1806? The?Journal?of?Clinical?Investigation? ? ? http://www.jci.org? ? ? Volume 115? ? ? Number 7? ? ? July 2005
HIF-1α expression regulates the bactericidal
capacity of phagocytes
Carole Peyssonnaux,1 Vivekanand Datta,2 Thorsten Cramer,1 Andrew Doedens,1
Emmanuel A. Theodorakis,3 Richard L. Gallo,2,4 Nancy Hurtado-Ziola,4 Victor Nizet,2
and Randall S. Johnson1
1Division of Biological Sciences, 2Department of Pediatrics, 3Department of Chemistry and Biochemistry, and 4Department of Medicine,
University of California, San Diego, La Jolla, California, USA.
The eradication of invading microorganisms depends initially on
innate immune mechanisms that preexist in all individuals and
act within minutes of infection. Phagocytic cell types, including
macrophages and neutrophils, play a key role in innate immunity
because they can recognize, ingest, and destroy many pathogens
without the aid of an adaptive immune response. The effective-
ness of myeloid cells in innate defense reflects their capacity to
function in low-oxygen environments. Whereas in healthy tissues
oxygen tension is generally 20–70 mm Hg?(i.e., 2.5–9% oxygen),
much lower levels (< 1% oxygen) have been described in wounds
and necrotic tissue foci (1–3).
The adaptive response of mammalian cells to the stress of oxy-
gen depletion is coordinated by the action of hypoxia-inducible
transcription factor 1 (HIF-1). HIF-1 is a heterodimer whose
expression is regulated by oxygen at the protein level. The pro-
tein stability of the α subunit (HIF-1α) is regulated by a family of
prolyl hydroxylases. This process is directed by the?interaction of
HIF-1α with the von Hippel–Lindau tumor-suppressor protein
(vHL). Under hypoxia, prolyl hydroxylase activity is inhibited, and
HIF-1α accumulates and translocates into the nucleus, where it
binds to HIF-1β, constitutively expressed. The heterodimer HIF-1
binds to the hypoxic response elements (HREs) of target gene reg-
ulatory sequences, resulting in the transcription of genes impli-
cated in the control of metabolism and angiogenesis as well as
apoptosis and cellular stress (4). Some of these direct target genes
include glucose transporters, glycolytic enzymes, erythropoietin,
and the angiogenic factor VEGF. Two additional HIF subunits
have subsequently been cloned and named HIF-2 (5–7) and HIF-3
(8), but their regulation is less well understood.
Confirmation that HIF-1α was expressed in activated
macrophages (9, 10) led us to explore the function of this tran-
scription factor in the myeloid cell lineage. Employing condition-
al gene targeting, we recently showed that HIF-1α control of the
metabolic shift to glycolysis plays an important role in myeloid
cell–mediated inflammatory responses (11). These studies also
provided preliminary in vitro evidence that deletion of HIF-1α
could impair myeloid cell bactericidal activity. The effectiveness
of neutrophils and macrophages in innate antibacterial defense
reflects a diverse array of highly specialized cellular functions
including phagocytic uptake of the bacterium, production of reac-
tive oxygen species, activation of iNOS, and release of antimicro-
bial peptides (e.g., cathelicidins, defensins) and granule proteases
(e.g., elastase, cathepsin). Here we perform a detailed analysis of
the underlying mechanisms by which HIF-1α transcriptional
control pathways contribute to the antibacterial function of
myeloid cells, and for the first time, to our knowledge, determine
the requirement of HIF-1α expression for myeloid cell–mediated
innate immune defense in vivo. Our results indicate a pivotal role
for HIF-1α in myeloid cell biology under both hypoxia and nor-
moxia and suggest that this transcription factor may represent a
unique therapeutic target for boosting immune defense function
in tissues compromised by bacterial infection.
Bacteria induce HIF-1α expression. Invasive pyogenic bacterial skin
and soft tissue infections generate localized tissue ischemia,
thrombosis, and necrosis and represent a formidable test of the
adaptiveness of neutrophils and macrophages in hypoxic microen-
vironments. In this regard, a strain of the Gram-positive pathogen
group A Streptococcus (GAS), isolated from a patient with necro-
tizing fasciitis (flesh-eating disease), was chosen as the primary
Nonstandard?abbreviations?used: AG, 1-amino-2-hydroxyguanidine, p-toluene-
sulfate; CoCl2, cobalt chloride; CRAMP, cathelicidin-related antimicrobial peptide;
fMLP, N-formyl-methionyl-leucyl-phenylalanine; GAS, group A Streptococcus; HIF-1α,
hypoxia-inducible factor 1, α subunit; HRE, hypoxic response element; L-Mim,
L-mimosine; MRSA, methicillin-resistant Staphylococcus aureus; NE, neutrophil
elastase; OH-pyridone, 3-hydroxy-1,2-dimethyl-4(1H)-pyridone; THB, Todd-Hewitt
broth; vHL, von Hippel–Lindau tumor-suppressor protein.
Conflict?of?interest: The authors have declared that no conflict of interest exists.
Citation?for?this?article: J. Clin. Invest. 115:1806–1815 (2005).
Related Commentary, page 1702
?The?Journal?of?Clinical?Investigation http://www.jci.org Volume 115 Number 7 July 2005
model organism for most in vitro and in vivo challenges. We found
that expression of HIF-1α was increased 4-fold in WT mouse
macrophages following exposure to GAS under normoxic condi-
tions (Figure 1A). Indeed, GAS represented a more potent stimu-
lus for HIF-1α induction than hypoxia itself. The phenomenon of
bacterial induction of HIF-1α under normoxia was also observed
with additional Gram-positive (methicillin-resistant Staphylococcus
aureus, hereafter S. aureus) and Gram-negative (Pseudomonas aerugi-
nosa, hereafter P. aeruginosa, and Salmonella typhimurium) bacterial
species of medical importance (Figure 1A).
We next evaluated whether the induction of HIF-1α protein by
GAS corresponded to an increase in HIF-1α transcriptional gene
activation. We measured HIF-1α–dependant transcription in
macrophages derived from HRE-luciferase transgenic mice, which
contain a luciferase reporter gene driven by 6 consecutive specific
HRE sequences. As shown in Figure 1B, a 3-fold increase in lucifer-
ase reporter activity was reached after incubating the macrophages
for 18 hours in 1% oxygen or in the presence of known pharmaco-
logical inducers of HIF-1α, including desferrioxamine mesylate,
cobalt chloride (CoCl2), and L-mimosine (L-Mim). Incubation of
the reported macrophages with live or heat-killed GAS bacteria at
normoxia stimulated luciferase activity to levels comparable to or
greater than those of hypoxia (Figure 1B).
HIF-1α regulates bactericidal capacity of myeloid cells. To assess the
functional consequences of HIF-1α activation, we used an anti-
biotic protection assay to calculate intracellular killing of GAS by
WT macrophages compared with killing by those derived from
the bone marrow of HIF1α-lysMcre mice (11). Here, targeted
deletion of the HIF-1α gene has been created via crosses into a
background of cre expression driven by the lysozyme M promoter
(lysMcre), allowing specific deletion of the transcription factor
in the myeloid lineage (11). As shown in Figure 2A, intracellular
killing of GAS by WT macrophages was increased under hypoxia,
providing initial indication that HIF-1α may be involved in the
bactericidal process. This result was especially notable because the
facultative GAS bacteria lack oxidative phosphorylation and grow
more rapidly under anaerobiasis (12). We found that, compared
to WT cells, macrophages from HIF-1α–null mice showed a 2-fold
decrease in GAS intracellular killing under normoxia and a 3-fold
decrease in GAS intracellular killing under hypoxia (Figure 2A).
