1 Multilocus sequence analysis reveals three distinct populations of 'Candidatus Phytoplasma 1
palmicola' with a specific geographical distribution on the African continent 2
3
Fabian Pilet#a, Robert Nketsia Quaicoeb, Isaac Jesuorobo Osagiec, Marcos Freired, and Xavier
4
Foissace
5 6
a
CIRAD, UMR PVBMT, F-97410 St Pierre, La Réunion, France
7 b
Council for Scientific Research Program - Oil Palm Research Institute (CSIR-OPRI), Sekondi,
8
Ghana
9 c
Nigerian Institute for Oil Palm Research (NIFOR), Benin City, Nigeria
10 d
Instituto de Investigaçao Agraria de Moçambique (IIAM), Maputo, Mozambique
11 e
UMR1332 Biologie du Fruit et Pathologie, INRA, Université de Bordeaux, CS20032, Villenave
12
d’Ornon CEDEX, France.
13 14
Running Head:
15
Ca. Phytoplasma palmicola multilocus sequence analysis
16 17
#)Address correspondence to Fabian Pilet, [email protected]
18 19 20
Copyright © 2019 Pilet et al.
This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license.
on February 17, 2019 by guest
http://aem.asm.org/
ABSTRACT
21
To sustain epidemiological studies on coconut lethal yellowing disease (CLYD), a devastating
22
disease in Africa caused by a phytoplasma, we developed a multilocus sequence typing (MLST)
23
scheme for ‘Candidatus Phytoplasma palmicola’ based on 8 housekeeping genes. At the
24
continental level, eight different sequence types were identified among 132 ‘Ca. Phytoplasma
25
palmicola-’ infected coconuts collected in Ghana, Nigeria and Mozambique where CLYD
26
epidemics are still very active.
27
‘Ca. Phytoplasma palmicola’ appeared to be a bacterium submitted to strong bottlenecks,
28
reducing the fixation of positively selected beneficial mutations into the bacterial population.
29
This phenomenon, as well as a limited plant host range, could explain the observed
country-30
specific distribution of the eight haplotypes. As an alternative means to increase fitness, bacteria
31
can also undergo genetic exchange; however, no evidence for such recombination events were
32
found for ‘Ca. Phytoplasma palmicola’. The implications for CLYD epidemiology and
33
prophylactic control are discussed. The usefulness of seven housekeeping genes to investigate the
34
genetic diversity in the genus 'Candidatus Phytoplasma' is underlined.
35
36
IMPORTANCE
37
Coconut is an important crop for both industry and small stakeholders in many intertropical
38
countries. Phytoplasma-associated lethal yellowing-like diseases have become one of the major
39
pests that limit coconut cultivation as they have emerged in different parts of the world. We
40
developed a multilocus sequence typing scheme for tracking epidemics of 'Ca. Phytoplasma
41
palmicola', which is responsible for coconut lethal yellowing disease (CLYD) on the African
42
continent. MLST analysis applied to diseased coconut samples collected in western and eastern
43
on February 17, 2019 by guest
http://aem.asm.org/
African countries also showed the existence of three distinct populations of 'Ca. Phytoplasma
44
palmicola' with low intrapopulation diversity. The reasons for the observed genetic bottlenecks
45
resulting in strong geographic patterns remain to be established but could result from the lethality
46
of CLYD and dominance of short distance insect-mediated transmission in its spread.
47 48
INTRODUCTION
49
Molecular epidemiology requires easy-to-use tools that can sufficiently discriminate at a
50
population level. Both multilocus sequence typing (MLST) and multilocus variable number
51
tandem repeat (VNTR) analysis (MLVA) have been developed to fulfill this objective, then were
52
applied to the surveillance of human, animal and plant pathogens (1, 2). Next-generation
53
sequencing (NGS) allows the study of bacterial diversity at the genomic level through
whole-54
genome sequencing (WGS). WGS considerably increased the resolution of subpopulations of
55
Staphylococcus aureus (3), Mycobacterium abscessus (4) and Mycoplasma pneumonia (5).
56
Mycoplasma like phytoplasma are bacteria of the class Mollicutes(6). Mycoplasma are restricted
57
to vertebrate hosts while phytoplasma are plant pathogens circulative-propagative vectored by
58
phloem-feeding insects of the order of Hemiptera (7, 8).
59
The inability to grow phytoplasmas in axenic culture remains a major constraint in developing
60
WGS strategies as it limits access to adequate amounts of pure phytoplasma DNA. Only six
full-61
chromosome sequences of phytoplasmas have been deciphered to date (9), and genome surveys
62
depend on fastidious and complex molecular strategies (10-13). However, metagenomic
63
approaches through NGS opened new opportunities and 17 additional phytoplasma genome draft
64
sequences have been deposited in GenBank (NCBI). Such genome-wide sequence analyses have
65
provided easy access to genetic markers, allowed unbiased definitions of species boundaries in
66
on February 17, 2019 by guest
http://aem.asm.org/
the genus 'Candidatus Phytoplasma’ and revealed the horizontal transfer of potential mobile units
67
and effectors (14, 15).
68
MLST schemes have been developed for many Mollicutes, almost exclusively Mycoplasma and
69
Ureaplasma because of their impacts on humans and animals (16-18). For phytoplasma, most of
70
the MLST schemes that have been developed for 'Candidatus Phytoplasma' species or 16S
71
taxonomic groups target both ribosomal RNAs, housekeeping, protein coding genes and also
72
positively selected genes involved in the interaction with the host (19, 20). A single study
73
following the classical requirement of an MLST scheme that includes at least 7 housekeeping
74
genes and targets a small number of samples is currently available (2, 21). Other studies that
75
targeted 3-5 housekeeping genes have addressed taxonomical or epidemiological questions (22,
76
23). Nevertheless, primers tend to be most of the time species-specific and MLST schemes are
77
not transferable to other phytoplasma species or taxonomic groups.
78
Among the numerous diseases caused by phytoplasmas, coconut lethal yellowing diseases
79
(CLYD), also known as lethal yellowing type syndromes (LYTS), are among the most
80
destructive, with significant economic and social impact (24). CLYD are present in Africa, the
81
Caribbean and Central America, and Oceania, where they are associated with different 'Ca.
82
Phytoplasma' species according the area considered (25)
83
In Africa, CLYD are responsible for the loss of millions of coconut trees since the beginning of
84
the 20th centuries. In east Africa, CLYD were first described in 1905 in Tanzania (26), then in
85
Kenya (27). The first description in Mozambique was in 1958 in Cabo Delgado (28), and that in
86
Zambezia was in 1972 (29), growing to an epidemic in the 1990's. In West Africa, CLYD was
87
described for the first time in Nigeria as Awka Wilt Disease in 1917 (30). From the 1930s,
88
similar diseases were described in the eastern Volta region of Ghana (31), in Togo(32), and in
89
Cameroon(33). In 1964, coconut lethal yellowing was observed in western region of Ghana (31),
90
on February 17, 2019 by guest
http://aem.asm.org/
and in the 2000's in Ivory Coast (34). Molecular characterization of the phytoplasmas associated
91
with these different syndromes was only performed in few countries (Fig. S1) and two different
92
'Candidatus Phytoplasma' species described exclusively on the African continent were identified:
93
'Ca. Phytoplasma cocostanzania' (CPctz) in Kenya and Tanzania (35) and North Mozambique
94
(36), and 'Ca. Phytoplasma palmicola' (CPpml) in Ghana, Nigeria (37) and Cote d'Ivoire (38) and
95
Mozambique (39).
