INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Legionella pneumophila, the etiologic agent of Legionnaires' disease, is a gram-negative, facultative intracellular bacterium (10). Normally, L. pneumophila is an inhabitant of fresh water, where it lives as part of biofilms or as a parasite of protozoa (28). However, it can reach the human respiratory tract following inhalation or aspiration of contaminated water (20, 26, 60). In the lung, the legionellae replicate within alveolar macrophages, eventually causing much cell death and damage to lung tissue (24, 60).

The identification of an L. pneumophila pilD gene (31, 32) afforded new insight into the molecular pathogenesis of legionellosis. In a variety of gram-negative bacteria, PilD is an inner membrane peptidase that cleaves and methylates the pilin and pilin-like proteins that will form the type IV pili (2, 3, 21, 38, 49, 57). In addition, PilD processes a second set of pilin-like molecules which contribute to the formation of a type II protein secretion system (9, 21, 22, 33, 43, 48, 56). The two functions of PilD have been confirmed in L. pneumophila; i.e., a mutation in pilD renders bacteria nonpiliated and deficient for protein secretion (5, 31). The existence of the Legionella type II secretion system was later confirmed by the identification of some of the genes encoding components of the secretion apparatus (23). Importantly, the L. pneumophila pilD mutant is severely defective (i.e., ca. 1,000-fold) for replication within freshwater protozoa and human macrophages (31). Although type IV pili are modestly involved in the attachment of the bacteria to host cells, they are not critical for intracellular growth and survival (31, 55). Consequently, it became more significant to identify and characterize the secreted activities that are absent in the pilD mutant. Several Legionella enzymatic activities have recently been shown to be secreted in a pilD-dependent manner, including a zinc metalloprotease, phospholipase A, acid phosphatase, p-nitrophenyl phosphorylcholine (pNPPC) hydrolase, RNase, and esterase/lipase (5, 31). Since mutants specifically defective for the production of either the pNPPC hydrolase activity or the metalloprotease proved not to be impaired for intracellular infection (5, 58), the critical pilD-dependent factors remain to be identified.

For three reasons, we targeted the acid phosphatase activity for further analysis. First, phosphatase activity is diminished in type II secretion mutants that are defective for intracellular growth (O. Rossier and N. P. Cianciotto, submitted for publication). Second, our observations in Legionella represent the first connection between an acid phosphatase and PilD or type II secretion. Third, in other pathogens, acid phosphatases have been postulated to have a role in intracellular infection (7, 11, 30, 45, 46, 50, 51). In this work, we provide evidence that L. pneumophila possesses two acid phosphatase activities and report the identification and mutation of the major, pilD-dependent acid phosphatase. Although the major acid phosphatase proved to be unique among bacterial acid phosphatases in terms of its sequence and secretion, it was not essential for the intracellular replication of L. pneumophila in macrophages or amoebae.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacteria and media. L. pneumophila serogroup 1 strain 130b (ATCC BAA 74), a virulent clinical isolate, and NU243, a direct derivative of 130b that contains a stable mini-Tn10 insertion in the Legionella pilD gene, were described previously (19, 31). A population of 130b randomly mutagenized with mini-Tn10 was also reported before (41). Legionellae were routinely cultured on buffered charcoal yeast extract (BCYE) agar for 3 days at 37°C (17). In order to determine the levels of secreted enzymatic activities and compare the growth rates between strains, bacteria were cultured in buffered yeast extract (BYE) broth (5). The extent of bacterial growth was assessed by measuring the optical density of the cultures at 660 nm (OD₆₆₀) (31). To ultimately screen for mutants deficient in acid phosphatase activity, colonies of randomly mutagenized 130b bacteria were taken from BCYE plates and inoculated into 100 µl of BYE contained in wells of 96-well plates. After overnight incubation at 37°C with shaking, 10-µl aliquots from each culture were assayed for acid phosphatase activity (see below). Escherichia coli DH5 was used as the host for recombinant plasmids (6). E. coli was grown in Luria-Bertani broth or agar, with ampicillin (100 µg/ml) added when needed.

Preparation of supernatants and cell lysates. To test for secreted or cell-associated enzymes, supernatants and lysates from L. pneumophila cultures were prepared as before (5). Briefly, supernatants were obtained by centrifugation followed by filtration (0.22 µm pore size) and in some cases were concentrated 100-fold through a Millipore YM10 ultrafiltration cell. Lysates were prepared by treatment with Triton X-100 and lysozyme.

