Genuswide Acid Tolerance Accounts for the Biogeographical Distribution of Soil Burkholderia Populations

bs_bs_banner

Environmental Microbiology (2014) 16(6), 1503–1512 doi:10.1111/1462-2920.12211

Genus-wide acid tolerance accounts for the biogeographical distribution of soil Burkholderia populations

Nejc Stopnisek,1,5 Natacha Bodenhausen,2 of the genus Burkholderia. Our results provide a Beat Frey,3 Noah Fierer,4 Leo Eberl1 and basis for building a predictive understanding of the Laure Weisskopf1,5* biogeographical patterns exhibited by Burkholderia 1Institute of Plant Biology, University of Zurich, Zürich sp. CH-8008, Switzerland. 2Institute of Microbiology, Swiss Federal Institute of Introduction Technology, Zürich CH-8093, Switzerland. 3Snow and Landscape Research, Swiss Federal The genus Burkholderia, which belongs to the β Institute for Forest, Birmensdorf CH-8903, Switzerland. -Proteobacteria class, currently comprises more than 60 4Department of Ecology and Evolutionary Biology, species that are widely distributed and frequently isolated University of Colorado, Boulder, CO 80309-0216, USA. from a large range of natural and clinical environments 5Swiss Federal Research Station for Agronomy and (Compant et al., 2008). The genus Burkholderia can be Nature, Agroscope Reckenholz-Tänikon, Zürich divided phylogenetically into two main clusters: the first CH-8046, Switzerland. one consists mainly of human, animal and plant pathogens, e.g. the Burkholderia cepacia complex (Bcc) and the rice pathogen B. glumae. However, it is important to Summary note that some strains belonging to Bcc, such as Bacteria belonging to the genus Burkholderia are B. ambifaria or B. lata, also show plant growth-promoting highly versatile with respect to their ecological niches abilities as well as biocontrol activities against and lifestyles, ranging from nodulating tropical plants phytopathogenic fungi. The other cluster consists mainly to causing melioidosis and fatal infections in cystic of plant-beneficial-environmental (PBE) Burkholderia fibrosis patients. Despite the clinical importance and species (Suárez-Moreno et al., 2012). The members of agronomical relevance of Burkholderia species, infor- the first cluster have been extensively studied because of mation about the factors influencing their occurrence, their medical importance, but recently the PBE cluster has abundance and diversity in the environment is been the focus of research efforts with the discovery that scarce. Recent findings have demonstrated that pH is various species of this cluster are able to fix nitrogen the main predictor of soil bacterial diversity and com- (Estrada-De Los Santos et al., 2001; Martínez-Aguilar munity structure, with the highest diversity observed et al., 2008) and to nodulate legumes (Moulin et al., in neutral pH soils. As many Burkholderia species 2001). Burkholderia from the PBE cluster have been have been isolated from low pH environments, we mainly isolated from plant rhizosphere, but they are also hypothesized that acid tolerance may be a general frequently detected in sediment and bulk soil (Salles feature of this genus, and pH a good predictor of their et al., 2002; Lazzaro et al., 2008; Lim et al., 2008; Lepleux occurrence in soils. Using a combination of environ- et al., 2012; Štursová et al., 2012). In addition to their mental surveys at trans-continental and local scales, nitrogen-fixing and nodulating abilities, their versatile as well as in vitro assays, we show that, unlike most metabolism also enables them to survive in harsh condi- bacteria, Burkholderia species have a competitive tions, such as nutrient-limited or polluted environments, advantage in acidic soils, but are outcompeted in and to degrade recalcitrant compounds (Pérez-Pantoja alkaline soils. Physiological assays and diversity et al., 2012). As such, they have been suggested as good analysis based on 16S rRNA clone libraries demon- candidates for use in biotechnology, e.g. for bio- or strate that pH tolerance is a general phenotypic trait phytoremediation, biocontrol and biofertilization. Despite the high relevance of the Burkholderia genus for human health, agronomy and biotechnology, surprisingly little is Received 17 May, 2013; revised 4 July, 2013; accepted 10 July, 2013. *For correspondence. E-mail [email protected]; known about the factors underlying their geographical Tel. (+41) 44 377 72 11; Fax (+41) 44 377 72 01. distribution.

A was high (8.7), the isolated strains were all able to grow at low pH (4.5). These reports provide anecdotal evidence that Burkholderia might be tolerant to low pH conditions, which enables members of this genus to thrive in niches where others would be inhibited. We, therefore, hypothesized (i) that low pH tolerance is an intrinsic phenotypic trait of the Burkholderia genus, and (ii) that the relative abundance and diversity of Burkholderia populations are highest in low pH soils, with the biogeography of Burkholderia predictable from soil pH. To test these pH scale Relative abundance (%) 3 − 4 6.70 hypotheses, we developed a novel quantitative polymer- 4 − 5 5 − 6 2.82 6 − 7 ase chain reaction (qPCR) protocol to analyze the relative 1.08 7 − 8 0.01 8 − 9 0.00 abundance of Burkholderia populations in soils at a trans- B continental and a local scale. Intrageneric diversity and community structure were determined by 16S rRNA- based clone libraries constructed from a selected subset 0 20406080 200 123456 − Relative abundance (%) abundance Relative 0 of the trans-continental scale soil samples. In addition, in 45678 pH vitro physiological assays were used to test the direct −160 −140 −120 −100 −80 −60 −40 effects of pH on Burkholderia species.

Fig. 1. Relative abundance of Burkholderia 16S rRNA genes in 44 Results soils. A. Representation of Burkholderia 16S rRNA gene relative Low pH tolerance, a genus-wide property of abundance at the different sites as assessed by qPCR. Relative abundance is represented by the circle size; the colour indicates Burkholderia that largely accounts for its relative the pH of the sampled soils. abundance in soils B. Influence of pH on the relative abundance of Burkholderia 16S rRNA genes. Circles represent the average of three replicates for To test whether the occurrence of Burkholderia species in each soil sample. acidic environments reflects an intrinsic capacity of this genus to tolerate low pH conditions, we tested the ability Soil pH has frequently been shown to be the main of 68 strains of Burkholderia belonging to 31 different predictor of overall soil bacterial community composition, species to grow at a pH range of 3.5–8. All Burkholderia diversity and the relative abundance of many individual strains that were tested in physiological assays grew in taxa (Tiedje et al., 1999; Fierer and Jackson, 2006; Rousk pH as low as pH 4.5. Out of 68 tested strains (31 different et al., 2010; Griffiths et al., 2011). Although it is not known species), 32 (18 species) were growing also at pH 4 and if Burkholderia distributions are related to soil pH, most 15 (8 species) even at pH 3.5, but no species-specific studies that have reported their presence in soil were tolerance could be observed under such conditions investigating acidic environments (Trân Van et al., 2000; (Table S1). Curtis et al., 2002; Salles et al., 2002; 2004; Belova et al., Based on this result, we hypothesized that members of 2006; Garau et al., 2009; Aizawa et al., 2010). For the Burkholderia genus would be favoured in low pH instance, Burkholderia unamae could only be isolated environments. We tested this hypothesis by analyzing the from the rhizosphere of plants growing in soils, with a pH relative abundance of soil Burkholderia in a trans- ranging from 4.5 to 7.1, but not from soils with a pH higher continental sample collection of 44 soils from a broad than 7.5 (Caballero-Mellado et al., 2004). Likewise, a array of ecosystem types that represent a wide range of survey of over 800 Australian soil samples revealed that soil and site characteristics (Table S2). The relative abun- B. pseudomallei was specifically associated with low pH dance of Burkholderia species greatly varied between soils, but not recovered from higher pH soils (Kaestli these different soils sampled across North and South et al., 2009). Burkholderia species have also been iso- America (Fig. 1A). Highest relative abundance was lated from acidic Sphagnum peat bogs (Belova et al., observed in moderately acidic soils (pH 5–pH 6), where 2006; Opelt et al., 2007a; 2007b), from root tissues of the up to 6.7% of the total bacterial population was repre- highly acidifying cluster rooted Lupinus albus (Weisskopf sented by Burkholderia species. It is worthwhile to notice et al., 2011) or from soils as acidic as pH 2.9 (Curtis et al., that within this pH range a large variability in Burkholderia 2002). To the best of our knowledge, only one study relative abundance was observed (standard devia- reported isolation of Burkholderia strains from an alkaline tion =±2.02), spanning from 0.04% to 6.25%, whereas in environment (Estrada-de los Santos et al., 2011). While more acidic soils (pH <4) relative abundances were the pH of the rhizosphere soil investigated in this study approximately 1% or higher. While high relative abun-

