AEM Accepted Manuscript Posted Online 5 June 2015 Appl. Environ. Microbiol. doi:10.1128/AEM.01399-15 Copyright © 2015, American Society for Microbiology. All Rights Reserved.
1 2 3 4 5 6
Revised AEM01399-15 Survival and Competitiveness of Bradyrhizobium japonicum Strains Twenty Years after Introduction into Field Locations in Poland
7 8
Dorota Narożnaa, Krzysztof Pudełkoa, Joanna Króliczaka, Barbara Golińskaa,
9
Masayuki Sugawarab, Cezary J. Mądrzaka,† and Michael J. Sadowskyb,*,†
10 11
a
12
11, 60-632 Poznań, Poland
Department of Biochemistry and Biotechnology, Poznań University of Life Sciences, Dojazd
13 14
b
15
Minnesota, St. Paul, Minnesota, USA
Department of Soil, Water, & Climate, and BioTechnology Institute, University of
16 17 18 19 20 21 22 23 24
Running Title: Soil Survival of Bradyrhizobium japonicum after 20 Years.
25 26
†
These authors contributed equally to the work.
27 28
*Address correspondence to Michael J. Sadowsky, BioTechnology Institute, 1479 Gortner
29
Ave, 140 Gortner Labs, St. Paul, MN 55108; Email: [email protected]; Phone (612) 624-
30
2706
31 32 33 34
Supplemental material for this article has been prepared.
35
1
36 37 38
Abstract It was previously demonstrated that there are no indigenous strains of Bradyrhizobium
39
japonicum forming nitrogen-fixing root nodule symbioses with soybean plants in arable field
40
soils in Poland. However, bacteria currently classified within this species are present (together
41
with B. canariense) as indigenous populations of strains specific for nodulation of legumes in
42
the Genisteae tribe. These rhizobia, infecting legumes such as lupins, are well established in
43
Polish soils. The studies described here were based on soybean nodulation field experiments,
44
established at the Poznan University of Life Sciences Experiment Station in Gorzyń, and
45
initiated in the spring of 1994. Long term research was then conducted in order to study the
46
relation between B. japonicum USDA 110 and USDA 123, introduced together into the same
47
location, where no soybean rhizobia were earlier detected, and nodulation and competitive
48
success were followed over time. Here we report the extra-long-term saprophytic survival of
49
B. japonicum strains nodulating soybeans that were introduced as inoculants 20 years earlier
50
and where soybeans were not grown for the last 17 years. The strains remained viable and
51
symbiotically competent, and molecular and immunochemical methods showed that strains
52
were undistinguishable from the original inoculum strains USDA110 and USDA123. We also
53
show that the strains had balanced numbers and their mobility in soil was low. To our
54
knowledge, this is the first report showing the extra long term persistence of soybean-
55
nodulating strains introduced into Polish soils and the first analyzing long term competitive
56
relations of USDA 110 and USDA 123 when both strains were not native and were introduced
57
to the environment almost two decades ago.
2
58 59
Introduction Soybean (Glycine max (L.) Merr.) is still an emerging crop in Poland and only about
60
2,800 hectares is presently being planted. However, there has been slow but systematic
61
increase in production and this is consistent with estimations made earlier by soybean
62
breeders (1). Grown only on a small scale in a few isolated locations, soybeans still remains
63
interesting for farmers and its cultivation is considered as a potential supplemental source of
64
plant proteins for feeding livestock.
65
Despite this, there are several registered soybean cultivars well suited to local
66
environmental conditions, including: Aldana and Augusta (from Poland), Merlin and
67
Lissabon (from Austria), and Annushka and Mavka (from the Ukraine). Recently, qualified
68
seed production of the above cultivars has dramatically increased from 2 ha in 2009 to almost
69
600 ha in 2013, which further indicates farmer’s interest in soybean production in Poland
70
(Jerzy Nawracała, personal communication).
71
Relatively little information is available concerning the presence of symbiotic,
72
nitrogen-fixing bradyrhizobia for soybean in Poland. There are no indigenous Bradyrhizobium
73
japonicum strains forming nitrogen-fixing, root nodule, symbioses with soybeans in arable
74
lands in Poland and thus far, only a few local populations have been identified in a few
75
locations where inocula were applied for growing soybean (1). It should be mentioned,
76
however, that bacteria initially classified as B. japonicum are present (together with B.
77
canariense) as indigenous populations of strains specific for legumes of the tribe Genisteae.
78
These rhizobia, infecting lupins, are well established in Polish soils. They have colonized the
79
area presumably migrating together with their hosts from the Mediterranean after the retreat
80
of glaciers during the last 10,000 years (2). These bacteria, however, cannot nodulate
81
soybeans.
3
82
The lack of naturally occurring soybean microsymbionts in Polish soils provides
83
unique opportunities to study the persistence, stability, competitiveness, and dispersal of
84
bradyrhizobia introduced as inocula. It also opens the possibility to analyze the survival of
85
rhizobia as soil saprophytes in the absence of host plants. This issue is particularly relevant
86
since despite the fact that rhizobia do not form resting stage spores (3), they still are
87
characterized as persistent autochthonous microbes in soils where they have been introduced.
88
Moreover, there are relatively few studies on the long term survival of rhizobia in the absence
89
of a legume host and in foreign environments, and those without naturally occurring host
90
plants. For soybeans, a few of these studies were done in France (4, 5) and Brazil (6, 7), but in
91
some there was repeated inoculation due to die off.
92
The competitiveness of strains infecting soybeans may be strongly dependent on
93
environmental conditions, and presuming that they are indigenous, are likely to be more fit to
94
interact with host plant genotypes than more recent bacterial and host additions to soils. Based
95
on these assumptions, we designed experiments to study the competitive relationship between
96
B. japonicum strains USDA 110 and USDA 123, introduced together into a field location
97
where no soybean rhizobia were previously detected. While it has been shown that USDA 123
98
outcompeted USDA110 in North American soils comprised of several Bradyhizobium
99
serotype strains (8, 9, 10), no information is available on how these bacteria compete for
100
nodulation in natural field soils devoid of indigenous soyean-nodulating bradyrhizobia.
101
Here we report results of field experiments initiated in 1994 where soybean was sown
102
into the soil after inoculation with B. japonicum USDA 123 and USDA 110, either separately
103
or as a mixed inoculant. Soybean was then grown in this location for three subsequent years
104
and the field was used for growing other crops for the next 17 years. After this time,
105
nodulation results showed the extra-long-term saprophytic survival of B. japonicum strains
4
106
nodulating soybeans and long term competitiveness of USDA 110 and USDA 123 in natural
107
soils initially devoid of soybean-nodulating bradyrhizobia.
108 109
Materials and Methods
110 111
Initial field experiments 1994-1997. A soybean nodulation field experiment was established
112
at the Poznan University of Life Sciences Experiment Station in Gorzyń, about 80 km North-
113
West from Poznań, Poland, in the spring of 1994 on soils characterized as Arenic Hapludalfs
114
or Albic Luvisols (Arenic) (11, 12). The selected field site was free of indigenous strains
115
capable of nodulating soybeans (1, and unpublished data). Prior to the first sowing of surface
116
sterilized seeds (G. max (L.) Merr. cv Naviko), separate plots were inoculated with USDA
117
123, USDA 110, or a mixture containing equal cell concentrations of both strains. Inoculation
118
was performed using a peat-based inoculum as previously described (13). The seeds were
119
inoculated using peat inoculum (100 g inoculum per 1 kg of seeds) and gum arabic as a
120
sticker. The mixed inoculation was performed using equal amounts of USDA 110 and USDA
121
123 at a final concentration of 2 x 106 cells per seed. In the case of mixed inoculation half of
122
this number was USDA 110 and half was USDA 123. An uninoculated control plot was also
123
included into the experiment and each experimental plot was replicated four times. No
124
inoculations were performed during subsequent experimental seasons. The experimental field
125
is a part of a larger experimental area, where field experiments were conducted for at least 50
126
years (Supplemental Fig. S1). The crops cultivated in this particular part of experimental
127
station were; wheat, barley, triticale, potatoes, and yellow and narrow-leafed lupins.
128
Nodulation of soybean plants was analyzed 6 wk after germination and both the
129
number and fresh weight of nodules were measured. Strain occupancy of nodules was
130
determined by using an immuno-spot-blot method and cross-adsorbed, strain-specific,
131
antibodies as previously described (1, 14). The experiment was repeated over four subsequent
132
seasons, until 1997, and after this time soybeans were not grown at this location.
133 134
Detection of soybean-nodulating bradyrhizobia 17 - 20 years after their introduction
135
into Gorzyń soil. Twelve soil samples were obtained from the original field experiment site 5
136
in September 2010. Surface sterilized soybean (G. max (L.) Merr. cv. Augusta) seeds were
137
sown in duplicate pots filled with these soil samples. Four seed were sown in each pot. Plants
138
were grown in the greenhouse for 5 wk.
139
Independently, a trap host approach (15) was used directly in the field to determine the
140
presence of soybean-nodulating bacteria and to determine the distance of their dispersal from
141
their release site. For this purpose, in the spring of 2011, surface sterilized soybean seeds (G.
