The effects of low pH on the properties of protochlorophyllide
oxidoreductase and the organization of prolamellar bodies
of maize (
Zea mays
)
Eva Selstam
1
, Jenny Schelin
1
, Tony Brain
2
and W. Patrick Williams
2
1
Umea
˚
Plant Science Center, Department of Plant Physiology, University of Umea
˚
, Sweden;
2
Life Sciences Division, King’s College,
University of London, UK
Prolamellar bodies (PLB) contain two photochemically
active forms of the enzyme protochlorophyllide oxido-
reductase POR-PChlide
640
and POR-PChlide
650
(the spec-
tral forms of POR-Chlide complexes with absorption
maxima at the indicated wavelengths). Resuspension of

M
.
Comparison of the two sets of changes suggest a close link
between the stability of the POR-PChlide
650
,membrane
organization and NADPH binding.
The low-pH driven spectral changes seen in maize PLB
are shown to be accelerated by adenosine AMP, ADP and
ATP. The significance of this is discussed in terms of current
suggestions of the possible involvement of phosphorylation
(or adenylation) in changes in the aggregational state of the
POR complex.
Keywords: protochlorophyllide oxidoreductase; prolamellar
body; protochlorophyllide; oxidoreductase; chlorophyllide.
Plant prolamellar bodies (PLB) found in the etioplasts of
dark-grown (etiolated) seedlings, are the precursors of the
chloroplast thylakoid membrane. The PLB membrane is
dominated by the presence of a single protein species,
protochlorophyllide oxidoreductase (EC 1.3.1.33) (POR)
that catalyses the light-driven, NADPH-dependent reduc-
tion of protochlorophyllide (PChlide) to chlorophyllide
(Chlide). Analyses of the absorption spectrum of PLB [1]
and low-temperature fluorescence spectra of etioplast inner
membrane preparations (EPIM) and PLB [2], indicate the
presence of three major pools of PChlide; a nonphotocon-
vertible form PChlide
628)633
and two photoconvertible
forms PChlide

shift occurs and the PLB lack the ability to undergo the
compositional and morphological changes seen in vivo.
The relationship between the two photoconvertible forms
of POR has been the subject of much discussion. A number
of lines of evidence suggest that POR-PChlide
640
and POR-
PChlide
650
are the less and more aggregated forms, respect-
ively, of the same enzyme [1,6–8] and Ryberg, Sundqvist and
their coworkers [9–11] have recently reported results
suggesting that this aggregation may be favoured by POR
phosphorylation. The idea of a possible phosphorylation
Correspondence to W. P. Williams, Life Sciences Division, King’s
College London, Franklin-Wilkins Building,
150 Stamford Street, London SE1 9NN.
Fax: + 44 20 7848 4450, Tel.: + 44 20 7848 4433,
E-mail:
Abbreviations: Chlide, chlorophyllide; PChlide, protochlorophyllide;
PLB, prolamellar body; POR, protochlorophyllide oxidoreductase;
POR-PChlide
635
,POR-PChlide
640
, POR-PChlide
650
,POR-
Chlide
676)677

the absorption maximum of the photoconverted enzyme
occurring under these conditions to be inhibited by ATP
and by the protein phosphatase inhibitor NaF. On the basis
of this study, and subsequent studies on the action of
protein kinase and protein phosphate inhibitors [10,11],
POR-PChlide
650
was suggested to be an aggregated form of
a phosphorylated (or possibly adenylated) ternary complex
of POR, its substrate PChlide, and NADPH. Dephospho-
rylation of the photoconverted form of this complex by an
endogenous phosphatase, it was further suggested, leads to
a disaggregation of POR and a dissociation of Chlide from
the POR complex, giving rise to the Shibata shift.
In this paper, the resuspension of maize PLB in media
with a pH below about pH 6.8, is shown to lead to a rapid
conversion of POR-PChlide
650
to POR-PChlide
640
.These
changes, which take place over an extremely narrow pH
range, are shown to be accompanied by marked decreases in
the ability of POR to bind NADPH and a rapid disassembly
of the PLB. The pH-driven spectral changes are compared
to those seen in aged PLB. The effects of adenosine, AMP,
ADP, ATP and NaF on the pH-driven changes are studied
and discussed in terms of their relevance to the POR
phosphorylation model.
MATERIALS AND METHODS

