Genomic RNAi screening in Drosophila S2 cells: What have we learned about host-pathogen interactions?

Abstract

The détente between pathogen and host has been of keen interest to researchers in spite of being exceedingly difficult to probe. Recently, new RNA interference (RNAi) technologies, in particular in Drosophila tissue culture cells, have made it possible to interrogate the genetics of host organisms rapidly, with nearly complete genomic coverage and high fidelity. Therefore, it is not surprising that the applications of RNAi to the study of host-pathogen interactions were amongst the first to be published, and have already revealed many new insights into the hosts’ role in infection. This review will highlight the application of RNAi screening to pathogen-host interactions in Drosophila cells and will reveal some of the lessons learned from this approach.

Introduction

All organisms are plagued by infectious pathogens including bacteria, fungi, parasites and viruses. In order to successfully invade and survive within their hosts, pathogens must exploit cellular pathways and establish their particular niche, as well as evade detection by the host-encoded immune system. Pathogens must also remodel and subvert host pathways to facilitate their own survival at the expense of the host. Viruses, as obligate intracellular pathogens with only a limited genome size, are even more dependent on host encoded factors for their replication cycle. The identification of host encoded factors involved in infection, intracellular replication and pathogenesis has historically been difficult due to the dearth of pathogen-host systems amenable to genetic screening, as well as the lack of a suitable in vivo model to test the interactions between the pathogen and host. To rapidly identify host encoded factors that affect microbial pathogenesis, recent studies have taken advantage of the genetically tractable Drosophila model system. Practical aspects of the Drosophila system enable unbiased approaches to the identification of host-encoded factors that impact the pathogen-host interface both at the cellular and organismal level.

Over the past several decades, studies in Drosophila have been central to our increasing understanding of various fundamental biological processes [1–3]. Drosophila have conserved developmental and cell biological processes making then a fruitful model for mammalian development and disease [4]. Moreover, many of the classic signal transduction systems were identified first in Drosophila using forward genetic screens. Because the Drosophila genome is compact with relatively low redundancy, single mutants are likely to reveal phenotypes of interest, in contrast to mammals where redundant gene families can make genetic analysis more complex. Recently, many insights into innate immunity have been obtained from both forward and reverse genetic studies in Drosophila [5–9]. Drosophila has no acquired immune system, and so relies exclusively on the innate immune system. The impact of Drosophila studies on our general understanding of innate immunity is underscored by the discovery and characterization of the single-pass, transmembrane-receptor Toll and its mammalian homologues, the Toll-like receptors, which play critical roles in innate immune responses to pathogens in both insects and mammals. Because of this conservation, and the ease of performing large-scale genetic screens, Drosophila has been instrumental in uncovering new connections between cellular factors, innate immunity and pathogens.

Host-Pathogen Interactions and Innate Immunity

The two main host-encoded processes that should be considered when discussing the interactions between pathogens and hosts are the host-encoded immune system used to combat the pathogen and the cellular factors used and/or manipulated by the pathogen for its survival. It is essential to understand both pathogen virulence, and the host immune response, and both can be studied in the Drosophila system. The Drosophila innate immune system involves both cellular and humoral responses, with both cell intrinsic and extrinsic components. The cellular arm of the innate immune system involves both specialized immune cells which recognize and respond to invaders as well as intrinsic immune mechanisms initiated within an infected cell. Drosophila have specialized professional phagocytic cells with properties similar to mammalian macrophages. These cells efficiently ingest particles and microbes and upon stimulation produce cytokines that modulate the downstream immune response. Although these specialized immune cells are an essential component of the innate immune response [9], it has become increasingly clear that all infected cells have some intrinsic capacity to respond to infections. A classical example is the production of interferons which can be produced by most virally-infected mammalian cell-types. In Drosophila these cell-intrinsic responses include apoptosis, autophagy and RNAi.

In Drosophila, the humoral arm of the innate system is characterized by the production of secreted factors that are produced by the activation of a variety of conserved signaling pathways to control infection. These cascades are activated downstream of pathogen recognition. Rather than specifically recognizing individual antigens unique to each pathogen as B- and T-cell Receptors do, the innate system responds to pathogen-associated molecular patterns (PAMPs)—motifs that are essential functional components of microorganisms, and are thus broadly expressed across classes of pathogens [10]. Drosophila uses many of the same recognition receptors and signaling pathways as mammals to initiate these anti-microbial cascades. For example, scavenger receptors, TLRs as NFkB signaling are essential components of both mammalian and insect responses.