142
max (L.) Merr. cv. Augusta) were sown in the spots forming two perpendicular lines of trap
143
hosts across the former field experiment site used in 1994. Plants were harvested and root
144
nodules were collected 6 wk after sowing. The experimental design is shown in Fig.1. In
145
2014, soybean was grown on the entire area of the initial experimental site as indicated above.
146 147
Serological identification of bacteria from soybean root nodules. Nodules were collected
148
from plants grown in the greenhouse after exposure to soil taken from the field site in
149
September 2010. Four plants were taken from each of 12 soil samples. All nodules from each
150
plant were separately collected and dried at 60oC for 48 hr. The 48 samples collectively
151
produced 1,266 nodules, all of which were analyzed for the presence of strains USDA 123
152
and USDA 110 by using the immuno-spot-blot method and cross-adsorbed, strain-specific,
153
antibodies as previously described (1, 14).
154 155
Isolation of bacteria from soybean root nodules. Bacteria were obtained from root nodules
156
surface sterilized with 95% ethanol and 3% sodium hypochlorite, according to Somasegaran
157
and Hoben (15). Purified bacterial isolates were obtained by streaking nodule suspensions
158
onto AG agar medium (16) containing 50 µl/ml cycloheximide, followed by two to three
159
successive isolations for single pure colonies on the same medium.
160
6
161
DNA isolation and PCR template preparation. Total genomic DNA was isolated from
162
single bacterial colonies by using the GenElute Bacterial Genomic DNA Kit (Sigma-Aldrich
163
Corporation, St. Louis, MO, USA) following the manufacturer's instructions. Templates for
164
PCR amplification were lysates of crushed soybean nodules stored in 25% glycerol at -80oC.
165
A 50 µl aliquot of PrepMan Ultra buffer (Applied Biosystems, Grand Island, NY, USA) was
166
added to 50 µl of crushed suspension. The resulting mixture was heated for 5 min at 95°C,
167
centrifuged for 5 min at 14,000 rpm, and the supernatant was used as a template for PCR.
168 169
Molecular identification of bacteria from soybean root nodules. Genomic DNA
170
preparations were obtained from bacteria isolated from nodules from the September, 2010
171
experiment. Three nodules were taken from each soil sample giving 36 DNA samples. Three
172
independent PCR analyses were performed for DNA from each isolate for the: (i) identi-
173
fication of USDA 110- specific sequence(using strain-specific primers), (ii) analysis of recA
174
gene sequences in order to distinguish between B. japonicum strains USDA 123 and USDA
175
110, and (iii) analysis of strain diversity using RP01 primers.
176
A B. japonicum USDA 110 strain-specific sequence was selected based on an in silico
177
search of the USDA 110 genome. Sequence BAC45285.1, within genomic position 17139-
178
18371, is apparently specific for B. japonicum strains USDA 110 and USDA 6. The PCR
179
primers were designed based on this sequence: Bj110F 5’-TAGGTGATGGGAT
180
AACGCTCT-3’ and Bj110R 5’-GTCCAGCTTCGCG ATCGACCGG-3’ (Suppl. Figure
181
S2). The primers amplify a 199 bp DNA fragment, between genomic positions 17151 and
182
17331, encoding a protein of unknown function.
183
In order to reconfirm the identification of USDA 110 vs USDA 123 genotypes, a
184
common phylogenetic marker - the sequence of recA was used. The nucleotide sequence of
185
the recA gene from USDA 110 and USDA 123 (collection strains) was determined.
7
186
Alignment of their 559 bp-long fragments allowed design of specific forward primers: recA
187
110 5’-GGGCTCGACATTGCGCTC-3’ and recA 123
188
3’ which are complementary to the fragment 135-151 of the recA sequence (Suppl. Fig. S3).
189
A common reverse primer, recA 549R 5’-TTGCGCAGC GCCTGGCTCAT-3’, was
190
designed for the conserved regions of many bradyrhizobia (nucleotide position 530 to 549).
5’-GGGCTCG ACATTGCACTG-
The verification of specificity of the recA primers was done for a collection of USDA
191 192
110 and USDA 123 strains and for randomly selected strains isolated from Gorzyń soil. The
193
diversity of isolates identified as related to USDA 110 or USDA 123 strains were analyzed
194
using the RP01 PCR (5’-AATTTTCAAGCGTCGTGCCA-3’) primer according to
195
Richardson et al. (17). The PCR was carried out using Allegro Tag DNA Polymerase
196
(Novazym, Poznań, Poland) according to the following conditions: initial denaturation 95°C
197
for 2 min , 30 cycles of 30 s at 95°C, 30 s at 56°C for Bj110F, Bj110R, recA 110, recA 123,
198
and recA 549R, or 52°C for RP01,1 min at 72°C, and a final elongation step of 5 min at
199
72°C.The PCR products were analyzed on 1% agarose gels.
200 201
Enumerating bacteria by using MPN analyses. Determination of the number of bacteria
202
capable of nodulating soybeans was performed by using the most probable number (MPN)
203
technique as described by Somasegaran and Hoben (15). Soil samples were collected in
204
Gorzyń on October 25, 2012 and stored at 4°C. An equivalent of 100 g dry mass of soil
205
(109.4 g fresh weigh) was suspended in 900 ml of water and shaken intensively for 5 min.
206
The resulting suspension was diluted and 1 ml of each dilutions was added to surface
207
sterilized soybean seeds sown in Leonard jar assemblies (15). Five replicates were used for
208
each dilution, and a negative control consisting of uninoculated surface sterilized seeds, was
209
used. Nodulation was recorded after 34 d and MPN values determined as previously described
210
(15).
8
211
Enumerating bacteria in soils using strain-specific fluorescent antibodies. The
212
identification and counting of the B. japonicum USDA 110 and USDA 123 cells were done
213
using strain-specific fluorescent antibodies (FA) as described by Somasegaran and Hoben
214
(15) and Schmidt et al (18). Two soil samples were analyzed and two filters were prepared by
215
filtering a 10 ml of solution from each sample (see Suppl. File 1). The number of bacteria
216
visible in 30 microscopic fields on each filter was determined, three times, using a Zeiss
217
Axiovert 200 Inverted microscope with a 100 x immersion lens and Filter set 09 (exc. BP
218
450-490, beam splitter FT 510, emission LP 515). No cross-reacting soil bacteria were
219
observed.
220 221
Statistical analyzes. Results of all studies were evaluated for significance by using the R
222
statistical software version 3.1.2 (19). ANOVA tests were followed by post hoc analysis with
223
Fisher's LSD test from agricolae package (20), at a significance level of α=0.05.
224 225 226
Results
227 228
Detection of soybean-nodulating bacteria 20 years after their introduction into Gorzyń
229
soil.
230
in1994 contained soybean rhizobia. Root nodules were found on all soybean plants grown in
231
the greenhouse in 24 pots containing 12 soil samples. Results of immuno-spot-blot analyses
232
revealed the presence of strains USDA 110 and USDA 123, in balanced numbers. Data shown
233
in Fig. 2A indicates that USDA 123 prevailed in soils taken from five locations (1, 3, 4, 6 and
234
12), with considerable domination in location 12 (72.4%). The soil samples from locations 5,
235
8, and 10 contained more USDA 110 than USDA 123 cells. Location 5 had a significant
236
prevalence of USDA 110 strains (85.6%). The remaining samples (from locations 2, 7, 9, and
All soil samples taken from the exact same location of the field experiment conducted
9
237
11) contained about equal numbers of both strains. Taken together, 587 nodules (46.4%) were
238
occupied by strain USDA110, 653 (51.6%) by USDA123, and 26 nodules (2%) contained
239
both strains (Fig. 2B). Only three locations did not contain any doubly occupied nodules
240
(Suppl. Table S1). Another 153 nodules were saved as crushed suspensions. Thirty-six
241
suspensions, representing nodules from three plants from each of the 12 soil samples, were
242
randomly chosen and used to obtain pure isolates. These isolates (UPP401 to UPP436) were
243
stored in 50% glycerol at -80oC.
244
A plant trap approach was used to determine the distance from the exact site of initial
245
inoculation where bradyrhizobia still could be found. As shown in Fig.1 and Table 1,
246
successful nodulation was observed in all tested soils samples, but not those located further
247
than about 40 m westward from the edge of former experiment site. This result suggests that
248
there was rather poor dispersal of inoculum strains. In 2014 soybean cultivation was
249
considered to serve as the ultimate confirmation of the presence of soybean-nodulating
250
bradyrhizobia. Nodulation was detected within the entire experimental area. No further
251
analyses were performed.
252
The identification of strains occupying nodules was performed using the PCR assay
253
with USDA 110-specific primers. Data in Table I shows that both strains were found
254
dispersed within the entire area, revealing a fairly balanced proportion of both bacterial
255
genotypes.
256 257
Molecular identification of bacteria from soybean root nodules. The identity and diversity
258
of each of the 36 purified isolates were determined by using PCR and using USDA 110 strain-
259
specific and RP01 primers. Results verifying the specificity of USDA 110-specific primers
260
are shown in Fig. 3. The amplicon of the expected size of 199 bp appeared when lysed cells of
261
B. japonicum USDA 110 were used as the template for PCR amplification (Fig. 3 lane 1). No
10
262
specific products were generated with DNAs of other bacteria tested, including B. japonicum
263
USDA 123 (Fig. 3, Lane 2), and lupin nodulating strains of B. japonicum UPP323 and
264
UPP344 (2), Bradyrhizobium sp. UPP331 and UPP370, Bradyrhizobium sp. USDA 3045
265
(Lupinus) as well as B. elkanii resembling strain UPP372 and B. yuangmingense, resembling
266
strain UPP253. Based on this data, primers Bj110F and Bj110R were considered to be
267
specific for B. japonicum strain USDA 110 and they were used to characterize purified
268
isolates obtained from root nodules of soybean plants grown in soil from Gorzyń. PCR results
269
in Fig. 4 show that 15 of 36 isolates (42%) could be presumed to be the descendants of the
270
original inoculum strain B. japonicum USDA 110.