Time-course studies showed that the killing defect observed in
HIF-1α–null macrophages increased over time, such that 15-fold
more viable bacteria were present within HIF-1α deleted cells by
the last time point of 120 minutes (Figure 2B). Macrophage killing
of the Gram-negative bacterium P. aeruginosa was likewise impaired
upon deletion of HIF-1α (Figure 2B).
As a complementary analysis of the linkage of myeloid cell
bactericidal functions with HIF-1α transcriptional control, we
explored the effects of increased HIF-1α activity on bacterial kill-
ing by using macrophages derived from vHL-null mice. vHL is a
key regulator of HIF-1α turnover; these mice have constitutively
high levels of HIF-1 activity in the deleted cell population (11). We
found that vHL-null macrophages showed increased intracellular
killing of GAS and P. aeruginosa compared with WT cells across
multiple time points (Figure 2C). Similar differences were
observed in macrophage bactericidal assays that omitted antibi-
otics and instead employed vigorous washing to quantify total
surviving cell-associated GAS or P. aeruginosa (not shown). Macro-
phage populations isolated from WT, HIF-1α–null, and vHL-null
mice both included more than 99.5% differentiated macrophages
by flow cytometric analysis, and Trypan blue straining showed
similar levels of macrophage viability (98–99%) throughout the
GAS-killing assays (not shown). These controls suggest that there
exists an intrinsic defect in the bactericidal activity of HIF-1α–
null cells that cannot be attributed to differences in the purity or
viability of the explanted cell populations.
Finally, we treated WT macrophages with a number of known
pharmacologic inducers of HIF-1α that each act directly or indi-
rectly to inhibit prolyl hydroxylase targeting of HIF-1α for ubiq-
uitination. These included the iron chelator desferrioxamine,
CoCl2, L-Mim, and 3-hydroxy-1,2-dimethyl-4(1H)-pyridone (OH-
pyridone) (13). The addition of each of these agents increased
intracellular killing of GAS by WT macrophages (Figure 2D).
Assays were performed using a concentration of each agonist and
exposure time that did not affect bacterial viability (not shown).
Myeloid cell HIF-1α production is important for control of GAS
infection in vivo. We chose an animal infection model of GAS-
induced necrotizing soft tissue infection for directly testing
myeloid cell bactericidal function in vivo. We introduced the
GAS inoculum subcutaneously into a shaved area on the flank
of WT and HIF-1α–/– male littermates and followed progression
of the infection over 96 hours. We found that mice with a tis-
sue-specific deletion of HIF-1α in macrophage and neutrophils
developed significantly larger necrotic skin lesions and experi-
enced greater weight loss than WT mice (Figure 3, A and B). Rep-
resentative gross appearance of the necrotic lesions in WT and
HIF-1α myeloid–null mice is shown in Figure 3C. We next asked
Bacteria increase HIF-1α protein expression and stimulate HIF-1α
transcriptional activity. (A and B) Macrophages were incubated under
hypoxia (0.1%) or with GAS, MRSA, S. typhimurium (ST), or P. aeru-
ginosa (PA) at an MOI equal to 5–10 under normoxic conditions for
4 hours. Expression of HIF-1α was normalized to β-actin levels and
quantified with ImageQuantTL software (Amersham Biosciences). (C)
HRE-luciferase BM-derived macrophages were incubated either with
GAS or heat-inactivated GAS at an MOI equal to 5–10 under hypoxia
(1%) or with the addition of mimosine (800 µM), desferrioxamine mesyl-
ate (150 µM), or CoCl2 (150 µM) for 18 hours. Statistical analyses were
performed using unpaired Student’s t test. **P < 0.01; ***P < 0.001.
1808?The?Journal?of?Clinical?Investigation http://www.jci.org Volume 115 Number 7 July 2005
whether myeloid cell production of HIF-1α was important in
limiting the ability of GAS to replicate within the necrotic skin
tissues and to disseminate from the initial focus of infection
into the bloodstream and systemic organs. Mice were sacri-
ficed at 96 hours after inoculation and quantitative bacterial
cultures performed on the skin ulcer (or site of inoculation if
no ulcer developed), blood, and spleen (Figure 3D). Approxi-
mately 1,660-fold greater quantities of GAS were present in the
skin biopsies of HIF-1α–null mice compared with those of WT
mice. Similarly, 27-fold (blood) or 85-fold (spleen) more bacteria
were isolated in systemic cultures from HIF-1α–null mice com-
pared with WT mice. Our findings indicate that the presence of
HIF-1α transcriptional control in neutrophils and macrophages
is important in limiting the extent of necrotic tissue damage and
preventing systemic spread of bacterial infection.
HIF-1α is not critical for neutrophil endothelial transcytosis or oxida-
tive burst function. We next began a series of experiments to probe
the potential cellular and molecular mechanisms through which
HIF-1α may support myeloid cell functional killing capacity in
vitro and in vivo. Although histopathologic examination of the
biopsies from the necrotic ulcers generated by GAS revealed clear
tissue ischemia by HypoxyProbe (Figure 4A), the observed immune
defect of HIF-1α–null animals did not appear to reflect impaired
phagocyte recruitment, since similar numbers of neutrophils were
observed on immunostaining of the skin tissue of WT compared
with that of HIF-1α–null mice at 6, 12, and 24 hours after infec-
tion (Figure 4B). The latter finding differed qualitatively from
our previous study, in which decreased neutrophil infiltration
was seen in skin tissue of HIF-1α after chemical irritation with
the phorbol ester tetradecanoyl phorbol acetate (11), and from the
prediction that might be derived from HIF-1α control of β2 inte-
grin expression (14). We speculate that the stimulus to neutrophil
migration elicited by bacterial infection is perhaps stronger and
more complex (i.e., involving more pathways) than that of chemi-
cal irritation such that the any contribution of HIF-1α may be
muted in comparison to its effects on bacterial killing. To explore
further whether the migratory capacity of WT and HIF-1α–null
neutrophils toward a bacterial stimulus was indeed unaffected,
we measured the rate of transcytosis across murine endothelial
cell monolayers following stimulation by GAS or the bacteria-
derived chemotactic peptide N-formyl-methionyl-leucyl-phenyl-
alanine (fMLP). In these assays, we also found no significant dif-
ference in transendothelial migration between the activated WT,
HIF-1α–null, and vHL-null murine neutrophils (Figure 4C).