3.5 to 6.8), C/N ratio, location and relative abundance of Burkholderia 16S rRNA genes were selected from the R = −0.76 trans-continental scale sampling set for phylogenetic analyses. Clone libraries targeting the 16S rRNA gene were constructed for each of the selected sites, and a total of 675 sequences (590 bp) were obtained, corresponding to 123 operational taxonomical units (OTUs) at a 98% identity threshold between Burkholderia 16S rRNA gene 0.05 0.10 0.15 0.20 sequences (Fig. 3). Diversity and richness of the soil Relative abundance (%) abundance Relative

0.00 Burkholderia communities were highly variable between 4.5 5.0 5.5 6.0 6.5 7.0 7.5 the sites, e.g. only one phylotype was found in KP2 or pH PE3, but CL4 and PE5 harboured 33 and 25 phylotypes respectively (Fig. 3). However, there was no significant Fig. 2. Relative abundance of Burkholderia 16S rRNA genes along correlation between pH and Burkholderia diversity. To test a local pH gradient in an agricultural field in Scotland. Relative abundance was quantified by qPCR in three soil samples for each whether pH or any other of the described environmental pH along the pH gradient. Abundance of Burkholderia is strongly parameters could influence Burkholderia community com- and negatively correlated with pH (Pearson’s product-moment position, a Mantel test was performed (see Table 1). Our correlation: R =−0.76, P < 0.001). data showed that pH had no correlation with community

structure (rM = 0.110, P = 0.204), while site elevation and spatial parameters did significantly positively correlate dance of Burkholderia 16S rRNA gene copy numbers was with Burkholderia communities (rM = 0.39, P = 0.002 and detected in soils with a pH lower than 7, the relative rM = 0.38, P = 0.038 respectively). Burkholderia commu- abundance was under the detection limit of our qPCR nity structure was also marginally influenced (P < 0.1) by −1 method (less than 100 copies reaction ) in neutral and climatic factors and soil chemistry (rM = 0.295, P = 0.077 alkaline soils (Fig. 1B). As expected, pH was a significant and rM = 0.260, P = 0.094 respectively). These results factor predicting Burkholderia relative abundance indicate that low pH would generally affect the relative (P = 0.03, R =−0.33), although the correlation was weak, abundance of the Burkholderia genus, but not the relative probably due to the high variability of Burkholderia relative abundances of individual species within this genus, which abundance in low pH soils. Moreover, pH is not the only is in line with our observations that low pH tolerance is a variable that changes across the soils analyzed, and pH genus-wide feature of Burkholderia sp. (Table S1). often correlates with other soil and site characteristics. The C/N ratio showed the best correlation (P = 0.0005, Burkholderia glathei: a major and widespread R = 0.50) with Burkholderia relative abundance, but since soil inhabitant pH and C/N ratio correlate (P = 0.003, R =−0.4380), we tested our hypothesis in a different experimental set-up, in Within the entire sequence set, those closely related to which the effect of pH could be discriminated from that of B. glathei were by far the most abundant and most widely C/N ratio. To this end, we analyzed the relative abundance distributed of all (approx. 40% of sequences). These of Burkholderia in an agricultural field with a pH gradient sequences comprised four OTUs, which contained 190, of 4.5–7.5 but a constant C/N ratio. Relative abundance of 33, 32 and 12 sequences respectively (Fig. 3). The most Burkholderia 16S rRNA was lower in this soil than in soils widespread and abundant OTU (190 sequences) was collected across North and South America. Highest rela- present at nine sites out of 14. Interestingly OTU 3, the tive abundance was detected at pH 4.5 (0.23–0.16%), next most abundant OTU of this B. glathei group, was and an almost linear decrease with increasing pH was present only at one site (HI3) and represented all of the 33 observed, reaching 0.01–0.008% of Burkholderia 16S sequences collected at this site. The next most abundant rRNA relative abundance in soil of pH 7.5 (P < 0.0001, OTU beside the B. glathei group was most closely related R =−0.76) (Fig. 2). to B. terricola (53 sequences). Unlike the B. glathei group, however, this OTU was only found in one site (KP2), where it was the only Burkholderia representative. Intrageneric diversity of Burkholderia soil populations Sequences closely related to B. phenazinium, does not depend on pH B. fungorum and B. terrae were very abundant as well, while other OTUs were represented by less than 20 To analyze the intrageneric diversity of soil Burkholderia sequences, and a high proportion (103 OTUs) consisted populations and to determine whether some groups of less than five sequences (Table S3). Despite the high showed any pH preference, 14 sites varying in pH (from diversity of Burkholderia in some of our soil samples,

Number of Closely related Burkholderia 100% sequences species from BLAST OTU 1 190 B. glathei; Hg 4 90% OTU 2 53 B. terricola (T); LMG 20594T OTU 3 33 B. glathei (T); LMG14190T OTU 4 32 B. glathei; Hg 5 80% OTU 5 29 B. phenazinium; Hg 8 OTU 6 27 B. fungorum; W566 OTU 7 26 B. terrae; KMY01 70% OTU 8 17 B. hospita; LMG 20574 OTU 9 15 Burkholderia cepacia complex OTU 10 14 B. sartisoli (T); RP007 60% OTU 11 13 B. phenazinium (T); LMG2247T OTU 12 12 uncultured eubacterium WD227 OTU 13 12 B. glathei; Hg 5 50% OTU 14 8 B. soli (T); GP25-8 OTU 15 8 Burkholderia sp. DM-Ni1 40% OTU 16 7 Burkholderia sp. TNFYE-5 OTU 17 7 B. phytofirmans (T); PsJN OTU 18 7 B. caribensis; MWAP71 30% OTU 19 6 B. mimosarum; Br3467 OTU 20 6 B. unamae (T); MTl-641 OTU 21 –123 5 < Burkholderia sp. 20%

10%

0% OTU 21-123 OTU 1 PE5 BF2 BF1 TL3 CL4 DF1 HJ1 SK1 CL1 SK3 SK2 KP2 HI3 SR1 pH 3.57 3.61 4.05 4.23 5.03 5.37 5.41 4.45 5.68 5.83 6.18 6.50 6.53 6.83

group No. of sequences 47 58 41 44 65 40 76 57 51 51 37 53 33 22 OTU 3 No. of a OTU 4 phylotypes 17 10 8 5 12 16 3 3 10 8 8 1 1 13 OTU 13 Coverage b OTU 2 0.45 0.86 0.91 0.91 0.68 0.50 1 0.95 0.86 0.73 0.82 1 1 0.50 Shannon c

B. glathei 2.78 2.19 1.95 1.18 2.26 2.69 0.84 0.49 2.19 1.43 1.62 0 0 2.22 Simpson d 0.93 0.8 0.84 0.62 0.87 0.93 0.48 0.24 0.88 0.62 0.71 0 0 0.84 a Number of phylotypes was calculated based on rarefied sequences number b Coverage was calculated based on rarefied sequences number c Shannon -Weaver diversity indexwas calculated based on rarefied sequences number d Simpson diversity index was calculated based on rarefied sequencesnumber

Fig. 3. Intrageneric diversity of soil Burkholderia. Pie chart represents the total number of OTUs obtained from 14 sites. Pieces represent OTUs that contain more than five sequences and are ordered by size (number of sequences). Dark green represents OTUs that contain less than five sequences. The largest group of sequences, which is represented by strains of B. glathei, is highlighted. The bar chart represents the relative abundance of OTUs per site, ordered by pH. The table contains the diversity indices calculated using rarefied sequence number (22 sequences) per site. All indices were calculated using 98% identity between sequences.

rarefaction curves reached a plateau in almost all soils Burkholderia are relatively more abundant analyzed (Fig. S1). The least diverse sites were HI3, KP2 in low pH soils and SK1 having less than four OTUs (Fig. 3). In all except Our results demonstrated a negative correlation between for KP2, the majority of sequences were affiliated with pH and relative abundance of Burkholderia 16S rRNA B. glathei (100% at site HI3 and 93% at site SK1), suggesting a preeminent role of this species in the soil. Table 1. Relationship of Burkholderia community structure to combined and individual environmental parameters revealed by Mantel Discussion test.

