Group Title: BMC Genomics
Title: A Subset of the diverse COG0523 family of putative metal chaperones is linked to zinc homeostasis in all kingdoms of life
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00099912/00001
 Material Information
Title: A Subset of the diverse COG0523 family of putative metal chaperones is linked to zinc homeostasis in all kingdoms of life
Physical Description: Book
Language: English
Creator: Haas, Crysten
Rodionov,Dmitry
Kropat, Janette
Malasarn, Davin
Merchant, Sabeeha
de Crecy-Lagard,Valerie
Publisher: BMC Genomics
Publication Date: 2009
 Notes
Abstract: BACKGROUND:COG0523 proteins are, like the nickel chaperones of the UreG family, part of the G3E family of GTPases linking them to metallocenter biosynthesis. Even though the first COG0523-encoding gene, cobW, was identified almost 20 years ago, little is known concerning the function of other members belonging to this ubiquitous family.RESULTS:Based on a combination of comparative genomics, literature and phylogenetic analyses and experimental validations, the COG0523 family can be separated into at least fifteen subgroups. The CobW subgroup involved in cobalamin synthesis represents only one small sub-fraction of the family. Another, larger subgroup, is suggested to play a predominant role in the response to zinc limitation based on the presence of the corresponding COG0523-encoding genes downstream from putative Zur binding sites in many bacterial genomes. Zur binding sites in these genomes are also associated with candidate zinc-independent paralogs of zinc-dependent enzymes. Finally, the potential role of COG0523 in zinc homeostasis is not limited to Bacteria. We have predicted a link between COG0523 and regulation by zinc in Archaea and show that two COG0523 genes are induced upon zinc depletion in a eukaryotic reference organism, Chlamydomonas reinhardtii.CONCLUSION:This work lays the foundation for the pursuit by experimental methods of the specific role of COG0523 members in metal trafficking. Based on phylogeny and comparative genomics, both the metal specificity and the protein target(s) might vary from one COG0523 subgroup to another. Additionally, Zur-dependent expression of COG0523 and putative paralogs of zinc-dependent proteins may represent a mechanism for hierarchal zinc distribution and zinc sparing in the face of inadequate zinc nutrition.
General Note: Periodical Abbreviation: BMC Genomics
General Note: Start page 470
General Note: M3: 10.1186/1471-2164-10-470
 Record Information
Bibliographic ID: UF00099912
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: Open Access: http://www.biomedcentral.com/info/about/openaccess/
Resource Identifier: issn - 1471-2164
http://www.biomedcentral.com/1471-2164/10/470

Downloads

This item has the following downloads:

PDF ( 2 MBs ) ( PDF )


Full Text


r)4


BMC Genomics BioMed



Research article

A subset of the diverse COG0523 family of putative metal
chaperones is linked to zinc homeostasis in all kingdoms of life
Crysten E Haas', Dmitry A Rodionov2,3, Janette Kropat4, Davin Malasarn4,
Sabeeha S Merchant4 and Valerie de Crecy-Lagard*1


Address: 'Department of Microbiology and Cell Science, University of Florida, Gainesville, FL, USA, 2Bumham Institute for Medical Research, La
Jolla, CA, USA, 3Institute for Information Transmission Problems (the Kharkevich Institute), RAS, Moscow, Russia and 4Department of Chemistry
and Biochemistry and Institute for Genomics and Proteomics, University of California at Los Angeles, Los Angeles, CA, USA
Email: Crysten E Haas crys66@ufl.edu; Dmitry A Rodionov rodionov@burnham.org; Janette Kropat- kropat@chem.ucla.edu;
Davin Malasarn dmalasarn@gmail.com; Sabeeha S Merchant merchant@chem.ucla.edu; Valerie de Crecy-Lagard* vcrecy@ufl.edu
* Corresponding author


Published: 12 October 2009
BMC Genomics 2009, 10:470 doi: 10.1 186/1471-2164-10-470


Received: 25 June 2009
Accepted: 12 October 2009


This article is available from: http://www.biomedcentral.com/1471-2164/10/470
2009 Haas et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.



Abstract
Background: COG0523 proteins are, like the nickel chaperones of the UreG family, part of the
G3E family of GTPases linking them to metallocenter biosynthesis. Even though the first COG0523-
encoding gene, cobW, was identified almost 20 years ago, little is known concerning the function of
other members belonging to this ubiquitous family.
Results: Based on a combination of comparative genomics, literature and phylogenetic analyses
and experimental validations, the COG0523 family can be separated into at least fifteen subgroups.
The CobW subgroup involved in cobalamin synthesis represents only one small sub-fraction of the
family. Another, larger subgroup, is suggested to play a predominant role in the response to zinc
limitation based on the presence of the corresponding COG0523-encoding genes downstream
from putative Zur binding sites in many bacterial genomes. Zur binding sites in these genomes are
also associated with candidate zinc-independent paralogs of zinc-dependent enzymes. Finally, the
potential role of COG0523 in zinc homeostasis is not limited to Bacteria. We have predicted a link
between COG0523 and regulation by zinc in Archaea and show that two COG0523 genes are
induced upon zinc depletion in a eukaryotic reference organism, Chlamydomonas reinhardtii.
Conclusion: This work lays the foundation for the pursuit by experimental methods of the specific
role of COG0523 members in metal trafficking. Based on phylogeny and comparative genomics,
both the metal specificity and the protein targets) might vary from one COG0523 subgroup to
another. Additionally, Zur-dependent expression of COG0523 and putative paralogs of zinc-
dependent proteins may represent a mechanism for hierarchal zinc distribution and zinc sparing in
the face of inadequate zinc nutrition.


Background
Transition metals perform vital roles in many chemical
reactions essential for life. A recent bioinformatic
approach suggests Zn-, non-heme Fe- and Cu-proteins


constitute 10% of bacterial and eukaryotic proteomes and
13% of archaeal proteomes [1-3]. The roles of these met-
als can be varied. In some oxidoreductases, for instance,
iron and copper are exploited for their ability to accept or


Page 1 of 21
(page number not for citation purposes)


central


ess







BMC Genomics 2009, 10:470


donate electrons, while in hemoglobin and hemocyanin,
these metals are used for oxygen transport [4-61. Zinc, on
the other hand, serves as an electrophile or Lewis acid in
many protein-catalyzed reactions. The activity of metallo-
proteins (many of which are essential proteins) is, conse-
quently, strictly dependent on the presence of a metal and
in most cases of a specific metal. Ensuring proper metal
allocation is therefore critical for survival.

It was initially assumed that free pools of metal ions were
available within the cell, such that a nascent polypeptide
would acquire its cofactor solely through the metal-affin-
ity of the chelating ligands. As discussed in several recent
reviews, this picture of metal metabolism was oversimpli-
fied [7,81, as: i) the ions are chelated intracellularly by
proteins and small molecule ligands, and ii) metal-bind-
ing ligands are not sufficiently selective to ensure that the
proper cofactor is loaded. Since the discovery of the first
copper metallochaperone, Atx 1[91, numerous protein fac-
tors involved in metallocenter biosynthesis have been
characterized. The mechanisms by which the cell ensures
the correct metal ions are loaded into metalloproteins are
just beginning to be understood.

Studies involving the maturation of Ni-urease and Ni-Fe
hydrogenase have provided the most extensive picture of
metallocenter biosynthesis (for a review see [101 and
[11]). These two nickel-containing proteins require a suite
of accessory proteins to properly insert Ni into the cata-
lytic site (only one exception has been found to date;
Bacillus subtilis encodes a functional urease in the absence
of the canonical accessory proteins [12]). In both cases, a
GTPase (UreG for urease or HypB for hydrogenase) is
involved in the incorporation of the Ni cofactor. These
two proteins belong to the G3E family of P-loop GTPases
as defined by Leipe and colleagues [13]. Other members
of this family include MeaB (ArgK), required for the acti-
vation of methylmalonyl-CoA mutase (a B12-dependent
enzyme) [14], and COG0523, a large and diverse sub-
family of proteins with poorly defined functions.

COG0523 proteins occur in all kingdoms of life, and most
sequenced genomes encode one or more homologs. The
first member of the COG0523 family was identified as
being involved in cobalamin biosynthesis in Pseudomonas
denitrificans and hence named CobW [15]. Other mem-
bers of COG0523 include the nitrile hydratase activator,
which is required for Fe-type nitrile hydratase activity
[16], and YciC of Bacillus subtilis. Due to repression by Zur,
a zinc-responsive transcription factor, yciC was originally
assumed to code for a low-affinity zinc transporter [17-
19]. Without the means to automatically distinguish
between these different functions, these annotations have
been propagated amongst all members of the family in
sequenced genomes. Therefore, as a result of these studies,


http://www. biomedcentral.com/1471-2164/10/470



genes encoding a COG0523 protein have been automati-
cally and arbitrarily assigned either a function in cobala-
min biosynthesis, in the activation of nitrile hydratase, or
as a low-affinity zinc transporter. Nevertheless, we note
that each of these functions is related in the general sense
to intracellular metal delivery. The diversity of the metals
putatively handled by COG0523, Ni, Fe, or Zn, suggests
that there might be different sub-groups identifiable
within the COG0523 superfamily.

The COG0523 family is a striking example of systematic,
homology-based mis-annotation. Although members are
frequently annotated as having specific functions, these
'functions' are based only on homology to a few family
members and are therefore suspect. The simplest way to
annotate a genome is based on sequence homology to
characterized genes. Sequence homology does not neces-
sarily equate to functional identity or even similarity.
Therefore, this approach to annotation is frequently inad-
equate as exemplified in the literature [20-22] and illus-
trated by the development of alternative paradigms for
functional annotation [23-26].

To provide an improved annotation for the various mem-
bers of this family and gain insight into the role members
of this protein family may perform, we conducted an
extensive comparative genomic analysis of the G3E family
of P-loop GTPases and more specifically of COG0523
members. By combining phylogenetic analysis, physical
clustering analysis, and regulatory site detection, we pre-
dict that the COG0523 family comprises subfamilies that
have specialized and distinct functions in metal metabo-
lism. We also hypothesize that several, but not all, of these
subfamilies have a role in survival under conditions of
poor zinc nutrition in both prokaryotic and eukaryotic
organisms.

Results and Discussion
Phylogenomic analysis of COG0523 proteins as members
of the G3E family of P-loop GTPases
Phylogenetic analysis first performed by Leipe, et al. [13]
and repeated here with a diverse set of family members
(see Figure 1) shows that COG0523 belongs to the G3E
family of P-loop GTPases (G3E family), which is sepa-
rated from the rest of the SIMIBI class of GTPases (for SIg-
nal recognition GTPases, MInD superfamily, and BIoD
superfamily) by a glutamate residue in the Walker B motif
and an intact NKXD motif (Figure 2A) [13]. Characterized
members of the G3E family perform two roles in metallo-
center assembly: 1) facilitating incorporation of the cofac-
tor in an energy- dependent manner into the target
protein's catalytic site (the insertase role) and, 2) storage
and delivery of a metal cofactor to a target metalloprotein
(the metallochaperone role). G3E family proteins have
been found to function as either metal-insertases or as a


Page 2 of 21
(page number not for citation purposes)






http://www. biomedcentral.com/1471-2164/10/470


- 1 (51) and 2 (12) -----------------------
pcd ppa
futB 0523 I 0432 taqCP 1
3 PM (6)
atzB uraA 0523 ubiX ubiD mh mh
4 B 0 M
wd40 0523 znuA znuB znuC
5 BJ (29)


0523 dacC ama
6 BS (13) 1


foIE2 yciC AB
1 AB
SC I DUF1826
\, 1 < o-- PAL
DUF1826
(5)' 0523-11 yciC dksA2 P F
| nhaA
nha3nhaB/ amdA AB


yciC BSU


Y


gatA potD D 0523 cih potA potD fabG fabG potC potB potA
7 .R L (5)
hslU hslV dksA 0523 zurznuC znuB znuA
8 BA (27)
0523 yjiX cstA2 COG0523
9 PA (19)
oppD oppC oppB oppA rtn spr IpxT 0523
10 YP (39)
DUF1826
0523 1 yciC
11 -'mj PP (6) members linked to
cbiH cbiL cbiC cobG cobN cobW cobU pduO cbiP zinc homeostasis
*'-L :: GB (51)


L 0523 ramA mtbA mtaA mtaC
13 MM (7)
I trxB feoB. 0523
14 4SA (4)
4989 0523
15 SS (8)


ureG ureE ureB
ureD \ ureF / ureC /reA
Fjoh 4832 u eA FJ
ureF ureB ureA UreG
ureG 1 ureE ureC lureJ/ ureD
Daci_1179 DA
hypB hypA
I -cce_2903 CY
hypE hypC HypB
SWS0791
r nmeaB mCmdo WB
sce2717 mm SC
meaB B, dom. MeaB
HQ2300A *HW


activation of Fe-NHase
*members linked to B12
biosynthesis
0 putative Zur binding site
# putative B12 riboswitch


32.8% prokaryotic genomes contain ureG
96% are in cluster with ure genes


29.5% prokaryotic genomes contain hypB
95% are in cluster with hyp genes


26% prokaryotic genomes contain meaB
63% are in cluster with MCM or the B12-
binding domain of MCM


Figure I (see legend on next page)


Page 3 of 21
(page number not for citation purposes)


I-


BMC Genornics 2009,10:470







http://www. biomedcentral.com/1471-2164/10/470


Figure I (see previous page)
Summary of phylogenomic analysis of G3E family and COG0523 members in prokaryotes. A, Phylogeny of
extracted GTPase domains from diverse members of the G3E family plus genome context for corresponding genes. B,
COG0523 distance tree. Each subgroup (3,4 and 6-15) was collapsed to its common node. Since subgroup 5 is paraphyletic
(the clade containing subgroup 5 also contains subgroup 4), branches were collapsed to three nodes. Representative gene
neighborhoods for each subfamily are shown as well as corresponding genome. Numbers in parentheses refer to the number
of species from which the gene cluster occurs. For subfamily I, this number refers to the number of species that contain a
putative Zur-regulated yciC. C, Representative branches from the subfamily I and 2 clade. Abbreviations: '0523', COG0523
homologs; 'yciC', subfamily I; 'cobW', subfamily 12; 'nha3', subfamily 2. For subfamily 1, '0523-1 I' refers to a subfamily II
COG0523 homolog that is found in the same gene cluster as yciC. All other gene abbreviations and genome abbreviations can be
found in Additional File 7.


dual function metallochaperone/insertase. For example,
MeaB appears to fulfill the role of an adenosylcobalamin
(Co2+)-insertase, facilitating the insertion of B12 into
methylmalonyl-CoA mutase (MCM) [271. A large struc-
tural rearrangement occurs upon interaction between
MeaB and its target [27], suggesting that MeaB may be
responsible for the structural changes that must occur for
B12 cofactor incorporation. In addition, MeaB also
appears to protect radical intermediates that are essential
for the activity of MCM [28,29]. In hydrogenase matura-
tion, HypB is thought to carry out both the insertase and
metallochaperone roles in most organisms due to the
presence of a histidine stretch located at the N-terminus of
these proteins. In Bradyrhizobium japonicum, this histidine
stretch was found to bind 18-Ni ions per dimer [30]. In
Escherichia coli, the common histidine stretch is missing
and SlyD is presumed to be the metallochaperone compo-


nent that delivers Ni to the assembly complex [31,32].
Most UreG proteins studied to date lack the histidine
stretch that is found in most HypB proteins. Accordingly,
in urease maturation, UreG appears to function as an
insertase and another accessory protein, UreE, functions
as the metallochaperone, delivering Ni to the maturation
complex [33-35].

