Full text of DOI:10.1007/s004250100605

Planta )2001) 214: 180±188
DOI 10.1007/s004250100605
ORIGINAL ARTICLE
Randy Dinkins á M.S. Srinivasa Reddy
Mei Leng á Glenn B. Collins
Overexpression of the Arabidopsis thaliana MinD1 gene alters
chloroplast size and number in transgenic tobacco plants
Received: 23 January 2001 / Accepted: 1 March 2001 / Published online: 30 June 2001
Ó Springer-Verlag 2001
Abstract The Arabidopsis thaliana )L.) Heynh. minD
Introduction
gene )AtMinD1) was isolated and constitutively ex-
pressed in tobacco )Nicotiana tabacum L.) plants using
The prokaryotic origin and endosymbiotic nature of
the CaMV 35S promoter. Confocal and electron-mi-
croscopic analysis of the AtMinD1 transgenic tobacco
lines revealed that the chloroplasts were abnormally
large and fewer in number compared with wild-type
tobacco plants. The abnormal chloroplasts were less
prevalent in guard cells than in mesophyll cells. Chloro-
plast and nuclear gene expression was not signi®cantly
di€erent in AtMinD1-overexpressing plants relative to
wild-type tobacco plants. Chloroplast DNA copy num-
ber was not a€ected, based on the relative level of the
rbcL gene in transgenic plants. Transgenic tobacco plants
constitutively overexpressing AtMinD1 were completely
normal phenotypically with respect to growth and
development, and also displayed normal photosynthetic
electron transport rates. These results show that the
Arabidopsis MinD1 gene also functions in a heterologous
system and con®rm the role of the MinD protein in
regulation of chloroplast division.
chloroplasts and mitochondria in plant cells is now an
accepted hypothesis. In higher plants, the chloroplast
genome encodes roughly 130 gene products, most of
which are involved in photosynthetic electron transport,
chloroplast DNA replication, transcription and trans-
lation. The nuclear genome encodes many of the pro-
teins involved in the photosynthetic apparatus and
controls many aspects of chloroplast gene expression
)Mullet 1988; Leon et al. 1998). Nuclear genes also
control whether the proplastid develops into a mature
photosynthesizing chloroplast or into a tissue-speci®c
organelle such as a chromoplast, etioplast or leucoplast
)Link 1991; Herrmann et al. 1992). Most cells maintain a
fairly constant number of chloroplasts based on cell
volume, although the molecular mechanism that con-
trols chloroplast number and division is not presently
understood )Pyke 1997, 1999). Single nuclear-recessive
mutants of Arabidopsis thaliana that a€ect number and
replication of chloroplasts )termed arc mutants) have
been isolated )Pyke and Leech 1992). Depending on the
mutation, the lesion results in many small chloroplasts
or a few large chloroplasts per cell. Although the gene
products of these arc mutants have not been described,
their presence suggests a role for nuclear control of
chloroplast division.
The proposed prokaryotic origin of the chloroplast
suggests that mechanisms for chloroplast division might
be similar to that of bacterial systems. The ®ndings of
Osteryoung et al. )1998) and, more recently, Colleti et al.
)2000) support this hypothesis. Osteryoung et al. )1998)
have shown that an A. thaliana homologue of the
ancestral tubulin cytoskeletal protein FtsZ )filament-
forming temperature sensitive), that is required in
bacterial cell division, is also required for chloroplast
division. In Escherichia coli, in addition to the FtsZ
protein, over 30 other gene products are required for cell
division )reviewed in: Bramhill 1997; Lutkenhaus and
Keywords Chloroplast division á
Bacterial cell division á AtMinD1 á Nicotiana
)chloroplast division) á Transgenic tobacco
Abbreviations 3D: three dimensional á F0, Fm, Fv: initial,
maximal and variable chlorophyll a ¯uorescence,
respectively á PCR: polymerase chain reaction á PSI,
PSII: photosystems I and II
The ®rst two authors contributed equally to this research
R. Dinkins )&) á M.S.S. Reddy á M. Leng á G.B. Collins
Department of Agronomy,
N109 Agriculture Science Center Building North,
University of Kentucky, Lexington,
KY 40546-0091, USA
E-mail: rddink1@pop.uky.edu
Fax: +1-859-2577125

181
Schardl et al. )1987). Initial experiments were done with T0 plants
that were transferred to the greenhouse. Self-fertilized seeds from
individual T0 plants were harvested and 10 T1 plants were assayed.
Addinall 1997; Roth®eld et al. 1999). The function of
most, but not all, of the bacterial genes has been eluci-
dated, and most of these gene products are components
of the mid-cell FtsZ ring or components of cell wall
synthesis in the septal assemblage.