96
Based on the sequence and RFLP pattern of the central part of its 16S rRNA gene, CPpml is
97
classified in the group 16SrXXII (40) and is divided into two subgroups that can be differentiated
98
by HaeIII enzymatic restriction. Subgroup 16SrXXII-A includes CPpml phytoplasmas from
99
Mozambique and Nigeria, and subgroup 16SrXXII-B corresponds to CPpml from Ghana and
100
Côte d'Ivoire (37). A few studies were conducted to differentiate between populations of CPpml
101
at the national level by using different genes. The sequence of the secA gene differentiated
102
between CPpml samples from Cote d'Ivoire, Ghana and Mozambique (41, 42), while the
103
rplV/rpsC genes were able to distinguish CPpml from different Ghanaian regions (43).
104
Propagation of CPpml depends on its insect vectors, dissemination potential, alternative host
105
plants and a still unproved seed transmission. In practice sampling CLYD and insects potentially
106
vectoring the disease is also challenging due to the difficulty to access to the several meters high
107
canopy of coconut trees. Despite years of epidemiological research, the insect vectors are still
108
unidentified. The rational of this study is based on the assumption that deciphering the genetic
109
structure of CPpml populations at regional, national and continental levels and mapping the
110
distribution of its genetic variants will give insights into the origin and pathways for spread of
111
CPpml. We therefore developed an eight genes MLST scheme that follows the general
112
requirements for genotyping bacterial agents and applied this scheme to a set of 132 CPpml
113
on February 17, 2019 by guest
http://aem.asm.org/
isolates collected from three African countries where active CLYD epidemics were previously
114
associated to the two CPpml taxonomic subgroups 16SrXXII-A and 16SrXXII-B.
115 116
RESULTS
117
Housekeeping gene selection and properties 118
Eight housekeeping genes, namely, dnaC, gyrB, leuS, lpd, secA, recA, rsmI and rplV, were
119
selected to investigate the genetic diversity among 132 samples of CPpml originating from three
120
of the main African countries affected by this phytoplasma, i.e., Ghana, Nigeria and
121
Mozambique. The primers designed for this study successfully amplified the respective gene
122
targets for all 132 DNA samples of CPpml of both the 16SrXXII-A and 16SrXXII-B groups,
123
irrespective of their collection dates or geographic origins. The PCR products observed on
124
agarose gels each showed a unique and clear DNA band from 553 to 983 bp depending on the
125
target gene (Table 1). None of the 1056 double-strand-sequenced PCR products displayed double
126
peaks or ambiguous bases, thus demonstrating the specificity of the primers and the absence of
127
mixed infections.
128
The GC content of CPpml housekeeping genes (cGC%) was between 26.3 (gyrB) and 30.4%
129
(recA) (Table 2). The average cGC% for the 8 genes was 27.80.
130
The Ka/Ks values, measuring the ratio of the number of nonsynonymous substitutions per
131
nonsynonymous site (Ka) to the number of synonymous substitutions per synonymous site (Ks)
132
are presented in Table 2. Low Ka/Ks values ranging from 0.01 (gyrB) to 0.17 (lpd), indicated that
133
all the selected genes underwent purifying selection.
134
Genetic diversity of 'Candidatus Phytoplasma palmicola' 135
on February 17, 2019 by guest
http://aem.asm.org/
The number of haplotypes (H/ST), the number of segregating sites (SS), the nucleotide diversity
136
(π), and the haplotype diversity (Hd) for each housekeeping gene and its concatenated sequence,
137
Concat, are presented in Table 2 according the infected-coconut country of origin.
138
The less discriminating gene, secA, revealed 3 haplotypes, HsecA, each corresponding to a
139
different country of origin. The more discriminating gene, lpd, differentiated into 6 haplotypes,
140
Hlpd.
141
The eight genes were concatenated to obtain a Concat sequence of 4,512 bp for each isolate.
142
With 205 segregating nucleotides, the Concat sequences allowed the differentiation of 8 sequence
143
types (ST) (Table 3, Fig. 1). This combination of housekeeping gene haplotypes showed that four
144
STs are present in Ghana (ST1 to ST4), three different STs (ST5 to ST7) in Mozambique and a
145
unique ST, ST8, in Nigeria (Table 3, Fig. 1). All but ST7 were represented by several isolates.
146
The 16S rRNA of fourty one samples representing each of the eight STs were sequenced (6
147
isolates ST1, 9 ST2, 4 ST3, 5 ST4, 6 ST5, 6 ST6, 1 ST7, 4 ST8). Two types of sequences were
148
identified. Seventeen samples corresponding to ST5 to ST8 originating from Nigeria and
149
Mozambique were all identical to the 16S rRNA sequence of the LDN strain from Nigeria
150
(accession number Y14175) (44, 45) and were therefore classified as members of the subgroup
151
16SrXXII-A. Twenty-four samples of ST1 to ST4 from Ghana were all identical to the 16S
152
rRNA of CSPW-dna19 and were therefore assigned to the subgroup 16SrXXII-B (accession
153
number KF364359) (37). Sequences from subgroups 16SrXXII-A and -B differed by 6 SNPs
154
over 1,505 bp (99.6% homology).
155
Geographic distribution of the 'Ca. Phytoplasma palmicola' Sequence Types 156
All the DNA samples were spatially referenced allowing the mapping of the distribution of their
157
corresponding STs. The geographic distribution of the STs illustrate their aggregation in each
158
country (Fig. 1B). Each ST presents a continuous distribution, each constituting a well-defined
159
on February 17, 2019 by guest
http://aem.asm.org/
focus, except for ST7, which was represented by the single isolate MZ12-187. The geographic
160
origins of the 132 samples and their corresponding STs are available (Supplementary Material 1).
161
In central Ghana (CR), ST2 was the most abundant and only type present, while in the western
162
region (WR), it coexisted with ST1, a single SNP variant of ST2 (Fig. 1). ST3 and ST4
163
constituted two foci in the eastern Volta region of Ghana (VR). In Nigeria, the second western
164
African country surveyed, only ST8 was found, but only four DNA samples were analyzed. In
165
Mozambique in East Africa, ST5 and ST6 were prevalent. ST5 was restricted to the province
166
Zambezia, while ST6 only appeared along the coast of the northern province Cabo Delgado. A
167
unique ST7 sample was also localized to Zambezia. ST7 exhibited 8 to 9 SNPs when compared
168
to the two other Mozambican ST5 and ST6 (Fig. 1A).
169
Population structure of 'Candidatus Phytoplasma palmicola' 170
Tajima's D and Fu and Li's D and F were calculated for each housekeeping gene, and the
171
concatenated sequences were calculated for each country and at the African continent level
172
(Table 2). Tajima's D was positive but no significant for the concatenated sequences and
173
sequences of all the genes except rsmI from Ghana. Even though they were not significant, the
174
calculated Tajima's Ds were negative using most of the Mozambican sequences, probably
175
because ST7 was represented by a unique sample. With the exception of dnaC and rplV, all
176
Tajima's D and Fu and Li's D and F values were significantly positive at the continental level,
177
suggesting bottleneck effects.
178
An integer neighbor-joining network was calculated using the concatenated sequences of the 132
179
isolates (Fig. 1A). Of over 205 segregating sites, 202 were parsimony informative. The structure
180
of the network illustrates both strong regional differentiation and the clonal structure of the
181
CPpml populations. From the network, we observed 3 distinct populations. MLSA-concatenated
182
sequences of samples from Ghana are similarly distant from the closer Nigerian and Mozambican
183
on February 17, 2019 by guest
http://aem.asm.org/
genotypes, from which they differed by 169 SNPs (3.75%) and 171 (3.79%) SNPs respectively.
184
While from the same 16SrXXII-A subgroup, Mozambican and Nigerian samples differ by at least
185
64 SNPs (1.42%). The within-country genetic diversity was low, with no SNPs for Nigeria, 6
186
SNPs for Ghana and 9 SNPs for Mozambique.