271
PCR amplifications done with recA-specific primers were used to reconfirm the
272
identification of B. japonicum isolates. Data in Fig. 5 show the specificity and usefulness of
273
these primers. The DNA of the USDA123 collection strain, and isolates identified as the
274
presumed descendants of this strain (UPP405 and UPP406 - compare Fig. 4 lanes 5 and 6)
275
did not generate PCR amplification products when the recA110 primer was used (lanes 4, 5,
276
and 6). Conversely, the recA123 primer did not produce PCR amplification products with
277
template DNA from USDA110 and their presumed descendants (UPP401 and UPP403).
278
The diversity of isolates identified as related to USDA 110 or USDA 123 strains was
279
analyzed by using the RP01 PCR and appeared to be low, since only two types of profiles
280
were found in Fig. 6 consistently representing the two strains identified. Among 36 strains
281
analyzed, UPP 403, UPP 404, UPP 407, UPP 408, UPP 401, UPP 409, UPP 410, UPP 412,
282
UPP 413, UPP 416, UPP 419, UPP 422, UPP 428, UPP 431, UPP 434, (shown in the lanes 1,
283
2, 5, 6, 9, 12, 13, 17, 18, 19, 20, 22, 28, 31, and 34, respectively) represent the profile typical
284
for strain USDA 110. The remaining profiles were characteristic of B. japonicum strain
285
USDA 123. The fragments in lane 16 are likely due to a nodule (used as template source) that
286
was occupied by both strains.
11
287
Enumerating bradyrhizoba in soils. The MPN technique was used to quantify the number
288
of bradyrhizobia in Gorzyń soil sample capable of nodulating soybeans. Results indicated that
289
this soil contained about 1 x103 bradyrhizobia per gram soil (dry weight); The number of
290
bradyrhizobia in duplicate Gorzyń soil samples was also evaluated by using direct
291
immunofluorescence (18). Results of fluorescent antibody (FA) analyses indicated that, on
292
average in 2012, the soil contained 1.2 x 104 and 6.9 x 103 cells per gram of USDA 123 and
293
USDA 110, respectively, after 15 years of no host plants (Suppl. Table S2).
294 295 296 297
Discussion These experiments were designed to study the interrelationship of well-known
298
competing B. japonicum strains (USDA123 and USDA 110), introduced to a foreign
299
environment, in 1994-1997. The mixed inoculation experiment revealed that initially USDA
300
110 prevailed over USDA 123. About 70% of nodules were occupied by this strain as was
301
established by using an immuno-spot-blot method and cross-adsorbed strain-specific
302
antibodies. However, good performance by USDA 110 was also observed during the
303
following years. Due to natural reasons, the areas occupied by both strains have expanded
304
leading to an overlap of their initial sites of inoculation. The domination of USDA 110 was
305
not as pronounced during the third and fourth year after inoculation, but both strains showed a
306
good survival rate (unpublished data). These observations allowed us to determine: (i) the
307
persistence (survival) of symbiotic bacteria (bradyrhizobia) without a host legume; (ii) the
308
genetic stability of inoculant bacteria in the soil environment, and (iii) the mobility of
309
particular genotypes – the spatial range of bacteria spread from the site of their introduction
310
into soil.
311
Addressing the first issue, our findings clearly indicate that both strains introduced as
312
the inocula remained viable and active as soybean microsymbionts for over 20 years after the 12
313
only inoculation in May 1994, and subsequent growth of soybean for three additional years
314
(1994-1997). This results is in contrast to that of Dunigan et al. (21) who showed that three
315
years of repeated inoculation was required to establish B. japonicum in a Louisiana soil. Also
316
the number of bacterial cells capable of nodulating soybean long time after inoculation was
317
surprisingly high for the introduced strains.
318
Their abundance estimated by the MPN method (~ 103 g-1) was very similar to the
319
number of Bradyrhizobium sp. strains nodulating lupins and native to Polish soils. Martyniuk
320
(22) classified soils containing over 1000 cells in 1g of soil as highly populated by these
321
rhizobia. This observation should be interpreted considering particular groups of diazotrophic,
322
bacterial, symbionts. One should remember that their presence in the soil is usually detected
323
by the plant-trap method. The difference in our results and those of previous studies may be
324
due to the fact that others have introduced rhizobia into soils already containing an indigenous
325
population of these bacteria.
326
Our estimation of cell numbers based on the FA method suggested 20-times more
327
rhizobial cells than that determined by MPN. This may be due to the presence of viable but
328
not culturable (infectable) bradyrhizobia. On the other hand, the immunofluorescence
329
technique tends to overestimate soil bacteria population sizes (23). This is perhaps the reason
330
for the relatively common opinion that the presence of host plants is necessary for the survival
331
of their respective microsymbionts. The chromosomal localization of symbiotic genes in the
332
genus Bradyrhizobium may explain the reason for the fact that the relative good persistence in
333
soil is reported for the symbiotic traits of bacteria belonging to this group. However, the
334
survival period of B. japonicum as saprophytes was reported to be 8-13 years (4, 5), and
335
environmental conditions also remains an important factor governing the presence of
336
indigenous strains. Very poor survival of inoculant strains of Bradyrhizobium sp. (Cajanus)
337
was reported (24). This was most probably due to high soil temperature and fluctuating soil
13
338
moisture. None of these factors were present in the Polish soil used in this study and there
339
were no indigenous strains of soybean-nodulating B. japonicum. Thus, as has been pointed
340
out by others, and is standard practice in some countries (e.g. New Zealand), decisions on the
341
application and type of inoculum strain should be made carefully. As has been discussed
342
earlier (25, 26), both edaphic conditions and existing bacterial populations strongly affect
343
inoculation success.
344
Considering the second issue, isolates which were currently obtained from the root
345
nodules of soybean plants grown in the soil from the experimental site in this study were very
346
likely descendants of the original inoculum strains (B. japonicum USDA110 and USDA123).
347
Thus, these strains appear to be very stable in their new environment. Similar observations
348
were reported for soils in France (5), whereas there was great variability in B. japonicum and
349
B. elkanii survivors 7 years after their introduction as inoculants into Brazilian soils (6).
350
Interestingly none of the tested strains outcompeted the other strain after long term persistence
351
in the Polish soil without host plants and strain abundance in the soil was almost equal. This is
352
in contrast to previous reports for US soils showing dominance of USDA123 over USDA110
353
strains (8, 9, 10). This may in part be due to a lack of other competing strains and strain
354
compatibility to the local environment, including local microbiota.
355
With respect to the issue of mobility – the spatial range of bacterial spread from the
356
site of inoculation, it was rather remarkable that soybean-nodulating bacteria were limited to a
357
rather small area. No soybean bradyrhizobia were found about 50 m from the original
358
experimental plots, as determined by using the MPN method. It should be noted, however,
359
that the original experiment in 1994 was done within the larger area of an experimental field,
360
were cultivation (ploughing and harrowing) was done in different directions along the entire
361
field. Moreover, even the short-distance mobility of rhizobia appeared to be limited to
362
mechanical movement (e.g. by harrowing) of soil particles, so the spatial range only likely
14
363
increased about twice during the course of the experiment. This is in contrast with finding of
364
symbiotic strains with characteristics similar to inoculant strains found in uncropped soils in
365
Brazil (7). While airborne soil has been considered one mechanism for movement of
366
compatible bacteria outside the area of host growth (27), this did not appear to be the case at
367
our field site.
368
In summary, results of this study showed that soybean-specific B. japonicum strains
369
remained viable and symbiotically competent in a foreign environment, originally free of
370
indigenous soybean microsymbionts, for at least 20 years after introducing them into the soil
371
as inocula. This occurred even though soybeans were not grown in the location for the last 17
372
years. We also found that the surviving isolates, which were determined by using molecular
373
and immunochemical methods, were undistinguishable from the original inoculant strains
374
USDA110 and USDA123. Moreover, these competing strains could be detected in fairly high
375
and balanced numbers, suggesting that the competitive dominance of strain USDA 123 in
376
U.S. soils may be due to situations where the inoculant strain was introduced into an
377
environment where well-established populations of indigenous strains exist. Alternately,
378
abiotic factors may have favored one strain over another. Lastly, the mobility of bradyrhizobia
379
was relatively low and nodulation competent soybean microsymbionts could be detected no
380
further than about 40 m from the limits of the original area of inoculation, despite intensive
381
soil cultivation at the study site. Taken together these results show that B. japonicum strains
382
maintain their symbiotically competence for decades in the absence of a legume host.
383
15
384
Acknowledgments
385 386
This study was supported, in part, by previous grant HRN-5600-G-00-2074-00 from
387
the U.S. Agency for International Development (to M.J.S. and C.J.M.) and by the University
388
of Minnesota Agricultural Experiment Station (M.J.S.).
