXECUT. This begins the scan. Some potential error conditions are
caught and this command is aborted; sometimes not, though.

LOAD. Complement of the WRITE option. Allows user to load in a
saved configuration command file. Not all of the variables from the previous
configuration are saved. More than one configuration file can exist.

WRITE. Complement of the LOAD option. This allows user to save
most of the current configuration in a file of choice. The first real data sequence,
thus, is #2. If user wants to plot the second sequence then #2 has to be
specified. User can write more than one configuration.

SETTE. This command causes entry into another menu which allows
user to define the configuration used for the timing pulses. See below,

"Configuration Display Screen for Timing Pulse Generation."

A232

REFRESH. This command refreshes the screen display; occasionally
things get a bit cluttered and/or out of place.

GP. Allows user to generate plots of up to six sequences at a time from
the previous scan. Graphic devices supported are usually some subset of the
MULPLT devices, and can be determined in a like fashion by typing:

GP? <cr>
This command results in the graphic devices supported, the referent number of
the graphic device, and the internal number of the graphics device to be printed.
Supposing that graphic device #1 was a tektronix graphics terminal and that
sequences 1,2,4, and 32 were to be plotted on this graphic device then the
following sequence would be entered:

GP 1, 1,2,4,32 <cr>
The first '1' is the graphic device number, and the following 4 numbers are the
sequences to be plotted. Additional options, in the form of switches, may also be
specified. As mentioned previously, options specified within a pair of "[...]" are
mutually exclusive, and non-exclusive options are delimited from one another by
"/." The examples below are given in the form:

GP{/...}/OPTIONn,m <cr>
or

GP{/...}/OPTION n,m <cr>

The "...{l...}..." refers to other options that may or not be present on the
command line. ..."/OPTION"... is the option being explained. Order of occurrence

in the command line is NOT significant.1

 

1the main difference in the two examples (above) is the separation of the
device and sequence(s) specifications from the body of the command switches.
In the case of where a switch will end with a numeric value it would be difficult
(impossible) to separate out the end of the switch's numeric value from the
beginning of the device number.

A233

[SUM AVG]
GP{I...}IAVG1,2 - plot the average of
' sequence 2 on device
[XSCALE]

GP{I...}IXSCALE:rrr 1,2 - plot the sequence Zon device 1, and
scale the x-axis by rrr.

[YSCALE]

GP{I...}IYSCALEzrrr 1,2,30 - plot the sequences 2 and
30 on devrce 1, and
scale the y-axis by rrr.

[BS] 2
GP{I...}IBS n,m - plot the sequences 2

if "GP?" is typed, the binary file option shows up, but choosing this result
causes the resulting graphics file to be processed in a special way.

SAVE. Allows user to save the results from the last scan. Again options
are available in a switch selectable form. When a scan is saved all of the
sequences in the last scan is saved, and, unlike other options where the user can
specify the input/output file names, the name for the saved scan file Is
determined by FAS.

Often when data files from an instrument are saved the user will come up
with unique names like:

FOO1.DAT
FOOZ.DAT
F003. DAT

and the names may or may not be relevant to the scan(s) completed. FAS

automatically generates a name based on the date and the elapsed time after

 

2Only valid if Bachround pulse if was given.

A234

midnight. This name, therefore, is unique (for a year at any rate), and cannot be
confused with another "FOOx.DAT"-type of filename. The naming and internal
structure of the saved data files are discussed in a later section.

The available switches are:

[SUM AVG]
SAVE/AVG - data in the sequences are saved in an
AVERAGED mode. This is the default.
[33]
SAVE/BS - sequences in the saved data file have the

base-line (as contained in sequence 1)
subtracted. The default is NOT BS.

REPORT. This command causes the results of the last scan to be

printed on the terminal.
.page

A.4.4.4. Configuration Display Screen for Timing Pulse Generation

A screen for displaying the timing pulse configuration is brought up when

the user specified "SETTE."

 

Define Time RETURN
DEVICE DELAY CONDITIONAL

1
2
3
SN

Only two commands exist for this screen (it is entirely possible that the

 

 

top line does not agree with what comes up in FAS, if that's so than what's said

here overrides what the program has displayedl).

A235
A.4.4.5. DT - Define Time

Allows user to define the timing parameters for a single pulse. A prompt

will come back after
DT<cr>
prompt:
Pulse #, Name, Time(s,m,u),{c_enable, beg, inc)
This is asking for the above elements to be given in a single line.

Pulse # - just the pulse being defined. Number cannot be any higher
than # of timing events defined in previous screen.

