INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Intracellular pathogens obtain shelter from the immune response and access to intracellular nutrients from within their hosts. However, to exploit this niche, these parasites must be able to evade the innate destructive mechanisms of their host cells. Organisms such as Mycobacteria tuberculosis (17), Chlamydia psittaci (14, 55), and Legionella pneumophila (26) achieve this evasion by preventing phagolysosomal fusion, while Shigellae (20, 49) and Listeria monocytogenes (18) escape from endosomes into the cytoplasm. Both mechanisms allow these bacteria to survive by avoiding the low pH, oxygen radicals, and degradative proteases of the phagolysosome.

Coxiella burnetii, however, employs no mechanisms to avoid the phagolysosome and indeed requires this harsh environment for replication. Vacuoles containing C. burnetii have characteristics of typical phagolysosomes, as shown by the presence of lysosomal membrane markers in vacuole membranes (23) as well as staining positively for the lysosomal enzymes, 5'-nucleotidase (8), and acid phosphatase (1, 8). The internal pH of vacuoles containing C. burnetii, about 4.8 (35), is identical to that of phagolysosomes. This acidic pH appears to be essential for C. burnetii metabolic activity. In vitro experiments find that C. burnetii is able to catabolize glucose and glutamate, generate ATP, and synthesize nucleic acids and proteins at low but not neutral pH. This is consistent with the in vivo observation that C. burnetii replication is inhibited by raising the phagolysosomal pH with lysomotrophic amines (19). Thus, C. burnetii is adapted to a phagolysosomal environment where it is able to flourish by means that are as yet poorly understood.

Current evidence indicates that upon infection C. burnetii is able to modify responses that accompany phagocytosis, perhaps ameliorating its environment to some degree. C. burnetii infection is known to reduce the respiratory burst of macrophages, an integral part of the bactericidal response during infection (3). Phagolysosomes containing C. burnetii appear to be more fusion permissive than phagolysosomes inhabited by another obligate intracellular pathogen, Leishmania amazonensis (42, 52, 53). Additionally, phagolysosomes containing C. burnetii coalesce into a single large vacuole, indicating fusion permissibility (22).

The present study documents another consequence of C. burnetii infection, an increase in transferrin receptor (TfR) expression. The significance of these observations is twofold. First, we find that intracellular iron is essential for C. burnetii replication, and thus upregulation of TfR may facilitate C. burnetii sequestration of this vital nutrient. Second, an increase in the intracellular iron in macrophages is known to depress killing of intracellular pathogens (37), and thus upregulation of TfR may promote C. burnetii survival.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Host cell lines and bacterial propagation. J774A.1, a murine macrophage-like cell line (TIB-67; American Type Culture Collection, Rockville, Md.), was maintained in RPMI 1640 medium (RPMI; Gibco BRL, Grand Island, N.Y.) supplemented with 10% heat-inactivated fetal bovine serum (FBS) in 5% CO₂ at 37°C. C. burnetii Nine Mile phase I (48) was propagated in J774A.1 cells in the same conditions as noted above. C. burnetii was harvested by differential centrifugation of cell culture supernatants first at 5,500 × g for 10 min to pellet host cell debris and then at 15,000 × g for 1 h to pellet the bacteria. Numbers of C. burnetii were estimated by measuring the turbidity of bacterial suspensions with a Klett Summerson photoelectric colorimeter (Klett Manufacturing Company Incorporated, New York, N.Y.), using a no. 42 filter (40); for a given experiment, densities were adjusted to provide the desired multiplicity of infection (MOI). C. burnetii was inactivated either chemically, by incubation for 12 h at room temperature with vigorous shaking in 2% glutaraldehyde-2% paraformaldehyde, or physically, by incubation for 2 h at 80°C. C. burnetii inactivated by both these methods were compared in all experiments, and no differences were seen (results not shown).

Immunoblotting. J774A.1 cells were seeded onto six-well Falcon plates (VWR Scientific Products, Seattle, Wash.) at 10⁶ cells per well, allowed to adhere for 12 h, and then incubated for 1 h with viable or inactivated C. burnetii at 10⁸ per ml in RPMI-10% FBS. Extracellular C. burnetii was removed by three washes with medium at 37°C. At designated time points, the medium was removed, cells were washed three times with phosphate-buffered saline (PBS; 0.1 M NaPO₄, 0.1 M NaCl [pH 7.4]), and cell numbers were determined. Cells were pelleted, suspended to 5 × 10⁶ per ml in deionized distilled water containing 2 mM phenylmethylsulfonyl fluoride, and frozen. Samples of 5 × 10⁴ cells were denatured by boiling in Laemmli's buffer (30), resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 10% polyacrylamide gels, then electrotransferred to polyvinylidene difluoride membranes by standard techniques. Membranes were blocked in Blotto (PBS with 10% skim milk powder and 0.1% Tween 20) for 1 h following protein transfer and then probed with rat anti-mouse TfR monoclonal antibodies (kindly provided by R. Hyman, Salk Institute, San Diego, Calif.) (32) followed by goat anti-rat immunoglobulin G conjugated to horseradish peroxidase (Pierce, Rockford, Ill.). Signal was visualized with an enhanced chemiluminescence kit (Amersham, Cleveland, Ohio), and images were analyzed with an Image Master scanning densitometer and associated software (Pharmacia Biotech, Uppsala, Sweden).