To support the proposition that the presence of a histidine
stretch in G3E family proteins could be indicative of met-
allochaperone activity, we analyzed the distribution of
genes encoding UreG and UreE (see 'G3E' subsystem in
the SEED database [36]). We found that several genomes
containing ureG homologs lack any recognizable ureE
homologs. In all but two of those genomes, UreG con-
tains an added histidine-rich motif at the N- or C-termi-
nus (Table 1). As has been suggested for the UreG of


Walker B
GCxCC hhhExxG NKxD
m- 7- 7-


Bacillus licheniformis
Bordetella pertussis


HFLEDHHHHHHDD
HDHHDHDHGHGHDHDHDHDHE
GECGAHCNHHHHHHHAHHDD


Prochlorococcus marinus HEHEHEHEHEHEHEHEHEHEH


GTPase domain Variable C-terminus domain


Figure 2
COG0523 amino acid conservation plot. A, Plot of amino acid conservation. The conserved GTPase motifs are high-
lighted in red. The conserved GCXCC motif is highlighted in yellow. The most common positions of His-stretches are shown.
B, Typical histidine-rich sequence found in COG0523 homologs from specified genomes.



Page 4 of 21
(page number not for citation purposes)


Walker A
GxxGxGK
I 1


BMC Genomics 2009,10:470








http://www. biomedcentral.com/1471-2164/10/470


Table I: Co-occurrence profile between ureE, slyD, and a His-stretch in UreG.


ureE UreG histidine stretch


Anaeromyxobacter sp. Fwl09-5
Arabidopsis thaliana
Bradyrhizobium japonicum USDA I 10
Cytophaga hutchinsonii ATCC 33406
Frankia sp. Ccl3
Gibberellazeae PH- I
Herpetosiphon aurantiacus ATCC 23779
Magnaporthe grisea 70-15
Mycobacterium bovis AF2122/97
Mycobacterium marinum M
Mycobacterium tuberculosis CDC I 55 I
Mycobacterium vanbaaleni vanbaalenii PYR- I
Neurospora crassa
Nocardia farcinica IFM 10152
Schizosaccharomyces pombe
Sorangium cellulosum So ce 56

Streptomyces avermitilis MA-4680
Streptomyces coelicolor A3
Verminephrobacter eiseniae EFO I -2
Bacillus cereus ATCC 10987
Corynebacterium glutamicum ATCC 13032
Haloarcula marismortui ATCC 43049
Rhizobium leguminosarum bv. viciae 3841
Ureaplasma urealyticum serovar 10
Helicobacter pylon 26695


HDHSLHSGHDHGLGPGSFHDRGAPH
HDHHHHHHDHEHDH

HLDHFDSPGHFHHRELIH

HSHDGQSHSHDGFNAQEHGHSH
HVHDDHHHHHHH (C-terminus)
HSHSHDGSAPHSHSHDGSTFNAQEHGHSH
HSHPHSH
HSHDHTHDHH
HSHPHSH
HFLDGQPHGH
HTHSHDHGDGGHHHHPHSHSHDFNSQSGFNAQEHGHSH
HDHAH
HKGGSDDSTHHHTHDYDHHNHDHHGHDHHSHDSSSNSSSEAARLQFIQEHGHSH
HDPGEHGHGRHDHDHDHDHVHDHDHDHDHVHGGGHRHAHEHEHAHEHAHG
HEHGHAHAHAHAHAHEHAHGHTHEHWAH
HLDHAHTH
HLDHHH
HHLHH
+
+
+
+
+
+


Mycobacterium tuberculosis, this His-stretch may be able to
compensate for the absence of UreE [37]. In addition, the
absence of ureE does not correlate with the presence of
slyD as would be expected if SlyD performs the metallo-
chaperone role in those organisms. Bradyrhizobium japoni-
cum USDA 110 and Frankia sp. Ccl3 are two exceptions to
this trend as they lack both ureE and a His-stretch exten-
sion in UreG. In these cases, the Ni-metallochaperone
involved in urease maturation could be HypB, which is
present in both of these organisms (see 'G3E' subsystem).
Indeed, it has been shown in Helicobacter pylori that HypB
is required for activity of both hydrogenase and urease
[38], and a physical interaction between UreG and HypB
has been verified [39]. Although ureE is present in the
genome of H. pylori, the corresponding protein lacks a
His-stretch. As expected, the addition of a His-stretch to
UreE was found to eliminate the need for HypB in the
maturation of urease [40].

We compared the amino acid sequence of 887 COG0523
proteins from all kingdoms (see 'G3E' subsystem for
sequences). We observed that like UreG and HypB
orthologs, COG0523 proteins are found with and without
His-stretches, suggesting a distribution of insertase and
metallochaperone activity among various members.
Approximately 40% of the sequences analyzed contain a


histidine-rich region, commonly found near the C-termi-
nus (Figure 2A and 2B); 365 COG0523 proteins contain
the minimal HxHxHxH motif, where x represents 0 4 res-
idues. Some proteins contain a His-stretch with up to 29
histidines, such as Ava_3717 [Genbank:75703646] from
Anabaena variabilis.

The region of highest similarity between COG0523 and
the other members of the G3E family is the GTPase
domain, defined by the canonical Walker A and Walker B
motifs (Figure 2) [13]. GTPase activity of HypB and UreG
has been shown to be essential for the metallocenter bio-
synthesis of hydrogenase and urease, respectively [41,42],
and postulated to be responsible for incorporation of B12
in MCM by MeaB [29]. GTPase activity has been verified
for YjiA, a COG0523 homolog from Escherichia coli, for
which the crystal structure was solved [43]. In addition to
the GTPase domain, all members of COG0523 have a
conserved, putative, metal-binding CXCC motif (Figure
2A). Analysis of the YjiA crystal structure reveals this motif
is found in the Switch I region of the protein, suggesting
that binding of GTP/GDP affects its conformation [43].
The same motif was found to be essential for the activity
of the nitrile hydratase activator protein, a member of
COG0523 assumed to be involved in the incorporation of
iron into Fe-type nitrile hydratase [44,45].



Page 5 of 21
(page number not for citation purposes)


Organism


BMC Genomics 2009,10:470







BMC Genomics 2009, 10:470


In addition to the GTPase domain, MeaB and most
COG0523 proteins contain an additional C-terminal
domain. On average, COG0523 is 99 and 147 amino
acids larger than HypB and UreG, respectively, and only
26 residues larger than MeaB. The smallest G3E protein,
UreG, is the GTPase component of a complex composed
of UreD and UreF, where the three proteins act together in
the activation of urease [46]. Activation of MCM appears
to only require delivery of the cofactor by adenosyltrans-
ferase and the activity of MeaB[47]. The size of G3E pro-
teins could be indicative of the number of other accessory
proteins required for activation of the target metallopro-
tein.

While the N-terminal GTPase domain is well conserved
among COG0523 members, the C-terminal region is
highly variable (Figure 2A). Indeed, COG0523 proteins
fall under the category of "segmentally variable genes
(SVGs)," as defined by Zheng et al [48]. SVG profiles for
four members of the family (HP0312, NMB1263,
VCA0527, yeiR) can be found at http://geneva.bu.edu.
SVGs are genes that code for proteins that have highly var-
iable regions interspersed with well-conserved regions.
The authors observed that SVGs encode proteins that are
involved in adaptation to environmental stresses and pro-
posed that highly variable domains are an indication of
protein-protein interaction specificity or specificity of
small molecule binding.

Finally, as summarized in Figure 1, while hypB genes are
consistently found in hydrogenase maturation gene clus-
ters, ureG genes in the urease maturation clusters and
meaB genes cluster with MCM-encoding genes, COG0523
genes are found in multiple gene clusters. Most genomes
contain only one homolog of hypB, ureG, or meaB. Con-
versely, up to 11 COG0523 genes can be found in a single
genome, as seen in Cyanothece sp. ATCC 51142. In addi-
tion, the available functional analyses of COG0523 mem-
bers suggest varied functions and an interaction with
various metals [15,18,45]. We predict that the different
gene clusters involving COG0523 represent distinct sub-
groups. In contrast to the HypB, UreG, and MeaB sub-
families, which are composed of chaperones for a single
protein, each COG0523 subgroup may perform a chaper-


Table 2: Literature reports of COG0523 mutant data.


http://www. biomedcentral.com/1471-2164/10/470



one role in different metallocenter biosynthesis of various
proteins.

In summary, this analysis suggests that like HypB or UreG,
COG0523 proteins are most certainly metal insertase and/
or metallochaperones. However, the metal substrate and
the metalloprotein targets) of most COG0523 family
proteins is not obvious, and the initial analysis of the
COG0523 gene neighborhoods suggests that there could
be a greater diversity of targets than observed for the other
G3E family subgroups. To investigate the presence of
diverse COG0523 subfamilies, we combined literature
analysis (Table 2 and Additional File 1) with predictions
of Zur and B12 regulation (see below) and phylogenetic
and gene neighborhood analyses. This approach led to the
identification of fifteen subfamilies (summarized in Fig-
ure 1, detailed in Additional Files 2 and 3). Each sub-
family is monophyletic and the corresponding genes
belong to similar genomic neighborhoods and/or share
conserved regulatory sites. Two exceptions are subfamilies
1 (Figure IC) and 5 (Figure 1B), which appear to be para-
phyletic; the clade composed of subfamily 1 also contains
subfamily 2 and the clade that contains subfamily 5 also
contains subfamily 4 (Additional File 2). Five subfamilies
(1, 2, 5, 12 and 13) are analyzed in more detail below. The
10 others are detailed in Additional File 3.

The CobW subfamily involved in cobalamin biosynthesis
CobW was the first member of COG0523 to be described
and so-named because the disruption of the correspond-
ing gene in Pseudomonas denitrificans resulted in the inabil-
ity to synthesize cobalamin (Table 2) [15]. Although
cobalaminn biosynthesis protein" is the most highly prop-
agated annotation for COG0523 members, our compara-
tive genomic and phylogenetic analysis reveals that true
CobW proteins (Subgroup 12, Figure 1) represent only
12.5% of the COG0523 family (for a list of putative cobW
genes, see 'Coenzyme B12 biosynthesis' subsystem). In
our previous genomic analysis, cobW genes were identi-
fied in Proteobacteria located within the cobalamin bio-
synthesis gene clusters under control of the B12 riboswitch,
a regulatory RNA element modulating gene expression in
response to changing B12 concentrations [49]. As many
more genome sequences have become available, we
updated this analysis and report that 54 out of 65 cobW


Organism


Bacillus subtilis
Brucella suis
Burkholderia pseudomallei
Pseudomonas denitrificans
Saccharomyces cerevisiae


COG0523

YciC
BRA0987

CobW
YNR029c


Phenotype


EDTA-sensitivity in a AycdH background
Deficiency in intramacrophagic replication
Inability to infect C. elegans
Cobalamin-minus
Salt- and heat-sensitivity
EGTA-sensitivity


Page 6 of 21
(page number not for citation purposes)


[17,18]
[70]
[71]
[15]
[89]
[88]







BMC Genomics 2009, 10:470


orthologs analyzed belong to B12-regulated gene clusters
in y-, 3-, and a-proteobacteria (Additional File 4). In three
a-proteobacteria from the Rhodospirillaceae family, cobW
genes belong to the cobalamin biosynthesis gene clusters
that are not preceded by B12 riboswitches. Finally, cobW
orthologs in cyanobacteria are neither clustered with B12
biosynthesis genes nor regulated by a B12 riboswitch.
However, these orthologs are highly similar to other
CobW proteins and the corresponding genes co-occur
with the cobalamin biosynthesis genes of the aerobic
pathway.