Name - allows user to associate a name with the pulse, like "laser",
"shutter", "die"...

Time (s,m,u) - specifies the elapsed time (real number) and the units
of the number before the pulse is triggered. A time of 1.0s would
cause a delay of 1 second before the pulse is triggered. A time of
12.5u would cause a delay of 12.5us. And a time of 0.01m would
cause a delay of 0.01 milliseconds. The screen display is updated
after every DT command, and sometimes the value entered is
changed. 0.01m would be (probably) modified to be 10.0u. The
minimum time allowable is 0.25u (250ns), and the maximum is
16.0s (16s).

{c_enable beg,lnc] - this is optional. Allows user to specify a
"conditional pulse." Normal pulses are fired every sequence and
every iteration. Conditional pulses can have their start time and
sequence increment fired every time. A command

DT1,SILL Y, 1.0U,CE_5,6
would cause pulse # 1 to be fired conditionally; first occurrence of

the pulse would be the 5th sequence of each iteration, and the next

A236

occurrence would be after 6 sequences had elapsed (the 12th). A
command

DT10,SILL Y,5.S, CE, 1, 1
would cause pulse #10 to be triggered an the first sequence of each
iteration, wait a sequence, and fire on the 3rd, wait, fire on the

fourth, etc.

A.5. ERRORS AND SUGGESTED RECOVERY METHODS

A.5.1. Fairgre to start UPLODR fully

A.5.1.1. Aborting UPLODR and/or FAS.

The MCR command only allows abortions of NAMED tasks. lf FAS was
invoked on user's terminal, call it 115:, by the user typing 'RUN LB1:[1,54]FAS'
then FAS's running name is 'TTS.‘ If run via CCL then its running name is
'FAST5.’ The appropriate abort commands would be, in order, 'ABORT TTS' and
'ABORT FAST5.‘ UPLODR is installed under the name UPLODR, and can then
be aborted via the same name. This is not a trivial point. KILL.TSK, unless
installed at a higher priority than UPLODR's (nominally 140 or 150) will not be

able to get in a kill these errant processesll).
A.5. 1 .2. Other Cases

Otherwise Normal Error Messages result in unexplainable l00ping. This
behaviour prevents graceful recovery. Some errors can be detected, and the

software does recover somewhat gracefully with appropriate error messages.

A237
A.5.2" interface at computer and not po_wered up.

 

UPLODR detects this, generates an error message that informs the user
this possibility should be checked out, and exits. Once the condition is remedied
(i.e, the power turned on), the user should make sure neither task is active from
the terminal, and attempt to restart.

If power was already on, then either power routing on the board(s) on the
panel has failed, a chip has gone down, or something similar. UPLODR
generates this message when it is unable to clear the DRV-11J input port, so the

interrupt generation circuitry should be checked out first.
A.6. DATA TRANSFORMATION

This is only included because it is important to understand how data is
transformed from 8-bit binary numbers relating to digitized voltage coming out of
the back end of a PMT to difference absorbance values.

Until recently, researchers typically reported optical changes in units of
dA ~~ dI/l ~ dVN

where I is the original light intensity upon the PMT face which is equivalent to the

voltage generated, V. dl is the small change of light intensity induced by the

photochemistry resulting from 9excitation, and this is also equivalent to dV, the

change in voltage coming out the PMT. dA is the change in absorbance.
Formally, though:

dA - log ((Voff + dV)Noff)

where Voff is the voltage detected BEFORE excitation, and dV is the change of
voltage detected from Voff.

A238

The thinking has been that as dVN is so small, 10'3, that it's roughly
equivalent to dA. If treatment of the data warrants this, then a potentially
complicated problem becomes tractable. When signals are small, as they are in
photosynthesis, it is necessary to average many scans, and the equations look
like:

dA (avg) - dA(tot)/N - SUM[dA(i)]lN

- SUM[Iog((Voff(i)+dV(i))Noff(i))]/N

where dA(avg) is the total absorption change averaged over N scans. dA(i) is the
absorption change of the ith scan. SUM[...] represents the summation sign over
the quantity in "[]." Voff(i) is the offset voltage for the ith scan, dV(i) is the small
voltage for the ith scan.

The above, however, is difficult to do in real-time, and requires two

channels over which data is collected. If the approximation :
dA(avg) ~ (SUM[dV(i)]/Voff(avg))/N

can be employed then Voff may be read off from a voltmeter at the beginning and
end of the scan, averaged, and used in the above equation. This is simple, and
only requires one data channel the computer has to worry about in real-time.