Low pH tolerance is a general property of the genus Environmental parameters rM P Burkholderia Soil chemistry 0.260 0.094 All Burkholderia strains that were tested on minimal pH 0.110 0.204 C/N ratio −0.197 0.916 medium showed tolerance to acid pH (pH 4.5). Similar % organic C 0.226 0.126 results were obtained by Estrada-de los Santos and Climatic 0.295 0.077 colleagues (2011), who reported that most of the 43 tested MAT 0.273 0.042* MAP 0.141 0.231 Burkholderia species were able to grow in a pH range of Soil 0.176 0.170 5–11, although the medium used in their study was neither % silt and clay 0.289 0.030* − buffered nor was growth quantitated. Our results are in Depth of O horizon 0.119 0.666 SMD 0.125 0.236 agreement with previous reports (Belova et al., 2006; Biological 0.227 0.125 Aizawa et al., 2010; 2011; Schmerk et al., 2011) and C mineralization rate 0.226 0.134 suggest that Burkholderia are acidotolerant rather than Spatial (longitude, latitude) 0.379 0.038* Site elevation 0.387 0.002* acidophilic. The unveiling of this genus-wide acid tolerance allows conclusions on the lifestyle and environmen- Parameters highlighted in bold represent combined matrices that tal adaptation of these bacteria, and also offers new were used and included factors highlighted in italic (used also sepa- rately). Bold values represent signiﬁcant P values (< 0.1). P values possibilities to select or enrich Burkholderia isolates from < 0.05 are indicated in bold and with an asterisk. rM, Mantel’s corre- complex environments. lation coefficient.

© 2013 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology, 16, 1503–1512 Preference of Burkholderia sp. for acid soils 1507 genes: Burkholderia relative abundance was higher in we observed very high relative abundance of B. glathei, acidic than in neutral soils and was absent or under the which was in this study by far the most abundant and detection limit in alkaline soils. The highest relative widespread OTU (Fig. 4). B. glathei has been previously abundance was observed in moderately acidic soils shown to be widely distributed across soils and (pH 5–6), where Burkholderia represented 6.25% of the rhizospheres (Belova et al., 2006; Uroz et al., 2007; total bacterial 16S gene copies. Similar relative abun- 2012). Here, we show that members of this species are dances of Burkholderia were previously observed in not only very abundant in different soil and ecosystem pyrosequencing studies investigating acidic bulk soils types but are also extremely widespread over diverse (1.2%), while abundances increased in the rhizosphere geographical sites. This suggests that B. glathei is a (1.96–3.08%) and even more in the mycorrhizosphere preeminent soil inhabitant, which is particularly well environments (3.30–8.33%) (Uroz et al., 2010; 2012). C/N adapted to this type of environment, although the specific ratio was another environmental parameter that signifi- functions responsible for the success of this species in cantly influenced the relative abundance of Burkholderia soil remain undetermined. In addition to site descriptors populations in soils, but this parameter was also, as often analyzed in this study, biological factors may have an observed, strongly correlated with pH (Kemmitt et al., important role in shaping Burkholderia community com- 2006). For this reason, we conducted a local-scale study position and might be responsible for the highly variable on an agricultural field with a pH gradient where the C/N intrageneric diversity observed in the selected soil ratio is fairly constant and where the aboveground plant samples. A good example of such biological factors is the community is the same. Our data showed that the effect of symbiotic association between nitrogen-fixing Burkhol- pH on Burkholderia relative abundance was even deria species and plants. For example, Burkholderia stronger than what was observed on the trans-continental mimosarum is capable of nodulating Mimosa plants, and scale, with an almost linear decrease with increasing pH, is therefore only found in areas where the plants are which strongly suggests that pH, rather than C/N ratio, endemic, such as tropical regions of South-eastern Asia influences the distribution of Burkholderia populations in and South America (Chen et al., 2006; Elliott et al., 2009). soil. In line with this, we detected B. mimosarum only at site PE5, which is located in Peruvian Amazonas. In summary, this study showed that low pH tolerance is Intrageneric diversity of Burkholderia soil populations a genus-wide feature of Burkholderia species. This does not depend on pH explains their presence in acidic soils but not their Previous studies investigating the phylogeny of absence from higher pH environments, especially consid- acidobacteria have shown that their relative abundance ering that under laboratory conditions, the majority of and intrageneric diversity are higher in low pH soils. Inter- Burkholderia strains are able to grow in neutral or even estingly, certain subgroups within this genus were alkaline culture media. This suggests that Burkholderia identified, which were only found in neutral or even in have developed pH tolerance mechanisms that enable alkaline soils (Lauber et al., 2008; Jones et al., 2009; them to survive and thrive in environmental niches where Griffiths et al., 2011). Since our results of Burkholderia many other taxa are inhibited, while they are outcompeted relative abundance are similar to the trends observed by faster growing microorganisms in less harsh condi- for acidobacteria, we investigated whether certain tions. Acid tolerance is a prerequisite for occurrence in low Burkholderia lineages would have a preference for soils pH soils, but it is tempting to postulate that the preference with a particular pH. However, our diversity analysis of Burkholderia for such niches is not only the conse- showed no correlation between pH and community com- quence of the ability to tolerate acidity, but the result of a position within the genus. While the intrageneric diversity multifaceted strategy involving both tolerance to abiotic varied greatly between the samples, no OTU was found stress factors (such as higher toxicity of heavy metals) that was specifically enriched in highly or moderately and to biological constraints (e.g. the predominance of acidic soils. This is in line with our in vitro low pH tolerance fungi) inherent to such environments. assays, which suggested that pH tolerance is a general feature of the genus Burkholderia. Interestingly, patho- Experimental procedures genic species, such as B. pseudomallei or B. mallei,or opportunistic pathogens, such as members of the Bcc, Testing growth of Burkholderia strains in vitro under were very rarely detected in our soil survey, indicating that different pH conditions while they have been reported to be major inhabitants To study the effect of pH on Burkholderia growth, an in vitro of maize (Bevivino et al., 2011) or sugar cane approach was used. To this end, 68 Burkholderia strains from (Castro-González et al., 2011) rhizospheres, they are not different isolation origins were used (Table S1). Before spot- commonly present in nutrient-limited bulk soil. In contrast, ting 20 μl aliquots of each culture on growth medium, over-

’Burkholderia sp. IMER

’Burkholderia sp. RKJ800’’unculturedsoilbacterium’ ’uncultured soil bacterium’ ’uncultured Burkholderia sp.’ ’Burkholderia sp. IMER ’Burkholderia sp.