In the majority of cases, cobW is found adjacent to the
cobalt chelatase component, cobN (Figure 1 and Addi-
tional File 3) and all CobW proteins analyzed contain a
His-stretch, which on average is composed of 7 histidines
(the least being 4 histidines and the most being 15). The
exact function of CobW is still not clear; it could be
involved in the presentation of the cobalt ion to the cobalt
chelatase, protection of the cofactor, or involved in insert-
ing a metal in a metal-dependent enzyme of the pathway,
such as Fe-dependent CobG [50].

The Nitrile hydratase activator subfamily
Based on our analysis, less then 0.7% of the COG0523
family is represented by the nitrile hydratase (NHase) acti-
vators (Subgroup 2, Figure 1). A complete list of identified
Fe-type NHase activators from both Genbank and SEED
databases is given in Additional File 3. In the literature,
these proteins are referred to as Nha3, P44K, or P47K,
depending on the organism in which the protein was
identified (Additional File 3). Here we refer to this sub-
group of COG0523 as Nha3. Nha3 is found clustered
exclusively with the genes encoding the two subunits of
the Fe-type NHase (Figure 1 and Additional File 3) and
has been found to be required for the in vivo activity of Fe-
type NHase [16]. NHases are enzymes that use either a
non-heme iron (III) or non-corrin cobalt (III) for the
hydration of nitriles to amides [51]. NHase types can be
differentiated by the strictly conserved metal binding
motifs CSLCSCT for Fe(III) and CTLCSCY for Co(II) [52].
Although, the same coordination geometry has been
determined for both Co(III)- and Fe(III)-binding sites
[53], the two types of NHases specifically incorporate the
correct metal. This specificity is thought to be due to acti-
vator proteins, which are required for full activity of their
respective NHase. For Co-type NHases, metallocenter bio-
synthesis is thought to occur via subunit exchange, a
mechanism called "self-subunit swapping" [54,55]. The
accessory protein in this case, NhlE, is a self-subunit swap-
ping chaperone and the corresponding gene is always
found adjacent with the target NHase genes (Additional
File 3). No sequence similarity is found between the Co-
type accessory protein and the Fe-type accessory protein


http://www. biomedcentral.com/1471-2164/10/470



supporting the conclusion that Co- and Fe-type metallo-
centers are assembled by different mechanisms.

Even if the involvement of Nha3 in Fe-type NHase activa-
tion is documented, its exact role is not known. It has
been postulated that it has an insertase role involved in
incorporation of iron into the active site of the hydratase
[45]. When the Fe-dependent NHase from Rhodococcus sp.
N-771 was expressed in E. coli without Nha3 in Co-sup-
plemented media, it incorporated Co instead of Fe [56].
Therefore, nitrile hydratase activator proteins may not
only be involved in incorporating Fe, but also in ensuring
that competing metal ions are excluded. In addition, the
coexpression of Nha3 with NHase was found to be unnec-
essary with the coexpression of the GroESL chaperones
[57]. This observation supports the hypothesis that
COG0523 proteins like the rest of the G3E family could
be involved in the structural rearrangements that must
take place to ensure the metal cofactor is incorporated
into the catalytic site.

Zur-regulated COG0523 proteins
Extensive analysis of the literature (Table 2 and Additional
File 1) reveals that members of COG0523 have been
implicated in the virulence of several pathogens whose
hosts are known to induce Zn-limitation as a defense
strategy. In 1973, Kochan introduced the concept of nutri-
tional immunity as a defense strategy against invading
pathogens [58]. The host organism actively deprives met-
als from the invaders inducing both hypoferremia and
hypozincemia (deficiency of iron and zinc, respectively,
in the blood) as part of the acute inflammatory response
[59-61]. Therefore, the mechanisms that enable a patho-
gen to overcome this host-induced Zn-starvation are con-
sidered essential to a pathogen's ability to cause infection
[62-64]. In Mycobacterium tuberculosis, a COG0523-like
gene, RV0106, (shown also to be repressed by Zur [65]) is
up-regulated during human macrophage infection [66]
(although RV0106 shows homology to COG0523, it is
missing both GTPase motifs, and the second cysteine of
the CXCC motif is not conserved). In the closely related
Mycobacterium avium subsp. paratuberculosis, this gene is
found on a pathogenicity island [66,67] and the corre-
sponding protein was the second strongest antigen con-
sistently reactive with cattle sera infected with M. avium or
Micobacterium bovis [68]. COG0523 is also found in a
pathogenicity island from Enterococcus faecalis [69]. The
loss of COG0523 in Brucella suis rendered this bacterium
incapable of intramacrophagic replication [70], while the
loss of COG0523 in Burkholderia pseudomallei results in the
inability to infect Caenorhabditis elegans (Table 2) [71]. An
ortholog from Francisella tularensis was expressed exclu-
sively in bacteria separated from infected murine spleen
tissue [72]. This gene is down-regulated in the Francisella
novicida ApmrA mutant [73]. PmrA is a transcription factor


Page 7 of 21
(page number not for citation purposes)







BMC Genomics 2009, 10:470


found to be essential for survival/growth inside human
and murine macrophage cell lines [731.

In plants, an opposing defense strategy may be employed,
as repression of zinc uptake machinery is required for full
virulence of the plant pathogens, Xanthomonas campestris
and Xanthomonas oryzae [74-761. In contrast to animal
pathogens and further supporting a role for COG0523 in
zinc homeostasis, two COG0523 homologs of Agrobacte-
rium tumefaciens as well as the genes encoding the high-
affinity zinc transporter, ZnuABC, are down-regulated in
response to plant signals (Additional File 1) [771.

The most extensive analysis of the zinc-dependent regula-
tion of a COG0523 gene has been performed on the
COG0523 (yciC) in Bacillus subtilis encoding a member of
subfamily 1 (Figure 1 and Additional File 3). The expres-
sion of yciC is controlled by the Zn-dependent Zur repres-
sor and is thus up-regulated under Zn-limiting conditions
[ 171. In addition to the work on yciC, an early comparative
genomic analysis had identified Zur-regulated yciC
orthologs in several Gram-positive bacteria (Bacillus, Sta-
phylococcus, Enterococcus) [621. As this initial analysis of
putative Zur-binding sites had been done when a limited
set of genomes was available and, as discussed above, scat-
tered observation links this family to zinc limitation, we
expanded the analysis to all currently complete bacterial
genomes (see Methods).

Sixty-eight yciC/COG0523 genes were found to be down-
stream of a potential Zur-binding site mainly in Firmi-
cutes and y-, 3-, and a-proteobacteria (Additional File 5).
Two COG0523 genes were found downstream of a puta-
tive Zur site in the cyanobacteria, Prochlorococcus marinus,
Nostoc sp. PCC 7120 and several Cyanothece species (Addi-
tional File 5). While most proteins encoded by Zur-regu-


http://www. biomedcentral.com/1471-2164/10/470



lated COG0523 members are found in subfamily 1
(75%), several paralogs are found in other subfamilies.
For instance, in Pseudomonas entomophila, Pseudomonas flu-
orescens, and Pseudomonas putida there are two COG0523
homologs per genome predicted to be downstream of a
Zur-binding site (Additional File 5). Our phylogenetic
analysis reveals that one paralog belongs to subfamily 1
while the other belongs to subfamily 11 (Additional File
2). Zur-regulated COG0523 paralogs are also found in
subfamily 5, 8, 10 and 14 (Additional File 2 and Addi-
tional File 5). One possibility is that the presence of sev-
eral Zur-regulated COG0523 subfamilies could be
indicative of more than one function of COG0523 under
zinc limitation (as discussed below).

Of the three COG0523 genes in Acinetobacter baylyi
ADPI, only one is regulated by Zur
To test the predictive power of our COG0523 phyloge-
nomic analysis, the regulation of the three COG0523
genes from Acinetobacter baylyi ADP1 was analyzed. The
first, ACIAD1614 [Genbank: ACIAD1614, 495307511, is
predicted to be a nha3 homolog (subfamily 2). The sec-
ond, ACIAD1025 [Genbank: ACIAD1025, 495302031, is
predicted to be most similar to subfamilies 10 and 11. The
third, ACIAD1741 [Genbank: ACIAD1741, 495308691, is
predicted to be regulated by Zur (subfamily 1). These
groupings suggest that the expression of only the
ACIAD 1741-encoding gene should be under Zur control.
Expression of the three A. baylyi COG0523 genes was ana-
lyzed by RT-PCR (see Material and Methods) in a WT
strain and in a Azur derivative (AACIAD0176). As shown
in Figure 3, the presence or absence of Zur does not affect
the expression of ACIAD1025 and ACIAD1614 under the
conditions tested. Nevertheless, as we predicted,
ACIAD1741 is only expressed in Azur background.


ACIAD1741


ACIAD1614


ACIAD1025


WT


zur::kan


WT


r_ __ r_ __ r_ __


zur. :kan


WT


r_ __ r_ __


zur. :kan
r___


M -RT +RT -RT +RT -RT +RT -RT +RT -RT+RT -RT+RT


1000 bp


500 bp


Figure 3
Differential expression of the three COG0523 genes of Acinetobacter baylyi ADP I. Amplification of ACIAD 1741
(lane 2, 3, 4 and 5), ACIAD I 1614 (lane 6, 7, 8 and 9) and ACIAD 1025 (lane 10, I I, 12 and 13) transcripts from A. baylyi ADP I
(WT) and A. baylyi Azur::kan (zur::kan). Abbreviations: M, base pair marker; -RT, reaction without reverse transcriptase; +RT,
reaction with reverse transcriptase.



Page 8 of 21
(page number not for citation purposes)







BMC Genomics 2009, 10:470


Identification of putative Zur-regulated genes encoding
paralogs of Zn-dependent enzymes
Genome context analysis revealed that a significant pro-
portion of the yciC genes (Zur-regulated COG0523) are
located within chromosomal gene clusters including
genes for zinc transport (e.g., znuABC), the zur repressor,
and various genes encoding paralogs of zinc-dependent
proteins (Figure 4 and 5). Nine families of Zn-dependent
enzymes whose paralogs belong to Zur regulons in y-, and
3-proteobacteria (Figure 5) were found. These Zn-depend-
ent enzymes include phosphoribosyl-AMP cyclohydro-
lase (HisI), dihydroorotase (PyrC), y-class carbonic
anhydrase (Cam), porphobilinogen synthase (HemB),
cysteinyl-tRNA synthetase (CysRS), threonyl-tRNA syn-
thetase (ThrRS), N-acetylmuramoyl-L-alanine amidase,
queuosine biosynthesis enzyme QueD, and C4-type zinc
finger regulator DksA (see 'Zinc regulated enzymes' sub-
system).

Differential regulation of distinct isofunctional genes by
co-factor availability is a known regulatory mechanism in
bacteria (for a review, see [78]) and in eukaryotes (as dis-
cussed below). For instance, the coenzyme B12-independ-
ent isozymes of methionine synthase and ribonucleotide
reductase are regulated by B12 riboswitches in the
genomes that encode both B12-dependent and -independ-
ent isozymes [49]. Likewise, a similar regulatory strategy
has been described for zinc availability. Zn-independent
proteins are negatively regulated by Zur and expressed


http://www. biomedcentral.com/1471-2164/10/470



under Zn-limiting conditions to replace the Zn-dependent
proteins. Examples include paralogs of ribosomal pro-
teins [62] and alternative isozymes of GTP cyclohydrolase
I (FolE1 and FolE2) [79]. In both cases, a Zn-dependent
protein is functionally replaced by a Zn-independent
counterpart during conditions of zinc deficiency.

Our comparative analysis of Zur regulons revealed co-reg-
ulation and frequent co-localization on the chromosome
between COG0523 and paralogs of these Zn-dependent
enzymes. For example, Cupriavidus metallidurans has a
Zur-regulated gene cluster encoding YciC, FolE2, and par-
alogs of CysS, QueD, Cam, and PyrC, whereas the Zur reg-
ulon in Pseudomonas fluorescens includes two COG0523
proteins, FolE2, and paralogs of AmiA, DksA, HisI, Cam,
and PyrC (Figure 4 and 5).

We hypothesize that these alternative enzymes could
require a metal other than Zn (or no metal) and are there-
fore expressed during Zn-limitation to replace or compen-
sate for the decreased activity of their Zn-dependent
counterparts. Indeed, the carbonic anhydrases found in
our analysis are members of the y-class. The y-class car-
bonic anhydrase from Methanosarcina thermophila exhib-
ited highest activity with Fe and, when purified under
anaerobic conditions, contained Fe and not Zn [80,81].
The Zur-regulated cam we have identified could therefore
encode an Fe-dependent carbonic anhydrase expressed to
compensate for the Zn-dependent carbonic anhydrases.


Pseudomonas fluorescens Pf-5
hisl2 dksA2 DUF1826
thrS2 pyrC2 cam, znuABC folE2 amiA2 | yciC I 0523-11
'A A k


Brucella melitensis 16M





Bacillus subtilis subsp. subtilis str. 168


folE2
- =>_


Cupriavidus metallidurans CH34
cam
yciC folE2 cysS2 queD2/ pyrC2


zur 0523-8 dksA2 yciC


OMR
4>


yciC

I Zur binding site


Figure 4
Representative gene clusters composed of Zur-regulated COG0523 members. Genes labeled yciC represent sub-
family I COG0523 members. COG0523-1 I and COG0523-8 refer to subfamilies I I and 8, respectively. Abbreviations not found
in text: DUF1826, Pfam protein family of unknown function; OMR, outer membrane protein.



