Diversity of Methylobacterium species associated with New Zealand native plants

Abstract Methylobacterium species are abundant colonizers of the phyllosphere due to the availability of methanol, a waste product of pectin metabolism during plant cell division. The phyllosphere is an extreme environment, with a landscape that is heterogeneous and continuously changing as the plant grows and is exposed to high levels of ultraviolet irradiation. Geographically, New Zealand (NZ) has been isolated for over a million years, has a biologically diverse flora, and is considered a biodiversity hotspot, with most native plants being endemic. We therefore hypothesize that the phyllosphere of NZ native plants harbor diverse groups of Methylobacterium species. Leaf imprinting using methanol-supplemented agar medium was used to isolate bacteria, and diversity was determined using ARDRA and 16S rRNA gene sequencing. Methylobacterium species were successfully isolated from the phyllosphere of 18 of the 20 native NZ plant species in this study, and six different species were identified: M. marchantiae, M. mesophilicum, M. adhaesivum, M. komagatae, M. extorquens, and M. phyllosphaerae. Other α, β, and γ-Proteobacteria, Actinomycetes, Bacteroidetes, and Firmicutes were also isolated, highlighting the presence of other potentially novel methanol utilizers within this ecosystem. This study identified that Methylobacterium are abundant members of the NZ phyllosphere, with species diversity and composition dependent on plant species.


Introduction
T he phyllosphere (abo ve ground parts of plants comprising mainly stems and leaves) is the Earth's biggest biological surface, estimated to be two times bigger than the land surface (Woodw ar d and Lomas 2004 ).The physicochemical environment of the phyllosphere is highly diverse due to fluctuating nutrient and water av ailability, temper atur e, wind pr essur e, exposur e to pollutants, UV radiation, and the variable biology of the waxy protective la yer (plant cuticle).T his diverse en vironment pro vides a habitat for many bacteria (Vorholt 2012 ), with the planetary phyllosphere bacterial population estimated to be as large as 10 26 cells (Lindow and Brandl 2003 ).Global processes of carbon, oxygen, and nitrogen cycling are greatly influenced by phyllosphere bacteria, as is plant health, with growth and productivity being enhanced (Trotsenko et al. 2001 , Lindow andBrandl 2003 ).The number and composition of phyllosphere bacteria are greatly influenced by plant species , sampling site , gr owing season, plant gr owth sta ge, location on the plant, leaf properties, and surrounding plant species (Kinkel et al. 2000, Redford et al. 2010, Finkel et al. 2011 ).
The genus Methylobacterium is a dominant group of Bacteria in the phyllosphere (Delmotte et al. 2009 ) and has been estimated at 10 4 -10 7 cells per gram of fresh plant material (Holland et al. 2002 ).The Methylobacterium association with plants can be epiphytic or endophytic (Corpe and Rheem 1989, Delmotte et al. 2009, Knief et al. 2010 ); or symbiotic (Sy et al. 2001 ); ho w e v er, ther e ar e no r eports of Methylobacterium as the cause of plant disease .T he association of Methylobacterium spp. with plants typically relies on methanol, a volatile organic compound (VOC), which is released by plants dur-ing gr owth thr ough stomatal por es in the epidermis (Galbally and Kirstine 2002 ).Ho w e v er, other plant-deriv ed carbon compounds may also support their colonization (Sy et al. 2005, Delmotte et al. 2009 ).Methanol is produced inside leaves as a byproduct of pectin metabolism during cell wall synthesis.During cell elongation and division pectin methylesterases catalyze the C-6 demethylation of homogalacturonan within plant cell walls, as a result, methanol is released (Korner et al. 2009 ).The global methanol emission from plants is estimated to be 100-128 Tg per year (Galbally and Kirstine 2002 ).Colonization of plants by methylotrophic bacteria, especially Methylobacterium species, is of interest because they play an important role in the atmospheric methanol c ycle b y utilizing methanol as their sole source of carbon and energy (Corpe and Rheem 1989 ).They also produce plant gr owth-pr omoting substances such as cytokinins , auxins , and vitamin B 12 (Tr otsenk o et al. 2001, Iv anov a et al. 2006 ) and are involved in seed germination, r oot de v elopment, and incr eased yield of a gricultur al plants (Meena et al. 2012 ).
Members of the genus Methylobacterium are pink-pigmented facultativ e methylotr ophs (PPFMs) (K ell y et al. 2014 ).They utilize one-carbon compounds, including methanol (CH 3 OH), methylamine (CH 3 NH 2 ), and formaldehyde (CH 2 O), and multi-carbon compounds containing no carbon-carbon bonds, as well as organic substrates with carbon-carbon bonds as the sole source of carbon and energy (Kelly et al. 2014 ).They are strict aerobes belonging to the α-Proteobacteria , order Rhizobiales , and are Gramnegativ e and r od-sha ped or ganisms.Methylobacterium ar e found worldwide on the leaves of many different plant species, and studies have shown that Methylobacterium species composition varies within and between plant species (Balachandar et al. 2008, Knief et al. 2008 ).Ho w e v er, geogr a phical location has been shown to have a stronger influence than plant species on both Methylobacterium and phyllosphere community composition (Knief et al. 2010, Wellner et al. 2011 ).
The objective of this study was to investigate the diversity of Methylobacterium species in the phyllosphere of native New Zealand (NZ) plants, via the isolation of methanol-utilizing bacteria.Amplified ribosomal DNA restriction analysis (ARDRA) was used to select r epr esentativ e isolates for sequencing.This study is significant as the first exploration of the species composition of the genus Methylobacterium in the phyllosphere of native NZ plants.

