Biochemical Genetics, Vol. 34, Nos.

7/8, 1996

Extensive C h r o m o s o m e Aberrations in Spiroplasma

citri Strain BR3
Fengchun Ye, 1 Ulrich Melcher, 2 John E. Rascoe, 1
and Jacqueline Fletcher 1,3

Received 9Jan. 1996--Final lO May 1996

Genetic variations in the plant pathogen, Spiroplasma citri strain BR3, were
characterized through physical genome mapping of the original isolate, BR3-3X,
and two derivatives, BR3- T and BR3- G, obtained after several years of different
maintenance conditions. BR3-T was transmitted from plant to plant via its
natural insect vector, the leafhopper Circulifer tenellus, while BR3-G was
maintained only in plants by periodic grafiing and has lost its ability to be insect
transmitted. By pulsed field gel electrophoresis (PFGE) analysis and DNA
hybridization, extensive changes in chromosomal DNA restriction patterns
relative to the parent, BR3-3X, were observed in both BR3- T and BR3-G, each of
which also had a larger genome size than the parent line. Genetic organization
was relatively conserved between BR3-T and BR3-3X. In contrast, a large
chromosomal inversion and deletions of approximately 10 kb near each of the
inversion borders were observed in BR3-G. One of the deletions, which included
several possibly functional genes, was closely linked to a SpVl-related trans-
posase gene. The locations of the deletion borders were also determined. The
results of this study demonstrated remarkable genome instability of spiroplasmas.

KEY WORDS: Spiroplasma citri; chromosome aberrations; SpVl-like sequences; transposase

and insect transmission.

INTRODUCTION
Spiroplasmas are wall-less prokaryotes (Mollicutes) having motility and a
helical morphology. Like other members of the class Mollicutes, they are

1 Department of Plant Pathology, Oklahoma State University, Stillwater, Oklahoma 74078.

2 Department of Biochemistry & Molecular Biology, Oklahoma State University, Stillwater,
Oklahoma 74078.
3 To whom correspondence should be addressed.

269
0006-2928/96/0800-0269509.50/0 ©1996PlenumPublishingCorporation
270 Ye, Melcher, Rascoe, and Fletcher

among the smallest and simplest organisms. Nevertheless, they are not
primitive prokaryotes but, apparently, were derived from Gram-positive
bacteria through successive genome reduction (Razin, 1989) and other
genetic changes such as a reduced (25 tool%) G + C content. The genome
reduction in these organisms may have led to the elimination of genes
involved in DNA repair systems that are commonly found in other bacteria
and a consequent high frequency of genetic mutation in mollicutes (Razin,
1989). Remarkable genetic variation, particularly variable genome sizes, has
been found in mycoplasmas (Christiansen et al., 1987), and a recent survey
indicated that the genome sizes of spiroplasma species also vary from 990 to
2200 kilobase pairs (kb) (Carle et al., 1995). However, significant genetic
variation within a single isolate of spiroplasma or genetic changes linked to a
particular phenotypic switch have not yet been reported.
In our laboratory, S. citri strain BR3 was isolated from horseradish
plants with brittle root disease (Fletcher et al., 1981). This isolate, triply
cloned and designated BR3-3X, was maintained in several different ways
over a 10-year period. A BR3-3X derivative cell line, designated BR3-G,
obtained through extended plant-to-plant transmission of the original iso-
late by grafting, underwent a phenotypic switch, losing its insect transmissi-
bility (Wayadande et al., 1993). During this same time period, the BR3-3X
isolate was also maintained by repeated transmission from turnip to turnip
via its natural insect vector, the leafhopper CircuIifer tenellus. The reisolated
S. citri, designated BR3-T, was still insect transmissible and lacked obvious
phenotypic changes. In this study, we report the physical mapping of the
genomes of S. citri strain BR3 and its derivatives and document for the first
time significant genetic variations within the same isolate of spiroplasmas
following maintenance under different conditions.

MATERIALS AND METHODS

Pulsed Field Gel Electrophoresis (PFGE) Analysis and Physical Mapping

of the Genome of S. citri BR3 Lines
S. citri BR3 lines BR3-3X, BR3-T, and BR3-G were grown to log phase in
LD8 broth medium (Davis, 1979) at 32°C. Cells were harvested by centrifu-
gation (12,000g, 30 min). Intact chromosomal DNA of S. citri was prepared
in low-melting point agarose as described previously (Ye et al., 1992) and
digested with restriction enzymes B s s H I I (Stratagene, La Jolla, CA) and
SalI (Promega, Madison, WI). The resulting fragments were separated by
pulsed field gel electrophoresis (PFGE) in 1% agarose using a TAFE system
(Geneline TM II, Beckman Instruments, Inc., Porterville, CA), as previously
reported (Ye et al., 1992, 1994a). In order to overlap B s s H I I and SalI
Chromosome Aberrations in Spiroplasma citri 271