Page 9 of 21
(page number not for citation purposes)









BMC Genomics 2009, 10:470


http://www. biomedcentral.com/1471-2164/10/470


y-proteobacteria Zur COG0523 FolE1 FolE2 CynT Cam PyrC lemE HisI DksA AmiA QueD ThrS
Pseudoalteromonas atlantica T6c + + +, + + + + +, + + + + +, + +
EscherichiacoliK12 + +, + + +, + + + + + +,+ +, + + +
Klebsiella pneumoniae MGH 78578 + +* +, +, +* + +* +, +, + +* + + +, +* +, +* +, +, +, + + +
Serratia marcescens Db11 + + + +, +, + + + + + + ++ +
Alcanivorax borkumensis SK2 + +, + + +, + + +, +, + + + +, + + +
Hahella chejuensis KCTC 2396 +, + +, + + + +, +, + +, +, + +, + +, + +, + + +, +
Chromohalobactersalexigens DSM3043 + + + + +, + + + + + +, + +
Marinomonas sp. MWYL1 + +, + + +, + +, + +, + +, + + + +, +
AcinetobacterbaumanniiATCC 17978 + +, + + +, + +, +, + + + + + +
Acinetobacterbaylyi ADP1 + +, +, + + + +, + + +, + + + + + + +
Psychrobacter sp. 273-4 + + + +, + +, + + + + + +
Azotobacter vinelandii + +, +, +, + +, + + +, + + + +, + +, +, + +, + +
Pseudomonas aeruginosa PA01 + +, +, + +, + + +, +, + + +, +, + + + +, + +, +, + + +
Pseudomonas entomophila L48 + +, +, + +, + + + + + +, +, + + + +, + + +, + + +
Pseudomonas fluorescens Pf-5 + +, +,+ + +, +, +,+ + + +, + +, + + +, +, + + +, +
Pseudomonas putida KT2440 + +, +,+, + +,+, + + +, + +, + + +, + +, + + +
Pseudomonas syringae B728a + +, +,+, + +, + + +, + + +, + + + +, + +, +, + + +
Francisella tularensis Schu 4 + + + +, + + + +
Vibrio alginolyticus 12G01 +, + +, +, +, + + +, +, + + +, + + + + + + +
Vibrio fischeri ES114 + + + + +, + + + + + + + + +, +
Xanthomonas axonopodis pv. citri str. 306 + + + + +, +, + + + + + +, +, + + +
Xanthomonas campestris ATCC 33913 + + + + +, +, + + + + + +, +, + + +
Xylella fastidiosa Temeculal + + + + +, + + + + + + + + +
Reinekea sp. MED297 +, + +, +, + + + +, + + + + + + +


R-nrnftnhafitnria


Bordetella pertussis Tohama I
Burkholderia cenocepacia AU 1054
Burkholderia cenocepacia J2315
Burkholderia cepacia R18194
Burkholderia cepacia R1808
Burkholderia fungorum
Burkholderia mallei ATCC 23344
Burkholderia xenovorans LB400
Burkholderia pseudomallei 1710b
Cupriavidus metallidurans CH34
Ralstonia eutropha JMP134
Ralstonia solanacearum GMI1000
Methylobacillus flagellatus KT
Chromobacterium violaceum ATCC 12472
Neisseria meningitidis Z2491
Azoarcussp. EbN1


nl_- A 1 .-. .-rJ o


B-rtobcei __ __ _ I-,, Y.-V,. j._ U :__


+, +
+, +, +,
+, +, +,
+, +, +,
+, +,
+, +,
+, +,
+, +,
+, +,
+, +, +
+, +, +
+, +
+, +, +
+
+
+


+
+
+
+, +
+
+, +, +
+
+, +, +
+
+
+

+, +

+


+, +
+, +, 4
+ + 4
+, +, 4

+ +
+, +
+ +
+, +
+, +
+, + 4
+, +
+, +
+, +
+
+
S+
"1- +1


Figure 5
Genomic co-localization of genes belonging to putative Zur regulon. Genomic distribution, candidate Zur-dependent
regulation and genomic co-localization of genes encoding Zur, COG0523 and Zn-dependent enzymes and their paralogs in y-
and P-proteobacteria. The presence of genes encoding the respective protein (columns) is shown by '+'. Multiple paralogs are
shown by '+' separated by a comma. Genes clustered on the chromosome (e.g. operons) are highlighted by a matching color.
Genes predicted to be regulated by Zur are marked in red. Zur-regulated gene cluster on the virulence plasmid, pLVPK, of
Klebsiella pneumoniae is marked by an asterisk.


The proteins of three other families downstream of puta-
tive Zur binding sites are missing the conserved zinc bind-
ing residues. As shown in Figure 5 and Additional File 6,
some genomes encode three PyrC paralogs. One paralog
is similar to the dihydroorotase from Escherichia coli.
These proteins have a binuclear zinc center chelated by the
conserved metal binding residues His 16, His 18, Lys 102,
Asp 250, His 139 and His 177 [82]. The second PyrC par-
alog is an inactive dihydroorotase, which is referred to in
the literature as PyrC' [831. Similar to PyrC', the zinc-bind-
ing residues are not conserved in the PyrC paralog whose


gene we predict to be regulated by Zur. However, unlike
PyrC', this PyrC paralog has previously been shown to dis-
play dihydroorotase activity [841. For porphobilinogen
synthase, the existence of zinc binding and non-zinc bind-
ing variants is documented in the literature [851. As
expected, HemB 1 contains the zinc chelating cysteine lig-
ands while those cysteines are not conserved in the pro-
tein, HemB2, encoded by the gene putatively regulated by
Zur in Pseudomonas putida (Additional File 6). HemB2 we
would accordingly expect to be active with magnesium
and/or potassium instead of zinc. In addition, the DksA


Page 10 of 21
(page number not for citation purposes)


F


r'nr'ncm I c^ic-i


ICO I "*%,." T 1







BMC Genomics 2009, 10:470


paralogs downstream of putative Zur binding sites are
missing the canonical C4-zinc finger motifs (Additional
File 6).

Not all paralogs seem to have lost their zinc-binding sites
as the zinc-binding residues are conserved in the HisI,
CysRS, ThrRS, QueD, and AmiA paralogs encoded by
genes predicted to be induced during zinc depletion. An
alternative explanation for the existence of these paralogs
could be to increase protein copy number during zinc
deficiency. The analysis of the metal content of some par-
alogs identified in this study is currently underway.

COG0523 in eukaryotes: Two Chlamydomonas
reinhardtii COG0523 homologs are induced under zinc
limitation
COG0523 is widespread in eukaryotes, with most organ-
isms containing one to four homologs (see 'G3E'subsys-
tem), and have been associated with stress phenotypes
(see Table 2 and Additional File 1). In Arabidopsis thaliana,
one of the three COG0523 genes (At1 g80480 [Genbank:
AT1g80480, 51536562]), which was isolated as a member
of the actively-transcribed plastid chromosome in mus-
tard seed [86], is induced under heat-stress [87]. Deletion
of COG0523 from Saccharomyces cerevisiae, YNR029c
[Genbank: 6324356], confers sensitivity to the metal che-
lator, glycol-bis (2-aminoethylether)-N,N,N',N'-tetraace-
tic acid (EGTA) [88] as well as salt-sensitive and heat-
sensitive phenotypes [89] (Table 2).

Gene clustering is not very informative in eukaryotes but
most eukaryotic COG0523 homologs including Homo
sapiens belong to subfamily 5 (Figure 6). The prokaryotic
members of subfamily 5 cluster on the genome with genes
that encode WD40-repeat proteins, which form a 3-pro-
peller structure thought to mediate protein-protein inter-
actions [90], and with znuABC and creatinase encoding
genes (Additional File 3). Several prokaryotic members of
subfamily 5 are also predicted to be downstream of a Zur-
binding site (Additional File 5), suggesting a role for bac-
terial members of subfamily 5 in the response to zinc lim-
itation.

Little if any work has been performed on the role of
COG0523 in eukaryotes, which do not encode a Zur
homolog. Therefore, we sought to investigate the regula-
tion of COG0523 during metal depletion in a eukaryotic
reference organism. Previous studies have established the
alga Chlamydomonas reinhardtii as a choice reference
organism for the study of trace metal homeostasis because
it is straightforward to deplete the medium of zinc, cop-
per, iron or manganese (as seen in [91-93]). Sentinel
genes for each of these deficiencies are known, such as
CYC6 for copper deficiency, FOX1 for iron-deficiency,
NRAMP1 for manganese deficiency and ZRT3 for Zn-defi-


http://www. biomedcentral.com/1471-2164/10/470



ciency [92,94,95]. Furthermore, Chlamydomonas has
retained many pathways present in the common ancestor
to the plant and animal lineages and displays the meta-
bolic flexibility of "back up" or alternate systems [96,97].
For instance, the replacement of B 12-independent methio-
nine synthase with a B12-dependent form when this cofac-
tor is available, the use of Mn-dependent superoxide
dismutase (SOD) in place of Fe-SOD in iron-limitation
and the replacement of plastocyanin with cytochrome c6
in copper-deficiency [93,98,99].

We identified 15 genes encoding proteins with COG0523
domains in versions 3.1 and 4.0 of the C. reinhardtii draft
genome [96] (see Additional File 7 for protein IDs). Of
these 15 gene models, only 10 gene models encoded full-
length COG0523 GTPase domains. Therefore, we tested
the expression of these 10 genes as a function of Zn, Cu,
Fe and Mn nutrition. Transcripts for two of these, encod-
ing proteins 123019 and 117458 (version 3.1 protein
IDs), are increased in abundance by several orders of mag-
nitude when cells are grown in zinc-limiting conditions as
opposed to zinc-replete conditions, only slightly induced
under copper limitation relative to copper-replete condi-
tions and unaffected by iron or manganese nutrition (Fig-
ure 7). The zinc sensors and regulatory factors responsible
for mediating this response to zinc depletion ARE yet
unknown in C. reinhardtii.

Our phylogenetic analysis reveals that protein 123019
belongs to subfamily 1, while protein 117458 belongs to
subfamily 5 (Figure 6 and Additional File 2). We therefore
substantiate the role of COG0523 family members in Zn
homeostasis in eukaryotes as well as in bacteria. Several
COG0523 proteins encoded by eukaryotic genomes
belong to subfamily 5 (Figure 6). We predict that the
expression of some of these other eukaryotic COG0523
proteins may also be regulated by zinc.

In addition, supporting the functional diversity revealed
by our gene neighborhood analysis, the expression of the
eight other COG0523 family members from C. reinhardtii
are not significantly affected by the deficiency of metals
tested.

COG0523 in Archaea
Although COG0523 was previously assumed to be miss-
ing from Archaea [13], the availability of recently
sequenced genomes reveals that out of 44 archaeal
genomes in the SEED database, eight genomes contain at
least one COG0523 homolog, with Methanosarcina acetivo-
rans C2A containing eight homologs. Most Archaeal
members belong to subfamily 13, members of which co-
localize with corrinoid-dependent methyltransferases
(Figure 1 and Additional File 3). In Methanosarcina barkeri,
Methanosarcina acetivorans, Methanosarcina mazei, and


Page 11 of 21
(page number not for citation purposes)







http://www. biomedcentral.com/1471-2164/10/470


* Enterococcus faecalis
* Oceanobacillus iheyensis
Chlamydomonas reinhardtii (106748)
Arabidopsis thaliana (At1 g15730)
Arabidopsis thaliana Atl g80480)
Chlamydomonas reinhardtii (195946
Chlamydomonas reinhardtii (122261
* Chlamydomonas reinhardtii (117458
Chlamydomonas reinhardtii (101629
Arabidopsis thaliana (At1 g26520)
Saccharomyces cerevisiae (YN R029c)
Taeniopygia guttata
Rattus norvegicus
Homo sapiens (CBWD2)
Homo sapiens (CBWD1)
* Sinorhizobium meliloti
* Caulobacter crescentus
* Bradyrhizobium japonicum
* Nitrobacter sp. Nb-311A
* Nitrobacter hamburgensis
* Rhodopseudomonas palustris
Bartonella henselae
Rhizobium leguminosarum
Serratia marcescens
Mesorhizobium sp. BNC1
Brucella suis
Synechococcus sp. WH 8102
* Nostoc sp. PCC 7120


Figure 6
Phylogeny of eukaryotic COG0523 members. Lightly shaded tree represents collapsed COG0523 tree. Branches are
labeled with corresponding subfamily number. Subfamily 5 Tree: branches representing eukaryotic homologs are colored. Blue
diamonds indicate putative Zur-binding sites upstream of corresponding genes. For the C. reinhardtii ortholog, the blue diamond
indicates confirmed induction of corresponding gene to zinc deficiency.


Methanococcus maripaludis S2, COG0523 clusters with
genes involved in methanol:CoM methylation: mtaA,
mtaB, (both are Zn-dependent [100]), mtaC (corrinoid
protein [101]) and ramM (iron-sulfur protein [102]) (Fig-
ure 1 and Additional File 3). Clustering between
COG0523 and methanogenesis genes is not limited to
Archaea but also found in Clostridium botulinum (Addi-
tional File 3). Another clostridium, Desulfitobacterium haf-
niense DCB-2, encodes a COG0523 that clusters with a
MeTr homolog (methyltetrahydrofolate:corrinoid/iron-
sulfur protein methyltransferase) (Additional File 3).
Also, proteome analysis of acetate- and methanol-grown
M. acetivoran scells revealed the presence of MA4382
(COG0523) in methanol-grown cells [103]. Finally,
MM1072 (M. mazei COG0523) is induced to the same


extent as its neighboring ramM homolog, MM1071, dur-
ing growth in high salt conditions (2.38 and 2.21 fold,
respectively) [104].

