A G-protein editor gates coenzyme B12 loading and is corrupted in methylmalonic aciduria.

Abstract

The mechanism by which docking fidelity is achieved for the multitude of cofactor-dependent enzymes is poorly understood. In this study, we demonstrate that delivery of coenzyme B(12) or 5'-deoxyadenosylcobalamin by adenosyltransferase to methylmalonyl-CoA mutase is gated by a small G protein, MeaB. While the GTP-binding energy is needed for the editing function; that is, to discriminate between active and inactive cofactor forms, the chemical energy of GTP hydrolysis is required for gating cofactor transfer. The G protein chaperone also exerts its editing function during turnover by using the binding energy of GTP to elicit release of inactive cofactor that is occasionally formed during the catalytic cycle of MCM. The physiological relevance of this mechanism is demonstrated by a patient mutation in methylmalonyl-CoA mutase that does not impair the activity of this enzyme per se but corrupts both the fidelity of the cofactor-loading process and the ejection of inactive cofactor that forms occasionally during catalysis. Consequently, cofactor in the incorrect oxidation state gains access to the mutase active site and is not released if generated during catalysis, leading, respectively, to assembly and accumulation of inactive enzyme and resulting in methylmalonic aciduria.

Full text

Page 1

A G-protein editor gates coenzyme B12loading
and is corrupted in methylmalonic aciduria
Dominique Padovani and Ruma Banerjee1
Department of Biological Chemistry, University of Michigan Medical Center, Ann Arbor, MI 48109-5606
Edited by Judith P. Klinman, University of California, Berkeley, CA, and approved October 2, 2009 (received for review July 21, 2009)
The mechanism by which docking fidelity is achieved for the
multitude of cofactor-dependent enzymes is poorly understood. In
this study, we demonstrate that delivery of coenzyme B12 or
5?-deoxyadenosylcobalamin by adenosyltransferase to methylma-
lonyl-CoA mutase is gated by a small G protein, MeaB. While the
GTP-binding energy is needed for the editing function; that is, to
discriminate between active and inactive cofactor forms, the chem-
ical energy of GTP hydrolysis is required for gating cofactor
transfer. The G protein chaperone also exerts its editing function
duringturnoverbyusingthebindingenergyofGTPtoelicitrelease
of inactive cofactor that is occasionally formed during the catalytic
cycle of MCM. The physiological relevance of this mechanism is
demonstrated by a patient mutation in methylmalonyl-CoA mu-
tase that does not impair the activity of this enzyme per se but
corrupts both the fidelity of the cofactor-loading process and the
ejection of inactive cofactor that forms occasionally during catal-
ysis. Consequently, cofactor in the incorrect oxidation state gains
access to the mutase active site and is not released if generated
during catalysis, leading, respectively, to assembly and accumula-
tion of inactive enzyme and resulting in methylmalonic aciduria.
cobalamin ? trafficking
C
zymes. With metalloproteins, intricate mechanisms are deployed
for receiving, elaborating [in the case of metal clusters (1)] and
delivering metal cofactors to target enzymes (2). Derivatives of
vitamin B12or cobalamin, are organometallic cofactors that are
used by only two mammalian enzymes (3, 4): the mitochondrial
methylmalonyl-CoA mutase (MCM) (5) and the cytoplasmic
methionine synthase (6). In circulation, cobalamin is associated
with a high-affinity binder, transcobalamin II (TCII) (7), which
is delivered to cells via a specific cell surface TCII receptor (8).
Following endocytosis into the acidic compartment of the lyso-
some, cobalamin is released, TCII is proteolyzed and the recep-
tor is recycled. It has been postulated (9, 10) that from its point
of egress into the cytoplasm, which is dependent on a lysosomal
membrane protein, LMBD1 (11), cobalamin is protein-bound
and escorted to the two B12-dependent enzymes found in
mammals. The strategy for escorted delivery of cobalamin to its
target enzymes would minimize adventitious reactions of a rare
and reactive cofactor while also ensuring specificity in a crowded
intracellular milieu. The cobalamin-trafficking pathway tailors
the inactive cofactor forms arriving in the cell into its biologically
active forms, methylcobalamin and 5?-deoxyadenosylcobalamin
(AdoCbl) needed by methionine synthase and MCM, respec-
tively, with the conversion of cob(II)alamin to methylcobalamin
occurring presumably on methionine synthase (10).
Following its arrival into the mitochondrion, the inactive
cob(II)alamin form of the cofactor is converted via the action of
adenosyltransferase (ATR) to the active AdoCbl form (12, 13).
ATR serves not only to tailor the active cofactor, but also to
deliver it to the target enzyme, MCM (14). ATR from Methyl-
obacterium extorquens AM1 is a trimeric enzyme that binds up
to two molecules of AdoCbl in an atypical ‘‘base-off’’ five-
coordinatestate(14,15)incontrasttoMCM,inwhichahistidine
ellular strategies for achieving fidelity during cofactor as-
sembly are needed to avert reconstitution of inactive en-
residue on the protein coordinates AdoCbl to give a base-off/
His-on six-coordinate state (Fig. 1). This difference in coordi-
nation number results in a large spectroscopic difference that is
useful for locating the cofactor. It is also exploited for direct
cofactor translocation during which the histidine ligand in MCM
plays a role in moving the cofactor from ATR to MCM (14). The
usage of only two of the three active sites in ATR is functionally
important for driving transfer of AdoCbl to MCM. Thus, ATR
operates via a rotary mechanism in which binding of ATP to the
vacant site in ATR leads to ejection of one equivalent of AdoCbl
from an adjacent active site (16). However, AdoCbl moves
reversibly between the active sites in this two-protein
ATR:MCM system with the equilibrium favoring the ATR site
(14), and insights into screening mechanisms that preclude
transfer of the inactive cob(II)alamin form have been lacking.
Isolated methylmalonic aciduria is a rare genetic disorder and
most commonly results from mutations in one of three loci: cblA,
cblB, and mut, encoding the mitochondrially localized proteins,
MMAA (for methylmalonic aciduria type A), ATR and MCM,
respectively (17). A G protein chaperone from M. extorquens
AM1, MeaB, which is orthologous to the human MMAA
protein, binds MCM with nanomolar affinity (18), suggesting
that the two proteins are likely to exist in a complex intracellu-
larly. MeaB exhibits low intrinsic GTPase activity, which is
enhanced approximately 100-fold in the MCM:MeaB complex.
Similarly, the presence of MeaB modulates MCM function,
protecting this radical enzyme from oxidative inactivation and
enhancing its turnover number approximately 2-fold (19).
In this study, we demonstrate that MeaB, a bacterial ortholog
of a mitochondrial protein involved in AdoCbl trafficking (20),
functionsasaneditor,discriminatingbetweeninactiveandactive
cofactor forms and permitting transfer only of AdoCbl, in a
process that is gated by GTP hydrolysis. The editing function of
MeaB is also used for rescuing inactivated mutase formed
occasionally during turnover. A patient mutation in MCM
resultsinmethylmalonicaciduriabycorruptingthefidelityofthe
cofactor loading process.
Results and Discussion
GTP-Dependent Gating by MeaB of AdoCbl Transfer from ATR to MCM.
We have previously shown that AdoCbl is transferred reversibly
between the active sites of ATR and MCM when the holo-form
of one protein is rapidly mixed with the apo-form of the other
(14). Remarkably, when AdoCbl-loaded ATR (holo-ATR) is
mixed with a stoichiometric complex of MCM:MeaB?GTP, no
cofactortransferisobserved(Fig.2A).Thisisinstrikingcontrast
to the ATP-independent transfer of cofactor observed previ-
ously in the absence of MeaB?GTP (14). Indeed, in the absence
Author contributions: D.P. and R.B. designed research; D.P. performed research; D.P. and
R.B. analyzed data; and D.P. and R.B. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
1To whom correspondence should be addressed. E-mail: rbanerje@umich.edu.
This article contains supporting information online at www.pnas.org/cgi/content/full/
0908106106/DCSupplemental.
www.pnas.org?cgi?doi?10.1073?pnas.0908106106PNAS ?
December 22, 2009 ?
vol. 106 ?
no. 51 ?
21567–21572
BIOCHEMISTRY