Phosphatase and other enzymatic assays. Phosphatase activity was measured as the ability of the sample to release p-nitrophenol (pNP) from p-nitrophenyl phosphate (pNPP) (Sigma Chemical, St. Louis, Mo.) (5). Bacterial supernatant or cell lysates were incubated with 7.6 mM pNPP in 50 mM citric acid buffer (pH 5) or 0.2 M acetate buffer (pH 5.5) at 37°C. Since pNP is colorless at acid pH, concentrated NaOH was added to the reactions after the appropriate incubation at 37°C, and the production of pNP was monitored at 410 nm. In some experiments, sodium tartrate and sodium molybdate, compounds that usually inhibit acid phosphatases, were assayed throughout a concentration range from 0.001 to 10 mM (15, 29, 50, 54). To measure alkaline phosphatase activity, the pNPP hydrolysis reactions were performed in 100 mM Tris (pH 10) (5). One unit of acid or alkaline phosphatase activity was defined as that which releases 1 nmol of pNP in 1 min. Acid phosphatase activity was also determined in samples electrophoresed through polyacrylamide gels (47, 52). Proteins in 100-fold-concentrated supernatants were separated in a discontinuous system with a 10% polyacrylamide separating gel under nondenaturing conditions. After the electrophoresis, the gel was equilibrated in 0.2 M acetate buffer (pH 5.5) and incubated with 7.6 mM pNPP in the same buffer at room temperature. To then detect the release of pNP, the solution was made alkaline by the addition of concentrated NaOH. The reactive band was isolated from the gel by electroelution under nondenaturing conditions following the manufacturer's recommendations (Bio-Rad Laboratories, Hercules, Calif.), and the proteins contained within were examined for pNPP hydrolysis and for size by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (27).

PCR, sequencing, and clone analysis. Genomic DNA from L. pneumophila was extracted as previously described (42). Based upon data from the L. pneumophila Philadelphia I genome project (http://genome3.cpmc.columbia.edu/~legion/), three DNA primers were designed for the amplification of an acid phosphatase gene: 1, 5'-GCCATCTTCCAAGGTATAGC, corresponding to a site 1,020 bp upstream of the gene; 2, 5'-ACCAACGGTGGCAAGATACG, from a site 74 bp upstream; and 3, 5'-ATTCCGAGCACGACCACAAC, from a site 201 bp downstream of the gene. To determine the approximate position of the mini-Tn10 in several acid phosphatase mutants, a standard PCR was performed using primer 3 and a primer designed from the end of the transposon (5'-CCTTAACTTAATGATTTTTAC) (32). PCR products obtained with wild-type DNA and primers 1 plus 3 and 2 plus 3 were used for sequencing the 130b acid phosphatase gene. Sequencing reactions were performed using the BigDye terminator cycle sequencing mix from PE Applied Biosystems (Foster City, Calif.). Primer synthesis and automated sequence analysis on an ABI Prism 373 DNA sequencer (Applied Biosystems) were performed at the Biotech Facility at Northwestern University Medical School, Chicago, Ill. Sequence database searches were performed using programs based on the BLAST algorithm (4). The nucleotide sequence was analyzed for promoter prediction (44). The predicted protein was analyzed with the SignalP program (36) and Psort (35) for a signal sequence and for protein motifs with the PROSITE database (8). The PCR fragments that were used in sequencing experiments were also cloned into the pGem-T Easy vector (Promega, Madison, Wis.). The 2.3-kb fragment derived using primers 1 and 3 generated plasmid pVA13, whereas the 1.3-kb fragment obtained with primers 2 and 3 generated pVA12.

Intracellular infection of U937 cells and Hartmannella amoebae. U937, a human cell line that differentiates into macrophage-like cells after treatment with phorbol esters, served as a host for in vitro infection by L. pneumophila (13). The cell line was infected as previously described (31), with some modifications. To quantitate intracellular growth, monolayers containing 10⁶ macrophages were inoculated with approximately 10⁵ CFU, incubated for 0, 6, 18, 24, 48, or 72 h, and then lysed. Serial dilutions of the lysates were plated on BCYE agar, supplemented with kanamycin for the mutants, to determine the number of bacteria per monolayer. To establish the cytopathic effect of L. pneumophila on U937 cells, the viability of infected monolayers was tested by their ability to reduce alamar blue, as recommended by the manufacturer (Biosource International, Vacaville, Calif.). Briefly, at 0, 24, 48, and 72 h, the wells were thoroughly washed to eliminate the extracellular bacteria. Medium (100 µl) with 10 µl of alamar blue was added to each well and incubated at 37°C for 4 h. After this time, the fluorescence (excitation, 540 nm; emission, 584 nm) was read in a Spectra Max Gemini fluorescence reader (Molecular Devices, Sunnyvale, Calif.). To examine the ability of legionellae to grow within a protozoan host, Hartmannella vermiformis was infected as previously indicated (14, 31). Approximately 10⁵ CFU were added to wells containing 10⁵ amoebae, and at 0, 24, 48, or 72 h postinoculation, the numbers of bacteria within the coculture were determined by serial dilution.