The other essential component of the host-pathogen interface is the factors that are subverted by the microbe to allow for its survival and replication. Drosophila uses highly conserved cell biological processes, but has less functional redundancy than mammals and other vertebrates. Thus, the use of genetic loss-of-function approaches to uncover the role of basic cellular functions in microbial pathogenesis is potentially more fruitful in Drosophila than in mammals. This is particularly important for intracellular pathogens and allows for the elucidation of pathways involved in intracellular growth of the invader. Pathogens are efficient cell biologists, reprogramming pathways and thereby teaching us both about the pathogen’s requirements and about essential processes at work in uninfected cells (i.e. splicing [11], transcription factors [12], etc. all discovered via studying virus infection). Importantly, because some host factors are essential for microbial growth, they may prove to be useful targets for anti-microbial therapeutics.

Drosophila Cell Culture Screening platform

While some aspects of the host-pathogen interface require organismal study, others can be studied at the cellular level. Cellular pathogen recognition and the downstream signaling pathways are readily assayed in Drosophila cells. Moreover, tissue-specific responses can be probed using cell lines derived from disparate cell types. For example, hemocyte-derived cells such as Schneider line 2 (S2) cells as well as Kc167 cells are responsive to PAMPS and can thus be used to study this aspect of immunity. Moreover, these cell lines are highly phagocytic and therefore have been used to study pathogen uptake and clearance. Intrinsic innate immune mechanisms including anti-viral RNAi, autophagy, and programmed cell death can also be studied in cultured cells. For intracellular pathogens, host factors required for pathogen replication and survival can be identified. The use of cell culture also allows the role of proteins and processes that are essential early in development to be assessed for an independent role in immunity.

Importantly, Drosophila cells can be infected by a wide variety of pathogens: intracellular and extracellular, fungal, bacterial and viral. For example, RNAi has been used to study factors involved in host interactions with pathogens including: Escherichia coli, Staphylococcus aureus, Mycobacterium sp.,Legionella pneumophila, Candida albicans, Chlamydia sp., Listeria monocytogenes, and Drosophila C virus (see below). Some of these are human pathogens while others are insect pathogens. The lessons learned from the study of these pathogens in Drosophila have revealed much about host-pathogen interactions and innate immunity in mammals. Table 1 presents an overview of the screens described in this review.

Table 1.

Host-Pathogen screens in Drosophila cells.This table describes the technical parameters of the screens described including: screen type, cell type, library used, reporter used, read-out, analysis and genes identified.

TABLE 1	Screen	Reference	Cell Line	Library	Inducer	Reporter	Mode	Analysis	Positive Regulators	Negative Regulators
Signaling	Imd Signaling	Foley et al	S2	7,216 conserved genes	LPS	diptericin-beta-glactosidase	microscope	visual	49	72
	Imd Signaling	Kleino et al	S2	6,700 hemocyte expressed ESTs	heat killed E.Coli	attacin-luciferase vs.control	plate-reader	statistical	4	nd
	JAK/Stat Signaling	Muller et al.	Kc176	genome-wide library	Upd Overexpression	Draf 10X STAT sites-luciferase vs. control	plate-reader	statistical	67	24
	JAK/Stat Signaling	Baeg et al.	S2	genome-wide library	Endogenous Activity	SOCS36E 10X STAT sites-luciferase vs. control	plate-reader	statistical	29	92
Phagocytosis	General Phagocytosis	Ramet et al.	S2	1,000 dsRNAs S2expressed library		GFP-E. coli, GFP-S.aureus	FACS	Fold	34	nd
	Serpent-dependent	Kocks, et al.	S2	45 differentially expressed genes		GFP-E. coli, GFP-S.aureus	FACS	Fold	1	nd
	Candida albicans	Stroschein-Stevenson, et al.	S2	7,216 conserved genes		GFP-C. albicans/anti-C.albicans	microscope	visual	184	nd
Live Bacterial	Mycobacterium Fortuitum	Philips, et al.	S2	genome-wide library		Induced GFP-M. fortuitum	microscope	statistical	85	nd
	Mycobacterium Marinum	Koo, et al.	S2	1000 genes		GFP-M. marinum	FACS	fold	nd	2
	Listeria Monocytogenes	Agaisse, et al.	S2	genome-wide library		GFP-L. monocytogenes	microscope	visual	167*	48
	Listeria Monocytogenes	Cheng, et al.	S2	7,216 conserved genes	(wt, LLO-minus, LLO-toxic)	anti-Listeria	microscope	visual	89,29,8	nd
	Legionella pneumophila	Dorer, et al.	Kc167	10dsRNAx73dsRN As trafficking		GFP-L. pneumophila	microscope	visual	12 singles, new combos	nd
	Chlamydia caviae	Derre, et al	S2	genome-wide library		anti-C. caviae	microscope	visual	54	nd
Virus	Drosophila C virus	Cherry, et al.	S2	genome-wide library		anti-Drosophila C virus	microscope	visual	110	nd