Page 2

of ATP, transfer of AdoCbl from ATR to MCM is inhibited at
increasing concentrations of MeaB until a stoichiometric com-
plex of MCM:MeaB?nucleotide is formed (Fig. 2B). Addition of
ATP drives the transfer of a single equivalent of AdoCbl from
ATR to MCM and is paralleled by a corresponding increase in
MCM activity under these conditions (Fig. 2A). Significantly,
when the same experiment is conducted in the presence of
GMPPNP, a non-hydrolysable analog of GTP, MCM activity is
undetectable and ATP triggers the release of AdoCbl from
holo-ATRintosolution.TheproductivetransferofAdoCblfrom
ATR is confirmed by enzyme-monitored turnover of MCM in
which approximately 30% of the cofactor exists as cob(II)alamin
under steady-state conditions when GTP but not GMPPNP is
used in the cofactor transfer experiment (Fig. 2C and Fig. S1).
These results demonstrate that MeaB utilizes the energy of GTP
hydrolysis to gate AdoCbl transfer from holo-ATR to MCM.
GTP hydrolysis is also used by UreG, for nickel delivery to
urease (21) and by HypB during maturation of the NiFe hydro-
genase (22). MeaB, UreG, and HypB all belong to the G3E
subfamily of P-loop GTPases (23).
Nucleotide-Dependent Editing of Cofactor Transfer from ATR to MCM.
Maturation of MCM appears to be governed by two nucleotide
switches: ATP binding to ATR and GTP hydrolysis by MeaB.
Next, we examined if either nucleotide switch is involved in
preventing transfer of the inactive cob(II)alamin form of the
cofactor from ATR to MCM. The M. extorquens ATR binds
cob(II)alamin in a mixture of the base-on and base-off states
(Fig. S2a). The base-off state found in ATR is further distin-
guished into five- or four-coordinate species in which the lower
axial position is occupied by a weak ligand such as water (-ATP)
or is vacant (?ATP), and yields EPR spectra that are similar to
those observed previously for human ATR (15, 24). In contrast,
MCM binds cob(II)alamin in a base-off/His-on five-coordinate
state in which the lower axial ligand is the imidazole side chain
of a histidine residue (25). A distinguishing feature of the EPR
spectra of cob(II)alamin with or without a nitrogenous axial
ligand is the appearance of triplet versus singlet structures
respectively in the octet of hyperfine lines that result from the
interactions between the unpaired electron and the cobalt
nucleus (I ? 7/2) (Fig. S2). When increasing concentrations of
MCM are incubated with ATR?cob(II)alamin, conversion of the
base-off to the base-off/His-on state is observed, consistent with
the transfer of the cofactor from ATR to MCM (Fig. S2b). The
transfer of cob(II)alamin is unaffected by ATP as evidenced by
the conversion of the four-coordinate base-off spectrum asso-
ciated with ATR to a five-coordinate base-off/His-on spectrum
associated with MCM (Fig. S2c). Under these conditions, co-
factor transfer is apparently unidirectional, as transfer from
holo-MCM to ATR was not detected ? ATP (Fig. S2d). ATP
does not induce release of cob(II)alamin from ATR in the
absence or presence of MCM:MeaB?GTP (Fig. S3). Importantly,
while MeaB or MeaB?GDP do not protect MCM from being
loaded with cob(II)alamin from ATR, MeaB?GTP or
MeaB?GMPPNP impede the process (Fig. 3A). Since GTP and
GMPPNP have similar effects, GTP binding rather than hydro-
lysis is used for cofactor editing by MeaB to shield MCM from
irreversible inactivation.
Fig. 1.
and their principal spectral features.
Schematic representation of AdoCbl conformations in ATR and MCM
Fig. 2.
transfer from holo-ATR (8 ?M final concentration) in the presence of 100 ?M
MCM:MeaB?GTP (F) or MCM:MeaB?GMPPNP (■) in Buffer A containing 15 mM
MgCl2at 20 °C. After each addition of ATP, MCM activity was assayed using the
radiolabeled assay (E,MCM:MeaB?GTP; ?, MCM:MeaB?GMPPNP). (B) MeaB in-
hibits ATP-dependent AdoCbl transfer from holo-ATR to wild-type but not to
R585C MCM. Cofactor transfer from 10–12 ?M holo-ATR to varying molar ratios
of[MeaB]/[MCM]?2mMnucleotides.Theamountofcofactortransferredfrom
holo-ATR to wild-type or R585C MCM in the absence of MeaB was 38–48%. (C)
Enzyme-monitored turnover of MCM loaded with AdoCbl from holo-ATR in the
presence of MeaB?nucleotide. The spectrum of 10 ?M holo-ATR in the presence
of a 10-fold molar excess of apo-MCM:MeaB?GTP (black) and after addition of 7
mM ATP (gray) and 7.2 mM (R and S)-methylmalonyl-CoA (dashed). The shift in
theabsorptionmaximafrom458nm(black)to525nm(gray)correspondstothe
conversionofbase-offAdoCbl?ATRtobase-off/His-onAdoCbl?MCM.Additionof
substrate results in the accumulation of cob(II)alamin (increase at 474 nm and a
decrease at 525 nm). (Inset) Quantification of cob(II)alamin present during turn-
over when MCM:MeaB?nucleotide (52–184 ?M, with 2.5 mM GTP or GMPPNP)
was reconstituted with holo-ATR (19–20 ?M in AdoCbl) (n ? 4).
Gating of coenzyme B12transfer. (A) ATP-dependent AdoCbl release/
21568 ?
www.pnas.org?cgi?doi?10.1073?pnas.0908106106Padovani and Banerjee