’Burkholderia sp. IMER

SK2 TL3 S SR1 SK3 K TL3 SK3 TL3 ’Burkholderia sp. R2G3’ HJ1 HJ1 7

HJ1 TL3 SK3 K 2 SK3 HJ1 E10 TL3 A C9 B2 B10 SK3 F8 E1 D9 C10 − 3 D10 F4 SK2 20 B6 − E5 SK2 16 −

− − − − D12 − SK3 12 − − − − − − − C − − − − G3 SK2 22 − − − − F1 2 − D8 SK3 20 F12 − − − C3 A12 C7

G4 E5 B SK3 14 SK2 4 F9 − F10 − SK3 15 − C4 A1 F3 1 − D10 H8 − − − K B7 B5 F9 − − − − SK1 SK2 15 1 1 B3 2 G7 D8 B1 − − G7 E8 H4 TL3 − TL3 SK1 HJ1 SR1 16 S B1 TL3 G3 TL3 A2 SK2 TL3 SK3 − SK2 − SK2 SK2 − B12 SK1 A4 SK1 − A1 TL3 TL3 TL3 H5 TL3 − CL4 TL3 − − SK3 − H10 − TL3 − A5I55’ BF2 A7 F7 − 31’ SK3 − CL4 D8 TL3 SR1 SK3 − H9 C3 SK3 − − H5 SK2 − H10 CL4 TL3 − F2 − A6 HJ1 SK3 − C1 C1 SK2 − HJ1 − SR1 2 − TL3 F10 − − SK3 G6 CL4 26 A5 TL3 − B9 B1 − D1 1 SK3 − SK1 − TL3 − C6 G7 SK1 A1 HJ1 − SK1 BF2 − ’unculturedbacterium’ B12 SK1 TL3 D11 ’Burkholderia sp. EBA04’ HJ1 − 42’ TL3 − C8 HJ1 − BF2 − SK1 − A3 − G5 17’ TL3 A4 − SK3 3 − HJ1 F3 SK3 2 − B8 HJ1 − G8 HJ1 − HJ1 − E6 SK1 − A4 SK1 C5 − G1 − G12 SK2 − D1 B3 TL3 SK1 − − − C12 − SK2 D4 SK1 C10 E9 HJ1 − SK1 SK1 − D11 HJ1 − B3 HJ1 A10 SK1 − C2 − SK1 − E8 SK1 − G1 HJ1 − E12 − − D2 HJ1 − C12 HI3 TL3 − C10 D6 HJ1 − F2 H1 HJ1 − H8 − SK1 − B8 HI3 − − D3 CL1 − A12 HI3 1 SK1 G10 A4 − C3 SK1 C10 HI3 HI3 SK1 − − − − F9 G9 SK1 HI3 G12 SK1 − A3 1 − − HJ1 − D10 HI3 G3 HI3 H12 SK3 − A1 − BF1 − D8 1 HI3 HI3 − H6 A5 HI3 SK1 − − B3 SK1 − E1 HI3 H9 − − HJ1 − B10 1 − C12 A5 HJ1 − HI3 C8 − C1 SK2 − E1 HI3 HI3 E1 HJ1 − B7 1 1 − SK2 − C7 HI3 − − C9 1 SK1 − 1 HI3 SK1 G1 D12 F1 − B1 HI3 − SK1 − HJ1 G6 HI3 − H7 − B12 ’Burkholderia glathei’ HI3 − C5 HJ1 − 1 C3 HJ1 H10 HI3 HI3 − − A1 B4 SK1 − − E9 HI3 − B9 SK1 1 − B. glathei group SK1 − H9 C10B7 HI3 SK2 − A1 HI3 − − HI3 D5 SK1 − SK1 C3 − A6 − − A1 SK1 HI3 A10 ’Burkholderiaglathei’D10 HI3 SK2 − ’Burkholderia sp. Y86’ HI3 − G3 HI3 C6 SR1 11 − A12 − − HI3 − F2 BF1 H6 − A1 BF1 − A12 E4 HI3 − 1 BF2 − F7 ’Burkholderia sp. NF23’ HI3 D1 BF2 − SR1 − 1 BF1 − A8 G5 C1 BF1 − − SK3 23 A6 SK3 25 A1 − G10 SK3 HJ1 − B5 TL3 HJ1 − − HJ1 H9 HJ1 − SK1 − B8 HJ1 H7 SK3 13 SK2 21 HJ1 E7 − SK2 17 − HJ1 − H6 HJ1 G4 ’Burkholderia sp. DF4MM3’ HJ1 F3 HJ1 − − − F8 ’unculturedsoilbacterium’C8 HJ1 B5 − HJ1 − TL3 − A5 TL3 H3 SK3 −18 A9 − F10 ’Burkholderia−B7 sp. DF4EH2’ HJ1 TL3 TL3 BF1 F4 −H4 − TL3 − SK3 8 B6 SK3 E3 DF1 − HJ1 −D4 − TL3 SK1 − D1 H12 SK3 21 −G4 SK3 − SK2 18 E1 SK3 B1 SK1 − − −A9 SK3 10 A7 HJ1 B6 −D6 − 1 SK1 SK1 F1 SK1 SK1 − SK1 −G2 −F5 HJ1 C7 −B2 − SK3 −E12 HJ1 H8 SK1 SK1 − DF1 −H6 −H5 − HJ1 B9 ’Pandoraea norimbergensis’DF1 − F8 HJ1 − ’unculturedbacterium’ H12 −F10 BF2 −C11 SK1 F6 − SK1 − HJ1 E12 −H2 CL1 HJ1 ’Pandoraea apista’ −E5 HJ1 −A7 −D7 HJ1 −E12 TL3 −F4 TL3 −E4 TL3 −B10 CL1 −B4 SK3 −A2 TL3 −B3 TL3 −B1 SK3 −E2 SK3 −E12 SK3 SR1 5−G5 SR1 SR1 −7E9 TL3 ’unculturedBurkholderiasp.’−D2 TL3 −E10 TL3 −D3 TL3 −D6 HJ1 −E4 HJ1 −D7 HJ1 −G9 SK1 −C1 HJ1 −E4 SK1 −B1 SK2 E2 DF1 − SK1 C8 DF1 −H10 − − HJ1 D3 DF1 E12 − − HJ1 −A8 DF1 D4 SK1 −E1 −H6 DF1 1 SK1 B4 DF1 −G2 − ’Burkholderia’Burkholderia sp. IMER terrae’− HJ1 −E10 DF1 G10 HJ1 −E10 −G3 HJ1 −G7 −A1 SK1 −D5 CL1 −4’ HJ1 D1 DF1 −A5 − − HJ1 −E8 DF1 B11 HJ1 −C2 ’Burkholderia andropogonis’ DF1− CL1 C7 SR1 13 −H6 ’unculturedbacterium’ BF1− −D1 BF2 F6 PE5 ’Burkholderia sp. EBA07’−G8 −F6 BF1 PE5 −E9 −D12 ’Burkholderia hospita’ PE5 G5 B. terrae PE5− DF1 CL4 31 −G3 CL4 7 CL1−E6 CL1 CL4 40 −D3 CL4 27 CL1−A4 CL1 CL4 32 −H10 CL4 17 CL1−D8 CL4−E8 CL1−G6 CL4−E4 CL1−F9 CL4−F8 DF1−G5 CL4−G6 CL1−B1 CL4 35 CL1−B11 CL4−H2 ’unculturedBurkholderiasp.’TL3−G4 CL4 25 ’Burkholderia sp. TFD2’ CL4 38 4−D−degradingbacteriumTFD2’ CL4 34 CL4 39 ’Burkholderia sp. BCB ’2 CL4−D3 ’Burkholderia phymatum’−6’ CL4 20 CL4 2 CL1−G5 CL4 24 CL1−G10 B. rhizoxinica CL4 30 CL1−A12 B. caribensis CL4−F9 DF1−B8 CL4 28 CL4−G3 CL4 29 ’Burkholderiacaribensis’ CL4−C1 ’Burkholderia sp. A3I14’ BF2−C7 CL1−F7 ’Burkholderia rhizoxinica’ ’Burkholderia caribensis’ CL4 19 DF1−G11 ’Burkholderia soli’ CL4−B3 CL4−H9 PE5−G6 ’Burkholderia thailandensis E264’ PE5−D7 ’unculturedbacterium’ CL4 23 CL1−B3 CL1−E9 ’Burkholderia sp. JPY157’ HJ1−H9 ’Burkholderia sp. hpigPE5 15.6’−A5 HJ1−D9 PE5−G7 HJ1−E2 PE5−C11 CL4 4 PE5−G3 CL4 15 PE5−G11 CL4−B11 PE5−H4 CL4 10 PE5−B5 CL4−D5 −H6 CL4 33 PE5 CL4 3 PE5−A10 −C6 CL4 18 CL1 ’Burkholderia ambifaria’ ’Burkholderia pseudomallei’ PE5−D9 −B11 ’Burkholderia gladioli’ PE5 ’Burkholderia gladioli pv ’Burkholderia sp.enrichment culture clone F36bPE5 06b’−H5 −B9 ’Burkholderia plantarii’ PE5 B. cepacia ’Burkholderia sp. PSB73’. gladioli’ −H8 ’Burkholderia