INFECTION AND IMMUNITY, June 2008, p. 2793–2801
Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Vol. 76, No. 6
Candida albicans Uses Multiple Mechanisms To Acquire the Essential
Metabolite Inositol during Infection?
Ying-Lien Chen, Sarah Kauffman, and Todd B. Reynolds*
Department of Microbiology, University of Tennessee, Knoxville, Tennessee 37996
Received 14 November 2007/Returned for modification 19 December 2007/Accepted 28 January 2008
Candida albicans is an important cause of life-threatening systemic bloodstream infections in immunocom-
promised patients. In order to cause infections, C. albicans must be able to synthesize the essential metabolite
inositol or acquire it from the host. Based on the similarity of C. albicans to Saccharomyces cerevisiae, it was
predicted that C. albicans may generate inositol de novo, import it from the environment, or both. The C.
albicans inositol synthesis gene INO1 (orf19.7585) and inositol transporter gene ITR1 (orf19.3526) were each
disrupted. The ino1?/ino1? mutant was an inositol auxotroph, and the itr1?/itr1? mutant was unable to
import inositol from the medium. Each of these mutants was fully virulent in a mouse model of systemic
infection. It was not possible to generate an ino1?/ino1? itr1?/itr1? double mutant, suggesting that in the
absence of these two genes, C. albicans could not acquire inositol and was nonviable. A conditional double
mutant was created by replacing the remaining wild-type allele of ITR1 in an ino1?/ino1? itr1?/ITR1 strain
with a conditionally expressed allele of ITR1 driven by the repressible MET3 promoter. The resulting ino1?/
ino1? itr1?/PMET3::ITR1 strain was found to be nonviable in medium containing methionine and cysteine
(which represses the PMET3promoter), and it was avirulent in the mouse model of systemic candidiasis. These
results suggest a model in which C. albicans has two equally effective mechanisms for obtaining inositol while
in the host. It can either generate inositol de novo through Ino1p, or it can import it from the host through
Candida albicans is a dimorphic yeast that can exist in a
human host either as a harmless commensal or as an oppor-
tunistic pathogen when the host’s immune system is impaired
(5). In patients that are neutropenic, Candida species are as-
sociated with severe and deadly systemic bloodstream infec-
tions with a high mortality rate (31.8%) (27). In fact, Candida
species are the fourth most common cause of catheter-related
bloodstream infections in hospitalized patients in intensive
care units, and C. albicans is the species most commonly iso-
lated from patients with these infections (4).
The ability of any pathogenic microbe to cause an infection
depends on the organism’s ability to acquire or generate es-
sential nutrients during residence within the host. There are
several examples of auxotrophies that compromise the viru-
lence of C. albicans (16). The most notable example is the
defect in uracil biosynthesis caused by a mutation in URA3.
Homozygous URA3 mutations (ura3?/ura3?) in C. albicans
block the ability to grow without uridine supplementation and
lead to avirulence. In fact, the level of URA3-encoded oroti-
dine 5?-monophosphate decarboxylase activity can be corre-
lated with growth and virulence (20). Homozygous mutations
in the ADE2 gene that block the ability of C. albicans to grow
in minimal medium (MM) or serum in the absence of exoge-
nous adenine also diminish virulence (10). In addition, ho-
mozygous mutations in the HEM3 gene leading to heme auxo-
trophy compromise virulence (16).
In contrast to these results, auxotrophies for a number of
amino acids, including serine, lysine, leucine, histidine, and
arginine, do not appear to compromise virulence (23, 32, 40).
For example, Noble and Johnson (32) demonstrated that C.
albicans his1?/his1?, his1?/his1? leu2?/leu2?, and his1?/his1?
arg4?/arg4? auxotrophic mutants exhibit wild-type virulence in
a mouse model of systemic candidiasis, and his1?/his1? leu2?/
leu2? arg4?/arg4? triply auxotrophic strains are only mildly
attenuated in virulence compared to the wild type. These re-
sults suggest that there are some metabolites, such as amino
acids, that C. albicans can scavenge from the host during an
infection and that there are other metabolites, such as uracil,
adenine, and heme, that it must make de novo.
myo-inositol (referred to hereinafter as inositol) is essential
for the growth of all eukaryotes and is needed by some bacteria
(12). Inositol is involved in many intracellular processes, in-
cluding growth regulation, membrane structure formation, os-
motolerance, signal transduction, and the formation of glyco-
sylphosphatidylinositol (GPI)-anchored proteins, which are
themselves essential (3, 7, 33, 41). Inositol has been implicated
in the pathogenicity of C. albicans because it is an essential
precursor of phospholipomannan, a GPI-anchored glycolipid
on the surfaces of Candida cells that is involved in pathoge-
Three potential sources of inositol available to C. albicans can
be suggested based on work that has been done with the related
yeast Saccharomyces cerevisiae: (i) de novo biosynthesis, in which
glucose 6-phosphate is converted into inositol 1-phosphate by the
inositol 1-phosphate synthase enzyme; (ii) inositol import from
the extracellular environment by an inositol transporter; and (iii)
the recycling of inositol from the dephosphorylation of inositol
polyphosphate products (7, 30, 41).
* Corresponding author. Mailing address: Department of Microbi-
ology, University of Tennessee, F321 Walters Life Sciences Building,
Knoxville, TN 37996. Phone: (865) 974-4025. Fax: (865) 974-4007.
?Published ahead of print on 11 February 2008.
In S. cerevisiae, de novo inositol biosynthesis and import
have both been well studied. The S. cerevisiae inositol 1-phos-
phate synthase is encoded by the ScINO1 gene (9). The Ino1p
enzyme converts glucose 6-phosphate into inositol 1-phos-
phate, and then a second enzyme encoded by the INM1 gene,
inositol monophosphatase, dephosphorylates inositol 1-phos-
phate to create inositol (22). S. cerevisiae carries two distinct
inositol transporter genes, S. cerevisiae ITR1 (ScITR1) and
ScITR2, the products of which are both capable of importing
inositol but at different efficiencies (31).