389 390 391
References
392
1. Mądrzak CJ, Golińska B, Króliczak J, Pudełko K, Łażewska D, Lampka B, Sadowsky
393
MJ. 1995. Diversity among field populations of Bradyrhizobium japonicum in Poland.
394
Appl Environ Microbiol 61:1194-1200.
395
2. Stępkowski T, Żak M, Moulin L, Króliczak J, Golińska B, Narożna D, Safronova
396
VI, Mądrzak CJ. 2011. Bradyrhizobium canariense and Bradyrhizobium japonicum are
397
the two dominant rhizobium species in root nodules of lupin and serradella plants growing
398
in Europe. Syst Appl Microbiol 34:368-375.
399
3. Hirsch AM. 2010. How Rhizobia Survive in The Absence of A Legume Host, A Stressful
400
World Indeed. p 377-391. In: Seckbach J, Grube M. (ed), Symbioses and Stress: Joint
401
Ventures in Biology. Springer Science+Business Media B.V.
402
4. Brunel B, Cleyet-Marel JC, Normand P, Bardin R. 1988. Stability of Bradyrhizobium
403
japonicum inoculants after introduction into soil. Appl Environ Microbiol 54:2636-2642.
404
5. Obaton M, Bouniols A, Guillaume P, Guillaume P, Vadez V. 2002. Are Bradyrhizobium
405 406
japonicum stable during a long stay in soil? Plant Soil 245:315-326. 6. Batista JSS, Hungria M, Barcellos FG, Ferreira MC, Mendes IC. 2007. Variability in
407
Bradyrhizobium japonicum and B. elkanii seven years after introduction of both the exotic
408
microsymbiont and the soybean host in a Cerrados soil. Microb Ecol 53:270-284.
16
409 410
7. Ferreira MC, Hungria M. 2002. Recovery of soybean inoculant strains from uncropped soils in Brazil. Field Crops Res 79:139-152.
411
8. Keyser HH, Cregan PB. 1987. Nodulation and competition for nodulation of selected
412
genotypes among Bradyrhizobium japonicum serogroup 123 isolates. Appl Environ
413
Microbiol 53:2631-2635.
414
9. Kosslak RM, Bohlool BB. 1985. Influence of environmental factors on interstrain
415
competition in Rhizobium japonicum. Appl Environ Microbiol 49:1128-1133.
416
10. Lohrke SM, Orf JH, Martínez-Romero E, Sadowsky MJ. 1995. Host-controlled
417
restriction of nodulation by Bradyrhizobium japonicum strains in serogroup 110. Appl
418
Environ Microbiol 61:2378-2383.
419 420 421
11. IUSS Working Group WRB 2006. 2006. World Reference Base for Soil Resources. Word Soil Resources Reports No. 103. FAO, Rome, 1-132pp. 12. Soil Taxonomy. 1999. A Basic System of Soil Classsification for Making and
422
Interpreting a Soil Surveys. Agricultural Handbook 436. U.S. Government Printing Office,
423
Washington, DC.
424
13. Singleton PW, Somasegaran P, Nakao PL, Keyser HH, Hoben HJ, Ferguson PI.
425
1990. Applied BNF Technology a Practical Guide for Extension Specialists. NifTAL
426
Project, University of Hawaii, US AID.
427
14. Cregan PB; Keyser HH; Sadowsky MJ. 1989. Host plant effects on nodulation and
428
competitiveness of the Bradyrhizobium japonicum serotype strains constituting serocluster
429
123. Appl Environ Microbiol 55:2532-2536.
430
15. Somasegaran P, Hoben HJ. 1994. Handbook for Rhizobia. Methods in legume -
431
Rhizobium technology. Springer-Verlag, New York, Berlin, Heidelberg, London, Paris,
432
Tokyo, Hong Kong, Barcelona, Budapest, 1-450pp.
17
433
16. Sadowsky MJ, Tully RE, Cregan PB, Keyser HH. 1987. Genetic diversity in
434
Bradyrhizobium japonicum serogroup 123 and its relation to genotype-specific nodulation
435
of soybean. Appl Environ Microbiol 53:2624-2630.
436
17. Richardson AE, Viccars LA, Watson JM, Gibson AH. 1995. Differentiation of
437
Rhizobium strains using the polymerase chain reaction with random and directed primers.
438
Soil Biol Biochem 27:515-524.
439 440 441 442 443 444 445
18. Schmidt EL, Bankole RO, Bohlool BB. 1968. Fluorescent antibody approach to study of rhizobia in soil. J Bacteriol 95:1987-1992 19. R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 20. De Mendiburu F. 2014. agricolae: Statistical Procedures for Agricultural Research. R package version 1.2-1. CRAN.R-project.org. 21. Dunigan EP, Bollich PK, Hutchinson RL, Hicks PM, Zaunbrecher FC, Scott SG.;
446
Mowers RP. 1984. Introduction and survival of an inoculant strain of Rhizobium
447
japonicum in soil, Agron J 76:463-466
448 449 450
22. Martyniuk S, Oroń J, Martyniuk M. 2005. Diversity and numbers of root nodule bacteria (rhizobia) in Polish soils. Acta Soc Bot Pol 74:83-86. 23. Tong Z, Sadowsky MJ. 1994. A selective medium for isolation and quantification of
451
Bradyrhizobium japonicum and Bradyrhizobium elkanii strains from soils and inoculants.
452
Appl Environ Microbiol 60:581-586.
453 454 455 456
24. Dudeja SS, Khurana AL. 1989. Persistence of Bradyrhizobium sp. (Cajanus) in a sandy loam. Soil Biol Biochem 21:709-713. 25. Botha WJ, Jaftha JB, Bloem JF, Habig JH, Law IJ. 2004. Effect of soil bradyrhizobia on the success of soybean inoculant strain CB 1809. Microbiol Res 159:219-231.
18
457
26. Brockwell J, Bottomley PJ, Thies JE. 1995. Manipulation of rhizobia microflora for
458
improving legume productivity and soil fertility: A critical assessment. Plant Soil 174:143-
459
180.
460 461
27. Hirsch PR. 1996. Population dynamics of indigenous and genetically modified rhizobia in the field. New Phytol 133:159-171.
462 463
19
464 465
Table 1. Identification of bacteria inducing nodules of soybean trap plants grown in field
466
locations shown in Figure 1.
467
1N
Number of nodules analyzed 10
Percent USDA 110 in nodulesa 6 (60)
Percent USDA 123 in nodulesb 4 (40)
2N
9
6 (66.7)
3 (33.3)
3N
5
2 (40)
3 (60)
4N
3
2 (66.7)
1 (33.3)
5N
0
-
-
Total
27
16
11
1S
9
0
9 (100)
2S
9
3 (33.3)
6 (66.7)
3S
9
8 (88.9)
1 (11.1)
4S
9
3 (33.3)
6 (66.7)
5S
9
2(22.2)
7(77.8)
6S
9
5 (55.5)
4 (44.5)
7S
9
3 (33.3)
6 (66.7)
8S
9
8 (88.9)
1 (11.1)
9S
9
7 (77.8)
2 (22.2)
Total
81
39
42
1W
ND
-
-
2W
9
6 (66.7)
3 (33.3)
3W
9
7 (77.8)
2 (22.2)
4W
9
7 (77.8)
2 (22.2)
Location
20
5W
6
2 (33.3)
4 (66.7)
6W
9
5 (55.5)
4 (44.5)
7W
1
1
0
8W
3
2
1
9W
0
-
-
10W – 22W
0
-
-
Total
46
33
16
1E
9
4 (44.5)
5 (55.5)
2E
9
1 (11.1)
8 (88.9)
3E
9
1 (11.1)
8 (88.9)
4E
9
3 (33.3)
6 (66.7)
5E
9
2 (22.2)
7(77.8)
6E
9
8 (88.9)
1 (11.1)
7E
9
8 (88.9)
1 (11.1)
8E
9
9 (100)
0 (0)
9E
9
9 (100)
0 (0)
Total
81
45
36
468 469
a
470
Bj110R).
471
b
472
sequences of recA from five randomly picked isolates were determined showing their identity
473
with USDA 123. ND = no plants germinated in this location
Identification is based on PCR analyses done using USDA 110-specific primers (Bj110F and
Bacteria which did not reveal the presence of USDA 110-specific amplicon. The nucleotide
474
21
475 476
Figure Legends
477
Figure 1. The design and topography of the field experiment of 2011. A number of plant
478
traps were established in order to determine the distance from the original experiment site
479
(grey squares) within which the nodulation of soybeans occurs. Plant traps are aligned E-W
480
and N-S every 4-5 m and marked with black dots which numbers starts from the crossing
481
point (1E, 2E..., 1W, 2W..., 2S, 3N..., and 2S, 3S, etc.). About 10 surface sterilized seeds were
482
sown in each spot. The area shaded light gray indicates the part of field where soybean was
483
sown in 2014. An asterisk (at 8W) marks the most westward located spot where soybean
484
bradyrhizobia were detected. The cross indicates the position of GPS coordinates which are: is
485
52º33’59.3” N and 15º54’47.8” E.
