Article

Selection Maintains Appar

ently DegenerateMetabolic Pathways due to Tradeoffs in UsingMethylamine for Carbon versus Nitrogen

Highlights

d Two degenerate methylamine oxidation modules have

distinct roles in methylotrophs

d Methylamine dehydrogenase enables rapid use of

methylamine as a growth substrate

d The N-methylglutamate pathway enables nitrogen

assimilation from methylamine

d Tradeoffs between ammonium toxicity and cellular

localization select for degeneracy

Nayak et al., 2016, Current Biology 26, 1416–1426June 6, 2016 ª 2016 Elsevier Ltd.http://dx.doi.org/10.1016/j.cub.2016.04.029

Authors

Dipti D. Nayak, Deepa Agashe,

Ming-Chun Lee, Christopher J. Marx

Correspondencecmarx@uidaho.edu

In Brief

Metabolic degeneracy is commonly

observed in microbial genomes;

however, an adaptive basis for this

phenomenon is rarely understood. Nayak

et al. uncover physiological tradeoffs

underlying the utilization of methylamine

as a growth substrate or as nitrogen

source that selects for two methylamine

oxidation pathways in methylotrophs.

Current Biology

Article

Selection Maintains Apparently DegenerateMetabolic Pathways due to Tradeoffsin Using Methylamine for Carbon versus NitrogenDipti D. Nayak,1,5 Deepa Agashe,1,6 Ming-Chun Lee,1,7 and Christopher J. Marx1,2,3,4,*1Organismic and Evolutionary Biology, Harvard University, Cambridge, MA 02138, USA2Department of Biological Sciences, University of Idaho, Moscow, ID 83844, USA3Institute for Bioinformatics and Evolutionary Studies, University of Idaho, Moscow, ID 83844, USA4Center for Modeling Complex Interactions, University of Idaho, Moscow, ID 83844, USA5Present address: The Carl R. Woese Institute for Genomic Biology, University of Illinois, Urbana, IL 61801, USA6Present address: National Centre for Biological Sciences, Bangalore 560065, India7Present address: Department of Biochemistry, University of Hong Kong, Pokfulam, Hong Kong*Correspondence: cmarx@uidaho.edu

http://dx.doi.org/10.1016/j.cub.2016.04.029

SUMMARY

Microorganisms often encode multiple non-ortholo-gous metabolic modules that catalyze the same re-action. However, little experimental evidence actu-ally demonstrates a selective basis for metabolicdegeneracy. Many methylotrophs—microorganismsthat grow on reduced single-carbon compounds—like Methylobacterium extorquens AM1 encode tworoutes for methylamine oxidation: the periplasmicmethylamine dehydrogenase (MaDH) and the cyto-plasmic N-methylglutamate (NMG) pathway. InMethylobacterium extorquens AM1, MaDH is essen-tial for methylamine growth, but the NMG pathwayhas no known physiological role. Here, we use exper-imental evolution of two isolates lacking (or inca-pable of using) MaDH to uncover the physiologicalchallenges that need to be overcome in order touse the NMG pathway for growth on methylamineas a carbon and energy source. Physiological char-acterization of the evolved isolates revealed regula-tory rewiring to increase expression of the NMGpathway and novel mechanisms to mitigate cyto-plasmic ammonia buildup. These adaptations ledus to infer and validate environmental conditionsunder which the NMG pathway is advantageouscompared to MaDH. The highly expressedMaDH en-ables rapid growth on high concentrations of methyl-amine as the primary carbon and energy substrate,whereas the energetically expensive NMG pathwayplays a pivotal role during growth with methylamineas the sole nitrogen source, which we demonstrateis especially true under limiting concentrations(<1 mM). Tradeoffs between cellular localizationand ammonium toxicity lead to selection for thisapparent degeneracy as it is beneficial to facultativemethylotrophs that have to switch between using

1416 Current Biology 26, 1416–1426, June 6, 2016 ª 2016 Elsevier L

methylamine as a carbon and energy source or justa nitrogen source.

INTRODUCTION

Multiple modules that apparently perform the same biological

function are often encoded in a genome [1–4]. In some instances,

these modules are closely related paralogs (genetic redun-

dancy), but in others, they are structurally and evolutionarily

distinct entities (functional degeneracy) [5]. Metabolic degener-

acy—non-orthologous enzymes and pathways that catalyze

the same metabolic transformation through distinct biochemical

reactions—is especially prevalent in bacterial genomes [1–4, 6].

Despite the cost, degeneracy (or even redundancy) at a cellular

level is proposed to increase robustness to environmental pertur-

bations [7], mutations [8], and stochasticity in cellular processes

[9]. Consider one of the classic examples of metabolic degener-

acy, the two alternate routes for assimilation of ammonium ions

(NH4+) into glutamate in Escherichia coli. Biochemical tradeoffs

between the energetically efficient glutamate dehydrogenase

(GDH)-mediated route and the higher-affinity route using gluta-

mine synthetase and glutamate synthase (GS-GOGAT) have

been proposed to allow optimal nitrogen assimilation across a

wide range of growth conditions and NH4+ concentrations [10].

Although there is some evidence that GDH is specifically benefi-

cial under free energy (or carbon) limitation [11], neither a growth

defect nor an overt NH4+ assimilation phenotype has been

consistently observed for mutants lacking GDH [10]. In fact,

the recurring proposition that GDH is a completely dispensable

enzyme really questions whether we understand the putative

selective basis for this particularly well-studied example of

metabolic degeneracy [10, 12]. Some other studies [2–4] have

invoked compensating tradeoffs between degenerate pathways

to explain stable coexistence, but there is surprisingly little direct

experimental evidence that explains the prevalence and mainte-

nance ofmetabolic degeneracy in bacterial genomes over evolu-

tionary timescales.

Methylotrophs are a polyphyletic group of microorganisms

that grow on reduced single-carbon (C1) compounds as the

td.

A B

periplasm

cytoplasm

CH3NH3+

CH3NH3+ HCHO

NH4+, 2e-

MaDH

HCHOFAESpontaneous

HCOO-

CO2

Biomass H4MPT oxidation pathway

H4Freduction pathway

periplasm

cytoplasm

CH3NH3+

CH3NH3+ HCHO

HCOO-

ATP ADP

NH4+ NADH + H+

CO2

H4MPT oxidation pathway

Biomass

NMG pathway

Figure 1. MethylamineMetabolism inMeth-

ylobacterium extorquens Strains

(A) A schematic representation of carbon flow

during methylamine growth mediated by MaDH

(orange) in M. extorquens AM1. The founding

ancestor A1 lacks the entire mau gene cluster

encoding MaDH. The founding ancestor A2 lacks

FAE (formaldehyde-activating enzyme), which

catalyzes the condensation of formaldehyde with

the C1 carrier molecule tetrahydromethanopterin

(H4MPT) and is only essential for methylamine

growth mediated by MaDH.

(B) Methylamine growth mediated by the NMG

pathway (blue) in other M. extorquens strains. The

spontaneous reaction between formaldehyde and

H4MPT, which occurs at a slower rate, is sufficient

for methylamine growth mediated by the NMG

pathway.

See also Figure S1. MaDH, methylamine dehy-

drogenase; NMG, N-methylglutamate.

sole carbon and energy source [13, 14]. Methylamine (CH3NH2),

a degradation product of proteins, nitrogenous pesticides, and

osmolytes like glycine betaine (N,N,N-trimethylglycine), can

serve as either a growth substrate or a nitrogen source for meth-

ylotrophs [15]. Here, we uncover selective pressures that lead to

the maintenance of two apparently degenerate modules for

methylamine oxidation in sequenced methylotrophs.

Two biochemically distinct pathways for methylamine

oxidation have been characterized in methylotrophic bacteria

[14]. Methylamine dehydrogenase (MaDH), along with ancil-

lary proteins required for proper assembly and localization, is

encoded by the mau gene cluster [16, 17]. MaDH catalyzes

the oxidation of methylamine to formaldehyde in a single step

in the periplasm, transferring a pair of electrons to cytochrome

c via an amicyanin electron acceptor (Figure 1A) [16, 18]. Unlike

MaDH, the cytoplasmic N-methylglutamate (NMG) pathway

mediates the primary oxidation of methylamine in three enzy-

matic steps (Figures 1B and S1), produces an NADH, and

also uses an ATP per mol of methylamine oxidized [19]. Further-

more, the flow of oxidized C1 units into assimilatory pathways

also differs during methylamine growth mediated by MaDH

versus the NMG pathway. All formaldehyde produced by

MaDH flows through the tetrahydromethanopterin (H4MPT)-

dependent formaldehyde oxidation pathway to formate [13,

20], which is partially oxidized to CO2 via a panel of four formate

dehydrogenases [21, 22] and partially assimilated into biomass

via the successive action of the tetrahydrofolate (H4F)-depen-

dent pathway and the serine cycle [14, 23]. In contrast, growth

mediated by the NMG pathway has no requirement for the H4F

pathway; C1 units are either dissimilated by the H4MPT-depen-

dent formaldehyde oxidation pathway or enter the serine cycle

directly [15]. Curiously, many methylotrophs encode MaDH as

well as the NMG pathway [24]. In Methylobacterium extorquens

AM1 (referred to as AM1 hereafter), an extensively studied

facultative methylotroph belonging to the Alphaproteobacteria

[13], the NMG pathway was observed to be incapable of

rescuing the methylamine growth defect of mutants lacking

MaDH [17]. A paradox thus arises: why does AM1 encode the

NMG pathway if it does not appear to contribute to growth on

methylamine?