If the basic approximation, instead of being:
dA ~ dV(i)Noff
were

dA~dVNoff_-1.0

A239

the results would then be within the correct order of magnitude, and for small (10’
3) changes or smaller, within experimental error. the results would have an error
of 1 0'4.

FASCI has two data channels availableand can perform the full blown
calculation; however, tests run with the 11/23 (the processor FASCI was
developed on) show that the intrinsic log function ran 10 times slower than an
equivalent Taylor series expansion which was accurate to 105. The same
performance ratio holds true for the 11/73 processor, though the latter is so much
faster the intrinsic log function could be put back in. A test case for this is given

in LOGTST1.FTN.
A.7. STRUCTURE OF THE SAVED DATA FILES

This section discusses the philosophy behind the name used for the data

files, and the structure by which they're written to disk.
A.7.1. Data File Names

As mentioned earlier, saved scan sets are given semi-unique names
(unique till the following year at any rate). These names are constructed from the
data and the time after midnight. An extension of ".FDF" (Flash Data File) is
added to further identify the file uniquely as being constructed from FASCI.

Within each data file is a header (see below), and in the header Is a
SCANID field. This SCANID contains the first 9 characters of the filename and
provides a degree of association in that fashion.

Plots which are directed to the line-printer are generated so that a
SCANID is written out on the plot itself, and also on the accompanying parameter

file.

A240

in this fashion data files which are generated (both internally in the
header and externally in the name itself), plots which are also derived from the
same data set by FASCI, and the parameter files which are written out are
associated fairly uniquely, i.e., chance for replication until the following year is

slim.
A.7.; Structgre of the Data File.

The data file has two parts :
Header, 'PL' data tag
DA TA, 'RD' data tag

As can be guessed by now, this file is in the binary MULPLT format. The
header contains information relating to the date and time of the scan, SCANID,
title of the scan (it one was given), configuration of the scan, timing configuration,
number of iterations and sequences, etc. Additionally, and quite important,
information relating as to how the data was saved (summed or averaged, base-
line zeroed or not, background subtraction) is also present and can be pulled out
to be examined.

The data field consists of the data tag, the x-value (in micro-seconds),
and then all of the sequences. It being plotted by MULPLT later, the format
command to access the x data and the first sequence would be :

FM#1,2
To access the x data and the 3rd sequence (if at least that many sequences were
specified):

FM#1,4

If a background subtraction was performed, then the first sequence data
(#2) Is the background shot. The first sequence of data containing the kinetic
data would be #3.

A241
A.8.3. Examination of the data file directly

Provisions have been made to examine separately the header.
BINDMP.TSK (made up from BINDMP.FI'N and BINDMP.CMD, there may be
additional files associated) will dump out the header of the saved scan file (it will

also work with data files collected by PERKIN).
A.8. GRAPHICS DEVICE 0 SELECTION

FASCI has been designed to take advantage of both 1) the RASTER
processing feature (from the MULPLT package), and 2) the batch environment of
RSX-1MPLUS to facilitate generation of graphics-hard copy while not interrupting
real-time experiments.

It is possible to run an experiment, save the data, send a graphics image
to a file, and continue running another experiment immediately. Meanwhile, a
batch job will have been started running at low priority which will reproduce the
plot on the printer specified, print the parameter file, delete the graphics file (fairly
large) and the parameter file (not large, but it gets to be annoying to have a lot of
them around).

This is all done by selecting "GPO,..." in the Go Plot command. Syntax for
specifying the sequences to be plotted are identical to the other GP commands.

A.8.1 . Method

Getting this accomplished is simple. Buried somewhere in FASCI is a
subroutine that gets called. This subroutine executes the command file
"RASOUE.CMD." RASOUE.CMD builds a batch file, "RASTER.BAT" that, after
successfully built, will be submitted to BAPO (the batch processor for RSX-
11M+). When the batch job, controlled by RASTER.BAT, is completed, the result

A242

of the job is written to a file RASTER.LOG. Unfortunately, LOTS of these will
acrue in user's area unless periodically purged.

Several files either have to be present in the user's area, or are
generated by RASOUE.CMD and/or RASTER.BAT (depending on version of
RASOUE.CMD). These files are:

ALLP.CMD - command file which runs under the indirect command
processor at low priority. It's purpose is to allocated the printer to the batch job.
This is to keep other print jobs from intervening between the plot and the
parameter file.

DEALLP.CMD - command file which deallocates the printer.

SEPARATETXT - a file which contains a line of "' to separate one
plotting set (defined by a plot and its parameter printout) from another.

FORMFEED.TXT - a file which contains the <FF> character to force a
formfeed.

"lililliiillll‘iilltilili