Chlorophyll and photosynthetic electron transport measurements
Genes that regulate the topological speci®city deter-
mining the placement of cell division in E. coli are en-
coded by the minicell )min) locus, the MinC, MinD and
MinE genes )de Boer et al. 1989). Deletion or incorrect
expression of the Min gene products results in placement
of the FtsZ ring at the wrong location, giving rise to
small minicells that lack chromosomes, or it inhibits
septation altogether, resulting in very long ®lamentous
cells. The MinD protein, a membrane ATPase, activates
MinC to inhibit cell division and is also required to in-
teract with MinE to specify the potential division sites
)de Boer et al. 1991, 1992). The mechanisms for the
MinD-MinE and MinC-MinD interactions in E. coli
have not been determined. Min genes have been found in
many bacteria and photosynthetic cyanobacteria in-
cluding Synechocystis. Postulated MinD homologues
have been identi®ed in the Chlorella plastid genome and
in the higher-plant Arabidopsis and rice sequence data-
bases. Recently Colletti et al. )2000) and Kanamaru et al.
)2000) have shown that the Arabidopsis MinD homo-
logue )named AtMinD1) has a role in chloroplast divi-
sion. In this report we show that overexpression of the
AtMinD1 gene in tobacco plants results in fewer chlo-
roplasts per cell and abnormal chloroplast morphology.
These results con®rm that the MinD protein is con-
served in higher plants and has a vital role in chloroplast
division.
Photosynthetic activity was measured via ¯uorescence using an
OS5-FL modulated chlorophyll ¯uorometer )Opti-Sciences,
Tyngsboro, Mass., USA). The initial modulation intensity was set
at 35 lmol quanta m±2 to calculate initial ¯uorescence )F0) and
actinic light )200 lmol quanta m±2) was run for 20 s to calculate the
ratio of variable to maximal chlorophyll a ¯uorescence )Fv/Fm).
Chlorophyll a and chlorophyll b were calculated as in Arnon
)1949).
Microscopy
Transmission electron microscopy
Leaf samples from greenhouse-grown AtMinD1 transgenic and
wild-type tobacco plants were trimmed into 1´1 mm2 squares, then
pre-®xed in 3.5% glutaraldehyde in 0.1 M Sorensons bu€er
)pH 6.8) for 2 h at room temperature. Samples were washed three
times with 0.1 M Sorensons bu€er )pH 6.8), and then post-®xed in
1% osmium tetroxide in the same bu€er for 2 h at room temper-
ature. Samples were rinsed in deionized water, then dehydrated
through an ethanol series, followed by treatment with Propylene
oxide, and in®ltrated and embedded in Spurr's resin. Polymeriza-
tion took place overnight at 70 °C. Ultrathin sections )approxi-
mately 0.07 lm) were cut with a diamond knife on a Reichert
Ultracut E ultra-microtome, recovered on a grid, and stained with
lead citrate. Stained sections were viewed in a Hitachi H-7000
transmission electron microscope at 75 kV.
Confocal scanning laser microscopy
Leaf samples from greenhouse-grown AtMinD1 transgenic and
wild-type plants were cut from fully developed mature leaves, and
all the major veins were removed. The leaf samples were directly
mounted on a microscope slide in a bu€er consisting of 90%
glycerol and 10% 0.05 M sodium phosphate )pH 6.0) to aid in
sample immobilization.
Confocal microscopy was performed with a Leica TCS )Leica
Microsystems Heidelberg, Heidelberg, Germany). The objective
lens was a Leica 100´ oil immersion with a numerical aperture of
1.2 and a working distance of 100±300 lm. An argon laser was
used to excite the chlorophyll molecules in the leaf at a wavelength
of 458‹20 nm. Two sets of ®lters were used to collect the data at
510±475 nm )to visualize plant cell wall phenolic compounds) and
greater than 620 nm )to monitor chlorophyll ¯uorescence). Sixteen
equally spaced optical slices were collected through the abaxial leaf
orientation, with the total distance through the leaf )z axis) for each
sample indicated in the ®gure legend.
The Leica TCS PowerScan software, run on a Windows NT
operating system, was used for imaging. The images were collected
using the medium speed setting with a 512´512-pixel resolution
with four passes through each of the 16 sections. Fluorescence in-
tensity was digitally coded using 256 levels of gray, with 0 repre-
senting the lowest intensity )black) and 255 the highest intensity
)white). The three-dimensional )3D) image was created using the
`extended focus' on the software program to view the apparent
depth, giving a false 3D image.
Materials and methods
DNA manipulations
Oligonucleotide primers were synthesized based on the sequence
found on BAC MZF18.5 on chromosome 5 [5¢ Forward: TTCT-
CGAGAATGGCGTCTCTGAGATTGTTC; 3¢ Back: TTCTAG-
ATTTGCCATTTAGCCGCCAAAG]. The primers were synthe-
sized to include the ATG start site and TAA stop codon )bold
above) with an addition of an XhoI and XbaI restriction
endonuclease site at the 5¢ and 3¢ ends, respectively. Total DNA
from the Arabidopsis thaliana )L.) Heynh. ecotype Columbia was
used for ampli®cation in a standard polymerase chain reaction
)PCR) reaction bu€er [20 mM Tris-HCl )pH 8.0), 2.0 mM MgCl2,
0.25 mM of each dNTP, 100 ng of each primer, 2 units of Taq
DNA polymerase )Gibco/BRL, Rockville, Md., USA)]. The DNA
fragment was cloned into pGEM T± )Promega, Madison, Wis.,
USA) and veri®ed by sequencing. The XhoI-XbaI AtMinD1 gene
fragment was excised from pGEM and cloned into the XhoI and
XbaI sites of the KYLX71:35S2 vector )Maiti et al. 1993). All E. coli
manipulations were carried out in the strain TB1. The
pKYLX71:AtMinD1 recombinant plasmid was mobilized into
Agrobacterium tumefaciens C58C1:pGV3850 by tri-parental mating
)Schardl et al. 1987).