The production of reactive oxygen metabolites generated by
lysosomal NADPH oxidases in a process known as the respiratory
HIF-1α regulates bactericidal activity of myeloid cells. (A) Intracellular killing of GAS by WT, HIF-1α–null, or vHL-null macrophages. BM-derived
macrophages were inoculated with GAS at an MOI equal to 2.5 and cultured under normoxic (white bars) or hypoxic (0.1%; black bars) condi-
tions for 1 hour after antibiotic treatment. Statistical analyses were performed using unpaired Student’s t test. *P < 0.05; **P < 0.01. (B) Loss of
HIF-1α in macrophages decreases intracellular killing of GAS and of P. aeruginosa. WT (black bars) or HIF-1α–null (white bars) BM-derived
macrophages were incubated with bacteria for 1 hour before antibiotics were added. Intracellular killing was analyzed by determination of viable
CFUs in macrophage lysates at the specified time points after bacterial uptake. Experiments were performed in triplicate. SEM is displayed.
Experiment shown is representative of 3 repeated studies. (C) Loss of vHL in BM-derived macrophages increases intracellular killing of GAS and
of P. aeruginosa. Experiments were performed in triplicate and are representative of 3 repeated studies. SEM is displayed. (D) Pharmacologic
agonists of HIF-1α increase myeloid cell bactericidal activity. Preincubation (5 hours) with desferrioxamine mesylate (DFO), CoCl2, OH-pyridone,
or Mim increased the intracellular killing capacity of WT macrophages against GAS. ***P < 0.001.
? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 115 Number 7 July 2005
burst is a major mechanism of bacterial killing. However, circulat-
ing neutrophils derived from HIF-1α–deficient or vHL-deficient
mice were similar to WT neutrophils in oxidative burst activity
(Figure 4D). Thus, the defect in innate immunity to GAS infection
observed in HIF-1α myeloid–null mice could not be attributed to
impairment in oxidative burst function.
Production of granule proteases and antimicrobial peptides is regulated
by HIF-1α. Granule proteases are increasingly recognized as an
important component of myeloid cell antimicrobial activity.
Neutrophil elastase (NE) and cathepsin G are abundant serine
proteases concentrated in the granules that are primarily des-
tined to fuse to phagosomes and form phagolysosomes. Gene
targeting of elastase in mice has directly supported a role of NE
in host innate immune defense (15), and accumulating evidence
suggests a similar role for cathepsin G (16–18). Patients with
Chediak-Higashi syndrome lack NE and suffer recurrent bac-
terial infections. To determine whether HIF-1α has an impact
on neutrophil production of granule proteases, we measured
NE and cathepsin G activity in WT, HIF-1α–null, and vHL-null
blood neutrophils. Protease activity was measured using a syn-
thetic peptide substrate containing recognition sites for each
molecule to allow either fluorometric (NE) (Figure 5A) or spec-
trophotometric (cathepsin G) detection (19) (Figure 5B). HIF-1α–
null neutrophils showed decreased enzymatic activity of each
granule protease compared with WT neutrophils while vHL-
null neutrophils exhibited increased protease activity. Mixing
experiments with WT and HIF-1α–null macrophages excluded
the possibility that HIF-1α–null neutrophils produce a greater
amount of an (unknown) inhibitor rather than less of the gran-
ule proteases (Figure 5, A and B).
The production of proteases by neutrophils may exert direct
antimicrobial effects or, alternatively, may serve to activate cat-
ionic antimicrobial peptides from their inactive precursor forms
(20, 21). An important component of innate immune defense in
mammals is the cathelicidin family of antimicrobial peptides?(22).
These gene-encoded “natural antibiotics” exhibit broad-spectrum
antimicrobial activity and are produced by several mammalian
species on epithelial surfaces and within the granules of phago-
cytic cells. Proteolytic cleavage of an inactive precursor form to
release the mature C terminal antimicrobial peptide is accom-
plished by proteases, such as elastase, upon degranulation of
activated neutrophils (23). Mice have a single cathelicidin, cathe-
licidin-related antimicrobial peptide (CRAMP), which closely
resembles the single human cathelicidin (LL-37). Importantly, we
demonstrated in earlier experiments using the murine model of
necrotizing skin infection that endogenous production of CRAMP
was essential for mammalian innate immunity to GAS (24). We
performed experiments to identify whether production or acti-
vation of CRAMP was under HIF-1α control. Lysates from WT,
HIF-1α–null, and vHL-null peritoneal neutrophils were analyzed
by SDS-PAGE and immunoblotted with a rabbit anti-mouse
CRAMP antibody against the CRAMP mature peptide. HIF-1α
deletion led to a dramatic reduction of the active mature form of
cathelicidin compared with WT neutrophils while CRAMP was
expressed at higher levels in vHL-deficient neutrophils (Figure 5C).
Regulation of cathelicidin expression occurred at least in part at
the mRNA level, as CRAMP transcript levels are reduced by 80% in
HIF-1α–null macrophages, and conversely increased with loss of
vHL (Figure 5D). As would be expected, CRAMP mRNA was also
increased by exposure of the neutrophils to hypoxia (Figure 5D).
HIF-1α deletion renders mice more susceptible to GAS infection. (A) Area of necrotic ulcer and (B) loss of weight in individual WT (squares) and
HIF-1α myeloid–null mice (triangles) 4 days after infection with GAS. (C) Representative appearance of GAS-induced necrotic skin ulcers in WT
and HIF-1α myeloid–null mice. A total of 11 mice in each group were tested in 3 paired experiments. (D) Bacterial counts in the blood, spleen,
and skin of WT and HIF-1α myeloid–null mice infected with GAS. The fold difference in quantitive GAS culture between WT and HIF-1α–null
animals is annotated. Statistical analyses were performed using unpaired Student’s t test. *P < 0.05; **P < 0.01.
1810? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 115 Number 7 July 2005
Thus, the production and activation of cathelicidin antimicrobial
peptides represents an additional myeloid cell killing mechanism
that is affected by alterations in the HIF-1α pathway.
HIF-1α is a principal regulator of NO production in response to bacterial
infection. NO is known to exert antimicrobial properties against a
variety of bacterial species (25). Nitric oxide is enzymatically pro-
duced by NOS through the oxidation of arginine, and mice defi-
cient in iNOS are more susceptible to bacterial infection (26, 27). It
has been well documented that HIF-1α is a transcriptional activa-
tor of iNOS expression (28–30), but no studies have examined this
linkage in the context of bacterial infection. Here we found that
exposure of macrophages to GAS increased iNOS mRNA produc-
tion approximately 250-fold (Figure 6A). Deletion of HIF-1α result-
ed in an approximately 70% reduction in iNOS gene transcription,
while deletion of vHL led to a marked increase in iNOS mRNA
levels (Figure 6A). Measurement of nitrite in cell culture superna-
tants confirmed that the observed differences in iNOS induction
translated directly to differences
in NO production (Figure 6B).
Addition of the NOS inhibitor
p-toluenesulfate (AG) signifi-
cantly inhibited the production
of NO in response to GAS (Fig-
ure 6B). WT macrophages treat-
ed with L-Mim, a pharmologi-
cal inducer of HIF-1α showed
greatly increased expression of
iNOS mRNA, but this increased
expression was very low in
HIF-1α–null cells (Figure 6C),
confirming a dependency of
the observed effect on the pres-
ence of the transcription factor.
These experiments indicate that
augmentation of iNOS expres-
sion and subsequent bacterial
killing (Figure 2C) can be phar-
macologically induced through
increased HIF-1α expression.
To establish the functional
importance of HIF-1α–induced
iNOS expression and NO pro-
duction, we performed mac-
rophage bactericidal assays in
the presence or absence of AG.