486 487
Figure 2. Nodule occupancy of B. japonicum strains in the experimental field locations (A)
488
and in the entire experimental field (B). Different lower case letters (a, b, c) indicate
489
significant differences among means (represented by crossed circles) according to Fisher's
490
LSD test (LSD=5.92; p
1 2 3 4 5 6
Revised AEM01399-15 Survival and Competitiveness of Bradyrhizobium japonicum Strains Twenty Years after Introduction into Field Locations in Poland
7 8
Dorota Narożnaa, Krzysztof Pudełkoa, Joanna Króliczaka, Barbara Golińskaa,
9
Masayuki Sugawarab, Cezary J. Mądrzaka,† and Michael J. Sadowskyb,*,†
10 11
a
12
11, 60-632 Poznań, Poland
Department of Biochemistry and Biotechnology, Poznań University of Life Sciences, Dojazd
13 14
b
15
Minnesota, St. Paul, Minnesota, USA
Department of Soil, Water, & Climate, and BioTechnology Institute, University of
16 17 18 19 20 21 22 23 24
Running Title: Soil Survival of Bradyrhizobium japonicum after 20 Years.
25 26
†
These authors contributed equally to the work.
27 28
*Address correspondence to Michael J. Sadowsky, BioTechnology Institute, 1479 Gortner
29
Ave, 140 Gortner Labs, St. Paul, MN 55108; Email: [email protected]; Phone (612) 624-
30
2706
31 32 33 34
Supplemental material for this article has been prepared.
35
1
36 37 38
Abstract It was previously demonstrated that there are no indigenous strains of Bradyrhizobium
39
japonicum forming nitrogen-fixing root nodule symbioses with soybean plants in arable field
40
soils in Poland. However, bacteria currently classified within this species are present (together
41
with B. canariense) as indigenous populations of strains specific for nodulation of legumes in
42
the Genisteae tribe. These rhizobia, infecting legumes such as lupins, are well established in
43
Polish soils. The studies described here were based on soybean nodulation field experiments,
44
established at the Poznan University of Life Sciences Experiment Station in Gorzyń, and
45
initiated in the spring of 1994. Long term research was then conducted in order to study the
46
relation between B. japonicum USDA 110 and USDA 123, introduced together into the same
47
location, where no soybean rhizobia were earlier detected, and nodulation and competitive
48
success were followed over time. Here we report the extra-long-term saprophytic survival of
49
B. japonicum strains nodulating soybeans that were introduced as inoculants 20 years earlier
50
and where soybeans were not grown for the last 17 years. The strains remained viable and
51
symbiotically competent, and molecular and immunochemical methods showed that strains
52
were undistinguishable from the original inoculum strains USDA110 and USDA123. We also
53
show that the strains had balanced numbers and their mobility in soil was low. To our
54
knowledge, this is the first report showing the extra long term persistence of soybean-
55
nodulating strains introduced into Polish soils and the first analyzing long term competitive
56
relations of USDA 110 and USDA 123 when both strains were not native and were introduced
57
to the environment almost two decades ago.
2
58 59
Introduction Soybean (Glycine max (L.) Merr.) is still an emerging crop in Poland and only about
60
2,800 hectares is presently being planted. However, there has been slow but systematic
61
increase in production and this is consistent with estimations made earlier by soybean
62
breeders (1). Grown only on a small scale in a few isolated locations, soybeans still remains
63
interesting for farmers and its cultivation is considered as a potential supplemental source of
64
plant proteins for feeding livestock.
65
Despite this, there are several registered soybean cultivars well suited to local
66
environmental conditions, including: Aldana and Augusta (from Poland), Merlin and
67
Lissabon (from Austria), and Annushka and Mavka (from the Ukraine). Recently, qualified
68
seed production of the above cultivars has dramatically increased from 2 ha in 2009 to almost
69
600 ha in 2013, which further indicates farmer’s interest in soybean production in Poland
70
(Jerzy Nawracała, personal communication).
71
Relatively little information is available concerning the presence of symbiotic,
72
nitrogen-fixing bradyrhizobia for soybean in Poland. There are no indigenous Bradyrhizobium
73
japonicum strains forming nitrogen-fixing, root nodule, symbioses with soybeans in arable
74
lands in Poland and thus far, only a few local populations have been identified in a few
75
locations where inocula were applied for growing soybean (1). It should be mentioned,
76
however, that bacteria initially classified as B. japonicum are present (together with B.
77
canariense) as indigenous populations of strains specific for legumes of the tribe Genisteae.
78
These rhizobia, infecting lupins, are well established in Polish soils. They have colonized the
79
area presumably migrating together with their hosts from the Mediterranean after the retreat
80
of glaciers during the last 10,000 years (2). These bacteria, however, cannot nodulate
81
soybeans.
3
82
The lack of naturally occurring soybean microsymbionts in Polish soils provides
83
unique opportunities to study the persistence, stability, competitiveness, and dispersal of
84
bradyrhizobia introduced as inocula. It also opens the possibility to analyze the survival of
85
rhizobia as soil saprophytes in the absence of host plants. This issue is particularly relevant
86
since despite the fact that rhizobia do not form resting stage spores (3), they still are
87
characterized as persistent autochthonous microbes in soils where they have been introduced.
88
Moreover, there are relatively few studies on the long term survival of rhizobia in the absence
89
of a legume host and in foreign environments, and those without naturally occurring host
90
plants. For soybeans, a few of these studies were done in France (4, 5) and Brazil (6, 7), but in
91
some there was repeated inoculation due to die off.
92
The competitiveness of strains infecting soybeans may be strongly dependent on
93
environmental conditions, and presuming that they are indigenous, are likely to be more fit to
94
interact with host plant genotypes than more recent bacterial and host additions to soils. Based
95
on these assumptions, we designed experiments to study the competitive relationship between
96
B. japonicum strains USDA 110 and USDA 123, introduced together into a field location
97
where no soybean rhizobia were previously detected. While it has been shown that USDA 123
98
outcompeted USDA110 in North American soils comprised of several Bradyhizobium
99
serotype strains (8, 9, 10), no information is available on how these bacteria compete for
100
nodulation in natural field soils devoid of indigenous soyean-nodulating bradyrhizobia.
101
Here we report results of field experiments initiated in 1994 where soybean was sown
102
into the soil after inoculation with B. japonicum USDA 123 and USDA 110, either separately
103
or as a mixed inoculant. Soybean was then grown in this location for three subsequent years
104
and the field was used for growing other crops for the next 17 years. After this time,
105
nodulation results showed the extra-long-term saprophytic survival of B. japonicum strains
4
106
nodulating soybeans and long term competitiveness of USDA 110 and USDA 123 in natural
107
soils initially devoid of soybean-nodulating bradyrhizobia.
108 109
Materials and Methods
110 111
Initial field experiments 1994-1997. A soybean nodulation field experiment was established
112
at the Poznan University of Life Sciences Experiment Station in Gorzyń, about 80 km North-
113
West from Poznań, Poland, in the spring of 1994 on soils characterized as Arenic Hapludalfs
114
or Albic Luvisols (Arenic) (11, 12). The selected field site was free of indigenous strains
115
capable of nodulating soybeans (1, and unpublished data). Prior to the first sowing of surface
116
sterilized seeds (G. max (L.) Merr. cv Naviko), separate plots were inoculated with USDA
117
123, USDA 110, or a mixture containing equal cell concentrations of both strains. Inoculation
118
was performed using a peat-based inoculum as previously described (13). The seeds were
119
inoculated using peat inoculum (100 g inoculum per 1 kg of seeds) and gum arabic as a
120
sticker. The mixed inoculation was performed using equal amounts of USDA 110 and USDA
121
123 at a final concentration of 2 x 106 cells per seed. In the case of mixed inoculation half of
122
this number was USDA 110 and half was USDA 123. An uninoculated control plot was also
123
included into the experiment and each experimental plot was replicated four times. No
124
inoculations were performed during subsequent experimental seasons. The experimental field
125
is a part of a larger experimental area, where field experiments were conducted for at least 50
126
years (Supplemental Fig. S1). The crops cultivated in this particular part of experimental
127
station were; wheat, barley, triticale, potatoes, and yellow and narrow-leafed lupins.
128
Nodulation of soybean plants was analyzed 6 wk after germination and both the
129
number and fresh weight of nodules were measured. Strain occupancy of nodules was
130
determined by using an immuno-spot-blot method and cross-adsorbed, strain-specific,
131
antibodies as previously described (1, 14). The experiment was repeated over four subsequent
132
seasons, until 1997, and after this time soybeans were not grown at this location.
133 134
Detection of soybean-nodulating bradyrhizobia 17 - 20 years after their introduction
135
into Gorzyń soil. Twelve soil samples were obtained from the original field experiment site 5
136
in September 2010. Surface sterilized soybean (G. max (L.) Merr. cv. Augusta) seeds were
137
sown in duplicate pots filled with these soil samples. Four seed were sown in each pot. Plants
138
were grown in the greenhouse for 5 wk.
139
Independently, a trap host approach (15) was used directly in the field to determine the
140
presence of soybean-nodulating bacteria and to determine the distance of their dispersal from
141
their release site. For this purpose, in the spring of 2011, surface sterilized soybean seeds (G.
142
max (L.) Merr. cv. Augusta) were sown in the spots forming two perpendicular lines of trap
143
hosts across the former field experiment site used in 1994. Plants were harvested and root
144
nodules were collected 6 wk after sowing. The experimental design is shown in Fig.1. In
145
2014, soybean was grown on the entire area of the initial experimental site as indicated above.