C. albicans carries a gene determined on the basis of se-
quence similarity to be a putative homolog of INO1; this gene
(GenBank accession no. L22737) was identified and sequenced
by Klig et al. (17, 18). C. albicans INO1 (CaINO1) shows 64%
identity to ScINO1 at the amino acid level (18). In addition, it
has been shown previously that C. albicans has potent inositol
transporter activity (15). It was demonstrated that the inositol
transporter in C. albicans has an apparent Kmof 240 ? 15 ?M,
and the transport system appears to be active and energy
dependent (15). C. albicans inositol transport activity differs in
substrate and cotransporter specificities from the human ino-
sitol-Na?transporter. However, the inositol transporter gene
in C. albicans was not identified in this previous study.
Recently, the importance of inositol biosynthesis in some
human infectious agents was demonstrated by studies of two
unrelated pathogens, Mycobacterium tuberculosis and Trypano-
soma brucei. In T. brucei, the INO1 homolog was shown to be
necessary for inositol biosynthesis, efficient GPI synthesis, and
growth in vitro (24, 25). It was found that the M. tuberculosis
INO1 homolog is required for de novo inositol biosynthesis
and full virulence in a mouse model of infection (29).
It is unknown what roles inositol import or biosynthesis play
in the virulence of C. albicans. This fungus may be able both to
synthesize and to import inositol. It became of interest to
determine if C. albicans would behave like M. tuberculosis and
T. brucei and require INO1 function for virulence and/or via-
bility or if, in contrast to these pathogens, C. albicans would be
able to utilize either de novo synthesis or inositol import to
support virulence thereafter, the C. albicans INO1 and ITR1
homologs are referred to as INO1 and ITR1, respectively.
MATERIALS AND METHODS
Strains and growth media. C. albicans strains used in this study are shown in
Table 1. The media used in this study include 1% yeast extract–2% peptone–2%
glucose (YPD) (42), defined medium 199 (Invitrogen), yeast carbon base-bovine
serum albumin (39), and MM (0.67% Difco yeast nitrogen base without amino
acids, 2% glucose) (42). MM was supplemented with 75 ?M inositol for the
growth of the C. albicans ino1?/ino1? mutant and 2.5 mM (each) methionine
and cysteine for MET3 promoter shutoff assays (6). The ino1?/ino1? itr1?
PMET3::ITR2-NAT1 strain was selected on MM medium containing no methi-
onine and cysteine, 75 ?M inositol, and 1 mg/ml nourseothricin. YPD containing
250 ?g/ml nourseothricin was used to select for other transformants. Inositol-
free medium (42) was made for testing the phenotype of the ino1?/ino1? mutant.
Agar plates were solidified with 2% agar (granulated; Fisher) for YPD and with
2% Bacto agar (which contains no residual inositol) for MM and inositol-free
Strain construction. The INO1 gene was disrupted by using the CaNAT1-FLP
cassette (39), whereas the ITR1 gene was disrupted by using the SAT1 flipper
For the INO1 disruption construct, the 564-bp 5? noncoding region (NCR) of
INO1 (5? INO1NCR) was amplified with primers JCO39 and JCO40 (Table 2),
which introduced KpnI and ApaI restriction sites, and was cloned into pJK863 in
the 5? direction from the CaNAT1-FLP cassette (see Fig. 1A). The 583-bp 3?
INO1NCRwas amplified with primers JCO41 and JCO42, which introduced SacII
and SacI sites, and was cloned into pJK863 in the 3? direction from the CaNAT1-
FLP cassette (see Fig. 1A). This procedure created the INO1 knockout construct
plasmid pYLC94 (Table 3; see Fig. 1A), which was cut with KpnI and SacI to
release the disruption construct, and the wild-type SC5314 strain was trans-
formed with the disruption construct by electroporation as described previously
(8, 35). The disruption construct was used to sequentially disrupt both alleles of
INO1 as previously described (39). The INO1 reconstitution construct was made
by amplifying a 2.1-kb fragment containing the INO1 open reading frame (ORF)
and the 5? NCR from SC5314 genomic DNA by using primers (JCO39 and
JCO47) that introduced KpnI and SalI sites. This fragment was ligated into the
pRS316 vector, along with a 1.7-kb fragment containing the NAT1-3? INO1NCR
fragment amplified from plasmid pYLC94 by using primers JCO50 and JCO42,
which introduced SalI and SacI sites. This procedure resulted in the INO1
reconstitution plasmid pYLC119 (Table 3; see Fig. 1B). The ino1?/ino1? mutant
YLC113 was transformed with the 3.8-kb KpnI-SacI fragment from pYLC119 in
order to create the reconstituted-INO1 (ino1?/ino1?::INO1) strain YLC120.
A similar approach was used to knock out ITR1 with the SAT1 flipper plasmid
pSFS2A (35). Approximately-500-bp 5? and 3? NCRs from ITR1 (5? ITR1NCR
and 3? ITR1NCR) were amplified using the primer pairs JCO89-JCO90 and
JCO93-JCO94, respectively, each of which introduced restriction sites into the
corresponding fragments. The ApaI-XhoI 5? ITR1NCRand NotI 3? ITR1NCR
fragments were ligated into pSFS2A, resulting in the ITR1 knockout construct
plasmid pYLC164 (Table 3; see Fig. 2A). The 5-kb ApaI-SacII fragment from
pYLC164 was introduced into SC5314 by electroporation and used to disrupt
both copies of ITR1 by sequential disruption steps (28, 35). For the reconstituted-
ITR1 construct, JCO89 and JCO111 primers, which added ApaI and EcoRI sites,
were used to amplify a 2-kb fragment containing the 5? ITR1NCRand the ITR1
ORF region, which was used to replace the ApaI-EcoRI fragment (containing
the 5? ITR1NCR-PMAL2-FLP fragment) in the pYLC164 plasmid, resulting in
pYLC208 (Table 3; see Fig. 2B). The 5-kb ApaI-SacII fragment from pYLC208
was used to reconstitute ITR1 in the itr1?/itr1? mutant strain YLC196, creating
the itr1?/itr1?::ITR1 strain YLC211.
The ITR1 conditional mutant was made as follows. With SC5314 genomic
DNA as the template, primers JCO118 and JCO119 (which introduced PstI and
NotI sites) were used to amplify a 1,362-bp fragment containing the CaMET3
promoter (6), and primers JCO120 and JCO121 (which introduced NotI and
TABLE 1. C. albicans strains
Description or genotype
Prototrophic wild type11
YLC181YLC113 This study
aAn ino1?/ino1? homozygous mutant derived separately from the other
bAn ino1?/ino1? itr1?/PMET3::ITR1 conditional strain derived separately
from the other conditional strain.