Figure 6D shows that AG inhib-
ited the bactericidal activity of
WT macrophages but did not
further suppress the poor bac-
tericidal activity of HIF-1α–null
macrophages. We found similar
results using the iNOS-specific
inhibitor 1400W (not shown). It
has recently been demonstrated
that NO, as well as certain reac-
tive oxygen species, cytokines,
and growth factors, can partici-
pate in stability regulation of
HIF-1α and HIF-1 transactiva-
tion during normoxia (31–36). As seen in Figure 6E, we found that
inhibition of iNOS by AG blocked the ability of GAS exposure to
generate increased levels of HIF-1α in WT macrophages. Thus,
HIF-1α induces the production of NO, which not only acts as a
key element in bacterial killing, but also serves as a regulatory mol-
ecule that further stabilizes HIF-1α. This places HIF-1α at the cen-
ter of an amplification loop during the innate immune response of
myeloid cells to bacterial infection.
HIF1-α regulates myeloid cell TNF-α production through a NO-depen-
dent process. We next examined the expression pattern of TNF-α, a
cytokine involved in the augmentation of inflammatory respons-
es to bacterial infection. Indeed, development of GAS-necrotiz-
ing fasciitis has been reported as a complication of anti–TNF-α
therapy (37). As shown in Figure 7A, GAS strongly induced TNF-α
mRNA production in WT macrophages. This transcriptional
response was severely diminished in HIF-1α–null macrophages
and upregulated in vHL-null macrophages. Whereas basal levels of
HIF-1α is not critical for neutrophil endothelial transcytosis or oxidative burst function. (A) Hypoxia is
present in lesions generated by GAS infection. Immunostaining for hypoxic markers in WT mouse skin
upon GAS infection. Magnification, ×100 (top); ×200 (bottom). The control corresponds to the omission of
primary antibody. (B) Similar numbers of neutrophils in WT and HIF-1α–null mouse skin tissue observed
by immunostaining at 6, 12, and 24 hours after infection. Magnification, ×100. (C) Migratory capacity of
activated neutrophils across endothelium is not affected by the deletion of HIF-1α. Count of neutrophils
transcytosing pulmonary endothelial monolayer toward GAS or fMLP stimulus is shown. (D) HIF-1α activ-
ity does not affect oxidative burst capacity. Flow cytometry of leukocytes derived from WT (squares),
HIF-1α–null (triangles) and vHL-null (inverted triangles) mice. Oxidative burst capacity as measured by
fluorescence before (0 seconds) and after the addition of a reagent designed to stimulate leukocyte phago-
cytic and oxidative activity as described in Methods. Data are representative of the results obtained for 4
individuals per genotype.
? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 115 Number 7 July 2005
TNF-α protein expression were similar in WT, HIF-1α, and vHL-
null macrophages, loss of HIF-1α also strongly depressed the rapid
secretion of TNF-α protein in response to GAS (Figure 7B). As NO
is markedly induced under GAS stimulation, TNF-α induction
by GAS may rely on HIF-1α–dependent NO production. ELISA
for secreted TNF-α demonstrated reduced amounts of TNF-α
protein in conditioned supernatants of WT, HIF-1α, and vHL-
deficient macrophages in the presence of the iNOS inhibitor AG
(Figure 7B). This finding indicates that NO production, acting in a
HIF-1α controlled manner, contributes significantly to the macro-
phage TNF-α response to bacterial infection.
Our studies use conditional gene targeting in the myeloid cell lin-
eage to demonstrate that HIF-1α transcriptional regulation plays
an important role in innate immunity to bacterial infection. Acti-
vation of HIF-1α under hypoxia enhances bactericidal activity,
and HIF-1α pathways are responsive to bacterial stimulation even
under normoxia. While certain myeloid cell functions, including
endothelial transmigration and respiratory burst activation, appear
to be independent of HIF-1α control, the transcription factor is
involved directly or indirectly in the regulation of specific immune
functions including NO, granule proteases (cathepsin G, NE),
and cathelicidin antimicrobial peptides. The marked reduction of
granule protease and cathelicidin expression in HIF-1α–deficient
neutrophils correlates with diminished bactericidal activity in vitro
and failure to control infection in vivo, lending support to recent
studies uncovering a key role for these neutrophil effectors in mam-
malian innate immunity (15, 24).
Successful control of infection in the peripheral tissues requires
that host myeloid phagocytic cells function effectively in hypoxic
environments. The challenge to immune defense is made more
critical when the microbial toxins or local edema damage host
cells and the vascular supply of oxygen to the tissues becomes fur-
ther compromised. The placement of essential microbial killing
functions of myeloid cells under regulation of HIF-1α therefore
represents an elegant controlled-response?system (Figure 8). Bac-
tericidal mechanisms can be maintained in an “off” state while the
myeloid cells circulate in the oxygen-rich bloodstream and then
be activated in response to the declining oxygen gradient encoun-
tered upon diapedesis and entry of the phagocytes into the infect-
ed tissues. Additional, more potent stimulation of the HIF-1α
transcriptional pathway is then provided by direct encounter with
the bacteria (Figure 1A). A regulatory mechanism by which HIF-1α
targets genes involved in microbial killing ensures that the corre-
sponding inflammatory mediators are expressed preferentially in
tissue foci of infection but not in healthy tissues where inflamma-
tory damage might otherwise harm host cells.
Our experiments also reveal that NO production is a myeloid
cell–killing mechanism principally regulated by HIF-1α during
bacterial infection. Further, we suggest that NO is likely to play
a key role in the amplification of the inflammatory response
through stimulation of TNF-α release. Although the effects of
inflammatory cytokines on regulating NO production have been
extensively studied (38–40), the reverse relationship pertaining to
the effect of NO on cytokines remains controversial (41–44). A
recent study demonstrated that suppression of NO could inhibit
LPS-induced TNF-α and IL-1 release and pinpointed such modu-
lation to the pretranslational level (45). We find here that macro-
phage production of TNF-α is dependent on NO levels controlled
in turn by HIF-1α transcriptional regulation of iNOS.
Recent data has established that HIF-1α is subjected to stabil-
ity regulation by soluble intracellular messengers, such as NO and
TNF-α (33, 34). With such processes at play, one can speculate that
HIF1-α is situated at the center of an amplification loop mechanism
for innate immune activation: stimulation of HIF-1α by oxygen
depletion and bacterial exposure induces the production of NO and
TNF-α, which function not only to generate inflammation and con-
trol bacterial proliferation but also as regulatory molecules to further
stabilize HIF-1α in myeloid cells recruited to the infectious focus.
The relative contributions of HIF-1 and HIF-2 to the regulation
of gene expression in hypoxic macrophages is still under debate.
Production of granule proteases and of
murine CRAMP is regulated by HIF-1α.
NE (A) and cathepsin G (B) activity in
WT, HIF-1α–null, vHL-null and in a mix
of WT and HIF–/– blood leukocytes. (C)
Neutrophils were processed for immu-
noblotting with anti-CRAMP antibody
(upper panels) or anti–β-actin antibody
(lower panels). (D) HIF-1α regulates
CRAMP at the mRNA level. Neutrophils
were cultured under normoxic or
hypoxic (0.1%) conditions. Total neu-
trophil RNA was extracted and mRNA
polyA+ isolated by an Oligotex mRNA
spin-column protocol (QIAGEN). WT
neutrophils were arbitrarily set to 1 unit
following normalization to β-actin RNA
levels. Statistical analyses were per-
formed using unpaired Student’s t test.