In this study, we use experimental evolution to uncover the

physiological challenges that need to be overcome in order to

use the NMG pathway for growth on methylamine as a carbon

and energy source and in the process discover the environ-

mental conditions under which the NMG pathway is advanta-

geous compared to MaDH. We began with two distinct strains

of AM1 that were incapable of growth on methylamine despite

an intact NMG pathway. After an extended period of time in

liquid medium with methylamine as the sole carbon and energy

source, one replicate for each genotype recovered slow growth.

Remarkably similar physiological hurdles were overcome in both

cases: an increase in the expression of the NMG pathway and

unique strategies to mitigate cytoplasmic ammonium buildup.

Inspired by these findings, we demonstrate that the presence

of the NMG pathway is a substantial advantage to cells in envi-

ronments where %1 mM methylamine is used as the sole nitro-

gen source. Thus, whereas MaDH is a highly expressed, poorly

regulated, and energetically favorable enzyme that enables rapid

growth of methylamine as a carbon source, the energetically

expensive NMG pathway enhances growth in environments

with limiting methylamine as the sole nitrogen source.

RESULTS

Fortuitous Evolution of NMG-Dependent Growth onMethylamineTwo strains of AM1, obtained from unrelated, independent

studies [25, 26], were previously observed to be incapable of

methylamine growth despite an intact NMG pathway. We

wanted to explore whether growth could be re-established by

experimental evolution on methylamine as the sole carbon and

energy source. The underlying rationale behind this experiment

was to use the beneficial mutations that enable methylamine

growth using the NMG pathway alone to help elucidate its phys-

iological role when present in a genome together with MaDH.

The first ancestral strain (referred to hereafter as ‘‘A1’’) was

isolated from the final time point from one of eight populations

of AM1 that were evolved by serial transfer for 1,500 generations

on a combination of methanol and succinate [25, 27]. Upon

sequencing the genome, besides other mutations not involved

Current Biology 26, 1416–1426, June 6, 2016 1417

in this study, we observed that A1 had lost genes encoding

MaDH due to homologous recombination between insertion se-

quences (ISs) flanking the 10-kb mau gene cluster (Table S1).

The second strain (referred to hereafter as ‘‘A2’’) contains an

in-frame, markerless deletion of fae [26] and was originally

used as a negative control for an evolution experiment involving

selection upon codons in synonymous variants of fae [28]. Form-

aldehyde-activating enzyme (FAE) catalyzes rapid condensa-

tion of formaldehyde and H4MPT, although this reaction is also

known to occur spontaneously in vivo at a slower rate [20, 29].

FAE is essential for rapid methylamine growth mediated by

MaDH [20]. Thus, even though A2 encodes a functionally active

MaDH, Dfae prevents methylamine growth due to the buildup of

toxic formaldehyde in the cell (Figure 1A) [20, 29]. Furthermore,

methylamine growth in other M. extorquens strains that only

contain the NMG pathway is at least 3-fold slower and can be

supported by the spontaneous reaction between formaldehyde

and H4MPT, which makes FAE dispensable (Figure 1B) [15].

A1 and A2 were each inoculated in minimal medium with

20 mM methylamine as the sole carbon and energy source for

2–4 weeks until a significant rise in optical density (DOD600 z0.05) was observed. After this initial rise in optical density, the

population initiated with A1 was plated to obtain single colonies

of evolved isolates, one of which we studied here (‘‘E1’’). On the

other hand, the population initiated with A2 was further propa-

gated by serial transfer for a further 150 generations and then

plated to obtain single colonies. An isolate from the final time

point of this evolution experiment was also studied further

(‘‘E2’’). E1 and E2 exhibited methylamine growth rates of

0.04 h�1 and 0.09 h�1 (average of three replicate cultures; Fig-

ure 2A), respectively. Despite this improvement in growth, the

growth rate of E1 and E2 was 79% (p < 0.0001; unpaired two-

sided Student’s t test comparing the mean growth rate of three

biological replicates each) and 55% (p < 0.0001) lower than

that observed for wild-type (WT). The yield (measured as

maximum OD600) of the two evolved isolates was 80% (p <

0.0001) and 15% (p = 0.0008) lower too (Figures 2A and 2B).

Mutations Re-establishing Growth Using the NMGPathwayThe genomes of E1 and E2 were sequenced to identify the

mutations that restored methylamine growth (Table S1). First,

we noted an IS-mediated recombination event in E2 that deleted

the mau gene cluster. Remarkably, this was exactly the same

event that had occurred during the evolution of A1 (Figure S2A).

Second, one mutation in each evolved isolate involved the NMG

pathway. E1 has a nonsense mutation (W173*) in Meta1_1544,

immediately upstream of genes encoding enzymes of the NMG

pathway. E2 has a tandem duplication of a 12-kb region of the

genome containing genes of the NMG pathway (Figure S2A).

In addition, E1 contains an IS insertion in the ykkC/yxkD RNA

element [30] located between two ABC transporters (Figure S2B)

and E2 contains a nonsense mutation (Q446*) in kefB, encoding

a glutathione-dependent K+/H+ antiporter [31] (Figure S2B).

Genetic Confirmation of NMG-Dependent MethylamineGrowthKey genes in dissimilatory pathways were deleted to establish

whether mutations in E1 and E2 rerouted primary oxidation

1418 Current Biology 26, 1416–1426, June 6, 2016

through the NMG pathway in the same manner as has been

established in other methylotrophs that encode only the

NMG pathway [15, 19]. First, we deleted NMG dehydrogenase

[15, 19] (encoded by mgdABCD), an enzyme that catalyzes the

last step of the NMG pathway (Figure S1), in WT, E1, and E2.

The DmgdABCDmutant of WT did not have a significant change

in growth rate (p = 0.11) but did have a 5.4% decrease in yield

(p = 0.05) on methylamine (Figures 2A and 2B). Neither the

DmgdABCD mutant of E1 nor a mutant strain of E2 lacking

both copies of the mgdABCD operon could grow on methyl-

amine. Next, in order to determine whether formaldehyde oxida-

tion by the H4MPT pathway was required for methylamine

growth [20, 32], the first dedicated gene for H4MPT biosynthesis

(mptG) [32] was deleted in E1 and E2. No methylamine growth

was observed in the DmptG mutant of either strain. These data

corroborate that methylamine growth mediated by the NMG

pathway involves the H4MPT-dependent formaldehyde oxida-

tion pathway [15].

Selective Coefficients of Evolved AllelesBy performing competition assays with strains containing one

or more evolved alleles against a fluorescently labeled WT

strain, we determined that both mutations in E1 were benefi-

cial in isolation (Figures 2C and 2D). Introducing the nonsense

mutation in Meta1_1544 (now termed nmgR; see below)

increased the fitness of A1 to 2.9% that of WT (p = 0.04),

whereas introducing the IS insertion in the ykkC/yxkD RNA

element increased the fitness of A1 to 13.3% that of the WT

(p = 0.0002). The fitness of E1 was indistinguishable from the

product of the fitness effect of each single mutant (W[E1] =

W[nmgRW173*] 3 W[IS insertion into ykkC/yxkD]; p = 0.6; Fig-

ures 2C and 2D).

In contrast to E1, adaptive mutations need to be obtained in a

particular order in E2 (Figures 2C and 2E). Re-introducing the

mau gene cluster on a plasmid (pAYC139) [17] in E2 resulted in

negative fitness indicating net death, likely due to formaldehyde

buildup (Figure 2E), suggesting that the Dmau mutation arose

first. In the Dmau background, a duplication of the genomic

region containing the NMG pathway increased methylamine

fitness by 9-fold (p < 0.0001), and when the Q446* nonsense

mutation in kefB was introduced, subsequently, the fitness

rose by another 16.3% (p = 0.003; Figure 2E). To test whether

this truncated allele behaves like a null mutation, the kefB locus

was deleted in E2. The methylamine fitness of the DkefB mutant

of E2 was significantly lower than E2 (8.5%; p = 0.002) yet was

significantly higher than a mutant strain of E2 containing the

ancestral kefBWT (7.1%; p = 0.04), indicating that kefBQ446* is

not a complete loss-of-function allele (Figure S3). Collectively,

these results confirm that the identified mutations are beneficial

and account for the improvement observed in the evolved

isolates.

NMGPathway Is Repressed duringMethylamine Growthin WTIn both E1 and E2, mutations changing the dosage of the NMG

pathwayorneighboringgeneswereacritical step toward restoring

methylamine growth. To quantify the change in expression of the

NMG pathway, the activity of NMG dehydrogenase (NMGDH)

was measured in crude cell extracts from methylamine-induced

A

0

0.05

0.1

0.15

0.2

0.25

WT A1 A2 E1 E2

Gro

wth

Rat

e (h

-1)

N.D.