Archaeal genomes sequenced to date lack any recogniza-
ble homolog of the Fur (Fe) or Zur (Zn) transcriptional
regulators. Alternatively, there is a large group of MntR/
DtxR-like regulators in Archaea (Methanosarcina, Pyrococci,
Archaeaglobus, Methanobacterium) that regulate iron home-
ostasis, whereas another small group of MntR/DtxR-like
repressors in Methanosarcina spp. named ZntR (e.g.
MA0022 in M. acetivorans), is predicted to regulate the
zinc uptake operon, znuABC (D.A.R., unpublished obser-
vation, see Additional File 8). Comparative genomic anal-
ysis of this novel zinc regulon in M. acetivorans reveals that


Page 12 of 21
(page number not for citation purposes)


BMC Genomics 2009,10:470








BMC Genomics 2009, 10:470


http://www.biomedcentral.com/1471-2164/10/470


10000

1000

100

10-

1

0.1


10000

1000

100

10

1 -

0.1


10000

1000

100

10

1

0.1


10000

1000

100

10-

1


a * *


* a i *a
n


I I


S, , T

MN<-LONWO(i)i-oNc


C0 <0( T--

N1 LL

Figure 7
Differential expression of the COG0523 genes of Chlamydomonas reinhardtii. Chlamydomonas strain 2137 was grown
under various metal-deficient or replete conditions in triplicate experiments, represented by squares, circles and triangles.
RNA was isolated from these cultures and analyzed by real time PCR. Each metal deficiency, zinc, copper, iron, and manganese,
is shown in a separate panel. RNA abundance is expressed relative to the metal-replete condition. Each data point represents
an independent experiment with each measurement representing the average of technical triplicates. CBLP was used as the ref-
erence gene. ZRT3, CYC6, FOXI and NRAMPI served as positive controls for Zn-, Cu-, Fe-, and Mn-deficiency, respectively.
(Note that Cu-deficiency is a secondary effect of Zn-deficiency (Malasarn, unpublished) and Fe-deficiency is a secondary effect
of Mn-deficiency [92].)





Page 13 of 21
(page number not for citation purposes)


1 tI I


I t


O







BMC Genomics 2009, 10:470


a COG0523 homolog, MA4381, is co-regulated with znu-
ABC based on the presence of candidate-binding sites of
the ZntR repressor (Additional File 8).

Conclusion
COG0523 is a diverse family of metal chaperones
Based on relatedness to the G3E family of GTPases, we
expect COG0523 to also be involved in metallocenter bio-
synthesis of target metalloproteins. The diversity of
genomic co-localization suggests that COG0523 is more
diverse than the other subfamilies of G3E. Both the metal
specificity and the protein targets) might vary from one
subgroup to another.

While the known roles in cobalamin biosynthesis and
response to zinc limitation predominate, our analysis
implies members of COG0523 are not limited to those
roles. Based on genome context (co-localization and/or
presence of a B12 riboswitch) and protein similarity analy-
ses, only 12.5% of sequenced COG0523 from the SEED
are true CobW proteins and assigned to the cobalamin
biosynthesis pathway. Only ~30% of COG0523 members
analyzed are linked to the zinc homeostasis either
through putative Zur sites (~8%) or co-localization with
genes involved in the response to zinc starvation (~20%).
In addition the third known role, NHase activator, only
applies to less than 1% of sequenced COG0523 genes.
Over half of COG0523 may perform a role in the activity
of unknown proteins.

A ubiquitous subset of COG0523 is linked to zinc
Although involvement in the response to zinc deficiency
applies to only a subset of COG0523, we show that this
function is not limited to Bacteria but also present in
Archaea and Eukaryota. Two C. reinhardtii COG0523
homologs, which belong to separate phylogenetic sub-
groups, are induced under zinc-deficient conditions com-
pared to zinc-replete conditions. In addition, the
expression of the eight other homologs was not signifi-
cantly affected by metal deprivation confirming the diver-
sity of COG0523.

Identification of novel zinc homeostasis mechanisms
The comparative genomic analysis of the zinc repressor
Zur regulons in Bacteria has revealed insights into previ-
ously unknown zinc homeostasis mechanisms. At least
nine protein families that are homologs or isozymes of
known Zn-dependent proteins were identified as candi-
date members of the reconstructed Zur regulons in y- and
3-proteobacteria, suggesting their up-regulation during
zinc limitation. Based on sequence analysis, four of these
protein families do not contain the canonical zinc bind-
ing residues. We propose, therefore, that these paralogs
may require a metal other than Zn for catalysis and are
involved in the adaptation to poor zinc nutrition. The


http://www. biomedcentral.com/1471-2164/10/470



presence of these paralogs could aid in compensating for
the loss in activity of the Zn-dependent protein analogs
and reducing in the total amount of Zn required by the
cell.

Putative roles of COG0523 in response to zinc limitation
At this stage, if the exact role of COG0523 members in
survival in low zinc conditions is still to be determined,
several hypotheses can be proposed. Our comparative
genomic analysis suggests that COG0523 may be a metal
chaperone for a protein that is also part of the Zur regu-
lon. The 'alternative enzymes' of Zn-dependent proteins
may require a metal other than zinc for catalysis and may
also require a metal chaperone for efficient cofactor acqui-
sition. However, in about half of the genomes analyzed,
yciC (Zur-regulated COG0523) appears to not belong to
any operon. For instance, although adjacent to folE2, yciC
is usually regulated by Zur independently (Figure 5).

Another possibility would be that COG0523 is involved
in the allocation and reallocation of zinc. Zinc is not an
essential cofactor for metabolic steps where zinc-inde-
pendent back-up proteins can substitute. Accordingly, in
conditions of poor zinc nutrition, we expect that zinc
delivery is prioritized to proteins that do not have zinc-
independent substitutes (and where zinc function is
hence essential). Induction of the C. reinhardtii genes,
123019 and 117458, containing the putative metal deliv-
ery COG0523 domain, may affect prioritized delivery to a
subset of zinc targets. These delivery factors might be par-
ticularly important in a compartmentalized eukaryotic
cell. In bacteria, COG0523 may also function as either a
zinc chaperone as proposed recently by Gabriel et al. [105]
or as a molecular chaperone that aids in the folding of
essential zinc metalloproteins ensuring that essential Zn-
proteins acquire zinc while nonessential proteins are
excluded (the possible existence of a zinc metallochaper-
one is discussed in the recent review [106]).

As a third hypothesis, some members of COG0523 may
be a chaperone involved in incorporating a metal other
than zinc into Zn-dependent enzyme(s) based on zinc
availability. In vitro, the activity of several Zn-dependent
enzymes is slightly less, the same, or in some cases higher
with a metal cofactor other than zinc (for recent examples
see [107-109]). Under zinc limitation and supplementa-
tion with cobalt, the zinc in carbonic anhydrase of the
marine diatom, Thalassiosira i,.. i- fl. e:i is substituted with
cobalt in vivo [1101. The genome of the closely related Tha-
lassiosira pseudonana encodes seven COG0523 proteins.
Interestingly, the genomes of cyanobacteria and algae
tend to encode relatively high numbers of COG0523 pro-
teins. Zinc-containing carbonic anhydrases are important
for assimilation of CO2, and algae tend to express multi-
ple isoforms in various organelles [111], which might


Page 14 of 21
(page number not for citation purposes)







BMC Genomics 2009, 10:470


require mechanisms for preferential metal delivery. The
symbiotic alga Chlorella sp. NC64A has twelve homologs
and the free-living Chlorella vulgaris C-169 has seven. The
genome of Micromonas sp. RCC299 encodes ten
COG0523 homologs and the smallest known free-living
eukaryote, Ostreococcus tauri, has four homologs. As stated
above, the cyanobacterium Cyanothece sp. ATCC 51142
has eleven homologs, while Anabaena variabilis has five
and Nostoc sp. PCC7120 and Prochlorococcus marinus susp.
marinus both have four homologs. This high number of
paralogs might reflect their particular lifestyles.

Lastly, there is some evidence from Magnetospirillum mag-
neticum AMB-1 that a MeaB homolog may function as a
cytoplasmic ATPase required for energizing Fe uptake
[112]. Indeed, neither MeaB homolog encoded in the M.
magneticum genome appears to be co-transcribed with a
gene encoding methylmalonyl-CoA mutase (also the case
for ~40% of meaB homologs; see 'G3E' subsystem). There-
fore, a role for some COG0523 members in affecting
metal transport cannot be ruled out at this point.

Further experimental work is now required to discrimi-
nate between these different potential roles. To compli-
cate the problem, there are up to three Zur-regulated
COG0523 paralogs in some genomes, therefore, a combi-
nation of the above functional hypotheses may prove to
be operational.

Methods
Comparative genomic analysis of COG0523 gene family
and G3E family
Analysis of 'COG0523,' 'G3E,' and 'Zinc regulated
enzymes' subsystems were performed in the SEED data-
base [23,36]. COG0523 gene sequences in the SEED data-
base were identified by homology to known COG0523
members and the presence of the conserved CXCC motif
and P-loop GTPase domain in the corresponding protein
sequences. cobW gene sequences were identified based on
homology to cobW from Pseudomonas denitrificans [15]
and occurrence within cobalamin biosynthesis operons
and/or downstream of a putative B12 riboswitch. Genomic
search for candidate B12 riboswitches was performed as
previously described [49].

Sequence Analysis
All COG0523, HypB, UreG, and MeaB sequences pre-
sented here were downloaded from the SEED or Genbank
[113] databases. The fig numbers (internal identifications
in the SEED) and Genbank accession numbers can be
found in Additional File 7. Identification of histidine
motifs was performed with Fuzzpro from the EMBOSS
software package [114]. Amino acid sequences were
aligned using the ClustalW2 algorithm with default
parameters [115]. For alignments of PyrC, DksA and


http://www. biomedcentral.com/1471-2164/10/470



HemB, ESPript 2.2 was also used [116]. Phylogenetic
analyses were carried out by employing the Phylip 3.67
program package [117]. Distance-based matrices were
generated between all pairs of sequences using the Jones-
Taylor-Thornton matrix as employed in Protdist (Phylip).
Phylogenetic trees were generated from these matrices
using the neighbor-joining method as implemented in
Neighbor (Phylip). Reliability of branches was deter-
mined with the bootstrap method of 1000 replicates using
Bootseq (Phylip).

For the G3E family distance tree, the GTPase domain was
extracted and aligned. The GTPase domain of CooC was
used as an outgroup. Although it has been previously
assumed that CooC is a member of the G3E family
[ 14,29 ], GTPase sequence motifs suggest that it is actually
a member of the closely related MinD/BioB family. The
COG0523 distance tree was built with 177 full-length
COG0523 sequences. Significant gene clusters between
prokaryotic COG0523 genes and neighboring genes were
identified in the SEED database. Members of COG0523
with the highest functional coupling score for each signif-
icant gene cluster were chosen for inclusion in the phylo-
genetic analysis. Functional coupling scores and
significant gene clusters were computed by the SEED data-
base. For an explanation of functional coupling scores
refer to [118]. In addition, COG0523 proteins whose
genes were identified through our analysis of the Zur reg-
ulon were also included. The COG0523 proteins from six
eukaryotes were also used including the 10 COG0523
homologs from C. reinhardtii whose transcript levels were
investigated by real-time PCR. RV0106 from Mycobacte-
rium tuberculosis CDC1551 was used as an outgroup in this
analysis. This protein, while having similarity to
COG0523, does not contain the canonical CXCC motif
(CXSC). In addition, it is missing the canonical Walker A
motif of the GTPase domain, suggesting that these
COG0523-like proteins do not have GTPase activity. Sub-
families were defined based on the following criteria.
Each subfamily had to be monophyletic. The exceptions
are subfamilies 1 and 5, which are paraphyletic. Sub-
family 1 becomes monophyletic with the subtraction of
the nitrile hydratase activators. Subfamily 5 becomes
monophyletic with the subtraction of subfamily 4. The
genes encoding proteins in each subfamily belong to sim-
ilar gene clusters and/or have shared regulatory sites.
Bootstrap values were all above 900. Tree illustration was
performed with Treedyn [119].

Plot of amino acid conservation
Thirty-two protein sequences representing the 15
COG0523 subfamilies were aligned using ClustalW2 and
default parameters. Accession numbers for sequences
used can be found in Additional File 7. Columns contain-
ing gaps in eight or more sequences were removed. Resi-


Page 15 of 21
(page number not for citation purposes)







BMC Genomics 2009, 10:470


due conservation at each position as determined by
Jalview was plotted [120].

Comparative genomic analysis of Zur regulons
Complete bacterial genomes were downloaded from Gen-
Bank [113]. The taxon-specific training sets for identifica-
tion of Zur-binding motifs were composed of the
previously identified Zur-binding sites in Firmicutes, a-,
3-, and y-proteobacteria [62] and the DNA motif search
profiles were constructed using the SignalX program. Ana-
lyzed genomes encoding both an ortholog of Zur regula-
tor and COG0523 proteins were scanned with the
constructed taxon-specific Zur motif profiles (see Addi-
tional File 9 for sequence logo motifs) using the Genome
Explorer software and the identified genes with candidate
Zur-binding sites were analyzed by the consistency check
comparative procedure as previously described [78]. Can-
didate ZntR-binding motif was obtained by applying the
SignalX program to the training set of the znuACB regula-
tory regions from methonogenic Archaea that have an
ortholog of the DtxR-like regulator ZntR (MA0022). Posi-
tional nucleotide weights in the recognition profile and Z
scores of candidate sites were calculated as the sum of the
respective positional nucleotide weights as described in
[121]. The threshold for the site search was defined as the
lowest score observed in the training set. Sequence logos
for DNA-binding sites were constructed using WebLogo
2.0 [1221.