Page 3

Postloading Editing by MeaB. During catalytic turnover, cobalt-
carbon bond homolysis results in formation of a biradical
intermediate: cob(II)alamin and the 5?-deoxyadenosyl radical.
Oxidative interception of the cob(II)alamin intermediate in the
presence of MeaB occurs with a kobsof 2.7 ? 10?3min?1at 20 °C
leading to formation of the inactive aquocobalamin (H2OCbl)
species (19). H2OCbl remains tightly bound to MCM and only a
smallfractionisreleasedintheabsence(3.0?0.6%)orpresence
of MeaB (5.4 ? 0.4%) or MeaB?GTP (11.0 ? 1.8%) (Fig. S4).
In the absence of a mechanism for releasing inactive cofactor,
this would eventually lead to accumulation of completely inac-
tive MCM.
Since MeaB in the presence of nucleotides is ineffective at
releasing H2OCbl from MCM, we assessed whether it can induce
release of cob(II)alamin that has become uncoupled from the
organic radical intermediates that form during turnover. While
the EPR spectrum of MCM reconstituted with cob(II)alamin is
not perturbed by the presence of MeaB or MeaB?GDP (Fig. 4A,
spectra i and ii), the addition of MeaB?GTP or MeaB?GMPPNP
elicit subtle spectral changes in the S-shaped absorption feature
at g ?2.28 (Fig. 4A, spectra iii and iv). Thus, the hyperfine
structure associated with this feature when the cofactor is bound
to the active site is lost, as observed for cob(II)alamin in solution
(Fig. 4A, spectrum v). To confirm this observation,
MCM?cob(II)alamin was mixed with MeaB?nucleotides and the
free cofactor was separated using a Centricon concentrator.
Analysis of the filtrate by absorption spectroscopy confirmed
that cob(II)alamin remains associated with MCM in the pres-
enceofMeaB?GDP,whichisinstrikingcontrasttothesituation
with MeaB and either GTP or GMPPNP (Fig. 4B). Since GTP
and GMPPNP elicit similar effects, it suggests that the energy of
GTP binding rather than hydrolysis is used for cofactor editing
following inactivation during catalytic turnover.
The ability of MeaB?GTP to catalyze expulsion of cob(I-
I)alamin from MCM begs the question as to how this activity is
suppressed during turnover conditions when cob(II)alamin is a
constituent of the biradical intermediate and the other radical is
either cofactor- [i.e., 5?-deoxyadenosyl-], substrate-, or product-
derived. Other B12enzymes have mechanisms either for the in
siturepairofoxidizedcofactorasinmethioninesynthase(26,27)
or for the exchange of inactive cofactor for AdoCbl that is
mediated via an ATP-dependent chaperone as in diol dehy-
dratase (28, 29). We have previously shown that the presence of
MeaB?GDP or MeaB?GTP confers protection against oxidative
interception of cob(II)alamin decelerating the rate of OH2Cbl
formation by 4- and 19-fold, respectively (19). Under catalytic
turnover conditions when MCM reconstituted with AdoCbl is
mixed with methylmalonyl-CoA in the absence or presence of
MeaB ?nucleotides, we could not detect release of any cofactor
over an approximate 40-min time period. This is consistent with
the very slow rate of inactivation of MCM under turnover
conditions. Similarly, when MCM was reconstituted with cob(I-
I)alamin and 5?-deoxyadenosine (as a mimic of the 5?-
deoxyadenosyl radical), in the presence or absence of the
substrate, methylmalonyl-CoA, it was resistant to MeaB?GTP-
induced release of cob(II)alamin (Fig. 4C). These results imply
that MeaB safeguards the radical intermediates during turnover
conditions and that the occasional loss of the deoxyadenosine
moiety rather than oxidation of cob(II)alamin signals formation
of inactive MCM, triggering release of the cofactor by
MeaB?GTP. Interestingly, substitution of 5?-deoxyadenosine by
adenosine in these experiments leads to a GTP hydrolysis
dependence of cob(II)alamin release from MCM (Fig. 4C),
indicating transmission of information about ligand-specific
interactions between the 5?-deoxyadenosine binding site in
MCM and MeaB.
Fig.3.
monitoring of cob(II)alamin transfer from ATR (40 ?M in cob(II)alamin, A and
B, i) to 1.6 equivalent MCM (A, ii) or 1.2 equivalent R585C MCM (B, ii) in the
presence of MeaB (A and B, ii), or MeaB with 2.5 mM GDP (A and B, iii), GTP
(A and B, iv), or GMPPNP (A and B, v). For the mutant, a molar ratio [MeaB]/
[R585C MCM] of 2.1 was used in comparison to a stoichiometric ratio for
wild-type MCM. The EPR spectra are the average of two scans.
EditingfunctionofMeaBduringcofactorassembly.EPRspectroscopic
Fig. 4.
bound to MCM in the presence of 1.5 molar excess of MeaB (i), MeaB?GDP (ii),
MeaB?GTP (iii) or MeaB?GMPPNP (iv), and free cob(II)alamin (v) in Buffer A. The
asterisks point to the hyperfine structures in the absorption feature when cob(I-
I)alamin is bound to MCM but not when it is in solution. (B) Comparison of the
percentage of cofactor remaining bound to MCM following addition of 1.5 (to
wild-type MCM) or 2.5 (to R585C MCM) molar excess of MeaB ? the indicated
nucleotides. (C) Comparison of the percentage of cofactor remaining bound to
MCM in the presence of 7.8 mM adenosine or 9 mM 5?-deoxyadenosine, and 14
mM(RandS)-methylmalonyl-CoAanda1.5molarexcessofMeaB?theindicated
nucleotides.
MeaB rescues inactive MCM. (A) EPR spectra of 50 ?M cob(II)alamin
Padovani and BanerjeePNAS ?
December 22, 2009 ?
vol. 106 ?
no. 51 ?
21569
BIOCHEMISTRY