Table I. Probes Used for Genome Mapping of S. cirri BR3 Lines

Probe Description Reference a

X2 1.8-kb E c o R I - S a l I fragment from S. citri BR3-3X

X3 3.1-kb E c o R I - S a l I fragment from S. cirri BR3-3X
X5 2.5-kb E c o R I - S a l I fragment from S. citri BR3-3X
X14 2.8-kb E c o R I - S a l I fragment from S. citri BR3-3X
X18 2.0-kb E c o R I fragment from S. cirri BR3-3X
X33 5.0-kb E c o R I fragment from S. cirri BR3-3X
A1 0.5-kb E c o R I fragment from S. citri BR3-3X
A4 1.5-kb E c o R I fragment from S. citri BR3-3X
A5 0.9-kb E c o R I fragment from S. citri BR3-3X
A8 2.0-kb E c o R I fragment from S. citri BR3-T
A9 1.2-kb E c o R I fragment from S. citri BR3-T
All 0.5-kb E c o R I fragment from S. citri BR3-3X
E46 4.0-kb E c o R I fragment from S. citri BR3-T
G25 0.8-kb E c o R I fragment from S. citri BR3-G
BsH 75-kb B s s H II fragment from S. citri BR3-3X
SII 83-kb SalI fragment from S. citri BR3-3X
or/C Chromosomal replication origin of S. citri R8A2 Ye et aL (1994)
spi Spiralin protein gene of S. citri R8A2 Chevalier et al. (1990)
rRNA (16S) gene of S. citri R8A2 Laigret et al. (1990)
pyrG-purA- CTP synthetase, adenylosuccinate synthetase, and Cini (1992)
purB lyase of S. citri R8A2

aMarkers without reference are from this work. All S. citri BR3 fragments were cloned in Blue-
script I1 SK + vector and maintained in E. coli DH5c~.

fragments of each line to assemble the genome map, we used randomly

selected plasmids containing cloned fragments of genomic D N A of BR3-3X,
BR3-T, or BR3-G in Escherichia coli strain DH5a and some known S. citri
genes (Table I) as probes to hybridize the PFGE-separated DNA. Some
small PFGE-separated fragments, such as BssHII H (75 kb) and SalI I (83
kb), were also purified from agarose gels and used as probes. The rod-
shaped spiroplasma virus SpV1-R8A2 was isolated from S. citri strain R8A2
by precipitation with 3.5% PEG (6000) containing 0.4 M NaC1. The 7-kb
circular single-stranded viral D N A (Renaudin et aL, 1990) was then purified
by phenol extraction and used as a probe to hybridize the restricted genomic
D N A of S. citri BR3. All the probe DNAs were used as templates for
labeling with [oL-32p]dATP, using a random priming labeling kit (GIBCO-
BRL, Gaithersburg, MD). For map assembly, BssHII and SaII fragments
reacting with the same probe were assumed to overlap each other. BssHII
fragments overlapping the same SalI fragment, or two SalI fragments
overlapping the same BssHII fragment, were assumed to be linked. Double
digestions of the chromosomal D N A with BssHII and SalI were also
conducted using the method of Pyle (Pyle and Finch, 1988). Data from both
272 Ye, Melcher,Rascoe, and Fletcher

Southern blot hybridizations and double digestions were used to establish

physical genome maps for BR3-3X, BR3-T, and BR3-G.

Cloning DNAs Mapped to the Deletion Regions of the BR3-G Genome

Differential genomic DNA screenings were carried out to clone fragments
deleted or rearranged in BR3-G. Briefly, total DNAs of BR3-3X and BR3-G
were digested to completion with EcoRI (GIBCO-BRL) and separated in
0.7% agarose. DNAs in different size ranges (0.5-2.0, 2.0-4.4, 4.4-6.6, and
6.6-9.3 kbp) were purified from the gel for both BR3-3X and BR3-G.
Purified BR3-3X DNAs of different sizes were separately ligated with
EcoRI-digested and calf intestinal phosphatase (Promega)-treated plasmid
Bluescript SK ÷ (Stratagene) using T4 DNA ligase (GIBCO-BRL). After
transformation of E. coli DH5oL, more than 300 colonies from each size
range DNA bank were screened using purified BR3-G DNA of the identical
size range as probe. Negative clones, which should contain plasmids having
fragments unique to BR3-3X in a specific size range, were selected. These
plasmids were then used as probes to hybridize EcoRI- and HindIII
(GIBCO-BRL)-digested genomic DNA of the different cell lines to confirm
their presence in BR3-3X and BR3-T and their absence in BR3-G.

Mapping and Nucleotide Sequencing of a DNA Segment

Corresponding to One of the BR3-G Deletion Regions
"Chromosomal walk" procedures were adapted to extend the cloning of one
of the regions (Deletion Area ]) deleted in BR3-G. The plasmid pE46,
resulting from the above differential screening and mapped to the deletion
region, was used as the initial probe. The cloned fragments mapped to the
deletion were subsequently "shot-gun" subcloned into M13mp18 and/or
M13mp19 vectors for DNA sequencing by the method of Sanger (Sanger et
al., 1977), using the PRISM Ready Reaction DyeDeoxy Terminator Cycle
Sequencing kit (Perkin Elmer Applied Biosystems Division, Norwalk, CT)
and an automatic sequencer (373 DNA Sequencer Stretch, Perkin Elmer).
DNA sequences were assembled using AssemblyLign software programs
(IBM-Kodak, Rochester, NY), and the consensus sequence was analyzed
using DNA Strider (Marck, 1988). The nucleotide sequence has been
deposited in GenBank as accession U44405. Potential transcription termina-
tors were identified using the Terminator application in the GCG sequence
analysis package (Madison, WI). Protein sequence databanks were searched
for sequences similar to those conceptually translated using BLASTP
(Altschul et al., 1990) and BLITZ (Sturrock and Collins, 1993). The approxi-
mate positions of the two S. citri BR3-G deletion borders were determined
Chromosome Aberrations in Spiroplasma citri 273

by Southern blot hybridization using DNA probes of the corresponding

regions. Upon hybridization, probes located within the deletion should
reveal bands present in BR3-3X and BR3-T but absent in BR3-G, whereas
probes spanning the deletion borders should reveal bands not missing but
having different sizes in BR3-G. The BR3-G fragments which were revealed
by the probes spanning the deletion borders were then cloned and se-
quenced for determination of the exact positions of the two deletion
borders.