glumae’ ’unculturedbacterium’PE5 −H11 PE5−H1 PE5 −B6 PE5− PE5 PE5 B2 −G2 −E1 PE5 A8 ’Burkholderia arboris’ PE5− E10 CL4− PE5− PE5 B9 −E11 PE5− complex ’BurkholderiaF4 ubonensis’ −A2 ’Burkholderia unamae’PE5 PE5 ’Burkholderiaunamae’−D5 −F1 CL1 −1’ ’Burkholderia1 sp. CNW17’ −B8 ’Burkholderia cepacia’ CL1 CL4 1 −B1 1 ’Burkholderia sp.BF2 H6 01’ CL4 14 − CL4 −E8 − PE5 CL4 C7 −C3 −H10 ’Burkholderia sp. OXPE5 CL4 CL4 −B7 −C6 − PE5 CL4 5E6 ’Burkholderia nodosa’ CL4 B. unamae, CL4 −F1 16’ −H5 G9 CL4 ’Burkholderia mimosarum’− CL4 −E1 PE5 − 1 ’Burkholderiamimosarum’ F1 CL4 C9 ’Burkholderia sp. Ms1 − BF1 −D8 PE5 −F9 − PE5 C5 BF1 E10 ’Burkholderia ferrariae’− BF1 −G7 PE5 C12 − − C5 B. nodosa ’BurkholderiaC3 sp. H801’ PE5 − ’unculturedbacterium’ BF1 ASS3’ ’Burkholderia sp. U1 −G5 ’uncultured eubacterium WD227’ CL4 −B7 B5 SK3 BF2 − DF1 −E4 ’Burkholderia sp. BF2 −C3 − 1’ SK3 F2 −4’ BF2 DF1 − C9 F5 − and B. tropica CL1 −B12 PE5 ASS7’ CL1 −C8 CL1 −A7 −F8 DF1 − ’Burkholderia sp. ASS1 F7 B5 PE5 − CL1 −C3 CL4 ’Burkholderia sp. 49’ SK3 −B2 JPY 389’− SK2 3− − −B1 −18’ H8 ’unculturedBurkholderiasp.’A1 SK3 7 − −F3 group ’Burkholderia phytofirmans’ PE5 ’Burkholderiaphytofirmans’ CL4 16 ’Burkholderia phytofirmansPsJN’ CL4 21−A5 DF1 BF1 ’Burkholderia sp. NKMU DF1 −H4 ’Burkholderia sp. IMER CL4− 6B8 2010 − ’Burkholderia sp. IMER E5 B. fungorum C7 PE5 − BF1 BF1 CL1 E4 −E8 − SK3 19− −B1 ’BurkholderiaG8 caledonica’ DF1 D9 DF1 − ’Burkholderia sp. R4M C4 B. phenazinium CL4 − BF2 C9 − 5’ ’Burkholderia− sp. Ch1 BF1 − ’BurkholderiaH1 sp. Ellin138’ DF1 1 ’unculturedBurkholderiasp.’ −F9 ’Burkholderia graminis’ ’Burkholderia sp. M14’ DF1−B12 −B4 PE5 BF2 G6 BF1 − CL1 − A3 − I’ − SR1 − D2 KP2 D4 ’BurkholderiaD2 ginsengisoli’ − KP2 − − 1’ H10 1 ’unculturedB6 betaproteobacterium’ ’Burkholderia sp. TNFYE KP2 − F1 DF1 − D8 BF2 KP2 − − KP2 −C4 BF2 B3 KP2 E7 CL1 −F10 − − KP2 B10 CL1 C8 KP2 − E1 CL1 −F2 − − D3 ’BurkholderiaC9 sp. LMG22945’ KP2 − A9 − KP2 − DF1 G1 KP2 D6 SK3 16 − DF1 − KP2 C10 B. sartisoli G4 KP2 − A7 B. terricola B. fungorum SK3 − − B2 SK3 D2 KP2 − B4 SK3 − KP2 − 1 − A4 DF1 E3 KP2 B1 SK3 − KP2 − E10 1 E9 − CL4 − C1 CL4 − H1 KP2 − G5 E8 KP2 − G4 CL4 − − ’unculturedbacterium’− A1 KP2 G9 C3 KP2 − C5 BF2 − KP2 − F8 BF2 E10 KP2 − BF2 − KP2 D12 BF2 A7 − C2 − KP2 − F9 BF2 − E2 − E4 BF1 − H9 KP2KP2 − F6 BF1 − G9 KP2 − D9 ’Burkholderia− E10sp. 2As’ KP2 − G3 ’Burkholderia− G4phenazinium’ KP2 − D2 BF2 H5 KP2 − H3 BF2 − H5 CL1 − KP2 ’unculturedBurkholderiasp.’ KP2 − B5 − H3 − H4 ’Burkholderia− sp.G12 SBH KP2 1 CL1 − 1 C1 KP2 H1 BF2 KP2 − E1 CL1 1 KP2 ’Burkholderia −sp. DM − E12 H9 G2 − CL1 − KP2 − C1 − A1 − F5 BF1 KP2 BF2 E8 − − KP2KP2 B12 BF2 G2 DF1 − C1 KP2 − − F7 − B5 KP2 − ’Burkholderia sp. CS2’ ’unculturedBurkholderiasp.’− C2 D10 D5 KP2 ’Burkholderiaphenazinium’− D3 KP2 − − F3 CL1 B6 KP2 − HJ1 F10 C6 1 SK2 9 − − 8’ KP2 KP2 − SK2 1 − KP2 D1 ’Burkholderia sp.N1US7’− D10 − A11 F2 DF1 F7 KP2 KP2 − − C9 ’Burkholderia sp. DM − − BF2 Cd1’ DF1 1 KP2 C12 − − BF2 KP2 SR1 BF2 H2 SR1 D1 BF2 − − − D3 BF2 C12 SK2 19 E7 BF2 − D12 SR1 SK2 12 − − BF2 − F3 E1 DF1 − B10 DF1 BF1 − D7 − G11 B8 DF1 − E8 BF1 − PE5 BF1 − D4 B. phenazinium − F9 PE5 SK3 22 − CL1 − F1 E3 − SK3 BF2 F3 − G7 BF2 − C10 BF2 B9 − 1 BF2 − BF1 D5 SK3 4 − BF1 − D8 BF1 D10 G2 CL1 − BF1 − H10 HJ1 − − BF1 D1 B12 1 − C10 − HJ1 BF1 − HJ1 − BF1 − F1 Ni1’ A10 BF2 A8 A1 CL1 − − − D5 ’Burkholderia terricola’ BF2 C11 ’Burkholderiasartisoli’ H9 HJ1 − C10 ’Burkholderia sartisoli’ − − A10 BF1 − F8 CL1 − BF1 − B9 D12 CL1 E3 − H12 F5 ’Burkholderia fungorum’ − − H7 ’Burkholderiafungorum’ − F5 BF2 − − G8 ’Burkholderiafungorum’ H5 BF2 BF2 D6 − 1 CL4 BF2 − ’Burkholderiafungorum’ − 44’ ’Burkholderiafungorum’ − BF1 ’unculturedBurkholderiasp.’ CL1 E1 − E5 − SR1 14 F1 F4 SR1 B3 CL1 SR1 A1 BF2 − C2 F1 SR1 CL1 − SR1 − E5 SR1 − CL1 SR1 − − SR1 − SK2 CL1 A3 SK2 − SK2 H10 SK2 C12 G4 1 SK1 E5 B11 D2 SK2 B2 BF1 ’unculturedbacterium’ − B6 ’Burkholderia sp. F4W’ TL3 ’Burkholderia sp. rif200871’ SK3 24 1 B

BF2 BF2 BF2 C5 BF1 − − − BF1 BF1 − − − SK1 − H2 52’ − − CL1 G7 F7 − SK1 − E10 A1 F − CL1 F2 E12 − − − SK1 F8 − F5 H6 − − E5 − DF1 − G8 − − E4 G1 1 − − D2 E10 −

− − E1 − − − − E3 BF2 G10 G9 F1 − − − − − DF1 BF2 − − − H3 BF2 BF2 B1 − H1 BF2 B6 F3 B7 BF2 −

B BF2 H2 D1 H11 SK2 1 D7 E1 C9 − CL1 CL1 DF1

DF1 1 SK2

BF2 BF1

BF2

’Burkholderia sediminicola’ BF1 0

SK2

BF2

’Burkholderia sp. UCT71’

’Burkholderia sp. II

’Burkholderiaphenazinium’ ’Burkholderiaphenazinium’

’unculturedBurkholderiasp.’