2794 CHEN ET AL.INFECT. IMMUN.
SacII sites) were used to amplify the ITR1 ORF. These fragments were used to
replace the PstI-SacII fragment in pYLC164, which deleted the 3?-end-flanking
FLP recombination target (FRT) site, resulting in the conditional ITR1 expres-
sion construct plasmid pYLC229 (Table 3; see Fig. 2C). The ino1?/ino1? itr1?/
ITR1 strain YLC184 was transformed with the 7.6-kb ApaI-SacII fragment from
pYLC229 to create the ino1?/ino1? itr1?/PMET3::ITR1 conditional strains
YLC261 and YLC266.
Northern blot analysis. Northern blotting for ITR1 expression was performed
as described previously (36), with the following exceptions. Strains grown in
liquid medium 199 at 37°C for 2 h were collected for total RNA extraction by the
hot-phenol method (36). A PCR product containing bp 24 to 750 of the ITR1
ORF (amplified with primers JCO102 and JCO103) was used as a probe. Ex-
pression was normalized against the expression of the CaACT1 gene, probed for
on the same membrane. The CaACT1 probe was generated with the primers
JCO48 and JCO49.
Southern blot analysis. Hybridization conditions for the Southern blot analysis
were similar to those for the Northern blot analysis, except that the Techne
Hybrigene oven was set to 60°C for the incubation step and 42 and 60°C for the
washing steps. The cells were grown in liquid YPD at 30°C overnight. The
genomic DNA was extracted using the Winston-Hoffman method (13), and 20 ?g
of genomic DNA was subjected to Southern blotting. The genomic DNA of ino1
mutants was cut by AflII and SphI restriction enzymes, while the genomic DNA
of itr1 mutants was cut by PstI. PCR products containing the ?500-bp 3?
INO1NCR(amplified with primers JCO41 and JCO42) and the 3? ITR1NCR
(amplified with primers JCO93 and JCO94) were used as probes.
Inositol uptake assays. The inositol uptake assay protocol was adapted in part
from a protocol of Jin and Seyfang (15). The wild-type, itr1?/ITR1, itr1?/itr1?,
and itr1?/itr1?::ITR1 strains were grown in YPD liquid cultures overnight at
30°C. Cells were diluted in YPD to an optical density at 600 nm of 1, grown at
30°C, and collected at an optical density at 600 nm of 5 by centrifugation at
2,600 ? g for 5 min. Cells were then washed twice with water at 4°C and
resuspended in 2% glucose to a final concentration of 2 ? 108cells/ml as
determined by a hemacytometer. From this time point on, cells were kept on ice
until being used for the actual assay. For the uptake assay, the reaction mixture
(250 ?l) contained 2% glucose, 40 mM citric acid-KH2PO4(pH 5.5), 0.15 ?M
[2-3H]inositol (1 ?Ci/?l; MP Biomedicals), and 200 ?M unlabeled inositol
(Alexis Biomedicals). Equal volumes of the reaction and cell mixtures (60 ?l
each) were warmed to 30°C and mixed for the uptake assay, which was per-
formed for 10 min at 30°C. As negative controls, mixtures were kept at 0°C (on
ice) during the 10-min incubation. Aliquots of 100 ?l were removed and trans-
ferred onto prewetted Metricel filters on a vacuum manifold. The filters were
washed four times each with 1 ml of ice-cold water. The washed filters were
removed and added to liquid scintillation vials for measurements on a
PerkinElmer TRI-CARB 2900TR scintillation counter. The uptake of radiola-
beled inositol over 10 min was calculated and plotted as a function of the
Mouse infection studies. Five- to 6-week-old male CD1 mice (18 to 20 g) from
Charles River Laboratories were used in this study. Mice were housed in groups
of five per cage. For infection, colonies from each C. albicans strain were inoc-
ulated into 20 ml of YPD or MM. Cultures were grown overnight and washed
TABLE 2. PCR primers
PrimerUseSequence (5? 3 3?)
Disruption of INO1
Disruption of INO1
Disruption of INO1
Disruption of INO1
Restoration of INO1
Restoration of INO1
Confirmation of ino1?
Confirmation of ino1?
Confirmation of ino1?
Disruption of ITR1
Disruption of ITR1
Disruption of ITR1
Disruption of ITR1
Confirmation of itr1?
Confirmation of itr1?
Confirmation of itr1?
Confirmation of itr1?
Restoration of ITR1
ITR1 conditional construct
ITR1 conditional construct
ITR1 conditional construct
ITR1 conditional construct
ITR1 Northern blot probe
ITR1 Northern blot probe
INO1 Northern blot probe
INO1 Northern blot probe
CaACT1 Northern blot probe
CaACT1 Northern blot probe
TABLE 3. Plasmids used in this study
Plasmid Relevant characteristic(s) Source or reference
CaNAT1-FLP cassette carrying nourseothricin resistance gene
SAT1 flipper carrying nourseothricin resistance gene
pJK863 flanked by 5? and 3? INO1NCRsequences for INO1 gene knockout
INO1 reconstitution construct
pSFS2A flanked by 5? and 3? ITR1NCRsequences for ITR1 gene knockout
ITR1 reconstitution construct
ITR1 conditional construct controlled by CaMET3 promoter
VOL. 76, 2008 MECHANISMS BY WHICH CANDIDA ALBICANS ACQUIRES INOSITOL2795
twice with 25 ml of sterile water, and cells were counted by using a hemacytom-
eter and resuspended in sterile water at 107cells per ml. The cells were then
plated onto YPD to determine the viability. Mice were injected via the tail vein
with 0.1 ml of the cell suspension (106cells) (43), and the course of infection was
monitored for up to 30 days. Survival was monitored twice daily, and moribund
mice were euthanized. All experimental procedures were carried out according
to the NIH guidelines for the ethical treatment of animals.
Statistics. The statistical analysis was done using Prism 4.0 software (Graph-
Pad Software). For the mouse model of systemic infection, Kaplan-Meier sur-
vival curves were compared for significance by using the Mantel-Haenszel log
rank test. The significance of differences in inositol uptake between strains was
determined using the two-tailed unpaired t test. Statistical significance was set at
a P value of ?0.05.
The INO1 gene in C. albicans is required for growth in the
absence of exogenous inositol. The C. albicans homolog of
ScINO1 identified by Klig et al. (18) was disrupted in order to
determine if it was required for inositol biosynthesis in C.
albicans. The C. albicans INO1 homolog (orf19.7585) was dis-
rupted by sequentially replacing both alleles of the gene with
the NAT1-FLP cassette, which contains the nourseothricin re-
sistance gene (39). The INO1 disruption construct is dia-
grammed in Fig. 1A. The INO1 gene was reintegrated into the
INO1 locus of the homozygous mutant (ino1?/ino1?), by using
the construct depicted in Fig. 1B, to verify the linkage of any
resulting phenotypes with the genotype. Wild-type (INO1/
INO1), heterozygous mutant (ino1?/INO1), homozygous mu-
tant (ino1?/?), and reconstituted-INO1 (ino1?/ino1?::INO1)
strains were analyzed by PCR (data not shown) and Southern
blotting (Fig. 1C) to confirm the deletion and reintegration of
the correct genes. The growth patterns of the wild-type, ino1?/
INO1, ino1?/ino1?, and ino1?/ino1?::INO1 strains on medium
lacking inositol (42) were then compared to determine if an
ino1?/ino1? mutation would compromise the ability of the
strain to grow in the absence of inositol. The ino1?/ino1?
strain was unable to grow on inositol-free medium, while the
wild-type, heterozygous, and reconstituted-INO1 strains grew
equally well. As expected, the ino1?/ino1? strain was able to
grow as well as the wild type on identical solid and liquid media
with inositol added (Fig. 1D and data not shown, respectively).
The INO1 gene is not required for virulence in C. albicans.
The INO1 gene does not appear to be required for virulence in
a mouse model of disseminated candidiasis. The wild-type,
ino1?/INO1, and ino1?/ino1?::INO1 strains and two sepa-
rately derived ino1?/ino1? strains were tested for virulence in
the mouse disseminated-infection model (43). These tests re-
vealed no difference in virulence between the wild type and the
ino1 mutant strains (see Fig. 5A). The ino1?/INO1 heterozy-
gote behaved like the wild type, and results for this strain are
The ITR1 gene is required for inositol transport in C. albi-
cans. The ability of the ino1?/ino1? mutant to grow on me-
dium containing inositol supports results from previous studies
that have shown that C. albicans harbors an inositol trans-
porter (15). A C. albicans homolog of the ScITR1 and ScITR2
inositol transporter genes was identified by a BLAST analysis
of the ScItr1p and ScItr2p protein sequences against the Can-
dida Genome Database (CGD; http://www.candidagenome
.org). This homolog is orf19.3526 and is currently referred to as
HGT15 in the CGD, but its alias in CGD is ITR2. This gene has
never been characterized functionally. Based on our BLAST
search of the CGD, C. albicans orf19.3526 is the closest ho-
molog of ScITR1 and ScITR2, and based on the corresponding
amino acid sequences, it is 51% identical to both S. cerevisiae
transporter genes over its full length (data not shown). The
protein product of C. albicans orf19.3526 also contains the
critical inositol transporter motif (D/E)(R/K)?GR(R/K) (38).
Based on the results described below, we will refer to
orf19.3526 as ITR1 hereinafter.
The two alleles of the ITR1 gene were sequentially disrupted
by replacing each ORF with the nourseothricin resistance cas-
sette from the SAT1 flipper (35). The itr1? disruption con-
struct is diagrammed in Fig. 2A. The ITR1 gene was reconsti-
tuted in the itr1?/itr1? strain by using the reconstitution
construct depicted in Fig. 2B. Gene deletions and replace-
ments were checked for accuracy by PCR (data not shown) and
Southern blotting (Fig. 2D). As a further test, Northern blot-
ting revealed that the ITR1 transcript could be detected only in
wild-type (ITR1/ITR1), heterozygous mutant (itr1?/ITR1), and
reconstituted-ITR1 (itr1?/itr1?::ITR1) strains but not in the
FIG. 1. The INO1 gene is required for growth in the absence of
exogenous inositol. (A) Structure of the INO1 disruption construct. Non-
coding DNA sequences of approximately 500 bp each flanking the 5? and
3? ends of the INO1 gene (5? and 3? INO1NCR, respectively) were cloned
onto either flank of the CaNAT1-FLP construct (39). The thick dark
arrows represent the FRT sites of the FLP recombinase. The ball-and-
stick symbol represents the ACT1 terminator (ACT1t), and the thinner
arrow on a raised line represents the SAP2 promoter (PSAP2). (B) INO1-
NAT1 construct used to reintegrate INO1 into the ino1?/ino1? mutant.
(C) Southern blotting was used to confirm the INO1 disruptions. Lanes: 1,
wild type; 2, ino1?::NAT1-FLP/INO1 strain; 3, ino1?/INO1 strain; 4,
ino1?/ino1?::INO1 strain. (D) The cells were streaked onto medium
containing either 0 or 75 ?M inositol and grown for 2 days at 30°C.
2796 CHEN ET AL.INFECT. IMMUN.
itr1?/itr1? strain (Fig. 2E). The expression levels in the itr1?/
ITR1 and itr1?/itr1?::ITR1 strains were lower than that in the
wild type, presumably because these strains contained only one
allele of the gene.
Analyses of inositol uptake by the wild-type, itr1?/ITR1,
itr1?/itr1?, and itr1?/itr1?::ITR1 strains revealed that ITR1 is
required for inositol uptake (Fig. 3). The level of inositol
uptake at 0°C was very low (15) (Fig. 3); hence, this level was
used as a negative control. The level of inositol uptake by the
wild type at 30°C was 58-fold higher than that at 0°C. In
contrast, the level of inositol uptake by the itr1?/itr1? mutant
at 30°C was only fourfold higher than that at 0°C, but a two-
tailed paired t test analysis revealed that there was no signifi-
cant difference in uptake levels between these two tempera-
tures (P ? 0.34). The itr1?/ITR1 mutant showed decreased
inositol import compared to the wild type, and the reconsti-
tuted-ITR1 strain had inositol import restored to a level similar
to that seen in the wild type.
The level of inositol uptake by the wild-type strain in 200 ?M
inositol was found to be 288 pmol per 5 ? 107cells during a
10-min uptake period. Jin and Seyfang showed a level of ap-
proximately 60 pmol per 5 ? 107cells during a single-minute
uptake period (15). The difference between their data and ours
is probably due to the saturation of the uptake system over a
10-min time course (15).
The ITR1 gene is not required for virulence in C. albicans.
The itr1?/itr1? strain did not appear to be any less virulent
than the wild type in a mouse model of systemic candidiasis.
The wild-type, itr1?/ITR1, itr1?/itr1?, and itr1?/itr1?::ITR1
strains were compared in the mouse model of disseminated
infection. There was not a significant difference between the
virulence of the wild type and that of the itr1?/itr1? strains
(P ? 0.95) (see Fig. 5B). The similarity in virulence between
wild-type and itr1?/itr1? strains was born out in two separate
infection experiments. The itr1?/ITR1 mutant behaved like the
wild type (see Fig. 5B) as well. In the group infected with the
reintegrated-ITR1 strain (itr1?/itr1?::ITR1), only half of
the mice succumbed to the infection. This result may be due to
FIG. 2. The ITR1 gene in C. albicans was disrupted. (A) Struc-
ture of the ITR1 disruption construct. Noncoding DNA sequences
of approximately 500 bp each flanking the 5? and 3? ends of the
ITR1 gene (5? and 3? ITR1NCR, respectively) were cloned into the
pSFS2A plasmid such that they flanked the SAT1 flipper cassette.
The thick dark arrows represent the FRT sites of the FLP recom-
binase. The ball-and-stick symbol represents the ACT1 terminator
(ACT1t), and the thinner arrow on a raised line represents the
MAL2 promoter (PMAL2). (B) ITR1-SAT1 construct used to reinte-
grate ITR1 into the itr1?/itr1? mutant. (C) PMET3::ITR1 construct
used to replace the ITR1 allele in the ino1?/ino1? itr1?/ITR1 strain
to generate the ITR1 conditional allele. (D) Southern blotting was
used to confirm the ITR1 disruptions. Lanes: 1, wild type; 2, itr1?::
SAT1-FLP/ITR1 strain; 3, itr1?/ITR1 strain; 4, itr1?/itr1?::SAT1-
FLP strain; 5, itr1?/itr1? strain; 6, itr1?/itr1?::ITR1-SAT1 strain;
and 7, ino1?/ino1? itr1?/PMET3::ITR1 strain. The bands correspond-
ing to itr1?::SAT1-FLP (the allele disrupted prior to the “flipping
out” of the SAT1-FLP construct) and ITR1-SAT1 (reintegrated
ITR1 marked with SAT1) ran together on the gel. (E) Northern
blotting revealed that the expression of ITR1 was lost in the itr1?/
itr1? mutant, but not the wild-type (WT), heterozygous mutant, and
reconstituted-ITR1 (itr1?/itr1?::ITR1) strains. Strains were grown
in defined medium 199 for 2 h at 37°C, and RNA was isolated,
subjected to Northern blotting, and probed for ITR1 expression.
ITR1 expression was normalized to CaACT1 expression on the same
FIG. 3. The ITR1 gene is required for inositol transport in C. albi-
cans. The inositol uptake assay revealed that the ability of the itr1?/
itr1? mutant to import [3H]myo-inositol was greatly reduced compared
to that of the wild type (WT) and that, along with the wild type, the
heterozygous mutant and reconstituted-ITR1 (itr1?/?::ITR1) strains
exhibited the ability to import inositol. Each strain was assayed at 0°C
(white bars) and 30°C (black bars).
VOL. 76, 2008MECHANISMS BY WHICH CANDIDA ALBICANS ACQUIRES INOSITOL2797
a technical error during the injection or an uncharacterized
mutation within the strain. The fact that the wild-type, itr1?/
ITR1, and itr1?/itr1? strains were all similarly virulent strongly
indicates that a lack of ITR1 does not impair virulence.
The INO1 and ITR1 genes show synthetic defects in growth
and virulence. Although neither ITR1 nor INO1 was required for
that each might compensate for the loss of the other during
infection since it has been shown previously that rat serum con-
tains 20 to 100 ?M inositol (14, 34) and mouse liver contains
approximately 100 ?M inositol (2). In order to test this hypoth-
esis, an attempt was made to disrupt both genes simultaneously.
One allele of ITR1 was disrupted in the ino1?/ino1? strain. How-
ever, it was not possible to disrupt the other allele of ITR1,
suggesting that a double mutation consisting of ino1?/ino1? and
itr1?/itr1? was synthetically lethal. Therefore, in the ino1?/ino1?
itr1?/ITR1 strain, the promoter of the remaining wild-type ITR1
allele was replaced with the CaMET3 conditional promoter (6) by
using the construct depicted in Fig. 2C. When the CaMET3 pro-
moter (PMET3) is used to replace the promoter of a target gene, it
represses the transcription of that gene in the presence of methi-
onine and cysteine in the medium. The correct insertion of the
PMET3::ITR1 allele into the chromosome was confirmed by PCR
analysis (data not shown) and Southern blotting (Fig. 2D, lane 7).
The resulting ino1?/ino1? itr1?/PMET3::ITR1 strain was tested for
growth in MM containing 75 ?M inositol and either 0 or 2.5 mM
(each) methionine and cysteine. It was found that unlike the
wild-type and ino1?/ino1? itr1?/ITR1 strains, two separately de-
rived ino1?/? itr1?/PMET3::ITR1 strains failed to grow in the pres-
ence of 2.5 mM (each) methionine and cysteine (Fig. 4).
The ino1?/ino1? itr1?/PMET3::ITR1 strain was tested in a
mouse model of systemic candidiasis to determine whether its
virulence was affected. Previous work had indicated that the
presence of the PMET3promoter on the essential gene CaFBA1
can compromise virulence (37). The methionine in the mouse
bloodstream is presumably able to prevent the expression of
the CaFBA1 gene sufficiently to compromise growth and viru-
lence in the mouse. The wild-type and ino1?/ino1? itr1?/ITR1
strains and two ino1?/ino1? itr1?/PMET3::ITR1 strains were
compared in the mouse model of systemic candidiasis. This
experiment revealed that the ino1?/ino1? itr1?/ITR1 strain
was attenuated in virulence and that the ino1?/ino1?
itr1?/PMET3::ITR1 strains were avirulent (Fig. 5C).