Southern and RNA blots
Plant tissue culture
Tobacco genomic DNA was isolated by homogenizing 100 mg leaf
tissue in a 1.5-ml microfuge tube containing 500 ll of extraction
Nicotiana tabacum L. cv. KY160 was used for all plant transfor- bu€er [100 mM Tris HCl )pH 8.0), 20 mM EDTA, 0.5 M NaCl,
mation experiments. Transformation protocols and media are as in 0.5% )w/v) SDS, 0.5% )v/v) b-mercaptoethanol). The ground

182
extract was treated with 500 ll of a phenol:chloroform:isoamyl
alcohol mixture )25:24:1, by vol.) and centrifuged at 12,000 g for
5 min. The aqueous phase was collected, mixed with 1 ll of
10 mg/ml RNase A and incubated at room temperature for
20 min. The samples were re-extracted with an equal volume of
phenol:chloroform:isoamyl alcohol )25:24:1, by vol.), followed by
two re-extractions with chloroform:isoamyl alcohol )24:1, v/v).
DNA was precipitated by adding 2.5 vol. of ethanol and spooled-
out. The spooled DNA was washed in 70% ethanol, dried, and
resuspended in 100 ll of sterile water.
Putative transgenic plants were screened for the presence of the
transgenes by Southern blot hybridization according to Sambrook
et al. )1989). Five micrograms of DNA from each plant was
digested overnight with XhoI, separated on a 0.8% agarose gel and
blotted onto `Zetaprobe' membrane )BioRad Laboratories,
Hercules, Calif., USA). XhoI cuts only once within the T-DNA.
Random-primed a-[32P]dCTP-labelled probes were prepared using
the Prime-It II Random Primer Labeling Kit )Stratagene, La Jolla,
Calif., USA). Hybridizations were carried out in a solution of
formamide:White Rain moisturizing formula hair shampoo )The
Gillette Company, Boston, Mass., USA):deionized water )5:4:1, by
vol.) overnight at 42 °C. The membrane was then washed three
times at room temperature in 0.2´SSC and 0.1% SDS and was
exposed in a phosphorimager cassette )Molecular Dynamics,
Sunnyvale, Calif., USA). Additional washings at higher tempera-
tures were carried out as necessary.
line. Chlorophyll analysis suggested that several of the
lines had less total chlorophyll, and that the ratios of
chlorophyll a to chlorophyll b were slightly reduced
)Table 1 and data not shown). Photosynthetic electron
transport rates measured by ¯uorescence kinetics were
slightly lower in some of the lines, but overall the values
for F0, Fm and Fv/Fm were not signi®cantly di€erent
from those of the wild type KY160 in most of the lines
)Table 1). Only three lines )8, 9 and 10) showed a sig-
ni®cant reduction in Fm but showed no di€erences in F0
or overall kinetics )Fv/Fm). Thus, while there may be
fewer functional photosystem II )PSII) centers in these
lines, photosynthetic electron transport is not limiting.
The abnormal chloroplast morphology has co-segre-
gated with the kanamycin resistance in all the AtMinD
lines for three generations of self-pollination, indicating
that the traits are linked. The abnormal chloroplast
phenotype can readily be observed shortly following
germination in the green cells in the cotyledons of the
seedlings )not shown).
Total RNA was isolated from 100 mg of fully expanded leaves
at the same time of day from greenhouse-grown plants of AtMinD1
tobacco using 1 ml of Trizol reagent )Gibco/BRL). Ten micro-
grams of total RNA was separated on 1.0% formaldehyde con-
taining agarose gels and transferred onto a Zetaprobe membrane
)BioRad). Hybridization and washing were done essentially as
described above for Southern blot hybridization.
Chloroplast morphology and ultrastructure
Confocal microscopic analysis con®rmed that the chlo-
roplasts in the overexpressing lines were abnormal. A
representative sample of the phenotypes observed is
presented in Fig. 1. Images were observed through both
the adaxial and abaxial orientations and the chloroplast
morphology was similar in both the palisade and spongy
mesophyll cells. Due to the compactness of the palisade
layer cells on the adaxial side and the presence of more
trichomes on the adaxial surface, which in tobacco
contain auto¯uorescing compounds that make imaging
more dicult, the abaxial leaf orientation was used in all
confocal imaging.