M
MgCl
2
,1m
M
EDTA
and 30 m
M
Hepes adjusted to pH 6.5 or pH 7.5. The
results reported in this paper were normally based on
measurements performed on at least three different PLB
preparations. Minor differences in the rates of the spectral
and structural changes were observed between different
preparations but the overall pattern of changes was
extremely consistent.
Absorption and fluorescence spectrophotometry
Aliquots (50 lL) of freshly thawed samples of PLB
containing  200 lg protein, were thoroughly washed in
cold assay medium at pH 7.5 to remove excess NADPH.
The washed pellet was then re-suspended in 1.0 mL of test
assay medium. Absorption spectra were normally measured
using a Shimadzu MPS 2000 spectrophotometer fitted with
a cuvette holder close to the photomultiplier to reduce light
scattering. A few measurements were made using a Philips
PU8720 spectrophotometer and a computer-generated
baseline used to minimize the effects of light scatter.
Photoconversion was brought about by exposure of the
sample to a defined number of flashes of bright white light
delivered by a Sunpak, Softlite 2000 A (Tocad, Tokyo,
Japan). When required, 40 lL samples were removed for

and a shift in absorption maximum, in the presence of
excess NADPH, to 684 nm. The 77 °K fluorescence emis-
sion spectrum of the nonphotoconverted PLB is dominated
by the 656-nm emission peak of POR-PChlide
650
(Fig. 1B).
Some emission from the nonphotoconvertible species
PChlide
628
is visible at about 630–635 nm but little or no
emission is seen from POR-PChlide
640
, reflecting the
efficient excitation energy transfer existing between this
species and POR-PChlide
650
[13]. Samples photoconverted
in the presence of excess NADPH are characterized by a
maximum at 696 nm typical of the Chlide derivative of
POR-PChlide
650
.
The results for maize PLB resuspended at pH 6.5 are
strikingly different. Under these conditions, there is a rapid
conversion of POR-PChlide
650
to POR-PChlide
640
(Fig. 1C). Exposure to bright white light leads to the
photoconversion of POR-PChlide

ble species POR-PChlide
635
lacking bound NADPH which
rapidly reconverts to POR-PChlide
640
in the presence of
added NADPH. POR-PChlide
635
has a 77 °K fluorescence
emission peak at 640 nm of similar intensity to the 648 nm
emission peak of POR-PChlide
640
(data not shown) clearly
distinguishing it from PChlide
628
, which emits at 633 nm
and remains nonphotoconverted even in the presence of
excess NADPH.
The pH-driven conversion of POR-PChlide
650
to POR-
PChlide
640
(POR-PChlide
635
in the absence of excess
NADPH) is very rapid, taking less than a minute at room
temperature and is complete in less than 10 min even at
0 °C (Fig. 3). As illustrated in Fig. 4, the process is
irreversible. Samples exposed to pH 6.5 were resuspended

characteristic of the low pH form (cf. Figs 2 and 4).
A potential complicating factor in these latter measure-
ments is the tendency of POR-PChlide
635
to break down to
yield a new PChlide absorption band with a maximum at
653 nm (PChlide
653
). Traces of this species are detectable in
the spectra shown in Fig. 4. This breakdown is more clearly
illustrated in Fig. 5, which shows the effects of ageing on the
absorption spectrum of maize PLB suspended in pH 6.5
assay medium in the absence of excess NADPH. PChlide
653
is easily distinguishable from POR-PChlide
650
as it is
nonphotoconvertible and gives rise to no obvious low-
temperature fluorescence. It is probably related to the
species PChlide
647
, attributed to aggregated protochloro-
phyllide, reported in dried etioplast membrane preparations
[19]. Care was taken in all measurements to restrict the time
of exposure of PLB samples to low pH in media lacking
excess NADPH to ensure that PChlide
653
formation was
avoided.
The pH-dependence of the conversion of POR-