*P < 0.05; **P < 0.01; ***P < 0.001.
1812?The?Journal?of?Clinical?Investigation http://www.jci.org Volume 115 Number 7 July 2005
Detectable levels of HIF-2α, but not HIF-1α, have been found in
a human promonocytic cell line following hypoxic induction in
vitro and in tumor-associated macrophages (46, 47). In contrast,
immunoreactive HIF-1α has been detected in human macrophages
in the hypoxic synovia of arthritic human joints (10), and human
macrophages accumulate higher levels of HIF-1 than of HIF-2 when
exposed to tumor-specific levels of hypoxia in vitro (9). Our present
results also clearly support a specific and independent action of
HIF-1α. Taken together, these findings suggest that HIF-1 may be
the major hypoxia-inducible transcription factor in macrophages.
In summary, our results demonstrate that HIF-1α not only helps
myeloid cells shift to glycolytic metabolism (11) but also functions
in coordinating a proper innate immune response for bacterial kill-
ing in hypoxic microenvironments. The in vivo studies confirm that
the HIF-1α pathway can play a critical role in controlling prolifera-
tion of a bacterial pathogen in compromised tissues. Recent com-
mentaries have suggested that downregulation of HIF-1α could
have a therapeutic effect in disease states characterized by chronic
inflammation (48, 49). We now have shown that medically impor-
tant bacterial species such as GAS, methicillin-resistant S. aureus
(MRSA), P. aeruginosa, and Salmonella species can trigger HIF-1α
expression. Thus, the present studies suggest that rational design
of pharmaceutical HIF-1α agonists (or vHL antagonists) to boost
myeloid cell microbicidal activity may likewise represent a novel
approach for adjunctive therapy of complicated infections due to
antibiotic-resistant pathogens or compromised host immunity.
All procedures involving animals were reviewed?by the University of Cali-
fornia San Diego Animal Care Committee, which serves to ensure that all
federal guidelines concerning animal experimentation are met.
Harvest of neutrophils, macrophages, and blood leukocytes. Neutrophils were
either isolated from the peritoneal cavity 3 hours after injection of thio-
glycollate as previously described (11, 50) or derived from bone marrow
as described (51). To isolate BM-derived macrophages, the marrow of
femurs and tibias of WT, HIF-1α myeloid–null, or vHL myeloid–null
mice were collected. Cells were plated in DMEM supplemented with 10%
heat-inactivated FBS and 30% conditioned medium (a 7-day superna-
tant of fibroblasts from cell line L-929 stably transfected with an M-CSF
expression vector). Mature adherent BM cells were harvested by gentle
scraping after 7 days in culture. To isolate blood leukocytes, 200–500
µl of whole blood was collected by retroorbital bleed into cold EDTA-
coated capillary tubes (Terumo Medical Corp.). Cells were centrifuged,
erythrocytes were lysed using ACK RBS lysis buffer (0.15 M NH4Cl, 10.0
mM KHCO3, 0.1 mM EDTA), and unlysed cells were washed once with
1 ml PBS 1% BSA.
Bacterial strains and media. GAS strain 5448 is an M1 serotype isolate from
a patient with necrotizing fasciitis and toxic shock syndrome (52). Addi-
tional bacterial strains were obtained from the ATCC Bacteriology collec-
tion, specifically methicillin-resistant S. aureus (ATCC 33591, designation
328), S. typhimurium (ATCC 1311), and P. aeruginosa (ATCC 27853, desig-
nation Boston 41501). GAS was propagated in Todd-Hewitt broth (THB)
(Difco; BD Diagnostics) and other strains in Luria-Bertani broth.
HIF-1α and vHL regulate NO production. (A)
Total RNA from WT, HIF-1α–/–, and vHL–/–
bone marrow–derived macrophages infect-
ed with GAS isolated 3 hours after antibiotic
treatment. iNOS mRNA was quantified by
RT-PCR. WT, nonstimulated macrophages
were arbitrarily set to 1 unit following nor-
malization to ribosomal RNA levels. (B) NO
production under GAS stimulation ± 1.5
mM AG (1-amino-2-hydroxyguanidine, p-
toluenesulfate; Calbiochem). BM-derived
macrophages were cultured for 20 hours,
conditioned supernatant collected, and NO
protein levels measured by the Griess assay.
(C) Mim enhances iNOS expression of WT
macrophages stimulated by GAS. Total
RNA from WT and HIF-1α–null BM-derived
macrophages infected with GAS ± Mim iso-
lated 3 hours after antibiotic treatment. iNOS
mRNA was quantified by RT-PCR. WT, non-
infected macrophages were arbitrarily set to
1 unit following normalization to ribosomal
RNA levels. Statistical analyses performed
by unpaired Student’s t test. **P < 0.01;
***P < 0.001. (D) Inhibition of iNOS by AG
blunts observed differences between HIF-
1α–null and WT microbicidal activity. (E)
Inhibition of iNOS prevents GAS-induced
HIF-1α expression. Expression of HIF-1α is
normalized to β-actin levels.
?The?Journal?of?Clinical?Investigation http://www.jci.org Volume 115 Number 7 July 2005
Bacterial killing assays. GAS were grown to logarithmic phase in THB to
OD600 = 1 × 108 cfu/ml. Bacteria were added to macrophages at an MOI
of 2.5 bacteria/cell and intracellular killing assessed using an antibiotic
protection assay (11, 53) or, alternatively, total cell-associated bacteria
measured by vigorous washing with PBS × 3 to remove nonadherent bac-
teria. At the end of the assay, total cell lysate was plated on THB agar for
enumeration of CFU. Comparable studies were performed with P. aerugi-
nosa at an?MOI of 25. To assess macrophage viability, the monolayers were
washed with PBS and incubated with 0.04% Trypan blue for 10 minutes at
37°C. As specified in the Figure 2 legend and in Results, macrophages were
preincubated with L-Mim (800 µM), OH-pyridone (150 µM), desferriox-
amine mesylate (100 µM), or CoCl2 (100 µM) for 5 hours prior to the kill-
ing assay; each drug level was known to be sufficient for HIF-1α induction
(13). Absence of bacterial inhibition was tested by incubating the drugs at
the above concentrations with GAS (∼105) at 37°C for 1–24 hours.
Mouse model of GAS infection. An established model of GAS subcutaneous
infection was adapted for our studies (24, 54). Briefly, 100 µl of a midloga-
rithmic growth phase (∼107 cfu) of GAS was mixed with an equal volume
of sterile Cytodex beads (Sigma-Aldrich) and injected subcutaneously into
a shaved area on the flank of 5- to 8-week-old male littermates. Mice were
weighed daily and monitored for development of necrotic skin lesions.
After 96 hours, skin lesions, spleen, and blood (via retroorbital bleeding)
were collected and homogenized in 1:1 mg/ml PBS. Serial dilutions of the
mixture were plated on THB agar plates for enumeration of CFUs.
Immunohistochemistry. Lesions were processed, embedded into paraffin,
and routine sections (5 µm) cut. Immunohistochemistry was performed
with an antibody specific for neutrophils (purified anti-mouse neutrophils
mAb; Accurate Chemical & Scientific Corp.) as described (55). To assess
development of hypoxic regions within the lesions, mice were injected
intraperitoneally with 60 mg/kg (weight/volume?in PBS) pimonidazole
(Hydroxyprobe-1, Natural Pharmacia International Inc.) 2 hours prior to
sacrifice. Immunohistochemistry was performed with Hydroyprobe-1 mouse
monoclonal antibody as reported (56).
Reverse transcription and real-time quantitative PCR. First-strand synthe-
sis was obtained from 1 µg of total RNA isolated with Trizol reagent
(Molecular Research Center Inc.) by the SuperScript system (Invitrogen
Corp.), employing random primers. For real-time PCR (RT-PCR) analyses,
cDNAs were diluted to a final concentration of 10 ng/µl and amplified
in a TaqMan Universal Master Mix, SYBR Green (Applied Biosystems).
cDNA (50 ng) was used as a template to determine the relative amount
of mRNA by RT-PCR in triplicate (ABI PRISM 7700 Sequence Detection
System; Applied Biosystems), using specific primers with the follow-
ing sequences: iNOS forward 5′-ACCCTAAGAGTCACAAAATGGC-3′;
iNOS reverse 5′-TTGATCCTCACATACTGTGGACG-3′; TNF-α forward
5′-CCATTCCTGAGTTCTGCAAAGG-3′; TNF-α reverse 5′-AGGTAG-
GAAGGCCTGAGATCTTATC-3′; TNF-α probe 5′-6[FAM]AGTGGT
CAGGTTGCCTCTGTCTCAGAATGA[BHQ]-3′; CRAMP forward 5′-
CTTCAACCAGCAGTCCCTAGACA-3′; CRAMP reverse 5′-TCCAGGTC-
CAGGAGACGGTA-3′; elastase forward 5′-TGGCACCATTCTCCCGAG-
HIF-1α and vHL regulate TNF-α production. (A) Total RNA from WT,
HIF-1α–/–, and vHL–/– bone marrow–derived macrophages infected
with GAS were isolated 3 hours after antibiotic treatment. TNF-α
mRNA was quantified by RT-PCR. WT, nonstimulated macrophages
were arbitrarily set to 1 unit following normalization to ribosomal RNA
levels. (B) Inhibition of iNOS decreases TNF-α production. BM-derived
macrophages were cultured for 1 hour after antibiotic treatment. Con-
ditioned supernatant was harvested and TNF-α protein analyzed by
ELISA (eBiosciences). Statistical analyses were performed using
unpaired Student’s t test. ***P < 0.001.
Model for the role of HIF-1α in myeloid cell
innate immune function. Bactericidal mecha-
nisms can be maintained in an “off” state
while myeloid cells circulate in the oxygen-
rich bloodstream. Transendothelial migration
toward an infectious focus occurs in a HIF-1α
independent fashion, but upon diapedesis,
specific bactericidal mechanisms are acti-
vated through HIF-1α induction in response to
the declining oxygen gradient. Further potent
stimulation of the HIF-1α transcriptional path-
way is provided after direct encounter with the
infecting bacterial pathogen. HIF-1α regulates
the generation of critical molecular effectors of
immune defense, including granule proteases,
antimicrobial peptides, and TNF-α. HIF-1α also
stimulates the production of NO, which not only
acts as an antimicrobial agent and inflamma-
tory mediator but further amplifies myeloid cell
bactericidal activity via HIF-1α stabilization.
1814? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 115 Number 7 July 2005
3′; elastase reverse 5′-CATAGTCCACAACCAGCAGGC-3′; β-actin forward
5′-AGGCCCAGAGCAAGAGAGG-3′; and β-actin reverse 5′-TACATGGCT-
Nitrite determination. The concentration of nitrite (NO2–), the stable oxi-
dized derivative of NO, was determined in 100-µl aliquots of cell culture
supernatants transferred to 96-well plates. Essentially, 100 µl of Griess
reagent (1% sulfanilamide, 0.1% naphthylene diamine dihydrochloride,
2% H3PO4) was added per well, and the absorbances were measured at
540 nm in a microplate reader. Sodium nitrite diluted in culture medium
was used as standard.
Elastase and cathepsin G assays. For elastase measurement, 100 µl of 0.2M
Tris-HCL (pH 8.5) containing 1M NaCl was mixed with 50 µg of blood
leukocytes lysed in HTAB buffer containing 0.1M Tris-Cl, pH 7.6, 0.15 M
NaCl, and 0.5% hexadecyltrimethylammonium bromide. Next, 100 µl of
MeOSuc-Ala-Ala-Pro-ValNmec dissolved in DMSO at 20 mM was added to
the buffered enzyme to start the reaction. The hydrolysis of the substrate
was monitoring spectrofluorometrically using excitation at 370 nm and
emission at 460 nm. For cathepsin G quantitation, 20 µl of Suc-Ala-Ala-
Pro-Phe-NphNO2 dissolved at 20 mM in DMSO was diluted to 180 µl with
0.1 M HEPES buffer, pH 7.5. The reaction was started by the introduction
of 10 µg of blood neutrophils lysed in HTAB buffer, and the increase in
A410 was monitored.
Western blot studies. Peritoneal neutrophils or bone marrow–derived
macrophages inoculated with GAS were harvested and washed with PBS,
and proteins were extracted with HTAB or RIPA?buffers. Protein concentra-
tion was calculated using the Bio-Rad assay (Bio-Rad Laboratories). Fifty mil-
ligrams of protein or nuclear extracts were loaded on a 10% Tris-tricine gel
in an MES buffer (Invitrogen Corp.) or 3–8% Tris-tricine gel in a Tris-acetate
buffer (Invitrogen Corp.) for CRAMP and HIF-1α Western blot respectively.
Proteins were transferred to a nitrocellulose membrane; the membrane was
blocked in 5% nonfat milk in 0.2% Tween TBS and then incubated in primary
Ab diluted in 5% nonfat milk. The primary Abs used were rabbit anti-mouse
CRAMP against the CRAMP mature peptide and rabbit anti-mouse HIF-1α
(Cayman Chemical Co.). The secondary Ab was peroxidase-conjugated goat
anti-rabbit (DAKO Corp.). Immunoreactive proteins were detected using the
ECL chemiluminescent system (Amersham Biosciences).
Reporter assay. Macrophages were derived from the marrow of femurs
and tibiae of transgenic HRE-luciferase mice as described above. The lucif-
erase reporter gene in these mice is driven by 6 specific HRE sequences.
Cells were incubated with GAS or heat-inactivated GAS for 18 hours. As a
positive control, macrophages were incubated under hypoxia (1%) or with
the addition of L-Mim (800 µM), desferrioxamine mesylate (150 µM), or
CoCl2 (150 µM) during the same period of time. Cells were then washed
out with PBS, and luciferase assay was performed by using the Bright-Glo
Luciferase Assay kit (Promega Corp.). Luciferase activities were measured
using a luminometer.
Oxidative burst assay. Isolated total blood leukocytes were resuspended
at 4°C in approximately 200 µl endotoxin- and pyrogen-free PBS, lack-
ing Ca2+ and Mg2+ but containing 5 mM glucose. Immediately before the
oxidative burst assay, 200 µl of PBS at 37°C containing 1.5 mM Mg2+ and
1.0 mM Ca2+ were added to the cell suspension. Oxidative burst activity
was measured by using the Fc OxyBURST Green assay reagent (Invitrogen
Corp.) according to the manufacturer’s instructions.
Endothelial cell transmigration assay. Thioglycolate-stimulated neutrophils
were added to the upper chamber of a Transwell membrane (Corning
HTS) coated with a primary murine pulmonary endothelial monolayer.
The chemokine fMLP (8 ng/µl) or GAS (MOI = 10:1) was added to the
lower well. The number of neutrophils migrating to the lower chamber was
counted after 1 hour of incubation at 37°C.
Reagents. AG and 1400W were purchased from EMD Biosciences.
L-Mim, OH-pyridone, desferrioxamine mesylate, and CoCl2 were purchased
This work was supported by NIH grants CA82515 (to R.S. John-
son) and AI48694 (to V. Nizet), a La Ligue Nationale Contre le
Cancer fellowship (to C. Peyssonnaux), a Deutsche Forschun-
gsgemeinschaft fellowship (to T. Cramer), and a grant from the
Edward J. Mallinckrodt, Jr. Foundation (to V. Nizet). We would
like to acknowledge Kelly Doran, Ronak Beigi, Wayne McNulty,
Dominique Sawka, and other members of the Johnson and Nizet
laboratories for helpful suggestions and assistance.
Received for publication November 9, 2004, and accepted in
revised form March 29, 2005.
Address correspondence to: Victor Nizet, Department of Pediat-
rics, University of California, San Diego, 9500 Gilman Drive, Mail
Code 0687, La Jolla, California 92093-0687, USA. Phone: (858)
534-7408; Fax: (858) 534-5611; E-mail: firstname.lastname@example.org. Or to:
Randall S. Johnson, Molecular Biology Section, University of Cali-
fornia, San Diego, 9500 Gilman Drive, Mail Code 0377, La Jolla,
California 92093-0377, USA. Phone: (858) 822-0509; Fax: (858)
822-5833; E-mail: email@example.com.
Thorsten Cramer’s present address is: Section of Molecular Gas-
troenterology, Charite-Hochschulmedizin Berlin, Campus Vir-
chow-Clinic, Berlin, Germany.
1. Vogelberg, K.H., and Konig, M. 1993. Hypoxia of
diabetic feet with abnormal arterial blood flow.
Clin. Investig. 71:466–470.
2. Arnold, F., West, D., and Kumar, S. 1987. Wound
healing: the effect of macrophage and tumour
derived angiogenesis factors on skin graft vascu-
larization. Br. J. Exp. Pathol. 68:569–574.
3. Negus, R.P., Stamp, G.W., Hadley, J., and Balkwill,
F.R. 1997. Quantitative assessment of the leuko-
cyte infiltrate in ovarian cancer and its relationship
to the expression of C-C chemokines. Am. J. Pathol.
4. Semenza, G.L. 2001. HIF-1 and mechanisms of
hypoxia sensing. Curr. Opin. Cell Biol. 13:167–171.
5. Flamme, I., et al. 1997. HRF, a putative basic helix-
loop-helix-PAS-domain transcription factor is
closely related to hypoxia-inducible factor-1α and
developmentally expressed in blood vessels. Mech.
6. Ema, M., et al. 1997. A novel bHLH-PAS factor with
close sequence similarity to hypoxia-inducible fac-
tor 1α regulates the VEGF expression and is poten-
tially involved in lung and vascular development.
Proc. Natl. Acad. Sci. U. S. A. 94:4273–4278.
7. Tian, H., McKnight, S.L., and Russell, D.W. 1997.
Endothelial PAS domain protein 1 (EPAS1), a tran-
scription factor selectively expressed in endothelial
cells. Genes Dev. 11:72–82.
8. Gu, Y.Z., Moran, S.M., Hogenesch, J.B., Wartman,
L., and Bradfield, C.A. 1998. Molecular character-
ization and chromosomal localization of a third
α-class hypoxia inducible factor subunit, HIF3α.
Gene Expr. 7:205–213.
9. Burke, B., et al. 2002. Expression of HIF-1α by
human macrophages: implications for the use of
macrophages in hypoxia-regulated cancer gene
therapy. J. Pathol. 196:204–212.
10. Hollander, A.P., Corke, K.P., Freemont, A.J., and
Lewis, C.E. 2001. Expression of hypoxia-induc-
ible factor 1α by macrophages in the rheumatoid
synovium: implications for targeting of therapeu-
tic genes to the inflamed joint. Arthritis Rheum.
11. Cramer, T., et al. 2003. HIF-1α is essential for
myeloid cell-mediated inflammation. Cell.
12. Schwartz, R.H., Gerber, M.A., and McCoy, P. 1985.
Effect of atmosphere of incubation on the isola-
tion of group A streptococci from throat cultures.
J. Lab. Clin. Med. 106:88–92.
13. Warnecke, C., et al. 2003. Activation of the hypox-
ia-inducible factor-pathway and stimulation of
angiogenesis by application of prolyl hydroxylase
inhibitors. FASEB J. 17:1186–1188.
14. Kong, T., Eltzschig, H.K., Karhausen, J., Colgan,
S.P., and Shelley, C.S. 2004. Leukocyte adhesion
during hypoxia is mediated by HIF-1-dependent
induction of β2 integrin gene expression. Proc. Natl.
Acad. Sci. U. S. A. 101:10440–10445.
15. Reeves, E.P., et al. 2002. Killing activity of
? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 115 Number 7 July 2005
neutrophils is mediated through activation of
proteases by K+ flux. Nature. 416:291–297.
16. Guyonnet, V., Johnson, J.K., Bangalore, N., Tra-
vis, J., and Long, P.L. 1991. In vitro activity of the
human neutrophil cathepsin G on Eimeria tenella
sporozoites. J. Parasitol. 77:775–779.
17. Miyasaki, K.T., and Bodeau, A.L. 1991. In vitro kill-
ing of oral Capnocytophaga by granule fractions of
human neutrophils is associated with cathepsin G
activity. J. Clin. Invest. 87:1585–1593.
18. Shafer, W.M., and Onunka, V.C. 1989. Mechanism
of staphylococcal resistance to non-oxidative antimi-
crobial action of neutrophils: importance of pH and
ionic strength in determining the bactericidal action
of cathepsin G. J. Gen. Microbiol. 135:825–830.
19. Barrett, A.J. 1981. Cathepsin G. Methods Enzymol.
20. Cole, A.M., et al. 2001. Inhibition of neutrophil
elastase prevents cathelicidin activation and
impairs clearance of bacteria from wounds. Blood.
21. Sorensen, O.E., et al. 2001. Human cathelicidin,
hCAP-18, is processed to the antimicrobial peptide
LL-37 by extracellular cleavage with proteinase 3.
22. Nizet, V., and Gallo, R.L. 2003. Cathelicidins and
innate defense against invasive bacterial infection.
Scand. J. Infect. Dis. 35:670–676.