The mechanism by which C. albicans acquires the essential
metabolite inositol during an infection has not been explored
previously. We have found that C. albicans is able to generate
inositol de novo via the INO1 gene product or import inositol
from the environment via the ITR1 gene product with efficien-
cies that allow it to establish an infection regardless of which
mechanism is employed. This conclusion applies only to blood-
stream infections, as a bloodstream infection model was the
only model tested. This result implies that the availability of
inositol in mice is sufficient to support an infection even if C.
albicans must acquire inositol solely by importing it. Although
our search of the literature did not reveal the estimated ino-
sitol content of mouse serum, the inositol levels found in rats
are 20 to 100 ?M (14, 34), which may be comparable to those
in mice and are similar to that found in humans (mean ?
standard deviation, 61.0 ? 12.4 ?M) (19).
The results found for C. albicans are in contrast to those
obtained previously for two important human pathogens, M.
tuberculosis and T. brucei, which require de novo inositol bio-
synthesis via INO1 homologs in order to be fully virulent in
mouse models (M. tuberculosis) or to be viable (T. brucei) (25,
29). Both of these pathogenic microbes are capable of import-
ing inositol, but import is not sufficient to allow for the viability
of T. brucei or the virulence of M. tuberculosis. In the case of T.
brucei, de novo-synthesized inositol is used to make GPI-an-
chored proteins, while imported inositol is used very ineffi-
ciently for this purpose (24). GPI-anchored proteins are re-
quired for the viability of T. brucei (21), so the disruption of
INO1 compromises viability. In the case of M. tuberculosis, the
inositol transporter is too inefficient to import inositol from the
host in order to support infection (29).
Unlike either of these pathogens, C. albicans possesses an
inositol transporter, encoded by ITR1, that is capable of trans-
porting inositol efficiently enough to allow full virulence in a
mouse model of systemic candidiasis even in the absence of de
novo inositol biosynthesis. In the case of both M. tuberculosis
and T. brucei, it has been suggested previously that the devel-
opment of selective inhibitors of Ino1p homologs may be an
effective way to generate antimicrobials (24, 29). The data
reported here indicate that this strategy would not be effective
for C. albicans, as it is fully virulent even in the absence of its
FIG. 4. The INO1 and ITR1 genes show synthetic growth defects on agar plates. The growth patterns of wild-type (WT), ino1?/ino1? itr1?/ITR1, and
two separately derived ino1?/ino1? itr1?/PMET3::ITR1 strains were compared in fivefold serial dilutions on MM containing 75?M inositol with or without
2.5 mM (each) methionine and cysteine. Numbers are cell populations determined by a hemacytometer.
2798 CHEN ET AL.INFECT. IMMUN.
Ino1 enzyme. As an alternative approach, it has been suggested
previously that toxic inositol analogs that are selectively taken
up by the C. albicans inositol transporter but not by the human
inositol-Na2?transporter may be effective drugs (15). This may
be a possibility, although it needs to be determined which of
these two mechanisms, de novo biosynthesis or import, is used
by wild-type C. albicans during an infection. If import is used
extensively, then this approach may work; however, since C.
albicans is fully virulent in the absence of ITR1, the develop-
ment of resistant mutants lacking the Itr1p transporter func-
tion may pose a problem. Nonetheless, a toxic analog may be
useful in combination with drugs affecting other targets, such
as azoles and polyenes (1).
Based on our results, it appears that Itr1p is the primary
and perhaps sole inositol importer in C. albicans both in
vitro and during infection. The disruption of ITR1 greatly
inhibited inositol uptake in vitro (Fig. 3), and in the absence
of INO1, a strain carrying only the PMET3::ITR1 allele of
ITR1 could not grow in medium containing cysteine and
methionine, even in the presence of 75 ?M extracellular
inositol (Fig. 4). The amounts of methionine and cysteine in
mouse serum are apparently sufficient to decrease expres-
sion from the MET3 promoter as well, which is consistent
with results obtained using a MET3-driven form of the
CaFBA1 gene (37). Our conclusion that ITR1 encodes the
primary or sole inositol transporter is consistent with
the findings of a previous study of C. albicans inositol trans-
port which concluded that there was one inositol transporter
in C. albicans that was responsible for all, or at least the vast
majority of, inositol transport activity (15).
The situation of C. albicans contrasts with that of S. cer-
evisiae, which has two inositol transporters that are ex-
pressed at widely different levels (31). In S. cerevisiae, the
ScItr1p transporter is the more highly expressed of the two
transporters and accounts for most of the transport activity.
The residual transport activity in S. cerevisiae is carried out
by ScItr2p, which is expressed at much lower levels. Based
on findings from BLAST searches of the CGD using the
ScITR1 or ScITR2 sequence as the query, the C. albicans
ITR1 gene was the closest homolog to the S. cerevisiae ino-
sitol transporter genes, showing 51% identity over the whole
sequence (based on the corresponding amino acids) to ei-
ther ScITR1 or ScITR2. C. albicans carries at least one other
homolog of the S. cerevisiae inositol transporter genes
(orf19.5447, or HGT19) that is predicted to encode 12 trans-
membrane domains and the (D/E)(R/K)?GR(R/K) motif
typical of inositol transporters. The predicted protein se-
quence corresponding to this gene bears 26 and 27% iden-
tity to ScItr1p and ScItr2p, respectively. It is possible that
orf19.5447 encodes an inositol transporter, but our results
FIG. 5. The INO1 and ITR1 genes show synthetic defects in viru-
lence. The survival of mice following intravenous challenge with 106C.