146 147
Serological identification of bacteria from soybean root nodules. Nodules were collected
148
from plants grown in the greenhouse after exposure to soil taken from the field site in
149
September 2010. Four plants were taken from each of 12 soil samples. All nodules from each
150
plant were separately collected and dried at 60oC for 48 hr. The 48 samples collectively
151
produced 1,266 nodules, all of which were analyzed for the presence of strains USDA 123
152
and USDA 110 by using the immuno-spot-blot method and cross-adsorbed, strain-specific,
153
antibodies as previously described (1, 14).
154 155
Isolation of bacteria from soybean root nodules. Bacteria were obtained from root nodules
156
surface sterilized with 95% ethanol and 3% sodium hypochlorite, according to Somasegaran
157
and Hoben (15). Purified bacterial isolates were obtained by streaking nodule suspensions
158
onto AG agar medium (16) containing 50 µl/ml cycloheximide, followed by two to three
159
successive isolations for single pure colonies on the same medium.
160
6
161
DNA isolation and PCR template preparation. Total genomic DNA was isolated from
162
single bacterial colonies by using the GenElute Bacterial Genomic DNA Kit (Sigma-Aldrich
163
Corporation, St. Louis, MO, USA) following the manufacturer's instructions. Templates for
164
PCR amplification were lysates of crushed soybean nodules stored in 25% glycerol at -80oC.
165
A 50 µl aliquot of PrepMan Ultra buffer (Applied Biosystems, Grand Island, NY, USA) was
166
added to 50 µl of crushed suspension. The resulting mixture was heated for 5 min at 95°C,
167
centrifuged for 5 min at 14,000 rpm, and the supernatant was used as a template for PCR.
168 169
Molecular identification of bacteria from soybean root nodules. Genomic DNA
170
preparations were obtained from bacteria isolated from nodules from the September, 2010
171
experiment. Three nodules were taken from each soil sample giving 36 DNA samples. Three
172
independent PCR analyses were performed for DNA from each isolate for the: (i) identi-
173
fication of USDA 110- specific sequence(using strain-specific primers), (ii) analysis of recA
174
gene sequences in order to distinguish between B. japonicum strains USDA 123 and USDA
175
110, and (iii) analysis of strain diversity using RP01 primers.
176
A B. japonicum USDA 110 strain-specific sequence was selected based on an in silico
177
search of the USDA 110 genome. Sequence BAC45285.1, within genomic position 17139-
178
18371, is apparently specific for B. japonicum strains USDA 110 and USDA 6. The PCR
179
primers were designed based on this sequence: Bj110F 5’-TAGGTGATGGGAT
180
AACGCTCT-3’ and Bj110R 5’-GTCCAGCTTCGCG ATCGACCGG-3’ (Suppl. Figure
181
S2). The primers amplify a 199 bp DNA fragment, between genomic positions 17151 and
182
17331, encoding a protein of unknown function.
183
In order to reconfirm the identification of USDA 110 vs USDA 123 genotypes, a
184
common phylogenetic marker - the sequence of recA was used. The nucleotide sequence of
185
the recA gene from USDA 110 and USDA 123 (collection strains) was determined.
7
186
Alignment of their 559 bp-long fragments allowed design of specific forward primers: recA
187
110 5’-GGGCTCGACATTGCGCTC-3’ and recA 123
188
3’ which are complementary to the fragment 135-151 of the recA sequence (Suppl. Fig. S3).
189
A common reverse primer, recA 549R 5’-TTGCGCAGC GCCTGGCTCAT-3’, was
190
designed for the conserved regions of many bradyrhizobia (nucleotide position 530 to 549).
5’-GGGCTCG ACATTGCACTG-
The verification of specificity of the recA primers was done for a collection of USDA
191 192
110 and USDA 123 strains and for randomly selected strains isolated from Gorzyń soil. The
193
diversity of isolates identified as related to USDA 110 or USDA 123 strains were analyzed
194
using the RP01 PCR (5’-AATTTTCAAGCGTCGTGCCA-3’) primer according to
195
Richardson et al. (17). The PCR was carried out using Allegro Tag DNA Polymerase
196
(Novazym, Poznań, Poland) according to the following conditions: initial denaturation 95°C
197
for 2 min , 30 cycles of 30 s at 95°C, 30 s at 56°C for Bj110F, Bj110R, recA 110, recA 123,
198
and recA 549R, or 52°C for RP01,1 min at 72°C, and a final elongation step of 5 min at
199
72°C.The PCR products were analyzed on 1% agarose gels.
200 201
Enumerating bacteria by using MPN analyses. Determination of the number of bacteria
202
capable of nodulating soybeans was performed by using the most probable number (MPN)
203
technique as described by Somasegaran and Hoben (15). Soil samples were collected in
204
Gorzyń on October 25, 2012 and stored at 4°C. An equivalent of 100 g dry mass of soil
205
(109.4 g fresh weigh) was suspended in 900 ml of water and shaken intensively for 5 min.
206
The resulting suspension was diluted and 1 ml of each dilutions was added to surface
207
sterilized soybean seeds sown in Leonard jar assemblies (15). Five replicates were used for
208
each dilution, and a negative control consisting of uninoculated surface sterilized seeds, was
209
used. Nodulation was recorded after 34 d and MPN values determined as previously described
210
(15).
8
211
Enumerating bacteria in soils using strain-specific fluorescent antibodies. The
212
identification and counting of the B. japonicum USDA 110 and USDA 123 cells were done
213
using strain-specific fluorescent antibodies (FA) as described by Somasegaran and Hoben
214
(15) and Schmidt et al (18). Two soil samples were analyzed and two filters were prepared by
215
filtering a 10 ml of solution from each sample (see Suppl. File 1). The number of bacteria
216
visible in 30 microscopic fields on each filter was determined, three times, using a Zeiss
217
Axiovert 200 Inverted microscope with a 100 x immersion lens and Filter set 09 (exc. BP
218
450-490, beam splitter FT 510, emission LP 515). No cross-reacting soil bacteria were
219
observed.
220 221
Statistical analyzes. Results of all studies were evaluated for significance by using the R
222
statistical software version 3.1.2 (19). ANOVA tests were followed by post hoc analysis with
223
Fisher's LSD test from agricolae package (20), at a significance level of α=0.05.
224 225 226
Results
227 228
Detection of soybean-nodulating bacteria 20 years after their introduction into Gorzyń
229
soil.
230
in1994 contained soybean rhizobia. Root nodules were found on all soybean plants grown in
231
the greenhouse in 24 pots containing 12 soil samples. Results of immuno-spot-blot analyses
232
revealed the presence of strains USDA 110 and USDA 123, in balanced numbers. Data shown
233
in Fig. 2A indicates that USDA 123 prevailed in soils taken from five locations (1, 3, 4, 6 and
234
12), with considerable domination in location 12 (72.4%). The soil samples from locations 5,
235
8, and 10 contained more USDA 110 than USDA 123 cells. Location 5 had a significant
236
prevalence of USDA 110 strains (85.6%). The remaining samples (from locations 2, 7, 9, and
All soil samples taken from the exact same location of the field experiment conducted
9
237
11) contained about equal numbers of both strains. Taken together, 587 nodules (46.4%) were
238
occupied by strain USDA110, 653 (51.6%) by USDA123, and 26 nodules (2%) contained
239
both strains (Fig. 2B). Only three locations did not contain any doubly occupied nodules
240
(Suppl. Table S1). Another 153 nodules were saved as crushed suspensions. Thirty-six
241
suspensions, representing nodules from three plants from each of the 12 soil samples, were
242
randomly chosen and used to obtain pure isolates. These isolates (UPP401 to UPP436) were
243
stored in 50% glycerol at -80oC.
244
A plant trap approach was used to determine the distance from the exact site of initial
245
inoculation where bradyrhizobia still could be found. As shown in Fig.1 and Table 1,
246
successful nodulation was observed in all tested soils samples, but not those located further
247
than about 40 m westward from the edge of former experiment site. This result suggests that
248
there was rather poor dispersal of inoculum strains. In 2014 soybean cultivation was
249
considered to serve as the ultimate confirmation of the presence of soybean-nodulating
250
bradyrhizobia. Nodulation was detected within the entire experimental area. No further
251
analyses were performed.
252
The identification of strains occupying nodules was performed using the PCR assay
253
with USDA 110-specific primers. Data in Table I shows that both strains were found
254
dispersed within the entire area, revealing a fairly balanced proportion of both bacterial
255
genotypes.
256 257
Molecular identification of bacteria from soybean root nodules. The identity and diversity
258
of each of the 36 purified isolates were determined by using PCR and using USDA 110 strain-
259
specific and RP01 primers. Results verifying the specificity of USDA 110-specific primers
260
are shown in Fig. 3. The amplicon of the expected size of 199 bp appeared when lysed cells of
261
B. japonicum USDA 110 were used as the template for PCR amplification (Fig. 3 lane 1). No
10
262
specific products were generated with DNAs of other bacteria tested, including B. japonicum
263
USDA 123 (Fig. 3, Lane 2), and lupin nodulating strains of B. japonicum UPP323 and
264
UPP344 (2), Bradyrhizobium sp. UPP331 and UPP370, Bradyrhizobium sp. USDA 3045
265
(Lupinus) as well as B. elkanii resembling strain UPP372 and B. yuangmingense, resembling
266
strain UPP253. Based on this data, primers Bj110F and Bj110R were considered to be
267
specific for B. japonicum strain USDA 110 and they were used to characterize purified
268
isolates obtained from root nodules of soybean plants grown in soil from Gorzyń. PCR results
269
in Fig. 4 show that 15 of 36 isolates (42%) could be presumed to be the descendants of the
270
original inoculum strain B. japonicum USDA 110.
