Iron acquisition is essential for the growth of most microorganisms and occurs primarily by two mechanisms: siderophore production and reductive iron assimilation 
. Both of these processes are induced by H. capsulatum
during iron limitation 
, but their role in cell growth and virulence has not been previously investigated. In this study, we determined the role of siderophore production in growth under conditions of iron limitation as well as during infection. We identified a siderophore biosynthetic gene cluster that was transcriptionally induced in response to iron limitation. Disruption of siderophore production resulted in iron-dependent growth in culture and during macrophage infection, and caused a growth defect in mice.
Presumably the sid1Δ
mutant shows a strong pulmonary colonization defect in competition with wild-type because it has a reduced capacity for iron acquisition in vivo
. The in vivo
growth defect caused by lack of siderophore production is most pronounced at 15 days post infection. Interestingly, the peak of IFN-γ production by T cells in response to H. capsulatum
infection occurs at day 14 
. Since one of the functions of IFN-γ is to restrict iron, perhaps siderophore production is primarily required for iron acquisition during the latter stages of infection, whereas redundant iron acquisition mechanisms, such as reductive iron assimilation, might play a role early in infection.
In other fungal pathogens, siderophore production and reductive iron assimilation, which is usually mediated by ferric reductase activity, play differential roles in pathogenesis. For example, siderophore production is essential for virulence in Aspergillus fumigatus
, but reductive iron assimilation is neither necessary nor sufficient for normal growth and survival in the host 
. Similarly, in the phytopathogens Cochliobolus miyabeanus
, Alternaria brassicicola, Cochliobolus heterostrophus
, and Fusarium graminearum
, extracellular siderophore production is required for full virulence 
. In contrast, deletion of the SID1
gene in U. maydis
does not affect virulence in maize 
; instead, reductive iron assimilation is required for virulence 
. This also true in Candida albicans
, where ferric reductase activity is required for systemic infection 
Ferric reductase activity has been clearly demonstrated in H. capsulatum 
, though the relevant genes that encode these activities have not been identified, and thus their role in virulence has not been assessed. Perusal of the genome revealed seven putative ferric reductase genes, though none of these candidates were observed to be upregulated by iron limitation in our experiments. Nonetheless, any or all of these genes could contribute to pathogenesis in the host.
In addition to identifying a role for siderophore production in virulence, this study revealed several interesting regulatory properties of siderophore-biosynthesis genes. First, inspection of the sequences upstream of each gene revealed a consensus sequence that was present at least once per gene. The expression of several fungal orthologs of SID1
is repressed by GATA-type negative regulators that recognize the HGATR motif 
. Interestingly, the consensus sequence we identified, 5′-(G/A)ATC(T/A)GATAA-3′, contained an HGATAR motif, and was shown to bind an H. capsulatum
GATA factor in vitro
(Chao et al., submitted). We are now investigating whether these regulatory sequences are necessary to confer gene regulation in response to iron levels in vivo
Second, the siderophore biosynthesis genes were located adjacent to each other in the genome, and thus comprise the first secondary metabolite gene cluster defined in H. capsulatum
. We observed similar clustering in the genome of the closely related systemic dimorphic fungal pathogen C. immitis
, but it is not evident to nearly the same extent for siderophore biosynthesis genes in other sequenced fungal genomes 
Third, we could not express significant levels of SID1 from an episomal plasmid, suggesting that integration of the gene may be critical for normal expression levels. Additionally, integration of SID1 at random sites in complementation strains resulted in partial complementation of siderophore production, suggesting that optimal function might be achieved only with expression from the original genomic locus. Unfortunately, targeted integration is extremely challenging in H. capsulatum, making it difficult to directly test this hypothesis.
The potential significance of the H. capsulatum
gene cluster with regards to gene regulation is intriguing. One possibility is that clustering in the genome facilitates local changes in chromatin structure that allow a regional change in promoter accessibility. This type of regulation is reminiscent of transcriptional control of secondary metabolite clusters in Aspergillus
species, where transcriptional accessibility of genes in the cluster is controlled by the putative methyltransferase LaeA 
. Perhaps local control of chromatin structure allows a rapid and coordinated transcriptional switch in response to changes in iron levels.