’Burkholderia sp. IMER

Fig. 4. Phylogenetic analysis of Burkholderia 16S rRNA gene sequences from 14 different sites. 675 sequences (590 bp) were aligned with additional reference sequences using ARB. Phylogeny was constructed using a maximum likelihood-based method. Bootstrap values (n = 1000) >50% are shown as circles. Colours indicate the affiliation with a given species. night liquid cultures were washed and resuspended in saline liquid conditions. Media were buffered with 0.1 M potassium buffer to optical density (OD600) of 1. As growth medium, AB hydrogen phthalate (C8H5KO4,pKa = 5.4) and 0.1 M HCl minimal medium (Clark and Maaløe, 1967) supplemented for pH 3, 0.1 M C8H5KO4 and water for pH 4, 0.1 M with glucose and agar was used, and the pH of the medium NaOH and 0.1 M C8H5KO4 for pH 5–6, and for media was adjusted to obtain a pH gradient of pH 4–7 in 0.5 unit higher than pH 7, 50 mM HEPES (4-(2-hydroxyethyl)-1- steps, with an additional medium of pH 8. To test growth piperazineethanesulfonic acid) was added. To detect under more acidic conditions, liquid AB medium supple- changes in pH caused by bacterial growth, resazurin was mented with glucose was used, adjusted to pH 3, pH 3.5 and added as a pH indicator. Plates were incubated for 4 days at pH 4. Liquid AB medium with pH 4 was used as a control to 30°C. Growth was assessed by inspecting the plates for the test if the growth patterns were the same between solid and formation of colonies.

Soil sampling and DNA extraction USA) and 2 μl of template DNA. PCR was carried out in a C1000 Thermal Cycler (Bio-Rad, United Kingdom). PCR prod- The dataset consisted of 44 soil samples distributed across ucts were confirmed by standard 1% agarose gel electropho- North and South America. The collected soils came from a resis and gel purified (Gel PCR purification kit, QIAGEN, broad range of ecosystems, climates and soil types Hilden, Germany). Burkholderia 16S rRNA clone libraries (Table S2). Soil collection protocol and methods for edaphic were made from 14 selected locations (Table S2). Purified and environmental properties have been described previ- PCR products were cloned into the pGEM-T Easy vector ously (Fierer and Jackson, 2006; Bates et al., 2011). In addi- (Promega, Southampton, United Kingdom). Selected clones tion, seven soil samples were collected in March 2011 from from 16S rRNA clone libraries were sequenced using the M13f an agricultural field divided into several plots with a pH gra- vector primer. dient of 4.5–7.5, which has been maintained since 1961 by the addition of either lime or aluminium sulphate, and where plots undergo an 8-year crop rotation cycle (Scottish Agricul- Sequence analysis tural College, Aberdeen, Scotland; grid reference NJ872104). Detailed soil characteristics are provided by Kemp and Sequences of chimeric origin were detected by analyzing colleagues (1992). The soils were sampled in triplicates from alignments using Chimera.Slayer and Chimera.UCHIME as the upper 20 cm soil layer, homogenized and stored at 4°C implemented by the MOTHUR software (Schloss et al., 2009; prior usage. Total nucleic acids were extracted from 0.5 g of Edgar et al., 2011; Haas et al., 2011). Sequences from short soil, as described by Griffiths and colleagues (2000) with or failed reads were excluded from analysis. Sequences were some modifications (Nicol et al., 2005). pH was measured in aligned using the SINA web aligner (Pruesse et al., 2007). The deionized water using a ratio of 1:2 soil : water (w/v), shaking alignments were merged into the SILVA SSU reference data- for 30 min and settling for 30 min before measurement. base release 106 using the ARB software package (Ludwig et al., 2004). Sequences were deposited to the National Center for Biotechnology Information database with acces- qPCR sion numbers KC353471 to KC354145. A 50% similarity filter was created for the dataset, based on the alignment, leaving To quantify Burkholderia 16S rRNA genes in soil samples, 590 nucleotides for 16S rRNA sequence alignments. The primers BKH812F (5′-CCC TAA ACG ATG TCA ACT AGT closest cultivated relatives were selected from the reference TG-3′) and BKH1249R (5′-ACC CTC TGT TCC GAC CAT-3′) dataset. Bootstrapped maximum likelihood trees (1000 rep- (Bergmark et al., 2012) were used. In their original publica- etitions) were calculated with sequences affiliated with the tion, Bergmark and colleagues (2012) observed that the groups of interest and close relatives on a dedicated RAxML designed primers were not specific, suggesting that the most web server (Stamatakis et al., 2008). likely explanation for lack of specificity was that the Tm value they used was too low. We, therefore, tested both primers for their specificity using higher Tm values on DNA isolated from Phylogenetic analysis soils with different pH. An annealing temperature of 64°C was found to efficiently amplify Burkholderia 16S rRNA genes with Distance matrices were exported to calculate rarefaction 100% specificity. This was tested by sequencing 270 clones curves and diversity indices with the MOTHUR software containing PCR products from soils with three different pH. (Schloss et al., 2009). Sequences were grouped into opera- For both bacterial and Burkholderia 16S rRNA gene, high tional taxonomic units (OTU) using the furthest-neighbour amplification efficiency was obtained by qPCR (93–100% approach, with an OTU defined as containing sequences that and 91–100%, respectively, and r2 values between 0.995 and are no more than 2% different from each other. This threshold 0.999). Relative abundance of Burkholderia was calculated of 98% identity was selected because of the high similarity as a ratio between Burkholderia gene copy numbers by bac- between Burkholderia 16S rRNA sequences over the rela- terial gene copy numbers (see supporting information for tively short read length used in the present study. Richness more details). and diversity were estimated from 16S rRNA gene clone libraries using the Shannon–Weaver diversity index (H) (Shannon and Weaver, 1963) and the Simpson diversity Amplification and cloning of Burkholderia 16S rRNA index (D) (Simpson, 1949). Good’s coverage (C) was calcu- gene sequences lated as C = 1–(n1/N), where n1 was the number of clones, which occurred only once in a library of N clones (Good, To study the diversity of Burkholderia in soil, Burkholderia 1953), and relative abundances of major phylogenetic groups 16S rRNA genes were amplified using the modified were determined. primers BKH143F (5′- TGGGGGATAGCYCGGCG −3′) and BKH1434R (5′- TGCGGTTAGRCTAGCYACT −3′) (Schönmann et al., 2009). Cycling conditions were 95°C for Statistical analysis 3 min, 40 cycles of 95°C for 60 s, 61.5°C for 60 s, and 72°C for 90 s, final extension at 72°C for 5 min. Reactions were per- Pearson’s product-moment correlations between formed in 50 μl volumes containing 1× reaction buffer contain- Burkholderia 16S rRNA gene relative abundance and envi- ing MgCl2 (1.5 mM) (Sigma-Aldrich, St. Louise, MO, USA), ronmental parameters were performed in R 2.12.0 (http:// 0.8 μM of each primer (Microsynth, Balgach, Switzerland), www.r-project.org/). For correlating Burkholderia trans- 0.2 mM dNTP mixture, 0.25 mg ml−1 of bovine serum albumin, continental distribution, we used the following soil and site 2U of Taq DNA Polymerase (Sigma-Aldrich, St. Louise, MO, characteristics: soil pH, organic C content, C/N ratio, C