N.D.

mgdABCD mutant

B

0

0.05

0.1

0.15

0.2

0.25

0.3

WT A1 A2 E1 E2

Yie

ld (

Max

imu

m O

D60

0)

Series2

N.D. N.D.

mgd ABCDmutant

C

0

0.2

0.4

0.6

0.8

1

1.2

WT A1 A2 E1 E2

Fit

nes

s

D

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

Fit

nes

s

A1

2

121 E1

nmgRW173*

IS insertion in ykkC/yxkD

E

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

Fit

nes

s

A2 1 1

21

2

3

3

E2

mau

kefB Q446*

12kb duplication

E2+

MaDH

Figure 2. Growth and Fitness of WT, Ancestral, and Evolved M. extorquens AM1 Isolates on Methylamine as the Sole Carbon and EnergySource

(A and B) Growth rate (h�1; A) and yield or maximum OD600 (B) on 20 mM methylamine for WT, the two ancestral genotypes (A1 and A2), and the two evolved

isolates (E1 and E2) with (gray) and without (blue) the NMG pathway.

(C) Competitive fitness of WT, the two ancestral strains (A1 and A2), and the two evolved isolates (E1 and E2) on 20 mM methylamine relative to a strain of

M. extorquens AM1 expressing the fluorescent protein mCherry.

(D) Competitive fitness of strains with different combinations of mutations from E1 on 20 mM methylamine.

(E) Competitive fitness of strains with different combinations of mutations from E2 on 20 mMmethylamine. The error bars represent the 95% confidence interval

(CI) of the mean of three independent fitness assays.

See also Figures S2 and S3. N.D., not detected.

cultures. The activity of NMGDH in A1 and A2 was indistinguish-

able from WT (p = 0.09 and p = 0.50; Figure 3A). The activity of

NMGDH in E1 and E2, however, was 5.2-fold (p < 0.0001) and

9.3-fold higher (p < 0.0001) than their respective ancestors (Fig-

ure 3A). Genomic proximity, along with phenotypic evidence that

a nonsense mutation in Mext_1544 increased NMGDH activity in

E1, indicated that the gene product plays a regulatory role influ-

encing expression of one or more enzymes of the NMG pathway.

Hence, Mext_1544 has been renamed NMG pathway regulator

(nmgR).

Cytoplasmic pH Homeostasis in E1 Achieved via UreaExcretionConsidering the stoichiometry of biomass, it became apparent

that, when used as a growth substrate, the 1:1 ratio of carbon:

nitrogen in methylamine greatly exceeds the cellular demand

for nitrogen (�5:1 carbon:nitrogen). This is further confounded

because 50% of the carbon from methylamine is respired as

CO2 [2, 23, 33]. Thus, even in the absence of extracellular

NH4+, only 10% of the NH4

+ generated during methylamine

growth will be assimilated and, to prevent the toxic effects of

Current Biology 26, 1416–1426, June 6, 2016 1419

0

10

20

30

40

50

60

70

80

90

WT A1 A2 E1 E2

NM

GD

HA

ctiv

ity

(nm

ol H

CH

O/m

g pr

otei

n/m

in)

B

0

20

40

60

80

100

120

140

A1 E1U

rea

in s

up

ern

atan

t (

M)

Succinate Succinate + MA

N. D. N. D. 0

0.05

0.1

0.15

0.2

0.25

0.3

5.5 6 6.5 7 7.5

Gro

wth

Rat

e (h

-1)

Extracellular pH

E1

E2

kefBAnc

E1nmgR

E2 kefB WT

A1 nmgRW173*

C

periplasm

cytoplasm

CH3NH3+

CH3NH3+ HCHO

HCOO-

NH4+

CO2

Biomass

Urea

Urea amidolyase

Urea efflux pump

A

D E

periplasm

cytoplasm

CH3NH3+

CH3NH3+ HCHO

HCOO-

NH4+

CO2

Biomass

H+

K+

H+

K+

H+

K+

KefBQ446*

FAE spontaneous

HCHO

NMG pathway NMG pathway

XΔMaDH

Figure 3. Physiological Adaptations in Evolved M. extorquens AM1 Isolates that Use the NMG Pathway for Growth on Methylamine as the

Sole Carbon and Energy Source

(A) Mean activity of the NMG dehydrogenase enzyme (nmol formaldehyde produced per mg protein per minute) in crude extracts from methylamine-induced

cultures of WT, the two ancestral strains (A1 and A2), and the two evolved isolates (E1 and E2).

(B) Mean concentration of urea in the supernatant for A1 and E1 during succinate growth (open) and after methylamine induction (filled).

(C) Growth rate of E1 (gray), E2 (black), A1 containing the evolved allele of nmgR (dashed gray), and E2 containing the ancestral allele of kefB (dashed black) in

minimal media with 3.5 mM succinate, buffered to pH values ranging from 5.5–7.5. The error bars represent the 95%CI of the mean of three biological replicates.

(D) Mutations in a putative repressor (nmgR) leading to overexpression of the NMG pathway and mitigation of NH4+ accumulation by urea excretion leads to

methylamine growth mediated by the NMG pathway in E1.

(E) Deletion of MaDH, a tandem duplication of the genes encoding the NMG pathway, and balancing cytoplasmic pH through a constitutive kefB mutant

(kefBQ446*) leads to methylamine growth mediated by the NMG pathway in E2.

See also Figure S4. MA, methylamine.

NH4+ buildup, the remainder must be released. Whereas the

periplasmic location of MaDHmay be an advantage for releasing

this excess NH4+, the cytoplasmic localization of the NMG

pathway might prove to be a growth-limiting constraint. There-

fore, we hypothesized that mitigating cytoplasmic NH4+ accu-

mulation, especially due to overexpression of the NMG pathway,

is likely to have been a target of other beneficial mutations that

restored methylamine growth in E1 and E2.

Similar to results in other organisms, prior studies in

M. extorquens have shown that IS insertions commonly change

1420 Current Biology 26, 1416–1426, June 6, 2016

expression of flanking genes [34, 35].We used qRT-PCR tomea-

sure the mRNA level for each of the ABC transporters flanking

the ykkC/yxkD RNA element in methylamine-induced cultures

of E1 and A1 (Figure S4A). The expression of the nitrate/sulfo-

nate/bicarbonate ABC efflux transporter in E1 was 58% higher

than in A1 (Meta1_4100 to Meta1_4102; p < 0.05), but there

was no significant change in expression of the glycine/betaine/

proline transporter (Meta1_4103 to Meta1_4105; 2% increase;

p = 0.86). The genomic proximity of the nitrate/sulfonate/

bicarbonate ABC efflux transporter to urea metabolism genes

A

0

0.05

0.1

0.15

0.2

0.25

WT

Gro

wth

Rat

e (h

-1)

WTnmgRW173

WTkefB Q446*

B

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

WTY

ield

(M

axim

um

OD

600)

WT

nmgRW173WT

kefB Q446*

Figure 4. Growth and Fitness of WT Strains

with Mutations that Enable Growth on NMG

Pathway

(A) Growth rate (h�1) and (B) yield or maximum

OD600 of WT, WT with the evolved allele of nmgR

from E1, and WT with the evolved allele of kefB

from E2. The error bars represent the 95% CI of

the mean of three biological replicates. See also

Figure S5.

(Figure S4A) further led to the hypothesis that it might serve as an

efflux pump for urea. No significant amount of urea (<20 mM) was

detected in the supernatant of A1 cultures during succinate

growth or after methylamine induction (Figure 3B). For E1, how-

ever, �100 mM urea was detected in the supernatant indepen-

dent of the growth conditions (OD600 of cultures varied by

<25%), suggesting that an IS insertion in the ykkC/yxkD RNA

element renders the neighboring nitrate/sulfonate/bicarbonate

ABC efflux transporter constitutively active (Figures 3B and 3D).

Truncated KefB Facilitates Growth of E2 in AlkalineMediaDue to the absence of any ABC transporter in the genomic prox-

imity of kefB, an alternate mechanism for suppressing the dele-

terious effects of NH4+ buildup is likely to have been adopted in

E2. Based on the location of the amber codon, the truncated

kefBQ446* allele encodes the H+/K+ exchanger domain without

the glutathione-dependent regulatory domain: when bound to

reduced glutathione, this domain represses the KefB H+/K+ anti-

porter activity to prevent acidification of the cytoplasm at the

expense of K+ ions (Figure S4B) [36]. KefBQ446* could thus func-

tion constitutively, and the constant proton inflow might buffer

the rise in cytoplasmic pH due to NH4+ accumulation. Extrapo-

lating this effect further, KefBQ446* might enable cells to maintain

pH homeostasis in the cytoplasm during growth in alkalinemedia

as well. When E2 and amutant strain of E2 containing the ances-

tral kefBWT allele were grown in minimal media buffered to

different pH values ranging from 5.5 to 7.5, the growth rate of

E2 was 2.3-fold greater (p < 0.0001) in medium buffered to

pH = 7.5 (Figure 3C) and indistinguishable at lower pH values.

Similarly, E1 also grew 37.2% (p < 0.0001) faster than a mutant

of A1 with the evolved nmgRW173* allele in media buffered to

pH = 7.5, but not at lower pH values (Figures 3C and 3E).