RESULTS

Analysis by Pulsed Field Gel Eiectrophoresis

of Genomic DNA of S. citri BR3 Lines
As shown in Fig. 1A, BssHII and SalI restriction patterns of both BR3-T and
BR3-G chromosomal DNA were quite different from each other and from
those of the parent cell line, BR3-3X. The restriction products resulting
from BssHII and SalI digestion of each BR3 line were alphabetically
designated in order of their sizes and are summarized in Table II. The
approximate genome sizes of the derivatives BR3-T (1750 kb) and BR3-G
(1870 kb) are 150-270 kb larger than that of BR3-3X (1600 kb). The genome
of S. citri BR3 contains multiple spiroplasma virus SpVl-like sequences.
Hybridization of BssHII- and SalI-restricted BR3-3X, BR3-T, and BR3-G
genomic DNAs with the virus SpV1-R8A2 DNA revealed several fragments
in each line, and resulted in very different patterns among the three
spiroplasma lines (Fig. 1B).

Construction and Comparison of Physical Maps

of the Genomes ofS. citri BR3 Lines
To explore the reason for the large differences in restriction patterns, we
mapped the genomes of these lines by reciprocal BssHII and SalI double
digestion and DNA hybridization, using 20 different markers. Except for
fragments SalI F and H of BR3-3X and BR3-T or SalI E and H of BR3-G,
which are located within a large BssHII fragment, the relative positions of all
other fragments could be determined. The results also allowed the assign-
ment of the 20 DNA markers to positions on the maps (Fig. 2). The orders of
the positioned DNA markers as well as that of most BssHII and SalI
restriction fragments of BR3-3X and BR3-T were almost identical. How-
ever, some fragments of BR3-T were larger than the corresponding frag-
ments in BR3-3X, making the whole genome of BR3-T (1750 kb) approxi-
mately 150 kb larger than that of BR3-3X (1600 kb).
274 Ye, Melcher, Rascoe, and Fletcher

QI

I--
~>
X
0
o.h
I-- z<.~
no X
?t o~N
0

.~ ~

X
II1
N~
I-
X
~ o ~

•~ ~.~

o~

o~
I-
X
~ .2 ,~- ~
m

e~ X

0
ChromosomeAberrations in Spiroplasma citri 275

Table II. BssHII and SalI Fragments of the Genome of S. citri BR3 Lines

BssH II BR3-3X BR3-T BR3-G SalI B R 3 - 3 X B R 3 - T BR3-G

(band) (kb) (kb) (kb) (band) (kb) (kb) (kb)

Aa 412 440 510 Aa 290 320 390

B 310 320 386 B 280 290 270
C 258 310 240 C 215 246 260
D 170 200 180 D 170 200 224
E 130 160 160 E 170 190 190
F 125 125 150 F 150 160 165
G 120 120 125 G 130 150 160
H 75 75 120 H 110 110 130
I 83 83 83
Total 1600 1750 1871 Total 1598 1749 1872

aIn BR3-3X, a band resulting from partial digestion between fragmentsBssHI1 A and BssHII H
often appeared above fragment BssHII A, and a band from partial digestion between frag-
ments SalI C and SalI I sometimes appeared above fragment SalI A (Fig. 1A).

The genome of B R 3 - G was 270 kb larger than that of BR3-3X. The

order of the markers between A9 or E46 and A8 or X33, and between
restriction fragments B s s H I I C and B s s H I I E, was opposite to that in
BR3-3X or BR3-T, suggesting a chromosomal inversion in this region. In
addition, the overlapping fragments B s s H I I C and SalI B of B R 3 - G were
both about 10-20 kb smaller than the corresponding fragments of BR3-3X
(Fig. 1C), suggesting that a deletion of that size occurred in this region
(Deletion A r e a I). Similarly, the occurrence of a 5- to 10-kb deletion
(Deletion A r e a II) near the other border of the large chromosome inversion
was suggested by the observation that fragments BssHII F and SalI F of
B R 3 - G were 5-10 kb smaller than their counterparts in BR3-3X and BR3-T.

Cloning DNAs Mapped to the Deletions in BR3-G

By differential screening, we isolated two plasmids containing D N A frag-
ments m a p p e d to the two deletion areas in BR3-G. As shown in Fig. 3A, the
plasmid pE46, carrying a 4-kb E c o R I fragment, revealed two E c o R I frag-
ments, of 4.0 and 15.0 kb, in BR3-3X and BR3-T but only the 15.0-kb
fragment in BR3-G. It also revealed two H i n d I I I fragments, of 8.0 and 16.0
kb, in BR3-3X and BR3-T, but the 8.0-kb H i n d I I I fragment was missing in
BR3-G. These findings suggest that the 4.0-kb E c o R I and the 8.0-kb H i n d I I I
fragments overlap each other and that both were deleted in BR3-G. Because
the pE46 probe reacted strongly with two E c o R I and two H i n d I I I fragments
in both BR3-3X and BR3-T, some sequences or the whole probe fragment
must be duplicated in the genome. In BR3-3X and BR3-T, this probe is
located in the region where the 10- to 20-kb deletion (Deletion A r e a I) was
276 Ye, M e l c h e r , Rascoe, a n d F l e t c h e r

o ~ t'~ 0
~o
v--4

6 ~
~D

- <1 x

oo
</ l
d
"<lm I
I
d
,--,i
m

l
~ -~_ ~
d

m
¢'4
d
l

~ ~7-
~7-
ChromosomeAberrationsin Spiroplasma citri 277

observed in B R 3 - G (Fig. 1C). The fact that only one BssHII or SalI fragment
was revealed by this probe suggests that the two reacting E c o R I a n d / o r
HindIII fragments with possibly duplicated sequences are both located in
the region covered by fragments BssHII C and SalI B in BR3-3X orBssHII C
and SalI A in BR3-T (Fig. 2). The other isolated plasmid, pA8, revealed
three E c o R I fragments and two HindIII fragments in BR3-3X and BR3-T,
but only two E c o R I fragments and a single HindIII fragment of larger size in
B R 3 - G (Fig. 3B). As with pE46, this probe also hybridized to several bands,
indicating that some sequences or the whole probe D N A were repeated in
the genome. Upon hybridization against PFGE-separated BR3 genomic DNA,
the probe pA8 mapped to the other deletion region (Deletion Area II).