RNA interference in Drosophila cells is relatively efficient and straight-forward, and thus has been extensively used to study many aspects of biology that can be modeled in cells in vitro [13]. Each step in the process must be carefully selected and optimized. A flow chart of the process is schematized in Figure 1. Bathing of many Drosophila cell lines with naked double-stranded RNA (dsRNA) 200–500bp in length results in uptake, processing and gene knock-down in greater than 95% of the cells. This allows for the simple generation of arrayed RNAi libraries which can be used to probe for factors that, when depleted, impact a phenotype of interest. A variety of dsRNA libraries have been used, and are typically employed in a 384 well format allowing for high-throughput screening [14]. Different libraries have been used to dissect host factor requirements, ranging from focused gene-sets to genome-wide libraries (Figure 1). Most assays allow 3–5 days for the RNAi to take effect, yielding hypomorphic loss of function phenotypes. Assays must be designed to give robust and reproducible results; optimization for the 384-well format is particular important. A number of different read-outs have been used ranging from microscope-based screens for microbial growth, to luciferase-based reporter gene assays for signaling pathways (See descriptions below). Each assay has been optimized to query specific steps in the biology of host-pathogen interactions. Knockdown of a given gene can either decrease or increase the read-out in question and potential positive candidates can be identified using statistical cut-offs or visual observation. Subsequent replication is used to filter out false positives and secondary assays are used to more fully characterize the role of the identified candidates in the biological process under study.

Figure 1.

RNAi screening workflow.This flow chart describes the critical decisions and optimizations that must be considered when developing and reading high-throughput cell-based RNAi screens.

Anti-Microbial Signaling Screens

The systemic response to bacterial and fungal infection in Drosophila is mediated by two major signaling pathways that lead to the production of anti-microbial factors: the Toll pathway, which primarily responds to Gram-positive and fungal pathogens and the Immune deficiency (IMD) pathway, which largely responds to Gram-negative bacteria [6,7,15]. Another pathway that has known signaling roles in innate immunity and in particular in anti-viral immunity is the JAK/Stat pathway [16–18]. Three large-scale RNAi screens have been performed, which probed the IMD signaling pathway in response to peptidoglycan [19–21] (Table 1). The first screen identified three classes of genes: positive regulators (49 DDRi genes), and two types of negative regulators: those that constitutively activated the NF-kB reporter in the absence of LPS (26 CDRi genes) and those that enhanced the responsiveness to peptidoglycan induction (46 EDRi genes) [19]. The other two screens identified additional (but fewer) genes, including two new positive regulators of the signaling pathway, Iap2 and TAB [20,21]. Iap2 has been subsequently shown to be required for an immune response to gram-negative bacteria in vivo [22].

Two genome-wide screens have been performed to identify immune-associated components and mediators of the JAK/Stat signaling pathway using luciferase reporter systems [23,24] (Table 1). Muller et al identified 67 positive regulators and 24 negative regulators of the pathway while Baeg et al identified 29 positive regulators and 92 negative regulators. While the overlap of these screens is negligible, both studies focused on a novel phosphatase, PTP61F, that has been shown to be a bona fide negative regulator of signaling in Drosophila both in vitro and in vivo. The differences in cell type (Kc167 vs S2) or assay conditions (overexpressed cytokine vs endogenous) may explain the lack of concordance between the screens.