Selective Coefficient of Adaptive Mutations IsLower in WTPrevious work in M. extorquens has demonstrated that benefi-

cial mutations that arise in the genomic context of strains with

distinct metabolic pathways are often deleterious in WT

[35, 37]. Therefore, we hypothesized that, if the NMG pathway

plays a distinct role from MaDH in AM1, the mutations in E1

and E2 may not be beneficial when moved into WT. When the

nmgRW173* mutation from E1 was introduced in WT, no signifi-

Current

cant difference in growth rate (p = 0.65)

or yield (p = 0.79) on methylamine was

observed and there was a marginally sig-

nificant 2.9% decrease in methylamine

fitness (p = 0.07; Figures 4A, 4B, and

S5). In contrast, when the kefBQ446* mutation from E2 was intro-

duced in WT, the resulting strain was 8.5% more fit (p < 0.0001)

and grew 12.0% faster (p < 0.0001) but did not have a significant

change in yield (p = 0.20; Figures 4A, 4B, and S5). However, this

fitness increase was only half that observed when the kefBQ446*

mutation was introduced in the A2 genomic background. These

results are consistent with the hypothesis that traits critical for

using the NMG pathway are at least somewhat distinct from

those needed for MaDH-dependent growth.

NMG Pathway Enables Growth on Methylamine as aNitrogen SourceHaving uncovered that the NMGpathway is tightly regulated and

may have inherent rate-limiting constraints on methylamine

metabolism due to cytoplasmic NH4+ accumulation, we devel-

oped the hypothesis that the NMG pathway might play a role

in using methylamine as a nitrogen source, especially at low

concentrations when NH4+ will not accumulate. In order to test

this idea, we performed two complementary experiments. In

the first experiment, we sought to determine whether increased

expression of the NMG pathway was beneficial during growth

with methylamine as the sole nitrogen source for E1 and E2.

As a metric for use of methylamine as a nitrogen source, we

compared the ratio of growth rates on succinate as a carbon

source with methylamine versus ammonia as the sole nitrogen

source (ksuccinate, methylamine/ksuccinate, ammonia). For E1, this ratio

was 18% greater than A1 (p = 0.002), and for E2, this ratio was

110% (p < 0.0001) greater than A2 (Figures 5A and 5B; Table

S2). Next, we tested whether AM1 strains lacking the NMG

pathway suffered a growth defect in media with succinate as

the carbon source and a wide concentration gradient of methyl-

amine as the sole nitrogen source. We observed a 30%–50%

decrease in growth rate when the methylamine concentra-

tion dropped to 1 mM or below (Figure 5C). This translates to a

2-fold advantage to the WT genotype possessing both degen-

erate pathways over the one solely containingMaDH. In addition,

the yield of the DmgdABCD mutant was significantly lower

across the entire concentration range tested (Figure 5D).

DISCUSSION

In this study, we demonstrate that, even though the NMG

pathway and MaDH each catalyze the oxidation of methylamine

Biology 26, 1416–1426, June 6, 2016 1421

A

0

0.05

0.1

0.15

0.2

0.25

0.3

WT A1 A2 E1 E2

Gro

wth

Rat

e (h

-1)

B

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

WT A1 A2 E1 E2

Yie

ld (

Max

imu

m O

D60

0)

C

0

0.5

1

1.5

2

2.5

0 2 4 6 8 10 12

Gro

wth

Rat

e (h

-1)

ofW

T /

Gro

wth

Rat

e (h

-1)

of

mg

d m

uta

nt

Methylamine Concentration (mM)

D

0

0.25

0.5

0.75

1

1.25

1.5

0 2 4 6 8 10 12

Yie

ld (

Max

. OD

600)

ofW

T/

Yie

ld (

Max

. OD

600)

of

mg

dm

uta

nt

Methylamine Concentration (mM)

periplasmcytoplasm

CH3NH3+

CH3NH3+ HCHO

NH4+, 2e-

MaDH

HCHOFAE

HCOO-

CO2

Biomass

E

periplasm

cytoplasm

F

CH3NH3+

CH3NH3+ HCHO

HCOO-

ATP ADP

NH4+

NADH + H+

CO2

Biomass

NMG pathway

Alternate Carbon Source

e-

Figure 5. A Role for the NMG Pathway during Growth in Medium with Methylamine as the Sole Nitrogen Source

(A and B) The growth rate (h�1; open; A) and yield or maximum OD600 (filled; B) of WT, the two ancestral strains (A1 and A2), and the two evolved isolates (E1 and

E2) in nitrogen-free media with 3.5 mM succinate as the carbon source and 10 mM methylamine as the sole nitrogen source.

(legend continued on next page)

1422 Current Biology 26, 1416–1426, June 6, 2016

in AM1, the former enables faster growth and higher yields on

limiting concentrations of methylamine as a nitrogen source

whereas the latter leads to rapid growth on methylamine as the

primary carbon and energy substrate.

Decades of research that had established that MaDH was the

sole pathway for the primary oxidation of methylamine in AM1

was recently challenged when a study reported that enzymes

of the NMG pathway are also encoded and expressed in this

strain [38]. To uncover the role of the NMG pathway, two strains

of AM1, previously known to be incapable of methylamine

growth but containing an intact NMG pathway [25, 26], were

experimentally evolved on methylamine as the sole carbon and

energy source. The first strain (A1) lacks the mau gene cluster

encoding MaDH (Figure 1A), whereas the second strain (A2)

has a lesion in the first gene (fae) of the formaldehyde oxidation

module (Figure 1A) that prevents methylamine growth mediated

by MaDH due to the buildup of toxic formaldehyde from methyl-

amine oxidation. Remarkably, the methylamine fitness of A1 and

A2 was significantly lower than other M. extorquens strains that

encode only the NMG pathway for methylamine oxidation [24].

Genomic, genetic, and biochemical analysis of the evolved

isolates revealed two classes of mutations that overcame the

regulatory and physiological constraints preventing methyl-

amine growth mediated by the NMG pathway. The first class

of mutations led to a significant increase in the expression level

of the NMG pathway (Figures S2 and 3A). Low expression of

enzymes of the NMG pathway in crude extracts from methyl-

amine-induced WT cells (Figure 3A) indicated that this pathway

is typically repressed, even in the presence of high concen-

trations of methylamine. Contrary to some other examples of

degeneracy, a significant growth or fitness advantage was not

observed when MaDH and the NMG pathway were simulta-

neously expressed during methylamine growth in WT (Figures

4A and S5) [1, 4]. Hence, these two degenerate pathways,

despite catalyzing the same biochemical transformation, seem

to be playing distinct roles in AM1.

The second class of adaptive mutations resulted in physiolog-

ical innovations to buffer the buildup of NH4+ ions in the cyto-

plasm, especially due to overexpression of the NMG pathway

in AM1, which typically uses a periplasmic MaDH for methyl-

amine growth. It is well established that eukaryotes mitigate

NH4+ buildup by converting it to the less-toxic, less-basic, nitrog-

enous compound urea, which is subsequently excreted (Figures

3B and 3D) [39]. To the best of our knowledge, this is the first

study to demonstrate urea excretion as a physiological response

to ameliorate the accumulation of NH4+ in the cytosol of bacterial

cells (Figure 3B). The opposite response of importing urea and

using urease to release NH4+ in order to increase the pH of the

cytoplasm is critical to how Helicobacter pylori thrives in the

acidic environment of the human stomach [40]. Additionally,

our physiological analyses also suggest that the ykkC/yxkD

RNA element represses the expression of the flanking efflux

pump unless bound to an unknown ligand. Whether this efflux

(C and D)Mean ratio of the (C) growth rate (h�1) and (D) yield ormaximumOD600 of

or 10 mM methylamine as the sole nitrogen source. The error bars represent the

(E) A schematic representing the metabolic route followed by methylamine when

(F) A schematic representing the metabolic route followed by methylamine wh

presence of other, likely more abundant, carbon substrate(s).

pump is specific for urea or has a wider substrate breath for

urea-based compounds, including many natural antimicrobial

agents, remains to be determined [41]. In E2, the kefBQ446* mu-

tation significantly enhances growth in alkalinemedia (Figure 3C);

likely, by removing the regulatory domain of the KefB K+/H+ anti-

porter and rendering it constitutively active, the H+ influx in the

cytoplasm is increased (Figures 3C and 3E). Until very recently,

cytoplasmic acidification by K+/H+ antiporters like KefB and

KefC has been singularly linked to a stress response mechanism

to counter the toxic effects of electrophiles like methylgloxal

in E. coli [31]. The results of this work broaden the cellular

role of KefB to the maintenance of pH homeostasis as well.

Interestingly, the kefBQ446* mutation in E2 mirrors the use of

dedicated monovalent cation/proton antiporters to maintain

cytoplasmic pH homeostasis under alkaline conditions in al-

kali-tolerant bacteria [42]. Furthermore, it was recently reported

that beneficial kefB alleles arose during experimental evolution of

a poor-growing M. extorquens AM1 strain engineered to use a

glutathione-based formaldehyde oxidation module, suggesting

that KefB activity may be more generally relevant to methylo-

trophy [37].

In nature, facultative methylotrophs like AM1 likely sense the

concentration of methylamine in its environment and the need

for alternative carbon or nitrogen sources in order to configure

the metabolic network accordingly to use methylamine either

as a growth substrate and/or as a nitrogen source. As a result,

the use of methylamine as a growth substrate and/or as a

nitrogen source might not be as simple as a direct correlation

with concentrations of methylamine in the environment. In envi-

ronments where methylamine is abundant and likely serves as

the preferred carbon and energy source, the highly expressed,

energetically efficient MaDH should lead to rapid growth and

the excess NH4+ produced can be released readily from the peri-

plasm (Figure 5E). In environments with higher concentrations of

other carbon sources but with limiting concentrations of methyl-

amine as the sole/preferred nitrogen source, the NMG pathway

is likely to be recruited for nitrogen assimilation (Figure 5F). Our

discovery of nmgR in regulating expression of the NMG pathway

enzymes is a start to understanding how M. extorquens AM1

make this decision. The concentration threshold for a strain

containing both pathways to have an advantage is 1 mMmethyl-

amine andmirrors the range ofmethylamine concentrations typi-

cally observed in different environments, thereby suggesting the

concentration-dependent advantage of degeneracy observed

here may be of great relevance in nature [43].