albicans blastospores was monitored. (A) Mice were injected with the
wild type (WT; n ? 10) and the following INO1 mutant strains: the
ino1?/ino1? strain YLC113 (n ? 10), the ino1?/ino1? strain YLC126
(n ? 11), and an ino1?/ino1?::INO1 strain (n ? 10). (B) Mice were
injected with the wild type (n ? 5) and the following ITR1 mutant
strains: an itr1?/ITR1 strain (n ? 5), an itr1?/itr1? strain (n ? 10), and
an itr1?/itr1?::ITR1 strain (n ? 6). (C) Mice were injected with the
wild type (n ? 10) and the following double mutant strains: the ino1?/
ino1? itr1?/ITR1 strain (n ? 10), the ino1?/ino1? itr1?/PMET3::ITR1
strain YLC261 (n ? 10), and the ino1?/ino1? itr1?/PMET3::ITR1 strain
YLC266 (n ? 10). Strains for the experiments described in the legends
to panels A and B were pregrown in YPD before injection, while
strains for the experiment described in the legend to panel C were
pregrown in MM lacking methionine and cysteine but containing 75
VOL. 76, 2008MECHANISMS BY WHICH CANDIDA ALBICANS ACQUIRES INOSITOL2799
suggest that if so, this other transporter is expressed at too
low a level to transport inositol efficiently or it is expressed
under different conditions from those tested in our experi-
ments. Alternatively, it is a very low affinity transporter like
that seen in M. tuberculosis (29).
Taken together, our results indicate that either Ino1p or
Itr1p can supply the inositol requirement during an infection,
but a number of questions remain to be answered. For wild-
type C. albicans, it is not known whether de novo inositol
biosynthesis or import is utilized during an infection or
whether a combination of both is in operation. The answer may
be variable, depending on the distribution of C. albicans cells
within the host. In addition, it is not known whether de novo
synthesis or import is favored during growth in host niches
other than the bloodstream, such as the gut, the oral mucosa,
and the vaginal tract. This answer again may be variable, de-
pending on the nutrient conditions of the particular host niches
and the locations of cells within those host sites. Further stud-
ies will be required to answer these questions.
We gratefully acknowledge Jeffrey Becker, Michael Lorenz, and
Pamela Small for their critical review of this work. We thank Julia
Ko ¨hler and Joachim Morschha ¨user for providing the CaNAT1-FLP
cassette and the SAT1 flipper, respectively. We are grateful to Melinda
Hauser and Li-Yin Huang for their assistance with the inositol uptake
assay and animal studies, respectively. We also thank all members of
the Reynolds, Kitazono, and Becker laboratories for many helpful
This work was funded in part by grant 1R03AI071863.
1. Anderson, J. B. 2005. Evolution of antifungal-drug resistance: mechanisms
and pathogen fitness. Nat. Rev. 3:547–556.
2. Berry, G. T., S. Wu, R. Buccafusca, J. Ren, L. W. Gonzales, P. L. Ballard,
J. A. Golden, M. J. Stevens, and J. J. Greer. 2003. Loss of murine Na?/myo-
inositol cotransporter leads to brain myo-inositol depletion and central ap-
nea. J. Biol. Chem. 278:18297–18302.
3. Betz, C., D. Zajonc, M. Moll, and E. Schweizer. 2002. ISC1-encoded inositol
phosphosphingolipid phospholipase C is involved in Na?/Li?halotolerance
of Saccharomyces cerevisiae. Eur. J. Biochem. 269:4033–4039.
4. Bustamante, C. I. 2005. Treatment of Candida infection: a view from the
trenches! Curr. Opin. Infect. Dis. 18:490–495.
5. Calderone, R. A. 2002. Candida and candidiasis. ASM Press, Washington,
6. Care, R. S., J. Trevethick, K. M. Binley, and P. E. Sudbery. 1999. The MET3
promoter: a new tool for Candida albicans molecular genetics. Mol. Micro-
7. Chen, M., L. C. Hancock, and J. M. Lopes. 2007. Transcriptional regulation
of yeast phospholipid biosynthetic genes. Biochim. Biophys. Acta 1771:310–
8. De Backer, M. D., D. Maes, S. Vandoninck, M. Logghe, R. Contreras, and
W. H. Luyten. 1999. Transformation of Candida albicans by electroporation.
Yeast (Chichester, England) 15:1609–1618.
9. Donahue, T. F., and S. A. Henry. 1981. myo-inositol-1-phosphate synthase.
Characteristics of the enzyme and identification of its structural gene in
yeast. J. Biol. Chem. 256:7077–7085.
10. Donovan, M., J. J. Schumuke, W. A. Fonzi, S. L. Bonar, K. Gheesling-Mullis,
G. S. Jacob, V. J. Davisson, and S. B. Dotson. 2001. Virulence of a phos-
phoribosylaminoimidazole carboxylase-deficient Candida albicans strain in
an immunosuppressed murine model of systemic candidiasis. Infect. Immun.
11. Gillum, A. M., E. Y. Tsay, and D. R. Kirsch. 1984. Isolation of the Candida
albicans gene for orotidine-5?-phosphate decarboxylase by complementation
of S. cerevisiae ura3 and E. coli pyrF mutations. Mol. Gen. Genet. 198:179–
12. Haites, R. E., Y. S. Morita, M. J. McConville, and H. Billman-Jacobe. 2005.
Function of phosphatidylinositol in mycobacteria. J. Biol. Chem. 280:10981–
13. Hoffman, C. S., and F. Winston. 1987. A ten-minute DNA preparation from
yeast efficiently releases autonomous plasmids for transformation of Esche-
richia coli. Gene 57:267–272.
14. Isaacks, R. E., A. S. Bender, C. Y. Kim, and M. D. Norenberg. 1997. Effect
of osmolality and myo-inositol deprivation on the transport properties of
myo-inositol in primary astrocyte cultures. Neurochem. Res. 22:1461–
15. Jin, J. H., and A. Seyfang. 2003. High-affinity myo-inositol transport in
Candida albicans: substrate specificity and pharmacology. Microbiology
(Reading, England) 149:3371–3381.
16. Kirsch, D. R., and R. R. Whitney. 1991. Pathogenicity of Candida albicans
auxotrophic mutants in experimental infections. Infect. Immun. 59:3297–
17. Klig, L. S., B. Antonsson, E. Schmid, and L. Friedli. 1991. Inositol biosyn-
thesis: Candida albicans and Saccharomyces cerevisiae genes share common
regulation. Yeast (Chichester, England) 7:325–336.
18. Klig, L. S., P. A. Zobel, C. G. Devry, and C. Losberger. 1994. Comparison of
INO1 gene sequences and products in Candida albicans and Saccharomyces
cerevisiae. Yeast (Chichester, England) 10:789–800.