Page 4

A Patient Mutation in MCM Corrupts the Editing Function of MeaB.
The physiological relevance of the cofactor-gating/editing model
is demonstrated by the R585C mutation in the M. extorquens
MCM, which was designed to mimic the pathogenic mutation,
R616C, identified in a patient with methylmalonic aciduria (30).
Homozygosity of this mutation results in an approximately 86%
reduction in intracellular AdoCbl levels (30). The arginine
residue resides at a solvent exposed edge of the B12 binding
domain and is distant from the cofactor binding and active sites
(Fig. 5A) and is involved in a network of hydrogen bonding
interactions with neighboring residues (Fig. S5). When the
kinetic properties of the mutant were evaluated in assays in
whichMCMisreconstitutedwithexogenousAdoCbl,noobvious
deficits were seen (Table 1). Similarly, in assays in which AdoCbl
was transferred from holo-ATR to the mutant, the cofactor
transfer efficiency was comparable to that of wild-type MCM
(Fig. S6). However, a profound phenotype was evident when the
interaction of R585C and its GTPase activating protein (GAP)
activity were interrogated in the presence of MeaB. The muta-
tion decreases the affinity of MCM for MeaB?GDP and
MeaB?GTP by approximately 80- and 9-fold respectively and
reduces the kcat/Kmfor GAP activity 280-fold (Table 1 and Fig.
5B).The diminishedaffinity
MeaB?nucleotide is accompanied by a large reduction in the
enthalpy of binding compared to wild-type MCM (??H ?
20.8–22.5 kcal/mol) (Fig. S7) and suggests a role for the arginine
residue in stabilizing the MCM:MeaB complex. GAPs often
furnish an arginine ‘‘finger’’ that stabilizes the transition state in
theGproteinactivesite(31).WespeculatethattheR585residue
might play this role in enhancing the GTPase activity of MeaB
or could modulate it indirectly.
The functional deficiencies in the R585C mutant were further
evident when holo-ATR was used as the cofactor source in the
presence of MeaB (Fig. 2B). In marked contrast to wild-type
MCM, the presence of MeaB stimulated the ATP-independent
ofR585C MCM for
transfer of AdoCbl from holo-ATR to R585C MCM. Further-
more, the mutation compromises the GTP-dependent editing
function of MeaB, resulting in its failure to safeguard the mutant
MCM from reconstitution with cob(II)alamin in the presence of
GTP or GMPPNP (Fig. 3B). The R585C mutation also impairs
release of inactive cofactor from MCM?cob(II)alamin in the
presence or absence of nucleotides (Fig. 4B). This loss of
function cannot simply be ascribed to the reduced binding
affinity between R585C MCM and MeaB, since the experiments
were performed under conditions where ?96% of the mutant
was present in complex with MeaB. Rather, the mutation
appears to impair the ability of the MCM:MeaB complex to
adopt and/or stabilize the editing conformation induced by GTP
Fig. 5.
R616 in the human and R585 in the M. extorquens sequences) in the crystal structure P. shermanii MCM (PDB: 1REQ). AdoCbl (red) and the arginine residue of
interest (navy) are shown in ball representation. The two subunits of MCM are differentiated by shades of blue. (B) Dependence on GTP concentration of the
rate of GTP hydrolysis by MeaB (?) MeaB:MCM (F) or MeaB:R585C MCM (?) in Buffer A at 37 °C. Michaelis-Menten analysis of the rate of GTP hydrolysis yielded
the kinetic parameters reported in Table 1. (C) Schematic model for the gating and editing functions of MeaB.
Model for gating and editing of cofactor transfer and its corruption by a pathogenic mutation. (A) Location of the arginine residue (corresponding to
Table 1. Comparison of the kinetic and thermodynamic
parameters and the GAP activity of wild-type and R585C MCM
Kinetic parameterWild-type MCMR585C MCM
kcat, s?1*
KM (R)-M-CoA, ?M*
kcat/KM, M?1s?1*
Kact AdoCbl, ?M*
Kd AdoCbl, ?M†
Kd MeaB?GDP, ?M†,‡
Kd MeaB?GMPPNP, ?M†,‡
kcat, min?1§
KM GTP, ?M§
kcat/KM, M?1min?1§
132 ? 16
86 ? 13
(1.5 ? 0.2) ? 106
1.6 ? 0.1
1.7 ? 0.3
0.11 ? 0.02
0.16 ? 0.02
9.5 ? 0.8
91 ? 11
(1.04 ? 0.04) ? 105
138 ? 5
93 ? 14
(1.5 ? 0.1) ? 106
2.4 ? 0.7
1.7 ? 0.06
8.6 ? 1.7
1.4 ? 0.4
0.2 ? 0.01
526 ? 82
376 ? 38
The values represent the average of at least two independent experiments
(SD). *Determined in the radiolabeled assay at 37°C. The data for wild-type
MCMisfromref.19.†DeterminedbyITC.‡Thedataforwild-typeMCMisfrom
ref. 18.§GAP activity of MCM was determined as described in Methods. The
correspondingparametersforMeaBare:kcat?0.050.01min?1,KMGTP?15,020
?M and kcat/KM? 31,316 M?1min?1(18).
21570 ?
www.pnas.org?cgi?doi?10.1073?pnas.0908106106 Padovani and Banerjee