These two functionally degenerate pathways for methylamine

oxidation coexist in one-third of sequenced methylotrophs with

at least one methylamine oxidation pathway [24]. Functionally

degenerate pathways specialized for unique ecological roles

would only coexist long-term in a genome if there were tradeoffs

associated with each pathway. The cytoplasmic localization

of the NMG pathway [19] might allow cells to capture a higher

fraction of NH4+ ions released from low concentrations of

WTand theDmgdABCDmutant inmedia with 3.5mMsuccinate and 0.5, 1.0, 5,

95% CI of the mean of three biological replicates.

M. extorquens AM1 uses methylamine as a growth substrate.

en M. extorquens AM1 uses methylamine as the sole nitrogen source in the

Current Biology 26, 1416–1426, June 6, 2016 1423

methylamine relative to MaDH, a periplasmic enzyme. On the

other hand,MaDH iswell situated tousemethylamineasagrowth

substrate because NH4+ excretion can occur without transport

costs from the periplasm [16]. Thus, despite catalyzing the

same biochemical reaction, the two modules play ecologically

distinct roles to enable methylotrophs to optimally utilize methyl-

amine either as a growth substrate or as a nitrogen source.

EXPERIMENTAL PROCEDURES

Chemicals and Media

All chemicals were purchased from Sigma-Aldrich. E. coli strains were grown

in Luria-Bertani broth at 37�C with the standard antibiotic concentrations.

Standard growth conditions for AM1 utilized a modified version of Hypho min-

imal medium described previously [44]. Nitrogen-free Hypho minimal media

was made with a nitrogen-free sulfate salts solution (7 g of MgSO4 , 7 H2O

in 1 l deionized water). All components were autoclaved separately before mix-

ing under sterile conditions. Filter-sterilized carbon sources were added just

prior to inoculation in liquid minimal media with a final concentration of

3.5 mM for sodium succinate and 20 mM for methylamine hydrochloride.

Experimental Evolution

Experimental Evolution I

M. extorquens AM1 was serially transferred for 1,500 generations in medium

with 1.75 mM disodium succinate and 7.5 mM methanol as described earlier

[25, 27]. An evolved isolate (CM1054; here, A1) that could not grow on methyl-

amine was subsequently grown on 20 mM methylamine hydrochloride as

described in Supplemental Experimental Procedures.

Experimental Evolution II

A single population of a Dfae Dcel strain ofMethylobacterium extorquens AM1

(CM2770; here, A2) was serially transferred in modified Hypho medium with

20 mM methylamine hydrochloride as described in Supplemental Experi-

mental Procedures.

Growth Rate Measurements

Cells were acclimated and grown in 48-well microtiter plates (CoStar-3548)

containing 640 ml of growthmedium, in a shaking incubator at 650 rpm (Liconic

USA LTX44 with custom fabricated cassettes), in a room that was constantly

maintained at 30�C and 80% humidity as described previously [45]. The spe-

cific growth rate of cultures was calculated from the log-linear growth phase

using an open source, custom-designed growth analysis software called

CurveFitter available at http://www.evolvedmicrobe.com/CurveFitter/.

Plasmids and Strains

Markerless chromosomal allele replacements or gene deletions were con-

ducted using the allelic exchange vector pCM433 [46]. Genomic regions

were amplified by PCR and ligated on the pCM433 backbone cut with NotI

using Gibson assembly [47]. Mutant strains ofM. extorquens AM1were gener-

ated through conjugation by a tri-parental mating technique described else-

where [46]. Mutants were confirmed by a diagnostic PCR and validated by

Sanger sequencing. All strains and plasmids used and generated for this study

are listed in Table S3.

Competitive Fitness Assays

Fitness measurements were estimated by competing each strain against

M. extorquens AM1Dcel-DkatA::Ptac-mCherry (CM3120) described elsewhere

[45] and counting the fraction of fluorescent to non-fluorescent cells using the

LSR Fortesssa (Becton Dickinson) as described in the Supplemental Experi-

mental Procedures section. The competitive fitness was calculated as

W = logðR1 � N=R0Þ=logðð1� R1Þ � N=ð1� R0ÞÞ, where R1 and R0 represent

the population fraction of the test strain before and after mixed growth and

N represents the fold increase in the population density.

NMG Dehydrogenase Assay

Methylamine-induced cultures were lysed using a FastPrep-24 (MP Bio)

and lysing matrix B (MP Bio). Lysates were centrifuged at 13,000 rpm for

5 min in a centrifuge maintained at 4�C, and total protein was quantified

1424 Current Biology 26, 1416–1426, June 6, 2016

using the Quick Start Bradford Assay (Bio-Rad) using BSA as a standard.

0.1 mg of protein was incubated in sodium phosphate buffer (pH = 7.6)

with 5 mM NMG and 0.5 mM NAD+ at room temperature for 40 min and at

58�C for 30 min after Nash reagent [48] was added. Formaldehyde produc-

tion was estimated by measuring the absorbance at 412 nm, and NMGDH

activity was measured as the nmol formaldehyde produced per mg protein

per minute.

Genome Sequencing

Cell pellets were lysed using a combination of lysozyme and heat shock at

90�C for 10 min. Lysed cells were treated with proteinase K and RNase A at

37�C, and genomic DNA was extracted using the phenol-chloroform extrac-

tion protocol. Single-end 50-bp reads for Illumina sequencing were prepared

using the TruSeq DNA sample prep Kit (Illumina) and sequenced using an Illu-

mina Hi-Seq at the Microarray and Genomics Core Facility, Huntsman Cancer

Institute. Reads were assembled and aligned to the M. extorquens AM1

genome [49] using Breseq version 0.21 [50], and mutations were verified by

PCR and Sanger sequencing.

qRT-PCR

RNA from methylamine-induced cultures was extracted using an RNeasy Kit

(QIAGEN) and quantified using a Nanodrop spectrophotometer ND-1000

(Thermo Fisher Scientific). One microgram RNAwas reverse transcribed using

SuperScript III (Life Technologies), and qPCR was performed using a CFX-96

qPCR machine (Bio-Rad) and EvaGreen qPCR mix (Biotium). Ribosomal pro-

tein RpsB was used as an internal standard, and the relative expression of

Meta1_4101 and Meta1_4103 was quantified using a protocol described else-

where [34, 51]. Three technical replicateswere performed for each of three bio-

logical replicates. A no-template control, a no-RT control, and a standard

curve with DNA concentrations from 1 ng to 10�6 ng was run for every primer

pair. Gene-specific primers designed for rpsB, Meta1_4101, and Meta1_4103

are listed in Table S4.

Urea Quantification Assay

Spent media from cultures was collected by centrifugation and filtration using

a sterile 0.2-mmmembrane filter (Millipore). Urea concentration was measured

using the Urea Assay Kit (Sigma-Aldrich) per the manufacturer’s instructions.

Absorbance was measured in 96-well, flat-bottom plates (CoStar-3595) using

the Tecan Safire2 plate reader (Tecan Group). Two technical replicates were

performed for each of three biological replicates.

ACCESSION NUMBERS

The accession number for the sequence data reported in this paper is Bio-

Project: PRJNA316889.

SUPPLEMENTAL INFORMATION

Supplemental Information includes five figures, four tables, and Supplemental

Experimental Procedures and can be found with this article online at http://dx.

doi.org/10.1016/j.cub.2016.04.029.

AUTHOR CONTRIBUTIONS

D.D.N. and C.J.M. designed experiments. D.D.N., D.A., and M.-C.L. per-

formed experiments. D.D.N. and C.J.M. analyzed data. D.D.N. and C.J.M.

wrote the manuscript.

ACKNOWLEDGMENTS

We would like to thank members of the C.J.M. lab for their feedback on the

manuscript. C.J.M. acknowledges financial support from the NIH (GM078209).

Received: December 29, 2015

Revised: February 17, 2016

Accepted: April 11, 2016

Published: May 19, 2016

REFERENCES

1. Marx, C.J., Miller, J.A., Chistoserdova, L., and Lidstrom, M.E. (2004).

Multiple formaldehyde oxidation/detoxification pathways in Burkholderia

fungorum LB400. J. Bacteriol. 186, 2173–2178.

2. Marx, C.J., Van Dien, S.J., and Lidstrom, M.E. (2005). Flux analysis un-

covers key role of functional redundancy in formaldehyde metabolism.

PLoS Biol. 3, e16.

3. Denef, V.J., Klappenbach, J.A., Patrauchan, M.A., Florizone, C.,

Rodrigues, J.L.M., Tsoi, T.V., Verstraete, W., Eltis, L.D., and Tiedje, J.M.

(2006). Genetic and genomic insights into the role of benzoate-catabolic

pathway redundancy in Burkholderia xenovorans LB400. Appl. Environ.

Microbiol. 72, 585–595.

4. Gerth, M.L., Ferla, M.P., and Rainey, P.B. (2012). The origin and ecological

significance of multiple branches for histidine utilization in Pseudomonas

aeruginosa PAO1. Environ. Microbiol. 14, 1929–1940.