Microbial Recognition and Survival Screens

Recognition/Phagocytosis/Uptake

Phagocytosis is required for the uptake and clearance of intracellular microbes [25]. Phagocytosis is initiated by the ligation of cell-surface receptors that either directly bind to the particle or to opsonins that are deposited on the particle’s surface.Professional immune cells such as hemocytes or hemocyte-derived S2 and Kc167 cells are particularly efficient at mediating uptake of a variety of microbes (and other particles), and have been used to screen for cellular factors required for phagocytosis.

To identify cellular mediators of phagocytosis of E. coli and S. aureus, 1000 randomly generated dsRNAs derived from an S2 cell cDNA library were screened using a FACS-based assay that monitored the uptake of FITC-labelled bacteria [26] (Table 1). 34 genes were required for efficient phagocytosis, but were not cytotoxic. Among these genes were four transcription factors, one of which encodes the master regulator of hematopoesis in flies, the GATA-factor Serpent. Another gene, PGRP-LC, was required for efficient uptake of E. coli but not S. aureus in this system. Further studies showed that PGRP-LC is the peptidoglycan receptor for Gram-negative bacteria which in turn activates the IMD signaling pathway. Building on the identification of Serpent as an important factor required for efficient phagocytosis in S2 cells, Kocks et al performed expression profiling and identified 45 genes that were down-regulated by depletion of Serpent [27] (Table 1). Among those genes was SR-C1, a phagocytic scavenger receptor [28]. RNAi against these Serpent-dependent genes also identified a novel phagocytic receptor, Eater. This was the first member of a class of EGF-like- repeat-containing phagocytic receptors that act on a broad range of bacteria. Another member of this family, Nimrod, was also shown to mediate bacterial uptake and adhesion [29].

Candida albicans is a major nocosomial fungal pathogen of humans [30]. Using phagocytic S2 cells infected with GFP-expressing Candida, Stroschein-Stevenson et al. screened a library of 7,000 conserved genes for factors required for engulfment [31](Table 1). They identified 184 genes that were required for efficient uptake. In particular, they identified Mcr (Macroglobulin-related protein), which is a member of a family of thioester-containing proteins, related to complement factors, called Teps, which act as opsonins. Mcr was shown to specifically opsonize Candida while TepII was required for efficient uptake of E.coli and TepIII for S. aureus. This is the first functional evidence for a specific role of Tep proteins in Drosophila opsinization.

Another screen following Mycobacteria fortuitum infection identified 54 genes, including a novel class B scavenger receptor peste [32](Table 1). Peste was dispensible for E. coli and S. aureus uptake in S2 cells but was required for L. monocytogenes infection in addition to M. fortuitum. [33]. Although Peste was dispensible in S2 cells for uptake of E. coli, expression of Peste in non-phagocytic human cells was sufficient to support uptake of E. coli or M. fortuitum, suggesting that there may be a larger number of functionally redundant receptors for E. coli than for other bacteria such as M. fortuitum and L. monocytogenes.

Altogether, the results from these screens demonstrate that different classes of entry receptors mediate uptake of different microbes, sometimes with overlapping and redundant specificities. These RNAi screen have identified many of these receptors, suggesting that this aspect of immunity is particularly amenable to such approaches. The RNAi screens listed above, along with additional screens using live bacterial infection, also identified genes required for the downstream steps of phagocytosis including actin remodeling and membrane reorganization (Table 1). Many genes identified in Drosophila were previously known to be required for phagocytosis in mammalian systems, validating this approach. For example, an important role for actin remodeling, including the Arp2/3 complex, and vesicular transport genes was uncovered in most of these screens. Taking advantage of this knowledge, Pielage, et al screened ~80 genes known to regulate the actin cytoskeleton to identify additional factors essential for Psudomonas aeruginosa entry [34]. This screen revealed an invasion pathway that is dependent upon the Abl signaling pathway both in Drosophila and mammalian cells.

In addition to known factors, RNAi screens were the first to identify the coatamer complexes (COPI and COPII) in uptake [26,31–33,35,36]. The COPI and COPII complexes are required for vesicular trafficking between the ER and golgi [37], are found to associate with the phagosome, and are required in S2 cells for the uptake of a wide variety of pathogens, although the mechanism of this requirement has not yet been established. Another novel and general finding of the RNAi screens was the requirement for the exocyst in phagocytosis [26,32,33,35,36]. This complex plays essential roles in secretory vesicle targeting and docking at the plasma membrane for exocytosis [38]. And again, these factors associate with the phagosome, and were identified as essential components of the phagocytic machinery in a number of RNAi screens. Other cell biological and metabolic pathways were identified in a number of screens including fatty acid metabolism (e.g., [17,32,39]). However, further study is needed to determine whether these pathways are directly or indirectly involved in membrane dynamics of phagocytosis.

Bacterial Survival and Growth

Several other RNAi screens, using live intracellular bacteria, were designed not only to monitor uptake into host cells, but also intracellular replication and survival of the microbe. These screens have revealed additional cellular pathways and processes that impact these steps in the pathogen lifecycle. Each intracellular microbe subverts a unique cellular compartment to its own advantage. For example, Listeria monocytogenes, a gram positive bacterium, exits from a phagosome to replicate in the cytoplasm, while Legionella pneumophila and Mycobacteria species modify the phagosome, preventing phagosome-lysosome fusion [40]. Given the unique niche of each microbe, these screens have identified novel cellular factors thought to control intracellular survival, many of which are unique to each pathogen screened (Table 1).

Two independent screens were performed for factors involved in L. monocytogenes infection [33,35]. Agaisse et al identified 160 genes that blocked infection, of which a large number are thought to be involved in vacuolar escape, although this was not directly tested. Cheng et al performed three screens. Screening with wild type L. monocytogenes they identified 89 genes required for infection; surprisingly, only a small subset of candidate genes overlapped with those found by Agaisse et al. In a parallel screen, they also assayed for factors that bypass the requirement for listeriolysin O (LLO), the bacterially-encoded enzyme required for exit from the phagosome, using an LLO-deficient L. monocytogenes strain. They identified 29 genes that specifically allow escape of this strain. In a third screen, the authors used a strain of Listeria carrying a mutant form of LLO with enhanced toxicity. This allowed the identification of 8 factors required for LLO toxicity. Among these genes were proteosomal components, which were also identified as required for infection in the Agaisse screen. Further experiments will be required to more fully clarify the roles of these factors in Listeria biology. Agaisse, et al also identified 48 genes that resulted in increased bacterial loads when targeted by RNAi, amongst which were a large number of cell cycle regulators. While this phenotype may be due to increased uptake, or increased bacterial growth, it is also possible that this group of genes were identified because their depletion was cytotoxic, leading to decreased cell number resulting in an increased multiplicity of infection (Table 1)

Mycobacterial species block phagosome-lysosome fusion in macrophages allowing for their replication and survival in that niche [41]. Two pathogenic species, M. fortuitum and M. marinum can infect and replicate in S2 cells. In addition the non-pathogenic strain M. smegmatis can invade but undergoes an abortive infection in S2 cells [32,42,43]. A genome-wide screen for factors required for M. fortuitum infection identified 32 genes that were required for survival but not uptake [32]. Amongst these genes were the ESCRT machinery and chromatin factors. In the absence of ESCRT (endosomal sorting complex required for transport) function, the phagosomal environment was altered resulting in differential bacterial growth whereby the non-pathogenic M. smegmatis strain is more successful, and the pathogenic M. fortuitum strain has altered gene expression profiles [44]. The ESCRT machinery was also identified as required for intracellular growth of Listeria suggesting that alterations in the phagosomal compartment may have more pleotropic effects on intracellular bacterial survival [33]. A screen of 1000 dsRNAs to identify host factors that control M. marinum infection in S2 cells identified 2 genes that restrict infection [43]. Further studies demonstrated that one factor identified in the screen, lysosomal beta-hexosaminidase, restricts bacterial growth both in insect cells and mammalian macrophages. The mycobactericidal effect is specific to M. marinum, as L. monocytogenes and S. typhimurium were unaffected, and may have significant effects in limiting the initial mycobacterial infection of macrophages.

L. pneumophila, the gram negative agent of Legionnaires disease, is an intracellular pathogen that creates an intracellular vacuolar replication niche [45]. Much is known about the bacterial regulators of this compartment while less is known about the host factors required for its generation. Dorer et al used a visual assay that monitored the size and intensity of Legionella-GFP vacuoles along with a combination of single and double knock-downs of 73 trafficking genes in Kc167 cells to identify an essential network of early secretory genes required for bacterial replication [46]. Redundant activities required the use of double mutants (combinations of dsRNAs) which would be difficult to perform in mammalian systems, again pointing to the utility of Drosophila to uncover important cellular factors subverted by pathogens.

Chlamydia spp. are gram negative obligate intracellular bacteria of which three species cause disease in humans [47]. Whereas the human pathogen C. trichomatis can initiate infection in S2 cells, it cannot complete its lifecycle [48]. A focused screen for early events in the lifecycle included actin regulators and led to the identification of 28 factors including the Abl kinase and components of the PDGFR signaling pathway as essential for early steps in the C. trichomatis infection both in insect and mammalian cells [49]. C. caviae, a guinea pig pathogen, productively infects S2 cells and was used to screen a genome-wide library, leading to the identification 31 genes that are specifically required for C. caviae replication [39]. Among these genes was the Tom complex, which regulates mitochondrial protein import [50] and is required both in Drosophila and mammalian cells for C. caviae but not C. trachomatis infection.

Viral screens

Viruses have been widely used as tools to probe cell biology. This is because they efficiently hijack and subvert cellular processes to remodel cells into viral production factories. This is accomplished with only a limited number of proteins-(actually, many viruses encode only 4 or 5 gene products). Importantly, insects are infected by a wide variety of viruses, some of which are insect-specific, but others also infect humans (arthropod-borne). Recent studies have demonstrated that many human viruses can productively infect Drosophila S2 cells including Vesicular Stomatitis Virus, Sindbis Virus, Rift Valley Fever Virus, Dengue virus, West Nile Virus ([51,52] and our unpublished data). The insect-specific viruses, Flock House Virus and Drosophila C virus (DCV) have also been extensively studied in fruit flies [51–56]. The first genome-wide RNAi screen for host factors that impact pathogen replication was performed using DCV infection of S2 cells [53]. DCV resembles human picornaviruses such as poliovirus and is a natural pathogen of fruit flies. The screen identified 110 genes that, when lost, block viral replication (Table 1). Two major categories of genes were identified in the screen. First, it was found that attenuation of the translation machinery blocked infection by DCV in Drosophila cells and in adult flies, as well as poliovirus replication in human cells. An unrelated virus, without an internal ribosome entry site, was not affected by depletion of the translation machinery. This demonstrates the power of the Drosophila system to uncover essential factors in human cells, and that the cross comparisons between different pathogens reveals specificity in the strategies used by different pathogens to exploit and subvert the host. The COPI coatamer was also identified as a cellular machine that is required for the generation of a vesicular compartment necessary for DCV replication as well as poliovirus replication in human cells [54]. Interestingly, the COPI coatamer blocks Vesicular Stomatitis Virus replication (VSV) although this virus does not remodel the membranes within the cell as picornaviruses do (Unpublished data). Therefore, it is likely to be required for a different aspect of VSV replication. The COPI coatamer was also found to be required for phagocytosis, suggesting that some cellular machines may play multiple different roles during infection, depending on the pathogen. Therefore, more mechanistic studies are necessary to assign a function to genes identified in these screens.

Future Directions

The recent development of genome-wide RNAi screening for host factors involved in microbial infection has identified hundreds of cellular genes that impact infection and opened up new avenues of research. It is clear that screening methodologies and technologies are in their infancy, and that these methodologies will continue to improve in efficacy. As the assays become more robust and as the downstream validation is streamlined, using richer more quantitative and statistically robust metrics, we will undoubtedly learn more about the host factors required for infection of particular pathogens. Further validation of candidate genes not only in cultured Drosophila cells but in flies and in mammalian systems is becoming simpler and will further our understanding of the biology. This will be facilitated by the availability of genome-wide libraries of transgenic flies that carry inverted repeat constructs against each gene allowing for directed RNAi in vivo in the animal, and the increasing specificity and robustness of siRNA methodologies. Mammalian RNAi screens are also beginning to contribute to our understanding of host-pathogen interactions [57–59], although technical limitations associated with siRNA and shRNA biology still make the Drosophila system ideal for RNAi screening. The comparisons between the host factors involved in pathogenesis in these disparate hosts will also inform the biology. Importantly, the more pathogens that are queried, the deeper our understanding will become of the generalities and specializations used by disparate microbes for infection and pathogenesis and the innate and intrinsic mechanisms used by hosts to inhibit infection.