271
PCR amplifications done with recA-specific primers were used to reconfirm the
272
identification of B. japonicum isolates. Data in Fig. 5 show the specificity and usefulness of
273
these primers. The DNA of the USDA123 collection strain, and isolates identified as the
274
presumed descendants of this strain (UPP405 and UPP406 - compare Fig. 4 lanes 5 and 6)
275
did not generate PCR amplification products when the recA110 primer was used (lanes 4, 5,
276
and 6). Conversely, the recA123 primer did not produce PCR amplification products with
277
template DNA from USDA110 and their presumed descendants (UPP401 and UPP403).
278
The diversity of isolates identified as related to USDA 110 or USDA 123 strains was
279
analyzed by using the RP01 PCR and appeared to be low, since only two types of profiles
280
were found in Fig. 6 consistently representing the two strains identified. Among 36 strains
281
analyzed, UPP 403, UPP 404, UPP 407, UPP 408, UPP 401, UPP 409, UPP 410, UPP 412,
282
UPP 413, UPP 416, UPP 419, UPP 422, UPP 428, UPP 431, UPP 434, (shown in the lanes 1,
283
2, 5, 6, 9, 12, 13, 17, 18, 19, 20, 22, 28, 31, and 34, respectively) represent the profile typical
284
for strain USDA 110. The remaining profiles were characteristic of B. japonicum strain
285
USDA 123. The fragments in lane 16 are likely due to a nodule (used as template source) that
286
was occupied by both strains.
11
287
Enumerating bradyrhizoba in soils. The MPN technique was used to quantify the number
288
of bradyrhizobia in Gorzyń soil sample capable of nodulating soybeans. Results indicated that
289
this soil contained about 1 x103 bradyrhizobia per gram soil (dry weight); The number of
290
bradyrhizobia in duplicate Gorzyń soil samples was also evaluated by using direct
291
immunofluorescence (18). Results of fluorescent antibody (FA) analyses indicated that, on
292
average in 2012, the soil contained 1.2 x 104 and 6.9 x 103 cells per gram of USDA 123 and
293
USDA 110, respectively, after 15 years of no host plants (Suppl. Table S2).
294 295 296 297
Discussion These experiments were designed to study the interrelationship of well-known
298
competing B. japonicum strains (USDA123 and USDA 110), introduced to a foreign
299
environment, in 1994-1997. The mixed inoculation experiment revealed that initially USDA
300
110 prevailed over USDA 123. About 70% of nodules were occupied by this strain as was
301
established by using an immuno-spot-blot method and cross-adsorbed strain-specific
302
antibodies. However, good performance by USDA 110 was also observed during the
303
following years. Due to natural reasons, the areas occupied by both strains have expanded
304
leading to an overlap of their initial sites of inoculation. The domination of USDA 110 was
305
not as pronounced during the third and fourth year after inoculation, but both strains showed a
306
good survival rate (unpublished data). These observations allowed us to determine: (i) the
307
persistence (survival) of symbiotic bacteria (bradyrhizobia) without a host legume; (ii) the
308
genetic stability of inoculant bacteria in the soil environment, and (iii) the mobility of
309
particular genotypes – the spatial range of bacteria spread from the site of their introduction
310
into soil.
311
Addressing the first issue, our findings clearly indicate that both strains introduced as
312
the inocula remained viable and active as soybean microsymbionts for over 20 years after the 12
313
only inoculation in May 1994, and subsequent growth of soybean for three additional years
314
(1994-1997). This results is in contrast to that of Dunigan et al. (21) who showed that three
315
years of repeated inoculation was required to establish B. japonicum in a Louisiana soil. Also
316
the number of bacterial cells capable of nodulating soybean long time after inoculation was
317
surprisingly high for the introduced strains.
318
Their abundance estimated by the MPN method (~ 103 g-1) was very similar to the
319
number of Bradyrhizobium sp. strains nodulating lupins and native to Polish soils. Martyniuk
320
(22) classified soils containing over 1000 cells in 1g of soil as highly populated by these
321
rhizobia. This observation should be interpreted considering particular groups of diazotrophic,
322
bacterial, symbionts. One should remember that their presence in the soil is usually detected
323
by the plant-trap method. The difference in our results and those of previous studies may be
324
due to the fact that others have introduced rhizobia into soils already containing an indigenous
325
population of these bacteria.
326
Our estimation of cell numbers based on the FA method suggested 20-times more
327
rhizobial cells than that determined by MPN. This may be due to the presence of viable but
328
not culturable (infectable) bradyrhizobia. On the other hand, the immunofluorescence
329
technique tends to overestimate soil bacteria population sizes (23). This is perhaps the reason
330
for the relatively common opinion that the presence of host plants is necessary for the survival
331
of their respective microsymbionts. The chromosomal localization of symbiotic genes in the
332
genus Bradyrhizobium may explain the reason for the fact that the relative good persistence in
333
soil is reported for the symbiotic traits of bacteria belonging to this group. However, the
334
survival period of B. japonicum as saprophytes was reported to be 8-13 years (4, 5), and
335
environmental conditions also remains an important factor governing the presence of
336
indigenous strains. Very poor survival of inoculant strains of Bradyrhizobium sp. (Cajanus)
337
was reported (24). This was most probably due to high soil temperature and fluctuating soil
13
338
moisture. None of these factors were present in the Polish soil used in this study and there
339
were no indigenous strains of soybean-nodulating B. japonicum. Thus, as has been pointed
340
out by others, and is standard practice in some countries (e.g. New Zealand), decisions on the
341
application and type of inoculum strain should be made carefully. As has been discussed
342
earlier (25, 26), both edaphic conditions and existing bacterial populations strongly affect
343
inoculation success.
344
Considering the second issue, isolates which were currently obtained from the root
345
nodules of soybean plants grown in the soil from the experimental site in this study were very
346
likely descendants of the original inoculum strains (B. japonicum USDA110 and USDA123).
347
Thus, these strains appear to be very stable in their new environment. Similar observations
348
were reported for soils in France (5), whereas there was great variability in B. japonicum and
349
B. elkanii survivors 7 years after their introduction as inoculants into Brazilian soils (6).
350
Interestingly none of the tested strains outcompeted the other strain after long term persistence
351
in the Polish soil without host plants and strain abundance in the soil was almost equal. This is
352
in contrast to previous reports for US soils showing dominance of USDA123 over USDA110
353
strains (8, 9, 10). This may in part be due to a lack of other competing strains and strain
354
compatibility to the local environment, including local microbiota.
355
With respect to the issue of mobility – the spatial range of bacterial spread from the
356
site of inoculation, it was rather remarkable that soybean-nodulating bacteria were limited to a
357
rather small area. No soybean bradyrhizobia were found about 50 m from the original
358
experimental plots, as determined by using the MPN method. It should be noted, however,
359
that the original experiment in 1994 was done within the larger area of an experimental field,
360
were cultivation (ploughing and harrowing) was done in different directions along the entire
361
field. Moreover, even the short-distance mobility of rhizobia appeared to be limited to
362
mechanical movement (e.g. by harrowing) of soil particles, so the spatial range only likely
14
363
increased about twice during the course of the experiment. This is in contrast with finding of
364
symbiotic strains with characteristics similar to inoculant strains found in uncropped soils in
365
Brazil (7). While airborne soil has been considered one mechanism for movement of
366
compatible bacteria outside the area of host growth (27), this did not appear to be the case at
367
our field site.
368
In summary, results of this study showed that soybean-specific B. japonicum strains
369
remained viable and symbiotically competent in a foreign environment, originally free of
370
indigenous soybean microsymbionts, for at least 20 years after introducing them into the soil
371
as inocula. This occurred even though soybeans were not grown in the location for the last 17
372
years. We also found that the surviving isolates, which were determined by using molecular
373
and immunochemical methods, were undistinguishable from the original inoculant strains
374
USDA110 and USDA123. Moreover, these competing strains could be detected in fairly high
375
and balanced numbers, suggesting that the competitive dominance of strain USDA 123 in
376
U.S. soils may be due to situations where the inoculant strain was introduced into an
377
environment where well-established populations of indigenous strains exist. Alternately,
378
abiotic factors may have favored one strain over another. Lastly, the mobility of bradyrhizobia
379
was relatively low and nodulation competent soybean microsymbionts could be detected no
380
further than about 40 m from the limits of the original area of inoculation, despite intensive
381
soil cultivation at the study site. Taken together these results show that B. japonicum strains
382
maintain their symbiotically competence for decades in the absence of a legume host.
383
15
384
Acknowledgments
385 386
This study was supported, in part, by previous grant HRN-5600-G-00-2074-00 from
387
the U.S. Agency for International Development (to M.J.S. and C.J.M.) and by the University
388
of Minnesota Agricultural Experiment Station (M.J.S.).