Acinetobacter RT-PCR
Acinetobacter baylyi ADP1 (ADP1) Azur:kanR was a gener-
ous gift from Veronique de Bernardinis (Genoscope, Insti-
tut de Genomique (CEA), Evry, France) [123]. Overnight
cultures of ADP1 and Azur:kanR cultured in Luria Broth
was used to inoculate 5 ml culture of Luria Broth supple-
mented with 50 gM ZnSO4 to ensure repression of tran-
scription by Zur. Samples (1 ml) were harvested in early
stationary phase, RNAprotect Bacteria Reagent (Qiagen)
was added and cells were frozen at negative 80C over-
night. Pellets were thawed and RNA was extracted using
TRIzol LS reagent (Invitrogen) followed by RNeasy mini
kit (Qiagen). Contaminating DNA was removed using
DNase I (RNA-free) (Ambion). RT-PCR reactions were
carried out with Superscript'T III One-Step RT-PCR System
with Platinum'Taq High Fidelity (Invitrogen). Reactions
were composed of 7.5 gl 2x reaction mix, 1 gl RNA (200
pg RNA), 0.3 gl Forward Primer (10 gM), 0.3 gl Reverse
Primer (10 gM), 0.3 gl Superscript'T III RT/Platinum Taq
High Fidelity enzyme mix, and water to a final volume of
15 gl. Reverse transcriptase minus controls were per-
formed using 5 PRIME Taq Master Mix (Fisher). Reactions
were composed of 6 gl 5 PRIME Master Mix, 1 gl RNA
(200 pg RNA), 0.3 gl Forward Primer (10 gM), 0.3 gl
Reverse Primer (10 gM), and water to a final volume of 15
gl. Growth of strains and RT-PCR were performed in


http://www. biomedcentral.com/1471-2164/10/470



experimental triplicate. Primer sequences used in this
analysis are available in Additional File 10.

Chlamydomonas RNA analysis
Cultures of Chlamydomonas reinhardtii wild-type strain
2137 were maintained in aerated Tris-Acetate-Phosphate
(TAP) medium with shaking in the light (60-100 gmol m-
2 s-1). To characterize Zn-responsive gene expression, cells
were initially grown to late exponential phase in TAP sup-
plemented with 2.5 gM zinc, followed by a round of
growth with no supplemental zinc, before they were inoc-
ulated into the experimental conditions at a density of 105
cells/mL. Characterization of the effects of copper, iron,
and manganese was performed similarly with the excep-
tion that cultures were grown under a second round of
metal deficiency prior to the experiment. The iron defi-
cient concentration used was 1 gM [95]. When cultures
reached mid- to late-exponential phase, total RNA was
prepared as described in [124]. cDNA preparation and
real-time PCR were performed as described in [92]with
CBLP used as the reference gene. All experiments were per-
formed in experimental triplicate. Additionally, all RT-
PCR analysis was performed in technical triplicate. Primer
sequences are available in Additional File 10. MIQE
checklist is available in Additional File 11.

List of abbreviations
MCM: methylmalonyl-CoAmutase; SVG: segmentally var-
iable gene; NHase: nitrile hydratase.

Authors' contributions
CEH carried out and designed the phylogenetic, compar-
ative genomic, and sequence analysis of COG0523 and
the G3E family, the RT-PCR analysis in A. baylyi, the
sequence analysis of the Zur-regulated back-up proteins,
the literature search and drafted the manuscript. DAR car-
ried out the analysis of putative Zur and ZntR binding
sites and B12 riboswitches, analysis of putative back-up
enzyme paralogs, and assisted in drafting the paper. JK
prepared C. reinhardtii RNA from Cu, Fe, Mn and Zn-defi-
cient cells, designed and tested primers for qRT-PCR, per-
formed the qRT- PCR and assisted in drafting the paper. JK
and DM established conditions for zinc-deficiency in
Chlamydomonas, isolated RNA, designed primers for qRT-
PCR, and assisted in drafting the paper. SSM and VDC
designed experiments, analyzed data and assisted in draft-
ing the paper. All authors read and approved the final
manuscript.










Page 16 of 21
(page number not for citation purposes)








http://www. biomedcentral.com/1471-2164/10/470


Additional material
Additional file 6
Sequence analysis of PyrC, DksA and HemB paralogs. Alignments of
Additional file 1 protein sequences encoded by genomes from Figure 4. / ...........f the
Literature reports of COG0523 expression data. three PyrC paralogs. The residues that chelate the a-metal are highlighted
Click here for file in blue while the p-metal ligands are highlighted in green. Lys1 02 serves
[http://www.biomedcentral.com/content/supplementary/1471- as a ligand for both metals ions. B, DksA paralogs. C, HemB paralogs.
2164-10-470-S1.PDF] Secondary structure as determined from the crystal structure of the
Escherichia coli homolog in each case is given (PDB identifiers: 1179
Additional file 2 (PyrC), I TIL (DksA), 1L6S (HemB)). For PyrC and DksA, the align-
Phylogeny of COG0523 subgroups. A, Each ,.1. ,, .1. ,..,,is ments show only the portion of the alignment containing the zinc binding
shaded and labeled. The branches representing proteins encoded by puta- residues. Columns containing the zinc chelating residues as determined
tive Zur-regulated genes are marked with a black square. The branches from the crystal structure are highlighted in yellow. Genome abbrevia-
representing C. reinhardtii COG0523 homologs encoded by the genes tions: EC, Escherichia coli; PE, Pseudomonas entomophila L48; PF,
induced by zinc deficiency are marked with a green square. Branches rep- Pseudomonas fluorescens Pf-5; PA, Pseudomonas aeruginosa
resenting the Pseudomonas paralogs discussed in the text are labeled. PAO1; ABa, Acinetobacter baylyi ADP1; VA, Vibrio alginolyticus
Protein IDs for each branch can be found in Additional File 7. 12G01; HC, Hahella chejuensis KCTC 2396; BCe, Burkholderia cen-
Click here for file ocepacia AU 1054; BC, Burkholderia cepacia R18194; CM, Cupria-
[http://www.biomedcentral.com/content/supplementary/1471- vidus metallidurans CH341; AB, Alcanivorax borkumensis SK2; KP,
2164-10-470-S2.PDF] Klebsiella pneumoniae MGH 78578 (on virulence plasmid pLVPK);
AV, Azotobacter vinelandii; BPe, Bordetella pertussis Tohama I; BF,
Additional file 3 Burkholderia fungorum; BM, Burkholderia malleiATCC23344; BX,
Burkholderia xenovorans LB400; BP, Burkholderia pseudomallei
Detailed description of each COG0523 subfamily. For subfamilies 2- Burkholderia xenovorans LB400; BP, Burkholderia pseudomallei
t . , 1710 Ob; RE, RalstoniaeutrophaJMP134; RS, Ralstoniasolanacearum
15, representative ..n. .... .. ... are given as is a list of the locus- GMI1000; MF, Methylobacillus flagelatus KT.
tags for members of each subfamily identified by physical clustering to Click here for file
common genes. [http://www.biomedcentral.com/content/supplementary/1471-
Click here for file 2164-10-470-S6.PDF]
[http://www.biomedcentral.com/content/supplementary/1471-
2164-10-470-S3.PDF] Additional file 7

Additional file 4 Protein IDs used in sequence analyses. Protein IDs used for various
., sequence analyses are .n .l l6i .11 1.1...1. .1. the indicated
cobW genes downstream of a putative B12 riboswitch. Presence/ sequence a analyses are se........ the indicated
absence of a putative B12riboswitch upstream of cobW is shown as well as Click here for file
the first gene in each putative cobalamin-regulated biosynthesis operon. [http www biomedcenra comconensuppemenary1471
Slocus tags r i bd B1 p l r d by [9 [http://www.biomedcentral.com/content/supplementary/1471-
CobW locus tags are in bold. B12riboswitches previously reported by [49] 2164-10-470- 7.XL S]
are marked with an asterisk.
Click here for file
[http://www.biomedcentral.com/content/supplementary/1471- Additional file 8
2164-10-470-S4.XLS] Putative ZntR-regulated genes in archaeal genomes. Candidate ZntR-
binding sites upstream of genes encoding COG0523, ZntR and ZnuABC
Additional file 5 in Archaea.
Click here for file
Putative Zur-binding sites in bacterial genomes. Candidate Zur-bind- Click here for file
[http://www.biomedcentral.com/content/supplementary/1471-
ing sites upstream of genes encoding COG0523 family proteins and par- 2164-10-470-S8.XLS] com/content/supplementa/
alogs of Zn-dependent enzymes in Bacilli, a-, f- and y-proteobacteria and
Cyanobacteria. Distance refers to the location of the putative Zur-binding
site upstream from the first gene in each operon. Additional file 9
Click here for file Sequence logos for DNA-binding motifs for candidate Zinc regulators.
[http://www.biomedcentral.com/content/supplementary/1471- The taxonomy-specific DNA motif logos were constructed using Zur- and
2164-10-470-S5.XLS] ZntR-binding sites identified for COG0523 and other zinc-responsive
genes described in the Additional Files 5 and 7.
Click here for file
[http://www.biomedcentral.com/content/supplementary/1471-
2164-10-470-S9.PDF]

Additional file 10
Primers used in transcription analyses. Sequences for primers used in
the transcription analysis of A. baylyi and C. reinhardtii COG0523
homologs.
Click here for file
[http://www.biomedcentral.com/content/supplementary/1471-
2164-10-470-S10.PDF]





Page 17 of 21
(page number not for citation purposes)


BMC Genomics 2009,10:470








BMC Genomics 2009, 10:470


Additional file 11
MIQE checklist for C. reinhardtii qRT-PCR.
Click here for file
[http://www.biomedcentral.com/content/supplementary/1471-
2164-10-470-S11.DOC]




Acknowledgements
We thank John Helmann for sparking our interest in the COG0523 family,
Ross Overbeek for generating the initial clustering analysis data in SEED,
Dirk Iwata-Reuyl for critical reading of the manuscript, Veronique de Ber-
nardinis for sending A. baylyi mutants. This work was supported by the U.S.
Department of Energy Grant (grant no. DE-FG02-07ER64498, to V. de C.-
L), National Institutes of Health (GM42143, to S.M. and I F32GM083562 to
D.M.), the Sol Leshin Program for BGU-UCLA Academic Cooperation (to
S.M.), and by the grant from the Russian Academy of Science (program
"Molecular and Cellular Biology", to D.A.R.).

References
I. Andreini C, Banci L, Bertini I, Rosato A: Zinc through the three
domains of life. J Proteome Res 2006, 5(1 I):3 173-3178.
2. Andreini C, Banci L, Bertini I, Elmi S, Rosato A: Non-heme iron
through the three domains of life. Proteins 2007, 67(2):317-324.
3. Andreini C, Banci L, Bertini I, Rosato A: Occurrence of copper
proteins through the three domains of life: a bioinformatic
approach. J Proteome Res 2008, 7(1):209-216.
4. Johnson DC, Dean DR, Smith AD, Johnson MK: Structure, func-
tion, and formation of biological iron-sulfur clusters. Annu Rev
Biochem 2005, 74:247-281.
5. De Rienzo F, Gabdoulline RR, Menziani MC, Wade RC: Blue copper
proteins: a comparative analysis of their molecular interac-
tion properties. Protein Sci 2000, 9(8):1439-1454.
6. Kurtz DM: Oxygen-carrying proteins: three solutions to a
common problem. Essays Biochem 1999, 34:85-100.
7. Finney L, O'Halloran T: Transition metal speciation in the cell:
insights from the chemistry of metal ion receptors. Science
2003, 300(5621 I):931-936.
8. Waldron KJ, Robinson NJ: How do bacterial cells ensure that
metalloproteins get the correct metal? Nat RevMicrobiol 2009,
7(1):25-35.
9. Pufahl RA, Singer CP, Peariso KL, Lin SJ, Schmidt PJ, Fahrni CJ, Culotta
VC, Penner-Hahn JE, O'Halloran TV: Metal ion chaperone func-
tion of the soluble Cu(l) receptor Atxl. Science 1997,
278(5339):853-856.
10. Kuchar J, Hausinger RP: Biosynthesis of metal sites. Chem Rev
2004, 104(2):509-525.
I I. Leach M, Zamble D: Metallocenter assembly of the hydroge-
nase enzymes. Curr Opin Chem Biol 2007, I I (2): 159-165.
12. KimJK, Mulrooney SB, Hausinger RP: Biosynthesis of active Bacil-
lus subtilis urease in the absence of known urease accessory
proteins. J Bacteriol 2005, 187(20):7150-7154.
13. Leipe DD, Wolf YI, Koonin EV, Aravind L: Classification and evo-
lution of P-loop GTPases and related ATPases. j Mol Biol 2002,
317(1):41-72.
14. Hubbard PA, Padovani D, Labunska T, Mahlstedt SA, Banerjee R,
Drennan CL: Crystal structure and mutagenesis of the metal-
lochaperone MeaB: insight into the causes of methylmalonic
aciduria. J Biol Chem 2007, 282(43):3 I1308-31316.
15. Crouzet J, Levy-Schil S, Cameron B, Cauchois L, Rigault S, Rouyez
MC, Blanche F, Debussche L, Thibaut D: Nucleotide sequence and
genetic analysis of a 13.1 -kilobase-pair Pseudomonas denitrif-
icans DNA fragment containing five cob genes and identifica-
tion of structural genes encoding Cob(l)alamin
adenosyltransferase, cobyric acid synthase, and bifunctional
cobinamide kinase-cobinamide phosphate guanylyltrans-
ferase. j Bacteriol 1991, 173(1 9):6074-6087.
16. Nojiri M, Yohda M, Odaka M, Matsushita Y, Tsujimura M, Yoshida T,
Dohmae N, Takio K, Endo I: Functional expression of nitrile
hydratase in Escherichia coli: requirement of a nitrile