187
Within the 16Sr subgroups, the genetic diversity was higher for subgroup 16SrXXII-A than for
188
subgroup 16SrXXII-B. In detail, the STs of subgroup 16srXXII-B split into two divergent
189
branches (Fig. 1A). Only 1 SNP differentiated STs within the same subregion, whereas
190
interregional diversity ranged from 4 to 6 SNPs. Within subgroup 16SrXXII-A, the two main
191
STs, ST5 and ST6, differed by only one mutation, while ST7 differed by 8 mutations compared
192
to the more closely related ST6 in Mozambique. In this subgroup, the Nigerian ST8 differed from
193
the Mozambican STs by a minimum of 64 SNPs, but was, however, more distant from the
194
Ghanaian STs of the other subgroup by a minimum of 138 SNPs.
195
Absence of recombination among 'Candidatus Phytoplasma palmicola' sequence types 196
Split networks were calculated from CPpml STs for each housekeeping gene (Fig. S2). The
split-197
graph structures observed for lpd and rsmI did not appear to exclude recombination phenomena,
198
phi tests of p=0.3405 and p=0.1417 respectively, were not significant irrespective of the gene
199
being considered. The phi test calculated for the Concat sequences were also not significant
200
(p=0.0748), and the RDP did not identify any recombination signals.
201
Interspecific comparison using the developed scheme 202
Seven housekeeping genes (dnaC, gyrB, leuS, lpd, secA, rsmI and rplV) and 16S rRNA gene
203
were found for 16 of 23 phytoplasma genomes or draft genomes. The 16S rRNA sequences of the
204
corresponding phytoplasmas were used to construct an ML phylogenetic tree after removing a
205
recombinant region of 155 bp. The ML phylogenetic tree presents an apparently similar topology
206
to that of the housekeeping genes but increased the strength of the phylogenetic tree (Fig. 2).
207
on February 17, 2019 by guest
http://aem.asm.org/
However, the Icong=1.1039 was not significant (P=0.35) indicating that the two calculated trees
208
were not more congruent than what is expected by chance (46).
209 210
Discussion 211
Multilocus sequence analysis highlights a genetic structure congruent with 16S rRNA 212
subgroups 213
To date, molecular characterization of CPpml was mainly based on amplification of the 16S-23S
214
rRNA genes operon using the universal phytoplasma 16S rRNA gene primers, P1/P7 (47, 48),
215
specific primers (49), or, alternatively, on the ribosomal protein gene (50) or the secA gene(51,
216
52). The MLST scheme we report is based on eight housekeeping genes and allows the
217
deciphering of CPpml genetic diversity as it could be applied to all CLYD diseased coconut
218
samples from the different country of origin and sampling dates. The protocol has been optimized
219
to avoid a nested PCR step, thereby reducing the risk of false positives. Because of their
220
specificity for CPpml and their efficiency in amplifying our entire sample set, each of the 7
221
primer pairs developed in this study could be used independently to confirm CPpml
222
diagnostically.
223
Intraspecific diversity among Mollicutes observed by MLST is variable, from very low for
224
Mycoplasma mycoides subsp. mycoides (53, 54) to high for M. agalactiae and M. bovis (18).
225
Even though available for some phytoplasmas, it has not yet been applied at large scale for
226
phytoplasmas (55, 56). Although the present MLSA scheme allowed the differentiation of only 8
227
STs of ‘Ca. Phytoplasma palmicola’, it clearly differentiated three distinct populations distributed
228
into two existing 16SrXXII subgroups. The current division of CPpml into two 16SrXXII distinct
229
subgroups (37) is consistent with the MLSA data because they correspond to the lowest similarity
230
value of the concatenated sequence of 96.21-96.25%, depending on the ST considered. The 16S
231
on February 17, 2019 by guest
http://aem.asm.org/
rRNA sequences do not allow the differentiation between the populations from Mozambique and
232
Nigeria (16SrXXII-A), while MLSA shows a dissimilarity of 1.4%. The complementarity of the
233
two methods, 16S rRNA sequencing and MLSA, for assigning strains to species or lineages has
234
been described previously (57).
235
'Candidatus Phytoplasma palmicola' is subject to bottlenecks leading to a strong geographic
236
structuration 237
Intracountry diversity of ‘Ca. P. palmicola’ measured by MLST is low, with the detection of 4
238
and 3 STs in Ghana and Mozambique, respectively. However, the identification of 3 STs in
239
Mozambique contrasts drastically with the findings of J. Bila et al. (42), who observed high
240
diversity and expanding populations of CPpml in Mozambique based on both the secA and 16S
241
rRNA genes, whereas those genes appear to be the most conserved in this study.
242
Except for the rare ST7, we observed 4 genetically related STs in Ghana versus 2 STs in
243
Mozambique. A unique ST8 was observed in Nigeria, the first African country where the disease
244
was described in 1917 (30). While only 4 Nigerian samples could be found and analyzed in this
245
study, those samples were collected from coconuts separated by distances of up to 300 km. It
246
contrasts to the situation in Ghana, where 3 to 4 STs were observed on similar distances. Only
247
genotyping a larger set of samples can confirm the lack of diversity in Nigeria.
248
The occurrence of strong bottlenecks could explain the very low diversity at the national or
249
regional scales. Coconut lethal yellowing phytoplasmas lead to the rapid and inescapable death of
250
the infected coconut in about one year. A new genetic phytoplasma variant own a short span of
251
time to appear and emerge out of the initial plant host. If such a variant appears, the probability
252
for it to be acquired by the insect vectors would be low when compared to that of the prevalent
253
original genotype unless this new variant acquires a higher multiplicative capacity in the coconut
254
or a greater ability to be acquired by the insect vector.
255
on February 17, 2019 by guest
http://aem.asm.org/
Geographic isolation of 'Candidatus Phytoplasma palmicola' disadvantages potential 256
intraspecific recombination. 257
Phytoplasma are considered to have a high level of plasticity through genome rearrangement and
258
recombination (58-60), and both the recA, when present, and PMU (Potential Mobile Units)
259
genes would play a role in adapting to different environments (61, 62). Despite the presence of
260
the recA gene in CPpml, no signs of recombination were observed among its different
261
populations. Since multiple infections have been observed for 'Ca. Phytoplasma mali' (63),
262
spatial aggregation of the different CPpml STs and low diversity at the regional level could
263
explain why we detected neither coinfection nor recombination. This distribution reduces the
264
probability that two STs would be in the same plant and hence reduces the probability of
265
observing signals of recombination. Detecting the occurrence of such a rare event of intraspecific
266
recombination would require extensive sampling where genetically distant STs share the same
267
geographic area, for example in the Zambezia province in Mozambique with ST6 an ST7.
268
A second phytoplasma infecting coconut trees in Africa and responsible for lethal decline (LD) in
269
Tanzania, 'Ca. Phytoplasma cocostanzania', has been described in both the Cabo Delgado (36)
270
and the Zambezia (42) provinces of Mozambique, where CPpml is present. Interspecific
271
coinfection has not yet been reported. The seven new primer pairs developed for the CPpml
272
MLST scheme were designed to be specific for CPpml and they are therefore not appropriate for
273
the detection of interspecific coinfections. However, interspecific coinfection could not be
274
detected through 16S rRNA or rplV gene amplification and sequencing, whilst the corresponding
275
primers can detect both phytoplasmas (50, 64).
276
New clues about 'Ca. Phytoplasma palmicola' epidemiology 277
Genotyping old historical samples would have been the best way to redraw routes of introduction
278
and evolution of CPpml. However, as phytoplasmas cannot be cultivated and as coconut cannot
279
on February 17, 2019 by guest
http://aem.asm.org/
be grafted, historical CPpml strains have not been maintained. Up to now, we could not found old
280
frozen CLYD plant material or nucleic acids.
