Cell, Vol. 70, 621-629, August 21, 1992, Copyright @ 1992 by Cell Press
In Situ Hybridization to the Crithidia fasciculata
Kinetoplast Reveals Two Antipodal Sites Involved
in Kinetoplast DNA Replication
David C. Ward,’ and Paul T. Englundt
*Departments of Molecular Biophysics
and Biochemistry and Genetics
New Haven, Connecticut 06510
tDepartment of Biological Chemistry
Johns Hopkins University School of Medicine
Baltimore, Maryland 21205
Al F. Torri,t
cles and maxicircles.
probes detected by digital fluorescence
has clarified the in vivo structureand
anism of the network.
tion kinetoplasts (with closed
cles on the periphery
center), and postreplication
with two peripheral structures containing free minicir-
cle replication intermediates
Replication may involve release of closed minicircles
from the center of the kinetoplast
to the peripheral structures,
minicircles therein, and then peripheral reattachment
of the progeny minicircles
DNA is a network of interlocked
In situ hybridization,
The probe recognizes
kinetoplasts (with nicked minicir-
and closed minicircles
and DNA polymerase.
and their migration
replication of the free
to the kinetoplast.
Both structurally and functionally, kinetoplast DNA(kDNA)
is possibly the most unusual DNA found in nature. kDNA
is a mitochondrial DNA, present only in trypanosomes and
related protozoan parasites, and it is organized in the form
of a network of several thousand minicircles and a few
dozen maxicircles, all of which are topologically
locked. Each cell has only one mitochondrion, which con-
tains a single kDNA network. Maxicircles are similar to
mitochondrial DNAs in higher eukaryotes in that they en-
code ribosomal RNAs and a handful of proteins involved
in mitochondrial energy transduction
However, the mechanism of gene expression is unique.
Recently there has been great interest in the fact that maxi-
circle mRNA precursors undergo the amazing process of
RNA editing, in which uridine residues are either inserted
or deleted at specific sites to create a functional transla-
tional reading frame (Benne et al., 1986; Feagin et al.,
1988; Simpson and Shaw, 1989; Feagin, 1990; Simpson,
1990). Both the maxicircles and the minicircles encode
guide RNAs, which control the specificity of RNA editing
(Blum et al., 1990; Sturm and Simpson, 1990; Pollard et
al., 1990). See Ryan et al. (1988) Simpson (1987), Ray
(1987), and Stuart (1983) for reviews on kDNA.
In the insect trypanosomatid Crithidia fasciculata, the
nonreplicating network contains about 5000 minicircles
(2.5 kb) and about 25 maxicircles (37 kb) (Englund, 1978;
Marini et al., 1980). As visualized by electron microscopy,
the isolated network is a cup-shaped sheet of DNA roughly
10 urn x 15 urn. However, inside the cell, as viewed by
electron microscopy of thin sections, the network is com-
pacted into a disk-shaped nucleoid body that we term the
kinetoplast. This structure is roughly 1 pm in diameter and
about 0.3 urn thick (Kusel et al., 1967; Anderson and Hill,
1969). There is very little known about how the minicircles
are organized within this disk in vivo, although there have
been a few speculations (Delain and Riou, 1969; Marini et
al., 1983; Silver et al., 1986; see Discussion).
We have therefore investigated the in vivo structure of
kDNA in C. fasciculata using in situ hybridization with
biotin-labeled minicircle probes. The probes were de-
tected by fluorescently labeled avidin allowing visualiza-
tion of the kinetoplast by confocal scanning and charge-
coupled device (CCD) fluorescence microscopy. In this
paper, we describe the in vivo organization of the minicir-
cle component of the kinetoplast, and we describe the
striking changes in kinetoplast structure that take place
during replication. We also demonstrate free minicircle
replication intermediates and DNA polymerase (detected
by immunofluorescence) in two small structures located
on opposite sides of the kinetoplast disk. These observa-
tions suggest a detailed model for kinetoplast replication
with Acridine Orange
We first used confocal microscopy to visualize the kineto-
plast in stationary phase C. fasciculata cells stained with
acridine orange (Figure 1A). Reconstruction of serial sec-
tions confirmed that the kinetoplast is in the form of a disk
about 1 urn in diameter and 0.4 urn thick, a size in close
agreement with that reported by electron microscopy (Ku-
sel et al., 1967; Anderson and Hill, 1969). When the cells
were treated under conditions used for in situ hybridization
(namely, fixation with paraformaldehyde, permeabilization
with Triton X-100, and incubation with 50% formamide
at 70%), there was no detectable change in kinetoplast
structure at this scale (data not shown).
of the Kinetoplast by Staining
by Protease Treatment
The kinetoplast disk is perpendicular to the base of the
flagellum; therefore, when the cell is lying on the micro-
scope slide, the disk is perpendicular to the slide surface
and must be visualized from its edge. Since our subse-
quent studies with DNA probes would be more informative
if the disk could be visualized from the top, we developed
a protocol for changing the orientation of the disk inside
the cell. We found that if fixed mounted cells were mildly
of the Kinetoplast Disk
Figure 1. Kinetoplast Dimensions Are Not Altered by Protease Treat-
ment and Hybridization Conditions
Ceils were from stationary phase cultures, and kinetoplasts
were virtually identical in structure.
(A) Confocal X-Z plane image section through the greatest diameter
of an acridine orange-stained kinetoplast from a fixed C. fasciculata.