5. Tononi, G., Sporns, O., and Edelman, G.M. (1999). Measures of degener-

acy and redundancy in biological networks. Proc. Natl. Acad. Sci. USA 96,

3257–3262.

6. Boucher, Y., Douady, C.J., Papke, R.T., Walsh, D.A., Boudreau, M.E.R.,

Nesbø, C.L., Case, R.J., and Doolittle, W.F. (2003). Lateral gene transfer

and the origins of prokaryotic groups. Annu. Rev. Genet. 37, 283–328.

7. Kafri, R., Springer, M., and Pilpel, Y. (2009). Genetic redundancy: new

tricks for old genes. Cell 136, 389–392.

8. Wang, Z., and Zhang, J. (2009). Abundant indispensable redundancies in

cellular metabolic networks. Genome Biol. Evol. 1, 23–33.

9. Stelling, J., Sauer, U., Szallasi, Z., Doyle, F.J., 3rd, and Doyle, J. (2004).

Robustness of cellular functions. Cell 118, 675–685.

10. van Heeswijk, W.C., Westerhoff, H.V., and Boogerd, F.C. (2013). Nitrogen

assimilation in Escherichia coli: putting molecular data into a systems

perspective. Microbiol. Mol. Biol. Rev. 77, 628–695.

11. Helling, R.B. (2002). Speed versus efficiency in microbial growth and the

role of parallel pathways. J. Bacteriol. 184, 1041–1045.

12. Brenchley, J.E., Prival, M.J., and Magasanik, B. (1973). Regulation of the

synthesis of enzymes responsible for glutamate formation in Klebsiella

aerogenes. J. Biol. Chem. 248, 6122–6128.

13. Chistoserdova, L., Chen, S.W., Lapidus, A., and Lidstrom, M.E. (2003).

Methylotrophy inMethylobacterium extorquensAM1 from a genomic point

of view. J. Bacteriol. 185, 2980–2987.

14. Chistoserdova, L. (2011). Modularity of methylotrophy, revisited. Environ.

Microbiol. 13, 2603–2622.

15. Nayak, D.D., and Marx, C.J. (2014). Methylamine utilization via the

N-methylglutamate pathway in Methylobacterium extorquens PA1 in-

volves a novel flow of carbon through C1 assimilation and dissimilation

pathways. J. Bacteriol. 196, 4130–4139.

16. McIntire, W.S., Wemmer, D.E., Chistoserdov, A., and Lidstrom, M.E.

(1991). A new cofactor in a prokaryotic enzyme: tryptophan tryptophylqui-

none as the redox prosthetic group in methylamine dehydrogenase.

Science 252, 817–824.

17. Chistoserdov, A.Y., Chistoserdova, L.V., McIntire, W.S., and Lidstrom,M.E.

(1994). Genetic organization of the mau gene cluster in Methylobacterium

extorquens AM1: complete nucleotide sequence and generation and

characteristics ofmaumutants. J. Bacteriol. 176, 4052–4065.

18. Chen, L., Durley, R.C., Mathews, F.S., andDavidson, V.L. (1994). Structure

of an electron transfer complex: methylamine dehydrogenase, amicyanin,

and cytochrome c551i. Science 264, 86–90.

19. Latypova, E., Yang, S., Wang, Y.S., Wang, T., Chavkin, T.A., Hackett, M.,

Schafer, H., and Kalyuzhnaya, M.G. (2010). Genetics of the glutamate-

mediated methylamine utilization pathway in the facultative methylo-

trophic beta-proteobacterium Methyloversatilis universalis FAM5. Mol.

Microbiol. 75, 426–439.

20. Marx, C.J., Chistoserdova, L., and Lidstrom, M.E. (2003). Formaldehyde-

detoxifying role of the tetrahydromethanopterin-linked pathway in

Methylobacterium extorquens AM1. J. Bacteriol. 185, 7160–7168.

21. Chistoserdova, L., Laukel, M., Portais, J.C., Vorholt, J.A., and Lidstrom,

M.E. (2004). Multiple formate dehydrogenase enzymes in the facultative

methylotroph Methylobacterium extorquens AM1 are dispensable for

growth on methanol. J. Bacteriol. 186, 22–28.

22. Chistoserdova, L., Crowther, G.J., Vorholt, J.A., Skovran, E., Portais,

J.-C., and Lidstrom, M.E. (2007). Identification of a fourth formate dehy-

drogenase in Methylobacterium extorquens AM1 and confirmation of the

essential role of formate oxidation in methylotrophy. J. Bacteriol. 189,

9076–9081.

23. Crowther, G.J., Kosaly, G., and Lidstrom, M.E. (2008). Formate as the

main branch point for methylotrophic metabolism in Methylobacterium

extorquens AM1. J. Bacteriol. 190, 5057–5062.

24. Nayak, D.D., and Marx, C.J. (2015). Experimental horizontal gene transfer

of methylamine dehydrogenase mimics prevalent exchange in nature and

overcomes the methylamine growth constraints posed by the sub-optimal

N-methylglutamate pathway. Microorganisms 3, 60–79.

25. Lee, M.C., and Marx, C.J. (2012). Repeated, selection-driven genome

reduction of accessory genes in experimental populations. PLoS Genet.

8, e1002651.

26. Agashe, D., Martinez-Gomez, N.C., Drummond, D.A., and Marx, C.J.

(2013). Good codons, bad transcript: large reductions in gene expression

and fitness arising from synonymousmutations in a key enzyme. Mol. Biol.

Evol. 30, 549–560.

27. Lee, M.C., Chou, H.H., and Marx, C.J. (2009). Asymmetric, bimodal trade-

offs during adaptation of Methylobacterium to distinct growth substrates.

Evolution 63, 2816–2830.

28. Agashe, D., Sane, M., Phalnikar, K., Diwan, G.D., Habibullah, A., Martinez-

Gomez, N.C., Sahasrabuddhe, V., Polachek, W., Wang, J., Chubiz, L.M.,

and Marx, C.J. (2016). Large-effect beneficial synonymous mutations

mediate rapid and parallel adaptation in a bacterium. Mol. Biol. Evol. 33,

1542–1553.

29. Vorholt, J.A., Marx, C.J., Lidstrom, M.E., and Thauer, R.K. (2000). Novel

formaldehyde-activating enzyme in Methylobacterium extorquens AM1

required for growth on methanol. J. Bacteriol. 182, 6645–6650.

30. Barrick, J.E., Corbino, K.A., Winkler, W.C., Nahvi, A., Mandal, M., Collins,

J., Lee, M., Roth, A., Sudarsan, N., Jona, I., et al. (2004). New RNA motifs

suggest an expanded scope for riboswitches in bacterial genetic control.

Proc. Natl. Acad. Sci. USA 101, 6421–6426.

31. Ferguson, G.P., Nikolaev, Y., McLaggan, D., Maclean, M., and Booth, I.R.

(1997). Survival during exposure to the electrophilic reagent N-ethylmalei-

mide in Escherichia coli: role of KefB and KefC potassium channels.

J. Bacteriol. 179, 1007–1012.

32. Rasche,M.E.,Havemann,S.A., andRosenzvaig,M. (2004).Characterization

of two methanopterin biosynthesis mutants of Methylobacterium extor-

quens AM1 by use of a tetrahydromethanopterin bioassay. J. Bacteriol.

186, 1565–1570.

33. Van Dien, S.J., and Lidstrom, M.E. (2002). Stoichiometric model for

evaluating the metabolic capabilities of the facultative methylotroph

Methylobacterium extorquens AM1, with application to reconstruction of

C(3) and C(4) metabolism. Biotechnol. Bioeng. 78, 296–312.

34. Chou, H.H., and Marx, C.J. (2012). Optimization of gene expression

through divergent mutational paths. Cell Rep. 1, 133–140.

35. Michener, J.K., CamargoNeves, A.A., Vuilleumier, S., Bringel, F., andMarx,

C.J. (2014). Effective use of a horizontally-transferred pathway for dichloro-

methane catabolism requires post-transfer refinement. eLife 3, 1–16.

36. Roosild, T.P., Castronovo, S., Miller, S., Li, C., Rasmussen, T., Bartlett, W.,

Gunasekera, B., Choe, S., and Booth, I.R. (2009). KTN (RCK) domains

regulate K+ channels and transporters by controlling the dimer-hinge

conformation. Structure 17, 893–903.

37. Carroll, S.M., Chubiz, L.M., Agashe, D., andMarx, C.J. (2015). Parallel and

divergent evolutionary solutions for the optimization of an engineered cen-

tral metabolism inMethylobacterium extorquens AM1. Microorganisms 3,

152–174.

Current Biology 26, 1416–1426, June 6, 2016 1425

38. Martinez-Gomez, N.C., Nguyen, S., and Lidstrom,M.E. (2013). Elucidation

of the role of the methylene-tetrahydromethanopterin dehydrogenase

MtdA in the tetrahydromethanopterin-dependent oxidation pathway in

Methylobacterium extorquens AM1. J. Bacteriol. 195, 2359–2367.

39. Singer, M.A. (2003). Do mammals, birds, reptiles and fish have similar ni-

trogen conserving systems? Comp. Biochem. Physiol. B Biochem. Mol.

Biol. 134, 543–558.