281
CPpml STs are geographically clustered and distributed as small or large foci, suggesting a
282
gradual spread. This observation is consistent with the previous observation of aggregated
283
disease patterns and predominant dissemination within short distances (65). A low dispersion
284
capacity of the unidentified CPpml insect vector or poor acquisition or transmission efficiencies
285
could explain this clustered geographic pattern. A longer-distance dissemination pattern would
286
imply extreme weather events or nut, seedling, alternative plant hosts or plant carrying
insect-287
vectors, human-assisted long-distance movements.
288
Regarding alternative plant hosts for palm phytoplasmas, Bahder et al. (66) emits the hypothesis
289
that 'Ca. Phytoplasma palmae’ (group 16SrIV) could have been introduced in one plot in Florida
290
by planting St Augusta grass, on which Haplaxus crudus, identified as insect vector of this
291
phytoplasma, could achieve part of its life cycle as it does on many grasses (67). Studies are
292
conflicting about possible CPpml alternative plant hosts. No alternative hosts were identified in
293
Ghana (68) whereas CPpml-related strains were detected in different weed plant families in Ivory
294
Coast (69). Surveys must be undertaken in other African countries to determine if these new
295
findings are or not restricted to this particular strain of CPpml and evaluate the potential role of
296
these plant species as reservoir for the short and long distance propagation of the disease.
297
'Ca. Phytoplasma palmicola' presents a strong geographical pattern, contrary to other
298
phytoplasmas, such as fruit tree phytoplasmas in Europe(19). In fruit trees, both grafting and
299
vegetative multiplication are enhancing dissemination and genetic uniformity; however, these
300
practices do not exist for coconut trees. In general, human dissemination of coconut varieties is
301
achieved through nuts or seedlings disseminated at the local or regional level. Dissemination at
302
the continental or intercontinental scale is achieved through nuts or, more recently, coconut
303
on February 17, 2019 by guest
http://aem.asm.org/
embryos. Phytoplasmas have been detected in coconut embryos (70, 71) and transmission to the
304
seedling has been observed through the in vitro germination of embryos (72). However, the
305
natural germination of phytoplasma-contaminated seeds and transmission of the phytoplasma to a
306
viable seedling have not yet been demonstrated. Seed transmission could theoretically explain the
307
occurrence of CPpml of the subgroup 16SrXXII-A in both eastern and western Africa because
308
some nuts may have been introduced from West to East Africa (73). However, as at least 60
309
SNPs are differentiating Nigerian and Ghanaian CPpml strains from those detected in
310
Mozambique, it excludes a recent introduction in Mozambique of strain originated from areas we
311
sampled in West Africa. It cannot be ruled out that the three genetically distinct African
312
populations of CPpml could have independently emerged due to a local plant host shift of a
313
unidentified African phytoplasma. However, our data do not allow to challenge such possibility.
314
In Mozambique, ST7 is represented by a single sample and differs by at least 8 SNPs to the other
315
Mozambican CPpml STs. It is presently unknown whether several distinct introductions or a
316
single introduction followed by a local diversification have occurred in this country.
317
‘Ca. Phytoplasma palmicola’ genetic diversity has implications on LYD management. 318
The only method for the large-scale control of LYD is early eradication (74-76), and/or
319
deployment of genetic resistance (77, 78). During the last decades, many trials have been
320
performed for the identification of promising resistant coconut varieties to CLYD in Ghana (78),
321
Mozambique or more recently Ivory Coast, where resistance trials are on-going. In Ghana, trials
322
that were conducted in different locations and coconut varieties showed different disease levels
323
between trials in Agona Junction and in Axim, where ST2 and ST1 are present, respectively (78).
324
Geographical differences in disease incidence could be explained by differences in the
325
aggressiveness or the epidemic propensity of the considered ST. Trials aimed to improve coconut
326
germplasm have to include the precise identification of the strain and a multisite establishment.
327
on February 17, 2019 by guest
http://aem.asm.org/
An MLST scheme for 'Candidatus Phytoplasma palmicola': a first step towards the 328
improved control of CLYD epidemics 329
The proposed MLST scheme will enhance the resolution for better tracking of CLYD epidemics
330
at the continental, regional or provincial level. It could be implemented to identify the CPpml
331
strain that recently emerged in the Mozambican Province of Inhambane. It could also confirm or
332
refute the recently suggested introduction of CPpml in Côte d'Ivoire from Western Ghana (41).
333
The development of an expanded MLST (eMLST) by including genes involved in the
334
pathogenicity and insect transmission to the present MLST could increase its resolution . Using
335
eMLST that was expanded by adding two putative virulence genes (ureG and mba-np1)
336
improved the discrimination of Ureaplasma urealyticum strains (79). Phytoplasma genes coding
337
for surface proteins under positive selection, such as AMP/STAMP, IMP or VMP, which were
338
proven to participate in various steps of insect vector colonization, are particularly effective in
339
phytoplasma molecular epidemiology (80-87). Only an IMP homolog was found to be encoded in
340
the CPpml genome draft. Such genes evolve faster by comparison to housekeeping and therefore
341
do not fulfill the requirement define by Maiden for MLST scheme. The imp gene will be
342
evaluated in a new eMLST scheme in further investigations.
343
On a smaller scale, MLST is clearly not sufficiently discriminating to trace CPpml populations at
344
a local or field scale. VNTRs and SSRs are widely used to investigate monomorphic or poorly
345
diversified bacterial species, such as Bacillus anthracis (88). If plastic Mollicutes genomes
346
harbor high quantities of DNA repeats (89), their types and distributions are variable among
347
phytoplasmas (11). Investigating the occurrence of SSRs in phytoplasmas revealed that the
348
repetition of short motifs are overrepresented in phytoplasmas compared to other prokaryotes,
349
while long motifs are underrepresented (90). Although difficult to achieve, the identification of
350
VNTRs applicable for inoculum tracing would benefit from a larger number of sequenced
351
on February 17, 2019 by guest
http://aem.asm.org/
phytoplasma genomes. The ultimate approach is whole-genome sequencing (WGS). WGS of
352
maize bushy stunt-phytoplasma isolates from the same field successfully revealed genetic
353
polymorphisms associated with functional traits (9).
354
In conclusion, the current MLST scheme will be valuable for assigning future outbreaks to
355
existing or presently unreported genetic clusters of CPpml. The interspecific comparisons
356
presented in the current paper were based on available draft genomes. Generalization of this
357
MLST scheme would require new genomes sequencing to develop new set of primers for each
358
group or 'Candidatus Phytoplasma' species. However, it would be a pertinent approach to
359
compare the genetic diversity and identify evolutionary constraints acting on the small genome of
360
these insect-transmitted plant pathogens.
361 362
Materials and Methods
363Housekeeping gene selection and primer design 364
Highly fragmented preliminary genome drafts obtained for one Mozambican isolate of the CPpml
365
subgroup, 16SrXXII-A, and one Ghanaian isolate of the CPpml subgroup, 16SrXXII-B
366
(unpublished data), were used to select housekeeping genes and to design the corresponding PCR
367
primers.
368
Housekeeping genes were selected for their availabilities on both drafts, and their locations on
369
different contigs to avoid contiguity. The two sequences of each selected gene from the two
370
CPpml genome drafts were aligned with BioEdit 7.0 Software
371
(http://www.mbio.ncsu.edu/BioEdit/bioedit.html) to produce a consensus sequence. Consensus
372
sequences served as templates to define specific PCR primers with predicted product sizes of 450
373
to 850 bp using Primer-Blast (https://www.ncbi.nlm.nih.gov/tools/primer-blast). The seven
374
on February 17, 2019 by guest
http://aem.asm.org/
housekeeping genes selected, primer sequences, annealing temperature and PCR product sizes are
375
shown in Table 1. An eighth gene, rplV coding for the 50S ribosomal protein L22 was included
376
in the MLSA scheme because of its capacity to differentiate two strains in Ghana (43).