© 2013 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology, 16, 1503–1512 1510 N. Stopnisek et al. mineralization rate, elevation, soil moisture deficit (SMD), Caballero-Mellado, J., Martínez-Aguilar, L., Paredes-Valdez, mean annual temperature (MAT) and mean annual precipita- G., and Santos, P.E. (2004) Burkholderia unamae sp. nov., tion (MAP). For correlation analysis at the local scale, pH was an N2-fixing rhizospheric and endophytic species. Int J the only factor used. Syst Evol Microbiol 54: 1165–1172. We used Spearman’s rank correlation to compare estimate Castro-González, R., Martínez-Aguilar, L., Ramírez-Trujillo, of Burkholderia composition with site elevation, soil chemistry A., Estrada-de los Santos, P., and Caballero-Mellado, J. (matrix including pH, C/N ratio and percentage of organic C) (2011) High diversity of culturable Burkholderia species and climatic (MAT, MAP), soil (percentage of silt and clay, associated with sugarcane. Plant Soil 345: 155–169. depth of O horizon and SMD), biological (C mineralization Chen, W.-M., James, E.K., Coenye, T., Chou, J.-H., Barrios, rate) and spatial (longitude, latitude) parameters. To estimate E., de Faria, S.M., et al. (2006) Burkholderia mimosarum the pairwise similarity in Burkholderia communities, we gen- sp. nov., isolated from root nodules of Mimosa spp. from erated Bray–Curtis dissimilarity matrices, using rarefied abun- Taiwan and South America. Int J Syst Evol Microbiol 56: dance table of Burkholderia phylotypes (OTUs) as an input (22 1847–1851. sequences per location). We used the Mantel test in R 2.12.0 Clark, D.J., and Maaløe, O. (1967) DNA replication and the to compare dissimilarity matrices to pairwise distances in division cycle in Escherichia coli. J Mol Biol 23: 99–112. environmental characteristics as estimated using normalized Compant, S., Nowak, J., Coenye, T., Clément, C., and Euclidean distances in the measured soil and site parameters. Ait Barka, E. (2008) Diversity and occurrence of Burkholderia spp. in the natural environment. FEMS Microbiol Rev 32: 607–626. Acknowledgements Curtis, P., Nakatsu, C., and Konopka, A. (2002) Aciduric proteobacteria isolated from pH 2.9 soil. Arch Microbiol The authors would like to thank Prof. Jim Prosser, Dr. 178: 65–70. Graeme Nicol and Dr. Cécile Gubry-Rangin for valuable dis- Edgar, R.C., Haas, B.J., Clemente, J.C., Quince, C., and cussions, Prof. Jakob Pernthaler, Esther Eckert and Stefan Knight, R. (2011) UCHIME improves sensitivity and speed Neuenschwander for help with the ARB software, as well as of chimera detection. Bioinformatics 27: 2194–2200. the Swiss National Science Foundation for funding (grant Elliott, G.N., Chou, J.-H., Chen, W.-M., Bloemberg, G.V., 31003A-130089 to L.W. and Marie Heim-Vögtlin fellowship to Bontemps, C., Martínez-Romero, E., et al. (2009) N.B.). Burkholderia spp. are the most competitive symbionts of Mimosa, particularly under N-limited conditions. Environ References Microbiol 11: 762–778. Estrada-De Los Santos, P., Bustillos-Cristales, R., and Aizawa, T., Bao Ve, N., Vijarnsorn, P., Nakajima, M., and Caballero-Mellado, J. (2001) Burkholderia, a genus rich in Sunairi, M. (2010) Burkholderia acidipaludis sp. nov., plant-associated nitrogen fixers with wide environmental aluminium-tolerant bacteria isolated from Chinese water and geographic distribution. Appl Environ Microbiol 67: chestnut (Eleocharis dulcis) growing in highly acidic 2790–2798. swamps in South-East Asia. Int J Syst Evol Microbiol 60: Estrada-de los Santos, P., Vacaseydel-Aceves, N., 2036–2041. Martínez-Aguilar, L., Cruz-Hernández, M., Mendoza- Aizawa, T., Vijarnsorn, P., Nakajima, M., and Sunairi, M. Herrera, A., and Caballero-Mellado, J. (2011) Cupriavidus (2011) Burkholderia bannensis sp. nov., an acid- and Burkholderia species associated with agricultural neutralizing bacterium isolated from torpedo grass plants that grow in alkaline soils. J Microbiol 49: 867–876. (Panicum repens) growing in highly acidic swamps. Int J Fierer, N., and Jackson, R.B. (2006) The diversity and bio- Syst Evol Microbiol 61: 1645–1650. geography of soil bacterial communities. Proc Natl Acad Bates, S.T., Berg-Lyons, D., Caporaso, J.G., Walters, W.A., Sci USA 103: 626–631. Knight, R., and Fierer, N. (2011) Examining the global Garau, G., Yates, R.J., Deiana, P., and Howieson, J.G. (2009) distribution of dominant archaeal populations in soil. ISME Novel strains of nodulating Burkholderia have a role in J 5: 908–917. nitrogen fixation with papilionoid herbaceous legumes Belova, S., Pankratov, T., and Dedysh, S. (2006) Bacteria of adapted to acid, infertile soils. Soil Biol Biochem 41: 125– the genus Burkholderia as a typical component of the 134. microbial community of Sphagnum peat bogs. Microbiol- Good, I.J. (1953) The population frequencies of species and ogy 75: 90–96. the estimation of population parameters. Biometrika 40: Bergmark, L., Poulsen, P.H.B., Al-Soud, W.A., Norman, A., 237–264. Hansen, L.H., and Sørensen, S.J. (2012) Assessment of Griffiths, R.I., Whiteley, A.S., O’Donnell, A.G., and Bailey, the specificity of Burkholderia and Pseudomonas qPCR M.J. (2000) Rapid method for coextraction of DNA and assays for detection of these genera in soil using 454 RNA from natural environments for analysis of ribosomal pyrosequencing. FEMS Microbiol Lett 333: 77–84. DNA- and rRNA-based microbial community composition. Bevivino, A., Costa, B., Cantale, C., Cesarini, S., Chiarini, L., Appl Environ Microbiol 66: 5488–5491. Tabacchioni, S., et al. (2011) Genetic relationships among Griffiths, R.I., Thomson, B.C., James, P., Bell, T., Bailey, M., Italian and Mexican maize-rhizosphere Burkholderia and Whiteley, A.S. (2011) The bacterial biogeography of cepacia complex (BCC) populations belonging to British soils. Environ Microbiol 13: 1642–1654. Burkholderia cenocepacia IIIB and BCC6 group. BMC Haas, B.J., Gevers, D., Earl, A.M., Feldgarden, M., Ward, Microbiol 11: 228. D.V., Giannoukos, G., et al. (2011) Chimeric 16S rRNA