40. Weeks, D.L., Eskandari, S., Scott, D.R., and Sachs, G. (2000). A H+-gated

urea channel: the link between Helicobacter pylori urease and gastric

colonization. Science 287, 482–485.

41. Tegos, G.P., Masago, K., Aziz, F., Higginbotham, A., Stermitz, F.R., and

Hamblin, M.R. (2008). Inhibitors of bacterial multidrug efflux pumps poten-

tiate antimicrobial photoinactivation. Antimicrob. Agents Chemother. 52,

3202–3209.

42. Padan, E., Bibi, E., Ito, M., and Krulwich, T.A. (2005). Alkaline pH homeo-

stasis in bacteria: new insights. Biochim. Biophys. Acta 1717, 67–88.

43. Ge, X., Wexler, A.S., and Clegg, S.L. (2011). Atmospheric amines - Part I. A

review. Atmos. Environ. 45, 524–546.

44. Nayak, D.D., and Marx, C.J. (2014). Genetic and phenotypic comparison

of facultative methylotrophy between Methylobacterium extorquens

strains PA1 and AM1. PLoS ONE 9, e107887.

45. Delaney, N.F., Kaczmarek, M.E., Ward, L.M., Swanson, P.K., Lee, M.C.,

and Marx, C.J. (2013). Development of an optimized medium, strain and

1426 Current Biology 26, 1416–1426, June 6, 2016

high-throughput culturing methods for Methylobacterium extorquens.

PLoS ONE 8, e62957.

46. Marx, C.J. (2008). Development of a broad-host-range sacB-based vector

for unmarked allelic exchange. BMC Res. Notes 1, 1.

47. Gibson, D.G., Young, L., Chuang, R.-Y., Venter, J.C., Hutchison, C.A., 3rd,

and Smith, H.O. (2009). Enzymatic assembly of DNA molecules up to

several hundred kilobases. Nat. Methods 6, 343–345.

48. Nash, T. (1953). The colorimetric estimation of formaldehyde by means of

the Hantzsch reaction. Biochem. J. 55, 416–421.

49. Vuilleumier, S., Chistoserdova, L., Lee, M.C., Bringel, F., Lajus, A., Zhou,

Y., Gourion, B., Barbe, V., Chang, J., Cruveiller, S., et al. (2009).

Methylobacterium genome sequences: a reference blueprint to investi-

gate microbial metabolism of C1 compounds from natural and industrial

sources. PLoS ONE 4, e5584.

50. Barrick, J.E., Colburn, G., Deatherage, D.E., Traverse, C.C., Strand, M.D.,

Borges, J.J., Knoester, D.B., Reba, A., and Meyer, A.G. (2014). Identifying

structural variation in haploid microbial genomes from short-read rese-

quencing data using breseq. BMC Genomics 15, 1039.

51. Chou, H.H., Berthet, J., and Marx, C.J. (2009). Fast growth increases the

selective advantage of a mutation arising recurrently during evolution un-

der metal limitation. PLoS Genet. 5, e1000652.

Current Biology, Volume 26

Supplemental Information

Selection Maintains Apparently Degenerate

Metabolic Pathways due to Tradeoffs

in Using Methylamine for Carbon versus Nitrogen

Dipti D. Nayak, Deepa Agashe, Ming-Chun Lee, and Christopher J. Marx

Figure S1: Related to Figure 1. A schematic of the N-methylglutamate pathway. The topology of the N-methylglutamate pathway in M. extroquens species has been shown to be semi-linear: N-methylglutamate synthase is capable of synthesizing N-methylglutamate from γ­-glutamylmethylamide (blue arrow) as well as methylamine (gray arrow) albeit only in the ∆gmas mutant. NMGDH: N-methylglutamate dehydrogenase, MGS: N-methylglutamate synthase, GMAS: γ­-glutamylmethylamide synthetase, NMG – N-methylglutamate, GMA - γ­-glutamylmethylamide

CH3NH2

NM

GS*

GM

AS Glutamate, ATP

ADP

(GMA)

HCHO

NM

GD

H

NH4+

H2O, NAD+

Glutamate, NADH

(NMG)

NM

GS

NH4+

Figure S2: Related to Figure 2. Mutations in E1 and E2. A schematic representation of the adaptive mutations in A) E1 and B) E2 that enable methylamine growth mediated by the N-methylglutamate pathway.

IS IS

mau gene cluster

IS

∆mau

kefB

Q446*

Truncated kefB

NMG pathway gene cluster

Duplication

1 2 3

1 2NMG pathway gene clusterG1617391A, W173*

nmgR

Truncated nmgR

ykkC-yxkD

ABC Transporter

IS

IS insertion in ykkC - yxkD

G/B/P TransporterIS

ABC Transporter G/B/P Transporter

A

B

Figure S3: Related to Figure 2. Fitness of kefB alleles the E2 genomic background. Competitive fitness of E2, a mutant of E2 with an in-frame deletion in the kefB CDS, and a mutant of E2 with the WT allele of kefB on 20 mM methylamine hydrochloride relative to CM3120 (M. extorquens AM1 ∆cel-∆katA::Ptac-mCherry). The error bars indicate the 95% CI of the mean fitness value of three replicate competition assays. Note: The y-axis does not start at 0 to highlight the fitness different between the three strains.

0.3

0.32

0.34

0.36

0.38

0.4

0.42

0.44

1 2 3

Fitn

ess

E2 E2 ∆kefB E2 kefBWT

Figure S4: Related to Figure 3. Mutations ameliorating cytoplasmic NH4+ accumulation in E1 and E2. A)

Genes flanking the ykkC/yxkD RNA element in M. extorquens AM1. In pink and blue are the two ABC transporters on either side of this RNA element. The ABC-type nitrate/sulfonate/bicarbonate efflux transporter (in pink) is commonly associated with the ykkC/yxkD RNA element. This efflux transporter is in close proximity to urea metabolism genes (in yellow) and other nitrogen metabolism genes as well (in gray).B) The amino acid sequence of kefB (Meta1_2712) showing the K+/H+ ion exchanger domain in green and the RCK (regulator of K+ conductance) domain in orange. The mutated residue in E2 is highlighted in blue.

ykkC-yxkD

ABC efflux transporter (Meta1_4100 to Meta1_4102)

Glycine/betaine/proline transporter(Meta1_4013 to Meta1_4015)

Creatininase

gmaS homolog

Urea amidolyase with urea carboxylase and allophanate hydrolase subunits

(Meta1_4095)

Putative regulator

A

MASATDHASFLPPVLTFLSAAVIGVPLVRLLGQSAVLGYLVAGVVIGPAGFSLIAEPETAASVAEIGVVLLLFIVGLELEISRLVSMRRDIFGLGAAQLALCSLVIAGAALAYGLTPAAASVVGIAIALSATAVALQLLEERGDLGSPYGGRSFAVLLFQDISVVAILALLPLLASAGATPKDGWLDEGLRSTGRAVVAVAGVVLVGRYGLNPFFRLLAAAGGREVMTAAALLVVLGTALVMEKAGLSMAMGAFLAGVMLAESNFRHQLEADIEPFRGMLLGLFFMSVGMSIDGGLLTNHWLALLAATMAAIAVKIALVAGLFRLFGSPWLDALRGAAVLAPAGEFAFVILPAAGDLRILASDVTRFCVALAALTMLVGPIAAKGLDALIARRPKEAEPPSDVGEAQESG

DTRVLVVGFGRFGQILAQVLLAEGINITVIDKDVEQ IRNATSFGFRIYYGDGTRLDVLRASGLAKADLICVCIDDAPAALKIVDIVHEEFPNVRTYVRAYDRTHAIELMNRDVDFQLRETVESALGFGRATLESLGLPAEAAARRVEDVRKRDVARLVLQQAGAMPDGSGWLRGTPELRPEPLTAPKSPSRALSAETRGLIEQQPHESAARALEPEDAEPNA*

B

Urea degradation associated genes (Meta1_4096 and Meta1_4097)

Figure S5: Related to Figure 4. Fitness of adaptive mutations in E1 and E2 in the WT background. Competitive fitness of WT, a mutant strain of WT containing the nmgRW173* allele from E1, and a mutant strain of WT containing the kefBQ446* allele from E2 relative on 20 mM methylamine hydrochloride. The error bars indicate the 95% CI of the mean fitness value of three replicate competition assays.

0

0.2

0.4

0.6

0.8

1

1.2

WT WT nmgRevo WT kefBevo

Fitn

ess

WTnmgRW173*

WTkefBQ446*

Table S1: Related to Figure 1. Mutations in A1, E1, and E2. A list of mutations in A1 relative to the M. extroquens AM1 reference genome and in E1 and E2 relative to their respective ancestors. Each of these mutations were verified by Sanger sequencing.