Statistical analysis. To study the statistical significance of the differences observed between samples, Student's t test was used. The P values are given when appropriate.

Nucleotide accession number. The L. pneumophila map locus sequence has been deposited in the GenBank database at the National Center for Biotechnology Information under accession number AF299349.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Isolation of L. pneumophila acid phosphatase mutants. With the purpose of determining the role of the acid phosphatase(s) in L. pneumophila intracellular infection, we sought to isolate and characterize mutants specifically deficient in this enzymatic activity. Toward that end, 1,767 colonies from a randomly mutagenized population of L. pneumophila 130b were inoculated into 100 µl of BYE in microtiter plates. After overnight growth at 37°C, an aliquot from each culture was tested for phosphatase activity with pNPP at pH 5. Five mutants with reduced activity, which was visually apparent, were selected and grown in standard culture tubes for retesting. The three mutants that appeared to produce a reduced level of acid phosphatase upon reexamination were designated NU254, NU255, and NU256. These mutants replicated in BYE broth with a growth profile similar to that of wild-type 130b (data not shown), indicating that the reduction in acid phosphatase activity was not due to a general growth defect.

View larger version (32K): [in this window] [in a new window]   FIG. 1.   Acid phosphatase activity produced by L. pneumophila strains. Late-log-phase (OD660 = 1.8 to 1.9) supernatants (A) and cell lysates (B) of wild-type strain 130b (WT), pilD mutant NU243, NU254, NU255, and NU256 were examined for their ability to release pNP from pNPP at pH 5. Bars represent the means ± standard deviation of the activity found in three cultures and are representative of the results seen in three independent experiments.

View larger version (52K): [in this window] [in a new window]   FIG. 2.   Lipase/esterase and pNPPC hydrolase activity secreted by L. pneumophila strains. Late-log-phase (OD660 = 1.8 to 1.9) supernatants of wild-type 130b (WT), NU254, NU255, and NU256 were tested for their ability to release pNP from p-nitrophenyl palmitate (A) and pNPPC (B). Bars represent the means ± standard deviation of the activity found in three cultures and are representative of the results obtained in three independent experiments. Similar to the results in panel A, the three mutants were not defective in the hydrolysis of p-nitrophenyl caprylate (data not shown)

Nucleotide sequence of the major L. pneumophila acid phosphatase (map) gene. To establish the genetic basis of the defect in NU254, NU255, and NU256, we sought to determine the sequences into which the transposon had inserted in each mutant. At the time we were starting this analysis using inverse PCR, as we have done in the past for other mutants (32), the sequence of an open reading frame (ORF) with homology to acid phosphatases became available in the developing L. pneumophila Philadelphia I genome database. To determine if this was the gene mutated in our 130b acid phosphatase mutants, PCR experiments were performed using a primer corresponding to sequences just 3' of the ORF and a primer for the mini-Tn10. The PCR fragments amplified from DNA of the three mutants indicated that this was the gene interrupted; i.e., we amplified fragments of 430 bp for NU254, 260 bp for NU255, and 960 bp for NU256. The exact position of the transposon in NU255 was confirmed by sequencing of the PCR fragment (see below).

View larger version (50K): [in this window] [in a new window]   FIG. 3.   Nucleotide sequence of the L. pneumophila 130b map locus. The deduced amino acid sequences of the various ORFs and the termination codons (#) are indicated. For the predicted Map protein, the putative signal peptide is indicated in bold, and conserved protein domains are shadowed. Within the conserved domains, asterisks indicate conserved amino acids, and bold asterisks indicate amino acids that are essential for activity in the E. coli periplasmic acid phosphatase. Putative promoter sequences are indicated by the 10 and 35 designations, and the Shine-Dalgarno sequence is indicated (SD). The mini-Tn10 insertions are indicated by an arrow alongside the name of the corresponding mutant. The positions of the insertions are approximate for NU254 and NU256. Primers 2 and 3 (see Materials and Methods) are shown in italics at their binding regions.

