Nitrogen assimilation in plant-associated bacteria

1
Nitrogen assimilation in plant-associated bacteria
Gail M. Preston Department of Plant
Sciences University of Oxford
2
Pseudomonas syringae
Pseudomonas fluorescens
Organic N High O2 Intimate association with plant
cells Low competition
Organic/inorganic N Med-low O2 Variable
association with diverse hosts High competition
S. Molin
M. Romantschuk
Endophyte / Leaf surface Plant Pathogen
Leaf surface / Roots Plant Growth-Promoting
3
Genome sequenced strains
P. aeruginosa PA01 P. aeruginosa PA14
P. entomophila L48 P. putida KT2440
P. syringae pv. tomato DC3000 P. syringae pv.
syringae B728a P. savastanoi pv.
phaseolicola 1448a
P. fluorescens Pf-5 P. fluorescens Pf0-1 P.
fluorescens SBW25
4
Why study nitrogen metabolism ?

• Nitrogen is essential for life frequently a
  limiting factor in natural environments
• Well characterised metabolic pathways (core
  metabolites and secondary metabolites)
• Environmental variability in nitrogen source and
  availability
• Environmental factors (pH, oxygen etc.) can
  affect nitrogen acquisition
• Environmental impact nitrogen fertilisers on
  natural ecosystems
• Variation in nitrogen metabolism across
  Pseudomonas

5
Leaves of specific plant species Leaf surface
and soil Soil Soil and animals
Why study Pseudomonas?
Ps1
P. syringae
Ps2
Ps3
Pf1
Pf2
P. fluorescens
Pf3
Pp1
P. putida
Pe1
P. entomophila
Pa1
P. aeruginosa
Niches vary in nutrient availability
environmental conditions pH, oxygen
host interactions (humans,
plants and simple animal models) Most strains can
grow on very minimal media salt, glucose, NH4
or nitrate
6
In silico predictions Using the Pfam database to
identify over and under-represented domains in P.
syringae
Amino acid transport
7
X P. syringae pv. tomato
8
Gene//Domain/Putative Function P. syringae pv. tomato P. syringae pv. tomato P. fluorescens SBW25 P. fluorescens SBW25
rpoN (sigma-54) PSPTO4453 Pflu0882
ntrB (NRII) PSPTO0353 Pflu0344
ntrC (NRI) PSPTO0352 Pflu0343
glnK (PII) amt-1 (ammonium transporter) PSPTO0217 PSPTO0218 Pflu5953 Pflu5952
gltB, gltD (glutamate synthase)(GOGAT) PSPTO5123/21 Pflu0414/5
glnA (glutamine synthase type I) PSPTO0359 Pflu0348
glnD (PII uridylyltransferase) PSPTO1532 Pflu1268
nac (nitrogen assimilation regulatory protein) PSPTO2923 Pflu4026
gdhA (glutamate dehydrogenase) No orthologous hit Pflu5326
nirB, nirD (nitrite reductase) PSPTO2302 - truncated nirB PSPTO3262/3 Pflu3425/4
Nitrate reductase Bifunctional nitrate reductase/sulfite reductase PSPTO2301 Pflu3426
Nitrate transporter PSPTO2304 Pflu4609
AA_permease domain proteins PSPTO5356, 1817, 2026 PSPTO5276 Pflu1674, 5187 Pflu5197, 1103, 0315, 2013, 5442 Pflu0368, 4870, 2264, 3375, 4890, Pflu4889, 3091, 3323, 3287, 3148, Pflu3094
Glutamine amidotransferase (class II) Glutamate synthase Ammonium transporter (amt-2) PSPTO2583 PSPTO2585 PSPTO2586 Pflu2324 Pflu2326 Pflu2327
Glutamine synthase (type II) PSPTO1921, 5309, 5310 Pflu1514, 2163, 3065, 5847, 5849
Ammonium transporter (amt-3) No orthologous hit Pflu1747
Glutamine synthase (type III) No orthologous hit Pflu2323
9
Predicting RpoN binding sites
10
Gene//Domain/Putative Function P. syringae pv. tomato P. syringae pv. tomato P. fluorescens SBW25 P. fluorescens SBW25
rpoN (sigma-54) PSPTO4453 ? Pflu0882 ?
ntrB (NRII) PSPTO0353 - Pflu0344 -
ntrC (NRI) PSPTO0352 - Pflu0343 -
glnK (PII) amt-1 (ammonium transporter) PSPTO0217 PSPTO0218 ? ? Pflu5953 Pflu5952 ? ?
gltB, gltD (glutamate synthase)(GOGAT) PSPTO5123/21 - Pflu0414/5 -
glnA (glutamine synthase type I) PSPTO0359 ? Pflu0348 ?
glnD (PII uridylyltransferase) PSPTO1532 - Pflu1268 -
nac (nitrogen assimilation regulatory protein) PSPTO2923 - Pflu4026 ?
gdhA (glutamate dehydrogenase) No orthologous hit Pflu5326 ?
nirB, nirD (nitrite reductase) PSPTO2302 - truncated nirB PSPTO3262/3 ? ? Pflu3425/4 ?
Nitrate reductase Bifunctional nitrate reductase/sulfite reductase PSPTO2301 ? Pflu3426 ?
Nitrate transporter PSPTO2304 ? Pflu4609 ?
AA_permease domain proteins PSPTO5356, 1817, 2026 PSPTO5276 ? - Pflu1674, 5187 Pflu5197, 1103, 0315, 2013, 5442 Pflu0368, 4870, 2264, 3375, 4890, Pflu4889, 3091, 3323, 3287, 3148, Pflu3094 ? ? - - -
Glutamine amidotransferase (class II) Glutamate synthase Ammonium transporter (amt-2) PSPTO2583 PSPTO2585 PSPTO2586 ? ?? Pflu2324 Pflu2326 Pflu2327 ?? ?
Glutamine synthase (type II) PSPTO1921, 5309, 5310 - Pflu1514, 2163, 3065, 5847, 5849 -
Ammonium transporter (amt-3) No orthologous hit Pflu1747 ?
Glutamine synthase (type III) No orthologous hit Pflu2323 ?
RpoN (s54) regulation of nitrogen metabolism
? intergenic s54 binding motif, ? intragenic
s54 binding motif, - no s54 binding motif
11
Phenoarrays
Nitrogen source utilisation by Pseudomonas
12
Pf56
Pa44
1
1
1
40
2
14
8
Overview of Pseudomonas utilisation of 96
nitrogen sources
Ps64
13
Amino acid utilisation by Pseudomonas
14
Nitrogen in natural habitats the leaf apoplast
Amino acid region of NMR spectra
glutamine
GABA
15