Sample collection and bacterial isolation
Twenty different native NZ plants were selected for isolation of Methylobacterium species (Table 1 ), plants were chosen to represent a diversity of plant types, including trees , shrubs , herbs , ferns , and flax.Leav es (thr ee) fr om fiv e plants of eac h plant species were collected in sterile containers, with the majority collected from the campus of the University of Waikato (Hamilton, NZ), wher e they gr ow natur all y.Leav es collected wer e healthy but not new growth, and were all collected in the morning.Leafimprinting was used to isolate bacteria from the leaf surface (Cor pe 1985 ). Immediatel y after collection, leav es wer e laid dir ectl y on the surface of 0.5% methanol-supplemented ammonium mineral salt (AMS) agar plates and impressed carefully, lar ge leav es wer e cut to the desired size and small leaves used whole, plates were then sealed with parafilm, and incubated at 30 • C for up to tw o w eeks (Holland and Polacco 1994 ).AMS agar medium contained: 0.7 g/l K 2 HPO 4 , 0.54 g/l NH 4 Cl, 0.1 mg/l ZnSO 4 .7H 2 O, 0.03 mg/l MnCl 2 .4H 2 O, 0.3 mg/l H 3 BO 3 , 0.2 mg/l CoCl 2 .6H 2 O, 0.01 mg/l CuCl 2 .2H 2 O, 0.02 mg/l NiCl 2 .6H 2 O, 0.06 mg/l Na 2 MoO 4 .2H 2 O, 15.0 g/l Difco TM agar (1.5%), in distilled water, pH was adjusted to 6.8 and the medium sterilized by autoclaving (Whittenbury et al. 1970 ).An amount of 0.5% methanol was ad ded ase pticall y to the autoclav ed medium upon cooling and mixed thor oughl y befor e the plates wer e pour ed.After incubation, colonies were chosen randomly from the plates and streaked.Colonies were re-streaked 5-6 times on fresh AMS agar plates to obtain a pure culture.

DN A extr action from pure cultures
For ARDRA analysis, DNA was extracted from bacterial colonies picked with a sterile toothpick into a 1.5-ml microfuge tube containing 1 ml of sterile H 2 O.The tube w as v ortexed vigor ousl y until cells were dispersed, boiled for 10 min, centrifuged for 5 min at 1000 rpm, then k e pt on ice .T he supernatant (5 μl) was used as a PCR template.
Prior to the extraction of DNA for sequencing (Marmur 1961 ), isolates wer e gr own in 50 ml AMS plus methanol broth at 30 • C, tr ansferr ed into 50 ml sterile falcon tubes, and centrifuged for 3 min at 3000 rpm.The supernatant was then removed and the cell pellet resuspended in 400 μl SET buffer (20% sucrose, 50 mM EDTA, 50 mM Tris.HCl).Lysozyme solution (50 mg/ml in TE plus 10 mM NaCl) was added (20 μl) and incubated at 37 • C for 1 hour.After incubation, 20 μl of 20% SDS and 10 μl of proteinase K solution (20 mg/ml in TE) were added to the lysates and incubated at 60 • C for 3 hours .T he digested l ysates wer e then purified by phe-nol c hlor oform extr action, concentr ated by ethanol pr ecipitation, and DN A w as stored at −20 • C. DNA concentration was quantified using a NanoDrop TM 1000 spectrophotometer and concentration was adjusted to 50-60 ng/ μl with TE buffer.
ARDRA is an established method for determining taxonomic relatedness between isolates (Fisher and Triplett 1999 ), which wer e scr eened to select r epr esentativ es for sequencing.The 16S rRNA gene PCR amplicon of each isolate was digested with Rsa I, and the products were analyzed by gel electrophoresis using 2% a gar ose in 1X T AE buffer .A 1 kb + DNA marker (Invitrogen) was run on e v ery gel to size the r estriction fr a gments.All isolates wer e compar ed visuall y for matc hing finger prints and gr ouped into differ ent r estriction types.Repr esentativ e isolates fr om eac h r estriction group were selected randomly for sequencing.