sequence formation and detection in Sanger and 454- Pérez-Pantoja, D., Donoso, R., Agulló, L., Córdova, M., pyrosequenced PCR amplicons. Genome Res 21: 494– Seeger, M., Pieper, D.H., and González, B. (2012) 504. Genomic analysis of the potential for aromatic compounds Jones, R.T., Robeson, M.S., Lauber, C.L., Hamady, M., biodegradation in Burkholderiales. Environ Microbiol 14: Knight, R., and Fierer, N. (2009) A comprehensive survey 1091–1117. of soil acidobacterial diversity using pyrosequencing and Pruesse, E., Quast, C., Knittel, K., Fuchs, B.M., Ludwig, W., clone library analyses. ISME J 3: 442–453. Peplies, J., and Glöckner, F.O. (2007) SILVA: a compre- Kaestli, M., Mayo, M., Harrington, G., Ward, L., Watt, F., Hill, hensive online resource for quality checked and aligned J.V., et al. (2009) Landscape changes influence the occur- ribosomal RNA sequence data compatible with ARB. rence of the melioidosis bacterium Burkholderia Nucleic Acids Res 35: 7188–7196. pseudomallei in soil in Northern Australia. PLoS Negl Trop Rousk, J., Baath, E., Brookes, P.C., Lauber, C.L., Lozupone, Dis 3: e364. C., Caporaso, J.G., et al. (2010) Soil bacterial and fungal Kemmitt, S.J., Wright, D., Goulding, K.W.T., and Jones, D.L. communities across a pH gradient in an arable soil. ISME (2006) pH regulation of carbon and nitrogen dynamics in J 4: 1340–1351. two agricultural soils. Soil Biol Biochem 38: 898–911. Salles, J.F., De Souza, F.A., and van Elsas, J.D. (2002) Kemp, J., Paterson, E., Gammack, S., Cresser, M., and Molecular method to assess the diversity of Burkholderia Killham, K. (1992) Leaching of genetically modified species in environmental samples. Appl Environ Microbiol Pseudomonas fluorescens through organic soils: influence 68: 1595–1603. of temperature, soil pH, and roots. Biol Fertil Soils 13: Salles, J.F., van Veen, J.A., and van Elsas, J.D. (2004) Mul- 218–224. tivariate analyses of Burkholderia species in soil: effect of Lauber, C.L., Strickland, M.S., Bradford, M.A., and Fierer, N. crop and land use history. Appl Environ Microbiol 70: (2008) The influence of soil properties on the structure of 4012–4020. bacterial and fungal communities across land-use types. Schloss, P.D., Westcott, S.L., Ryabin, T., Hall, J.R., Soil Biol Biochem 40: 2407–2415. Hartmann, M., Hollister, E.B., et al. (2009) Introducing Lazzaro, A., Widmer, F., Sperisen, C., and Frey, B. (2008) mothur: open-source, platform-independent, community- Identification of dominant bacterial phylotypes in a supported software for describing and comparing microbial cadmium-treated forest soil. FEMS Microbiol Ecol 63: 143– communities. Appl Environ Microbiol 75: 7537–7541. 155. Schmerk, C.L., Bernards, M.A., and Valvano, M.A. (2011) Lepleux, C., Turpault, M.P., Oger, P., Frey-Klett, P., and Hopanoid production is required for low-pH tolerance, anti- Uroz, S. (2012) Correlation of the abundance of microbial resistance, and motility in Burkholderia betaproteobacteria on mineral surfaces with mineral cenocepacia. J Bacteriol 193: 6712–6723. weathering in forest soils. Appl Environ Microbiol 78: 7114– Schönmann, S., Loy, A., Wimmersberger, C., Sobek, J., 7119. Aquino, C., Vandamme, P., et al. (2009) 16S rRNA gene- Lim, J.H., Baek, S.-H., and Lee, S.-T. (2008) Burkholderia based phylogenetic microarray for simultaneous identifica- sediminicola sp. nov., isolated from freshwater sediment. tion of members of the genus Burkholderia. Environ Int J Syst Evol Microbiol 58: 565–569. Microbiol 11: 779–800. Ludwig, W., Strunk, O., Westram, R., Richter, L., Meier, H., Shannon, C.E., and Weaver, W. (1963) The Mathematical et al. (2004) ARB: a software environment for sequence Theory of Communication. Urbana, IL: University of Illinois data. Nucleic Acids Res 32: 1363–1371. Press. Martínez-Aguilar, L., Díaz, R., Peña-Cabriales, J.J., Simpson, E.H. (1949) Measurement of diversity. Nature 163: Estrada-de los Santos, P., Dunn, M.F., and Caballero- 688–688. Mellado, J. (2008) Multichromosomal genome structure Stamatakis, A., Hoover, P., and Rougemont, J. (2008) A rapid and confirmation of diazotrophy in novel plant-associated bootstrap algorithm for the RAxML web servers. Syst Biol Burkholderia species. Appl Environ Microbiol 74: 4574– 57: 758–771. 4579. Štursová, M., Žifcˇáková, L., Leigh, M.B., Burgess, R., and Moulin, L., Munive, A., Dreyfus, B., and Boivin-Masson, C. Baldrian, P. (2012) Cellulose utilization in forest litter and (2001) Nodulation of legumes by members of the soil: identification of bacterial and fungal decomposers. β-subclass of proteobacteria. Nature 411: 948–950. FEMS Microbiol Ecol 80: 735–746. Nicol, G.W., Tscherko, D., Embley, T.M., and Prosser, J.I. Suárez-Moreno, Z., Caballero-Mellado, J., Coutinho, B., (2005) Primary succession of soil Crenarchaeota across Mendonça-Previato, L., James, E., and Venturi, V. (2012) a receding glacier foreland. Environ Microbiol 7: 337– Common features of environmental and potentially benefi- 347. cial plant-associated Burkholderia. Microb Ecol 63: 249– Opelt, K., Berg, C., Schonmann, S., Eberl, L., and Berg, G. 266. (2007a) High specificity but contrasting biodiversity of Tiedje, J.M., Asuming-Brempong, S., Nüsslein, K., Marsh, Sphagnum-associated bacterial and plant communities in T.L., and Flynn, S.J. (1999) Opening the black box of soil bog ecosystems independent of the geographical region. microbial diversity. Appl Soil Ecol 13: 109–122. ISME J 1: 502–516. Trân Van, V., Berge, O., Ngô Kê, S., Balandreau, J., and Opelt, K., Chobot, V., Hadacek, F., Schönmann, S., Eberl, L., Heulin, T. (2000) Repeated beneficial effects of rice inocu- and Berg, G. (2007b) Investigations of the structure and lation with a strain of Burkholderia vietnamiensis on early function of bacterial communities associated with Sphag- and late yield components in low fertility sulphate acid soils num mosses. Environ Microbiol 9: 2795–2809. of Vietnam. Plant Soil 218: 273–284.

Uroz, S., Calvaruso, C., Turpault, M.P., Pierrat, J.C., Mustin, Appendix S1. Detailed experimental procedure. C., and Frey-Klett, P. (2007) Effect of the mycorrhizosphere Fig. S1. Rarefaction curves for Burkholderia 16S rRNA gene on the genotypic and metabolic diversity of the bacterial libraries from 14 different samples. The OTUs were formed at communities involved in mineral weathering in a forest soil. a 98% identity threshold. Dotted vertical line represents the Appl Environ Microbiol 73: 3019–3027. sequence number threshold used for OTU analyses (22 Uroz, S., Buée, M., Murat, C., Frey-Klett, P., and Martin, F. sequences). (2010) Pyrosequencing reveals a contrasted bacterial Table S1. In vitro tests of acid tolerance. 68 Burkholderia diversity between oak rhizosphere and surrounding soil. strains were tested on AB minimal medium supplemented Environ Microbiol Rep 2: 281–288. with glucose and agar. Colony formation was selected as Uroz, S., Oger, P., Morin, E., and Frey-Klett, P. (2012) Distinct growth criterion. Presence or absence of colony formation ectomycorrhizospheres share similar bacterial communi- under various pH conditions is represented in the table with + ties as revealed by pyrosequencing-based analysis of 16S and – signs respectively. rRNA genes. Appl Environ Microbiol 78: 3020–3024. Table S2. Site description of selected sampling locations Weisskopf, L., Heller, S., and Eberl, L. (2011) Burkholderia from the trans-continental scale study. species are major inhabitants of white lupin cluster roots. Table S3. OTU table with closest relatives. Sequences from Appl Environ Microbiol 77: 7715–7720. each OTU were compared with the database on NCBI using the BLAST tool, and the best hits were selected as closest relatives. Supporting information

Additional Supporting Information may be found in the online version of this article at the publisher’s web-site: