- Email: [email protected]
Naip5 Affects Host Susceptibility to the Intracellular Pathogen Legionella pneumophila
Current Biology, Vol. 13, 27–36, January 8, 2003, 2003 Elsevier Science Ltd. All rights reserved.
PII S0960-9822(02)01359-3
Naip5 Affects Host Susceptibility to the Intracellular Pathogen Legionella pneumophila Edward K. Wright, Jr.,1 Sheryl A. Goodart,1 Joseph D. Growney,2 Vey Hadinoto,2 Matthew G. Endrizzi,2 E. Michelle Long,1 Keyvan Sadigh,2 Andrew L. Abney,2 Isaac Bernstein-Hanley,2 and William F. Dietrich1,2,* 1 Howard Hughes Medical Institute 2 Department of Genetics Harvard Medical School Boston, Massachusetts 02115
Summary Background: Legionella pneumophila is a gram-negative bacterial pathogen that is the cause of Legionnaires’ Disease. Legionella produces disease because it can replicate inside a specialized compartment of host macrophages. Macrophages isolated from various inbred mice exhibit large differences in permissiveness for intracellular replication of Legionella. A locus affecting this host-resistance phenotype, Lgn1, has been mapped to chromosome 13, but the responsible gene has not been identified. Results: Here, we report that Naip5 (also known as Birc1e) influences susceptibility to Legionella. Naip5 encodes a protein that is homologous to plant innate immunity (so-called “resistance”) proteins and has been implicated in signaling pathways related to apoptosis regulation. Detailed recombination mapping and analysis of expression implicates Naip5 in the Legionella permissiveness differences among mouse strains. A bacterial artificial chromosome (BAC) transgenic line expressing a nonpermissive allele of Naip5 exhibits a reduction in macrophage Legionella permissiveness. In addition, morpholino-based antisense inhibition of Naip5 causes an increase in the Legionella permissiveness of macrophages. Conclusions: We conclude that polymorphisms in Naip5 are involved in the permissiveness differences of mouse macrophages for intracellular Legionella replication. We speculate that Naip5 is a functional mammalian homolog of plant “resistance” proteins that monitor for, and initiate host response to, the presence of secreted bacterial virulence proteins. Introduction Legionella pneumophila is a gram-negative bacterial pathogen that is the cause of a severe pneumonia called Legionnaires’ Disease [1]. Some studies suggest that Legionella infection may account for as many as 5%– 10% of community-acquired pneumonia cases [2]. Contaminated fresh water supplies are typically the sources of human Legionella infections, with aspiration or inhalation of infectious aerosols believed to be the usual inoculation route [3, 4]. *Correspondence: [email protected]
Legionella is capable of intracellular growth inside of eukaryotic host cells, including mammalian macrophages, and the ability of Legionella to replicate inside of host macrophages is a crucial aspect of its virulence in animal models [5, 6]. In order to replicate efficiently in macrophages, Legionella must actively block or alter normal cellular endocytic physiology, allowing it to gain access to an unusual rough ER-bounded phagosome [7, 8]. The intracellular trafficking of Legionella depends on the function of the dot/icm genes [9]. The Dot/Icm proteins comprise a specialized secretion system that serves to transfer bacterial proteins into the cytosol of target host cells [10–12]. Several studies suggest that a functional Dot/Icm apparatus is needed within minutes of bacterial uptake to mediate intracellular trafficking decisions that determine the outcome of the infection [10–12]. Recently, it has been shown that Legionella, in a Dot/ Icm-dependent manner, secretes a guanine nucleotide exchange factor (RalF) into the host cell. RalF apparently alters the activity of host ADP ribosylation factor 1 (ARF1), thus allowing the bacterium to recruit ERderived membranes as part of its strategy to remodel its vacuole [12]. Since RalF is not essential for virulence of Legionella in mammalian cells, it is likely that there are other Dot/Icm-secreted effector proteins that are important for growth in macrophages. The future identification of these additional Dot/Icm-secreted proteins and the host pathways that they affect will be crucial for a full understanding of the pathogenesis of Legionella. Mammalian genetics is a potentially useful approach that can be used to understand the pathogenesis of Legionnaires’ Disease. It has been noted that in vitrocultured macrophages from two different inbred mouse strains (A/J and C57BL/6J) differ in their permissiveness for intracellular Legionella replication [13]. This phenotypic difference segregates in a Mendelian fashion [14, 15], and the gene responsible for the macrophage permissiveness difference maps to a locus called Lgn1 on mouse chromosome 13 [16, 17]. Detailed characterization of the mouse Lgn1 interval has demonstrated that it consists of a series of 80- to 100-kb direct repeats, each containing a paralogous member of the Neuronal Apoptosis Inhibitory Protein (Naip, also known as the Birc1) gene family [18, 19]. More recently, we have shown that the Lgn1 mutation must lie within an interval that contains only the first 12 exons of Naip2 (Birc1b), the entirety of Naip5 (Birc1e), and a fragmentary ⌬Naip (Birc1c-ps1) pseudogene (Figure 1; [20–22]). Most of the mouse Naip paralogs are functional transcription units that encode proteins having 85% identity to each other [23]. Much research has focused on the potential role of Naip proteins in apoptotic signaling [24]. However, it has been shown that Lgn1 (that is, either Naip2 or Naip5) alleles can influence very early events of intracellular trafficking of Legionella, suggesting that the Lgn1 gene product likely has some function(s) that
Current Biology 28
Figure 1. Gene Content of Lgn1 Critical Interval A depiction of the map of the Lgn1 interval, along with the positions of relevant markers and genomic clones, is shown. The right of the figure points away from the centromere of chromosome 13. The size of the Lgn1 interval, in kb, is indicated by the scale at the top of the figure [20, 22]. The bold horizontal line depicts the extent of the C57BL/6J-derived BAC clone 235. The map position of this clone was previously reported [19]. The three lines with arrowheads depict the position and orientation of genes (Naip2 [Birc1b] and Naip5 [Birc1e]) and gene fragments (⌬Naip) in the Lgn1 interval [19, 23]. The exon positions and numbers for Naip2 (Birc1b) and Naip5 (Birc1e) are indicated by small, labeled lines drawn above a thin horizontal line that depicts the chromosome. Below the chromosome are the names and approximate positions of relevant microsatellite markers. The pattern-filled boxes shown below the chromosome indicate the extent of the minimal genetic intervals for Lgn1. The box labeled “Lgn1 Prior” shows the minimal Lgn1 interval that was previously determined [21]. “Lgn1 Current” depicts the minimal Lgn1 interval after utilizing Naip2 (Birc1b) missense polymorphisms to refine the location of the centromere-proximal recombination events (Figure 2C). The transgene content of BAC 235 transgenic lines are indicated at the bottom of the figure, with “⫹” indicating that the named strain contains the C57BL/6J allele of that marker in its genome; “⫺” indicates that the strain does not have the C57BL/6J allele of the marker.
are unrelated to its putative role in apoptosis regulation [25]. It has been suggested that the Naip genes may actually be mammalian homologs of plant genes that are known to mediate innate immunity to bacterial pathogens (the so-called “resistance” or “R” genes) [26–29]. Here, we report data demonstrating that polymorphisms in Naip5 cause alterations in the Lgn1 phenotype. This conclusion is based primarily on our genetic and molecular analyses of Naip2 and Naip5, on our alteration of the Lgn1 phenotype with a bacterial artificial chromosome (BAC) transgene expressing only Naip5, and on our ability to increase macrophage permissiveness with Naip5 antisense reagents. Our identification of a gene affecting the Lgn1 phenotype provides a framework for future studies aimed at identifying the specific Lgn1 mutation(s) and at understanding the molecular role that is played by Naip5 in host resistance to Legionella pathogenesis. Results The FvB/N Mouse Is Permissive Due to an Allele of Lgn1 As a starting point for distinguishing between the Lgn1 candidate genes (Figure 1), we undertook the phenotypic characterization of additional inbred mouse strains. As previously reported, macrophages from A/J
mice support vigorous intracellular Legionella growth, while those from C57BL/6J mice rarely yield more than 2 ⫻ 104 colony-forming units (cfu) at 6 days postinfection (Figure 2A). In addition, we found that the FvB/NJ (FvB) mouse strain is clearly permissive (see the Experimental Procedures). We believe that FvB is permissive because of an allele of Lgn1. The permissiveness phenotype of FvB is recessive (Figure S1A) and appears to be in the same complementation group as the Lgn1 allele found in A/J, since macrophages from A/J ⫻ FvB/N F1 hybrids are permissive (Figure 2A). In addition, we found that ten (A/J ⫻ FvB/NJ) ⫻ A/J backcross animals were all permissive, making it unlikely that an FvB allele unlinked to Lgn1 accounts for the phenotype of A/J ⫻ FvB/N F1 hybrids. A Modifier of Permissiveness from FvB Maps to Lgn1 The 144-hr bacterial yields from macrophages of A ⫻ FvB F1 animals are reproducibly at least 5-fold lower than those seen in A/J macrophages (Figure 2A). Interestingly, six of the previously described (A/J ⫻ FvB/ NJ) ⫻ A/J backcross animals exhibited an “A ⫻ FvB F1like” permissiveness phenotype, while four others were “A/J-like” (Figure 2B). The segregation of this qualitative difference in permissiveness correlated perfectly with the inheritance of a marker in Naip5, D13Die40, establishing linkage of this trait to the Lgn1 locus (LOD score ⫽
Naip5 Affects Resistance to Legionella pneumophila 29
Figure 2. Characterization and Mapping of Legionella Permissiveness of Mouse Macrophages (A) Bone marrow-derived macrophages from each of the indicated strains were infected with Legionella inocula prepared as described [45, 46]. A ⫻ FvB F1 is an F1 hybrid between A/J and FvB/NJ. Shown is the average log10 of the number of bacterial colony-forming units for each of the samples at different times postinfection. The error bars indicate the standard deviation of the average. A/J, n ⫽ 11; C57BL/6J, n ⫽ 12; FvB/NJ, n ⫽ 6; A ⫻ FvB F1, n ⫽ 12. (B) Data from a 10-animal (A/J ⫻ FvB) ⫻ A/J cross. The genotype and phenotype of A ⫻ FvB F1 hybrid (F1) and A/J animals are shown by the two leftmost columns. The three rightmost columns indicate the genotypes and phenotypes of all the progeny, including the single recombinant animal. In our experience, the behavior of the A ⫻ FvB F1 macrophages is very consistent, yielding 6.58 ⫾ 0.13 log10 cfu at 144 hr postinfection. Accordingly, the phenotype of the animals was declared to be A ⫻ FvB F1-like if the total bacterial yield at 144 hr was within three standard deviations of this average. Importantly, animal #2880 has a clear “A ⫻ FvB F1-like” permissiveness phenotype (144-hr yield of 6.6 log10 cfu, which was 10-fold less than the yield of the contemporaneous A/J control), yet it is an A/J homozygote for markers within Naip2 (see the Supplementary Experimental Procedures). (C) Data from a 466-animal (A/J ⫻ C57BL/6J) ⫻ A/J cross [18]. The genotype and phenotype of A ⫻ B F1 hybrid (F1) and A/J animals are shown by the two leftmost columns. The four rightmost columns indicate the genotypes and phenotypes of the four informative backcross animals. The assignment of phenotypes was documented in a previous publication [18].
3). However, this variation in susceptibility cannot be accounted for by missense polymorphisms in Naip2, since a single recombinant animal (#2880) excludes the entire Naip2 ORF (which begins in exon 3; see Figure 1). Furthermore, we do not observe differences in Naip2 mRNA levels between A/J, FvB, and C57BL/6J (Figure 3A), eliminating the possibility that Naip2 regulatory mutations result in this modifying effect. The Lgn1 Locus of Other Inbred Strains We also characterized the phenotypes of several other inbred strains. The P/J mouse exhibits a level of nonpermissiveness similar to C57BL/6J (Figure S1B). In contrast, C3H/HeJ (C3H), BALB/cJ (BALB), and 129S1 (129) macrophages behave similarly to those of FvB (Figures S1A and S1C). We mapped the Legionella permissiveness phenotype of the 129 strain to the Lgn1 locus (LOD score ⫽ 6.7) by infecting macrophages isolated from a 29-animal (129S1 ⫻ C57BL/6J) F2 cross and by analyzing the yield of bacterial colony-forming units at 144 hr as a quantitative trait (see the Supplementary Experimental Procedures available with this article online).
We determined the genomic and cDNA sequences of the Naip2 and Naip5 genes from some of the permissive and nonpermissive strains described above (Table 1). These sequences exhibit a high single nucleotide polymorphism (SNP) rate that is in the vicinity of 0.3%–0.5% [30]. Overall, no nonsense or rearrangement mutations that would obviously abolish Naip RNA or protein expression were observed, but comparison of Naip2 and Naip5 cDNA sequences from all the strains allowed us to identify candidate Lgn1 missense mutations (Tables 2 and 3). Interestingly, the near identity of the P/J and C57BL/6J sequences suggests that these nonpermissive mice share an Lgn1 haplotype that is derived from a more recent common ancestor than is found among the Legionella permissive strains (Tables 2 and 3). Refined Genetic Mapping of Lgn1 In addition to potentially contributing to the interstrain variation in Legionella susceptibility, the Naip gene polymorphisms can be used as genetic markers to refine the localization of the Lgn1 interval. Therefore, we used our previously characterized Lgn1 mapping cross (an [A/J ⫻ C57BL/6J] ⫻ A/J backcross) to position the Naip2
Current Biology 30
Figure 3. Expression of Naip (Birc1) Genes in Inbred Strain Macrophages (A) Naip2 (Birc1b) mRNA levels. Shown is a bone marrow macrophage total RNA Northern blot (with 15 g RNA per lane) hybridized with a probe that specifically hybridizes to Naip2 (upper panel) and then subsequently hybridized to a Gapdh probe as a loading control (lower panel). Each inbred strain is represented by a pair of lanes, with the leftmost of the pair containing RNA from an uninfected macrophage sample (⫺) and the rightmost lane containing RNA from macrophages that have been infected for 6.5 hr with Legionella (⫹) at an multiplicity of infection (m.o.i.) of 1 (see the Experimental Procedures for details). While there seems to be a slight upregulation of Naip2, there are no substantial differences in the levels or induction of Naip2 mRNA between A/J, C57BL/6J, and FvB. We have observed similar data from several other Naip2 Northern experiments with samples obtained from A/J and C57BL/6J macrophages and infected for 1.5, 5.5, 9.5, and 13.5 hr at m.o.i.s of 1 and 5 (data not shown). (B) Specificity of Naip antisera. Shown is a Western blot of protein isolated from 293T cells transfected with the indicated Naip gene expression constructs (see the Experimental Procedures). The labels on the right indicate the antiserum used to probe the blot; the labels on the left indicate the approximate position and mass (in kD) of protein size standards. M, mock transfected; N5, Naip5 construct; N2, Naip2 construct; N6, Naip6 construct. Although it is not shown, neither antiserum crossreacts with Naip1. (C) Expression of Naip proteins in A/J and C57BL/6J macrophages. Shown are a Western blot of total cellular protein probed for Naip2 and a Western blot of Naip5 antibody-immunoprecipitated protein probed for Naip5 (see the Supplementary Experimental Procedures). A/J, protein from A/J primary macrophages; B6, protein from C57BL/6J primary macrophages; N2, 293T cells transfected with Naip2 expression construct; N5, 293T cells transfected with Naip5 expression construct. (D) Expression of the BAC transgene in the Full and N5 transgenic lines. C57BL/6J allele-specific RT-PCR assays were performed on the following cDNA samples obtained from bone marrow macrophage total RNA (see the Experimental Procedures): Full, Full transgene carrier (on A ⫻ FvB F1 background); N5, N5 transgene carrier (on A ⫻ FvB F1 background); NT, nontransgenic A ⫻ FvB F1; B6, C57BL/6J; H2O, water; ⫹, B10.A (identical in sequence to C57BL/6J) Naip2 (Birc1b) plasmid or Naip5 (Birc1e) plasmid, as appropriate (AF135490 and AF135492; [23]).
and Naip5 missense polymorphisms relative to Lgn1. As can be seen in Figure 2C, the polymorphism in codon 862 of Naip2 recombines away from the Lgn1 phenotype (in two separate animals, #661 and #549), while the more 5⬘ polymorphisms do not. This result effectively excludes all but seven Naip2 polymorphisms from the Lgn1 interval (Figures 1 and 2C; Table 2).
Naip2 Is Not Differentially Expressed between A/J and C57BL/6J To investigate if Naip2 or Naip5 protein levels could possibly play a role in determining the Lgn1 phenotype, we developed two different antisera that recognize Naip2 and Naip5 (Figure 3B). We used these antisera in
Western blot and immunoprecipitation experiments to evaluate the levels of Naip2 and Naip5 proteins in macrophages from A/J and C57BL/6J mice. While the levels of Naip2 protein in these two strains are equivalent, the levels of protein recognized by the Naip5 antibody are substantially higher in C57BL/6J (Figure 3C). Since the Naip5 antibody crossreacts with Naip6 (Figure 3B), the signal detected by this antibody in the macrophage lysates is likely comprised of a mixture of Naip5 and Naip6. However, the mRNA for Naip5 is expressed at higher levels than Naip6 [23], suggesting that the majority of what we detect is Naip5. The magnitude of this differential expression appears similar to the 4-fold differences in Naip expression reported previously [31]; this finding suggests that the differences that these authors ob-
Table 1. Genomic and cDNA Sequences of Lgn1 Candidates Strain
Genomic
Naip2 (Birc1b) cDNA
Naip5 (Birc1e) cDNA
References
C57BL/6J P/J FvB/N C3H/HeJ BALB/cJ A/J 129S1
AF367967, AF367968, AF367969 None None None None AF367966 AF131205
AY147001 AY147005 AY147004 AY147003 AY147002 AF381770 AY147000
AY146995 AY146999 AY146998 AY146997 AY146996 AF381771 AY146994
This work This work This work This work This work This work [20] and this work
The GenBank accession numbers of the indicated sequences are shown.
Naip5 Affects Resistance to Legionella pneumophila 31
Table 2. Comprehensive List of Amino Acid Polymorphisms in Naip2 (Birc1b) AA#
Exon
C57BL/6J
P/J
Nonpermissive 377 403 478 540 556 558 561 862 1079 1080 1089 1115 1122 1136 1157 1160 1167 1186 1192 1271 1272
10 11 11 11 11 11 11 11 11 11 11 11 11 11 12 12 12 12 12 14 14
D L L Y G E K K S D R K T D S D G S V F G
FvB/N
C3H/HeJ
BALB/cJ
A/J
129S1
D F L N G E K N S D R K T D S N G Y V F A
D F L N G E K N S D R K T D S N G S V F A
G F L N D G N N F N C E E D G N R S V F G
G F I Y G E K N F N C E E E G D R S V C G
Permissive D L L N G E K K S D R K T D S D G S V F A
D F L Y G E N N S D R K T D S N G S L F A
Polymorphisms and the exonic location of the indicated amino acids (AA#) are shown. The amino acid polymorphisms that distinguish specific strains from all other strains having the opposite phenotype are underlined. The total length of the Naip2 (Birc1b) protein is 1447 amino acids [23].
served were due to variations in Naip5 (and possibly Naip6) expression. A Naip5 Transgene Affects the Lgn1 Phenotype Given the previous data, we sought additional positive evidence implicating Naip5 in affecting the Lgn1 phenotype. To that end, we attempted to complement the Lgn1 phenotype of FvB animals inheriting a C57BL/6J BAC clone transgene containing the Lgn1 candidate genes Naip2 and Naip5. From our injections of this clone (named 235), we obtained two male FvB founder animals that sired progeny inheriting useful portions of the BAC clone. One transgene insertion (named Full) contained polymorphic markers from throughout the BAC clone insert, while the other (named N5) contained polymorphic markers from the entirety of Naip5 and only a portion of Naip2 (Figure 1). These two transgene insertions were stably inherited and were found to express Naip mRNA and protein in macrophages in a manner consistent with the marker content of the transgenes (Figure 3D and data not shown). After backcrossing each transgene twice to FvB and then inbreeding for 4–5 generations (see the Experimental Procedures), we assessed the macrophages of mice from these FvB transgenic lines for evidence of complementation of the Legionella permissiveness phenotype. Surprisingly, the Full transgene has no effect on macrophage permissiveness (Figure 4). In contrast, the growth supported by the N5 transgenic macrophages was always lower than that of contemporaneous FvB control macrophages. While the N5 transgene does not cause the macrophages to become as nonpermissive as C57BL/6J, statistically significant decreases in bacterial yield become apparent by 72 hr postinfection (Figure 4). By 144 hr postinfection, the FvB-N5 transgenic animals
typically yield 5- to 10-fold fewer bacterial colony-forming units than the FvB controls. We also wanted to determine if the BAC transgenes exhibited any effect on the phenotype of more permissive genetic backgrounds. Therefore, we backcrossed the N5 and Full transgenes to the A/J strain and examined the effects of the heterozygous transgenes on the phenotypes of A ⫻ FvB F1 and (A ⫻ FvB) ⫻ A animals. Again, the Full transgene has no significant effect (data not shown). The N5 transgene causes decreases in the 144-hr bacterial yield in the A ⫻ FvB F1 background (a 7-fold average decrease, n ⫽ 3, p ⫽ 0.003) and in the (A ⫻ FvB) ⫻ A-Lgn1A/A background (a 4-fold average decrease, n ⫽ 3, p ⫽ 0.032). Antisense Inhibition of Naip5 Increases Legionella Permissiveness The reduction of permissiveness caused by the N5 transgene is important evidence in favor of Naip5 playing an important role in regulating permissiveness. Nevertheless, we pursued a complementary approach to assess the function of Naip5 with respect to Legionella pathogenesis. We decided to inhibit the function of Naip5 by using morpholinos, which are antisense compounds comprised of an uncharged backbone whose mechanism of action is to inhibit the translation of the target mRNA [32, 33]. As described in the Experimental Procedures, we designed a morpholino that would hybridize to the 5⬘ UTR of Naip5 transcripts from all mouse strains and therefore should inhibit the translation of these transcripts. Since the N5 transgenic line exhibited partial rescue of the permissiveness phenotype, we reasoned that it would be a sensitized background in which we could observe an effect of the morpholino treatment, even if
Current Biology 32
Table 3. Comprehensive List of Amino Acid Polymorphisms in Naip5 (Birc1e) AA#
Exon
C57BL/6J
P/J
Nonpermissive 92 144 234 242 368 472 496 512 514 516 517 521 533 538 647 755 855 952 1021 1092 1116 1123 1137 1140 1228 1241/42 1275/76
3 3 5 5 9 11 11 11 11 11 11 11 11 11 11 11 11 11 11 12 12 12 12 12 14 15 15
R R E S T T Y D G A N A V S A V S S M E N R Q T E V D
FvB/N
C3H/HeJ
BALB/cJ
A/J
129S1
R R K S M A Y D G A N A A I A V T T I D D G R R AQ V D
R R K S M A Y D G A N A A I A V T T I D D G R R AQ V D
R R E S T A N G E A K A A I T M S S M D D G Q T E I N
K S E G T A Y D G D N T A I A V S S M D D G Q T E I N
Permissive R R E S T T Y D G A N A V S A V S S M E N R Q T AQ V D
R R K S M A Y D G A N A A I A V T T I D D G R R AQ V D
Polymorphisms and the exonic location of the indicated amino acids (AA#) are shown. The amino acid polymorphisms that distinguish specific strains from all other strains having the opposite phenotype are underlined. The total length of the Naip5 (Birc1e) protein is 1402 amino acids [23].
the translational inhibition was incomplete. Accordingly, macrophages from the FvB-N5 line that are pretreated with this morpholino have higher bacterial yields than untreated control macrophages at 72 and 120 hr postinfection (Figure 4). This relative increase in permissiveness is somewhat diminished by 144 hr postinfection, suggesting that the inhibitory effect of the morpholinos is transient. In addition, preliminary experiments with C57BL/6J macrophages suggest that treatment with this morpholino can also modestly increase permissiveness in this nonpermissive background (Figure S2). Discussion Positional cloning is a straightforward, if not always easy, process. Genetic linkage provides a discrete region of the genome to be searched. It then becomes a simple matter of searching through regional genes for variations in sequence or expression that correlate with phenotype, or of using tests of complementation to establish a functional effect of a candidate gene. Unfortunately, positional cloning of phenotypes that map to clusters of highly related paralogous genes, such as Lgn1, are much more resistant to these simple strategies. Despite the difficulties presented by the region, we have accumulated evidence favoring a role for Naip5 in modulating permissiveness to Legionella. For example, the entire Naip5 ORF resides in the Lgn1 interval, and there are large differences in Naip5 (and possibly Naip6)
protein levels between A/J and C57BL/6J macrophages. In contrast, our measurements of Naip2 mRNA and protein levels show no differences between permissive and nonpermissive strains. Furthermore, our fine-structure genetic map of Lgn1 eliminates nearly half the open reading frame of Naip2 from the critical interval, leaving only the polymorphism at amino acid 403 as a realistic candidate mutation (Table 2). In addition, we have mapped a modifier of Legionella permissiveness to Lgn1 and have shown that this modifier cannot be caused by variations in Naip2. While it is possible that the modifying allele is in a Naip locus extragenic to Lgn1, the simplest explanation is that there are different permissive alleles of Lgn1 that permit distinct levels of bacterial replication. A casual inspection of Table 3 reveals ample missense polymorphisms in the Naip5 gene that could easily account for variations among permissive strains. The existence of closely linked, highly similar genes presents serious complications in the interpretation of Lgn1 complementation experiments. Nevertheless, we have identified a transgenic line (N5) expressing the C57BL/6J allele of Naip5 that exhibits a reduction in the permissiveness of its macrophages. In addition, we are able to induce an increase in the permissiveness of the N5 line macrophages after treatment with a morpholino designed to inhibit the translation of Naip5 protein. This effect on the transgenic lines is a strong indication that the effect of the N5 transgene on permissiveness is due to the expression of Naip5, and not to disruption of some other gene during transgene integration. While the
Naip5 Affects Resistance to Legionella pneumophila 33
Figure 4. Intramacrophage Growth of Legionella in Transgenic Mice Bone marrow-derived macrophages from transgenic and control strains were infected with Legionella (see the Experimental Procedures). Shown is the log10 of the number of bacterial colony-forming units for each of the samples at different times postinfection. (See the Experimental Procedures). Each symbol represents the average cfu from duplicate infected macrophage samples obtained from a single animal. The experiments shown were performed in small groups over several independent trials. The average of each group of animals is indicated by the short, bold horizontal line. The brackets indicate the p values of comparisons between the controls (on left) and the N5 transgenic or morpholino-treated macrophages. The two-tailed p values were calculated by using the Student’s t test. Control treatment with a nonspecific morpholino sequence caused no effect on the permissiveness of these macrophages (data not shown).
N5 transgene does not cause a complete reversion to nonpermissiveness, preliminary RT-PCR data suggest that the levels of transgenic Naip5 expression in the N5 line are low (data not shown), providing a possible explanation for this observation. The behavior of the Full line, which expresses appreciable levels of both Naip2 and Naip5, but does not exhibit diminished permissiveness, is unexplained. We could not account for the lack of complementation by examining levels of expression of Naip5 or by finding mutations in the Full transgene (data not shown). Since Naip2 and Naip5 are very closely related, it is possible that the two proteins utilize some common cofactor that is present in limiting quantities, and superphysiologic expression of both Naip2 and Naip5 may prevent us from observing the complementing activity of Naip5. We do not know if the Legionella permissiveness phenotype differences that we see are caused by missense polymorphisms in Naip5, alterations in its expression level, or both. However, the candidate mutations that we have identified should be important subjects of a future study using knock-in strategies. In any case, a structure-function study of the candidate mutations that
we have found in the Naip genes will be very important, especially since the molecular functions of the Naip genes are far from clear. There are currently two hypotheses about the molecular function of the Naip proteins. The first Naip family member to be discovered was found in the human genome, and, based on the fact that it encoded proteinprotein-interacting BIR domains (Figure S3), it was hypothesized to be a functional relative of baculoviral IAP genes that can regulate apoptosis [34]. However, not all BIR domain-containing proteins (BIRPs) have a normal physiologic function relevant to apoptosis [35]. Nevertheless, many investigations have uncovered potential roles for human and mouse Naip in apoptotic signaling [36–40, 24]. The other hypothesis regarding Naip function is based solely on its striking architectural similarity to plant resistance or “R” proteins [27, 28]. The plant R proteins characteristically have N-terminal protein-protein-interaction motifs, a centrally located nucleotide binding site, or NBS, and a series of C-terminal leucine-rich repeats [41]. As depicted in Figure S3, the Naip proteins also harbor N-terminal protein-protein-interacting motifs (the
Current Biology 34
BIR domains), a nucleotide binding site (the NACHT domain; [26]), and cysteine-containing leucine-rich repeats near the C terminus. Since R proteins are important components of a surveillance system used by plants to detect and respond to secreted bacterial virulence proteins [41], it is possible that the Naip proteins may serve a similar function in detecting Legionella virulence proteins. It has been suggested by others that Dot/Icm-dependent apoptosis of macrophages plays an important role in Legionella pathogenesis [4, 42]. Nevertheless, we favor the theory that Naip5 regulates Legionella permissiveness by detecting the presence of Legionellasecreted virulence factors and initiating a (unknown) response of the cell that prevents the bacterium from gaining access to the replicative phagosome. Our reasons for this preference are simple: we have never been able to observe differential effects of macrophage apoptosis at the low multiplicities of infection that we use in our assays (data not shown), and we have previously documented effects of Lgn1 alleles on intracellular Legionella targeting [25]. Of course, the apparent effect of Naip5 on intracellular targeting of Legionella does not preclude it from having a function in apoptosis at later points of the infection. In that regard, it is interesting to note that many plant R proteins initiate an apoptotic response upon detecting bacterial virulence factors [41]. Future work to resolve the remaining questions of the function of Naip5 in Legionella pathogenesis will provide important information to distinguish among these and other possibilities. Conclusions We have used genetics, molecular biology, and functional studies to show that Naip5 (Birc1e) alleles play a role in determining macrophage permissiveness to Legionella replication. This observation opens up many new avenues of research regarding the role of apoptosis and/or innate intracellular response in resistance to Legionella pathogenesis. Should the Naip genes be functional mammalian homologs of plant R genes, it would seem likely that the other Naip gene family members are maintained in the genome to detect and respond to the presence of other types of infections. Experimental Procedures Macrophage Culture Primary macrophages were differentiated from bone marrow cells following a published protocol [43], as described previously [17]. Bacterial Strains and Media The strain of Legionella pneumophila used in this work is Lp02, a streptomycin-resistant, thymine-auxotrophic derivative of Philadelphia-1 [44]. Lp02 can be propagated either on specially constructed solid [44] or liquid [45] media or inside of host cells. Since Lp02 is a thymidine auxotroph, all bacterial or cell culture media with Legionella was supplemented with thymidine to 100 g/ml. Preparation of Legionella Inocula Our method is based on the observation that Legionella undergoes a profound developmental change to a highly virulent, motile form upon amino acid starvation [45, 46]. Such highly virulent inocula are required to see permissiveness in many strains. See the Supplementary Experimental Procedures for details.
Determination of Legionella Permissiveness Phenotypes After harvesting, the primary macrophages were replated in 24-well tissue culture plates at 2.5 ⫻ 105 cells per well in 0.5 ml BMM, without penicillin and streptomycin (BMM-P/S). Each macrophage sample was plated in duplicate in seven distinct replicates. These cells were then allowed to adhere overnight before infection. Based on the OD600 of the chosen Lp02 culture, BMM-P/S would be prepared that had a bacterial concentration of approximately 10 cfu per l (BMM-P/S ⫹ Lp02). The media on the replated macrophages was then aspirated and replaced with 0.5 ml BMM-P/S ⫹ Lp02. After 1.5 hr of incubation in the tissue culture incubator, the infected replicates were processed as previously described [17]. Historically, nonpermissive strains such as C57BL/6J have exhibited bacterial yields at 144 hr of 13,000 ⫾ 13,000 colony-forming units. We have used this mean plus three standard deviations (52,000) as the upper boundary for bacterial yields in a nonpermissive strain. Therefore, any sample that exceeds this bacterial yield is considered to be permissive. Mouse Strains, DNAs, and Crosses These methods are described in the Supplementary Experimental Procedures. Determination of Genomic Sequences The genomic sequence of the Lgn1 interval from C57BL/6J was done by sequencing a BAC clone (286d6) with methods that have been extensively described elsewhere [20]. Since no A/J genomic libraries were easily available, the genomic sequence of the Lgn1 interval from A/J was determined by shearing, cloning, and approximately 7⫻ shotgun sequencing (as described in [20]) of an overlapping set of 16 PCR fragments, as described in the Supplementary Experimental Procedures. Genetic Analysis of Naip2 (Birc1b) Missense Polymorphisms Genomic DNA from critical recombinant animals from the (A/J ⫻ C57BL/6J) ⫻ A/J backcross were amplified (as described in the Supplementary Experimental Procedures) by using the LPCR #9B assay in Table S1. The fragments were purified from agarose gels and were sequenced directly. The sequence traces of the recombinant animals were compared to contemporaneous A/J and A ⫻ C57BL/6J animals by using Sequencher 3.1.1 (GeneCodes) to determine if the polymorphic positions contained only the A/J base or a mixture of A/J and C57BL/6J bases. These data were compiled and compared to the known genotypes and order of nearby polymorphic markers to define the location of recombination breakpoints (Figure 2C). Isolation of Macrophage RNA Total RNA was generated from approximately 2 ⫻ 106 primary mouse macrophages at 6 days of differentiation by using standard procedures. The RNA for the Northern blot shown in Figure 3A was isolated as above from differentiated macrophages that were either uninfected or infected for 6.5 hr with Lp02 at a multiplicity of infection of 1. Northern Blotting A total of 15 g of total RNA was separated on a formaldehyde gel and blotted as described by others [47], except that we used 0.2 M formaldehyde in the gel. The filters were hybridized with 32Plabeled probes generated by using the following PCR primers: Naip2 probe: 5⬘-AAGCCCAGGAACTTGAACCT-3⬘ and 5⬘-CTGTGGAGGGA AGATAGGTCC-3⬘, Gapdh: 5⬘-CCACCCAGAAGACTGTGGAT-3⬘ and 5⬘-AGGGTTTCTTACTCCTTGGAGG-3⬘. The Naip2 probe hybridizes to a portion of exon 11 that is present only in Naip2 [23]. cDNA Synthesis Oligo dT-primed cDNA was made by using standard procedures. Allele-Specific RT-PCR of Naip2 (Birc1b) and Naip5 (Birc1e) The primers used in Figure 3D are as follows: Naip2: 5⬘-CCCGAG GAAGAAGCAGGAAC-3⬘ and 5⬘-CTCTACAGAGAACAGACACC-3⬘, Naip5: 5⬘-GAAGTGTGTCTGAGCAGCAG-3⬘ and 5⬘-ATAGATGCACA ACCACATCAAGGC-3⬘. Allele-specific RT-PCR was then performed
Naip5 Affects Resistance to Legionella pneumophila 35
by using Platinum Taq (Invitrogen) with minor modifications to the manufacturer’s instructions, as described in the Supplementary Experimental Procedures. Determination of cDNA Sequences Intact Naip2 and Naip5 cDNA was amplified as described above from macrophage total cDNA by using the primers shown in Table S1. Screening and Propagation of Transgenic Lines Transgenic founder and carrier animals were identified by screening tail biopsy DNA (as previously described in [17]) for the previously published markers depicted in Figure 1 [18, 19, 21]. The only previously unpublished primer pair is D13Die40 5⬘-CACAGGAACT GAGGTGGGTT-3⬘ and 5⬘-CCTGCATCTGTAGGTAAAAGGG-3⬘. All the transgenes described in this work were backcrossed twice onto a pure FvB background prior to inbreeding the transgene. The experiments performed on animals of the FvB background (Figure 4) were done with animals obtained at inbreeding generations 4–6. As described for Figure 3D, the transgenes were also backcrossed onto an A/J background, by using N2 backcross animals to FvB as the parents.
Supplementary Material Supplementary Material including more detailed Experimental Procedures and additional supporting data is available at http://images. cellpress.com/supmat/supmatin.htm. Acknowledgments The authors thank Julie Sprague and the staff of the Biopolymers Facility at Harvard Medical School for technical assistance. We also thank Victor Boyartchuk, Karl Broman, James Watters, Rebecca Mosher, Eric Boyden, Mark Daly, David Reich, and David Altshuler for helpful discussions. This work was supported in part by grants from the Muscular Dystrophy Association and the National Institutes of Health (AI49987) to W.F.D. W.F.D. is an Assistant Investigator of the Howard Hughes Medical Institute. Received: September 9, 2002 Revised: November 13, 2002 Accepted: November 15, 2002 Published: January 8, 2003 References
Antibodies Affinity-purified antisera was obtained from Zymed Laboratories, following immunization of rabbits with the following peptides: Naip2: CSVLKRQQEEDHKTR, Naip5: CRRNFAQDEEIIKNYENIR. Naip Expression Constructs Naip cDNAs were amplified from A/J RNA by using the following primers: Naip1 and Naip6: 5⬘-CGGGATCCAAAATGGCTGAGCAT-3⬘ and 5⬘-AAATATGCGGCCGCTCCCTCCAGGACAACAGGAG-3⬘, Naip2: 5⬘-CGGGATCCAAAATGGCAGCCCAG-3⬘ and 5⬘-AAATATGCGGCC GCTCCCTTCTGAATGACAGGAG-3⬘, Naip5: 5⬘-CGGGATCCAAAAT GGCTGAGCAT-3⬘ and 5⬘-AAATATGCGGCCGCTCCCTCCAGGAT AACAGGAG-3⬘. Following TA cloning, sequencing, and correction using site-directed mutagenesis, NotI/BamHI fragments were cloned into NotI/BamHI-digested pcDNA3.1 myc/HIS vector (Invitrogen). Transfections The 293T protein lysates used in Figures 3B and 3C were obtained as described in the Supplementary Experimental Procedures. Immunoprecipitations Immunoprecipitations were carried out by using standard procedures, as described in the Supplementary Experimental Procedures. SDS-PAGE/Western Western blots were performed by using standard procedures, as described in the Supplementary Experimental Procedures. Morpholino Treatment of Macrophages Following differentiation, the primary macrophages were replated into 24-well plates as described above for the Legionella infection assays. After an overnight incubation, the cells were treated with Special Delivery morpholinos (Gene Tools, LLC) by using the manufacturer’s Delivery Solution and procedures, with the following minor modification: the morpholino treatment was carried out in media containing only 0.77% fetal bovine serum. Following the incubation of the cells with the morpholino in the low serum media, the media was replaced with regular BMM, and the cells were allowed to recover overnight before the Legionella infections were performed. The sequence of morpholino used to inhibit Naip5 translation is TATGATGAGCACAATGACCAAAGCT. This antisense morpholino is designed to anneal to the 5⬘ UTR of Naip5, just prior to the initiating methionine. While this morpholino has no significant homology to Naip2, it can also potentially inhibit the translation of Naip6 (present in the genome of A/J and C57BL/6J) and Naip7 (not present in the A/J and C57BL/6J genomes), which are closely related to Naip5 [19]. In addition, we could find no significant BLAST homology of this sequence to any gene in GenBank, other than the aforementioned Naips.
1. Bernstein, M.S., and Locksley, R.M. (1991). Legionella infections. In Harrison’s Principles of Internal Medicine, 12th Edition, J.D. Wilson, E. Braunwald, K.J. Isselbacher, R.G. Petersdorf, J.B. Martin, A.S. Fauci, and R.K. Root, eds. (New York: McGrawHill), pp. 634–637. 2. Fang, G.D., Fine, M., Orloff, J., Arisumi, D., Yu, V.L., Kapoor, W., Grayston, J.T., Wang, S.P., Kohler, R., Muder, R.R., et al. (1990). New and emerging etiologies for community-acquired pneumonia with implications for therapy: a prospective multicenter study of 359 cases. Medicine (Baltimore) 69, 307–316. 3. Atlas, R.A. (1999). Legionella: from environmental habitats to disease pathology, detection and control. Environ. Microbiol. 1, 283–293. 4. Harb, O.S., Gao, L.Y., and Abu Kwaik, Y. (2000). From protozoa to mammalian cells: a new paradigm in the life cycle of intracellular bacterial pathogens. Environ. Microbiol. 2, 251–265. 5. Marra, A., and Shuman, H.A. (1992). Genetics of Legionella pneumophila virulence. Annu. Rev. Genet. 26, 51–69. 6. Ciancotto, N., Eisenstein, B.I., Engleberg, N.C., and Shuman, H. (1989). Genetics and molecular pathogenesis of Legionella pneumophila, an intracellular parasite of macrophages. Mol. Biol. Med. 6, 409–424. 7. Swanson, M.S., and Fernandez-Moreia, E. (2002). A microbial strategy to multiply in macrophages: the pregnant pause. Traffic 3, 170–177. 8. Roy, C.R., and Tilney, L.G. (2002). The road less traveled: transport of Legionella to the endoplasmic reticulum. J. Cell Biol. 158, 415–419. 9. Segal, G., Purcell, M., and Shuman, H.A. (1998). Host cell killing and bacterial conjugation require overlapping sets of genes within a 22-kb region of the Legionella pneumophila genome. Proc. Natl. Acad. Sci. USA 95, 1669–1674. 10. Vogel, J.P., Andrews, H.L., Wong, S.K., and Isberg, R.R. (1998). Conjugative transfer by the virulence system of Legionella pneumophila. Science 279, 873–876. 11. Sexton, J.A., and Vogel, J.P. (2002). Type IVB secretion by intracellular pathogens. Traffic 3, 178–185. 12. Nagai, H., Kagan, J.C., Zhu, X., Kahn, R.A., and Roy, C.R. (2002). A bacterial guanine nucleotide exchange factor activates ARF on Legionella phagosomes. Science 295, 679–682. 13. Yamamoto, Y., Klein, T.W., Newton, C.A., Widen, R., and Friedman, H. (1988). Growth of Legionella pneumophila in thioglycollate-elicited peritoneal macrophages from A/J mice. Infect. Immun. 56, 370–375. 14. Yamamoto, Y., Klein, T.W., and Friedman, H. (1991). Legionella pneumophila growth in macrophages from susceptible mice is genetically controlled. Proc. Exp. Biol. Med. 196, 405–409. 15. Yoshida, S.I., Goto, Y., Mizuguchi, Y., Nomoto, K., and Skamene, E. (1991). Genetic control of natural resistance in mouse macro-
Current Biology 36
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33. 34.
phages regulating intracellular Legionella pneumophila multiplication in vitro. Infect. Immun. 59, 428–432. Beckers, M.C., Yoshida, S., Morgan, K., Skamene, E., and Gros, P. (1995). Natural resistance to infection with Legionella pneumophila: chromosomal localization of the Lgn1 susceptibility gene. Mamm. Genome 6, 540–545. Dietrich, W.F., Damron, D.M., Isberg, R.R., Lander, E.S., and Swanson, M.S. (1995). Lgn1, a gene that determines susceptibility to Legionella pneumophila, maps to mouse chromosome 13. Genomics 26, 443–450. Scharf, J.M., Damron, D., Frisella, A., Bruno, S., Beggs, A.H., Kunkel, L.M., and Dietrich, W.F. (1996). The mouse region syntenic for human spinal muscular atrophy lies within the Lgn1 critical interval and contains multiple copies of Naip exon 5. Genomics 38, 405–417. Growney, J.D., Scharf, J.M., Kunkel, L.M., and Dietrich, W.F. (2000). Evolutionary divergence of the mouse and human Lgn1/ SMA repeat structures. Genomics 64, 62–81. Endrizzi, M., Huang, S., Scharf, J.M., Kelter, A.R., Wirth, B., Kunkel, L.M., Miller, W., and Dietrich, W.F. (1999). Comparative sequence analysis of the mouse and human Lgn1/SMA interval. Genomics 60, 137–151. Growney, J.D., and Dietrich, W.F. (2000). High-resolution genetic and physical map of the Lgn1 interval in C57BL/6J implicates Naip2 or Naip5 in Legionella pneumophila pathogenesis. Genome Res. 10, 1158–1171. Endrizzi, M.G., Hadinoto, V., Growney, J.D., Miller, W., and Dietrich, W.F. (2000). Genomic sequence analysis of the mouse Naip gene array. Genome Res. 10, 1095–1102. Huang, S., Scharf, J.M., Growney, J.D., Endrizzi, M.G., and Dietrich, W.F. (1999). The mouse Naip gene cluster on Chromosome 13 encodes several distinct functional transcripts. Mamm. Genome 10, 1032–1035. Maier, J.K., Lahoua, Z., Gendron, N.H., Fetni, R., Johnston, A., Davoodi, J., Rasper, D., Roy, S., Slack, R.S., Nicholson, D.W., et al. (2002). The neuronal apoptosis inhibitory protein is a direct inhibitor of caspases 3 and 7. J. Neurosci. 22, 2035–2043. Watarai, M., Derre, I., Kirby, J., Growney, J.D., Dietrich, W.F., and Isberg, R.R. (2001). Legionella pneumophila is internalized by a macropinocytotic uptake pathway controlled by the Dot/ Icm system and the mouse Lgn1 locus. J. Exp. Med. 194, 1081– 1096. Koonin, E.V., and Aravind, L. (2000). The NACHT family—a new group of predicted NTPases implicated in apoptosis and MHC transcription activation. Trends Biochem. Sci. 25, 223–224. Inohara, N., and Nun˜ez, G. (2001). The NOD: a signaling module that regulates apoptosis and host defense against pathogens. Oncogene 20, 6473–6481. Inohara, N., Ogura, Y., and Nun˜ez, G. (2002). Nods: a family of cytosolic proteins that regulate the host response to pathogens. Curr. Opin. Microbiol. 5, 76–80. Girardin, S.E., Sansonetti, P.J., and Philpott, D.J. (2002). Intracellular vs extracellular recognition of pathogens—common concepts in mammals and flies. Trends Microbiol. 10, 193–199. Lindblad-Toh, K., Winchester, E., Daly, M.J., Wang, D.G., Hirschhorn, J.N, Laviolette, J.P., Ardlie, K., Reich, D.E., Robinson, E., Sklar, P., et al. (2000). Large-scale discovery and genotyping of single-nucleotide polymorphisms in the mouse. Nat. Genet. 24, 381–386. Diez, E., Yaraghi, Z., MacKenzie, A., and Gros, P. (2000). The neuronal apoptosis inhibitory protein (Naip) is expressed in macrophages and is modulated after phagocytosis and during intracellular infection with Legionella pneumophila. J. Immunol. 164, 1470–1477. Summerton, J., and Weller, D. (1997). Morpholino antisense oligomers: design, preparation, and properties. Antisense Nucleic Acid Drug Dev. 7, 187–195. Morcos, P.A. (2001). Achieving efficient delivery of morpholino oligos in cultured cells. Genesis 30, 94–102. Roy, N., Mahadevan, M.S., McLean, M., Shutler, G., Yaraghi, Z., Farahani, R., Baird, S., Besner-Johnston, A., Lefebvre, C., Kang, X., et al. (1995). The gene for neuronal apoptosis inhibitory protein is partially deleted in individuals with spinal muscular atrophy. Cell 80, 167–178.
35. Miller, L.K. (1999). An exegesis of IAPs: salvation and surprises from BIR motifs. Trends Cell Biol. 9, 323–328. 36. Liston, P., Roy, N., Tamai, K., Lefebvre, C., Baird, S., ChertonHorvat, G., Farahani, R., McLean, M., Ikeda, J.E., MacKenzie, A., et al. (1996). Suppression of apoptosis in mammalian cells by NAIP and a related family of IAP genes. Nature 379, 349–353. 37. Gotz, R., Karch, C., Digby, M.R., Troppmair, J., Rapp, U.R., and Sendtner, M. (2000). The neuronal apoptosis inhibitory protein suppresses neuronal differentiation and apoptosis in PC12 cells. Hum. Mol. Genet. 9, 2479–2489. 38. Mercer, E.A., Korhonen, L., Skoglosa, Y., Olsson, P.A., Kukkonen, J.P., and Lindholm, D. (2000). NAIP interacts with hippocalcin and protects neurons against calcium-induced cell death through caspase-3 dependent and independent pathways. EMBO J. 19, 3597–3607. 39. Xu, D.G., Crocker, S.J., Doucet, J.P., St-Jean, M., Tamai, K., Hakim, A.M., Ikeda, J.E., Liston, P., Thompson, C.S., Korneluk, R.G., et al. (1997). Elevation of neuronal expression of NAIP reduces ischemic damage in the rat hippocampus. Nat. Med. 3, 997–1004. 40. Holcik, M., Thompson, C.S., Yaraghi, Z., Lefebvre, C.A., MacKenzie, A.E., and Korneluk, R.G. (2000). The hippocampal neurons of neuronal apoptosis inhibitory protein 1 (NAIP1)-deleted mice display increased vulnerability to kainic acid-induced injury. Proc. Natl. Acad. Sci. USA 97, 2286–2290. 41. Staskawicz, B.J., Mudgett, M.B., Dangl, J.L., and Galan, J.E. (2001). Common and contrasting themes of plant and animal diseases. Science 292, 2285–2289. 42. Zink, S.D., Pedersen, L., Ciancotto, N.P., and Abu-Kwaik, Y. (2002). The Dot/Icm Type IV secretion system of Legionella pneumophila is essential for the induction of apoptosis in human macrophages. Infect. Immun. 70, 1657–1663. 43. Celada, A., Gray, P., Rinderknecht, E., and Schreiber, R. (1984). Evidence for a gamma-interferon receptor that regulates macrophage tumoricidal activity. J. Exp. Med. 160, 55–74. 44. Berger, K.H., and Isberg, R.R. (1993). Two distinct defects in intracellular growth complemented by a single genetic locus in Legionella pneumophila. Mol. Microbiol. 7, 7–19. 45. Byrne, B., and Swanson, M.S. (1998). Expression of Legionella pneumophila virulence traits in response to growth conditions. Infect. Immun. 66, 3029–3034. 46. Hammer, B.K., Tateda, E.S., and Swanson, M.S. (2002). A twocomponent regulator induces the transmission phenotype of stationary-phase Legionella pneumophila. Mol. Microbiol. 44, 107–118. 47. Brown, I., and Mackey, K. (1997). Analysis of RNA by Northern and Slot blot hybridization. In Current Protocols in Molecular Biology, F.M. Ausubel, R. Brent, R.E. Kingston, D.D. Moore, J.G. Seidman, J.A. Smith, and K. Struhl, eds. (New York, New York: John Wiley and Sons), pp. 4.9.1–4.9.16. Note Added in Proof We have become aware of data from Diez, Lee, Gauthier, Yaraghi, Tremblay, Vidal, and Gros that is generally consistent with what we report here. Their paper is in press at Nature Genetics (doi: 10.1038/ ng1065).
PII S0960-9822(02)01359-3
Naip5 Affects Host Susceptibility to the Intracellular Pathogen Legionella pneumophila Edward K. Wright, Jr.,1 Sheryl A. Goodart,1 Joseph D. Growney,2 Vey Hadinoto,2 Matthew G. Endrizzi,2 E. Michelle Long,1 Keyvan Sadigh,2 Andrew L. Abney,2 Isaac Bernstein-Hanley,2 and William F. Dietrich1,2,* 1 Howard Hughes Medical Institute 2 Department of Genetics Harvard Medical School Boston, Massachusetts 02115
Summary Background: Legionella pneumophila is a gram-negative bacterial pathogen that is the cause of Legionnaires’ Disease. Legionella produces disease because it can replicate inside a specialized compartment of host macrophages. Macrophages isolated from various inbred mice exhibit large differences in permissiveness for intracellular replication of Legionella. A locus affecting this host-resistance phenotype, Lgn1, has been mapped to chromosome 13, but the responsible gene has not been identified. Results: Here, we report that Naip5 (also known as Birc1e) influences susceptibility to Legionella. Naip5 encodes a protein that is homologous to plant innate immunity (so-called “resistance”) proteins and has been implicated in signaling pathways related to apoptosis regulation. Detailed recombination mapping and analysis of expression implicates Naip5 in the Legionella permissiveness differences among mouse strains. A bacterial artificial chromosome (BAC) transgenic line expressing a nonpermissive allele of Naip5 exhibits a reduction in macrophage Legionella permissiveness. In addition, morpholino-based antisense inhibition of Naip5 causes an increase in the Legionella permissiveness of macrophages. Conclusions: We conclude that polymorphisms in Naip5 are involved in the permissiveness differences of mouse macrophages for intracellular Legionella replication. We speculate that Naip5 is a functional mammalian homolog of plant “resistance” proteins that monitor for, and initiate host response to, the presence of secreted bacterial virulence proteins. Introduction Legionella pneumophila is a gram-negative bacterial pathogen that is the cause of a severe pneumonia called Legionnaires’ Disease [1]. Some studies suggest that Legionella infection may account for as many as 5%– 10% of community-acquired pneumonia cases [2]. Contaminated fresh water supplies are typically the sources of human Legionella infections, with aspiration or inhalation of infectious aerosols believed to be the usual inoculation route [3, 4]. *Correspondence: [email protected]
Legionella is capable of intracellular growth inside of eukaryotic host cells, including mammalian macrophages, and the ability of Legionella to replicate inside of host macrophages is a crucial aspect of its virulence in animal models [5, 6]. In order to replicate efficiently in macrophages, Legionella must actively block or alter normal cellular endocytic physiology, allowing it to gain access to an unusual rough ER-bounded phagosome [7, 8]. The intracellular trafficking of Legionella depends on the function of the dot/icm genes [9]. The Dot/Icm proteins comprise a specialized secretion system that serves to transfer bacterial proteins into the cytosol of target host cells [10–12]. Several studies suggest that a functional Dot/Icm apparatus is needed within minutes of bacterial uptake to mediate intracellular trafficking decisions that determine the outcome of the infection [10–12]. Recently, it has been shown that Legionella, in a Dot/ Icm-dependent manner, secretes a guanine nucleotide exchange factor (RalF) into the host cell. RalF apparently alters the activity of host ADP ribosylation factor 1 (ARF1), thus allowing the bacterium to recruit ERderived membranes as part of its strategy to remodel its vacuole [12]. Since RalF is not essential for virulence of Legionella in mammalian cells, it is likely that there are other Dot/Icm-secreted effector proteins that are important for growth in macrophages. The future identification of these additional Dot/Icm-secreted proteins and the host pathways that they affect will be crucial for a full understanding of the pathogenesis of Legionella. Mammalian genetics is a potentially useful approach that can be used to understand the pathogenesis of Legionnaires’ Disease. It has been noted that in vitrocultured macrophages from two different inbred mouse strains (A/J and C57BL/6J) differ in their permissiveness for intracellular Legionella replication [13]. This phenotypic difference segregates in a Mendelian fashion [14, 15], and the gene responsible for the macrophage permissiveness difference maps to a locus called Lgn1 on mouse chromosome 13 [16, 17]. Detailed characterization of the mouse Lgn1 interval has demonstrated that it consists of a series of 80- to 100-kb direct repeats, each containing a paralogous member of the Neuronal Apoptosis Inhibitory Protein (Naip, also known as the Birc1) gene family [18, 19]. More recently, we have shown that the Lgn1 mutation must lie within an interval that contains only the first 12 exons of Naip2 (Birc1b), the entirety of Naip5 (Birc1e), and a fragmentary ⌬Naip (Birc1c-ps1) pseudogene (Figure 1; [20–22]). Most of the mouse Naip paralogs are functional transcription units that encode proteins having 85% identity to each other [23]. Much research has focused on the potential role of Naip proteins in apoptotic signaling [24]. However, it has been shown that Lgn1 (that is, either Naip2 or Naip5) alleles can influence very early events of intracellular trafficking of Legionella, suggesting that the Lgn1 gene product likely has some function(s) that
Current Biology 28
Figure 1. Gene Content of Lgn1 Critical Interval A depiction of the map of the Lgn1 interval, along with the positions of relevant markers and genomic clones, is shown. The right of the figure points away from the centromere of chromosome 13. The size of the Lgn1 interval, in kb, is indicated by the scale at the top of the figure [20, 22]. The bold horizontal line depicts the extent of the C57BL/6J-derived BAC clone 235. The map position of this clone was previously reported [19]. The three lines with arrowheads depict the position and orientation of genes (Naip2 [Birc1b] and Naip5 [Birc1e]) and gene fragments (⌬Naip) in the Lgn1 interval [19, 23]. The exon positions and numbers for Naip2 (Birc1b) and Naip5 (Birc1e) are indicated by small, labeled lines drawn above a thin horizontal line that depicts the chromosome. Below the chromosome are the names and approximate positions of relevant microsatellite markers. The pattern-filled boxes shown below the chromosome indicate the extent of the minimal genetic intervals for Lgn1. The box labeled “Lgn1 Prior” shows the minimal Lgn1 interval that was previously determined [21]. “Lgn1 Current” depicts the minimal Lgn1 interval after utilizing Naip2 (Birc1b) missense polymorphisms to refine the location of the centromere-proximal recombination events (Figure 2C). The transgene content of BAC 235 transgenic lines are indicated at the bottom of the figure, with “⫹” indicating that the named strain contains the C57BL/6J allele of that marker in its genome; “⫺” indicates that the strain does not have the C57BL/6J allele of the marker.
are unrelated to its putative role in apoptosis regulation [25]. It has been suggested that the Naip genes may actually be mammalian homologs of plant genes that are known to mediate innate immunity to bacterial pathogens (the so-called “resistance” or “R” genes) [26–29]. Here, we report data demonstrating that polymorphisms in Naip5 cause alterations in the Lgn1 phenotype. This conclusion is based primarily on our genetic and molecular analyses of Naip2 and Naip5, on our alteration of the Lgn1 phenotype with a bacterial artificial chromosome (BAC) transgene expressing only Naip5, and on our ability to increase macrophage permissiveness with Naip5 antisense reagents. Our identification of a gene affecting the Lgn1 phenotype provides a framework for future studies aimed at identifying the specific Lgn1 mutation(s) and at understanding the molecular role that is played by Naip5 in host resistance to Legionella pathogenesis. Results The FvB/N Mouse Is Permissive Due to an Allele of Lgn1 As a starting point for distinguishing between the Lgn1 candidate genes (Figure 1), we undertook the phenotypic characterization of additional inbred mouse strains. As previously reported, macrophages from A/J
mice support vigorous intracellular Legionella growth, while those from C57BL/6J mice rarely yield more than 2 ⫻ 104 colony-forming units (cfu) at 6 days postinfection (Figure 2A). In addition, we found that the FvB/NJ (FvB) mouse strain is clearly permissive (see the Experimental Procedures). We believe that FvB is permissive because of an allele of Lgn1. The permissiveness phenotype of FvB is recessive (Figure S1A) and appears to be in the same complementation group as the Lgn1 allele found in A/J, since macrophages from A/J ⫻ FvB/N F1 hybrids are permissive (Figure 2A). In addition, we found that ten (A/J ⫻ FvB/NJ) ⫻ A/J backcross animals were all permissive, making it unlikely that an FvB allele unlinked to Lgn1 accounts for the phenotype of A/J ⫻ FvB/N F1 hybrids. A Modifier of Permissiveness from FvB Maps to Lgn1 The 144-hr bacterial yields from macrophages of A ⫻ FvB F1 animals are reproducibly at least 5-fold lower than those seen in A/J macrophages (Figure 2A). Interestingly, six of the previously described (A/J ⫻ FvB/ NJ) ⫻ A/J backcross animals exhibited an “A ⫻ FvB F1like” permissiveness phenotype, while four others were “A/J-like” (Figure 2B). The segregation of this qualitative difference in permissiveness correlated perfectly with the inheritance of a marker in Naip5, D13Die40, establishing linkage of this trait to the Lgn1 locus (LOD score ⫽
Naip5 Affects Resistance to Legionella pneumophila 29
Figure 2. Characterization and Mapping of Legionella Permissiveness of Mouse Macrophages (A) Bone marrow-derived macrophages from each of the indicated strains were infected with Legionella inocula prepared as described [45, 46]. A ⫻ FvB F1 is an F1 hybrid between A/J and FvB/NJ. Shown is the average log10 of the number of bacterial colony-forming units for each of the samples at different times postinfection. The error bars indicate the standard deviation of the average. A/J, n ⫽ 11; C57BL/6J, n ⫽ 12; FvB/NJ, n ⫽ 6; A ⫻ FvB F1, n ⫽ 12. (B) Data from a 10-animal (A/J ⫻ FvB) ⫻ A/J cross. The genotype and phenotype of A ⫻ FvB F1 hybrid (F1) and A/J animals are shown by the two leftmost columns. The three rightmost columns indicate the genotypes and phenotypes of all the progeny, including the single recombinant animal. In our experience, the behavior of the A ⫻ FvB F1 macrophages is very consistent, yielding 6.58 ⫾ 0.13 log10 cfu at 144 hr postinfection. Accordingly, the phenotype of the animals was declared to be A ⫻ FvB F1-like if the total bacterial yield at 144 hr was within three standard deviations of this average. Importantly, animal #2880 has a clear “A ⫻ FvB F1-like” permissiveness phenotype (144-hr yield of 6.6 log10 cfu, which was 10-fold less than the yield of the contemporaneous A/J control), yet it is an A/J homozygote for markers within Naip2 (see the Supplementary Experimental Procedures). (C) Data from a 466-animal (A/J ⫻ C57BL/6J) ⫻ A/J cross [18]. The genotype and phenotype of A ⫻ B F1 hybrid (F1) and A/J animals are shown by the two leftmost columns. The four rightmost columns indicate the genotypes and phenotypes of the four informative backcross animals. The assignment of phenotypes was documented in a previous publication [18].
3). However, this variation in susceptibility cannot be accounted for by missense polymorphisms in Naip2, since a single recombinant animal (#2880) excludes the entire Naip2 ORF (which begins in exon 3; see Figure 1). Furthermore, we do not observe differences in Naip2 mRNA levels between A/J, FvB, and C57BL/6J (Figure 3A), eliminating the possibility that Naip2 regulatory mutations result in this modifying effect. The Lgn1 Locus of Other Inbred Strains We also characterized the phenotypes of several other inbred strains. The P/J mouse exhibits a level of nonpermissiveness similar to C57BL/6J (Figure S1B). In contrast, C3H/HeJ (C3H), BALB/cJ (BALB), and 129S1 (129) macrophages behave similarly to those of FvB (Figures S1A and S1C). We mapped the Legionella permissiveness phenotype of the 129 strain to the Lgn1 locus (LOD score ⫽ 6.7) by infecting macrophages isolated from a 29-animal (129S1 ⫻ C57BL/6J) F2 cross and by analyzing the yield of bacterial colony-forming units at 144 hr as a quantitative trait (see the Supplementary Experimental Procedures available with this article online).
We determined the genomic and cDNA sequences of the Naip2 and Naip5 genes from some of the permissive and nonpermissive strains described above (Table 1). These sequences exhibit a high single nucleotide polymorphism (SNP) rate that is in the vicinity of 0.3%–0.5% [30]. Overall, no nonsense or rearrangement mutations that would obviously abolish Naip RNA or protein expression were observed, but comparison of Naip2 and Naip5 cDNA sequences from all the strains allowed us to identify candidate Lgn1 missense mutations (Tables 2 and 3). Interestingly, the near identity of the P/J and C57BL/6J sequences suggests that these nonpermissive mice share an Lgn1 haplotype that is derived from a more recent common ancestor than is found among the Legionella permissive strains (Tables 2 and 3). Refined Genetic Mapping of Lgn1 In addition to potentially contributing to the interstrain variation in Legionella susceptibility, the Naip gene polymorphisms can be used as genetic markers to refine the localization of the Lgn1 interval. Therefore, we used our previously characterized Lgn1 mapping cross (an [A/J ⫻ C57BL/6J] ⫻ A/J backcross) to position the Naip2
Current Biology 30
Figure 3. Expression of Naip (Birc1) Genes in Inbred Strain Macrophages (A) Naip2 (Birc1b) mRNA levels. Shown is a bone marrow macrophage total RNA Northern blot (with 15 g RNA per lane) hybridized with a probe that specifically hybridizes to Naip2 (upper panel) and then subsequently hybridized to a Gapdh probe as a loading control (lower panel). Each inbred strain is represented by a pair of lanes, with the leftmost of the pair containing RNA from an uninfected macrophage sample (⫺) and the rightmost lane containing RNA from macrophages that have been infected for 6.5 hr with Legionella (⫹) at an multiplicity of infection (m.o.i.) of 1 (see the Experimental Procedures for details). While there seems to be a slight upregulation of Naip2, there are no substantial differences in the levels or induction of Naip2 mRNA between A/J, C57BL/6J, and FvB. We have observed similar data from several other Naip2 Northern experiments with samples obtained from A/J and C57BL/6J macrophages and infected for 1.5, 5.5, 9.5, and 13.5 hr at m.o.i.s of 1 and 5 (data not shown). (B) Specificity of Naip antisera. Shown is a Western blot of protein isolated from 293T cells transfected with the indicated Naip gene expression constructs (see the Experimental Procedures). The labels on the right indicate the antiserum used to probe the blot; the labels on the left indicate the approximate position and mass (in kD) of protein size standards. M, mock transfected; N5, Naip5 construct; N2, Naip2 construct; N6, Naip6 construct. Although it is not shown, neither antiserum crossreacts with Naip1. (C) Expression of Naip proteins in A/J and C57BL/6J macrophages. Shown are a Western blot of total cellular protein probed for Naip2 and a Western blot of Naip5 antibody-immunoprecipitated protein probed for Naip5 (see the Supplementary Experimental Procedures). A/J, protein from A/J primary macrophages; B6, protein from C57BL/6J primary macrophages; N2, 293T cells transfected with Naip2 expression construct; N5, 293T cells transfected with Naip5 expression construct. (D) Expression of the BAC transgene in the Full and N5 transgenic lines. C57BL/6J allele-specific RT-PCR assays were performed on the following cDNA samples obtained from bone marrow macrophage total RNA (see the Experimental Procedures): Full, Full transgene carrier (on A ⫻ FvB F1 background); N5, N5 transgene carrier (on A ⫻ FvB F1 background); NT, nontransgenic A ⫻ FvB F1; B6, C57BL/6J; H2O, water; ⫹, B10.A (identical in sequence to C57BL/6J) Naip2 (Birc1b) plasmid or Naip5 (Birc1e) plasmid, as appropriate (AF135490 and AF135492; [23]).
and Naip5 missense polymorphisms relative to Lgn1. As can be seen in Figure 2C, the polymorphism in codon 862 of Naip2 recombines away from the Lgn1 phenotype (in two separate animals, #661 and #549), while the more 5⬘ polymorphisms do not. This result effectively excludes all but seven Naip2 polymorphisms from the Lgn1 interval (Figures 1 and 2C; Table 2).
Naip2 Is Not Differentially Expressed between A/J and C57BL/6J To investigate if Naip2 or Naip5 protein levels could possibly play a role in determining the Lgn1 phenotype, we developed two different antisera that recognize Naip2 and Naip5 (Figure 3B). We used these antisera in
Western blot and immunoprecipitation experiments to evaluate the levels of Naip2 and Naip5 proteins in macrophages from A/J and C57BL/6J mice. While the levels of Naip2 protein in these two strains are equivalent, the levels of protein recognized by the Naip5 antibody are substantially higher in C57BL/6J (Figure 3C). Since the Naip5 antibody crossreacts with Naip6 (Figure 3B), the signal detected by this antibody in the macrophage lysates is likely comprised of a mixture of Naip5 and Naip6. However, the mRNA for Naip5 is expressed at higher levels than Naip6 [23], suggesting that the majority of what we detect is Naip5. The magnitude of this differential expression appears similar to the 4-fold differences in Naip expression reported previously [31]; this finding suggests that the differences that these authors ob-
Table 1. Genomic and cDNA Sequences of Lgn1 Candidates Strain
Genomic
Naip2 (Birc1b) cDNA
Naip5 (Birc1e) cDNA
References
C57BL/6J P/J FvB/N C3H/HeJ BALB/cJ A/J 129S1
AF367967, AF367968, AF367969 None None None None AF367966 AF131205
AY147001 AY147005 AY147004 AY147003 AY147002 AF381770 AY147000
AY146995 AY146999 AY146998 AY146997 AY146996 AF381771 AY146994
This work This work This work This work This work This work [20] and this work
The GenBank accession numbers of the indicated sequences are shown.
Naip5 Affects Resistance to Legionella pneumophila 31
Table 2. Comprehensive List of Amino Acid Polymorphisms in Naip2 (Birc1b) AA#
Exon
C57BL/6J
P/J
Nonpermissive 377 403 478 540 556 558 561 862 1079 1080 1089 1115 1122 1136 1157 1160 1167 1186 1192 1271 1272
10 11 11 11 11 11 11 11 11 11 11 11 11 11 12 12 12 12 12 14 14
D L L Y G E K K S D R K T D S D G S V F G
FvB/N
C3H/HeJ
BALB/cJ
A/J
129S1
D F L N G E K N S D R K T D S N G Y V F A
D F L N G E K N S D R K T D S N G S V F A
G F L N D G N N F N C E E D G N R S V F G
G F I Y G E K N F N C E E E G D R S V C G
Permissive D L L N G E K K S D R K T D S D G S V F A
D F L Y G E N N S D R K T D S N G S L F A
Polymorphisms and the exonic location of the indicated amino acids (AA#) are shown. The amino acid polymorphisms that distinguish specific strains from all other strains having the opposite phenotype are underlined. The total length of the Naip2 (Birc1b) protein is 1447 amino acids [23].
served were due to variations in Naip5 (and possibly Naip6) expression. A Naip5 Transgene Affects the Lgn1 Phenotype Given the previous data, we sought additional positive evidence implicating Naip5 in affecting the Lgn1 phenotype. To that end, we attempted to complement the Lgn1 phenotype of FvB animals inheriting a C57BL/6J BAC clone transgene containing the Lgn1 candidate genes Naip2 and Naip5. From our injections of this clone (named 235), we obtained two male FvB founder animals that sired progeny inheriting useful portions of the BAC clone. One transgene insertion (named Full) contained polymorphic markers from throughout the BAC clone insert, while the other (named N5) contained polymorphic markers from the entirety of Naip5 and only a portion of Naip2 (Figure 1). These two transgene insertions were stably inherited and were found to express Naip mRNA and protein in macrophages in a manner consistent with the marker content of the transgenes (Figure 3D and data not shown). After backcrossing each transgene twice to FvB and then inbreeding for 4–5 generations (see the Experimental Procedures), we assessed the macrophages of mice from these FvB transgenic lines for evidence of complementation of the Legionella permissiveness phenotype. Surprisingly, the Full transgene has no effect on macrophage permissiveness (Figure 4). In contrast, the growth supported by the N5 transgenic macrophages was always lower than that of contemporaneous FvB control macrophages. While the N5 transgene does not cause the macrophages to become as nonpermissive as C57BL/6J, statistically significant decreases in bacterial yield become apparent by 72 hr postinfection (Figure 4). By 144 hr postinfection, the FvB-N5 transgenic animals
typically yield 5- to 10-fold fewer bacterial colony-forming units than the FvB controls. We also wanted to determine if the BAC transgenes exhibited any effect on the phenotype of more permissive genetic backgrounds. Therefore, we backcrossed the N5 and Full transgenes to the A/J strain and examined the effects of the heterozygous transgenes on the phenotypes of A ⫻ FvB F1 and (A ⫻ FvB) ⫻ A animals. Again, the Full transgene has no significant effect (data not shown). The N5 transgene causes decreases in the 144-hr bacterial yield in the A ⫻ FvB F1 background (a 7-fold average decrease, n ⫽ 3, p ⫽ 0.003) and in the (A ⫻ FvB) ⫻ A-Lgn1A/A background (a 4-fold average decrease, n ⫽ 3, p ⫽ 0.032). Antisense Inhibition of Naip5 Increases Legionella Permissiveness The reduction of permissiveness caused by the N5 transgene is important evidence in favor of Naip5 playing an important role in regulating permissiveness. Nevertheless, we pursued a complementary approach to assess the function of Naip5 with respect to Legionella pathogenesis. We decided to inhibit the function of Naip5 by using morpholinos, which are antisense compounds comprised of an uncharged backbone whose mechanism of action is to inhibit the translation of the target mRNA [32, 33]. As described in the Experimental Procedures, we designed a morpholino that would hybridize to the 5⬘ UTR of Naip5 transcripts from all mouse strains and therefore should inhibit the translation of these transcripts. Since the N5 transgenic line exhibited partial rescue of the permissiveness phenotype, we reasoned that it would be a sensitized background in which we could observe an effect of the morpholino treatment, even if
Current Biology 32
Table 3. Comprehensive List of Amino Acid Polymorphisms in Naip5 (Birc1e) AA#
Exon
C57BL/6J
P/J
Nonpermissive 92 144 234 242 368 472 496 512 514 516 517 521 533 538 647 755 855 952 1021 1092 1116 1123 1137 1140 1228 1241/42 1275/76
3 3 5 5 9 11 11 11 11 11 11 11 11 11 11 11 11 11 11 12 12 12 12 12 14 15 15
R R E S T T Y D G A N A V S A V S S M E N R Q T E V D
FvB/N
C3H/HeJ
BALB/cJ
A/J
129S1
R R K S M A Y D G A N A A I A V T T I D D G R R AQ V D
R R K S M A Y D G A N A A I A V T T I D D G R R AQ V D
R R E S T A N G E A K A A I T M S S M D D G Q T E I N
K S E G T A Y D G D N T A I A V S S M D D G Q T E I N
Permissive R R E S T T Y D G A N A V S A V S S M E N R Q T AQ V D
R R K S M A Y D G A N A A I A V T T I D D G R R AQ V D
Polymorphisms and the exonic location of the indicated amino acids (AA#) are shown. The amino acid polymorphisms that distinguish specific strains from all other strains having the opposite phenotype are underlined. The total length of the Naip5 (Birc1e) protein is 1402 amino acids [23].
the translational inhibition was incomplete. Accordingly, macrophages from the FvB-N5 line that are pretreated with this morpholino have higher bacterial yields than untreated control macrophages at 72 and 120 hr postinfection (Figure 4). This relative increase in permissiveness is somewhat diminished by 144 hr postinfection, suggesting that the inhibitory effect of the morpholinos is transient. In addition, preliminary experiments with C57BL/6J macrophages suggest that treatment with this morpholino can also modestly increase permissiveness in this nonpermissive background (Figure S2). Discussion Positional cloning is a straightforward, if not always easy, process. Genetic linkage provides a discrete region of the genome to be searched. It then becomes a simple matter of searching through regional genes for variations in sequence or expression that correlate with phenotype, or of using tests of complementation to establish a functional effect of a candidate gene. Unfortunately, positional cloning of phenotypes that map to clusters of highly related paralogous genes, such as Lgn1, are much more resistant to these simple strategies. Despite the difficulties presented by the region, we have accumulated evidence favoring a role for Naip5 in modulating permissiveness to Legionella. For example, the entire Naip5 ORF resides in the Lgn1 interval, and there are large differences in Naip5 (and possibly Naip6)
protein levels between A/J and C57BL/6J macrophages. In contrast, our measurements of Naip2 mRNA and protein levels show no differences between permissive and nonpermissive strains. Furthermore, our fine-structure genetic map of Lgn1 eliminates nearly half the open reading frame of Naip2 from the critical interval, leaving only the polymorphism at amino acid 403 as a realistic candidate mutation (Table 2). In addition, we have mapped a modifier of Legionella permissiveness to Lgn1 and have shown that this modifier cannot be caused by variations in Naip2. While it is possible that the modifying allele is in a Naip locus extragenic to Lgn1, the simplest explanation is that there are different permissive alleles of Lgn1 that permit distinct levels of bacterial replication. A casual inspection of Table 3 reveals ample missense polymorphisms in the Naip5 gene that could easily account for variations among permissive strains. The existence of closely linked, highly similar genes presents serious complications in the interpretation of Lgn1 complementation experiments. Nevertheless, we have identified a transgenic line (N5) expressing the C57BL/6J allele of Naip5 that exhibits a reduction in the permissiveness of its macrophages. In addition, we are able to induce an increase in the permissiveness of the N5 line macrophages after treatment with a morpholino designed to inhibit the translation of Naip5 protein. This effect on the transgenic lines is a strong indication that the effect of the N5 transgene on permissiveness is due to the expression of Naip5, and not to disruption of some other gene during transgene integration. While the
Naip5 Affects Resistance to Legionella pneumophila 33
Figure 4. Intramacrophage Growth of Legionella in Transgenic Mice Bone marrow-derived macrophages from transgenic and control strains were infected with Legionella (see the Experimental Procedures). Shown is the log10 of the number of bacterial colony-forming units for each of the samples at different times postinfection. (See the Experimental Procedures). Each symbol represents the average cfu from duplicate infected macrophage samples obtained from a single animal. The experiments shown were performed in small groups over several independent trials. The average of each group of animals is indicated by the short, bold horizontal line. The brackets indicate the p values of comparisons between the controls (on left) and the N5 transgenic or morpholino-treated macrophages. The two-tailed p values were calculated by using the Student’s t test. Control treatment with a nonspecific morpholino sequence caused no effect on the permissiveness of these macrophages (data not shown).
N5 transgene does not cause a complete reversion to nonpermissiveness, preliminary RT-PCR data suggest that the levels of transgenic Naip5 expression in the N5 line are low (data not shown), providing a possible explanation for this observation. The behavior of the Full line, which expresses appreciable levels of both Naip2 and Naip5, but does not exhibit diminished permissiveness, is unexplained. We could not account for the lack of complementation by examining levels of expression of Naip5 or by finding mutations in the Full transgene (data not shown). Since Naip2 and Naip5 are very closely related, it is possible that the two proteins utilize some common cofactor that is present in limiting quantities, and superphysiologic expression of both Naip2 and Naip5 may prevent us from observing the complementing activity of Naip5. We do not know if the Legionella permissiveness phenotype differences that we see are caused by missense polymorphisms in Naip5, alterations in its expression level, or both. However, the candidate mutations that we have identified should be important subjects of a future study using knock-in strategies. In any case, a structure-function study of the candidate mutations that
we have found in the Naip genes will be very important, especially since the molecular functions of the Naip genes are far from clear. There are currently two hypotheses about the molecular function of the Naip proteins. The first Naip family member to be discovered was found in the human genome, and, based on the fact that it encoded proteinprotein-interacting BIR domains (Figure S3), it was hypothesized to be a functional relative of baculoviral IAP genes that can regulate apoptosis [34]. However, not all BIR domain-containing proteins (BIRPs) have a normal physiologic function relevant to apoptosis [35]. Nevertheless, many investigations have uncovered potential roles for human and mouse Naip in apoptotic signaling [36–40, 24]. The other hypothesis regarding Naip function is based solely on its striking architectural similarity to plant resistance or “R” proteins [27, 28]. The plant R proteins characteristically have N-terminal protein-protein-interaction motifs, a centrally located nucleotide binding site, or NBS, and a series of C-terminal leucine-rich repeats [41]. As depicted in Figure S3, the Naip proteins also harbor N-terminal protein-protein-interacting motifs (the
Current Biology 34
BIR domains), a nucleotide binding site (the NACHT domain; [26]), and cysteine-containing leucine-rich repeats near the C terminus. Since R proteins are important components of a surveillance system used by plants to detect and respond to secreted bacterial virulence proteins [41], it is possible that the Naip proteins may serve a similar function in detecting Legionella virulence proteins. It has been suggested by others that Dot/Icm-dependent apoptosis of macrophages plays an important role in Legionella pathogenesis [4, 42]. Nevertheless, we favor the theory that Naip5 regulates Legionella permissiveness by detecting the presence of Legionellasecreted virulence factors and initiating a (unknown) response of the cell that prevents the bacterium from gaining access to the replicative phagosome. Our reasons for this preference are simple: we have never been able to observe differential effects of macrophage apoptosis at the low multiplicities of infection that we use in our assays (data not shown), and we have previously documented effects of Lgn1 alleles on intracellular Legionella targeting [25]. Of course, the apparent effect of Naip5 on intracellular targeting of Legionella does not preclude it from having a function in apoptosis at later points of the infection. In that regard, it is interesting to note that many plant R proteins initiate an apoptotic response upon detecting bacterial virulence factors [41]. Future work to resolve the remaining questions of the function of Naip5 in Legionella pathogenesis will provide important information to distinguish among these and other possibilities. Conclusions We have used genetics, molecular biology, and functional studies to show that Naip5 (Birc1e) alleles play a role in determining macrophage permissiveness to Legionella replication. This observation opens up many new avenues of research regarding the role of apoptosis and/or innate intracellular response in resistance to Legionella pathogenesis. Should the Naip genes be functional mammalian homologs of plant R genes, it would seem likely that the other Naip gene family members are maintained in the genome to detect and respond to the presence of other types of infections. Experimental Procedures Macrophage Culture Primary macrophages were differentiated from bone marrow cells following a published protocol [43], as described previously [17]. Bacterial Strains and Media The strain of Legionella pneumophila used in this work is Lp02, a streptomycin-resistant, thymine-auxotrophic derivative of Philadelphia-1 [44]. Lp02 can be propagated either on specially constructed solid [44] or liquid [45] media or inside of host cells. Since Lp02 is a thymidine auxotroph, all bacterial or cell culture media with Legionella was supplemented with thymidine to 100 g/ml. Preparation of Legionella Inocula Our method is based on the observation that Legionella undergoes a profound developmental change to a highly virulent, motile form upon amino acid starvation [45, 46]. Such highly virulent inocula are required to see permissiveness in many strains. See the Supplementary Experimental Procedures for details.
Determination of Legionella Permissiveness Phenotypes After harvesting, the primary macrophages were replated in 24-well tissue culture plates at 2.5 ⫻ 105 cells per well in 0.5 ml BMM, without penicillin and streptomycin (BMM-P/S). Each macrophage sample was plated in duplicate in seven distinct replicates. These cells were then allowed to adhere overnight before infection. Based on the OD600 of the chosen Lp02 culture, BMM-P/S would be prepared that had a bacterial concentration of approximately 10 cfu per l (BMM-P/S ⫹ Lp02). The media on the replated macrophages was then aspirated and replaced with 0.5 ml BMM-P/S ⫹ Lp02. After 1.5 hr of incubation in the tissue culture incubator, the infected replicates were processed as previously described [17]. Historically, nonpermissive strains such as C57BL/6J have exhibited bacterial yields at 144 hr of 13,000 ⫾ 13,000 colony-forming units. We have used this mean plus three standard deviations (52,000) as the upper boundary for bacterial yields in a nonpermissive strain. Therefore, any sample that exceeds this bacterial yield is considered to be permissive. Mouse Strains, DNAs, and Crosses These methods are described in the Supplementary Experimental Procedures. Determination of Genomic Sequences The genomic sequence of the Lgn1 interval from C57BL/6J was done by sequencing a BAC clone (286d6) with methods that have been extensively described elsewhere [20]. Since no A/J genomic libraries were easily available, the genomic sequence of the Lgn1 interval from A/J was determined by shearing, cloning, and approximately 7⫻ shotgun sequencing (as described in [20]) of an overlapping set of 16 PCR fragments, as described in the Supplementary Experimental Procedures. Genetic Analysis of Naip2 (Birc1b) Missense Polymorphisms Genomic DNA from critical recombinant animals from the (A/J ⫻ C57BL/6J) ⫻ A/J backcross were amplified (as described in the Supplementary Experimental Procedures) by using the LPCR #9B assay in Table S1. The fragments were purified from agarose gels and were sequenced directly. The sequence traces of the recombinant animals were compared to contemporaneous A/J and A ⫻ C57BL/6J animals by using Sequencher 3.1.1 (GeneCodes) to determine if the polymorphic positions contained only the A/J base or a mixture of A/J and C57BL/6J bases. These data were compiled and compared to the known genotypes and order of nearby polymorphic markers to define the location of recombination breakpoints (Figure 2C). Isolation of Macrophage RNA Total RNA was generated from approximately 2 ⫻ 106 primary mouse macrophages at 6 days of differentiation by using standard procedures. The RNA for the Northern blot shown in Figure 3A was isolated as above from differentiated macrophages that were either uninfected or infected for 6.5 hr with Lp02 at a multiplicity of infection of 1. Northern Blotting A total of 15 g of total RNA was separated on a formaldehyde gel and blotted as described by others [47], except that we used 0.2 M formaldehyde in the gel. The filters were hybridized with 32Plabeled probes generated by using the following PCR primers: Naip2 probe: 5⬘-AAGCCCAGGAACTTGAACCT-3⬘ and 5⬘-CTGTGGAGGGA AGATAGGTCC-3⬘, Gapdh: 5⬘-CCACCCAGAAGACTGTGGAT-3⬘ and 5⬘-AGGGTTTCTTACTCCTTGGAGG-3⬘. The Naip2 probe hybridizes to a portion of exon 11 that is present only in Naip2 [23]. cDNA Synthesis Oligo dT-primed cDNA was made by using standard procedures. Allele-Specific RT-PCR of Naip2 (Birc1b) and Naip5 (Birc1e) The primers used in Figure 3D are as follows: Naip2: 5⬘-CCCGAG GAAGAAGCAGGAAC-3⬘ and 5⬘-CTCTACAGAGAACAGACACC-3⬘, Naip5: 5⬘-GAAGTGTGTCTGAGCAGCAG-3⬘ and 5⬘-ATAGATGCACA ACCACATCAAGGC-3⬘. Allele-specific RT-PCR was then performed
Naip5 Affects Resistance to Legionella pneumophila 35
by using Platinum Taq (Invitrogen) with minor modifications to the manufacturer’s instructions, as described in the Supplementary Experimental Procedures. Determination of cDNA Sequences Intact Naip2 and Naip5 cDNA was amplified as described above from macrophage total cDNA by using the primers shown in Table S1. Screening and Propagation of Transgenic Lines Transgenic founder and carrier animals were identified by screening tail biopsy DNA (as previously described in [17]) for the previously published markers depicted in Figure 1 [18, 19, 21]. The only previously unpublished primer pair is D13Die40 5⬘-CACAGGAACT GAGGTGGGTT-3⬘ and 5⬘-CCTGCATCTGTAGGTAAAAGGG-3⬘. All the transgenes described in this work were backcrossed twice onto a pure FvB background prior to inbreeding the transgene. The experiments performed on animals of the FvB background (Figure 4) were done with animals obtained at inbreeding generations 4–6. As described for Figure 3D, the transgenes were also backcrossed onto an A/J background, by using N2 backcross animals to FvB as the parents.
Supplementary Material Supplementary Material including more detailed Experimental Procedures and additional supporting data is available at http://images. cellpress.com/supmat/supmatin.htm. Acknowledgments The authors thank Julie Sprague and the staff of the Biopolymers Facility at Harvard Medical School for technical assistance. We also thank Victor Boyartchuk, Karl Broman, James Watters, Rebecca Mosher, Eric Boyden, Mark Daly, David Reich, and David Altshuler for helpful discussions. This work was supported in part by grants from the Muscular Dystrophy Association and the National Institutes of Health (AI49987) to W.F.D. W.F.D. is an Assistant Investigator of the Howard Hughes Medical Institute. Received: September 9, 2002 Revised: November 13, 2002 Accepted: November 15, 2002 Published: January 8, 2003 References
Antibodies Affinity-purified antisera was obtained from Zymed Laboratories, following immunization of rabbits with the following peptides: Naip2: CSVLKRQQEEDHKTR, Naip5: CRRNFAQDEEIIKNYENIR. Naip Expression Constructs Naip cDNAs were amplified from A/J RNA by using the following primers: Naip1 and Naip6: 5⬘-CGGGATCCAAAATGGCTGAGCAT-3⬘ and 5⬘-AAATATGCGGCCGCTCCCTCCAGGACAACAGGAG-3⬘, Naip2: 5⬘-CGGGATCCAAAATGGCAGCCCAG-3⬘ and 5⬘-AAATATGCGGCC GCTCCCTTCTGAATGACAGGAG-3⬘, Naip5: 5⬘-CGGGATCCAAAAT GGCTGAGCAT-3⬘ and 5⬘-AAATATGCGGCCGCTCCCTCCAGGAT AACAGGAG-3⬘. Following TA cloning, sequencing, and correction using site-directed mutagenesis, NotI/BamHI fragments were cloned into NotI/BamHI-digested pcDNA3.1 myc/HIS vector (Invitrogen). Transfections The 293T protein lysates used in Figures 3B and 3C were obtained as described in the Supplementary Experimental Procedures. Immunoprecipitations Immunoprecipitations were carried out by using standard procedures, as described in the Supplementary Experimental Procedures. SDS-PAGE/Western Western blots were performed by using standard procedures, as described in the Supplementary Experimental Procedures. Morpholino Treatment of Macrophages Following differentiation, the primary macrophages were replated into 24-well plates as described above for the Legionella infection assays. After an overnight incubation, the cells were treated with Special Delivery morpholinos (Gene Tools, LLC) by using the manufacturer’s Delivery Solution and procedures, with the following minor modification: the morpholino treatment was carried out in media containing only 0.77% fetal bovine serum. Following the incubation of the cells with the morpholino in the low serum media, the media was replaced with regular BMM, and the cells were allowed to recover overnight before the Legionella infections were performed. The sequence of morpholino used to inhibit Naip5 translation is TATGATGAGCACAATGACCAAAGCT. This antisense morpholino is designed to anneal to the 5⬘ UTR of Naip5, just prior to the initiating methionine. While this morpholino has no significant homology to Naip2, it can also potentially inhibit the translation of Naip6 (present in the genome of A/J and C57BL/6J) and Naip7 (not present in the A/J and C57BL/6J genomes), which are closely related to Naip5 [19]. In addition, we could find no significant BLAST homology of this sequence to any gene in GenBank, other than the aforementioned Naips.
1. Bernstein, M.S., and Locksley, R.M. (1991). Legionella infections. In Harrison’s Principles of Internal Medicine, 12th Edition, J.D. Wilson, E. Braunwald, K.J. Isselbacher, R.G. Petersdorf, J.B. Martin, A.S. Fauci, and R.K. Root, eds. (New York: McGrawHill), pp. 634–637. 2. Fang, G.D., Fine, M., Orloff, J., Arisumi, D., Yu, V.L., Kapoor, W., Grayston, J.T., Wang, S.P., Kohler, R., Muder, R.R., et al. (1990). New and emerging etiologies for community-acquired pneumonia with implications for therapy: a prospective multicenter study of 359 cases. Medicine (Baltimore) 69, 307–316. 3. Atlas, R.A. (1999). Legionella: from environmental habitats to disease pathology, detection and control. Environ. Microbiol. 1, 283–293. 4. Harb, O.S., Gao, L.Y., and Abu Kwaik, Y. (2000). From protozoa to mammalian cells: a new paradigm in the life cycle of intracellular bacterial pathogens. Environ. Microbiol. 2, 251–265. 5. Marra, A., and Shuman, H.A. (1992). Genetics of Legionella pneumophila virulence. Annu. Rev. Genet. 26, 51–69. 6. Ciancotto, N., Eisenstein, B.I., Engleberg, N.C., and Shuman, H. (1989). Genetics and molecular pathogenesis of Legionella pneumophila, an intracellular parasite of macrophages. Mol. Biol. Med. 6, 409–424. 7. Swanson, M.S., and Fernandez-Moreia, E. (2002). A microbial strategy to multiply in macrophages: the pregnant pause. Traffic 3, 170–177. 8. Roy, C.R., and Tilney, L.G. (2002). The road less traveled: transport of Legionella to the endoplasmic reticulum. J. Cell Biol. 158, 415–419. 9. Segal, G., Purcell, M., and Shuman, H.A. (1998). Host cell killing and bacterial conjugation require overlapping sets of genes within a 22-kb region of the Legionella pneumophila genome. Proc. Natl. Acad. Sci. USA 95, 1669–1674. 10. Vogel, J.P., Andrews, H.L., Wong, S.K., and Isberg, R.R. (1998). Conjugative transfer by the virulence system of Legionella pneumophila. Science 279, 873–876. 11. Sexton, J.A., and Vogel, J.P. (2002). Type IVB secretion by intracellular pathogens. Traffic 3, 178–185. 12. Nagai, H., Kagan, J.C., Zhu, X., Kahn, R.A., and Roy, C.R. (2002). A bacterial guanine nucleotide exchange factor activates ARF on Legionella phagosomes. Science 295, 679–682. 13. Yamamoto, Y., Klein, T.W., Newton, C.A., Widen, R., and Friedman, H. (1988). Growth of Legionella pneumophila in thioglycollate-elicited peritoneal macrophages from A/J mice. Infect. Immun. 56, 370–375. 14. Yamamoto, Y., Klein, T.W., and Friedman, H. (1991). Legionella pneumophila growth in macrophages from susceptible mice is genetically controlled. Proc. Exp. Biol. Med. 196, 405–409. 15. Yoshida, S.I., Goto, Y., Mizuguchi, Y., Nomoto, K., and Skamene, E. (1991). Genetic control of natural resistance in mouse macro-
Current Biology 36
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33. 34.
phages regulating intracellular Legionella pneumophila multiplication in vitro. Infect. Immun. 59, 428–432. Beckers, M.C., Yoshida, S., Morgan, K., Skamene, E., and Gros, P. (1995). Natural resistance to infection with Legionella pneumophila: chromosomal localization of the Lgn1 susceptibility gene. Mamm. Genome 6, 540–545. Dietrich, W.F., Damron, D.M., Isberg, R.R., Lander, E.S., and Swanson, M.S. (1995). Lgn1, a gene that determines susceptibility to Legionella pneumophila, maps to mouse chromosome 13. Genomics 26, 443–450. Scharf, J.M., Damron, D., Frisella, A., Bruno, S., Beggs, A.H., Kunkel, L.M., and Dietrich, W.F. (1996). The mouse region syntenic for human spinal muscular atrophy lies within the Lgn1 critical interval and contains multiple copies of Naip exon 5. Genomics 38, 405–417. Growney, J.D., Scharf, J.M., Kunkel, L.M., and Dietrich, W.F. (2000). Evolutionary divergence of the mouse and human Lgn1/ SMA repeat structures. Genomics 64, 62–81. Endrizzi, M., Huang, S., Scharf, J.M., Kelter, A.R., Wirth, B., Kunkel, L.M., Miller, W., and Dietrich, W.F. (1999). Comparative sequence analysis of the mouse and human Lgn1/SMA interval. Genomics 60, 137–151. Growney, J.D., and Dietrich, W.F. (2000). High-resolution genetic and physical map of the Lgn1 interval in C57BL/6J implicates Naip2 or Naip5 in Legionella pneumophila pathogenesis. Genome Res. 10, 1158–1171. Endrizzi, M.G., Hadinoto, V., Growney, J.D., Miller, W., and Dietrich, W.F. (2000). Genomic sequence analysis of the mouse Naip gene array. Genome Res. 10, 1095–1102. Huang, S., Scharf, J.M., Growney, J.D., Endrizzi, M.G., and Dietrich, W.F. (1999). The mouse Naip gene cluster on Chromosome 13 encodes several distinct functional transcripts. Mamm. Genome 10, 1032–1035. Maier, J.K., Lahoua, Z., Gendron, N.H., Fetni, R., Johnston, A., Davoodi, J., Rasper, D., Roy, S., Slack, R.S., Nicholson, D.W., et al. (2002). The neuronal apoptosis inhibitory protein is a direct inhibitor of caspases 3 and 7. J. Neurosci. 22, 2035–2043. Watarai, M., Derre, I., Kirby, J., Growney, J.D., Dietrich, W.F., and Isberg, R.R. (2001). Legionella pneumophila is internalized by a macropinocytotic uptake pathway controlled by the Dot/ Icm system and the mouse Lgn1 locus. J. Exp. Med. 194, 1081– 1096. Koonin, E.V., and Aravind, L. (2000). The NACHT family—a new group of predicted NTPases implicated in apoptosis and MHC transcription activation. Trends Biochem. Sci. 25, 223–224. Inohara, N., and Nun˜ez, G. (2001). The NOD: a signaling module that regulates apoptosis and host defense against pathogens. Oncogene 20, 6473–6481. Inohara, N., Ogura, Y., and Nun˜ez, G. (2002). Nods: a family of cytosolic proteins that regulate the host response to pathogens. Curr. Opin. Microbiol. 5, 76–80. Girardin, S.E., Sansonetti, P.J., and Philpott, D.J. (2002). Intracellular vs extracellular recognition of pathogens—common concepts in mammals and flies. Trends Microbiol. 10, 193–199. Lindblad-Toh, K., Winchester, E., Daly, M.J., Wang, D.G., Hirschhorn, J.N, Laviolette, J.P., Ardlie, K., Reich, D.E., Robinson, E., Sklar, P., et al. (2000). Large-scale discovery and genotyping of single-nucleotide polymorphisms in the mouse. Nat. Genet. 24, 381–386. Diez, E., Yaraghi, Z., MacKenzie, A., and Gros, P. (2000). The neuronal apoptosis inhibitory protein (Naip) is expressed in macrophages and is modulated after phagocytosis and during intracellular infection with Legionella pneumophila. J. Immunol. 164, 1470–1477. Summerton, J., and Weller, D. (1997). Morpholino antisense oligomers: design, preparation, and properties. Antisense Nucleic Acid Drug Dev. 7, 187–195. Morcos, P.A. (2001). Achieving efficient delivery of morpholino oligos in cultured cells. Genesis 30, 94–102. Roy, N., Mahadevan, M.S., McLean, M., Shutler, G., Yaraghi, Z., Farahani, R., Baird, S., Besner-Johnston, A., Lefebvre, C., Kang, X., et al. (1995). The gene for neuronal apoptosis inhibitory protein is partially deleted in individuals with spinal muscular atrophy. Cell 80, 167–178.
35. Miller, L.K. (1999). An exegesis of IAPs: salvation and surprises from BIR motifs. Trends Cell Biol. 9, 323–328. 36. Liston, P., Roy, N., Tamai, K., Lefebvre, C., Baird, S., ChertonHorvat, G., Farahani, R., McLean, M., Ikeda, J.E., MacKenzie, A., et al. (1996). Suppression of apoptosis in mammalian cells by NAIP and a related family of IAP genes. Nature 379, 349–353. 37. Gotz, R., Karch, C., Digby, M.R., Troppmair, J., Rapp, U.R., and Sendtner, M. (2000). The neuronal apoptosis inhibitory protein suppresses neuronal differentiation and apoptosis in PC12 cells. Hum. Mol. Genet. 9, 2479–2489. 38. Mercer, E.A., Korhonen, L., Skoglosa, Y., Olsson, P.A., Kukkonen, J.P., and Lindholm, D. (2000). NAIP interacts with hippocalcin and protects neurons against calcium-induced cell death through caspase-3 dependent and independent pathways. EMBO J. 19, 3597–3607. 39. Xu, D.G., Crocker, S.J., Doucet, J.P., St-Jean, M., Tamai, K., Hakim, A.M., Ikeda, J.E., Liston, P., Thompson, C.S., Korneluk, R.G., et al. (1997). Elevation of neuronal expression of NAIP reduces ischemic damage in the rat hippocampus. Nat. Med. 3, 997–1004. 40. Holcik, M., Thompson, C.S., Yaraghi, Z., Lefebvre, C.A., MacKenzie, A.E., and Korneluk, R.G. (2000). The hippocampal neurons of neuronal apoptosis inhibitory protein 1 (NAIP1)-deleted mice display increased vulnerability to kainic acid-induced injury. Proc. Natl. Acad. Sci. USA 97, 2286–2290. 41. Staskawicz, B.J., Mudgett, M.B., Dangl, J.L., and Galan, J.E. (2001). Common and contrasting themes of plant and animal diseases. Science 292, 2285–2289. 42. Zink, S.D., Pedersen, L., Ciancotto, N.P., and Abu-Kwaik, Y. (2002). The Dot/Icm Type IV secretion system of Legionella pneumophila is essential for the induction of apoptosis in human macrophages. Infect. Immun. 70, 1657–1663. 43. Celada, A., Gray, P., Rinderknecht, E., and Schreiber, R. (1984). Evidence for a gamma-interferon receptor that regulates macrophage tumoricidal activity. J. Exp. Med. 160, 55–74. 44. Berger, K.H., and Isberg, R.R. (1993). Two distinct defects in intracellular growth complemented by a single genetic locus in Legionella pneumophila. Mol. Microbiol. 7, 7–19. 45. Byrne, B., and Swanson, M.S. (1998). Expression of Legionella pneumophila virulence traits in response to growth conditions. Infect. Immun. 66, 3029–3034. 46. Hammer, B.K., Tateda, E.S., and Swanson, M.S. (2002). A twocomponent regulator induces the transmission phenotype of stationary-phase Legionella pneumophila. Mol. Microbiol. 44, 107–118. 47. Brown, I., and Mackey, K. (1997). Analysis of RNA by Northern and Slot blot hybridization. In Current Protocols in Molecular Biology, F.M. Ausubel, R. Brent, R.E. Kingston, D.D. Moore, J.G. Seidman, J.A. Smith, and K. Struhl, eds. (New York, New York: John Wiley and Sons), pp. 4.9.1–4.9.16. Note Added in Proof We have become aware of data from Diez, Lee, Gauthier, Yaraghi, Tremblay, Vidal, and Gros that is generally consistent with what we report here. Their paper is in press at Nature Genetics (doi: 10.1038/ ng1065).