Fluorescent labeling of TfR. Confocal microscopy was used to compare TfR expression seen during C. burnetii infection with that seen in response to soluble factors in conditioned medium. J774A.1 cells were seeded on 15-mm-diameter Thermonox cover slips (Fisher Scientific, Santa Barbara, Calif.) in 24-well plates at 2.5 × 10⁵ per well and allowed to adhere for 12 h. Then they were inoculated with 10⁸ viable or inactivated C. burnetii (both in RPMI-10% FBS) or exposed to filter-sterilized conditioned medium (harvested from a 3-day culture of persistently infected J774A.1 cells) for an additional 16 h. At the end of this period, more than 95% of the cells were found to have taken up the inoculum. Following treatment, coverslips were rinsed with PBS, fixed in 0.1 M lysine HCl-2% paraformaldehyde-7.25 mM sodium periodate in 0.1 M NaPO₄ buffer (pH 7.4) for 1 h, and washed three times in PBS. The samples were then blocked and permeabilized with 0.1% gelatin-0.3% saponin in PBS for 15 min, incubated with rat anti-mouse TfR monoclonal antibody for 1 h, washed three times, and probed with a fluorescein isothiocyanate-labeled rat secondary antibody (Organon Teknika Corp., West Chester, Pa.) for 1 h. Secondary antibody was washed off once for 10 min with blocking solution and twice for 10 min with PBS. Samples were fixed for 1 h at room temperature in PBS containing 4% paraformaldehyde. A Prolong antifade kit (Molecular Probes Inc., Eugene, Oreg.) was used as directed. A Bio-Rad 1024 laser scanning confocal microscopy imaging system equipped with a krypton-argon laser (Bio-Rad) on a Nikon inverted microscope with a 60× objective was used for laser scanning confocal microscopy. Images were processed with Adobe Photoshop 4.0 (Adobe Systems, Inc., Mountain View, Calif.).

Determination of intracellular iron content. J774A.1 cells were seeded onto six-well plates at 10⁶ cells per well, allowed to adhere for 12 h, and then incubated for 1 h with viable or inactivated C. burnetii at 10⁸ per ml in RPMI-10% FBS. Removal of extracellular C. burnetii was facilitated by three washes with 37°C medium. At designated time points, the medium was removed, cells were washed three times with 0.7% saline solution, and cells were removed by scraping. Following determination of cell numbers and viability, cells were pelleted, saline was removed, and cells were stored at 80°C. For analysis, samples were suspended in 100 µl of 6 N HNO₃ at 5 × 10⁶ cells per ml, incubated at room temperature for 12 h, heated to 90°C for 3 h, and then diluted to 1 ml with deionized distilled H₂O. Iron content was then determined by inductively coupled plasma emission, using a Jobin Yvon JY24 instrument (Instrumental Services, Livermore, Calif.).

Chelation of intracellular free iron. J774A.1 cells were seeded at 4 × 10⁵ cells per well onto six-well plates and allowed to adhere for 12 h. Cells were incubated for 24 h with C. burnetii resuspended at 10⁸ per ml in RPMI-10% FBS containing fresh 0, 20, or 100 µM desferrioxamine (DfX; Sigma, St. Louis, Mo.). Extracellular C. burnetii was removed by washing, and incubations continued in medium containing DfX. Cultures were evaluated at 24, 48, and 72 h postinfection. Cells were stained with the viability indicator trypan blue, and total cell numbers were determined by hemocytometer. Infection levels were determined by examining J774A.1 cells stained with Gimenez reagent (16) that were mounted on glass slides with a Cytospin centrifuge (Shandon Inc., Pittsburgh, Pa.). For each treatment, at least 300 cells were scored as having 0, 0 to 10, 10 to 20, 20 to 50, 50 to 100 or more than 100 C. burnetii bacteria per cell. From these numbers, the C. burnetii count per well was estimated for each treatment. Cell numbers and viability were also determined for identically maintained uninoculated J774A.1 cells.