Page 5

binding. This loss of editing function can explain the biochemical
phenotype of the patient cell line (30) since capture of cob(I-
I)alamin by the mutant would lead to inactive MCM, with
consequent reduction in propionate incorporation, and dimin-
ished AdoCbl levels as observed in patient fibroblasts (30) due
to depletion of the precursor cob(II)alamin.
A Model for MeaB-Dependent Cofactor Editing and Gating. Based on
the results from this study, we propose a model for AdoCbl
delivery to MCM (Fig. 5C). Cob(II)alamin delivered to the
mitochondrion is received by ATR by a transport mechanism
that awaits elucidation. The nearly equivalent affinity of MeaB
for GTP versus GDP (18) implies that the active site of this
protein is predominantly loaded with GTP, which is much more
abundant in the cell. The conformation of the idle
MCM:MMAA?GTP complex prevents hijacking of the entering
cob(II)alamin, which is instead converted by ATR to AdoCbl.
The latter is directly transferred from ATR to the
MCM:MMAA?GTP complex in a process that is gated by two
nucleotide-dependent switches: ATP binding to ATR and GTP
hydrolysis. During turnover of MCM, occasional loss of the
deoxyadenosine moiety leads to enzyme inactivation as AdoCbl
cannot be regenerated at the end of the catalytic cycle. This in
turn, elicits a postloading editing function of MMAA in which
the GTP binding energy is used to trigger release of inactive
cob(II)alamin from the active site of MCM. Pathogenic muta-
tions can potentially interfere with any one of the finely orches-
trated steps that govern the fidelity of MCM maturation and lead
to disease. Based on our studies on the R585C mutation in the
M. extorquens MCM, we predict that the R616C mutation in
human MCM has pleiotropic consequences on the integrity of
the gating/editing functions of the MCM:MMAA complex and
leads to cob(II)alamin capture by and inactivity of MCM and to
reduced levels of AdoCbl. The responsiveness of the WG3118
patient cell line harboring the R616C-carrying allele to OHCbl
supplementation as assessed by the propionate incorporation
assay (30), might be explained by the increased availability of
cob(II)alamin for AdoCbl synthesis by ATR versus cob(I-
I)alamin hijacking by the MCM:MMAA complex. Interestingly,
all patients with mutations in MMAA exhibit in addition to
reduced MCM activity, diminished intracellular AdoCbl levels
(32), consistent with competition for the limiting mitochondrial
cob(II)alamin supply by MCM. Genetic disruption of the meaB
gene in M. extorquens leads to complete loss of MCM activity
consistent with the presence of the inactive cob(II)alamin-
loaded enzyme (20). However, this organism can biosynthesize
B12(33) and AdoCbl levels are unaffected in the ?meaB strain,
presumably because cob(II)alamin is not limiting.
In summary, our studies reveal an unprecedented role for a G
protein trafficking chaperone that uses the binding energy
derived from GTP for editing and the chemical energy released
from GTP hydrolysis for gating transfer of AdoCbl from its site
of synthesis to its site of utilization. These studies furnish a
mechanistic framework for understanding how pathogenic mu-
tations in the mitochondrial trafficking pathway might corrupt
this process leading to methylmalonic aciduria.
Materials and Methods
Materials. AdoCbl, OHCbl, nucleotides and other reagent grade chemicals
were purchased from Sigma. [14C]-methylmalonyl-CoA (56 Ci/mol) was pur-
chased from New England Nuclear and methylmalonyl-CoA was synthesized
as described (34).
ConstructionoftheR585CMutant.PlasmidsexpressingtheM.extorquensAM1
MCM, MeaB, and ATR were generous gifts from Mary Lidstrom (University of
Washington, Seattle). The R585C mutant of MCM was created using the
QuikChange kit (Stratagene) and the following sense primer: 5?-CAAGAT-
GGGCCAGGACGGGAACGACCGCGGCCAGAAGGTG-3?. The antisense muta-
genic primer had the complementary sequence. Following PCR amplification,
the mutation was confirmed by nucleotide sequence determination at the
Genomics Core Facility (University of Nebraska-Lincoln).
Enzyme Assays. The wild-type and R585C MCM, MeaB and ATR were purified
as described (14, 18). The specific activity of MCM was determined in the
radiolabeled assay as described (19). The GTPase activity of MeaB was deter-
mined by HPLC as described (18). Each experiment was performed at least in
duplicate.
Kinetic parameters for wild-type and R585C MCM. The kinetic parameters for