Mapping, Nucleotide Sequencing, and Genetic Organization of the DNA

Segment Corresponding to Deletion Area I in BR3-G
By using the plasmid pE46 as a probe for further screening of BR3-3X
libraries, three other plasmids--pS4, carrying a 4.0-kb S a u 3 A fragment;
pSH5, carrying a 5.0-kb Sau3A-HindlII fragment; and pA9, carrying a
1.8-kb E c o R I f r a g m e n t - - w e r e obtained. The inserts of these plasmids
covered the whole Deletion A r e a I (Fig. 4A). By Southern hybridizations,
the left-side deletion border (Deletion Border I) was located within the
1.5-kb Sau3A-HindIII fragment, because this probe recognized a 4.2-kb
S a u 3 A band in B R 3 - G rather than the 4.0-kb S a u 3 A band in BR3-3X and
BR3-T and the adjacent 4.0-kb E c o R I fragment revealed E c o R I and HindIII
bands present in BR3-3X and BR3-T but absent in BR3-G. The right-side
deletion border (Deletion Border II) was positioned within the 0.4-kb
E c o R I - X b a I fragment, because this probe revealed a 3.2-kb E c o R I frag-
ment in B R 3 - G rather than the 1.8-kb E c o R I fragment in BR3-3X and
BR3-T and the adjacent 2.6-kb E c o R I fragment revealed E c o R I and HindIII
bands present in BR3-3X and BR3-T but absent in BR3-G. Nucleotide
sequences of the four overlapping plasmids have been determined. As

Fig. 2. Physicalmaps of the genomes of S. citri BR3-3X and its derivativesBR3-T and BR3-G.
Letters symbolizethe restriction fragments listed in Table II, and positions of the DNA markers
listed in Table I are indicated above or below the restriction maps. The diamond between
fragments Sall F and SalI H in BR3-3X and BR3-T, and between SalI H and Sall E in BR3-G,
indicates undetermined relative positions of these SalI fragments within the corresponding
BssHII fragment. The two BR3-G deletions, Deletion Area I and Deletion Area II, were based
on the smaller sizesof the corresponding fragments in BR3-G with respect to their counterparts
in BR3-3X. The crossed dashed lines indicate the relative positions of the chromosomal
inversion in BR3-G. The symbolsAAIB and CA2D (preinversion borders) define the BR3-3X
segments corresponding to Deletion Area 1 and Deletion Area II and their borders before
genetic events leading to the deletions and inversion in BR3-G, and the symbolsAC and BD
(postinversionborders) define the results from exchangejoints between the borders of Deletion
Area I and Deletion Area II.
278 Ye, Melcher, Rascoe, and Fletcher

<

3:
X
o3

o
o
X
e~
O O

!
e-
,i

X 3:

n-
O
O
tlJ

O. O.

0 0
Chromosome Aberrations in Spiroplasma citri 279

illustrated in Fig. 4B, at least seven open reading frames (ORFs) were
identified within this continuous 9.6-kb sequence. Four of these ORFs, all
on the same DNA strand covering 6.6 kb (CCR) and encoding proteins of 58
kDa (p58), 12 kDa (p12), 54 kDa (p54), and 123 kDa (p123), were included
within the deletion area. A database search revealed no definitive homo-
logues for the proteins encoded by these ORFs. The incomplete, uninter-
rupted ORF at the left end (LCR) of the 9.6-kb sequence can encode 240
amino acids, 61% identical to the N-terminal sequence of the 324-amino
acid transposase polypeptide encoded by ORF3 of spiroplasma virus SpV1-
R8A2 (Renaudin et aL, 1990). The similarity in amino acid sequence
between these two ORFs is reflected in an even higher identity (66%) at the
nucleotide (nt) level. In BR3-3X, the putative initiating AUG follows 9 nt
after a GGAG sequence, a possible ribosome binding site. This site occurs
10 nt after a probable promoter sequence, TTGTCA (-35) and TATAAA
( - 10). Alignment of the predicted amino acid sequence with those of the
ORF3 of SpV1 and of related transposases from insertion elements of other
bacteria revealed that residues conserved among the transposases were also
conserved in this ORF (Fig. 5). The conservation of residues and transcrip-
tion and translation signals suggests that the ORF upstream of the deletion
in BR3-G encodes a functional transposase. However, no sequences resem-
bling the putative target sites (Renaudin et al., 1990) of the SpV1-R8A2
transposase were found in a search of the 9.6-kb sequence. An ORF
encoding a protein of 18 kDa was identified at the right end (RCR) of the
9.6-kb sequence; no definitive homologue was found for this protein in the
data banks.
No large stretches of homologous sequences were found between the
two deletion borders or their nearby sequences. In contrast, noncoding
regions of 425 bp (IGL) and 1154 bp (IGR) were found flanking the central
6.6-kb coding region (Fig. 4B). IGL contains three tightly linked oligoT-
oligoA sequences composed of TsA4, TsA6, and TsA3, and the first two runs
of these sequences can form an imperfect stem-loop structure having a stem
with 12 of 14 bp matched and a five-residue loop. Two additional possible
stem-loop structures were identified in this region. The right-side noncoding
region (IGR) contains a 208-nt stretch that has a 45% G + C content.
Embedded in this section is a 40-nt stretch of 65% G + C content. Two
sequences more than 10 residues long are repeated in this IGR region. One,
TAATTTCTTTAT, was located at nt 7999 and 8131 and had no discernible
unusual features. The other one, AAAAAAGAGAT, or minor variations of
it, occurred at three locations in that orientation and at two locations in
inverted orientation (Fig. 6). Because of the potential significance of these
repeated structures, the whole 9.6-kb sequence was searched for sequences
on either strand with a 90% or greater match to AAAAAAGAGA. One
280 Ye, Melcher, Rascoe, and Fletcher