Incorporation of [³⁵S]methionine-[³⁵S]cysteine in vitro. C. burnetii bacteria harvested from the supernatant of persistently infected cell cultures were divided equally into six samples (approximately 10⁹/sample). Each sample was incubated at 37°C in citrate buffer (25 mM citric acid, 100 mM glycine, 65 mM KCl, 10 mM NaCl, 5 mM glutamate, 150 µM K₂HPO₄, 150 µM MgCl₂, 500 µM glucose, 20 µM proline, 50 µg cycloheximide/ml [pH 4.5 or 7.0]) containing FeSO₄ at 0, 0.1, or 0.2 mM. After 1 h, buffers were replaced and 50 µCi of [³⁵S]methionine-[³⁵S]cysteine (Express; NEN, Boston, Mass.) was added to each incubation and incubated a further 2 h. Bacteria were washed in PBS and precipitated in acetone-1 mM HCl.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

TfR expression in macrophages during infection by C. burnetii. Intracellular trafficking of iron in mammalian systems utilizes the iron carrier protein transferrin. Certain extracellular parasites such as Neisseria and Haemophilus spp. exploit this system by producing their own receptors for transferrin, through which they harvest sufficient iron to meet their metabolic needs (11). More recently, the intracellular parasite Ehrlichia chaffeensis was found to inhabit an endosome that accumulates TfRs in conjunction with an increase in TfR mRNA (5). In the present study, immunoblot analysis of SDS-PAGE-resolved lysates from uninfected and C. burnetii-infected cells were examined by using a monoclonal antibody specific for the murine TfR to determine if C. burnetii similarly modified TfR expression.

View larger version (21K): [in this window] [in a new window]   FIG. 1.   (A) To establish that TfR upregulation was not due to the large numbers of internalized C. burnetii necessary for visualization, J774A.1 cultures were infected with viable C. burnetii at the MOI indicated at the bottom. At 12 h postinfection, TfR expression was examined by immunoblot analysis. The optical density of bands on immunoblots was determined by scanning densitometry; the results shown are the means of three experiments, with error bars indicating standard errors. Student t tests indicate that there is a significant difference in the optical density of TfR bands from cultures infected at an MOI of 100 with P  0.05. (B) Representative immunoblot used to generate the data shown in panel A.

View larger version (30K): [in this window] [in a new window]   FIG. 2.   (A) The time course of TfR upregulation by C. burnetii infection was determined by immunoblotting. J774A.1 cells were inoculated with viable or inactivated C. burnetii. At designated time points, TfR expression was examined by immunoblot analysis. The optical density of bands was determined by scanning densitometry. TfR expression in cultures inoculated with inactivated C. burnetii show no increase in density, while infection with viable C. burnetii causes densities to increase by 24 h to 4.5 times that seen in uninoculated cells. Results shown are the mean of four experiments, with error bars indicating standard errors. Student t tests indicate that there is a significant difference in the optical density of TfR bands from infected samples taken at 9, 12, and 24 h and that for P < 0.1 (a) for P < 0.05 (b) and for P < 0.001 (c). (B) Representative immunoblots used to generate the data shown in panel A. Lane 12d shows TfR expression at 12 h by uninoculated J774A.1 cells.

TfR distribution and subcellular location. Confocal microscopy was used to visualize the increase in TfR after C. burnetii infection (Fig. 3). The intensity of fluorescence in cells infected with C. burnetii was uniformly distributed and 1.5-fold greater than that seen in either cells treated with inactivated C. burnetii or uninoculated control cells. Both infected cells and cells treated with conditioned medium exhibited a morphological change from spread out to rounded up (Fig. 3A and 3B). However, this morphological change did not correlate with increased TfR expression, as there was no increase in the fluorescence of cells incubated in conditioned medium, demonstrating that soluble factors were not responsible for TfR upregulation. The distribution of fluorescent label on the outside of C. burnetii-infected J774A.1 cells and a subcellular examination by transmission electron microscopy of TfR distribution (not shown) indicate that C. burnetii infection did not disturb the normal TfR trafficking.

View larger version (132K): [in this window] [in a new window]   FIG. 3.   J774A.1 cells were grown on Thermonox coverslips; TfRs were labeled with an anti-TfR monoclonal antibody and examined by confocal microscopy for TfR expression. (A) J774A.1 cells infected for 16 h with viable C. burnetii show increased TfR expression as well as a rounding up of infected cells. (B) J774A.1 cells incubated in conditioned medium from persistently infected J774A.1 cells. These cells show no upregulation of TfR but do exhibit the rounding up seen in infected cells. (C) J774A.1 cells inoculated with inactivated C. burnetii and (D) uninoculated J774A.1 cells exhibit a basal level of TfR expression and no morphological changes.