http://www. biomedcentral.com/1471-2164/10/470




hydratase activator and post-translational modification of a
ligand cysteine. j Biochem 1999, 125(4):696-704.
17. Gaballa A, Helmann JD: Identification of a zinc-specific metal-
loregulatory protein, Zur, controlling zinc transport operons
in Bacillus subtilis. J Bacteriol 1998, I 80(22):5815-5821.
18. Gaballa A, Wang T, Ye RW, Helmann JD: Functional analysis of
the Bacillus subtilis Zur regulon. J Bacteriol 2002,
184(23):6508-6514.
19. Smith KF, Bibb LA, Schmitt MP, Oram DM: Regulation and activity
of a zinc uptake regulator, Zur, in Corynebacterium diphthe-
riae. J Bacteriol 2009, I 91 (5): 1595-1603.
20. Galperin MY, Koonin EV: Sources of systematic error in func-
tional annotation of genomes: domain rearrangement, non-
orthologous gene displacement and operon disruption. In Sil-
ico Biol 1998, l(I):55-67.
21. Brenner SE: Errors in genome annotation. Trends Genet 1999,
15(4):132-133.
22. Devos D, Valencia A: Intrinsic errors in genome annotation.
Trends Genet 2001, 1 7(8):429-43 1.
23. Overbeek R, Begley T, Butler RM, Choudhuri JV, Chuang HY,
Cohoon M, de Crecy-Lagard V, Diaz N, Disz T, Edwards R, et al.: The
subsystems approach to genome annotation and its use in
the project to annotate 1000 genomes. Nucleic Acids Res 2005,
33(1 7):5691-5702.
24. Glasner JD, Liss P, Plunkett G, Darling A, Prasad T, Rusch M, Byrnes
A, Gilson M, Biehl B, Blattner FR, et at.: ASAP, a systematic anno-
tation package for community analysis of genomes. Nucleic
Acids Res 2003, 31 (I):147-151.
25. Mao X, Cai T, OlyarchukJG, Wei L: Automated genome annota-
tion and pathway identification using the KEGG Orthology
(KO) as a controlled vocabulary. Bioinformatics 2005,
21(1 9):3787-3793.
26. Vallenet D, Labarre L, Rouy Z, Barbe V, Bocs S, Cruveiller S, Lajus A,
Pascal G, Scarpelli C, Medigue C: MaGe: a microbial genome
annotation system supported by synteny results. Nucleic Acids
Res 2006, 34(1):53-65.
27. Padovani D, Labunska T, Banerjee R: Energetics of interaction
between the G-protein chaperone, MeaB, and BI 2-depend-
ent methylmalonyl-CoA mutase. ] Biol Chem 2006,
281(26):17838-17844.
28. Korotkova N, Lidstrom ME: MeaB is a component of the meth-
ylmalonyl-CoA mutase complex required for protection of
the enzyme from inactivation. J Biol Chem 2004,
279(14): I 13652-13658.
29. Padovani D, Banerjee R: Assembly and protection of the radical
enzyme, methylmalonyl-CoA mutase, by its chaperone. Bio-
chemistry 2006, 45(30):9300-9306.
30. Fu C, Olson JW, Maier RJ: HypB protein of Bradyrhizobium
japonicum is a metal-binding GTPase capable of binding 18
divalent nickel ions per dimer. Proc Natl Acad Sci USA 1995,
92(6):2333-2337.
31. ZhangJW, Butland G, GreenblattJF, Emili A, Zamble DB: A role for
SlyD in the Escherichia coli hydrogenase biosynthetic path-
way. j Biol Chem 2005, 280(6):4360-4366.
32. Leach MR, Zhang JW, Zamble DB: The role of complex forma-
tion between the Escherichia coli hydrogenase accessory fac-
tors HypB and SlyD. J Biol Chem 2007, 282(22): 16177-16186.
33. Soriano A, Colpas GJ, Hausinger RP: UreE stimulation of GTP-
dependent urease activation in the UreD-UreF-UreG-ure-
ase apoprotein complex. Biochemistry 2000,
39(40): 12435-12440.
34. Remaut H, Safarov N, Ciurli S, Van Beeumen J: Structural basis for
Ni(2+) transport and assembly of the urease active site by
the metallochaperone UreE from Bacillus pasteurii. J Biol
Chem 2001, 276(52):49365-49370.
35. Song HK, Mulrooney SB, Huber r H RP: Crystal structure
of Klebsiella aerogenes UreE, a nickel-binding metallochaper-
one for urease activation. j Biol Chem 2001,
276(52):49359-49364.
36. The SEED: an Annotation/Analysis Tool Provided by FIG
[http://theseed.uchicago.edu/FIG/index.cgi]
37. Zambelli B, Musiani F, Savini M, Tucker P, Ciurli S: Biochemical
studies on Mycobacterium tuberculosis UreG and compara-
tive modeling reveal structural and functional conservation
among the bacterial UreG family. Biochemistry 2007,
46(1 1):3171-3182.



Page 18 of 21
(page number not for citation purposes)








BMC Genomics 2009, 10:470


38. Olson JW, Mehta NS, Maier RJ: Requirement of nickel metabo-
lism proteins HypA and HypB for full activity of both hydro-
genase and urease in Helicobacter pylori. Mol Microbiol 2001,
39(1):176-182.
39. Stingl K, Schauer K, Ecobichon C, Labigne A, Lenormand P, Rousselle
JC, Namane A, de Reuse H: In vivo interactome of Helicobacter
pylori urease revealed by tandem affinity purification. Mol Cell
Proteomics 2008, 7(12):2429-2441.
40. Benoit S, Maier RJ: Dependence of Helicobacter pylori urease
activity on the nickel-sequestering ability of the UreE acces-
sory protein. j Bacteriol 2003, 185(16):4787-4795.
41. Mehta N, Benoit S, Maier RJ: Roles of conserved nucleotide-bind-
ing domains in accessory proteins, HypB and UreG, in the
maturation of nickel-enzymes required for efficient Helico-
bacter pylori colonization. Microb Pathog 2003, 35(5):229-234.
42. Olson JW, Maier RJ: Dual roles of Bradyrhizobium japonicum
nickelin protein in nickel storage and GTP-dependent Ni
mobilization. j Bacteriol 2000, 182(6): 1702-1705.
43. Khil PP, Obmolova G, Teplyakov A, Howard AJ, Gilliland GL, Cam-
erini-Otero RD: Crystal structure of the Escherichia coli YjiA
protein suggests a GTP-dependent regulatory function. Pro-
teins 2004, 54(2):371-374.
44. Hashimoto Y, Nishiyama M, Horinouchi S, Beppu T: Nitrile
hydratase gene from Rhodococcus sp. N-774 requirement for
its downstream region for efficient expression. Biosci Biotechnol
Biochem 1994, 58(10): 1859-1865.
45. Lu J, Zheng Y, Yamagishi H, Odaka M, Tsujimura M, Maeda M, Endo I:
Motif CXCC in nitrile hydratase activator is critical for
NHase biogenesis in vivo. FEBS Lett 2003, 553(3):391-396.
46. Soriano A, Hausinger RP: GTP-dependent activation of urease
apoprotein in complex with the UreD, UreF, and UreG
accessory proteins. Proc Natl Acad Sci USA 1999,
96(20):1 I140-1 1 144.
47. Padovani D, Labunska T, Palfey BA, Ballou DP, Banerjee R: Adeno-
syltransferase tailors and delivers coenzyme Bi2. Nat Chem
Biol 2008, 4(3): 194-196.
48. Zheng Y, Roberts RJ, Kasif S: Segmentally variable genes: a new
perspective on adaptation. PLoS Biol 2004, 2(4):E8 1.
49. Rodionov DA, Vitreschak AG, Mironov AA, Gelfand MS: Compara-
tive genomics of the vitamin Bi2 metabolism and regulation
in prokaryotes. j Biol Chem 2003, 278(42):41148-41159.
50. Schroeder S, Lawrence AD, Biedendieck R, Rose RS, Deery E, Gra-
ham RM, McLean KJ, Munro AW, Rigby SE, Warren MJ: Demonstra-
tion that CobG, the monooxygenase associated with the ring
contraction process of the aerobic cobalamin (vitamin Bi2)
biosynthetic pathway, contains an Fe-S center and a mono-
nuclear non-heme iron center. J Biol Chem 2009,
284(8):4796-4805.
51. Mascharak PK: Structural and functional models of nitrile
hydratase. Coord Chem Rev 2002, 225(1-2):201-214.
52. Banerjee A, Sharma R, Banerjee UC: The nitrile-degrading
enzymes: current status and future prospects. AppI Microbiol
Biotechnol 2002, 60(1 -2):33-44.
53. Endo I, Nojiri M, Tsujimura M, Nakasako M, Nagashima S, Yohda M,
Odaka M: Fe-type nitrile hydratase. J Inorg Biochem 2001,
83(4):247-253.
54. Zhou Z, Hashimoto Y, Shiraki K, Kobayashi M: Discovery of post-
translational maturation by self-subunit swapping. Proc Natt
Acad Sci USA 2008, 105(39):14849-14854.
55. Zhou Z, Hashimoto Y, Kobayashi M: Self-subunit swapping chap-
erone needed for the maturation of multimeric metalloen-
zyme nitrile hydratase by a subunit exchange mechanism
also carries out the oxidation of the metal ligand cysteine
residues and insertion of cobalt. j Biol Chem 2009,
284(22): 14930-14938.
56. Nojiri M, Nakayama H, Odaka M, Yohda M, Takio K, Endo I: Cobalt-
substituted Fe-type nitrile hydratase of Rhodococcus sp. N-
771. FEBS Lett 2000, 465(2-3): 173-177.
57. Stevens JM, Rao Saroja N,Jaouen M, Belghazi M, SchmitterJM, Mansuy
D, Artaud I, Sari MA: Chaperone-assisted expression, purifica-
tion, and characterization of recombinant nitrile hydratase
Nil from Comamonas testosteroni. Protein Expr Purif 2003,
29(1):70-76.
58. Kochan I: The role of iron in bacterial infections, with special
consideration of host-tubercle bacillus interaction. Curr Top
Microbiol Immunol 1973, 60:1-30.


http://www. biomedcentral.com/1471-2164/10/470




59. Weinberg E: Metal starvation of pathogens by hosts. Bioscience
1975, 25:314-318.
60. Liuzzi JP, Lichten LA, Rivera S, Blanchard RK, Aydemir TB, Knutson
MD, Ganz T, Cousins RJ: Interleukin-6 regulates the zinc trans-
porter Zip 14 in liver and contributes to the hypozincemia of
the acute-phase response. Proc Natl Acad Sci USA 2005,
102(1 9):6843-6848.
61. Motley ST, Morrow BJ, Liu X, Dodge IL, Vitiello A, Ward CK, Shaw
KJ: Simultaneous analysis of host and pathogen interactions
during an in vivo infection reveals local induction of host
acute phase response proteins, a novel bacterial stress
response, and evidence of a host-imposed metal ion limited
environment. Cell Microbiol 2004, 6(9):849-865.
62. Panina EM, Mironov AA, Gelfand MS: Comparative genomics of
bacterial zinc regulons: enhanced ion transport, pathogene-
sis, and rearrangement of ribosomal proteins. Proc Natl Acad
Sci USA 2003, 100(1 7):9912-9917.
63. Kim S, Watanabe K, Shirahata T, Watarai M: Zinc uptake system
(znuA locus) of Brucella abortus is essential for intracellular
survival and virulence in mice. j Vet Med Sci 2004,
66(9): 1059-1063.
64. Pasquali P, Ammendola S, Pistoia C, Petrucci P, Tarantino M, Valente
C, Marenzoni ML, Rotilio G, Battistoni A: Attenuated Salmonella
enterica serovar Typhimurium lacking the ZnuABC trans-
porter confers immune-based protection against challenge
infections in mice. Vaccine 2008, 26(27-28):3421-3426.
65. Maciag A, Dainese E, Rodriguez GM, Milano A, Provvedi R, Pasca MR,
Smith I, Pald G, Riccardi G, Manganelli R: Global analysis of the
Mycobacterium tuberculosis Zur (FurB) regulon. J Bacteriol
2007, 189(3):730-740.
66. Cappelli G, Volpe E, Grassi M, Liseo B, Colizzi V, Mariani F: Profiling
of Mycobacterium tuberculosis gene expression during human
macrophage infection: upregulation of the alternative sigma
factor G, a group of transcriptional regulators, and proteins
with unknown function. Res Microbiol 2006, 157(5):445-455.
67. Stratmann J, Strommenger B, Goethe R, Dohmann K, Gerlach GF,
Stevenson K, Li LL, Zhang Q, Kapur V, Bull TJ: A 38-kilobase path-
ogenicity island specific for Mycobacterium avium subsp.
paratuberculosis encodes cell surface proteins expressed in
the host. Infect Immun 2004, 72(3): 1265-1274.
68. Bannantine JP, Paustian ML, Waters WR, Stabel JR, Palmer MV, Li L,
Kapur V: Profiling bovine antibody responses to Mycobacte-
rium avium subsp. paratuberculosis infection by using protein
arrays. Infect Immun 2008, 76(2):739-749.
69. McBride SM, Coburn PS, Baghdayan AS, Willems RJ, Grande MJ,
Shankar N, Gilmore MS: Genetic variation and evolution of the
pathogenicity island of Enterococcus faecalis. J Bacteriol 2009,
191(10):3392-3402.
70. Kohler S, Foulongne V, Ouahrani-Bettache S, Bourg G, Teyssier J,
Ramuz M, Liautard JP: The analysis of the intramacrophagic vir-
ulome of Brucella suis deciphers the environment encoun-
tered by the pathogen inside the macrophage host cell. Proc
Natl Acad Sci USA 2002, 99(24): 1571 I -15716.
71. Gan YH, Chua KL, Chua HH, Liu B, Hii CS, Chong HL, Tan P: Char-
acterization of Burkholderia pseudomallei infection and iden-
tification of novel virulence factors using a Caenorhabditis
elegans host system. Mol Microbiol 2002, 44(5): 1185-1197.
72. Twine SM, Mykytczuk NC, Petit MD, Shen H, Sjostedt A, Wayne
Conlan Kelly JF: In vivo proteomic analysis of the intracellular
bacterial pathogen, Francisella tularensis, isolated from
mouse spleen. Biochem Biophys Res Commun 2006,
345(4):1621-1633.
73. Mohapatra NP, Soni S, Bell BL, Warren R, Ernst RK, Muszynski A,
Carlson RW, Gunn JS: Identification of an orphan response reg-
ulator required for the virulence of Francisella spp. and tran-
scription of pathogenicity island genes. Infect Immun 2007,
75(7):3305-3314.
74. Tang DJ, Li XJ, He YQ, Feng JX, Chen B, Tang JL: The zinc uptake
regulator Zur is essential for the full virulence of Xan-
thomonas campestris pv. campestris. Mol Plant Microbe Interact
2005, 18(7):652-658.
75. Huang DL, Tang DJ, Liao Q, Li HC, Chen Q, He YQ, Feng JX, Jiang
BL, Lu GT, Chen B, et al.: The Zur of Xanthomonas campestris
functions as a repressor and an activator of putative zinc
homeostasis genes via recognizing two distinct sequences