t.o rw
e~ W d
m
,e- rw

Vl
l o - ¢- E
O
m
GO --
a
r~ \:!
if-)
w 4

l.l.l m

t.O
fig__ El-

N -- 0_

3=
El.I--

I¢.O_
O
m
¢--

D
!
Chromosome Aberrations in Spiroplasma citri 281

other sequence with paired direct and inverted repeats was identified in IGL
(Rep3; nt 844-878) and was 75% identical to either of the IGR repeats in
the central 28 nt. The sequence identified corresponds to the putative
stem-loop structure identified in the analysis of IGL.
Though there are no ATG-initiated open reading frames in IGR, there
is a 0.5-kb open frame lacking an ATG in the strand complementary to the
central coding region's coding strand. The peptide sequence that could be
encoded by this stretch, were initiation possible, contains two motifs resem-
bling sequences of the coat protein of spiroplasma virus SpV1-R8A2 (Renau-
din et aL, 1990). The first has 11 of 20 residues identical to that of the viral
coat protein. The second, within four residues of the end of the pseudo-
ORF (near Repl), has 12 of 23 residues identical. Alignment of the two
segments requires a 41-residue insertion in the pseudo ORF. At the
nucleotide level, the sequence similarity appeared to extend further, the first
stretch being 50% identical over 164 residues and the second 49% identical
over 126 residues. These identity values suggest that the sequence is a highly
diverged copy of part of the coat protein gene of SpV1-R8A2 or a related
virus.
In order to determine the positions of the two deletion borders, 4.2-kb
Sau3A and 3.2-kb E c o R I fragments (pGS4 and pGE3; Fig. 4A), containing
part of the left and right ends of the 9.6-kb sequence, were cloned from
BR3-G and partially sequenced. Sequence analysis showed that the left
1.0-kb S a u 3 A - E c o R I subfragment of pGS4 and the left end of the 9.6-kb
sequence had no mismatch until nucleotide residue 1023, after which the
two sequences began to contain mismatches. Perfect match between the
right part of pGE3 and the right end of the 9.6-kb sequence began after
residue 8198; the left part of pGE3 had no significant match to the 9.6-kb
sequence. The nonmatching sequences were not derived from or closely
related to any sequence within the 9.6-kb BR3-3X sequence, as revealed by
dot matrix comparison (Marck, 1988). These data defined the Deletion

Fig. 4. Restriction map (A) and genetic organization (B) of the 9.6-kb segment cloned from S.
citri BR3-3X and mapped to Deletion Area I in the BR3-G genome. The plasmids covering the
deletion region are represented by dark lines. The dark lines of the plasmids pGS4 and pGE3
symbolize the parts with perfect matches to the ends of the 9.6-kb sequence. The putative open
reading frames (ORFs) are shown by rectangles, with their transcription directions marked by
large arrows and the characters of gene products labeled with letters and numbers. The three
pairs of short repetitive sequences with the possibility to form stem-loop structures in the
left-side noncoding region are represented by shaded boxes, and the inverted repeat pairs in
both the left-side and the right-side noncoding regions (Repl, Rep2, and Rep3) as well as the
direct repeat not paired with an inverted repeat (Rep0.5) are indicated by small arrows.
Positions of the two deletion borders are marked. E, EcoRI; S, Sau3A; H, HindIII; X, XbaI.
LCR, left coding region; RCR, right coding region; CCR, central coding region; IGL, intergenic
region of the left side; IGR, intergenic region of the right side; Tn'ase, transposase; SpV1 CP
sim, SpV1 coat protein similar sequence.
282 Ye, Melcher, Rascoe, and Fletcher

.......... M K D , ~ , ~ < Y D E R N L F K D ~ J ~ D S CK~Hfq~fI-~S~RQT S, citri BR3

............ ~ _ ',S~AjSF M D K V K L E ~ S , , ~ I F '~KNG~Q~.j I ~ K Y L SpV1-RSA2
........... M T Y K H L T I D E L T M I E S Y Y ........ L Q H N K P V E I ~ I R M IS6770
MNMS TNYS T T N Q S Y K H L S E A E R G E I EAYL ........ S V G L K P A E ~ R R L ISII61
............ M S K H I T E E Q R Y A I S M M L ........ QI P M S K K i ~ E A I IS4351
........ M T R T K Y Q Q L Q PEERMIRI E IWK ........ A E D V S L ~ R L ISI086

[~,'SVIk~I~EI'I<-I~F}<]NI ED,%~I[~I EAP~J[Y~RKK'~I ............. S. citriBR3

~[}IR S',~,I[I~EI KRF ~T I D ~J_jS"~'~IK S I ) ~ ' _ ~ R K ~ ............. SpV1-RSA2
GRAI QTI Y N V V N K F K - - Q G K T A L D Y W H Q Y K E N K K K C G . . . . . . . . . . . . . IS6770
G R N R S T I T R E I N R G S I T Q V K K V N G A K G L L PTLLCRCCS I T V I R H A R E A S Y ISl161
GVDKS TVYRE I K R N C D A R S GSYSMELAQRIq-gDRRKQQ . . . . . . . . . . . . . IS4351
GRAP S T L M R E L R R N A T A R G G Y G A M S A Q A C R T Q R L K A S ............. IS1086