23. Panyutich, A., Shi, J., Boutz, P.L., Zhao, C., and
Ganz, T. 1997. Porcine polymorphonuclear leuko-
cytes generate extracellular microbicidal activity by
elastase-mediated activation of secreted proproteg-
rins. Infect. Immun. 65:978–985.
24. Nizet, V., et al. 2001. Innate antimicrobial peptide
protects the skin from invasive bacterial infection.
25. Lirk, P., Hoffmann, G., and Rieder, J. 2002. Induc-
ible nitric oxide synthase–time for reappraisal.
Curr. Drug Targets Inflamm. Allergy. 1:89–108.
26. Gyurko, R., et al. 2003. Mice lacking inducible
nitric oxide synthase demonstrate impaired kill-
ing of Porphyromonas gingivalis. Infect. Immun.
27. Alam, M.S., et al. 2002. Role of nitric oxide in host
defense in murine salmonellosis as a function of
its antibacterial and antiapoptotic activities. Infect.
28. Melillo, G., et al. 1997. Functional requirement of
the hypoxia-responsive element in the activation
of the inducible nitric oxide synthase promoter
by the iron chelator desferrioxamine. J. Biol. Chem.
29. Jung, F., Palmer, L.A., Zhou, N., and Johns, R.A.
2000. Hypoxic regulation of inducible nitric oxide
synthase via hypoxia inducible factor-1 in cardiac
myocytes. Circ. Res. 86:319–325.
30. Hu, R., Dai, A., and Tan, S. 2002. Hypoxia-induc-
ible factor 1α upregulates the expression of induc-
ible nitric oxide synthase gene in pulmonary arter-
ies of hypoxic rat. Chin. Med. J. 115:1833–1837.
31. Thornton, R.D., et al. 2000. Interleukin 1 induces
hypoxia-inducible factor 1 in human gingival and
synovial fibroblasts. Biochem. J. 350:307–312.
32. Hellwig-Burgel, T., Rutkowski, K., Metzen, E., Fan-
drey, J., and Jelkmann, W. 1999. Interleukin-1β and
tumor necrosis factor-α stimulate DNA binding of
hypoxia-inducible factor-1. Blood. 94:1561–1567.
33. Sandau, K.B., Fandrey, J., and Brune, B. 2001. Accu-
mulation of HIF-1α under the influence of nitric
oxide. Blood. 97:1009–1015.
34. Zhou, J., Fandrey, J., Schumann, J., Tiegs, G., and
Brune, B. 2003. NO and TNF-alpha released from
activated macrophages stabilize HIF-1alpha in
resting tubular LLC-PK1 cells. Am. J. Physiol. Cell
35. Zhou, J., Schmid, T., and Brune, B. 2003. Tumor
necrosis factor-alpha causes accumulation of a
ubiquitinated form of hypoxia inducible factor-1α
through a nuclear factor-kappaB-dependent path-
way. Mol. Biol. Cell. 14:2216–2225.
36. Kasuno, K., et al. 2004. Nitric oxide induces hypox-
ia-inducible factor 1 activation that is dependent
on MAPK and phosphatidylinositol 3-kinase sig-
naling. J. Biol. Chem. 279:2550–2558.
37. Chan, A.T., Cleeve, V., and Daymond, T.J. 2002.
Necrotising fasciitis in a patient receiving inflix-
imab for rheumatoid arthritis. Postgrad. Med. J.
38. Nemoto, Y., et al. 1999. Differential effects of
interleukin-4 and interleukin-10 on nitric oxide
production by murine macrophages. Inflamm. Res.
39. Drapier, J.C., Wietzerbin, J., and Hibbs, J.B., Jr.
1988. Interferon-gamma and tumor necrosis factor
induce the L-arginine-dependent cytotoxic effector
mechanism in murine macrophages. Eur. J. Immu-
40. Corradin, S.B., et al. 1993. Induction of macrophage
nitric oxide production by interferon-gamma and
tumor necrosis factor-alpha is enhanced by inter-
leukin-10. Eur. J. Immunol. 23:2045–2048.
41. Eigler, A., Moeller, J., and Endres, S. 1995. Exoge-
nous and endogenous nitric oxide attenuates tumor
necrosis factor synthesis in the murine macrophage
cell line RAW 264.7. J. Immunol. 154:4048–4054.
42. Marcinkiewicz, J., Grabowska, A., and Chain,
B. 1995. Nitric oxide up-regulates the release of
inflammatory mediators by mouse macrophages.
Eur. J. Immunol. 25:947–951.
43. Kuo, H.P., et al. 2000. Nitric oxide modulates inter-
leukin-1beta and tumor necrosis factor-alpha syn-
thesis by alveolar macrophages in pulmonary tuber-
culosis. Am. J. Respir. Crit. Care Med. 161:192–199.
44. Iuvone, T., D’Acquisto, F., Carnuccio, R., and Di
Rosa, M. 1996. Nitric oxide inhibits LPS-induced
tumor necrosis factor synthesis in vitro and in vivo.
Life Sci. 59:PL207–PL211.
45. Wu, C.H., et al. 2003. Nitric oxide modulates pro-
and anti-inflammatory cytokines in lipopolysaccha-
ride-activated macrophages. J. Trauma. 55:540–545.
46. Griffiths, L., et al. 2000. The macrophage - a novel
system to deliver gene therapy to pathological
hypoxia. Gene Ther. 7:255–262.
47. Talks, K.L., et al. 2000. The expression and distri-
bution of the hypoxia-inducible factors HIF-1α
and HIF-2α in normal human tissues, cancers,
and tumor-associated macrophages. Am. J. Pathol.
48. Nathan, C. 2003. Immunology: oxygen and the
inflammatory cell. Nature. 422:675–676.
49. Strieter, R.M. 2003. Mastering innate immunity.
Nat. Med. 9:512–513.
50. Clausen, B.E., Burkhardt, C., Reith, W., Renkawitz,
R., and Forster, I. 1999. Conditional gene targeting
in macrophages and granulocytes using LysMcre
mice. Transgenic Res. 8:265–277.
51. Walmsley, S.R., et al. 2005. Hypoxia-induced neu-
trophil survival is mediated by HIF-1α-dependent
NF-κB activity. J. Exp. Med. 201:105–115.
52. Kansal, R.G., McGeer, A., Low, D.E., Norrby-
Teglund, A., and Kotb, M. 2000. Inverse relation
between disease severity and expression of the
streptococcal cysteine protease, SpeB, among clon-
al M1T1 isolates recovered from invasive group
A streptococcal infection cases. Infect. Immun.
53. Liu, G.Y., et al. 2004. Sword and shield: linked
group B streptococcal β-hemolysin/cytolysin
and carotenoid pigment function to subvert host
phagocyte defense. Proc. Natl. Acad. Sci. U. S. A.
54. Humar, D., et al. 2002. Streptolysin S and necrotis-
ing infections produced by group G streptococcus.
55. Grunstein, J., Roberts, W.G., Mathieu-Costello,
O., Hanahan, D., and Johnson, R.S. 1999. Tumor-
derived expression of vascular endothelial growth
factor is a critical factor in tumor expansion and
vascular function. Cancer Res. 59:1592–1598.
56. Blouw, B., et al. 2003. The hypoxic response of
tumors is dependent on their microenvironment.
Cancer Cell. 4:133–146.