Mounted C. fasciculata were treated with RNAase A at 1 ug/ml for 1
hrat 37’C prior to staining with 0.01% acridineorange
orange was used because it is a DNA stain that can be visualized by
confocal microscopy using the same wavelength filters as required for
FITC. This image shows the edge of a vertically oriented kinetoplast
(B) CCD image of an X-Y plane DAPI image of a kinetoplast that has
undergone both protease digestion and formamide denaturation.
image shows the flat surface of a horizontally oriented kinetoplast disk.
Bar is 1 urn.
in all cells
in PBS. Acridine
digested by proteinase K, the kinetoplast disk would “fall
over,” allowing visualization from its planar side. This treat-
ment resulted in little visible change to the structure of the
cell except for the kinetoplast reorientation; the structure
of the kinetoplast itself was not detectably different as ob-
served by CCD or confocal fluorescence microscopy. To-
tal nucleic acid staining revealed the same dimensions
(Figure lB), and subsequent hybridization studies con-
firmed that proteinase K treatment did not alter the kineto-
plast structure (see legend to Figure 2).
We first used in situ hybridization to visualize the kineto-
plast in cells from an asynchronous culture of logarithmi-
cally growing C. fasciculata (8 x 10” cells/ml). This popula-
tion should include cells with kinetoplasts at all stages of
replication. This experiment involved a double fluores-
cence label, in which each kinetoplast was visualized with
probe detected with fluorescein isothiocyanate (FITC). Al-
though all kinetoplasts stained uniformly with DAPI, we
visualized three strikingly different types of kinetoplast
structures using the minicircle probe. Type A kinetoplasts,
present in about 42% of log phase cells, hybridized ex-
tremely weakly to the minicircle probe (Figure 2A; note that
Probes Reveal Three Types
(DAPI) and a minicircle
7.0*10 ’ 65 12
Figure 2. Visualization of the C. fasciculata Ki-
netoplasts by DAPI and by Hybridization
Minicircle Probes Reveals Three Distinct Popu-
C. fasciculata cells were fixed, treated with pro-
tease, and hybridized.
then visualized by DAPI fluorescence
cle hybridization patterns
three types. (A) shows type A kinetoplasts.
These are not visually detectable with the mini-
circle probe under standard conditions. These
two imageswere obtained by allowing the CCD
camera to accumulate
longer than for those in (B) and (C). (B) shows
type B kinetoplasts. With the minicircle probe
they resemble donuts with a hole of varying
size. Arrows indicate protrusions, which will be
discussed below. (C) shows type C kineto-
plasts, which are uniformly fluorescent with the
minicircle probe. (D) shows the proportion
each type of kinetoplast when measured in cul-
tures of the indicated density. All kinetoplasts
were counted in 20 adjacent microscope view-
ing fields. Populations
kinetoplasts were evaluated from cultures of
6.1 x 10B/ml (mid log phase), 7.0 x lO’/ml
(near stationary phase), and 1.2 x 108/ml (sta-
tionary phase), respectively.
Each kinetoplast was
were divided into
photons for 15 times
of 286, 453, and 167
Scale bar is 1 urn.
For in situ hybridization
probes gave indistinguishable
horizontally oriented kinetoplasts)
to the “conserved
of this type, we used three different minicircle probes. One covered the region of the minicircle bent helix,
sequence” about 90° from the bent helix, while the third contained total minicircle sequences. The three
results except that the total minicircle probe resulted in higher signal intensities.
in which type B kinetoplasts were compared in nonproteased
cells by confocal fluorescence microscopy indicated that the donut structure
(i.e., vertically oriented kinetoplasts) and proteased (i.e.,
is not altered by the protease
Replication In Vivo
Figure 3. DNAase
Probes to Type A Kineloplasts
Images of kinetoplasts
minicircle probe. Prior to hybridization
DNAase I, as described in Experimental
used were 0.01 nglml (A), 0.1 nglml (B), 1 .O nglml (C), 10 nglml (D),
and 100 nglml (not shown). The 100 rig/ml treated slide showed no
recognizable structures, either by minicircle hybridization
fluorescence, presumably because the kinetoplasts
I Nicking Increases
Phase C. fasciculata
phase cells hybridized
the samples were treated with
from stationary with
or by DAPI
had been com-
these two minicircle-FITC
over the others in this figure). Type B kinetoplasts, also
present in about 42% of the cells, had an unexpected
structure. Although DAPI revealed uniform fluorescence,
the minicircle probe detected only the peripheral region of
the disk, resulting in a donut-shaped structure. In some
type B kinetoplasts, the fluorescence did not extend com-
pletely around the periphery but was concentrated into two
peripheral zones on opposite sides of the kinetoplast (see
below). The type B kinetoplasts displayed a continuum of
structures. The smaller type B kinetoplasts hybridized only
in the extreme peripheral region of the disk, with larger
“donut holes” (Figure 28, upper panels). The larger type
B structures had larger peripheral regions of fluorescence
and smaller holes (Figure 28, lower panels). Type C kineto-
plasts, constituting about 16% of the total, hybridized uni-
formly with the minicircle probe (Figure 2C). In contrast to
logarithmically growing cells, those from near stationary
phase cultures (7 x 10’ cells/ml) had kinetoplasts that
images were enhanced 1Bfold
were 65% type A, and those in full stationary phase (1.2
x lOa cells/ml) had kinetoplasts that were 99% type A
The average area of the top of the kinetoplast disk, cal-
culated from DAPI fluorescence, was 0.54 f 0.08 pm* for
type A kinetoplasts and 1 .16 f 0.12 pm* for type C. Type
B kinetoplasts displayed a range of sizes intermediate be-
tween type A and type C; the average area of type B struc-
tures with a donut hole approximately one-third the diame-
ter of the donut (like that in Figure 28, lower left panel) was
0.89 + 0.09 pm*. All three kinetoplast types were identical
in disk thickness, as assessed from measurements
DAPI-stained kinetoplasts in the vertical (nonproteased)
orientation. Some kinetoplasts showed two small protru-
sions of increased signal intensity oriented on opposite
sides of the disk (arrows in Figure 28); the structure and
biological significance of these will be discussed below.
Note that the larger type B kinetoplasts and all the type C
kinetoplasts are elongated disks, as visualized by either
DAPI staining or minicircle hybridization (Figures 2B and
2C). With type B and C kinetoplasts, the outer edge of the
minicircle hybridization corresponded to the edge of the
DAPI fluorescence. The same is probably true for type A
kinetoplasts, although the weak signal from the minicircle
probe made this point difficult to confirm.
Absence of Minicircle
Is Caused by Inefficient
Stationary phase C. fasciculata contain kDNA networks
composed exclusively of covalently closed minicircles(En-
glund, 1978). In contrast, partially replicated networks, iso-
lated from cells in logarithmically growing cuftures, contain
a central zone of covalentfy closed minicircles and a pe-
ripheral zone of nicked or gapped minicircles (Englund,
1978). It therefore seemed possible that covalently closed
minicircles might not be accessible to hybridization and
that the minicircle-specific fluorescence detected by CCD
fluorescence microscopy might reflect the location of
nicked or gapped minicircles within the kinetoplast in vivo.
We used two different procedures to evaluate this hy-
pothesis. First, we mildly digested protease-treated sta-
tionary phase C. fasciculata with various concentrations
of DNAase I (Figure 3). Increasing concentrations of this
enzyme resulted in increased hybridization of the minicir-
cle probe, presumably owing to nicking of the minicircles.
Eventually, the fluorescence signal reached a plateau
equivalent in size and uniformity to the DAPI-stained ki-
netoplast (Figure 3D). Further increases in enzyme con-
centration resulted in degradation of the network and a
diminution of the hybridization signal. Second, we hybrid-
ized the minicircle probe to kinetoplastsfrom both logarith-
mically growing and stationary-growth phase C. fascicu-
lata with denaturation at 100°C (in contrast with the
standard 70°C). This treatment also resulted in a uniform
and strong fluorescence of the kinetoplast (data not
shown). Presumably the high temperature of denaturation
introduces nicks in the minicircles or converts them to a
form accessible to the probe.
Fluorescence in Type A
and in the Center of Type B Kinetopkts
Hybridiion to Covakntly
Figure 4. Protrusions Are Found on Most Type
Examples of type B kinetoplasts
sions. Each panel shows 12 kinetoplasts,
a DAPI image shown above an imageof minicir-
hybridization to kinetoplastssubjected
denaturation prior to hybridization.
the minicircle hybridization
turation of the target DNA. Bar is 1 pm.
(A) shows minicircle
without prior dena-
We also treated kinetoplasts from log phase cells with
DNAase I. This treatment resulted in loss of the donut-
shaped kinetoplasts. It appeared that the center of the
donut became fluorescent (owing to nicking of minicircles),
but the peripheral region, previously fluorescent, was lost
(data not shown). This loss was not unexpected, since
some of the minicircles on the periphery of partly replicated
networks are already highly nicked or gapped (Kitchin et
al., 1985, 1984; Birkenmeyer et al., 1987).
As shown in Figure 28 (see arrows), some kinetoplasts, as
visualized by minicircle hybridization, have two associated
fluorescent protrusions. These protrusions are found fre-
quently on type B kinetoplasts (128 of 145) never on type
A kinetoplasts (0 of 152), and only occasionally on type C
kinetoplasts (8 of 28). Figure 4A shows examples of type
B kinetoplasts. The protrusions are found in structures
Are Found on Most Type B Kinetoplasts
with both large and small donut holes. They are always
about 180” apart, but they are distributed randomly rela-
tive to the long axis of larger type B kinetoplast disks. In
cells not treated with protease (i.e., those with a vertically
oriented kinetoplast), the two protrusions are almost al-
ways in the same plane, which is parallel to the surface of
the slide. This fact implies both that the C. fasciculatacells
have a preferred orientation on the slide (possibly owing
to a flattened surface) and that the protrusions have a fixed
orientation relative to the flattened surface.
The protrusions found on the type B kinetoplasts are remi-
niscent of kinetoplast-associated structures demonstrated
previously to contain topoisomerase
1988). By immunofluorescence,
two structures (roughly the same size as the protrusions)
on Type B Kinetoplasts
II (Melendy et al.,
this enzyme localizes in
;;betoplast Replication In Vivo
Figure 5. Type B Protrusions
Log phase C. fasciculata were hybridized with
panel contains a minicircle-FITC
(right) and the corresponding
cence (left). In the DAPI images, the nucleus is
the structure below the kinetoplast; in the FITC
images, the nonspecific
cence is due to incomplete washing out of the
riboprobes. Images in the left column (A, C, and
E) derive from hybridization
probe. Images in the right column (6, D, and
F) derive from hybridization
probe. (A) and(B) are nondenaturing
tions to non-protease-treated
(vertically oriented kinetoplasts).
are nondenaturing hybridizations
(E) and (F) show images from denaturing
bridizations to protease-treated
is 1 flm.