--'~KLPT[~Q[,'NF'II~, - ' ~ F f ~ H [ ~ P ~ I [ [ f ~ O j ~ L I , ~ F , , ~ I ' ~ - - ~ - C ] V ~ , S, citri BR3

- - ' . ~ - F K~_..,q~..QL N.F_I.,H~T- ~ F ~ H ~ .P.,'S.~,I:~'~f
'_ F I~.._,F[KV~..'- ~ _~,_qV._}j SpV1-R8A2
--RKVIQL PAHEVDYI KE- K - V T L G ~ P D V I IGR .... K E R P - - ~ S C G M R IS6770
YLKLD SVS D D F M R A F T D A M R E K P R V ~ T F V H T Y R L Q H V D A - -~PSTK ISI161
- - K H R K E V L T P A M R K R I I - KLLKKGF~!PEQIVGRSRLE- -GI - - ~ S H E IS4351
- -R P V A K L A P D G V L W G V V R H F L D Q K ~ P Q E I SGTL KRAF PDQ P D ~ N V S H E ISI086

' ~ .... F ~ I ~ ~ N C ~ N . . . . . R ~ K R ~ G - K ~ D E '~,, S. citri BR3

. . . . P~..._~F,~ , . . ~_~.~ H I~_ . ~ _ . . . . . . K'P~..GKS~ _NRG.K_.L~rD -,.~j SpVI-R8A2
~ L ~ R L ~ .... S K G I F - - D I D T ~ P M K G K R K P - - N G H Q E K R G K Q Q Y Q R S I-H IS6770
~L~ i.... H Q G L L E I K V I D ~ P R R V R I R K K F T K R P S T K K H L G K - - S I-E ISl161
~I~R~i~EDKRRGGK - - LHKY~RRQGRRYAKRG S KNAGRGF I PGRVD I - D IS4351
~I~N~YAYPRGELRRQL I A C ~ R Q A R T K R L P R S R G T D R R G Q I PDMVS I - H IS1086

'~S]['W-f)~--K~S~Z~'~I~Hb'I~K-SL-:'~V~VE~S~M~-E~, ' S. citri BR3

Z SI W D ! ~ S ~ ~ K D H KS - - ~ V ~ v E Q~K~..A~L E~,, SpV1-R8A2

ERPEE~NN- R S R F ~ D W ~ I ~ S ~ L ~ G K T I GE P S I ~ L ~ V E R Q ~ T ~ K ~ L V E ISl161
~RPE i ~ E i z ~ v ~ r ~ I ~ r ~ i ~ i [ ~ m < ~ - - ~ ~ s ~ C ~ i ~ s ~ ~s4~s~
VR P P E ~ D - R LMP ~ [ H W ~ G ~ I ~ G N Q S - - A~GVI~VE ~ . A ~ L V ~ PD IS1086

~ T ~ E ~ ~ K D Ii~HZ~k~I'~ ~ ~ R ~ S KWRE~EMFAET Q ~ SpV1-RSA2

R ~ I ~ ] 3 ~ T X ~ Q ~ S RF P~{Ni F F ~ ] T ' F ~ C G ~ S NWKI<~ISNQHDI D ~ IS6770
KK~E~QA.~:~- - -~ECMKLY - P I K S ~ T A ~ G ~ S S L S~ E - - - G L D ~ ISl161

A T ~ A S ~ L A G ~ T G ~ Q S LVA P - LRQ ~ T Y ~ Q G ~ H A E ~ S AATGV~ IS1086

9?????????????????9??????????????????????????????? S. citri BR3

DAG S P Q ~ P L I ~ E L~I- H ~ TD~,NK~S~KQ~~KT~ SpV1-R8A2
D PGTPS~PLN~S~I L~NG~SMD~RE~TF~S S ~ S N Q R ~ I P~ IS6770
HAYS S ~ G T ~ N F ~ L L ~ - E~GC S~KE~NL~ED~TKAI~ER P~ ISl161
]<PYH S ~ G A N ~ L I~- ~ G KD~S E ~ K Q ~ E N K L ~ N R P~ IS4351
D pH S p W ~ G T C ~ N ~ L L~- Q Y ~ T D ~ S ~ S ~ ] E E~DA]~AD S L~GR P~ IS1086

?????????????????????????? S. citri BR3

P C L ~~.S K.~E ~ L O~i
N M--9 " ~. .~. .~. .~. ~. .~ ~ ~ ~ SpV1-R8A2
KSL~PIE I~SYVQEAFYSNLI ? IS6770
RIHG~,Q~KKT,~.LTQTA?? ? ? ? ? ? ? ISn61
KRL~L~PNE~KQI INQNSVAFAS ? IS4351
KTL~PLQ~AQVLANPTDRL PVQ IS10S6
Chromosome Aberrations in Spiroplasma citri 283

844 TTACAAAAAAAAGACTTGAAAAAGTCTTTTTTTAA 878

II I [rJ[ll[l I l II III llllll
7710 ATAAAGAAAAAAGAGATTATTAAATCTCTTTTTAT 7744
[ I[ [ilrllrJlll lllllIllllll[ll i
8271 ACAATAAAA_AAAGAGATAATTAAATCTCTTTTTTT 8305
I lllllllII[llll Ill I
8848 AGAATAAAAAAAGAGAGGTTTATAAA 8868
Fig. 6. Nucleotide sequence alignment of the repeat pairs ~und in the leR (IGL) and right
(IGR) noncoding regions of the 9.6-kb sequence. The inverted repeats are underlined.
Conse~ed bases are indicated by vertical lines.