• Nitrogen metabolism
• Enzymes and metabolites well-defined
• 10 Pseudomonas genome sequences available
• Diverse ecological niches and selection
  pressures
• Diversity in nitrogen metabolism
• Experimentally tractable
• Evolving in response to

• Internal selection (network, flux, regulation)
• External selection (nutrient availability,
  environment (e.g. pH, oxygen), host interactions

16
Modelling the evolution of metabolic networks

• Which principle of evolutionary reconstruction
  should we apply?
• How do we represent metabolism?
• Which events can happen to a metabolism
• How can we generate models with biological
  relevance?

17
Which principle of evolutionary reconstruction
are we to apply?
Parsimony evolution has taken the shortest
possible path
Likelihood evolution has taken the most likely
path based on modelling of all possible
evolutionary events
In practice often give similar results Begin
with parsimony? easier to implement
18
Evolutionary Metabolic Network Models
Metabolites Nodes Reactions - Edges
Adjacency Matrix Each metabolite is a node (n1,
n2, n3, n4) For any two nodes I and j Aij 1
if there is an edge going from I to j
2 if there
is no edge between I and j

• Dynamical rules for evolution
• Take two nodes at random
• Perform a creation or deletion of edges with
  probability µ

19
Computational Challenges
Basic question Computing likelihoods What is
the probability of two observed homologous
metabolic networks
Principal answer Sum over all possible
evolutionary histories
Problem Computationally intensive!

• Potential strategies
• Develop recursive relations and dynamic
  programming algorithms
• Markov Chain Monte Carlo methods

20
Illustrated Metabolism

Network Model
Metabolism Network
21

• Adding biological relevance
• Define initial network according to biological
  model
• Define core metabolism label nodes that cannot
  be deleted or nodes that are omnipresent
  (environmental metabolite sources)
• Define constraints (e.g. preserve connectedness)
  label nodes with allowed changes
• Restrict changes to nodes with at least one
  allowed change
• Add directionality to connections
• Relate to biological data and evolutionary
  models
• Network structural features scale free? How
  many metabolites?

22
Ps1
P. syringae
Ps2
Ps3
Pf1
Pf2
P. fluorescens
Pf3
Pp1
P. putida
Pa1
P. aeruginosa
One metabolism accurate graph Two metabolisms
one metabolism changes into another Three
metabolisms define ancestral metabolism Four
metabolisms analysis is phylogeny dependent
23
Relating model evolution to organismal evolution

• Do nodes (metabolites) and edges (enzymes)
  evolve at the same rate ?
• Is it reasonable to assume a fixed rate of
  evolutionary change?
• Is it reasonable to assume that networks are
  scale free?
• Detect and exclude non-functional metabolisms to
  produce credible results. What criteria should we
  use to define non-functional metabolisms ?