19. Kouzuma, T., M. Takahashi, T. Endoh, R. Kaneko, N. Ura, K. Shimamoto,
and N. Watanabe. 2001. An enzymatic cycling method for the measurement
of myo-inositol in biological samples. Clin. Chim. Acta 312:143–151.
20. Lay, J., L. K. Henry, J. Clifford, Y. Koltin, C. E. Bulawa, and J. M. Becker.
1998. Altered expression of selectable marker URA3 in gene-disrupted Can-
dida albicans strains complicates interpretation of virulence studies. Infect.
21. Lillico, S., M. C. Field, P. Blundell, G. H. Coombs, and J. C. Mottram. 2003.
Essential roles for GPI-anchored proteins in African trypanosomes revealed
using mutants deficient in GPI8. Mol. Biol. Cell 14:1182–1194.
22. Lopez, F., M. Leube, R. Gil-Mascarell, J. P. Navarro-Avino, and R.
Serrano. 1999. The yeast inositol monophosphatase is a lithium- and
sodium-sensitive enzyme encoded by a non-essential gene pair. Mol.
23. Manning, M., C. B. Snoddy, and R. A. Fromtling. 1984. Comparative patho-
genicity of auxotrophic mutants of Candida albicans. Can. J. Microbiol.
24. Martin, K. L., and T. K. Smith. 2006. The glycosylphosphatidylinositol (GPI)
biosynthetic pathway of bloodstream-form Trypanosoma brucei is depen-
dent on the de novo synthesis of inositol. Mol. Microbiol. 61:89–105.
25. Martin, K. L., and T. K. Smith. 2005. The myo-inositol-1-phosphate synthase
gene is essential in Trypanosoma brucei. Biochem. Soc. Trans. 33:983–985.
26. Mille, C., G. Janbon, F. Delplace, S. Ibata-Ombetta, C. Gaillardin, G.
Strecker, T. Jouault, P. A. Trinel, and D. Poulain. 2004. Inactivation of
CaMIT1 inhibits Candida albicans phospholipomannan beta-mannosylation,
reduces virulence, and alters cell wall protein beta-mannosylation. J. Biol.
27. Morrell, M., V. J. Fraser, and M. H. Kollef. 2005. Delaying the empiric
treatment of Candida bloodstream infection until positive blood culture
results are obtained: a potential risk factor for hospital mortality. Antimi-
crob. Agents Chemother. 49:3640–3645.
28. Morschhauser, J., S. Michel, and P. Staib. 1999. Sequential gene disruption
in Candida albicans by FLP-mediated site-specific recombination. Mol. Mi-
29. Movahedzadeh, F., D. A. Smith, R. A. Norman, P. Dinadayala, J. Murray-
Rust, D. G. Russell, S. L. Kendall, S. C. Rison, M. S. McAlister, G. J.
Bancroft, N. Q. McDonald, M. Daffe, Y. Av-Gay, and N. G. Stoker. 2004. The
Mycobacterium tuberculosis ino1 gene is essential for growth and virulence.
Mol. Microbiol. 51:1003–1014.
30. Nikawa, J., T. Nagumo, and S. Yamashita. 1982. myo-inositol transport in
Saccharomyces cerevisiae. J. Bacteriol. 150:441–446.
31. Nikawa, J., Y. Tsukagoshi, and S. Yamashita. 1991. Isolation and charac-
terization of two distinct myo-inositol transporter genes of Saccharomyces
cerevisiae. J. Biol. Chem. 266:11184–11191.
32. Noble, S. M., and A. D. Johnson. 2005. Strains and strategies for large-scale
gene deletion studies of the diploid human fungal pathogen Candida albi-
cans. Eukaryot. Cell 4:298–309.
33. Orlean, P., and A. K. Menon. 2007. Thematic review series: lipid posttrans-
lational modifications. GPI anchoring of protein in yeast and mammalian
cells, or: how we learned to stop worrying and love glycophospholipids. J.
Lipid Res. 48:993–1011.
34. Palmano, K. P., P. H. Whiting, and J. N. Hawthorne. 1977. Free and lipid
myo-inositol in tissues from rats with acute and less severe streptozotocin-
induced diabetes. Biochem. J. 167:229–235.
35. Reuss, O., A. Vik, R. Kolter, and J. Morschhauser. 2004. The SAT1 flipper,
an optimized tool for gene disruption in Candida albicans. Gene 341:119–
36. Reynolds, T. B. 2006. The Opi1p transcription factor affects expression of
FLO11, mat formation, and invasive growth in Saccharomyces cerevisiae.
Eukaryot. Cell 5:1266–1275.
37. Rodaki, A., T. Young, and A. J. Brown. 2006. Effects of depleting the essen-
tial central metabolic enzyme fructose-1,6-bisphosphate aldolase on the
growth and viability of Candida albicans: implications for antifungal drug
target discovery. Eukaryot. Cell 5:1371–1377.
38. Seyfang, A., and S. M. Landfear. 2000. Four conserved cytoplasmic sequence
2800CHEN ET AL.INFECT. IMMUN.
motifs are important for transport function of the Leishmania inositol/H? Download full-text
symporter. J. Biol. Chem. 275:5687–5693.
39. Shen, J., W. Guo, and J. R. Kohler. 2005. CaNAT1, a heterologous dominant
selectable marker for transformation of Candida albicans and other patho-
genic Candida species. Infect. Immun. 73:1239–1242.
40. Shepherd, M. G. 1985. Pathogenicity of morphological and auxotrophic
mutants of Candida albicans in experimental infections. Infect. Immun. 50:
41. Strahl, T., and J. Thorner. 2007. Synthesis and function of membrane phos-
phoinositides in budding yeast, Saccharomyces cerevisiae. Biochim. Biophys.
42. Styles, C. 2002. How to set up a yeast laboratory. Methods Enzymol. 350:
43. Warenda, A. J., S. Kauffman, T. P. Sherrill, J. M. Becker, and J. B. Konopka.
2003. Candida albicans septin mutants are defective for invasive growth and
virulence. Infect. Immun. 71:4045–4051.
Editor: A. Casadevall
VOL. 76, 2008 MECHANISMS BY WHICH CANDIDA ALBICANS ACQUIRES INOSITOL2801