Border I at least at residue 1023 and the Deletion Border II at most at 8198,
both of which were within the two noncoding regions (IGL and IGR).

DISCUSSION
Genomic heterogeneity has been reported among different species of spiro-
plasmas, but the occurrence of extensive genomic aberrations in organisms
derived from the same isolate, such as S. cirri BR3, during a documented
time period is unexpected. The mollicute S. cirri is a plant pathogen which
specifically parasitizes both the plant and the leafhopper vectors by which it
is transmitted from plant to plant. During the 10-year period of differential
spiroplasma maintenance, BR3-T was frequently exposed to both plant and
insect, whereas BR3-G was exposed only to the plant. Whether the different
conditions for spiroplasma maintenance had some impact on the chromo-
some aberrations of S. cirri BR3 still remains uncertain, as extensive rear-
rangements occurred in both BR3-T and BR3-G. However, the organization
of the positioned markers and most of the BssHII and SalI sites are relatively
conserved between BR3-3X and BR3-T. In contrast, the large chromosomal
inversion and deletions near the two inversion borders in BR3-G make its
genome quite different from those of BR3-X and BR3-T. Data from DNA
sequencing suggested that one of the regions deleted in BR3-G contained
several functional genes. The other deletion region in BR3-G may also

Fig, 5. Alignment of selected transposase sequences with the putative transposase fragment of
S. cirri BR3. Sequences shown are of the putative transposases of spiroplasma virus SpV1-R8A2
(SwissProt P15894) and of insertion elements from Enterococcus faecalis (IS6770, GenBank
L28754), Streptococcus salivarius (ISll61, SwissProt P37245), Bacteroides fragilis (IS4351,
SwissProt P37247), and Alcaligenes eutrophus (IS1086, SwissProt P37248). Sequences from
plasmid pHKKT01 (GenBank L38972), Streptococcus salivaris (ISl139, PIR $37438), and
Alcaligenes salmonicida (GenBank L27157) were included in the alignment but omitted from
the figure due to their close similarity to depicted sequences. Residues identical in all eight
comparison sequences are indicated by reverse type. Residues similar in all eight comparison
sequences are shaded. Boxes denoted by dashed lines enclose identities between the S. cirri
BR3 transposase sequence and that of the SpV1-R8A2 virus.
284 Ye, Meicher, Rascoe, and Fletcher

contain functional genes, but that was not investigated in this study. Actu-
ally, a comparison of total proteins of S. citri BR3 lines showed that at least
one protein of 144 kDa was missing in BR3-G (Fletcher et al., 1996).
Investigation of the possibility that genetic rearrangements and DNA dele-
tions in BR3-G are related to the loss of its insect transmissibility is currently
in progress.
The extensive genetic rearrangements in S. citri BR3 lines suggest an
extreme instability of the spiroplasma genome. In addition, the genomes of
both BR3-T and BR3-G are larger than that of their parent, BR3-3X, a fact
that seems contradictory to the general concept of mollicute evolution
through genome reduction. The origin of the additional DNA in BR3-3X
derivatives is still unknown. Two possibilities are the repetition of large
DNA segments in some regions of the genome, or the multiple insertions of
extrachromosomal DNA, including the spiroplasma virus SpV1 sequence,
into the host genome. This rod-shaped virus has a circular single-stranded
DNA genome which carries a gene (ORF3 of SpV1-R8A2) encoding a
transposase resembling that of some bacterial insertion elements (IS).
SpVl-like sequences have been found at multiple sites in the genomes of
most S. citri strains (Renaudin and Bove, 1994). It was speculated that these
SpVl-like sequences in the genome of S. citri might function as repeated
elements and play a role in large-scale genomic rearrangements such as
inversions, transpositions, and deletions of large DNA segments (Ye et al.,
1992). As demonstrated in this work, the genome of S. citri BR3 also
contains multiple SpVl-like sequences. Hybridization of restricted genomic
DNA of BR3 with the viral DNA probe resulted in multiple bands and very
different patterns for the different lines, suggesting that these SpVl-like
sequences may have been directly or indirectly involved in the extensive
chromosome aberrations.
The large inversion and the deletions that led to the BR3-G genome
likely are coupled events. A simple deletion of the BR3-3X CCR should
result in a junction between IGL and IGR sequences. A subsequent
inversion should preserve such a junction. However, in BR3-G no IGL-IGR
junction was found in either the IGL- or the IGR-containing fragment
(pGS4 and pGE3). Similarily, the BR3-3X probe that recognized the pGS4
insert did not react with pGE3, and conversely the BR3-3X probe that
reacted with pGE3 did not recognize pGS4. Thus deletion followed coinci-
dentally by inversion is unlikely. Coupling could occur by the inversion
creating target sites for the deletions or by the deletions creating target sites
for the inversion. Perhaps more likely, though, is that the deletions were
created in the process of inversion. The process envisioned is one in which
sequence B of the left preinversion border (Fig. 2) joins to sequence D of the
right preinversion border as its sequence C joins to A of the left preinversion
Chromosome Aberrations in Spiroplasma citri 285

border. This possibility cannot be confirmed until the right preinversion

border and its rearranged products are characterized.
The molecular mechanisms by which the chromosomal inversion and
deletions in BR3-G occurred remain to be determined. Several short,
paired, inverted repeats were found in the two non-coding regions (IGL and
IGR). These repeats resemble the recognition sites for site-specific recombi-
nases, including invertases and resolvases found in some transposons (for a
review, see Raju and Smith, 1988). This observation suggests the possibility
that site-specific intrachromosomal recombination could have resulted in
the deletions and inversion in BR3-G. Nevertheless, site-specific recombina-
tion, either at the short inverted repeats or catalyzed by the SpVl-like
transposase, is unlikely.
Recombination junctions did not occur within the spacer regions of the
inverted repeats. A target sequence for the SpV1-R8A2 transposase was not
found within the 9.6-kb sequence. These observations do not rule out an
auxilliary role, such as binding recombination factors or aligning DNA
segments, for the SpVl-like sequences or the inverted repeats. That the two
junctions characterized are within 300 and 100 bp, respectively, of both an
inverted repeat and SpV1 similarity supports some role for these sequences
in recombination.