377
DNA Phytoplasma: Origin of the samples and DNA Extraction 378
In West Africa, Ghana and Nigeria are the only countries where the CLYD epidemics associated
379
to CPpml was active at the period of sampling (2004-2013). Sampling was extensively performed
380
in Ghana during the 2004-2009 period as the CLYD epidemics was very active. There was no
381
restriction of sampling in any country. Ghana and Nigeria were subjected to extensive surveys in
382
2009 and 2012 respectively. In Nigeria, where the disease was first reported in 1932, four
383
samples, collected in four different areas, resulted positive for CPpml. In Ivory Coast, CPpml had
384
not yet been reported at the time of the survey. A survey performed in Togo in 2006 was
385
unsuccessful and the disease was not reported during the 2004-2013 period. Similar feedback
386
were obtained from Cameroon. In East Africa, LYTS is associated to ‘Ca. Phytoplasma
387
cocostanzania’ except in Mozambique where both CPpml and ‘Ca. Phytoplasma cocostanzania’
388
could be associated to distinct disease cases. In Mozambique, survey in Cabo Delgado Province
389
was done proposedly with systematic sampling of symptomatic coconut trees. Samples from the
390
Zambezia province of Mozambique where collected during a time limited survey.
391
The MLST scheme was applied to the 132 CPpml DNA samples (Table 4). The DNA samples
392
were extracted from stem sawdust or inflorescences of LYD-infected coconuts collected between
393
2004 and 2013 in Ghana (96 samples), Nigeria (4 samples) and Mozambique (32 samples). DNA
394
was extracted using a modified CTAB protocol (39) or the DNeasy Plant Mini Kit (Qiagen,
395
Hilden, Germany).
396
Housekeeping genes and 16S rRNA gene amplification and sequencing 397
on February 17, 2019 by guest
http://aem.asm.org/
PCR amplification was performed for each of the 7 selected housekeeping genes (i.e., DNA
398
recombinase A (recA), DNA gyrase subunit B (gyrB), leucyl-tRNA synthetase (leuS),
399
dihydrolipoyl dehydrogenase (lpd), DNA replication protein (dnaC), protein translocase subunit
400
(secA), and ribosomal RNA small subunit methyltransferase I (rsmI)) plus the rplV gene for the
401
132 isolates. The rplV gene was amplified with the rpLYF/rpLYR primer pairs (50). In a subset
402
of 41 CPpml samples, 16S rRNA gene was amplified for comparison. Amplification of the locus
403
sequences were obtained by a single PCR reaction performed in 25 µl containing a final
404
concentration of 1x PCR buffer, 1.5 mM MgCl2, 0.2 mM of each deoxynucleoside triphosphate,
405
0.25 µM of each forward and reverse primer and 1.25 U of GoTaq® Flexi DNA Polymerase
406
(Promega, Madison, Wisconsin, USA). The 16S rRNA gene was amplified with the classic P1/P7
407
(48) primers. Both phytoplasma-free coconut-tree DNA and water controls were systematically
408
included as negative controls in all tests. Amplifications were performed with the GeneAmp™
409
PCR System 9700 (Thermo Fisher Scientific, Waltham, USA) with an initial denaturation at
410
95°C for 2 min followed by 35 cycles of 30 s at 94°C, 50 s at the optimal annealing temperature
411
presented in Table 1, 1 min 30 s at 72°C, and a final 6 min extension at 72°C. For P1/P7
412
amplification, the annealing temperature was 58°C for 50 s with an extension step of 1 min 30 s
413
at 72°C. The quality and concentrations of the amplified products were estimated on 1% agarose
414
gels. PCR primers were used directly for double-strand sequencing of the PCR products by
415
Genewiz (Takeley, UK). Two internal primers were designed for each strand to complete the
416
P1/P7 sequencing.
417
Forward and reverse chromatograms were assembled using Geneious R8 (Biomatters Ltd.,
418
Auckland, New Zealand), extremities were trimmed to identical lengths and to start at the first
419
base of a codon, and finally consensus sequences were exported for phylogenetic analyses.
420
Because rpLYF/rpLYR primers amplify both the rplV coding for the 50S ribosomal protein L22,
421
on February 17, 2019 by guest
http://aem.asm.org/
and for the rpsC genes coding for the 30S ribosomal protein, S3, the sequences were trimmed to
422
eliminate the short intergenic and rpsC regions. The P1/P7 16S rRNA sequences were trimmed to
423
eliminate the ITS1-tRNA-Ile-ITS2 and 23S rRNA gene sequences.
424
'Candidatus Phytoplasma palmicola' diversity analysis.
425
The number of polymorphic sites, nucleotide diversity (π), and the average cGC% were
426
calculated by using DnaSP Version 6.10. The Ka/Ks ratio, measuring the level of selection by
427
calculating the ratio of the number of nonsynonymous substitutions per nonsynonymous site
428
(Ka), to the number of synonymous substitutions per synonymous site (Ks), was calculated for
429
each housekeeping gene by using MEGA version 7.0.26 (megasoftware.net). An integer
430
neighbor-joining network (reticulation=0.5) of the 132 concatenated sequences was constructed
431
using PopART v1.7 software (91).
432
Split network and recombination analysis 433
A split network of the sequence types observed for each housekeeping gene was generated by
434
using the neighbor-net method (92) of SplitTree4 v4.14 (93) and recombination was tested for
435
each gene and for the concatenated sequences by performing the pairwise homoplasy index (phi)
436
test (94). Recombination was also estimated on concatenated sequences by RDP4 (95).
437
Interspecific analysis 438
To conduct interspecific analysis, a total of 23 full and incomplete phytoplasma chromosomes
439
available in GenBank were screened for the availability of all of the 8 MLST-selected
440
housekeeping genes and the 16SrRNA operon. As some of the drafts lack complete recA
441
sequences, the analysis was restricted to the seven other housekeeping genes (ie. dnaC, gyrB,
442
leuS, lpd, secA, rsmI and rplV).
443
The data set included ‘Ca. Phytoplasma asteris’ strains onion yellows OY-M (NC_005303.2),
444
wheat blue dwarf (AVAO01000003.1), chrysanthemum yellows strain CYP (JSWH01000164.1),
445
on February 17, 2019 by guest
http://aem.asm.org/
rice orange leaf phytoplasma (NZ_MIEP), maize bushy stunt (NZ_CP015149.1), and aster
446
yellow witches broom (NC_007716-20); ‘Ca. Phytoplasma mali’ strain AT (CU469464.1); ‘Ca.
447
Phytoplasma australiense’ strains Australian grapevine yellows (NC_010544.1) and strawberry
448
lethal yellows NZSb11 from New Zealand (NC_021236.1); ‘Ca. Phytoplasma pruni’ strains
449
milkweed yellows-MW1 (AKIL01000055.1) and Italian clover phyllody MA1
450
(AKIM01000008.1), vaccinium witches'-broom VAC (AKIN01000002.1); 16SrII-A
451
phytoplasma strain peanut witches'-broom phytoplasma NTU2011 (AMWZ01000011.1); ‘Ca.
452
Phytoplasma solani’ strain 284/09 (FO393427.1); ‘Ca. Phytoplasma pruni’ (LHCF01000001.1);
453
and ‘Ca. Phytoplasma phoenicum’ (NZ_JPSQ000). Acholeplasma palmae (NC_022538) was
454
used as an outgroup.