flashes of bright white light followed by (e) a 60-s exposure to full room
light.
Fig. 5. Absorption spectra showing the formation of PChlide
653
in PLB
suspended in pH 6.5 assay medium incubated at room temperature for
(a) 22, (b) 37, (c) 46, (d) 67, (e) 80, (f) 100 and (g) 125 min.
Fig. 6. Plots showing the pH dependence of the absorption maxima of
(A) PChlide in nonphotoconverted PLB samples (B) Chlide formed after
photoconversion by two flashes of bright white light measured 2 min after
conversion. Measurements were made in the presence (j)andthe
absence (d)of1.0m
M
NADPH.
Ó FEBS 2002 pH-dependent changes in PLB organization (Eur. J. Biochem. 269) 2339
presented in Fig. 7. At pH 7.5 (Fig. 7A), the majority of the
PLB are in the form of highly ordered paracrystalline arrays
based on networks of interconnected tubular tetrapodal
membrane units forming a bicontinuous diamond cubic
(Fd3m) lattice [21,22]. Resuspension at pH 6.5 (Fig. 7B),
however, leads to complete loss of such ordered structures
and their replacement by highly disordered arrays of
entangled tubes. Parallel X-ray diffraction measurements
(data not shown) confirmed that resuspension of PLB in
low pH media leads to a complete loss of long-range order
in the samples.
This breakdown in PLB structure, like the conversion of
the POR complex, is irreversible. There is no evidence of the
reformation of organized PLB if the pH of the low pH
sample is returned to pH 7.5 by the addition of small

ble POR-PChlide
635
. The higher yield achieved by multiple
flashes, indicates the re-establishment of this equilibrium in
the dark period between flashes. The equilibrium between
the two forms is also reflected in the measurements of the
NADPH dependence of the red-shift in the PChlide
absorption maximum shown in Fig. 8B. In both cases,
half-saturation of these changes occurs at  0.25 m
M
NADPH, indicating that POR-PChlide
640
binds NADPH
comparatively weakly at pH 6.5. Interestingly, even at high
levels of NADPH, a single saturating flash was unable to
photoconvert all the pigment present at low pH. The
reasons for this are unclear.
This contrasts strongly with the binding of NADPH to
POR-PChlide
640
at pH 7.5. Attempts to remove NADPH
from POR-PChlide
640
and/or POR-PChlide
650
by repeated
Fig. 7. Typical electron micrographs of PLB
samples suspended in assay medium at (A)
pH 7.5 and (B) pH 6.5. Magnification bar
200 nm.

by NADPH [4,5]. The samples
were first photoconverted in the absence of excess NADPH
to form the NADP
+
-bound enzyme. The extent of the red-
shift following the addition of different concentrations of
NADPH was then used to estimate the efficiency of
NADPH binding. The results of these measurements,
performed at 5 °C to slow down other possible changes in
the photoconverted enzyme, are shown in Fig. 9. In the
absence of added NADPH, the wavelength of the absorp-
tion maximum measured 10–20 s after photoconversion
was at  680–681 nm falling to  677–676 nm after about
5 min. The full red-shift was obtained even if the addition of
the NADPH was delayed until the wavelength had stabil-
ized at this shorter wavelength, indicating that this initial
decline is not linked to a loss of Chlide (Fig. 9A).
Approximately 2 l
M
NADPH was sufficient to drive the
full shift (Fig. 9B). This approach, unfortunately, cannot be
adopted at low pH as the enzyme is essentially nonphoto-
convertible in the absence of excess NADPH and the red-
shift, if one exists at all, is negligibly small.
Comparison with the effects of ageing
A similar, but much slower, conversion of POR-PChlide
650
to shorter wavelength forms is seen in aged etioplasts and
PLB [14–17]. Maize PLB prewashed in NADPH-free assay
medium (pH 7.5) were aged in the dark at room tempera-

and its subsequent photoconversion to POR-Chlide
684
,are
all clearly visible in the difference spectra shown in
Fig. 10(C,D). These changes are similar to those reported
byBrodersen[17],whoworkedwithagedbarleyPLB.
The NADPH dependence of the reformation process was
estimated from measurements of the amounts of reformed
POR-PChlide
650
available for photoconversion from differ-
ence spectra of the type shown in Fig. 10D. The half-
saturation value for NADPH binding to POR-PChlide
635
at
pH 7.5 estimated on this basis is  1 l
M
(Fig. 11). In
agreement with the findings of Griffiths [16] for water-lysed
etioplasts, we found no requirement for ATP in these
changes. TEM measurements (not shown) indicated a
decrease in the overall degree of order of the PLB with
ageing but no dramatic structural changes of the type seen
on exposure to low pH. Re-addition of NADPH had no
obvious effects on structure.
Adenyl nucleotide and fluoride sensitivity
Ryberg & Sundquist and their coworkers [9–11] have
presented a number of lines of evidence suggesting the
Fig. 9. Changes in the wavelength of the Chlide absorption maximum
following photoconversion of washed PLB samples suspended in pH 7.5

aliquots (70 lL) of POR-PChlide
650
suspended in assay
medium at pH 7.5 to a much larger volume (1 mL) of
pH 6.5 assay medium containing 1 m
M
NADPH, immedi-
ately photoconverting the sample by exposure to a satur-
ating flash of white light and then monitoring the changes in
wavelength of the Chlide absorption maximum. All meas-
urements were performed at 5 °C to reduce the rate of the
pH-driven conversion between the long- and short-wave-
length forms of the enzyme. Additions of NaF, adenyl
nucleotides and adenosine were made 2 min after photo-
conversion to ensure the formation of POR-Chlide
684
prior
to their addition. Measurements were restricted to the
changes seen directly after initial photoconversion as no
reformation of photoconvertible PChlide occurs in the
maize PLB system. There is, however, no obvious reason
why the stability of the reformed pigment complex should
differ from that originally present.
Fig. 10. Regeneration of POR-PChlide
650
in PLB aged in NADPH-
free assay medium at pH 7.5. (A) The initial sample (thin line); aged
sample before (thick line) and after (medium line) exposure to light.
(B) Illuminated sample before (thin line) and after (thick line) dark
incubation with 50 l

(d). 5 m
M
ATP (h), 5 m
M
ADP (s), 5 m
M
AMP (e), or 5 m
M
adenosine (n) in the presence or absence of 10 m
M
NaF as indicated.
2342 E. Selstam et al. (Eur. J. Biochem. 269) Ó FEBS 2002
In contrast to the study on wheat EPIM [9], the addition
of ATP (or adenosine and the other adenyl nucleotides)
accelerated rather than inhibited the blue shift. An inhibi-
tion was observed if both ATP and NaF were present. A
similar inhibition, however, was also observed for ADP,
AMP and adenosine under these conditions indicating that
in maize PLB at least this inhibition is not ATP-specific. In
contrast to the study on wheat EPIM, no significant
difference was seen between the rate of the blue shift in the
presence or absence of NaF alone.
The NADPH-binding efficiency of POR-Chlide
684
,at
low pH is unknown. It is thus hard to disentangle the effects
of a possible loss of bound NADPH (leading to a reversal of
the NADPH-dependent red shift seen in Fig. 11) from those
of a physical dissociation of Chlide and/or conformational
changes associated with the pH-dependent conversion of

(1 m
M
) NADPH in the presence and the absence of 5 m
M
adenosine or the adenyl nucleotides. Little or no difference
was observed, indicating that although they had a marked
effect on the stability of POR-PChlide
650
, they had little
effect on the final level of NADPH binding to POR-
PChlide
640
.
Control measurements indicated that the effects of
adenosine and the adenyl nucleotides were limited to the
low pH range. No significant changes on the absorption
spectra of nonphotoconverted PLB containing POR-
PChlide
650
, or PLB photoconverted in the presence of
excess NADPH to form POR-Chlide
684
, were observed at
pH 7.5. However, as shown in Fig. 13, changes were seen, if
the measurements on the photoconverted enzyme were
made in the absence of excess NADPH. Under these
conditions, the absorption maximum of the freshly photo-
converted Chlide shows the usual decline from  679 to
676–677 nm. Addition of ATP leads to a rapid decrease in
the wavelength to 673–674 nm. If NADPH is then added,

mutant of pea [24], which are both deficient in POR-
PChlide
650
and have been shown to be characterized by a
parallel deficiency in organized PLB.
The ability of the PLB to form a bicontinuous cubic
phase is linked to the high proportion of the nonbilayer
forming lipid monogalactosyldiacylglycerol (MGDG) pre-
sent in the membranes. MGDG normally accounts for
 50 mol% of the membrane lipids in the PLB membrane
[25]. The presence of such a high content of nonbilayer
forming lipid, however, is a necessary but not a sufficient
cause for cubic phase formation. Whilst cubic structures can
be formed in model systems containing high proportions
Fig. 13. Plots of the time dependence of the blue-shift in the absorption
maxima of Chlide following photoconversion of washed PLB suspended
in assay medium pH 7.5 containing no excess NADPH. Samples con-
tained no additions (j), or 5 m
M
adenosine or adenyl nucleotide in the
presence (d)or(s) absence of 10 m
M
NaF.
Ó FEBS 2002 pH-dependent changes in PLB organization (Eur. J. Biochem. 269) 2343
of MGDG, they are stable under only very limited ranges
of composition and hydration [26]. The existence of stable
cubic structure in the PLB is dependent on the membrane
protein content where POR is by far the dominant
component. The mechanism by which POR-PChlide
650