24
Exploring the impact of natural selection on
metabolic networks

• Is it valid to assume a fixed pool of
  metabolites over evolutionary time and have just
  the reactions changing ?
• Can we explore the role of niche-specific
  conditions in network evolution by defining core
  available metabolites ?
• Can we develop theories about how and why
  selection has acted on networks by modulating
  selected variables (e.g. nitrogen source and
  availability)

25
Pathogenic Pseudomonas show clonal population
dynamics
Apoplast
Dissemination
Defined Niche
Infection
Modulation of plant/host physiology
Impact on other organisms in ecosystem
26
Rhizosphere
Dissemination
Heterogenous Niche
Infection
Modulation of plant/host physiology
Impact on other organisms in ecosystem
27
Relating network models to evolutionary models
Are parsimony and maximum likelihood equally
valid principles for studying network evolution
? Can we use network models as a basis for
phylogenetic trees ?
28
Archaea
Mycoplasma
Mycoplasma/Ureaplasma species
ONION YELLOWS PHYTOPLASMA
Borrelia burgdorferi
Treponema pallidum
Chlamydia
Chlamydia species
Wigglesworthia glossinidis
Buchnera species
Candidatus Blochmannia floridanus
Tropheryma whipplei
a
Bartonella species
?
Rickettsia species
Wolbachia pipientis
?
Coxiella burnetii
Haemophilus ducreyi
Pasteurella multocida
Haemophilus influenzae
?ß
Nitrosomonas aerogenes
Neisseria meningitidis
XYLELLA FASTIDIOSA
XYLELLA FASTIDIOSA Temecula1
Caulobacter crescentus
a
Brucella melitensis
Rhodopseudomonas palustris
BRADYRHIZOBIUM JAPONICUM
AGROBACTERIUM TUMEFACIENS
?
SINORHIZOBIUM MELILOTI
MESORHIZOBIUM LOTI
Acinetobacter species
ß
Bordetella species
XANTHOMONAS CAMPESTRIS
?ß
XANTHOMONAS AXONOPODIS
Chromobacterium violaceum
RALSTONIA SOLANACEARUM
PSEUDOMONAS SYRINGAE
?ß
Pseudomonas putida
Pseudomonas aeruginosa
Photorhabdus luminescens
ERWINIA CAROTOVORA
Yersinia pestis KIM
?
Salmonella species
Escherichia coli
Shigella flexneri
Shewanella oneidensis
Vibrio cholerae
Photobacterium profundum
Vibrio vulnificus
Vibrio parahaemolyticus
Deinococcus radiodurans
Gram ve
Firmicutes (Low GC Gram positives)
Actinomycetes (High GC Gram positives)
Thermotoga maritima
Thermotoga denticola
Fusobacterium nucleatum
Bacteroides thetaiotamicron (Low GC Gram
positives)
Porphrymonas gingivalis
Chlorobium tepidum
Desulfovibrio vulgaris
Geobacter sulfurreducens
Consensus tree of 100 jacknife trials based on
presence or absence of 7677 Pfam domain families
Epsilon Proteobacteria
Aquifex aeolicus
Cyanobacteria
Cyanobacteria
Rhodopirellula baltica
Leptospira interrogans
Bdellovibrio bacteriovorans
29
Oxford Jotun Hein Jon Churchill Andrea
Rocco David Studholme (Sainsbury Laboratory
Norwich)
30
(No Transcript)
31

• Adaptation of nitrogen assimilation networks may
  be influenced by
• Nitrogen source availability and type
• Ability to release nitrogen from complex
  macromolecules
• Ability to obtain nitrogen through host
  interactions
• Short and long term variation in nitrogen
  availability
• Other metabolic factors (e.g. respiration)
• Optimisation of energy consumption
• Consequences of nitrogen utilisation for
  bacteria-host interactions (mutually beneficial
  symbiosis, induction of host defences)
• Evasion of / adaptation to anti-microbial
  factors (e.g. anti-microbial peptides transported
  by N-transporters or inhibitors of N assimilation
  enzymes)

32
Are all events possible? Are all events equally
likely?
G
C
D
B
A
F
E

• Maintain functionality in long term (e.g. retain
  intermediate metabolism)

• Maintain core functionality (e.g. retain
  certain core metabolites and reactions)

33
The process

• Define universal/maximal metabolism all
  observed reactions and metabolites
• Extant and ancestral metabolisms represent
  subset of universal metabolism
• Metabolisms evolve by having reactions added or
  deleted
• Define properties of metabolites (nodes) and
  enzymes (edges)
• Estimate probabilities of metabolisms one
  evolutionary event away
• Analyse evolution of metabolisms