Most of the AtMinD lines contained only one or a few
large chloroplasts per cell. The degree of abnormality in
chloroplast morphology correlated to the level of
expression of the AtMinD1 gene. In most of the AtMinD1
transgenic lines, the chloroplasts in the guard cells
appeared to be somewhat larger than normal in size but
were normal with respect to their number )Fig. 1B, C).
Some of the transgenic lines contained guard cells with
both normal and abnormal chloroplasts. Only the lines
Results
In an e€ort to establish the role for the AtMinD1 gene in
tobacco, we isolated the Arabidopsis thaliana MinD1
gene by PCR and cloned it into the pKYLX71 binary
vector containing the CaMV 35S2 promoter for con-
stitutive overexpression )Maiti et al. 1993). The
KYLX71:AtMinD1 plasmid was introduced into to-
bacco by Agrobacterium tumefaciens-mediated transfor-
mation, and kanamycin-resistant regenerated plants
were grown to maturity in the greenhouse. At the whole-
plant level, all of the AtMinD transgenic tobacco lines
appeared essentially normal with respect to plant size
and fertility under greenhouse conditions. Some of the
AtMinD1-overexpressing tobacco plants were slightly
pale and grew slower than the normal KY160 parental
Table 1 Chlorophyll and
Line
Chlorophyll content
Fluorescence measurements
¯uorescence measurements of
wild-type )WT) and AtMinD1-
overexpressing tobacco )Nicoti-
ana tabacum) plants.
Total Chl
)lg/mg)
Chl a/b
)Ratio)
Fo
)Relative units)
Fm
)Relative units)
Fv/Fm
)Relative units)
Chlorophyll measurements are
averaged from four samples,
while ¯uorescence measure-
ments are averaged from six to
eight plants
WT
1.90‹0.09
1.74‹0.08
1.62‹0.05
1.51‹0.07
1.66‹0.08
1.58‹0.06
1.63‹0.04
1.50‹0.08
3.11
2.64
3.01
3.07
3.00
2.95
3.07
2.71
137.4‹12.0
135.7‹11.8
136.5‹17.1
128.5‹32.3
125.5‹19.9
136.5‹11.3
113.6‹17.2
139.5‹20.6
616.6‹34.0
636.2‹27.1
534.9‹66.1
489.9‹78.6
520.5‹58.3
543.1‹14.3
444.1‹58.2
564.9‹32.7
0.777‹0.015
0.787‹0.017
0.757‹0.020
0.741‹0.037
0.759‹0.018
0.748‹0.025
0.743‹0.017
0.756‹0.032
AtMin1 4
AtMin1 5
AtMin1 8
AtMin1 9
AtMin1 10
AtMin1 14
AtMin1 17

183
Fig. 1A±D Confocal microsco-
py of wild-type and AtMinD1-
overexpressing tobacco
)Nicotiana tabacum) lines.
A Wild type KY160, B At-
MinD1 line 5, C AtMinD1 line
14, D AtMinD1 line 17. Each
plate is a composite of 16
frames using the extended focus
from the Leica TCS PowerScan
software to give a false 3D
image. All images were taken
using a 100´ objective lens
[512´512 pixels )x-y axis) which
equals 100 lm]. The z axis was
composed of 16 frames equally
spaced through 52.5 lm )A),
31.5 lm )B), 49.5 lm )C) and
31.5 lm )D) of the abaxial leaf
surface. St Stomatal guard cells.
Bar = 10 lm
with high AtMinD1 expression had all the guard cells with enzyme XhoI that cuts once within the T-DNA generates
abnormal chloroplasts )Fig. 1D). Thus the guard cells bands that are indicative of the number of AtMinD1 in-
appear to be less a€ected than the mesophyll cells by the serts into the genome. Most of the lines had a single copy
CaMV 35S constitutive expression of the AtMinD gene.
of the T-DNA, one line had two copies )AtMinD1 line 10)
In order to evaluate the integrity of the thylakoid and two lines had multiple copies )AtMinD1 lines 3 and 5).
membranes in the overexpressing lines, transmission The moderate-stringency wash showed that the AtMinD1
electron microscopy was performed )Fig. 2). The conti- gene did not cross-hybridize with a tobacco MinD ho-
nuity of the thylakoid membranes could be readily mologue. Southern blot hybridization was also done us-
observed throughout the entire chloroplast in all of the ing the rbcL gene as a probe and no di€erences other than
overexpressing lines )Fig. 2B). The size of the chloroplast DNA loading were observed )Fig. 3). This suggests that
does not a€ect the thylakoid membrane structure, as it while chloroplast division is not normal in the transgenic
appears normal with respect to the presence of both plants, the AtMinD1 overexpression has no e€ect on
grana and stromal lamellae, which overlap continually chloroplast DNA replication.
throughout the entire length of the chloroplast. Although
an extensive survey was not conducted, the few mito-
chondria that were seen in the electron micrographs were Chloroplast gene expression
normal irrespective of the AtMinD1 expression level.