455
The 7 housekeeping gene sequences were aligned using Muscle (Codon) implemented in MEGA
456
version 7.0.26, then were concatenated. The maximum likelihood (ML) phylogenetic tree was
457
constructed by using the best-fitting model option of RDP4 (95), with recombinant regions
458
removed when detected by more than three methods. Bootstrap values were calculated from 1000
459
replications. The analysis was repeated with one of the two 16Sr RNA gene sequences of each
460
genome. The congruence of the phylogenetic trees from the two sets of sequences (MLSA and
461
16S) were calculated by using the Congruency index Icong (46) (
http://max2.ese.u-462
psud.fr/icong/index.help.html).
463
Accession number(s). The 1056 sequences obtained are available on the European Nucleotide
464
Archive (ENA) (https://www.ebi.ac.uk/ena/) under the project PRJEB27044 (Study ERP109081).
465
The sequences can be found under the Accession Number LR028084-LR029139
466
(http://www.ebi.ac.uk/ena/data/view/LR028084-LR029139). The 16S rRNA sequences were
467
deposited on the ENA under the project PRJEB27044 with the Accession Number
LR028043-468 LR028083 (http://www.ebi.ac.uk/ena/data/view/LR028043-LR028083). 469
on February 17, 2019 by guest
http://aem.asm.org/
Downloaded from
Acknowledgements 470
We thank O. Pruvost for helpful discussions during manuscript preparation; C. Boyer, K. Boyer,
471
S. Javegny, N. Munguambe; and S. Sisteron for technical support at the Plant Protection Platform
472
(3P, IBISA). We thank S Hogenhout, M. Kube, C. Siewert, D. Severac, and S. Carrere for their
473
contributions to obtaining the first LY phytoplasma genome drafts under the Phytoplasma
474
Sequencing Initiative of the European COST action FA0807. This work was co-funded by the
475
European Union: European regional development fund (ERDF GURDT I2016-1731-0006632),
476
the Conseil Régional de La Réunion and the Centre de Coopération Internationale en Recherche
477
Agronomique pour le Développement (CIRAD).
478 479 480 References 481 482
1. Achtman M. 2008. Evolution, population structure, and phylogeography of genetically
483
monomorphic bacterial pathogens. Annual Review of Microbiology 62:53-70.
484
2. Maiden MCJ. 2006. Multilocus sequence typing of bacteria. Annual Review of
485
Microbiology 60:561-588.
486
3. Aanensen DM, Feil EJ, Holden MTG, Dordel J, Yeats CA, Fedosejev A, Goater R,
487
Castillo-Ramírez S, Corander J, Colijn C, Chlebowicz MA, Schouls L, Heck M, Pluister
488
G, Ruimy R, Kahlmeter G, Åhman J, Matuschek E, Friedrich AW, Parkhill J, Bentley SD,
489
Spratt BG, Grundmann H. 2016. Whole-genome sequencing for routine pathogen
490
surveillance in public health: a population snapshot of invasive Staphylococcus aureus in
491 Europe. mBio 7. 492
on February 17, 2019 by guest
http://aem.asm.org/
Downloaded from
4. Bryant JM, Grogono DM, Greaves D, Foweraker J, Roddick I, Inns T, Reacher M,
493
Haworth CS, Curran MD, Harris SR, Peacock SJ, Parkhill J, Floto RA. 2013.
Whole-494
genome sequencing to identify transmission of Mycobacterium abscessus between
495
patients with cystic fibrosis: a retrospective cohort study. The Lancet 381:1551-1560.
496
5. Diaz MH, Desai HP, Morrison SS, Benitez AJ, Wolff BJ, Caravas J, Read TD, Dean D,
497
Winchell JM. 2017. Comprehensive bioinformatics analysis of Mycoplasma pneumoniae
498
genomes to investigate underlying population structure and type-specific determinants.
499
PLOS ONE 12:e0174701.
500
6. Razin S, Yogev D, Naot Y. 1998. Molecular biology and pathogenicity of Mycoplasmas.
501
Microbiology and Molecular Biology Reviews 62:1094-1156.
502
7. Chen L-L, Chung W-C, Lin C-P, Kuo C-H. 2012. Comparative analysis of gene content
503
evolution in phytoplasmas and mycoplasmas. PLoS ONE 7:e34407.
504
8. Strauss E. 2009. Phytoplasma research begins to bloom. Science 325:388-390.
505
9. Orlovskis Z, Canale MC, Haryono M, Lopes JRS, Kuo C-H, Hogenhout SA. 2017. A few
506
sequence polymorphisms among isolates of Maize bushy stunt phytoplasma associate
507
with organ proliferation symptoms of infected maize plants. Annals of Botany
119:869-508
884.
509
10. Garcia-Chapa M, Batlle A, Rekab D, Rosquete MR, Firrao G. 2004. PCR-mediated whole
510
genome amplification of phytoplasmas. Journal of Microbiological Methods 56:231-242.
511
11. Cimerman A, Arnaud G, Foissac X. 2006. Stolbur phytoplasma genome survey achieved
512
using a suppression subtractive hybridization approach with high specificity. Applied and
513
environmental microbiology 72:3274-3283.
514
12. Tran-Nguyen LTT, Gibb KS. 2007. Optimizing Phytoplasma DNA Purification for
515
Genome Analysis. Journal of Biomolecular Techniques : JBT 18:104-112.
516
on February 17, 2019 by guest
http://aem.asm.org/
13. Liefting LW, Kirkpatrick BC. 2003. Cosmid cloning and sample sequencing of the
517
genome of the uncultivable mollicute, Western X‐disease phytoplasma, using DNA
518
purified by pulsed‐field gel electrophoresis. FEMS Microbiology Letters 221:203-211.
519
14. Chung W-C, Chen L-L, Lo W-S, Lin C-P, Kuo C-H. 2013. Comparative analysis of the
520
peanut witches'-broom phytoplasma genome reveals horizontal transfer of potential
521
mobile units and effectors. PLOS ONE 8:e62770.
522
15. Firrao G, Martini M, Ermacora P, Loi N, Torelli E, Foissac X, Carle P, Kirkpatrick BC,
523
Liefting L, Schneider B, Marzachi C, Palmano S. 2013. Genome wide sequence analysis
524
grants unbiased definition of species boundaries in "Candidatus Phytoplasma". Syst Appl
525
Microbiol 36:539-48.
526
16. Brown RJ, Holden MTG, Spiller OB, Chalker VJ. 2015. Development of a multilocus
527
sequence typing scheme for molecular typing of Mycoplasma pneumoniae. Journal of
528
Clinical Microbiology 53:3195-3203.
529
17. Ghanem M, El-Gazzar M. 2016. Development of multilocus sequence typing (MLST)
530
assay for Mycoplasma iowae. Veterinary Microbiology 195:2-8.
531
18. Manso-Silván L, Dupuy V, Lysnyansky I, Ozdemir U, Thiaucourt F. 2012. Phylogeny
532
and molecular typing of Mycoplasma agalactiae and Mycoplasma bovis by multilocus
533
sequencing. Veterinary microbiology 161:104-112.
534
19. Danet JL, Balakishiyeva G, Cimerman A, Sauvion N, Marie-Jeanne V, Labonne G,
535
Laviňa A, Batlle A, Križanac I, Škorić D, Ermacora P, Serçe ÇU, Çağlayan K, Jarausch
536
W, Foissac X. 2011. Multilocus sequence analysis reveals the genetic diversity of
537
European fruit tree phytoplasmas and supports the existence of inter-species
538 recombination. Microbiology 157:438-450. 539
on February 17, 2019 by guest
http://aem.asm.org/
Downloaded from
20. Quaglino F, Kube M, Jawhari M, Abou-Jawdah Y, Siewert C, Choueiri E, Sobh H, Casati
540
P, Tedeschi R, Lova MM, Alma A, Bianco PA. 2015. ‘Candidatus Phytoplasma
541
phoenicium’ associated with almond witches’-broom disease: from draft genome to
542
genetic diversity among strain populations. BMC Microbiology 15:148.