ACKNOWLEDGMENTS
This work was supported by the National Science Foundation (Grant E H R
9108771) and the Oklahoma Agricultural Experiment Station (Grant 2052).
Special thanks go to Dr. Matthias Ullrich, Dr. Nancy Kirkpatrick, and Dr.
Kelly Chenault for their kindness in reading the manuscript.

REFERENCES
Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990). Basic local
alignment search tool. J. Mol. BioL 215:403.
Carle, P., Laigret, F., Tully, J. G., and Bove, J. M. (1995). Heterogeneity of genome sizes within
the genus Spiroplasma. Int. J. Syst. Bacteriol. 45:187.
Chevalier, C., Saillard, C., and Bove, J. M. (1990). Organization and nucleotide sequences of
the Spiroplasma citri genes for ribosomal protein $2, elongation factor Ts, spiralin,
phosphofructokinase, pyruvate kinase, and an unidentified protein. J. Bacteriol. 172:2693.
Christiansen, G., Andersen, H., Birkelund, S., and Freundt, E. A. (1987). Genomic and gene
variation in Mycoplasma hominis strains. Isr. J. Med. Sci. 113:595.
Citti, C. (1992). Contribution a l'etude de l 'organization du genome de Spiroplasma citri: Caracteri-
sation de deux tRNAs"P chez S. citri et des genes correspondants, organization de genes de
tRNAs et identification des genes, pyrG, purA et purB, Ph.D. thesis, University II, Bordeaux,
France.
Davis, R. E. (1979). Spiroplasmas: Helical cell wall-free prokaryotes in diverse habitats. In Proc.
ROC-US Crop. Sci. Semin. Mycoplasma Dis. Plants, Natl. Sci. Council, Taiwan, ROC, pp.
59-64.
286 Ye, Melcher, Rascoe, and Fletcher

Fletcher, J., Shaw, M. E., Baker, G. R., Dugan, K. J., Ye, F., Sha, Y. H., Zuck, P., and Myers,
G. D. (1996). Protein and DNA profiles of Spiroplasma citri BR3 lines which differ in
transmissibilityby the leafhopper Circulifer tenellus. Can. J. Microbiol. 42"124.
Fletcher, J., Schultz, G. A., Davis, R. E., Eastman, C. E., and Goodman, R. M. (1981). Brittle
root disease of horseradish: Evidence for an etiological role of Spiroplasma citri. Phytopa-
thology 71:1073.
Laigret, F., Grau, O., and Bove, J. M. (1990). Comparison of 16S rDNA sequences of various
mollicutes. In Recent Advances in Mycoplasmology, Zbl. Bakt. Sup., Gustav Fisher Verlag,
Stuttgart/New York, Vol. 20, pp. 435-440.
Marck, C. (1988). "DNA strider": A "C" program for the fast analysis of DNA and protein
sequences on the Apple Macintosh family of computers. Nucleic Acids Res. 16-1829.
Pyle, L. E., and Finch, L. R. (1988). A physical map of the genome of Mycoplasma mycoides
subspecies mycoides Y with some functional loci. Nucleic Acids Res. 16:6027.
Raju, K., and Smith, G. R. (1988). Genetic Recombination, American Society for Microbiology,
Washington, DC.
Razin, S. (1989). Molecular approach to mycoplasma phylogeny. In Whitcomb, R. F., and Tully,
J. G. (eds.), The Mycoplasmas, Vol. V, Academic Press, New York, pp. 38-69.
Renaudin, J., Aullo, P., Vignault, J. C., and Bove, J. M. (1990). Complete nucleotide sequence
of the genome ofSpiroplasma citri virus SpV1-R8A2. Nucleic Acids Res. 18:1293.
Renaudin, J., and Bove, J. M. (1994). SpV1 and SpV4, spiroplasma viruses with circular,
single-stranded DNA genomes, and their contribution to the molecular biology of spiroplas-
mas.Adv. Virus Res. 44:429.
Sanger, F., Nicklen, S., and Coulson, A. R. (1977). DNA sequencing with chain-terminating
inhibitors. Proc. Natl. Acad. Sci. USA 74:5463.
Sturrock, S. S., and Collins, J. F. (1993). MPsrch Version 1.5, Biocomputing Research Unit,
University of Edinburgh, UK.
Wayadande, A. C., Shaw, M. E., and Fletcher, J. (1993). Tests of differential transmission of
three Spiroplasma citri lines by the leafhopper, Circulifer tenellus. Phytopathology 83:468.
Ye, F., Laigret, F., Whiteley, J., Citti, C., Finch, L., Carle, P., Renaudin, J., and Bove, J. M.
(1992). A physical and genetic map of the Spiroplasma citri genome. Nucleic Acids Res.
20:1559.
Ye, F., Laigret, F., and Bove, J. M. (1994a). A physical and genomic map of the prokaryote
Spiroplasma melliferum and its comparison with the Spiroplasma citri map. C.R. Acad. Sci.
Paris 317:392.
Ye, F., Renaudin, J., Bove, J. M., and Laigret, F. (1994b). Cloning and sequencing of the
replication origin (or/C) of the Spiroplasma cirri chromosome and construction of autono-
mously replicating artificial plasmids. Curr. MicrobioL 29:23.