The number of transgene insertions did not appear to be
a factor in the expression of the AtMinD1 transgene in
Southern blot analysis
the tobacco lines, except for AtMinD1 line 5, where
multiple inserts were present, yet AtMinD1 line 5 had the
A Southern blot analysis was done to determine copy lowest expression level )Fig. 4). This was probably due
number of the inserted AtMinD1 transgene )Fig. 3). to gene silencing of the AtMinD transgenes. A correlation
Digestion of total genomic DNA with the restriction was observed between AtMinD1 expression and the

184
Fig. 2A, B Transmission
electron microscopy of
wild-type tobacco and
AtMinD1-overexpressing
tobacco lines. A Wild type
KY160, B AtMinD1 line 14.
Bar = 2 lm
number and size of the abnormal chloroplasts. The either normal, or in some cases, both normal and
AtMinD1 line 5 had the lowest transgene expression, and abnormal. Two AtMinD1-expressing lines )6 and 17)
most of the cells contained normal chloroplasts )Fig. 1B, had no normal chloroplasts in either the mesophyll or
and data not shown). Some of the cells of this line guard cells )Fig. 1D, and data not shown) and appeared
contained abnormal chloroplasts, but these abnormali- to have the highest expression of the AtMinD1.
ties were the exception rather than the rule. Most of the
Variations were observed in the steady-state RNA
AtMinD1 lines had abnormal chloroplasts in the meso- levels of several nuclear-encoded chloroplast-directed
phyll cells, but the chloroplasts in the guard cells were gene products )cab8, cab10 and rbcS) and chloroplast

185
Fig. 3 Southern blot of wild-type tobacco and AtMinD1-overex-
pressing tobacco lines. W Wild type KY160 )two di€erent lines).
Each number corresponds to the AtMinD1 tobacco line described
in the text. Probes indicated are the tobacco rbcL gene )PCR-
ampli®ed from genomic DNA) and Arabidopsis thaliana MinD1
gene )described in the text)
Fig. 4 RNA blot of wild type and AtMinD1-overexpressing
tobacco lines. W Wild type KY160 )two di€erent lines). Each
number corresponds to the AtMinD1 tobacco line described in the
text. Probes indicated are: minD, the A. thaliana MinD1 gene
described in the text; cab8, a 550-bp internal HindIII fragment of
the tomato PSI chlorophyll a/b-binding protein gene 8 )Schwarz
and Pichersky 1990); cab10B, a 700-bp internal PstI-XbaI fragment
of the tomato PSII chlorophyll a/b-binding protein )CP24) gene
10B )Pichersky et al. 1989); rbcS, the small subunit of the spinach
ribulose bisphosphate carboxylase )gift from R. Houtz, USA); 28S,
the 28S ribosomal RNA gene from maize; rbcL, PCR-ampli®ed
large subunit of ribulose bisphosphate carboxylase from tobacco
chloroplast; psbA, PCR-ampli®ed PSII subunit D1 from tobacco
chloroplast DNA; psbB, PCR-ampli®ed PSII chlorophyll a-binding
47-kDa protein from tobacco chloroplast DNA; psbD, PCR-
ampli®ed PSII subunit D2 from tobacco chloroplast DNA. The
PCR-ampli®ed probes encompass the entire coding region of the
gene. Each gene was sequenced using the PCR primers prior to
operons )atpA, atpB/E, psaB, psbA, psbB, psbD, petA
and rbcL), but there were no apparent or consistent
di€erences seen due to the overexpression of the At-
MinD1 transgene in these plants )Fig. 4, and data not
shown). The uniformity of the 28S ribosomal gene for all
the lines in Fig. 4 demonstrated that fairly equal RNA
loading was done. The transcripts that displayed the
most variability in the AtMinD1 tobacco lines were
rbcL, psbA and psbB. More than a 2-fold di€erence was
observed in the rbcL gene expression, with the AtMinD1
lines 4, 6, 7 and 11 displaying the highest levels. The
transgenic AtMinD1 line 6 had one of the highest levels
of AtMinD1 transgene expression, whereas the other labeling
high-AtMinD1-expressing line, AtMinD1 #17, had only
a moderately high level of rbcL transcript. High levels of
the psbB transcript were observed in AtMinD1 lines 5, 7
and 11. AtMinD1 tobacco lines 7 and 11 had a high rbcL Discussion
transcript level; however, AtMinD1 line 6 also exhibited
high rbcL transcript levels but had low levels of the psbB Since the chloroplast is presumed to be of prokaryotic
transcript. AtMinD1 line 5 possessed one of the highest origin )McFadden 2001), it has been postulated that
levels of psbB expression, yet had the lowest AtMinD1 components of chloroplast division could be similar to
transgene expression of any of the lines evaluated. Thus, those of bacteria. The results of Osteryoung et al. )1998)
while variation was observed in chloroplast transcript showed that the plant FtsZ homologue was required for
levels in the di€erent AtMinD1 tobacco plants, there was chloroplast division, demonstrating the similarity to
no discernable pattern in transcript level with respect to bacterial division mechanisms. Recently, it has been
AtMinD1 transgene expression. No hybridization was shown that the MinD gene product is also required for
observed in the lane that contained wild-type tobacco normal chloroplast development in A. thaliana )Colletti
RNA when the Arabidopsis MinD1 gene was used as a et al. 2000; Kanamaru et al. 2000). Our research result
probe, con®rming that the AtMinD1 gene does not demonstrated that the A. thaliana MinD1 gene disrupts
cross-hybridize with a tobacco homologue.