543
21. Yong L, Chun‐gen P, Guo‐zhong T, Zhi‐xin L, Min‐wei G, Cai‐li L, Xi‐zhuo W. 2014.
544
Multilocus sequences confirm the close genetic relationship of four phytoplasmas of
545
peanut witches'‐broom group 16SrII‐A. Journal of Basic Microbiology 54:818-827.
546
22. Arnaud G, Malembic-Maher S, Salar P, Bonnet P, Maixner M, Marcone C,
Boudon-547
Padieu E, Foissac X. 2007. Multilocus sequence typing confirms the close genetic
548
interrelatedness of three distinct Flavescence Dorée phytoplasma strain clusters and group
549
16SrV phytoplasmas infecting grapevine and alder in Europe. Applied and Environmental
550
Microbiology 73:4001-4010.
551
23. Malembic-Maher S, Salar P, Filippin L, Carle P, Angelini E, Foissac X. 2011. Genetic
552
diversity of European phytoplasmas of the 16SrV taxonomic group and proposal of
553
‘Candidatus Phytoplasma rubi’. International Journal of Systematic and Evolutionary
554
Microbiology 61:2129-2134.
555
24. Gurr GM, Johnson AC, Ash GJ, Wilson BAL, Ero MM, Pilotti CA, Dewhurst CF, You
556
MS. 2016. Coconut Lethal Yellowing Diseases: A phytoplasma threat to palms of global
557
economic and social significance. Frontiers in Plant Science 7.
558
25. Yankey EN, Bila J, Rosete YA, Oropeza C, Pilet F. 2018. Phytoplasma Diseases of
559
Palms, p 267-285. In Rao GP, Bertaccini A, Fiore N, Liefting LW (ed), Phytoplasmas:
560
Plant Pathogenic Bacteria - I: Characterisation and Epidemiology of Phytoplasma -
561
Associated Diseases doi:10.1007/978-981-13-0119-3_9. Springer Singapore, Singapore.
562
on February 17, 2019 by guest
http://aem.asm.org/
26. Stein H. 1905. Die Kokosnuss und deren Bearbeitung in Deutsch-Ostafrika. Der
563
Tropenpflanzer 9:195-201.
564
27. Dowson W. 1921. Some problems of economic biology in East Africa (Kenya Colony).
565
Annals Applied Biology 8:85-87.
566
28. de Carvalho T, Mendes O. 1958. Doenças de plantas em Moçambique. Província de
567
Moçambique, Direcção de agricultura e florestas, repartição de sanidade vegetal, secçaõ
568
de patologia vegetal.
569
29. Santana Quadros A. 1972. A "doença desconhecida” do coqueiro na Zambézia. Revista
570
Agricola 14:33-4.
571
30. Johnson WH. 1918. Annual report of the agricultural department of southern provinces
572
Nigeria for the year 1917, Ibadan, 25th. Government Publication:14 p.
573
31. Chona BL, Addoh PG. 1970. Cape St. Paul wilt: the present position, p 3-8, Coconut in
574
Ghana Bulletin no 3, Chona, B.L. and Adansi, M.A. ed. Crops Research Institute. Council
575
for Scientific and Industrial Research (Ghana).
576
32. Meiffren M. 1951. Note préliminaire sur l’étude de la maladie des cocotiers au Togo.
577
Agronomie tropicale 6:163-174.
578
33. Grimaldi J, Monveiller G. 1965. La Maladie des cocotiers de Kribi. Cameroun agricole
579
pastoral et forestier 90:18-27.
580
34. Konan Konan J, Allou K, Atta Diallo H, Saraka Yao D, Koua B, Kouassi N, Benabid R,
581
Michelutti R, Scott J, Arocha-Rosete Y. 2013. First report on the molecular identification
582
of the phytoplasma associated with a lethal yellowing-type disease of coconut palms in
583
Cote d'Ivoire. New Disease Reports 28:3.
584
on February 17, 2019 by guest
http://aem.asm.org/
35. IRPCM phytoplasma/spiroplasma working group. 2004. 'Candidatus Phytoplasma', a
585
taxon for the wall-less, non-helical prokaryotes that colonize plant phloem and insects.
586
International Journal of Systematic and Evolutionary Microbiology 54:1243-1255.
587
36. Dollet M, Lourenço E, Macome F, Vaz A, Fabre S. 2012. In Mozambique, at least two
588
different phytoplasmas induce Lethal Yellowing Type Syndromes in coconut palms,
589
Congress of the International Organization for Mycoplasmology, Toulouse, France,
2012-590
07-15 / 2012-07-20. s.n., public.
591
37. Harrison NA, Davis RE, Oropeza C, Helmick EE, Narváez M, Eden-Green S, Dollet M,
592
Dickinson M. 2014. ‘Candidatus Phytoplasma palmicola’, associated with a lethal
593
yellowing-type disease of coconut (Cocos nucifera L.) in Mozambique. International
594
Journal of Systematic and Evolutionary Microbiology 64:1890-1899.
595
38. Arocha-Rosete Y, Konan Konan JL, Diallo AH, Allou K, Scott JA. 2014. Identification
596
and molecular characterization of the phytoplasma associated with a lethal yellowing-type
597
disease of coconut in Côte d’Ivoire. Canadian Journal of Plant Pathology 36:141-150.
598
39. Mpunami AA, Tymon A, Jones P, Dickinson MJ. 1999. Genetic diversity in the coconut
599
lethal yellowing disease phytoplasmas of East Africa. Plant Pathology 48:109-114.
600
40. Wei W, Davis RE, Lee I-M, Zhao Y. 2007. Computer-simulated RFLP analysis of 16S
601
rRNA genes: identification of ten new phytoplasma groups. International Journal of
602
Systematic and Evolutionary Microbiology 57:1855-1867.
603
41. Rosete YA, Diallo HA, Konan Konan JL, Yankey N, Saleh M, Pilet F, Contaldo N,
604
Paltrinieri S, Bertaccini A, Scott J. 2017. Detection and differentiation of the coconut
605
lethal yellowing phytoplasma in coconut‐growing villages of Grand‐Lahou, Côte d'Ivoire.
606
Annals of Applied Biology 170:333-347.
607
on February 17, 2019 by guest
http://aem.asm.org/
42. Bila J, Mondjana A, Samils B, Högberg N. 2015. High diversity, expanding populations
608
and purifying selection in phytoplasmas causing coconut lethal yellowing in
609
Mozambique. Plant Pathology 64:597-604.
610
43. Pilet F, Poulin L, Nkansah-Poku J, Quaicoe RN. 2011. Ribosomal protein gene sequences
611
reveal a geographical differentiation between CSPWD phytoplasmas in Ghana. Bulletin
612
of Insectology 64 (suppl.) S219-S220.
613
44. Osagie IJ, Ojomo EE, Pilet F. 2015. Occurrence of Awka wilt disease of coconut in
614
Nigeria for one century. Phytopathogenic Mollicutes 5:S61-S62.
615
45. Tymon AM, Jones P, Harrison NA. 1998. Phylogenetic relationships of coconut
616
phytoplasmas and the development of specific oligonucleotide PCR primers. Annals of
617
Applied Biology 132:437-452.
618
46. de Vienne DM, Giraud T, Martin OC. 2007. A congruence index for testing topological
619
similarity between trees. Bioinformatics 23:3119-3124.