16S rRNA gene sequencing and analysis
PCR products of isolates selected for sequencing were purified using ExoSAP (Affymetrix, Ohio, USA), according to the manufactur er's pr otocol.Sequencing r eactions wer e performed at the Waikato DNA Sequencing Facility with primers 27F and 1492R on an ABI 3130XL (Applied Biosystems).Each 16S rRNA gene sequence was analyzed for closest identities using BLASTN in the NCBI database (Altschul et al. 1990 ).

Results
To investigate Methylobacterium species diversity on NZ native plants, a total of 20 plant species fr om differ ent plant types (tree, shrub , herb , fern, and flax) were selected.After 10-14 days incubation of leaf-imprinted agar plates at 30 • C, colonies were selected for streaking on fresh plates, and over an extended period, these were re-streaked 4-5 times, resulting in the isolation of 245 pure cultures of methanol-grown strains .T he majority of isolates (83%) were pale to vivid pink pigmented, with the remaining isolates either cream (10%) or dark orange to red (7%).Methylotr ophs wer e isolated fr om the leav es of e v ery plant species, but the number isolated varied between species (Table 2 ).Most isolates wer e fr om Micropiper excelsum (21.6%), follo w ed b y P .regius (20%), P .tenax (18.8%),P .cookianum (11.4%), and A .oblongifolium (7.3%), with the lo w est numbers from B .novea-zealandiae , Cyathea dealbata, C .medullaris , and K .excelsa (0.4%).