R585C MCM were determined at 37 °C in the radiolabeled assay (19). KMand
Vmaxwere determined in the presence of 100 ?M AdoCbl and varying con-
centrations of (R,S)-[14C]-methylmalonyl-CoA (80–4,000 ?M). The Kact for
AdoCbl was obtained by varying the concentration of the cofactor (1–50 ?M)
and a saturating concentration of the substrate {4 mM (R,S)-[14C]-methylma-
lonyl-CoA}.
MCM activity after cofactor transfer in the presence of MeaB and nucleotides. To
avoid cofactor dilution and release in solution during the experiments, the
concentration of enzymes in the assay was increased approximately 1,000-
fold. As a consequence, after the AdoCbl transfer experiments, the activity
measurements were performed on ice to be under initial velocity conditions
intheradiolabeledMCMassay.AdoCbltransferwasaccomplishedat20 °Cfor
5 min by adding various concentrations of ATP (0–10 mM) to a reaction
mixture containing 8 ?M holo-ATR (16 ?M bound AdoCbl) preincubated with
MCM (100 ?M) in complex with MeaB (110 ?M) and 2 mM GTP or 2 mM
GMPPNPinBufferA(50mMHEPESbuffer,pH8.0,containing0.3MKCl,15mM
MgCl2,and5%glycerol).Then,thesolutionswereincubatedfor20minonice
before starting the assay by addition of 6 mM ice-cold (R,S)-[14C]-methylma-
lonyl-CoA. After 1 min incubation, the reaction mixtures were quenched and
the samples treated as described (14). Control experiments lacking the com-
plex (negative control) or MeaB (positive control) were performed in parallel.
A specific activity of 0.053 ? 0.007 ?mol min?1mg?1on ice was measured in
the absence of MeaB and denotes 100% activity.
Modulation of MeaB activity by MCM. The effect of MCM on the GTPase activity
of MeaB was determined at 37 °C by incubating various concentrations of
wild-typeorR585CMCM(0–100?M)inthepresenceofMeaB(10?M)andGTP
(5 mM) in Buffer A. At various time points, 100-?L aliquots were quenched
with 20 ?L of 2M tricholoroacetic acid. The samples were then analyzed as
described(18).TheKMandVmaxweredeterminedinthepresenceofwild-type
MCM (2 ?M MeaB, 2 ?M MCM) or R585C MCM (10 ?M MeaB, 100 ?M R585C)
and varying concentrations of GTP (0.05–5 mM) at 37 °C in Buffer A.
Cofactor Transfer Assays. Cofactor transfer between ATR and R585C MCM. The
transfer of AdoCbl between the two enzyme active sites was monitored in
BufferAinthepresenceof0–8.3mMATPbyabsorptionspectroscopyat20 °C
as described (16).
Requirement of GTP hydrolysis for cofactor transfer between Holo-ATR and
MCM:MeaB?GTP. Holo-ATR (9.5–10 ?M) was preincubated for 5 min with 51.6–
184 ?M MCM:MeaB?GTP/GMPPNP (2.5 mM nucleotide) in Buffer A at 20 °C.
Cofactor transfer/release was initiated by addition of 6 mM ATP and the
amount of cofactor transferred to MCM or released into solution was esti-
matedusinga??525 nm?7.75mM?1cm?1(transfer)ora??525 nm?6.69mM?1
cm?1(release), respectively. After 10 min incubation, 7.2 mM (R,S)-
methylmalonyl-CoA was added to the reaction mixture and the amount of
cob(II)alamin generated under steady-state conditions was estimated using a
??525 nm?-4.8mM?1cm?1.Analysisofthefiltrateasdescribed(16)confirmed
thepresenceofAdoCblinsolutionwhenGMPPNP(51.1?2.7%,n?4)butnot
GTP (4.1 ? 2.9%, n ? 4) was used.
ATP dependence of AdoCbl transfer from ATR to wild-type or R585C MCM. The ATP
dependence of cofactor transfer was investigated by incubating holo-ATR
(8.5–11 ?M) with MCM (30–37 ?M) in the presence of increasing concentra-
tions of MeaB (18.8–357 ?M), MeaB?GTP (7–180 ?M), or MeaB?GDP (9.1–137
?M) in Buffer A at 20 °C. The data were acquired 5 min after mixing. The
amountofAdoCbltransferredwasestimatedfromtheincreaseinabsorbance
at 525 nm (??525 nm? 7.75 mM?1cm?1). The same experiments were per-
formed with R585C MCM (27.5–29.5 ?M) in the presence of increasing con-
centrations of MeaB (5–298 ?M), MeaB?GTP (5–296 ?M), or MeaB?GDP (5–296
?M). When used, nucleotides were 2.5 mM.
Cofactor Release Assays. MeaB-dependent release of cob(II)alamin from MCM.
Wild-typeorR585CMCM(60–70?M)wasincubatedfor10minwith45–50?M
cob(II)alamin in Buffer A under strictly anaerobic conditions. Then, a 1.5
(MCM)or2.5(R585CMCM)molarexcessofapo-MeaB,MeaB?GDP,MeaB?GTP,
or MeaB?GMPPNP (2–5 mM nucleotides) in Buffer A was added to the MCM
solution, incubated for 15 min and frozen for EPR spectroscopy. Following
accumulation of the EPR spectra, the samples were thawed, and subjected to
Padovani and BanerjeePNAS ?
December 22, 2009 ?
vol. 106 ?
no. 51 ?
21571
BIOCHEMISTRY