normal chloroplast division when overexpressed in

186
tobacco. Thus, the MinD gene appears to be one of the understood and are somewhat contradictory to what
hinge mechanisms for correct chloroplast division across would be expected from electron micrographs. Confocal
di€erent plant species. In this experiment, we overex- images are from biologically active chloroplasts, while
pressed the AtMinD1 in tobacco rather than in Arabid- electron microscopy requires ®xation, embedding and
opsis, as endogenous transgene overexpression may staining that may not accurately re¯ect what is going on
silence both the trans- and endogenous genes due to in vivo. Although the results obtained from the confocal
the phenomenon of co-suppression )Depicker and van images of the AtMinD1 overexpressing lines are not
Montagu 1997; Smyth 1999). Initial Southern hybrid- completely understood at this time, these images con®rm
ization results )similar to Fig. 3) showed that AtMinD1 that the chloroplasts in the AtMinD lines were not
did not cross-hybridize to its counterpart in tobacco, dividing and in some cases the thylakoid membranes
indicating the lack of homology at the nucleotide level. extended the entire circumference of the cell.
Moreover, overexpression of AtMinD1 resulted in inhi-
In spite of the di€erences in chloroplast morphology,
bition of septum formation, leading to ®lamentous cells there did not appear to be any signi®cant di€erences in
in E. coli )data not shown). These results show that the photosynthetic electron transport )Table 1) nor any
function of this gene is conserved among di€erent plant consistent di€erences in chloroplast gene expression or
species, and between bacteria and higher plants.
nuclear gene expression of chloroplast-directed proteins
Variations in the number of chloroplasts per cell were in the overexpressing plants )Fig. 4). None of the
observed in some of the transgenic tobacco lines. Most di€erences observed could be attributed to higher or
of the lines had few large chloroplasts per cell compared lower AtMinD1 RNA levels in the transgenic lines. It
with normal tobacco control plants. Some of the varia- has been previously shown that much of the regulation
tion appeared to be due to the level of AtMinD1 trans- of chloroplast genes occurs post-transcriptionally
gene expression in the lines, and a correlation was )Gruissem et al. 1988; Rochaix 2001). We have not been
observed between the AtMinD1 RNA level and number able to look at the protein level of these gene products
of abnormal chloroplasts in the overexpressing plants. due to the lack of antisera, but it will be interesting to
This was especially evident in the number and size of the determine if any changes can be detected in accumula-
chloroplasts in the guard cells. The mesophyll cells for tion of the proteins. No di€erences were seen in the
the most part contained abnormally large chloroplasts in number of chloroplast DNA molecules, as measured by
all the overexpressing lines except for AtMinD1 line 5, the relative level of the rbcL gene )Fig. 3). These results
which contained both normal and abnormal chloro- suggest that chloroplast DNA replication was normal
plasts )Fig. 1B). This line also had the lowest level of even though chloroplast division was inhibited in the
AtMinD1 expression, probably a result of silencing due AtMinD1 transgenic plants.
to multiple copies of the transgene )Fig. 4). Most of the
No cross-hybridization was observed with a tobacco
AtMinD1 tobacco plants had abnormal mesophyll cell MinD homologue when using the Arabidopsis gene as a
chloroplasts and normal guard-cell chloroplasts or in probe. This was somewhat surprising since the washes
some cases the guard-cell chloroplasts were both were done at moderately low stringency, and there
abnormal and normal. In the high-expressing lines, appears to be a high degree of conservation at the amino
all mesophyll and guard cells had large abnormal acid level across diverse species )Kanamaru et al. 2000).
chloroplasts )Fig. 1D). Although we have not seen We have not attempted to use lower stringency to detect
documentation of CaMV 35S-promoter-driven di€er- the tobacco gene. The RNA blots also did not detect a
ential expression in the guard cells compared to meso- tobacco homologue. However, this can be explained by
phyll cells, variation in expression in di€erent cell types the probability that endogenous MinD transcripts are in
has been shown )Benfey and Chua 1990).
Electron microscopy revealed that the thylakoid expressed transgene.
low abundance compared to the CaMV 35S2-promoter-
membranes were continuous and normal throughout the
The discovery of the role of the FtsZ protein in
chloroplast in the AtMinD1 tobacco plants. Confocal chloroplast division has suggested a mechanism similar
microscopy of the live tissue corroborated the normal to that in bacteria )Osteryoung et al. 1998; Strepp et al.
functioning of the AtMinD1 tobacco chloroplasts as 1998). The bacterial model for higher-plant chloroplast
chlorophyll ¯uorescence was used to view chloroplast division is no doubt an oversimpli®cation. Osteryoung
structure. None of the plants displayed the very high et al. )1998) have demonstrated that the FtsZ proteins
¯uorescence that is usually observed when photosyn- are encoded by a small gene family consisting of two
thetic electron transport is abnormal or disrupted )Miles forms, FtsZ1 and FtsZ2, with both required for chlo-
1994). In the tobacco AtMinD1 lines, the chlorophyll roplast division. The FtsZ1 protein product was found
¯uorescence appears as sheets layered on top of each to be targeted into the chloroplast and probably
other throughout the length of the entire chloroplast. provides a similar function to the bacterial FtsZ protein
This type of ¯uorescence is more commonly observed in in E. coli. The FtsZ2 gene product does not appear to
the agranal bundle-sheath chloroplasts of C4 plants such have a chloroplast targeting sequence and is speculated
as maize or sorghum )Mehta et al. 1999). The results to function on the outer envelope in chloroplast division
obtained from studies using confocal microscopy to )Osteryoung et al. 1998). Thus, it appears that two
study chloroplast ¯uorescence are not yet completely components of the Z ring are needed in higher plants for

187
correct chloroplast division to occur. Plant cells have a
dual membrane system that separates the chloroplast,
allowing additional nuclear factors to be postulated.
Many algae, including the model photosynthetic or-
ganism Chlamydomonas reinhardtii, have only a single
chloroplast within the cell. Single chloroplasts per cell
are the exception rather than the rule and not a feature
in higher plants. However, the genes controlling chlo-
roplast division in organisms with a single chloroplast
per cell, such as Chlamydomonas, have not been deter-
mined. No Min genes have been documented in the
databases from Chlamydomonas to date; however, in
order to maintain a chloroplast in each cell, chloroplast
division must occur. At no time was sectoring observed
in the developing seedlings of the AtMinD lines, indi-
cating that all of the cells contained functional chloro-
plasts in spite of their large size. The mechanism for
partitioning during cell division under the conditions
References
Arnon DI )1949) Copper enzyme in isolated chloroplasts:
polyphenol oxidase in Beta vulgaris. Plant Physiol 24:1±15
Benfey PN, Chua N-H )1990) The cauli¯ower mosaic virus 35S
promoter: combinatorial regulation of transcription in plants.
Science 250:959±966
Bramhill D )1997) Bacterial cell division. Annu Rev Cell Dev Biol
13:395±424
de Boer PAJ, Crossley RE, Roth®eld LI )1989) A division inhibitor
and topological speci®city factor coded for by the minicell locus
determine proper placement of the division septum in E. coli.
Cell 56:641±649
de Boer PAJ, Crossley RE, Hand AR, Roth®eld LI )1991) The
MinD protein is a membrane ATPase required for the correct
placement of the Escherichia coli division site. EMBO
J 10:4371±4380
de Boer PAJ, Crossley RE, Hand AR, Roth®eld LI )1992) Roles
of MinC and MinD in the site-speci®c septation block medi-
ated by the MinCDE system of Escherichia coli. J Bacteriol
174:63±70
where a single large chloroplast is present is not known. Colleti KS, Tattersall EA, Pyke KA, Froelich JE, Stokes KD,
Osteryoung KW )2000) A homologue of the bacterial cell di-
vision site-determining factor MinD mediates placement of the
chloroplast division apparatus. Curr Biol 10:507±516
Depicker A, van Montagu M )1997) Post-transcriptional gene
Pyke et al. )1994) have speculated that in the arc6
mutant of Arabidopsis, which has only one or two large
chloroplasts per cell, chloroplast segregation is due to
breakage and re-annealing during cytokinesis. However,
it is also possible that additional nuclear gene products
direct chloroplast division during cell division, ensuring
that chloroplast segregation is accomplished. Analysis of
chloroplast-division genes in organisms with only a
single chloroplast, such as Chlamydomonas, should aid
in the analysis of this question.
silencing in plants. Curr Opin Cell Biol 9:373±382
Gruissem W, Barkan A, Deng X, Stern D )1988) Transcriptional
and post-transcriptional control of plastid mRNA levels in
higher plants. Trends Genet 4:258±263
Herrmann RG, Westho€ P, Link G )1992) Biogenesis of plastids
in higher plants. In: Herrmann RG )ed) Plant gene rese-
arch, cell organelles. Springer, Berlin Heidelberg New York,
pp 276±349
Kanamaru K, Fujiwara M, Kim M, Nagashima A, Nakazato E,
Tanaka, Takahashi H )2000) Chloroplast targeting, distribution
and transcriptional ¯uctuation of AtMinD1, a Eubacteria-type
factor critical for chloroplast division. Plant Cell Physiol
41:1119±1128
Leon P, Arroyo A, Mackenzie S )1998) Nuclear control of plastid
and mitochondrial development in higher plants. Annu Rev
Plant Physiol Plant Mol Biol 49:453±580
In addition to the Arabidopsis MinD gene, a rice
MinD homologue has been identi®ed )GenBank
Accession Number AF149810). Neither the MinE nor
the MinC plant homologues have been identi®ed. The
MinD gene and the other gene products of the Min
operon have been identi®ed in several photosynthetic
organisms. The MinD and MinE gene have been
identi®ed in the Chlorella chloroplast genome, and the
MinE gene is found adjacent to MinD in a similar
arrangement to that in E. coli. It is postulated that the
MinD and MinE genes are co-translated in E. coli to
ensure that equal protein levels are present )de Boer
et al. 1989). Although the MinD/MinE gene co-regula-
tion has not been shown in Chlorella chloroplasts, the
gene arrangement makes this an attractive hypothesis.