Iron concentration in macrophages during infection by C. burnetii. To determine if increased levels of TfR in C. burnetii-infected J774A.1 cells resulted in an increase in intracellular iron, cell lysates were analyzed for iron content by inductively coupled plasma emission (Fig. 4). Inoculation with inactivated C. burnetii did not alter the intracellular iron levels of J774A.1 cells. However, consistent with the observed increase in TfR expression, inoculation with viable C. burnetii resulted in a 2.5-fold increase in intracellular iron during the first 12 h. By 24 h postinfection, intracellular iron concentrations returned to preinfection levels.

View larger version (19K): [in this window] [in a new window]   FIG. 4.   An increase in intracellular iron would be expected to result from an upregulation of TfRs. The graph shows the intracellular iron content determined by inductively coupled plasma emission of J774A.1 cell cultures inoculated with viable (open circles) and inactivated C. burnetii (filled circles) as well as an uninoculated control (triangles). The results shown are the means for an experiment done in triplicate, and error bars indicate standard deviation. There are no error bars for two points at 36 h, as one data point was available for cells inoculated with inactivated C. burnetii and two were available for uninoculated cells.

Iron requirement for C. burnetii. The observed increase in TfR expression and total intracellular iron following C. burnetii infection suggests that C. burnetii may acquire iron through accessing the increased pool of intracellular iron. To verify a role for free intracellular iron, the cell-permeant iron chelator DfX was used to reduce the intracellular free iron pool. Cell cultures were treated with 0, 20, and 100 µM DfX were infected and then sampled at 24 and 72 h postinfection (Table 1). J774A.1 cells were scored by microscopic examination as having 0, 1 to 10, 11 to 20, 21 to 50, 51 to 100, or more than 100 C. burnetii organisms per cell. Less than 1% of scored J774A.1 cells had more than 50 intracellular C. burnetii organisms. Averages of each group, 0, 5, 15, 35, 75, or 200, were used to calculate average C. burnetii number per cell. Numbers and viability (above 80% at all time points) of the infected cells were determined for each well. By 72 h after infection, the number of C. burnetii organisms in cells treated with 100 µM DfX had decreased, whereas the number in untreated J774A.1 cells had increased sevenfold. The growth rate and viability of uninfected J774A.1 cells treated with 0, 20, or 100 µM DfX were the same as for similarly treated infected cells. These results suggest that the free intracellular iron pool is a source of iron for C. burnetii. However, as one would expect, removal of iron also slowed (but did not halt) host cell replication; thus, we cannot at this point exclude the possibility that stress on the host cell contributed to the observed depression in the numbers of C. burnetii organisms.

Effect of iron concentration on in vitro translation by C. burnetii. Experiments with DfX suggested that C. burnetii accesses free intracellular iron. If this is the case, then C. burnetii should be able to access iron in the absence of transferrin or any other host-derived iron carrier. To test this possibility, the effect of free iron on C. burnetii metabolism was assayed in vitro, using purified organisms. C. burnetii was incubated in a citrate buffer containing [³⁵S]methionine-[³⁵S]cysteine at pH 4.5 for 2 h after a 1-h preincubation in unlabeled buffers. We compared results of three independent experiments in which three aliquots of C. burnetii were incubated at 0, 0.1, or 0.2 mM iron. We found that at 0 mM iron, total incorporation of [³⁵S]methionine-[³⁵S]cysteine was 62% ± 9% that seen at 0.2 and 0.1 mM iron and that at 0.2 mM iron, two major proteins, 17 and 20 kDa, showed decreased expression (Fig. 5). These results demonstrated a C. burnetii response to free soluble iron, supporting the hypothesis that C. burnetii is able to obtain iron from the intracellular pool of free iron.