Page 6

air oxidation for 7 h at room temperature. During this time, the MCM:MeaB
complex is stable. The reaction mixtures were then filtered through a Centri-
conYM30concentrator(5,000rpm,45min,4 °C)andthefiltratewasanalyzed
by UV-visible spectroscopy to estimate the amount of cofactor released from
the mutase, ?356 nm? 20.6 mM?1cm?1for OH2Cbl.
Toassesstheeffectofoccupancyatthedeoxyadenosinebindingsiteonthe
releaseofcob(II)alaminfromMCM,thesameexperimentswererepeatedwith
MCM (50–60 ?M) preincubated for 15 min with 7.5–10 mM 5?-deoxyade-
nosine or adenosine ? 6–9 mM (R)-methylmalonyl-CoA before addition of
45–50 ?M cob(II)alamin. Then, a 1.5 molar excess of MeaB, MeaB?GDP,
MeaB?GTP,orMeaB?GMPPNPinBufferAwasaddedtothemutasesampleand
processed as described above. Each experiment was performed at least in
duplicate.
Analysis of AdoCbl release from MCM under catalytic turnover conditions in the
presence of MeaB ? nucleotide. MCM (14–55 ?M cofactor bound, ?525 nm? 9.06
mM?1cm?1) was incubated at 20 °C for 15 min in Buffer A with a 1.5 molar
excess of apo-MeaB, MeaB?GDP, MeaB?GTP, or MeaB?GMPPNP ? 5–9 mM
(R)-methylmalonyl-CoA. The order of addition of methylmalonyl-CoA and
MeaB did not affect the results. The amount of cofactor released from MCM
was quantified using a centricon YM30 filter and analyzing the filtrate by
absorption spectroscopy using ?525 nm? 8 mM?1cm?1. The experiments were
performed at least in quadruplicate.
Isothermal Titration Calorimetry (ITC). ITC experiments were performed to
study(i)bindingofAdoCbltowild-typeorR585CMCMand(ii)theinteraction
between R585C MCM with MeaB ? nucleotides as described (18, 19). Each
experiment was performed at least in triplicate in Buffer A at 20 °C.
EPR Spectroscopy. EPR spectra were recorded on a Bruker EMX spectrometer
(BrukerBiospinCorp.),equippedwithanOxfordITC4temperaturecontroller,
a Hewlett-Packard model 5340 automatic frequency counter, and Bruker
gaussmeter. Unless otherwise noted, the EPR spectroscopic parameters in-
cluded the following: temperature, 100 K; microwave power, 25 mW; micro-
wave frequency, 9.38 GHz; receiver gain, 2 ? 105; modulation amplitude, 10
G; modulation frequency, 100 kHz.
ACKNOWLEDGMENTS.WethankTanyaLabunska(UniversityofNebraska)for
generating the R585C mutant of MCM and Markos Koutmos (University of
Michigan) for generating Fig. S5. This work was supported in part by the
National Institutes of Health (DK45776).
1. Lill R, Muhlenhoff U (2006) Iron-sulfur protein biogenesis in eukaryotes: Components
and mechanisms. Annu Rev Cell Dev Biol 22:457–486.
2. Finney LA, O’Halloran TV (2003) Transition metal speciation in the cell: Insights from
the chemistry of metal ion receptors. Science 300:931–936.
3. Kolhouse JF, Allen RH (1977) Recognition of two intracellular cobalamin binding
proteins and their recognition as methylmalonyl-CoA mutase and methionine syn-
thetase. Proc Natl Acad Sci USA 74:921–925.
4. Mellman IS, Youngdahl-Turner P, Huntington FW, Rosenberg LE (1977) Intracellular
binding of radioactive hydroxocobalamin to cobalamin-dependent apoenzyme in rat
liver. Proc Natl Acad Sci USA 74:916–920.
5. Banerjee R (2003) Radical carbon skeleton rearrangements: Catalysis by coenzyme
B12-dependent mutases. Chem Rev 103:2083–2094.
6. MatthewsRG(2001)Cobalamin-dependentmethyltransferases.AccChemRes34:681–
689.
7. FedosovSN,BerglundL,FedosovaNU,NexoE,PetersenTE(2002)Comparativeanalysis
of cobalamin binding kinetics and ligand protection for intrinsic factor, transcobal-
amin, and haptocorrin. J Biol Chem 277:9989–9996.
8. Quadros EV, Nakayama Y, Sequeira JM (2009) The protein and the gene encoding the
receptorforthecellularuptakeoftranscobalamin-boundcobalamin.Blood113:186–192.
9. Banerjee R (2006) B12 trafficking in Mammals: A case for coenzyme escort service. ACS
Chem Biol 1:149–159.
10. BanerjeeR,GherasimC,PadovaniD(2009)TheTinker,tailor,soldierinintracellularB12
trafficking. Current Op Chem Biol. 13:477–484.
11. RutschF,etal.(2009)Identificationofaputativelysosomalcobalaminexporteraltered