The conservation of the MinD protein, and its inter-
action with MinE and MinC suggests that it might
interact with additional proteins as well. It remains to
be determined what these gene products are in higher-
plant chloroplast division.
Link
G )1991) Photoregulated development of chloroplasts
)including C4). In: Bogorad L, Vasil I )eds) Cell culture and
somatic cell genetics of plants, vol 7, the molecular biology of
plastids and the photosynthetic apparatus. Academic Press,
New York, pp 365±394
Lutkenhaus J, Addinall SG )1997) Bacterial cell division and the Z
ring. Annu Rev Biochem 66:93±116
Maiti I, Murphy JF, Shaw JG, Hunt AG )1993) Plants that express
a potyvirus proteinase gene are resistant to virus infection. Proc
Natl Acad Sci USA 90:6110±6114
McFadden GI )2001) Chloroplast origin and integration. Plant
Physiol 125:50±53
Mehta M, Sara®s V, Critchley C )1999) Thylakoid membrane
architecture by confocal microscopy. Aust J Plant Physiol
26:709±716
Miles D )1994) The role of high chlorophyll ¯uorescence photo-
synthesis mutants in the analysis of chloroplast thylakoid
membrane assembly and function. Maydica 39:35±45
Mullet JE )1988) Chloroplast development and gene expression.
Annu Rev Plant Physiol Plant Mol Biol 39:475±502
Osteryoung KW, Stokes KD, Rutherford SM, Percival AL, Lee
WY )1998) Chloroplast division in higher plants requires
members of two functionally divergent gene families with
homology to bacterial FtsZ. Plant Cell 10:1991±2004
Picherscky E, Brock TG, Nguyen D, Ho€man NE, Piechulla B,
Tanksley SD, Green BR )1989) A new member of the CAB gene
family: structure, expression and chromosomal location of Cab-
8, the tomato gene encoding the type III chlorophyll a/b-bind-
ing polypeptide of photosystem I. Plant Mol Biol 12:257±270
Acknowledgments The authors thank Curtis Meurer, Michael
Hayes and Ray Stevens for technical assistance in the laboratory
and greenhouse; Matt Turnbull for assistance with the confocal
scanning laser microscopy; Mary Gail Engle and Mary Jennes at
the University of Kentucky Electron Microscopy and Imaging
Facility for technical assistance with the transmission electron mi-
croscopy; George Cheniae, Sharyn Perry, Bruce Downie and
Robert Houtz for manuscript reviews. This research was supported
by KAES Hatch funds to G.B. Collins. This paper )No. 00-06-67)
is published with the approval of the Director of the Kentucky
Agricultural Experiment Station.

188
Pyke KA )1997) The genetic control of plastid division in higher Sambrook J, Fritsch EF, Maniatis T )1989) Molecular cloning: a
plants. Am J Bot 84:1017±1027
laboratory manual, 2nd edn. Cold Spring Harbor Laboratory
Press, Cold Spring Harbor, NY
Pyke KA )1999) Plastid division and development. Plant Cell
11:549±556
Pyke KA, Leech RM )1992). Chloroplast division and expansion is
radically altered by nuclear mutations in Arabidopsis thaliana.
Plant Physiol 99:1005±1008
Pyke KA, Rutherford SM, Robertson EJ, Leech RM )1994) arc6, a
fertile Arabidopsis mutant with only two mesophyll cell chlo-
roplasts. Plant Physiol 106:1169±1177
Schardl CI, Byrd AD, Benzion G, Altschuler M, Hilderbrand DF,
Hunt AG )1987) Design and construction of a versatile system
for the expression of foreign genes in plants. Gene 61:1±11
Schwartz E, Pichersky E )1990) Sequence of two tomato nuclear
genes encoding chlorophyll a/b-binding proteins of CP24, a
PSII antenna component. Plant Mol Biol 15:157±160
Smyth DR )1999) Gene silencing: cosuppression at a distance. Curr
Biol 7:793±795
Rochaix J-D )2001) Posttranscriptional control of chloroplast gene
expression: from RNA to photosynthetic complex. Plant Strepp R, Schotz S, Kruse S, Speth V, Reski R )1998) Plant nuclear
Physiol 125:142±144
Roth®eld L, Justice S, Garcia-Lara J )1999) Bacterial cell division.
Annu Rev Genet 33:423±448
gene knockout reveals a role in plastid division for the homo-
logue of the bacterial cell division protein FtsZ, an ancestral
tubulin. Proc Natl Acad Sci USA 95:4368±4373