View larger version (48K): [in this window] [in a new window]   FIG. 5.   To determine the effect of soluble iron on metabolic activity, C. burnetii was labeled with [35S]methionine for 2 h in a citrate buffer (pH 4.5) containing 0, 0.1, or 0.2 mM iron. Proteins expressed during this in vitro incubation were visualized by autoradiography to compare expression. Equal numbers of cells were loaded onto all lanes. Total expression is reduced by 38% in an average of five experiments at 0 mM iron, while expression of a 20-kDa and a 17-kDa protein (indicated by arrows) is reduced at 0.2 mM iron.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In living organisms, the intracellular iron concentration must be tightly regulated. Iron is essential for a variety of cellular processes, including electron transport, numerous redox reactions, and DNA biosynthesis. Iron is also toxic, catalyzing the generation of oxygen radicals via the Haber reaction, damaging DNA and cellular membranes (2, 36, 47). Iron uptake in mammalian cells is controlled by the expression of TfRs and the concentration of extracellular iron-loaded transferrin. TfRs are upregulated in response to the depletion of intracellular iron (24, 54). This regulation occurs at the posttranscriptional level by stabilization of TfR mRNA through the interaction of iron-responsive proteins with iron-responsive elements on mRNA (41). There appear to be additional uncharacterized pathways, as exemplified by cytokine-induced control of TfR expression (54). In this study, an increase in TfR expression was observed when J774A.1 cells were infected with viable C. burnetii. Mechanisms that elicit this response are unknown. Soluble factors released by infected cells did not appear to be responsible for TfR upregulation. It is possible that C. burnetii interacts directly with regulation pathways that lead to TfR upregulation. On the other hand, host cells may respond to a decrease in available iron as a result of iron sequestration by C. burnetii.

To move iron into a mammalian cell, TfR binds transferrin, an iron binding serum protein, at the plasma membrane of the cell. The TfR-transferrin complex is internalized via a clathrin-coated pit, which is then incorporated into an early endosome (7, 21). Here vacuolar ATPases lower the internal pH of the endosome to about 5.5 (50), and in response to the lowered pH, the transferrin releases the iron while remaining bound to the TfR (33, 39). The empty transferrin (or apotransferrin)-TfR complex is returned to the plasma membrane; here the neutral pH causes the TfR to lose its affinity for the apotransferrin, and it is released into the serum (12, 28). Thus, TfRs are cycled from the plasma membrane to endosomes (25) and back to the plasma membrane constitutively carrying iron into the cell as long as iron-loaded transferrin is available.

The TfRs upregulated during C. burnetii infection appear to cycle normally, as demonstrated by the fluorescent labeling seen in Fig. 3, and one would expect to see an increase in intracellular iron levels. We found a doubling of intracellular iron by 12 h postinfection, but by 24 h it returned to the preinfection level. Because TfR levels do not return to preinfection levels at this same rate, we speculate that homeostatic mechanisms maintaining intracellular iron balance may achieve this balance (43). Establishment of increased iron traffic may provide an iron source for replicating C. burnetii.

In addition to enhancing the availability of an essential nutrient, the increase in intracellular iron associated with C. burnetii upregulation of TfRs may allow the proliferation of these intracellular pathogens within macrophages. A diet rich in iron enhances infection by intracellular pathogens, while an iron-deficient diet reduces infection (31, 37, 46). In vitro studies have shown that increased intracellular iron levels allow the proliferation of intracellular pathogens (13, 15, 34, 44), some by reversing the gamma interferon stimulation of macrophages that results in the killing of intracellular pathogens (4, 9, 29). The mechanisms of killing by gamma interferon-activated macrophages include an enhancement of the oxidative burst (38), an increase in nitric oxide levels (27), and the downregulation of TfRs accompanied by a reduction in intracellular iron (6, 9, 51). In macrophages infected with C. burnetii, the oxidative burst is reduced (1) and TfR levels are upregulated as reported here, possibly diminishing the ability of macrophages to eliminate these intracellular pathogens.

Certain intracellular pathogens appear to access free iron pools and indeed may require this free iron pool (5, 45). DfX removes intracellular free iron or iron which is not bound by host-chelating proteins. Treatment of infected mammalian cells with DfX halts proliferation of intracellular Plasmodium falciparum (44), E. chaffeensis (4), L. pneumophila (10, 15), and F. tularensis (13), and we found that C. burnetii proliferation was likewise halted by DfX treatment of infected cells. C. burnetii is unique among these pathogens in that it resides within a mature phagolysosome, where the availability of iron is unknown. These studies suggest that C. burnetii is able to exploit the intracellular free iron pool from within the phagolysosome. In vitro experiments support the hypothesis that C. burnetii is able to utilize free iron, as shown by reduced expression of all proteins when iron is limiting and reduced expression of two major proteins when iron concentration is doubled.

This study has shown for the first time that infection of J774A.1 cells with viable but not inactivated C. burnetii is accompanied by an upregulation of TfR expression and that this increase in TfR expression was concurrent with a transient increase in intracellular iron. Microscopy has shown that TfRs cycle normally, carrying iron into the host cells, as supported by an increase in intracellular iron. An increase in intracellular iron may provide an iron source for C. burnetii and/or result in modulation of the immune response to C. burnetii infection; thus, TfR upregulation may be an important virulence factor for C. burnetii.