389 390 391
References
392
1. Mądrzak CJ, Golińska B, Króliczak J, Pudełko K, Łażewska D, Lampka B, Sadowsky
393
MJ. 1995. Diversity among field populations of Bradyrhizobium japonicum in Poland.
394
Appl Environ Microbiol 61:1194-1200.
395
2. Stępkowski T, Żak M, Moulin L, Króliczak J, Golińska B, Narożna D, Safronova
396
VI, Mądrzak CJ. 2011. Bradyrhizobium canariense and Bradyrhizobium japonicum are
397
the two dominant rhizobium species in root nodules of lupin and serradella plants growing
398
in Europe. Syst Appl Microbiol 34:368-375.
399
3. Hirsch AM. 2010. How Rhizobia Survive in The Absence of A Legume Host, A Stressful
400
World Indeed. p 377-391. In: Seckbach J, Grube M. (ed), Symbioses and Stress: Joint
401
Ventures in Biology. Springer Science+Business Media B.V.
402
4. Brunel B, Cleyet-Marel JC, Normand P, Bardin R. 1988. Stability of Bradyrhizobium
403
japonicum inoculants after introduction into soil. Appl Environ Microbiol 54:2636-2642.
404
5. Obaton M, Bouniols A, Guillaume P, Guillaume P, Vadez V. 2002. Are Bradyrhizobium
405 406
japonicum stable during a long stay in soil? Plant Soil 245:315-326. 6. Batista JSS, Hungria M, Barcellos FG, Ferreira MC, Mendes IC. 2007. Variability in
407
Bradyrhizobium japonicum and B. elkanii seven years after introduction of both the exotic
408
microsymbiont and the soybean host in a Cerrados soil. Microb Ecol 53:270-284.
16
409 410
7. Ferreira MC, Hungria M. 2002. Recovery of soybean inoculant strains from uncropped soils in Brazil. Field Crops Res 79:139-152.
411
8. Keyser HH, Cregan PB. 1987. Nodulation and competition for nodulation of selected
412
genotypes among Bradyrhizobium japonicum serogroup 123 isolates. Appl Environ
413
Microbiol 53:2631-2635.
414
9. Kosslak RM, Bohlool BB. 1985. Influence of environmental factors on interstrain
415
competition in Rhizobium japonicum. Appl Environ Microbiol 49:1128-1133.
416
10. Lohrke SM, Orf JH, Martínez-Romero E, Sadowsky MJ. 1995. Host-controlled
417
restriction of nodulation by Bradyrhizobium japonicum strains in serogroup 110. Appl
418
Environ Microbiol 61:2378-2383.
419 420 421
11. IUSS Working Group WRB 2006. 2006. World Reference Base for Soil Resources. Word Soil Resources Reports No. 103. FAO, Rome, 1-132pp. 12. Soil Taxonomy. 1999. A Basic System of Soil Classsification for Making and
422
Interpreting a Soil Surveys. Agricultural Handbook 436. U.S. Government Printing Office,
423
Washington, DC.
424
13. Singleton PW, Somasegaran P, Nakao PL, Keyser HH, Hoben HJ, Ferguson PI.
425
1990. Applied BNF Technology a Practical Guide for Extension Specialists. NifTAL
426
Project, University of Hawaii, US AID.
427
14. Cregan PB; Keyser HH; Sadowsky MJ. 1989. Host plant effects on nodulation and
428
competitiveness of the Bradyrhizobium japonicum serotype strains constituting serocluster
429
123. Appl Environ Microbiol 55:2532-2536.
430
15. Somasegaran P, Hoben HJ. 1994. Handbook for Rhizobia. Methods in legume -
431
Rhizobium technology. Springer-Verlag, New York, Berlin, Heidelberg, London, Paris,
432
Tokyo, Hong Kong, Barcelona, Budapest, 1-450pp.
17
433
16. Sadowsky MJ, Tully RE, Cregan PB, Keyser HH. 1987. Genetic diversity in
434
Bradyrhizobium japonicum serogroup 123 and its relation to genotype-specific nodulation
435
of soybean. Appl Environ Microbiol 53:2624-2630.
436
17. Richardson AE, Viccars LA, Watson JM, Gibson AH. 1995. Differentiation of
437
Rhizobium strains using the polymerase chain reaction with random and directed primers.
438
Soil Biol Biochem 27:515-524.
439 440 441 442 443 444 445
18. Schmidt EL, Bankole RO, Bohlool BB. 1968. Fluorescent antibody approach to study of rhizobia in soil. J Bacteriol 95:1987-1992 19. R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 20. De Mendiburu F. 2014. agricolae: Statistical Procedures for Agricultural Research. R package version 1.2-1. CRAN.R-project.org. 21. Dunigan EP, Bollich PK, Hutchinson RL, Hicks PM, Zaunbrecher FC, Scott SG.;
446
Mowers RP. 1984. Introduction and survival of an inoculant strain of Rhizobium
447
japonicum in soil, Agron J 76:463-466
448 449 450
22. Martyniuk S, Oroń J, Martyniuk M. 2005. Diversity and numbers of root nodule bacteria (rhizobia) in Polish soils. Acta Soc Bot Pol 74:83-86. 23. Tong Z, Sadowsky MJ. 1994. A selective medium for isolation and quantification of
451
Bradyrhizobium japonicum and Bradyrhizobium elkanii strains from soils and inoculants.
452
Appl Environ Microbiol 60:581-586.
453 454 455 456
24. Dudeja SS, Khurana AL. 1989. Persistence of Bradyrhizobium sp. (Cajanus) in a sandy loam. Soil Biol Biochem 21:709-713. 25. Botha WJ, Jaftha JB, Bloem JF, Habig JH, Law IJ. 2004. Effect of soil bradyrhizobia on the success of soybean inoculant strain CB 1809. Microbiol Res 159:219-231.
18
457
26. Brockwell J, Bottomley PJ, Thies JE. 1995. Manipulation of rhizobia microflora for
458
improving legume productivity and soil fertility: A critical assessment. Plant Soil 174:143-
459
180.
460 461
27. Hirsch PR. 1996. Population dynamics of indigenous and genetically modified rhizobia in the field. New Phytol 133:159-171.
462 463
19
464 465
Table 1. Identification of bacteria inducing nodules of soybean trap plants grown in field
466
locations shown in Figure 1.
467
1N
Number of nodules analyzed 10
Percent USDA 110 in nodulesa 6 (60)
Percent USDA 123 in nodulesb 4 (40)
2N
9
6 (66.7)
3 (33.3)
3N
5
2 (40)
3 (60)
4N
3
2 (66.7)
1 (33.3)
5N
0
-
-
Total
27
16
11
1S
9
0
9 (100)
2S
9
3 (33.3)
6 (66.7)
3S
9
8 (88.9)
1 (11.1)
4S
9
3 (33.3)
6 (66.7)
5S
9
2(22.2)
7(77.8)
6S
9
5 (55.5)
4 (44.5)
7S
9
3 (33.3)
6 (66.7)
8S
9
8 (88.9)
1 (11.1)
9S
9
7 (77.8)
2 (22.2)
Total
81
39
42
1W
ND
-
-
2W
9
6 (66.7)
3 (33.3)
3W
9
7 (77.8)
2 (22.2)
4W
9
7 (77.8)
2 (22.2)
Location
20
5W
6
2 (33.3)
4 (66.7)
6W
9
5 (55.5)
4 (44.5)
7W
1
1
0
8W
3
2
1
9W
0
-
-
10W – 22W
0
-
-
Total
46
33
16
1E
9
4 (44.5)
5 (55.5)
2E
9
1 (11.1)
8 (88.9)
3E
9
1 (11.1)
8 (88.9)
4E
9
3 (33.3)
6 (66.7)
5E
9
2 (22.2)
7(77.8)
6E
9
8 (88.9)
1 (11.1)
7E
9
8 (88.9)
1 (11.1)
8E
9
9 (100)
0 (0)
9E
9
9 (100)
0 (0)
Total
81
45
36
468 469
a
470
Bj110R).
471
b
472
sequences of recA from five randomly picked isolates were determined showing their identity
473
with USDA 123. ND = no plants germinated in this location
Identification is based on PCR analyses done using USDA 110-specific primers (Bj110F and
Bacteria which did not reveal the presence of USDA 110-specific amplicon. The nucleotide
474
21
475 476
Figure Legends
477
Figure 1. The design and topography of the field experiment of 2011. A number of plant
478
traps were established in order to determine the distance from the original experiment site
479
(grey squares) within which the nodulation of soybeans occurs. Plant traps are aligned E-W
480
and N-S every 4-5 m and marked with black dots which numbers starts from the crossing
481
point (1E, 2E..., 1W, 2W..., 2S, 3N..., and 2S, 3S, etc.). About 10 surface sterilized seeds were
482
sown in each spot. The area shaded light gray indicates the part of field where soybean was
483
sown in 2014. An asterisk (at 8W) marks the most westward located spot where soybean
484
bradyrhizobia were detected. The cross indicates the position of GPS coordinates which are: is
485
52º33’59.3” N and 15º54’47.8” E.
486 487
Figure 2. Nodule occupancy of B. japonicum strains in the experimental field locations (A)
488
and in the entire experimental field (B). Different lower case letters (a, b, c) indicate
489
significant differences among means (represented by crossed circles) according to Fisher's
490
LSD test (LSD=5.92; p