Discussion
Methylobacterium species diversity on the leaves of native NZ plants was investigated using leaf imprinting on methanolsupplemented AMS agar.Methanol is a k e y substrate for the gro wth of Meth ylobacterium, and in plants, methanol is produced during plant cell division as the waste product of pectin metabolism (Galbally and Kirstine 2002 ).A total of 245 methylotr ophs wer e isolated fr om the leav es of 20 differ ent nativ e NZ plants, but the bacterial species and number of isolates varied between plants.Interspecies variability is fr equentl y seen in other studies of the phyllosphere community (Yang et al. 2001, Lambais et al. 2006, Whipps et al. 2008, Redford et al. 2010 ).In this study, isolates were selected for sequencing using ARDRA, which identified sixteen different groups of isolates.Sequencing of repr esentativ e isolates fr om eac h r estriction gr oup r e v ealed that the majority of isolates (79.5%) were Methylobacterium species commonly found to colonize plants (Delmotte et al. 2009, Wellner et al. 2011, Knief et al. 2012a, Knief et al. 2012b ).Sixteen different phyllosphere Methylobacterium species have been identified (Balachandar et al. 2008, Knief et al. 2008 ), with M .extorquens being a ubiquitous colonizer of the phyllosphere of many plants (Del-  motte et al. 2009 ).This study identified six different Methylobacterium species isolated from the leaves of native NZ plants .T his low number of species may be because the isolation medium used was highly selective, with only methanol as a carbon source, and Methylobacterium species ar e commonl y differ entiated according to their carbon utilization ability (K ell y et al. 2014 ).Another reason could be the degree of association between NZ plants and Methylobacterium sp., with other studies showing that Methylobacterium association can be epiphytic (Omer et al. 2004), endophytic (Lacava et al. 2004), or symbiotic (Jourand et al. 2004 ).The Methylobacterium species identified in this study ( M .marchantiae , M .adhaesivum , M .komogatae , M .mesophilicum , M .extorquens , and M .phyllosphaerae ) are common leaf epiphytes (Knief et al. 2008, Verginer et al. 2010, Schauer et al. 2011, Tani et al. 2012 ), not sur prisingl y given that leaf imprinting was used to isolate the bacteria.The final reason could be geogr a phic location, whic h has been shown to be an important determinant in shaping Methylobacterium colonization of the phyllosphere (Knief et al. 2010 ); ho w e v er, this could not be addressed in this study.
Another factor to be considered was the effect of leaf texture and structure on the ease of the isolation of bacteria from different plants .T he lea ves of plants that yielded the most isolates ( P .regius, M .excelsum , P .tenax , P .cookianum , and A .oblongifolium ) are thin and easily lie on the medium (see Fig. 1 ), making imprinting easier than for other plants in this study that ha ve lea ves that are thick, glossy, hard, shiny, spiky, or v elv ety ( A .australis , A .excelsus , C .robusta , G .littoralis , H . elliptica , K .excelsa , M .excelsa , O .albida , O .traversii , and P .tenuifolium ).Leaf texture and structure have been shown to be important factors in obtaining epiphytes from the leaf surface using imprinting (Holland et al. 2000 ).The low number of Methylobacterium isolates from B .novea -zealandiae and C .cunninghamii ma y ha ve been due to the fact that only small, young plants were sampled.But generally, in this study, few Methylobacterium species were isolated from ferns, and none were isolated fr om C .dealbata (blac k fern) or C .medullaris (silv er fern), whic h may indicate that either leaf structure, leaf age (fern leaves have a shorter life span than other NZ native plants), or leaf chemistry which may restrict colonization by methylobacteria.Howe v er, plant species hav e been shown to be the main driver for the comm unity structur e of Methylobacterium species in se v er al studies (Kinkel et al. 2000, Omer et al. 2004, Knief et al. 2008, Knief et al. 2010, Redford et al. 2010, Wellner et al. 2011 ) and may ther efor e be significant for NZ native plant species.
The most fr equentl y isolated Methylobacterium colonizer was M. mesophilicum, which was isolated from 17 plant species and was the only Methylobacterium sp. to be isolated fr om se v en plant species ( A .australis , B .novea-zealandiae , C .robusta , G .littoralis , K .excelsa , M .excelsa , and O .traversii ), possibly indicating some specialization.Similarl y, onl y M. marc hantiae was isolated from O. albida .From this study, it is difficult to explain the reasons for the association of individual Methylobacterium sp. with specific plants.Ho w e v er, plant species and the generalist nature of some Methylobacterium species have been found to play a combined role in colonization (Dourado et al. 2012 ).
Other α-Proteobacteria ( Hyphomicrobium , Methylopila , and Rhizobium ), β-Proteobacteria ( Alcaligenes , Methylophilus , and Ramlibacter ), γ -Proteobacteria ( Xanthomonas ), Actinomycetes ( Janibacter ), Bacteroidetes ( Niastella ), and Firmicutes ( Paenibacillus ) were also isolated, with similar genera seen in culture-independent studies of the phyllospher e (Jac kson et al. 2006, Lambais et al. 2006, Redford et al. 2010, Rastogi et al. 2012 ).Se v er al studies hav e demonstr ated the association of some of these Bacteria with plants, including Hyphomicrobium sp. from the Arabidopsis phyllospher e (Reisber g et al. 2013 ), Methylopila sp. from banana fruit (Doronina et al. 2013 ), R .endophyticum from the green bean (Lopez-Lopez et al. 2010 ), and A .f aecalis fr om gr een ash leav es (Sandhu et al. 2009 ).Methylophilus has been found in the phyllosphere (Wellner et al. 2011 ), Xanthomonas sp. have been identified on plant leaves and stems (Corpe andRheem 1989 , Sheng et al. 2011 ), Janibacter have been detected in the rhizosphere (Guiñazú et al. 2013 ), N .populi from the soil of a Euphr ates Poplar for est (Zhang et al. 2010 ), and P .lautus from the rhizosphere of wild grass (Sharma et al. 2010 ).While a number of these gener a ar e known methanol utilizers ( Meth ylophila , Meth ylophilus , and Hypomicrobium ), to date there is no clear evidence for methanol utilization by most of the other genera of Bacteria ( Alcaligenes , Janibacter , Niastella , Ramlibacter , Rhizobium , Paenibacillus , and Xanthomonas ), although they did r epr esent onl y a small number of all the isolates in this study.Further studies are therefore required to understand more about the utilization of carbon substrates by these isolates.

Table 1 .
NZ native plant species used in this study.

Table 2 .
Isolates from each plant species grouped by ARDRA analysis.

Table 3 .
BLAST analysis of 16S rRNA gene sequences of isolates from each OTU.