Probes without Minicircle
of Target DNA
with the H strand
with the L strand
(C) and (D)
cells. Scale bar
on opposite sides of the kinetoplast disk. If minicircle fluo-
rescence in the protrusions is associated with topoisomer-
ase II, it would be possible that this fluorescence derives
from free minicircle replication intermediates
about to be reattached to the network. Free minicircles are
molecules that have decatenated from the network for the
purpose of replication (Englund, 1979).
To test this possibility, we exploited the fact that the
free minicircle population includes some molecules with L
strand single-stranded regions (Englund et al., 1982),
which should be detectable without denaturation. Indeed,
hybridization in the absence of denaturation revealed only
the protrusions and not the kinetoplast disk (Figure 46).
To distinguish whether the target sequences were L strand
or H strand, we generated strand-specific riboprobes from
a pGEM 32 vector containing a 312 bp segment of the
minicircle. This sequence contains a conserved region
with one of the replication origins (Sugisaki and Ray,
1987). Only the H strand riboprobe detected any target
in a nondenaturing hybridization, illuminating two foci on
opposite sides of the kinetoplast disk (Figures 5A-5D);
therefore, the target sequences must be exclusively L
strand. Both the H and L strand riboprobes hybridized to
kinetoplasts after denaturation, revealing the donut struc-
ture (Figures 5E and 5F). The signal types observed from
such a denaturing hybridization, although less intense,
were the same as those from the total minicircle probe
hybridized under the same conditions.
Antibody to DNA Polymerase
To explore the possibility that the kinetoplast protrusions
might be sites of minicircle replication, we used an immu-
nofluorescence technique to localize a recently isolated
C. fasciculata mitochondrial DNA polymerase (Torri and
Englund, 1992). As shown in Figure 6, antibodies to this
Colocalizes with the
Figure 6. DNA Polymerase Is Localized to
Same Positions Detected by the Minicircle
A mouse antiserum was used for immunofluo-
rescence localization of mitochondrial DNA
pofymerase (Toni and Englund, 1992) in log
phase C. fasciculata. Each panel contains an
an&DNA pofymerase-FlTC image (right side)
compostted with its corresponding DAPI image
(left side). The images are oriented with the
kinetopfast above the nucleus. Identical results
were obtained with the rabbit antibody de-
scribed previously (Torn and Englund, 1992).
protein illuminate two foci on opposite sides of the kineto-
plast disk, a location virtually indistinguishable
of the minicircle protrusions. (For this experiment we used
nonproteased C. fasciculata; therefore, the images in Fig-
ure 6 should be compared with the one in Figure 5A.) As
with the nondenaturing minicircle hybridizations,
signals from the DNA polymerase antibody are almost al-
ways located in a plane parallel to the slide surface, owing
to the apparent preferred orientation of the cell on the slide.
Therefore, we can confidently state that the minicircle and
polymerase signals colocalize.
kDNA replication occurs during a discrete S phase of the
C. fasciculata cell cycle (Cosgrove and Skeen, 1970).
There is already considerable information available about
the structure of isolated networks from different stages of
the replication cycle (Englund, 1978). Prior to replication,
in the Gl phase of the cycle, the network contains about
5000 covalently closed minicircles.
work, a structure also present in stationary phase cells.
When the S phase begins, minicircles are released from
the network by a topoisomerase for the purpose of replica-
tion as B-type intermediates
progeny of each parental free minicircle, which contain
nicks or gaps, are reattached to the periphery of the net-
work in another topoisomerase
1974; Kitchin et al., 1985, 1984). This process results in
the network growing in size. The replicating network con-
tains two zones, a peripheral zone of nicked or gapped
minicircles that have undergone
This is a form I net-
(Englund, 1979). The two
reaction (Simpson et al.,
replication, and a central
zone containing covalently closed minicircles
not. When all minicircles have replicated, the network has
doubled in size and all of its minicircles
gapped (form II networks). When the form II network splits
in two and the minicircle nicks and gaps are repaired, the
products aretwoform I networks, each of which is identical
to the parent network (see Ryan et al.  and Ray
[1987) for reviews of kDNA replication). All of this informa-
tion on kDNA replication has been derived from investiga-
tion of isolated networks, which are giant two-dimensional
sheets of DNA; there has been virtually no information on
how these various types of networks are organized into
the much smaller kinetoplast disks found in vivo.
In log phase cells, we observed three types of kineto-
plast structures, types A, B, and C (Figure 2). In stationary
phase cells, all kinetoplasts were type A. It is likely that the
three different kinetoplast structures represent different
stages in the replication cycle. The most interesting struc-
ture was type B, in which minicircle hybridization resem-
bled a donut. This structure at first seemed inconsistent
with the uniform DAPI staining and the uniform kinetoplast
structure observed by electron microscopy of C. fascicu-
lata thin sections. However, DNAase I nicking (Figure 3)
and 100% treatment indicated that the nonhybridizing
type A networks (as well as the donut holes) must be invisi-
ble, since they contain covalently closed minicircles that
are inaccessible to the minicircle probe. Therefore, the
pattern of minicircle hybridization within the type B kineto-
plast simply reveals the distribution of nicked and cova-
lently closed minicircles. By this criterion, type A kineto-
plasts must be composed of form I networks, type B must
be composed of replicating
are nicked or
networks (with nicked or
Replication In Vivo
gapped minicircles at the periphery), and type C must be
composed of form II networks. The isolated replicating
network is a two-dimensional sheet greater than 10 urn x
15 pm in size; however, the type B kinetoplast is a disk
roughly 1 .l urn in diameter and about 0.4 urn thick. There-
fore, the kDNA network must be compacted in vivo in a
manner that conserves the peripheral location of the
nicked or gapped minicircles.
Our in situ hybridization results correlate in other ways
with the structure of isolated networks. Just as the size of
isolated networks increases during the replication cycle,
from form I, to replicating, to form II (Englund, 1978), the
size of the kinetoplast in vivo progresses from type A to
type B to type C. Furthermore, isolated replicating net-
works from different stages of replication differ in the rela-
tive sizes of the peripheral ring of nicked or gapped minicir-
cles. Early replicative forms have a narrow peripheral ring
and later forms have a much broader ring (Englund, 1978;
unpublished data). Similar stages are seen in type B ki-
netoplasts (Figure 28). Isolated form II networks, which
are double size, have an elongated structure (Englund,
1978). Type C kinetoplasts also have an elongated struc-
ture (Figure 2C). Finally, the relative frequency of types A,
B, and C kinetoplasts (42%, 42%, and 16%, respectively)
observed in this study correlates well with the ratio of form
I, replicating, and form II networks (50%, 30%, and 200/o,
respectively) isolated by cesium chloride-propidium
dide gradients from logarithmically growing C. fasciculata
There is little known about how the massive kDNA net-
work is condensed into the in vivo kinetoplast disk. Over
20 years ago, Delain and Riou noted that in Trypanosoma
cruzi the thickness of the kinetoplast disk, as visualized by
electron microscopy of thin sections, is roughly half the
circumference of a minicircle; they also noted that the DNA
fibers seem to be oriented parallel to the axis of the disk
(Delain and Riou, 1969). They proposed that each minicir-
cle is elongated like a rubber band and interlocked with
several neighboring minicircles. In this way, the DNA net-
work would be condensed into a disk-shaped structure.
(See Marini et al.  for further discussion of this model
and Silver et al.  for a variation of it.) Our data are
consistent with this model, which is shown in Figure 7.
This diagram is asection through a replicating kinetoplast.
Most importantly, in this model the zone of nicked or
gapped minicircles, known to be on the periphery of iso-
lated replicating networks, is also located on the periphery
of the replicating kinetoplast in vivo. Our data completely
rule out other bizarre patterns for folding of the kDNA net-
work in which the periphery of the network does not coin-
cide with the periphery of the kinetoplast in vivo.
Of special interest are the protrusions, detected by mini-
circle probes, found on opposite sides of some kinetoplast
disks (Figures 2,4, and 5). These are found frequently with
type B kinetoplasts, rarely with type C, and never with type
A. Since it is possible that the protrusion-associated
C kinetoplasts are in fact late-stage type Bs, it is likely that
the protrusions are associated exclusively with kineto-
plasts undergoing replication. The fact that minicircles
within the protrusions include some that contain single-
stranded regions (as judged by hybridization without prior
denaturation of the target DNA) is consistent with the pos-
sibility that these minicircles are free minicircle replication
intermediates. Strong support for this contention was de-
rived from hybridization with strand-specific riboprobes.
These experiments demonstrated that the single-strand
minicircle sequences in the protrusions are exclusively L
strand. It is likely that the molecules detected by the probe
include 8 structures with single-strand regions and possi-
bly single-strand circles (Englund et al., 1982; Kitchin et
al., 1985; Sheline et al., 1989).
Because the protrusions apparently contain not only
free minicircle replication intermediates but also a mito-
chondrial DNA polymerase (Figure 6) and a topoisomer-
ase II previously studied by Melendy et al. (1988), we sug-
gest that the two protrusions may be the sites of minicircle
replication. If so, other replication proteins are also likely
to be organized within these structures. The localization
of these enzymes differs from that of other trypanosomatid
mitochondrial proteins. A heat shock protein (Engman et
al., 1989; P. N. Effron, D. M. Engman, J. E. Donelson,
and P. T. E., submitted) localizes in a halo around the
kinetoplast, and a kinetoplast-associated
izes with the T. cruzi kinetoplast (Gonzalez et al., 1990).
We envision that minicircle replication occurs by the
mechanism shown in Figure 7. Covalently closed minicir-
cles are released by a topoisomerase from the center of
the kinetoplast disk. The released free minicircles then
diffuse (or are transported) to one of the two complexes of
replication enzymes where they undergo replication via 0
Free Minicircle prior
Figure 7. Model for In Vivo Organization
Replication of kDNA
This diagram, a section through
of the kinetoplast disk, shows stretched-out
minicircles interlocked with their neighbors on
either side. The complexes
zymes contain free minicircle replication inter-
mediates, a DNA polymerase,
erase II. Nicked, newly replicated
are shown in bold. The disk axis is indicated
by the vertical line. See Discussion
of replication en-
and a topoisom-
’ Newly Replicated Minicircle
intermediates; presumably the mitochondrial DNA poly-
merase participates in this process. The products of free
minicircle replication, which contain nicks or gaps, are
then reattached to the network periphery. The topoisomer-
ase II described by Melendy, Sheline, and Ray (Melendy
et al., 1988), apparently localized within this complex,
could be responsible for unlinking of the parental minicir-
cle strands during replication and for reattachment of the
minicircle progeny to the network. The presence of the
two complexes on opposite sides of the kinetoplast disk
explains the autoradiographic
Simpson (1976). They found that newly synthesized mini-
circles, labeled with [3H]thymidine in a very short pulse,
are localized in two discrete foci, 1 80° apart, on the periph-
ery of the network. This radiolabeling pattern would be
expected for minicircles that had just been reattached to
the network periphery following replication within one of
the complexes. In some type B kinetoplasts, minicircle
fluorescence is localized exclusively in the complexes or
the immediate flanking regions (Figure 4, first and third
examples). These kinetoplasts are probably in the earliest
stages of replication, and the newly synthesized nicked
minicircles are not yet evenly distributed around the whole
Our data do not rule out an alternative possibility in which
only the final stages of replication occur in the protein
complexes (e.g., gap repair by the polymerase and reat-
tachment to the network by the topoisomerase II). In this
case, the early stages of replication could occur at some
unidentified site. However, we favor the model shown in
Figure 7, as the free minicircle replication intermediates
detectable by our hybridization probe under nondenatur-
ing conditions (e.g., 8 structures) probably are early repli-
The model presented in Figure 7 raises challenging
minicircle reattachment to the kDNA network occurs only
adjacent to the two complexes of replication enzymes, it
will be important to determine how newly replicated nicked
minicircles are ultimately distributed uniformly around the
network periphery. Another important question concerns
minicircle inheritance: if sister progeny minicircles are re-
attached at neighboring sites adjacent to a complex, do
they ultimately segregate into different daughter net-
works? We are currently seeking answers to these ques-
tions as well as trying to isolate intact functional com-
studies of Simpson and
For example, if progeny
Fixation and Protease Treatment of C. faaciculata Cells
C. fasciculata cells were cultured at 27OC in brain-heart
dium (Englund, 1976). Log or stationary
fuged, washed in phosphate-buffered
and resuspended in PBS containing
0.5% glutaraldehyde. After 10 min at room temperature,
was added to make a final concentration
was incubated for another 10 min. Cells destined for immunofluores-
cence detection were incubated for 10 min in 0.1 M glycine. The cell
suspension was then centrifuged,
stored at a concentration of about 5 x 10’ cells/ml in PBS at 4’C for
up to 2 weeks.
phase cultures were centri-
saline (PBS), centrifuged
of O.l%, and the suspension
washed three times in PBS, and
Cells were mounted by placing 200 pl of suspension (approximately
108 cells) on the slide (previously
or VectaBond [Vector Labs] to enhance adherence).
covered with an 16 x 16 mm coverslip, and the cells settled onto the
slide surface by gravity for at least 30 min. The coverslip was removed
by tilting the slide, which could then be stored in a Coplin jar with PBS.
To reorient the kinetoplast disk so that its flat surface was parallel to
the slide surface, the mounted cells were treated with 160 pl of protein-
ase K (Sigma; the optimal concentration
HCI (pH 6.0) 5 mM EDTA, 150 mM NaCI, and 0.5% SDS. The reaction
mix was covered with a 22 x 40 mm coverslip and the slide was placed
horizontally in a humid 37W chamber for 60 min. Care was taken not
to disturb the slides during this process. After removal of the coverslip,
slides were washed three times and stored in PBS. Nuclease digestion
of specimens was carried out using the same procedure
concentrations of DNAase I in 50 mM Tris-HCI
Prior to probe application and hybridization,
were equilibrated for at least 30 min in 50% formamide, 2 x SSC (1 x
SSC is 0.15 M NaCl plus 0.015 M sodium citrate).
coated with poly-b-lysine [Sigma]
The drop was
was 0.5 pglml) in 10 mM Tris-
(pH 6.0) 5 mM MgCb.
the PBS-washed slides
Three different minicircle probes were used. pPK2011CAT (Kitchin et
al., 1966) contains 219 bp of the minicircle bent region. pDP312, pre-
pared by Dr. David Perez-Morga, contains a Klenow-repaired
Ncol-Xbal fragment (Sugisaki and Ray, 1967) cloned into the Smal
siteof pGEM 3Z(Promega). It containsoneof
regions with a replication origin. The third probe was a gel-isolated
2.5 kb Xhol fragment of kDNA networks (Xhol cleaves nearly all C.
fasciculata minicircles once). This probe, containing
sequences, was used for all hybridizations
for the one in Figure 5.
The DNA probes were labeled by nick translation
modified deoxyuridine triphosphate
1990). In the case of plasmid probes, the entire plasmid was used. It
is important to tailor reaction conditions to produce probes less than
500 bases in length; longer probes result in nonspecific
RNA strand-specific probes were generated from the pDP312 plasmid
by incorporating Biotin-21-UTP (Clonetech) with T7 or SP6 RNA poly
merase according to the manufacturer’s
shown in this paper except
using standard protocols (Boyle,
Hybridization and Development
Hybridization was carried out as previously
1966) with minor modifications.
of 2 nglml in 10 nl of hybridization
10% dextran sulfate, and 5 ng of sheared [< 500 bp] salmon sperm
DNA). The probe droplet was placed on a well-drained
16 mm coverslip was placed over the droplet, and finally its edges were
sealed with rubber cement. The slides were then incubated at 70°C
for 6 min to denature the probe and target. They were then transferred
to a prewarmed 37OC humid incubation chamber for overnight hybrid-
ization. The 70°C denaturation step was omitted for nondenaturing
After 12 hr, the rubber cement was peeled off and the slides were
washed four times at 40°C (or 35OC when the minicircle bent region
probe was used) in 2x SSC, 50% formamide,
washes were done in Coplin jars; the first wash usually removed the
coverslip. The slides were then washed three times in 2x SSC, 0.1%
Tween 20 at 6OOC. Slides were blocked with 5% bovine serum albumin
in 4x SSC, 0.1% Tween 20 at 37OC for 30-60 min. The detection
reagent avidin-FITC (Vector Labs) was applied in 160 PI of 1% bovine
serum albumin, 4 x SSC, 0.1% Tween 20 at 37OC for 30-60 min. The
final wash included three changes of 4x
45OC. The slides were then well drained and covered with 25 t.d of an
anti-fading solution (20 mM Tris-HCI
vol] diazabicyclo[2,2,2]octane [Sigma]) containing 0.1 ttglml DAPI. A
22 x 40 mm coverslip was sealed on top with nail polish.
described (Lichter et al.,
The probe was used at aconcentration
solution (50% formamide, 2 x SSC,
slide, an 16 x
0.1% Tween 20. All
SSC, 0.1% Tween 20 at
[pH 8.0],90% glycerol [2.3% WV
lmmunofluorescent Detectlon of DNA Polymerase
Localization of the DNA polymerase was done using standard immuno-
cytochemical techniques. After fixation as described above, cells were
blocked in 20% goat serum in PBS at 37OC for 45 min. Mouse serum
(Torri and Englund, 1992) was diluted 1:lO in 5% goat serum in PBS
and incubated at 37W for 45 min. The slides were then washed in
Kinetoplast Download full-text
Replication In Vivo
three changes of PBS. The primaly antibody was detected using an
FITC-conjugated anti-mouse immunoglobulin
ary antibody was applied at 10 uglml under the same conditions as the
primary antibody and followed by similar washes. Finally, an anti-
fading solution containing DAPI was applied and sealed under a cov-
G (Sigma). The second-
Two devices were used for fluorescent
MRC-500 laser scanning confocal microscope
Optiphot was used on samples requiring serial section acquisition.
Photometrics PM512 cooled CCD camera attached to a Zeiss Axi-
oskop was used when the kinetoplasts
protease treatment. The CCD system has the advantage of being able
to detect very low signal intensity. Apple Macintosh computers
used for camera control and image processing.
signal detection. A Bio-Rad
mounted onto a Nikon
had been “tipped over” by the
We thank David Perez-Morga for providing the DNA probes and Terry
Shapiro and Kathy Ryan for comments on the manuscript. This research
was supported by grants from NIH (GM-401 15 to D. C. W. and GM-27606
to P. T. E.) and the MacArthur Foundation (to P. T. E.). A. F. T. was
supported by an NIH postdoctoral fellowship (GM13604-02).
The costs of publication of this article were defrayed
the payment of page charges. This article must therefore be hereby
marked “advertisement” in accordance
solely to indicate this fact.
in part by
with 18 USC Section 1734
Received December 9, 1991; revised June 23, 1992.
Anderson, W., and Hill, G. C. (1989). Division and DNA synthesis in
the kinetoplast of Crithidia fasciculara. J. Cell Sci. 4, 61 l-620.
Benne, R., Van Den Burg, J., Brakenhoff, J. P. J., Sloof, P.. Van Boom,
J. H., and Tromp, M. C. (1986). Major transcript
cox /I gene from trypanosome mitochondria
that are not encoded in the DNA. Cell 46, 819-826.
Birkenmeyer, L., Sugisaki, H., and Ray, D. S. (1987). Structural charac-
terization of site-specific discontinuities
gins of minicircle DNA from Crifhidia fasciculata.
Blum, B., Bakalara, N., and Simpson,
editing in kinetoplastid mitochondria:
scribed from maxicircle DNA provide the edited information.
Boyle, A. R. (1990). Nonisotopic labeling of DNA probes using nick
translation. In Current Protocols in Molecular Biology, Supplement 12,
Section V (New York: John Wiley and Sons), pp. 3.18.1/3.18.7.
Cosgrove, W. B., and Skeen, M. J. (1970). The cell cycle in Crithidia
fasciculafa: temporal relationships
cleic acid in the nucleus and in the kinetoplast. J. Protozool. 17, 172-
Delain, E., and Riou, G. (1969). DNA ultrastructure
of Trypanosoma cruzi cultivated in vitro. CR Acad. Sci. 268, 1225-
Englund, P. T. (1978). The replication of kinetoplast DNA networks in
Crithidia fasciculata. Cell 74, 157-168.
of the frameshifted
contains four nucleotides
associated with replication ori-
J. Biol. Chem. 262,
L. (1990). A model for RNA
“guide” RNA molecules tran-
between synthesis of deoxyribonu-
of the kinetoplast
Englund, P. T. (1979). Free minicircles of kinetoplast DNA in Crithidia
fasciculata. J. Biol. Chem. 254, 4895-4900.
Englund, P. T., Hajduk, S. L., Marini, J. C., and Plunkett, M. L. (1982).
Replication of kinetoplast DNA. In Mitochondrial
P. Eorst, and G. Attardi, eds. (Cold Spring Harbor, New York: Cold
Spring Harbor Laboratory), pp. 423-433.
Engman, D. M., Kirchhoff, L. V., and Donelson. J. E. (1989). Molecular
cloning of mtp70, a mitochondrial
Cell. Biol. 9, 5163-5168.
Feagin, J. E. (1990). RNAediting in kinetoplastid
Chem. 265, 19373-19376.
Genes, P. Slonimski,
member of the hsp70 family. Mol.
mitochondria. J. Biol.
Feagin, J. E., Abraham, J. M., and Stuart, K. (1988). Extensive editing
of the cytochrome c oxidase Ill transcript in Trypanosoma
Gonzalez, A., Resales, J. L., Ley, V., and Diaz, C. (1990). Cloning and
characterization of a gene coding for a protein (KAP) associated with
the kinetoplast of epimastigotes and amastigotes
cruzi. Mol. Biochem. Parasitol. 40, 233-244.
Kitchin, P. A., Klein, V. A., Fein, B. I., and Englund, P. T. (1984).
Gapped minicircles. A novel replication
DNA. J. Biol. Chem. 259, 15532-15539.
Kitchin, P. A., Klein, V. A., and Englund, P. T. (1985). Intermediates
in the replication of kinetoplast DNA minicircles.
Kitchin, P. A., Klein, V. A., Ryan, K. A., Gann, K. L., Rauch, C. A.,
Kang, D. S., Wells, R. D., and Englund, P. T. (1986). A highly bent
fragment of Crithidia fasciculata kinetoplast DNA. J. Biol. Chem. 261,
Kusel, J. P., Moore, K. E., and Weber, M. M. (1967). The ultrastructure
of Crithidia fasciculata and morphological
in acriflavin. J. Protozool. 74, 283-296.
Lichter, P., Cremer, T., Borden, J., Manuelidis,
(1988). Delineation of individual human chromosomes
and interphase cells by in situ suppression
binant DNA libraries. Hum. Genet. 80, 224-234.
Marini, J. C., Miller, K. G., and Englund, P. T. (1980). Decatenation
kinetoplast DNA by topoisomerases.
Marini, J. C., Levene, S. D., Crothers, D. M., and Englund, P. T. (1983).
A bent helix in kinetoplast DNA. Cold Spring Harbor Symp. Quant.
Biol. 47, 279-283.
Melendy, T., Sheline, C., and Ray, D. S. (1988). Localization of a type
II DNA topoisomerase to two sites at the periphery of the kinetoplast
DNA of Crithidia fasciculata. Cell 55, 1083-1088.
Pollard, V. W., Rohrer, S. P., Michelotti, E. F., Hancock, K., and Haj-
duk, S. L. (1990). Organization of minicircle genes for guide RNAs in
Trypanosoma brucei. Cell 63, 783-790.
Ray, D. S. (1987). Kinetoplast DNA minicircles:
mitochondrial plasmids. Plasmid 77, 177-190.
Ryan, K. A., Shapiro, T. A., Rauch, C. A., and Englund, P. T. (1988).
The replication of kinetoplast DNA in trypanosomes.
biol. 42, 339-358.
Sheline, C., Melendy, T., and Ray, D. S. (1989). Replication of DNA
minicircles in kinetoplasts isolated from Crithidia fasciculata: structure
of nascent minicircles. Mol. Cell. Biol. 9, 169-176.
Silver, L. E., Torri, A. F., and Hajduk, S. L. (1966). Organized packaging
of kinetoplast DNA networks. Cell 47, 537-543.
Simpson, A. M., and Simpson, L. (1976). Pulse-labeling of kinetoplast
DNA: localization of 2 sites of synthesis within the networks and kinet-
ics of labeling of closed minicircles.
Simpson, L. (1987). The mitochondrial
zoa: genomic organization, transcription,
Annu. Rev. Microbial. 47, 363-382.
Simpson, L. (1990). RNA editing: a novel genetic phenomenon?
ence 250, 512-513.
Simpson, L., and Shaw, J. (1989). RNA editing and the mitochondrial
cryptogenes of kinetoplastid protozoa. Cell 57, 355-366.
Simpson, L., Simpson, A. M., and Wesley, R. D. (1974). Replication of
the kinetoplast DNAof Leishmania tarentolae and Crithidia fasciculata.
Biochim. Biophys. Acta 349, 161-172.
Stuart, K. (1983). Kinetoplast DNA, mitochondrial
ence. Mol. Biochem. Parasitol. 9, 93-104.
Sturm, N. R., and Simpson, L. (1990). Kinetoplast
encode guide RNAs for editing of cytochrome
mRNA. Cell 67, 879-884.
Sugisaki, H., and Ray, D. S. (1987). DNA sequence of Crithidia fascico-
lata kinetoplast minicircles. Mol. Biochem. Parasitol. 23, 253-263.
Torri. A. F., and Englund, P. T. (1992). Purification of a mitochondrial
DNA polymerase from Cfithidia fasciculata. J. Biol. Chem. 267, 4786-
intermediate of kinetoplast
J. Biol. Chem. 260,
changes induced by growth
L., and Ward, D. C.
hybridization using recom-
J. Biol. Chem. 255, 4976-4979.
Annu. Rev. Micro-
J. Protozool. 23, 583-587.
genome of kinetoplastid
DNA with a differ-
oxidase subunit Ill