A1 E1 E2

G420551C (R73P)Meta_0397

G1617391A (W173*)Meta1_1544 or nmgR

G2837255A(Q446*)kefB

2308994 (+T)Meta1_2237

ISmex4 insertion in ykkC-yxkD

Deletion (28679 bp-30272 bp)Meta1_2751-Meta1_2783

(Δmau gene cluster)

G3724230A(intergenic)Meta1_3582/Meta1_3583

New JunctionPosition 162893, Meta1_1554

or nmgR, and position 1616982, Meta1_1544 or

nmgR

A4125088G (intergenic)Meta1_4038/rffH

A11165G(R31R)Meta2_1054

T144186G(Q24H)Meta2_0154

C5903G (intergenic)p2meta_006/p2meta_007

Deletion (28679bp -30281 bp)Meta1_2751 -Meta1_2783

(Δmau gene cluster)

Deletion (61658bp -61678bp)Meta1_3749-Meta1_tRNA33

Deletion (~0.6Mb)Meta2_0933 – Meta1_0277

ISmex4 insertionMeta1_2101/Meta1_2102

ISmex4 insertionglnE

ISmex4 insertionureA/icuA

Table S2: Related to Figure 5. Growth on ammonium versus methylamine as nitrogen source. Growth rate (h-1) of WT, E1, and E2 in media with 3.5 mM disodium succinate and either 7.66 mM ammonia as the sole nitrogen source (blue column) or 7.66 mM methylamine as the sole nitrogen source (green column). The values indicate the mean of three replicate growth measurements and the 95% CI.

Strain k(succinate, ammonia)

k(succinate, methylamine)

Ratio k(succinate,

methylamine)k(succinate, ammonia)

WT 0.207±0.001 0.220±0.001 1.059±0.013

E1 0.271±0.001 0.245±0.001 0.906±0.011

E2 0.251±0.004 0.222±0.001 0.884±0.021

Table S3. M.extorquens strains and plasmids used in this study

Strains/Plasmids Description Reference

Strains

CM501 M. extorquens AM1 S1

CM1054 A1 S2

CM2351 Δmau in M. extorquens AM1 This study

CM2720 Δcel in M. extorquens AM1 S3

CM2770 Δfae in CM2720 (A2) This study

CM2986 E2 This study

CM3014 E1 This study

CM3488 CM2986 + pAYC139 This study

CM3807 ΔmptG in CM2720 This study

CM3811 ΔmptG in CM2986 This study

CM3813 ΔmptG in CM3014 This study

CM4135 kefBQ446* in CM2770 This study

CM4140 ΔkefB in CM2986 This study

CM4142 kefBWT in CM2986 This study

CM4149 ΔmgdABCD in CM2986 This study

CM4152 ΔmgdABCD in CM3014 This study

CM4153 nmgRWT in CM3014 This study

CM4168 nmgRW173* in CM1054 This study

CM4169 kefBQ446* in CM2720 This study

CM4187 ΔmgdABCD in CM4149 This study

CM4189 ΔmgdABCD in CM2720 This study

CM4194 nmgRW173* in CM2720

This study

CM4311 Δfae in CM2351 This study

Plasmids

pCM433 Allelic exchange vector (TetR, SucroseS) S1

pDN50 pCM433 with Δ fae upstream and downstream flanks S4

pDN69 pCM433 with ΔmptG upstream and downstream flanks This study

pDN112 pCM433 with ΔmgdABCD upstream and downstream flanks This study

pDN136 pCM433 with kefB allele from E2 This study

pDN137 pCM433 with kefB allele from WT This study

pDN142 pCM433 with ΔkefB upstream and downstream flanks This study

pDN150 pCM433 with nmgR allele from E1 This study

pDN151 pCM433 with nmgR allele from WT This study

pAYC139 Plasmid with mau gene cluster from M. extorquens AM1 S5

pRK2073 Conjugative helper plasmid (IncP, tra, StrR) S6

Table S4 List of primers used for q-RT PCR in this study

Primer Description Sequence

rpsB_qPCR_f (internal standard) TGACCAACTGGAAGACCATCTCC

rpsB_qPCR_r (internal standard) TTGGTGTCGATCACGAACAGCAG

qRT-DNA_contamination_check_f TCTGAGGTTCCAGGCTCGCTC

qRT-DNA_contamination_check_r TGCGACACTACGCCTTGGCAC

Meta1_4101_f GTGATCGAGTTCGTGCGCTA

Meta1_4101_r GAAGAAGGTGCCGATGACGA

Meta1_4103_f AGGGCTGGGAGACGAAGTAT

Meta1_4103_r GTCTTCTTGAGCTTGCGCTG

Supplemental Experimental Procedures

Experimental Evolution

Experimental Evolution I: Eight replicate populations of Methylobacterium extorquens AM1 (CM501 and CM502) were evolved in Hypho medium with 1.75 mM disodium succinate and 7.5 mM methanol for 1500 generations in an orbital, shaking incubator maintained at a constant speed (225 rpm) and temperature (30 ºC) [S7]. An evolved isolate (CM1054 or A1) from population C1 that could not grow on methylamine [S7] was subsequently evolved in 10 mL of modified Hypho minimal medium with 20 mM methylamine hydrochloride in an orbital, shaking incubator maintained at a constant speed (225 rpm) and temperature (30 ºC). The culture was allowed to grow for ~14 days till the media turned cloudy and was then plated on Hypho, 20 mM methylamine, agar (2% w/v) plates to obtain single colonies. Strain CM3014 (E1) was isolated from this plate and was used for further studies.

Experimental Evolution II: A single population of a ∆fae ∆cel strain of Methylobacterium extorquens AM1 (CM2770 or A2) was experimentally evolved in one well of a 48-well microtiter plates (CoStar-3548) containing 640 μl of modified Hypho minimal medium containing 20 mM methylamine hydrochloride. Plates were maintained in a shaking incubator at 650 rpm (Liconic USA LTX44 with custom fabricated cassettes) in a room that was constantly maintained at 30 ºC and 80% humidity and the culture was allowed to grow for ~7 days till the media turned cloudy. A 1/32 dilution was serially passaged every 3.5 days for 150 generations. Strain CM2986 (i.e. E2) was isolated as a single colony when the evolved population was plated on Hypho, 20 mM methylamine, agar (2% w/v).

Competitive fitness assays

Growth of the test strain and M. extorquens AM1 ∆cel-∆katA::Ptac-mCherry (CM3120) was initiated by inoculating 10 µL of the freezer stock of each strain into 10 mL of Hypho medium with 3.5 mM disodium succinate. Cultures were incubated in an orbital, shaking incubator maintained at a constant speed (225 rpm) and temperature (30 ºC). Upon reaching stationary phase, cultures were transferred 1:64 in 48-well microtiter plates (CoStar-3548) containing 640 μl of the modified Hypho medium with 20 mM methylamine hydrochloride. Plates were maintained in a shaking incubator at 650 rpm (Liconic USA LTX44 with custom fabricated cassettes) in a room that was constantly maintained at 30 ºC and 80% humidity and allowed to reach saturation in this ‘acclimation’ phase. The acclimation phase for strains that did not (or could barely) grow on methylamine consisted of growth in modified Hypho medium with 3.5 mM disodium succinate and 20 mM methylamine hydrochloride. At the end of the acclimation phase, CM3120 and the test strain were mixed in equal proportions by volume and a 64-fold dilution of this initial mix (T0) was transferred into fresh 48-well microtiter plates (CoStar-3548) containing 640 μl of the modified Hypho medium with 20 mM methylamine hydrochloride. The remaining 450 µL of the T0 sample was mixed with 10% DMSO and frozen at -80 ºC. Plates containing the T0 mix were incubated under the same conditions as the ‘acclimation’ phase. At the onset of saturation phase, a 500 µL sample of the culture (T1) was collected. The ratio of CM3120 and the test strain before and after growth was ascertained using flow cytometry. Cells were diluted appropriately such that at a flow rate of 0.5 μl/sec on the LSR Fortesssa (BD, Franklin Lakes, NJ) ~1000 events/second would be recorded. Fluorescent mCherry was excited at 561

nm and measured at 620/10 nm. The competitive fitness was calculated as 𝑊 =!"# (!!∗!!! )

!"# ( !!!! ∗!!!!! )

where R1 and

R0 represent the population fraction of the test strain before and after mixed growth, and N represents the fold increase in the population density.

Supplemental References

S1. Marx CJ. Development of a broad-host-range sacB-based vector for unmarked allelic exchange. BMC

Res Notes 2008;1:1. doi:10.1186/1756-0500-1-1.

S2. Lee MC, Marx CJ. Repeated, selection-driven genome reduction of accessory genes in experimental populations. PLoS Genet 2012;8:2–9. doi:10.1371/journal.pgen.1002651.

S3. Delaney NF, Kaczmarek ME, Ward LM, Swanson PK, Lee MC, Marx CJ. Development of an

optimized medium, strain and high-throughput culturing methods for Methylobacterium extorquens. PLoS One 2013;8. doi:10.1371/journal.pone.0062957.

S4. Nayak DD, Marx CJ. Genetic and phenotypic comparison of facultative methylotrophy between

Methylobacterium extorquens strains PA1 and AM1. PLoS One 2014;9:e107887. doi:10.1371/journal.pone.0107887.

S5. Chistoserdov AY, Chistoserdova L V., McIntire WS, Lidstrom ME. Genetic organization of the mau

gene cluster in Methylobacterium extorquens AM1: Complete nucleotide sequence and generation and characteristics of mau mutants. J Bacteriol 1994;176:4052–65.

S6. Figurski DH, Helinski DR. Replication of an origin-containing derivative of plasmid RK2 dependent

on a plasmid function provided in trans. Proc Natl Acad Sci U S A 1979;76:1648–52. S7. Lee MC, Chou HH, Marx CJ. Asymmetric, bimodal trade-offs during adaptation of Methylobacterium

to distinct growth substrates. Evolution 2009;63:2816–30. doi:10.1111/j.1558-5646.2009.00757.x.