Identification of a minor acid phosphatase activity produced by L. pneumophila. Although an acid phosphatase gene was interrupted in NU254, NU255, and NU256, the supernatants from the mutants always showed a low level of activity in the pNPP assay (Fig. 1A) which could be eliminated by 10 min of incubation at 100°C. To determine whether this activity was simply cell-associated alkaline phosphatase that was released by spontaneous lysis of some cells in the culture (5, 25), supernatants and cell lysates from the wild type and map mutants were assayed for hydrolysis of pNPP at both pH 5 and 10 (Fig. 4). The supernatants showed a small amount of phosphatase activity at pH 10. However, the cell lysates, which were very rich in alkaline phosphatase, showed a minimal pNPP reaction at pH 5, indicating that alkaline phosphatase cannot be responsible for the activity in the supernatants from the mutants. Since two of the mutants had transposon insertions in domains of the protein that are likely essential for activity (Fig. 3), and all of them have the same level of pNPP hydrolytic activity (Fig. 1A and 4), it is unlikely that the residual activity is due to low levels of truncated Map. Consequently, we suspected that the phosphatase activity remaining in the supernatants of the map mutants was due to another (minor) acid phosphatase activity, since bacteria can have multiple acid phosphatases (47, 50).

View larger version (24K): [in this window] [in a new window]   FIG. 4.   Acid and alkaline phosphatase activity in L. pneumophila cultures. Supernatants and cell lysates from the wild type (bar 1) and map mutants NU254 (bar 2), NU255 (bar 3), and NU256 (bar 4) were examined for their ability to hydrolyze pNPP at pH 5 and 10. Bars represent the means ± standard deviation of three cultures. The only significant difference between the wild type and mutants was observed with supernatants tested at pH 5 (P < 0.01).

View larger version (19K): [in this window] [in a new window]   FIG. 5.   L. pneumophila map mutants lack a tartrate-sensitive acid phosphatase. Supernatants from wild-type 130b (), the pilD mutant NU243 (), and the mutants NU254 () and NU255 () were examined for their ability to hydrolyze pNPP in the presence of different concentrations of molybdate or tartrate. The isolated symbols represent the level of activity without inhibitor. The results are the means ± standard deviation of the activity found in three cultures. Similar results were obtained in two independent experiments using concentrated supernatants (data not shown).

View larger version (17K): [in this window] [in a new window]   FIG. 6.   Effect of molybdate and tartrate on the cloned L. pneumophila Map activity. Whole cultures of E. coli DH5(pVA12) () and DH5(pVA13) () were tested for acid phosphatase activity in acetate buffer (pH 5.5) in the presence of different concentrations of sodium molybdate (top panel) or sodium tartrate (bottom panel). The results are expressed as the percentage of refractive activity and are the means ± standard deviation of the activity found in three cultures and representative of two independent experiments. The error bars are too small to be seen.

Intracellular infection by L. pneumophila map mutants. To determine the role of the major acid phosphatase in intracellular infection, we tested the map mutants for their relative ability to infect macrophages and protozoa. It is well documented that the intracellular growth of L. pneumophila results in the destruction of the macrophage host (1, 12, 40). Therefore, the acid phosphatase mutants, together with the wild type and the pilD mutant, were first inoculated into U937 cells and examined for cytopathic effects at different times after infection (Fig. 7). The macrophages infected with the wild type showed a substantial decrease in viability after 48 and 72 h of infection, only 19 and 9% viable macrophages, respectively (P < 0.001). Contrary to these results, the macrophages infected with the pilD mutant showed little, if any, reduction in viability after 72 h, as has been seen previously (5, 31). The map mutants showed a pattern of cytopathicity that was identical to that of the wild type (Fig. 7). To determine if this cytopathic effect correlated with a normal capacity to replicate inside macrophages, the mutants were inoculated into U937 monolayers, and the number of bacteria was recorded after various incubation intervals (Fig. 8A). Wild-type 130b showed the typical intracellular growth pattern, increasing 10,000- to 100,000-fold by 48 h after infection. Interestingly, the same results were obtained with the acid phosphatase mutants, indicating that these strains are not defective for intracellular replication in U937 cells. Moreover, the mutants did not seem to have a defect in entry, since the numbers recovered after the uptake period were the same as for the wild type (see t = 0 in Fig. 8A). To determine the role of the major acid phosphatase in amoeba infection, H. vermiformis cultures were inoculated with the wild type and map mutants, and the number of bacteria was recorded at different times. As in the macrophage infection, the same numbers of bacteria were recovered from cocultures with the wild type or with the mutants (Fig. 8B). Together, these results indicate that the major acid phosphatase is not essential for macrophage or amoeba infection by L. pneumophila.

View larger version (17K): [in this window] [in a new window]   FIG. 7.   Cytopathic effect of L. pneumophila on U937 cells. At different times after inoculation, the viability of the macrophage monolayer that had been infected at a multiplicity of infection of 0.1 with 130b (), pilD mutant NU243 (), NU254 (), or NU255 () was measured by fluorescence associated with alamar blue reduction. As a control, an uninfected monolayer () was processed in parallel. Results represent the means ± standard deviation of triplicate wells and are representative of two independent experiments.

View larger version (17K): [in this window] [in a new window]   FIG. 8.   Intracellular infection by wild-type and mutant L. pneumophila. U937 cells (A) and H. vermiformis amoebae (B) were infected at multiplicities of infection of 0.1 and 1, respectively, with wild-type 130b (), NU254 (), or NU255 (). The number of bacteria in each well was quantitated at 0, 24, 48, and 72 h by plating aliquots on BCYE agar. Results represent the means ± standard deviation of triplicate wells and are representative of two independent experiments. In a third U937 experiment, the numbers of bacteria were also determined at 6 and 18 h (inset).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

An acid phosphatase activity was first described in 1981 for the Legionella genus (34). That activity was detected in all the strains tested, i.e., representatives of L. pneumophila serogroups 1 to 6 and Legionella species L. bozemanii, L. dumoffii, L. gormanii, L. longbeachae, and L. micdadei (34, 37, 59). Subsequently, two cell-associated acid phosphatases, ACP₁ and ACP₂, were purified from L. micdadei (50). Recently, we demonstrated that the acid phosphatase activity of L. pneumophila is actually secreted in a pilD-dependent fashion (5). Our current data indicate that the secreted acid phosphatase activity of L. pneumophila is the result of at least two different enzymes, which can be discriminated by their different sensitivity to tartrate. Mutant and sequence analyses indicated that the majority of activity is due to a tartrate-sensitive histidine acid phosphatase, which we have named Map. The Map protein is notable among bacterial acid phosphatases in two aspects. First, it is an extracellularly secreted enzyme, while the rest of the bacterial acid phosphatases characterized to date are periplasmic or membrane-bound proteins (47). Second, Map aligned with significant homology only with eukaryotic histidine acid phosphatases that are also secreted. Although several bacterial acid phosphatases have been sequenced (47), they did not show homology with L. pneumophila Map as determined by the NCBI BLAST programs. Thus, our data represent the first evidence for a bacterial extracellularly secreted histidine acid phosphatase.

L. pneumophila mutants containing insertions in map did not show a defect in infection of U937 cells or amoebae. Therefore, Map either has no role in L. pneumophila intracellular replication or has a dispensable one. As we have demonstrated, L. pneumophila has at least one other acid phosphatase, and this, or other enzymes, may compensate for the lack of Map. Also, even though Map is not essential for U937 cell infection, it may play a role in Legionella survival within other host cells, such as activated macrophages or neutrophils. Although there is no bacterial precedent for Map, two secretory histidine acid phosphatases, SAcP-1 and SAcP-2, have been identified in the macrophage parasite Leishmania donovani (53). It is believed that the parasitic phosphatases do have a role in infection, because they are produced during human infection (18). Since Map and the SAcP enzymes showed significant homology, they may have a common target or function during intracellular infection. On the other hand, it is entirely possible that Map and its parasite analog promote extracellular survival and spread in vivo.

Although the loss of Map secretion was not the reason for the intracellular infectivity defect of the pilD mutant, it is possible that the pilD-dependent tartrate-resistant acid phosphatase is important for U937 cell or amoeba infection. Indeed, in addition to L. micdadei ACP₂ noted above, tartrate-resistant acid phosphatases from Francisella tulariensis, Coxiella burnetii, and Leishmania donovani have been implicated in the resistance of these intracellular pathogens to the oxidative burst of phagocytes; the purified enzymes inhibit superoxide production by fMet-Leu-Phe-activated cells (7, 30, 45, 46, 50). In the case of L. micdadei and Leishmania donovani, the enzymes interfere in phosphoinositide metabolism by degrading phosphatidylinositol diphosphate and inositol trisphosphate (16, 51). Although the tartrate-resistant activity is a minor one in L. pneumophila BYE culture supernatants, it may be a major secreted factor when the bacteria are exposed to host cells or tissue. Thus, the tartrate-resistant acid phosphatase activity that we have uncovered, as well as other pilD-dependent factors, may be quite relevant for the pathogenesis of L. pneumophila.