620
47. Deng S, Hiruki C. 1991. Amplification of 16S rRNA genes from culturable and
621
nonculturable Mollicutes. Journal of Microbiological Methods 14:53-61.
622
48. Schneider B, Cousin MT, Klinkong S, Seemüller E. 1995. Taxonomic relatedness and
623
phylogenetic positions of phytoplasmas associated with diseases of faba bean, sunnhemp,
624
sesame, soybean, and eggplant. Zeitschrift für Pflanzenkrankheiten und Pflanzenschutz
625
102:225-232.
626
49. Dollet M, Fabre S, Pilet F, De Almeida Marinho VL, Quaicoe RN. 2006. Diagnosis and
627
variability in coconut lethal yellowing phytoplasmas in Africa, Annual Meeting American
628
Phytopthological Society, Cartegena, Colombie, 2006-09-12 / 2006-09-16. s.n., none.
629
on February 17, 2019 by guest
http://aem.asm.org/
50. Marinho VLA, Fabre S, Dollet M. 2006. Diagnóstico do “Coconut lethal yellowing
630
phytoplasma”, praga de quarentena para o Brasil, baseado no estudo de genes da proteína
631
ribosomal. Fitopatologia Brasileira 31:S363–364.
632
51. Obeng‐Darko SA, Quaicoe RN, Yankey EN, Twumasi P. 2018. Comparative PCR
633
analyses for the detection of the Cape St. Paul wilt disease phytoplasma in coconut palms
634
in Ghana. Journal of Phytopathology 0:1-7.
635
52. Yankey EN, Swarbrick PJ, Nipah JO, Quaicoe RN, Dickinson MJ. 2014. Detection of the
636
Cape St. Paul wilt phytoplasma in coconut palms in Ghana through the combination of
637
visual symptoms assessment and molecular diagnosis using a secA gene based assay. v.
638
96.
639
53. Dupuy V, Manso-Silvan L, Barbe V, Thebault P, Dordet-Frisoni E, Citti C, Poumarat F,
640
Blanchard A, Breton M, Sirand-Pugnet P, Thiaucourt F. 2012. Evolutionary history of
641
contagious bovine pleuropneumonia using next generation sequencing of Mycoplasma
642
mycoides Subsp. mycoides "Small Colony". PLoS One 7:e46821.
643
54. Yaya A, Manso-Silván L, Blanchard A, Thiaucourt F. 2008. Genotyping of Mycoplasma
644
mycoides subsp. mycoides SC by multilocus sequence analysis allows molecular
645
epidemiology of contagious bovine pleuropneumonia. Vet Res 39:14.
646
55. El-Sisi Y, Omar AF, Sidaros SA, Elsharkawy MM, Foissac X. 2018. Multilocus sequence
647
analysis supports a low genetic diversity among ‘Candidatus Phytoplasma australasia’
648
related strains infecting vegetable crops and periwinkle in Egypt. European Journal of
649
Plant Pathology 150:779-784.
650
56. Kakizawa S, Kamagata Y. 2014. A Multiplex-PCR Method for Strain Identification and
651
Detailed Phylogenetic Analysis of AY-Group Phytoplasmas. Plant Disease 98:299-305.
652
on February 17, 2019 by guest
http://aem.asm.org/
57. Gevers D, Cohan FM, Lawrence JG, Spratt BG, Coenye T, Feil EJ, Stackebrandt E, de
653
Peer YV, Vandamme P, Thompson FL, Swings J. 2005. Re-evaluating prokaryotic
654
species. Nature Reviews Microbiology 3:733.
655
58. Andersen MT, Liefting LW, Havukkala I, Beever RE. 2013. Comparison of the complete
656
genome sequence of two closely related isolates of ‘Candidatus Phytoplasma
657
australiense’ reveals genome plasticity. BMC Genomics 14:529.
658
59. Siampour M, Izadpanah K, Marzachi C, Salehi Abarkoohi M. 2015. Identification and
659
characterization of conserved and variable regions of lime witches’ broom phytoplasma
660
genome. Microbiology 161:1741-1751.
661
60. Wei W, Davis RE, Jomantiene R, Yan Z. 2008. Ancient, recurrent phage attacks and
662
recombination shaped dynamic sequence-variable mosaics at the root of phytoplasma
663
genome evolution. Proceedings of the National Academy of Sciences of the United States
664
of America 105:11827-11832.
665
61. Bai X, Zhang J, Ewing A, Miller SA, Jancso Radek A, Shevchenko DV, Tsukerman K,
666
Walunas T, Lapidus A, Campbell JW, Hogenhout SA. 2006. Living with genome
667
instability: the adaptation of phytoplasmas to diverse environments of their insect and
668
plant hosts. Journal of Bacteriology 188:3682-3696.
669
62. Tran-Nguyen LTT, Kube M, Schneider B, Reinhardt R, Gibb KS. 2008. Comparative
670
genome analysis of 'Candidatus Phytoplasma australiense' (subgroup tuf-Australia I;
rp-671
A) and "Ca. Phytoplasma asteris" strains OY-M and AY-WB. Journal of Bacteriology
672
190:3979-3991.
673
63. Seemüller E, Kiss E, Sule S, Schneider B. 2010. Multiple infection of apple trees by
674
distinct strains of ‘Candidatus Phytoplasma mali’ and its pathological relevance.
675 Phytopathology 100:863-870. 676
on February 17, 2019 by guest
http://aem.asm.org/
Downloaded from
64. Marinho VLA, Fabre S, Dollet M. 2008. Genetic variability among isolates of Coconut
677
lethal yellowing phytoplasmas determined by Heteroduplex Mobility Assay (HMA).
678
Tropical plant pathology 33:377-380.
679
65. Bonnot F, de Franqueville H, Lourenco E. 2010. Spatial and spatiotemporal pattern
680
analysis of Coconut Lethal Yellowing in Mozambique. Phytopathology 100:300-312.
681
66. Bahder BW, Helmick EE, Chakrabarti S, Osorio S, Soto N, Chouvenc T, Harrison NA.
682
2018. Disease progression of a lethal decline caused by the 16SrIV-D phytoplasma in
683
Florida palms. Plant Pathology 67:1821-1828.
684
67. Ramos Hernández E, Magaña Alejandro MA, Ortiz García CF, Oropeza Salín C, Lesher
685
Gordillo JM, Sánchez Soto S. 2018. The coconut pathosystem: weed hosts of nymphs of
686
the American palm Cixiid Haplaxius crudus (Hemiptera: Fulgoroidea). Journal of Natural
687
History 52:255-268.
688
68. Yankey EN, Pilet F, Quaicoe RN, Dery SK, Dollet M, Dzogbefia VP. 2009. Search for
689
alternate hosts of the coconut Cape Saint Paul Wilt Disease pathogen. OCL 16:123-126.
690
69. Rosete YA, Diallo HA, Konan Konan JL, Kouamé AEP, Séka K, Kra KD, Toualy MN,
691
Kwadjo KE, Daramcoum WAMP, Beugré NI, Ouattara BWM, Kouadjo Zaka CG, Allou
692
K, Fursy-Rodelec ND, Doudjo-Ouattara ON, Yankey N, Dery S, Maharaj A, Saleh M,
693
Summerbell R, Contaldo N, Paltrinieri S, Bertaccini A, Scott J. 2016. Detection and
694
identification of the coconut lethal yellowing phytoplasma in weeds growing in coconut
695
farms in Côte d’Ivoire. Canadian Journal of Plant Pathology 38:164-173.
696
70. Cordova I, Jones P, Harrison NA, Oropeza C. 2003. In situ PCR detection of phytoplasma
697
DNA in embryos from coconut palms with lethal yellowing disease. Molecular Plant
698 Pathology 4:99-108. 699