650
, and/or changes in the ionization of groups
associated with POR. These changes then lead to the
destabilization of POR-PChlide
650
and membrane reorgan-
ization.
The suspension of total polar lipid extracts of chloroplast
membranes, which have a very similar lipid composition to
the PLB membrane, in low pH media favours membrane
fusion and formation of nonlamellar structures. The pK
a
for
this process is  pH 4.5 [28], indicating that it reflects the
protonation of the acidic lipids present in the mixture. The
critical pH for the changes reported in the present study is
close to pH 7, suggesting that the initial changes are unlikely
to be directly related to changes in lipid headgroup
ionization.
Our results can be explained by attributing the effects of
low pH to a reduction in the strength of NADPH binding to
POR-PChlide
650
that triggers its relaxation to POR-
PChlide
640
/POR-PChlide
635
, which then destabilizes the
cubic structure of the membrane. This reduced NADPH

650
with no obvious effect on membrane organiza-
tion.
Adenylate-sensitivity of pH effects
POR, in common with many NADPH-binding enzymes,
contains the characteristic motif Gly-X-X-X-Gly-X-Gly
associated with the bA-aA-bB binding domain known as
the Rossmann fold [29]. The detailed organization of the
NADPH-binding pocket has yet to be established for POR
but it has been determined in other members of the short-
chain dehydrogenase/reductase family [30]. In common
with Rossmann folds in general, these sites contain two
mononucleotide binding sites; one for the nicotinamide and
one for the adenosine moiety [31].
2¢-Adenyl nucleotides, of the type found in NADPH, and
the 5¢-nucleotides used in this study, are both known to bind
within the adenosine site of such folds and can act as
Fig. 14. Model illustrating the relationship
between the different POR-PChlide and
POR-Chlide complexes studied in this paper.
2344 E. Selstam et al. (Eur. J. Biochem. 269) Ó FEBS 2002
inhibitors interfering with NAD(P)
+
binding [32,33]. A
possible explanation of the acceleration of the low-pH
driven spectral changes seen for the photoconverted and
nonphotoconverted enzymes on addition of the adenyl
nucleotides or adenosine (Fig. 12) is that these compounds
are able to compete for the NADPH binding site under low
pH conditions destabilizing the aggregated POR-PChlide

POR-PChlide
650
from pre-existing POR-PChlide
635
(P-630),
is solely dependent upon the presence of NADPH and does
not appear to require ATP. This contrasts the earlier studies
of Horton & Leech on aged maize etioplasts [14,15] where
this conversion appeared to be ATP-driven. The possibility
that the preparations employed by Griffiths and ourselves
have lost the putative kinase during the course of prepar-
ation cannot be excluded, but that still leaves unanswered
the question of how the addition of micromolar concentra-
tions of NADPH suffice to drive the conversion of POR-
PChlide
635
to POR-Chlide
650
in its absence. It is noteworthy
that the preparations used in the earlier studies contained
sufficient excess NADPH to allow NADP
+
/NADPH
exchange in the photoconverted enzyme leading to the
formation of POR-Chlide
685
[15] raising the possibility that
these results might reflect an ATP-dependent perturbation
in NADPH binding efficiency rather than a direct POR-
phosphorylation step.

lipids. Again, no phosphorylation step is involved in
pigment loading. The precise role of any possible POR
kinase therefore still remains unclear.
The main line of evidence for the existence of a POR-
phosphatase is the observation by Wiktorsson et al. [9] of an
inhibition, by NaF and ATP, of the loss of the long-
wavelength form of Chlide following photoconversion of
reformed phototransformable PChlide in wheat EPIM. In
agreement with these findings, we observed related inhibi-
tions of the pH-driven blue shift in Chlide and PChlide
absorption in the presence of ATP and NaF (Fig. 12). The
presence of NaF clearly perturbs the PLB system in some
way. However, it must be emphasized that these effects, in
maize PLB at least, are only observed under low-tempera-
ture conditions and when adenosine or adenyl nucleotides
are present. The possibility of other explanations of these
effects cannot therefore be excluded.
CONCLUSIONS
The present study emphasizes the close relationship existing
between the local organization of the PLB membrane and
the stability of different POR-pigment complexes. It dem-
onstrates the extreme sensitivity of these complexes, in PLB
at least, to small changes in pH. It also confirms the central
role of NADPH in the reformation of POR-PChlide
650
in
aged PLB and raises the question that the sensitivity of
spectral changes to the presence of adenyl nucleotides
may reflect their effects on NADPH binding rather than
their effects on specific phosphorylation (or adenylation)

non-irradiated and irradiated prolamellar bodies. Physiol. Plant
85, 659–669.
8. Wiktorsson, B., Engdahl, S., Zhong, L., Bo
¨
ddi, B., Ryberg, M.
& Sundqvist, C. (1993) The effect of cross-linking of the
subunits of NADPH-protochlorophyllide oxidoreductase on the
aggregational state of protochlorophyllide. Photosynthetica 29,
205–218.
9. Wiktorsson, B., Ryberg, M. & Sundqvist, C. (1996) Aggregation
of NADPH-protochlorophyllide oxidoreductase-pigment is
favoured by protein phosphorylation. Plant Physiol. Biochem. 34,
23–34.
10. Ryberg, M., Kovacheva, S. & Sundqvist, C. (1998) Accumulation
of photo-transformable protochlorophyllide is controlled by
protein phosphorylation – a hypothesis. In Photosynthesis:
Mechanisms and Effects. Vol. IV. (Garab, G., ed.) pp. 3197–3202.
Kluwer Academic, the Netherlands.
11. Kovacheva, S., Ryberg, M. & Sundqvist, C. (2000) ATP/ADP and
protein phodphorylation dependence of phototransformable
protochlorophyllide in isolated etioplast membranes. Photosyn.
Res. 64, 127–136.
12. Widell-Wigge, A. & Selstam, E. (1990) Effects of salt wash on the
structure of the prolamellar body membrane and the membrane
binding of NADPH-protochlorophyllide oxidoreductase. Physiol.
Plant 78, 315–323.
13. Kahn, A., Boardman, N. & Thorne, S.W. (1970) Energy
transfer between protochlorophyllide molecules: evidence
for multiple chromophores in the photoactive proto-
chlorophyllide–protein complex in vivo and in vitro. J. Mol. Biol.

Three-dimensional structure of the prolamellar body in squash
etioplasts. Protoplasma 128, 147–156.
22. Williams, W.P., Selstam, E. & Brain, A.P.R. (1998) X-ray dif-
fraction studies of the structural organisation of prolamellar
bodies isolated from Zea mays. FEBS Lett. 422, 252–254.
23. Sperling,U.,Franck,F.,vanCleve,B.,Frick,G.,Apel,K.&
Armstrong, G.A. (1998) Etioplast differentiation in Arabidopsis:
both PORA and PORB restore the prolamellar body and photo-
active protochlorophyllide-F655 to the cop1 photomorphogenic
mutant. Plant Cell 10, 283–296.
24. Seyyedi, M., Timko, M.P. & Sundquist, C. (1999) Proto-
chlorophyllide, NADPH-protochlorophyllide oxidoreductase,
and chlorophyll formation in the lip1mutantofpea.Physiol.
Plant. 106, 344–354.
25. Selstam, E. & Sandelius, A. (1984) A comparison between pro-
lamellar bodies and prothylakoid membranes of etioplasts of
dark-grown wheat concerning lipid and polypetide composition.
Plant Physiol. 76, 1036–1040.
26. Brentel, I., Selstam, E. & Lindblom, G. (1985) Phase equilibria of
mixtures of plant galactolipids. The formation of a bicontinuous
cubic phase. Biochim. Biophys. Acta 812, 816–826.
27. Grevby, C., Engdahl, S., Ryberg, M. & Sundqvist, C. (1989)
Binding-properties of NADPH-protochlorophyllide oxido-
reductase as revealed by detergent and ion treatments of isolated
and immobilized prolamellar bodies. Physiol. Plant. 77, 493–503.
28. Gounaris, K., Sen, A., Brain, A.P.R., Quinn, P.J. & Williams,
W.P. (1983) The formation of non-bilayer structures in total polar
lipid extracts of chloroplast membranes. Biochim. Biophys. Acta
728, 129–139.
29. Wilks, H.M. & Timko, M.P. (1995) A light-dependent com-

and protochlorophyllide-NADPH complexes and their
regeneration after flash illumination in leaves and etioplast mem-
branes of dark-grown wheat. Photosynth Res. 59, 53–61.
36. Klement, H., Oster. U. & Ru
¨
diger, W. (2000) The influence of
glycerol and chloroplast lipids on the spectral shifts of pigments
associated with NADPH: protochlorophyllide oxidoreductase
from Avena sativa L. FEBS Lett. 480, 306–310.
2346 E. Selstam et al. (Eur. J. Biochem. 269) Ó FEBS 2002