in the cblF defect of vitamin B12 metabolism. Nat Genet 41:234–239.
12. Dobson CM, et al. (2002) Identification of the gene responsible for the cblB comple-
mentation group of vitamin B12-dependent methylmalonic aciduria. Hum Mol Genet
11:3361–3369.
13. Leal NA, Park SD, Kima PE, Bobik TA (2003) Identification of the human and bovine
ATP:Cob(I)alamin adenosyltransferase cDNAs based on complementation of a bacte-
rial mutant. J Biol Chem 278:9227–9234.
14. Padovani D, Labunska T, Palfey BA, Ballou DP, Banerjee R (2008) Adenosyltransferase
tailors and delivers coenzyme B12. Nat Chem Biol 4:194–196.
15. Yamanishi M, Labunska T, Banerjee R (2005) Mirror ‘‘base-off’’ conformation of
coenzyme B12in human adenosyltransferase and its downstream target, methylma-
lonyl-CoA mutase. J Am Chem Soc 127:526–527.
16. Padovani D, Banerjee R (2009) A rotary mechanism for Coenzyme B12 synthesis by
adenosyltransferase. Biochemistry 48:5350–5357.
17. Shevell MI, Rosenblatt DS (1992) The neurology of cobalamin. Can J Neurol Sci
19:472–486.
18. Padovani D, Labunska T, Banerjee R (2006) Energetics of interaction between the
G-protein chaperone, MeaB and B12-dependent methylmalonyl-CoA mutase. J Biol
Chem 281:17838–17844.
19. Padovani D, Banerjee R (2006) Assembly and protection of the radical enzyme, meth-
ylmalonyl-CoA mutase, by its chaperone. Biochemistry 45:9300–9306.
20. Korotkova N, Lidstrom ME (2004) MeaB is a component of the methylmalonyl-CoA
mutase complex required for protection of the enzyme from inactivation. J Biol Chem
279:13652–13658.
21. Soriano A, Hausinger RP (1999) GTP-dependent activation of urease apoprotein in
complex with the UreD, UreF, and UreG accessory proteins. Proc Natl Acad Sci USA
96:11140–11144.
22. Maier T, Jacobi A, Sauter M, Bock A (1993) The product of the hypB gene, which is
required for nickel incorporation into hydrogenases, is a novel guanine nucleotide-
binding protein. J Bacteriol 175:630–635.
23. Leipe DD, Wolf YI, Koonin EV, Aravind L (2002) Classification and evolution of P-loop
GTPases and related ATPases. J Mol Biol 317:41–72.
24. Stich TA, Yamanishi M, Banerjee R, Brunold TC (2005) Spectroscopic evidence for the
formation of a four-coordinate Co2?cobalamin species upon binding to the human
ATP:cobalamin adenosyltransferase. J Am Chem Soc 127:7660–7661.
25. Padmakumar R, Padmakumar R, Banerjee R (1997) Evidence that cobalt-carbon bond
homolysis is coupled to hydrogen atom abstraction from substrate in methylmalonyl-
CoA mutase. Biochemistry 36:3713–3718.
26. Fujii K, Huennekens FM (1974) Activation of methionine synthase by a reduced
triphosphopyridinenucleotide-dependentflavoproteinsystem.JBiolChem249:6745–
6753.
27. Banerjee RV, Harder S, R., Ragsdale SW, Matthews RG (1990) Mechanism of reductive
activation of cobalamin-dependent methionine synthase: An electron paramagnetic
resonance spectroelectrochemical study. Biochemistry 29:1129–1137.
28. Honda S, Toraya T, Fukui S (1980) In situ reactivation of glycerol-inactivated coenzyme
B12-dependent enzymes, glycerol dehydratase and diol dehydratase. J Bacteriol
143:1458–1465.
29. Toraya T, Mori K (1999) A reactivating factor for coenzyme B12-dependent diol dehy-
dratase. J Biol Chem 274:3372–3377.
30. Worgan LC, et al. (2006) Spectrum of mutations in mut methylmalonic acidemia and
identification of a common Hispanic mutation and haplotype. Hum Mutat 27:31–43.
31. Scheffzek K, et al. (1997) The Ras-RasGAP complex: Structural basis for GTPase activa-
tion and its loss in oncogenic Ras mutants. Science 277:333–338.
32. Lerner-Ellis JP, et al. (2004) Mutations in the MMAA gene in patients with the cblA
disorder of vitamin B12metabolism. Hum Mutat 24:509–516.
33. Ivanova EG, Federov DN, Doronina NV, Trotensko YA (2006) Production of vitamin B12
in aerobic methylotrophic bacteria. Microbiology 75:494–496.
34. Padovani D, Banerjee R (2006) Alternative pathways for radical dissipation in an active
site mutant of B12-dependent methylmalonyl-CoA mutase. Biochemistry 45:2951–
2959.
21572 ?
www.pnas.org?cgi?doi?10.1073?pnas.0908106106 Padovani and Banerjee