Page 19 of 21
(page number not for citation purposes)








BMC Genomics 2009, 10:470


within its target promoters. Nucleic Acids Res 2008,
36(13):4295-4309.
76. Yang W, Liu Y, Chen L, Gao T, Hu B, Zhang D, Liu F: Zinc uptake
regulator (zur) gene involved in zinc homeostasis and viru-
lence of Xanthomonas oryzae pv. oryzae in rice. Curr Microbiol
2007, 54(4):307-314.
77. Yuan ZC, Haudecoeur E, Faure D, Kerr KF, Nester EW: Compara-
tive transcriptome analysis of Agrobacterium tumefaciens in
response to plant signal salicylic acid, indole-3-acetic acid
and gamma-amino butyric acid reveals signalling cross-talk
and Agrobacterium--plant co-evolution. Cell Microbiol 2008,
10(1 1):2339-2354.
78. Rodionov DA: Comparative genomic reconstruction of tran-
scriptional regulatory networks in bacteria. Chem Rev 2007,
107(8):3467-3497.
79. Sankara SB, Shah K, Gabriel S, Reddy R, Schimmel P, Rodionov DA,
de V, Crecy-Lagard Helmann JD, Iwata-Reuyl D, Swairjo MA: Zinc-
independent folate biosynthesis: Genetic, biochemical, and
structural investigations reveal new metal dependence for
GTP Cyclohydrolase IB. journal of Bacteriology in press.
80. Tripp BC, Bell CB, Cruz F, Krebs C, Ferry JG: A role for iron in an
ancient carbonic anhydrase. ] Biol Chem 2004,
279(8):6683-6687.
81. Macauley SR, Zimmerman SA, Apolinario EE, Evilia C, Hou YM, Ferry
JG, Sowers KR: The archetype gamma-class carbonic anhy-
drase (Cam) contains iron when synthesized in vivo. Biochem-
istry 2009, 48(5):817-819.
82. Thoden J, Phillips GJ, Neal T, Raushel F, Holden H: Molecular struc-
ture of dihydroorotase: a paradigm for catalysis through the
use of a binuclear metal center. Biochemistry 2001,
40(24):6989-6997.
83. Schurr M, Vickrey J, Kumar A, Campbell A, Cunin R, Benjamin R,
Shanley M, O'Donovan G: Aspartate transcarbamoylase genes
of Pseudomonas putida: requirement for an inactive dihy-
droorotase for assembly into the dodecameric holoenzyme.
j Bacteriol 1995, 177(7): 1751-1759.
84. Brichta DM, Azad KN, Ralli P, O'Donovan GA: Pseudomonas aeru-
ginosa dihydroorotases: a tale of three pyrCs. Arch Microbiol
2004, 182(1):7-17.
85. Jaffe E: An unusual phylogenetic variation in the metal ion
binding sites of porphobilinogen synthase. Chem Biol 2003,
10(1):25-34.
86. Pfalz J, Liere K, Kandlbinder A, Dietz KJ, Oelmuller R: pTAC2, -6,
and -I 2 are components of the transcriptionally active plas-
tid chromosome that are required for plastid gene expres-
sion. Plant Cell 2006, 18(1):176-197.
87. TAIR: The Arabidopsis Information Resource. [http://
www.arabidopsis.org].
88. Kumar A, Cheung KH, Tosches N, Masiar P, Liu Y, Miller P, Snyder
M: The TRIPLES database: a community resource for yeast
molecular biology. Nucleic Acids Res 2002, 30(1):73-75.
89. de Jesus Ferreira MC, Bao X, Laize V, Hohmann S: Transposon
mutagenesis reveals novel loci affecting tolerance to salt
stress and growth at low temperature. Curr Genet 2001,
40(1):27-39.
90. Smith TF, Gaitatzes C, Saxena K, Neer EJ: The WD repeat: a com-
mon architecture for diverse functions. Trends Biochem Sci
1999, 24(5):181-185.
91. Merchant S, Bogorad L: Rapid degradation of apoplastocyanin in
Cu(ll)-deficient cells of Chlamydomonas reinhardtii. J Biol Chem
1986, 261(34): 15850-15853.
92. Allen MD, Kropat J, Tottey S, Del Campo JA, Merchant SS: Manga-
nese deficiency in Chlamydomonas results in loss of photosys-
tem II and MnSOD function, sensitivity to peroxides, and
secondary phosphorus and iron deficiency. Plant Physiol 2007,
143(1):263-277.
93. Merchant SS, Allen MD, KropatJ, MoseleyJL, Long JC, Tottey S, Ter-
auchi AM: Between a rock and a hard place: trace element
nutrition in Chlamydomonas. Biochim Biophys Acta 2006,
1763(7):578-594.
94. Quinn JM, Merchant S: Two copper-responsive elements asso-
ciated with the Chlamydomonas Cyc6 gene function as tar-
gets for transcriptional activators. Plant Cell 1995, 7(5):623-628.
95. Allen MD, del Campo JA, Kropat J, Merchant SS: FEAI, FEA2, and
FRE I, encoding two homologous secreted proteins and a
candidate ferrireductase, are expressed coordinately with


http://www. biomedcentral.com/1471-2164/10/470




FOXI and FTRI in iron-deficient Chlamydomonas reinhardtii.
Eukaryot Cell 2007, 6(10): 1841 -1852.
96. Merchant SS, Prochnik SE, Vallon 0, Harris EH, Karpowicz SJ, Wit-
man GB, Terry A, Salamov A, Fritz-Laylin LK, Marechal-Drouard L, et
al.: The Chlamydomonas genome reveals the evolution of key
animal and plant functions. Science 2007, 318(5848):245-250.
97. Grossman AR, Croft M, Gladyshev VN, Merchant SS, Posewitz MC,
Prochnik S, Spalding MH: Novel metabolism in Chlamydomonas
through the lens of genomics. Curr Opin Plant Biol 2007,
10(2):190-198.
98. Croft MT, Warren MJ, Smith AG: Algae need their vitamins.
Eukaryot Cell 2006, 5(8): I 175-1 183.
99. Merchant S, Bogorad L: Regulation by copper of the expression
of plastocyanin and cytochrome c552 in Chlamydomonas rein-
hardi. Mol Cell Biol 1986, 6(2):462-469.
100. Sauer K, Thauer R: Methanol:coenzyme M methyltransferase
from Methanosarcina barker. Zinc dependence and thermo-
dynamics of the methanol:cob(I)alamin methyltransferase
reaction. Eur Biochem 1997, 249(l):280-285.
101. Sauer K, Harms U, Thauer RK: Methanol:coenzyme M methyl-
transferase from Methanosarcina barker. Purification, prop-
erties and encoding genes of the corrinoid protein MTI. Eur
J Biochem 1997, 243(3):670-677.
102. Ferguson T, Soares JA, Lienard T, Gottschalk G, Krzycki JA: RamA,
a protein required for reductive activation of corrinoid-
dependent methylamine methyltransferase reactions in
methanogenic archaea. J Biol Chem 2009, 284(4):2285-2295.
103. Li Q, Li L, Rejtar T, Karger BL, Ferry JG: Proteome of Methanosa-
rcina acetivorans Part I: an expanded view of the biology of
the cell. J Proteome Res 2005, 4(l):112-128.
104. Pfluger K, Ehrenreich A, Salmon K, Gunsalus RP, Deppenmeier U,
Gottschalk G, Muller V: Identification of genes involved in salt
adaptation in the archaeon Methanosarcina mazei GoI using
genome-wide gene expression profiling. FEMS Microbiol Lett
2007, 277(1):79-89.
105. Gabriel SE, Miyagi F, Gaballa A, Helmann JD: Regulation of the
Bacillus subtilis yciC gene and insights into the DNA-binding
specificity of the zinc-sensing metalloregulator Zur. J Bacteriol
2008, 190(10):3482-3488.
106. Eide D: Zinc transporters and the cellular trafficking of zinc.
Biochim Biophys Acta 2006, 1763(7):71 1-722.
107. Campos-Bermudez VA, Leite NR, Krog R, Costa-Filho AJ, Soncini FC,
Oliva G, Vila AJ: Biochemical and structural characterization
of Salmonella typhimurium glyoxalase II: new insights into
metal ion selectivity. Biochemistry 2007, 46(39):l 1069-1 1079.
108. Cimara B, Marin M, Schlomann M, Hecht HJ, Junca H, Pieper DH:
trans-Dienelactone hydrolase from Pseudomonas reinekei
MTI, a novel zinc-dependent hydrolase. Biochem Biophys Res
Commun 2008, 376(2):423-428.
109. Hall RS, Xiang DF, Xu C, Raushel FM: N-Acetyl-D-glucosamine-6-
phosphate deacetylase: substrate activation via a single diva-
lent metal ion. Biochemistry 2007, 46(27):7942-7952.
I 10. Yee D, Morel FMM: In vivo substitution of zinc by cobalt in car-
bonic anhydrase of a marine diatom. Limnol Oceanogr 1996,
41 (3):573-577.
II I. Moroney J, Ynalvez R: Proposed carbon dioxide concentrating
mechanism in Chlamydomonas reinhardtii. Eukaryot Cell 2007,
6(8):1251-1259.
I12. Suzuki T, Okamura Y, Arakaki A, Takeyama H, Matsunaga T: Cyto-
plasmic ATPase involved in ferrous ion uptake from magne-
totactic bacterium Magnetospirillum magneticum AMB-1.
FEBS Lett 2007, 581 (18):3443-3448.
II 3. Benson DA, Karsch-Mizrachi I, Lipman DJ, Ostell J, Sayers EW: Gen-
Bank. Nucleic Acids Res 2009:D26-3 I.
I14. Rice P, Longden I, Bleasby A: EMBOSS: the European Molecular
Biology Open Software Suite. Trends Genet 2000,
16(6):276-277.
I15. Thompson JD, Gibson TJ, Higgins DG: Multiple sequence align-
ment using ClustalW and ClustalX. Curr Protoc Bioinformatics
2002, Chapter 2(Unit 2.3):.
I16. Gouet P, Courcelle E, Stuart D, Metoz F: ESPript: analysis of mul-
tiple sequence alignments in PostScript. Bioinformatics 1999,
I 5(4):305-308.
117. PHYLIP (Phylogeny Inference Package) version 3.67 [http://
evolution.genetics.washington.edu/phylip.html]




Page 20 of 21
(page number not for citation purposes)








http://www. biomedcentral.com/1471-2164/10/470


118. Overbeek R, Fonstein M, D'Souza M, Pusch G, Maltsev N: The use
of gene clusters to infer functional coupling. Proc Nat! Acad Sci
USA 1999, 96(6):2896-2901.
119. Chevenet F, Brun C, Bahuls A, Jacq B, Christen R: TreeDyn:
towards dynamic graphics and annotations for analyses of
trees. BMC Bioinformatics 2006, 7:439.
120. Waterhouse AM, Procter JB, Martin DM, Clamp M, Barton GJ:
jalview Version 2 a multiple sequence alignment editor and
analysis workbench. Bioinformatics 2009, 25(9): 1189-1 191.
121. Mironov AA, Koonin EV, Roytberg MA, Gelfand MS: Computer
analysis of transcription regulatory patterns in completely
sequenced bacterial genomes. Nucleic Acids Res 1999,
27(1 4):2981-2989.
122. Crooks G, Hon G, Chandonia J, Brenner S: WebLogo: a sequence
logo generator. Genome Res 2004, 14(6): 188-1 190.
123. de Berardinis V, Vallenet D, Castelli V, Besnard M, Pinet A, Cruaud C,
Samair S, Lechaplais C, Gyapay G, Richez C, et al.: A complete col-
lection of single-gene deletion mutants of Acinetobacter bay-
lyi ADP I. Mol Syst Biol 2008, 4:174.
124. Quinn JM, Merchant S: Copper-responsive gene expression dur-
ing adaptation to copper deficiency. Methods Enzymol 1998,
297:263-279.


Page 21 of 21
(page number not for citation purposes)


Publish with BioMed Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical research in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours you keep the copyright
Submit your manuscript here: BioMedcentral
http://www.biomedcentral.com/info/publishing adv.asp


BMC Genomics 2009,10:470




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs