Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Pathways to zoonotic spillover

Pathways to zoonotic spillover PERSPECTIVES Although many recent articles have OPINION examined the fields of zoonoses or emerging 2,3,10–15 pathogens , a synthetic mechanistic understanding of animal-to-human 14,16 transmission is lacking . Much attention has been dedicated to the characterization Raina K. Plowright, Colin R. Parrish, Hamish McCallum, Peter J. Hudson, 3,11,12,15 of emerging infections ; for example, Albert I. Ko, Andrea L. Graham and James O. Lloyd-Smith the high frequency of zoonoses among 3,12 emerging infections , their socio-economic, Abstract Zoonotic spillover, which is the transmission of a pathogen from environmental and ecological 2,13,17,18 a vertebrate animal to a human, presents a global public health burden but is a drivers , and their phylogenetic and poorly understood phenomenon. Zoonotic spillover requires several factors to geographical distribution . Similarly, the phases of zoonotic emergence in the align, including the ecological, epidemiological and behavioural determinants of 11,14,18 human population , adaptation and pathogen exposure, and the within-human factors that affect susceptibility to 10,11,19 compatibility of zoonoses in humans , infection. In this Opinion article, we propose a synthetic framework for and approaches to modelling the animal-to-human transmission that integrates the relevant mechanisms. This 14,16 transmission of zoonoses , have also framework reveals that all zoonotic pathogens must overcome a hierarchical series been addressed in the literature. However, a comprehensive understanding of the of barriers to cause spillover infections in humans. Understanding how these processes that enable a pathogen from a barriers are functionally and quantitatively linked, and how they interact in space vertebrate animal to establish infection and time, will substantially improve our ability to predict or prevent spillover in a human, and how these processes are events. This work provides a foundation for transdisciplinary investigation of hierarchically, functionally and quantitatively spillover and synthetic theory on zoonotic transmission. linked, remains a fundamental deficit in 14,16 research on zoonoses . In this Opinion The phenomenon of cross-species spillover host distribution, pathogen prevalence and article, we present a mechanistic structure is the defining characteristic of pathogens pathogen release from the reservoir host, that integrates the determinants of spillover that transmit from vertebrate animals to followed by pathogen survival, development and the interactions among them (FIG. 1). humans (zoonoses). The public health and dissemination outside of the reservoir However, we do not address broader burden that is presented by zoonoses hosts. Second, human and vector determinants of pathogen emergence or includes outbreaks of pathogens such as behaviour determine pathogen exposure; factors that affect disease severity or onward Ebola virus, influenza A virus (H1N1) specifically, the likelihood, route and dose of transmission in humans. pdm09 and Middle East respiratory exposure. Third, genetic, physiological and Although many of the individual syndrome coronavirus (MERS-CoV), immunological attributes of the recipient determinants of spillover are subjects of as well as the ongoing transmission of human host, together with the dose and intensive study, each is usually addressed in endemic pathogens, such as Salmonella route of exposure, affect the probability and isolation in a specialized discipline (FIG. 2). spp., Leptospira spp., Trypanosoma spp., severity of infection. Accordingly, the better-characterized 1–6 Mycobacterium spp. and West Nile virus . Each phase presents multiple barriers factors become the focus of public health Spillover transmission is promoted by to the flow of a pathogen from a reservoir interventions. For example, reservoir hosts successive processes that enable an animal host to a recipient host. Spillover requires or vectors are often targeted for control pathogen to establish infection in a human. the pathogen to pass every barrier and before the concatenation and relative The probability of zoonotic spillover is thus can only occur when gaps align influence of processes that lead to spillover determined by interactions among several in each successive barrier within an are understood, which sometimes leads factors, including disease dynamics in appropriate window in space and time to inefficient or even counterproductive the reservoir host, pathogen exposure (FIG. 2). Consequently, zoonotic spillover is a interventions . In other cases, multiple and the within-human factors that affect relatively rare event, and although humans mechanisms are aggregated in analyses that susceptibility to infections. These factors are continually exposed to many potentially obscure the interactions or heterogeneities can be partitioned into three functional infectious pathogens that are derived from that drive risk. Although the aggregation of phases that describe all major routes of other species, most of these microorganisms mechanisms may be appropriate at times, 7–10 transmission (FIG. 1). In the first phase, cannot infect or cause disease in humans . identifying discrete mechanisms and how the amount of pathogen available to the This Opinion article focuses on spillover they interact to drive spillover is essential to human host at a given point in space and transmission, strictly defined as the processes recognize the assumptions that are implicit time, known as the pathogen pressure, is that enable a pathogen from a vertebrate in simpler models, and to clarify which determined by interactions among reservoir animal to establish infection in a human. processes must be modelled explicitly and 502 | AUGUST 2017 | VOLUME 15 w w w.nature.com/nrmicro © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES which can be combined. For example, • Reservoir host distribution does assessment of the risk of acquiring a Distribution and • Reservoir host density intensity of infection zoonotic infection require the measurement • Prevalence of infection in reservoir hosts • Intensity of infection of the pathogen burden carried by individual reservoir hosts, or is it sufficient to estimate the cumulative abundance of a pathogen Excretion Slaughter Vector borne in the environment over time? This is a key question for pathogens such as Leptospira Pathogen release Shedding rate Harvest rate Biting rate (vector– from reservoir host reservoir host) interrogans, Giardia spp., and Escherichia coli O157, and the answer may depend on modes Pathogen survival, Pathogen survival Pathogen survival Vector survival of contact and dose–response relationships development and and movement and transport of and movement dissemination meat in humans (see below). Models that integrate data from experiments, the field Pathogen pressure and epidemiological studies, even if only Human exposure Human behavior Butchering, Biting rate partially parameterized, may be necessary to pathogen that leads to contact preparation and (vector–human) to make such determinations. with pathogen eating We describe how pathogens overcome a series of barriers to pass from reservoir hosts to humans. Crucially, nonlinear interactions Dose and route of exposure among the barriers create bottlenecks in the flow of a pathogen between species. Such bottlenecks provide opportunities for public health interventions that could lead to substantial reductions in the risk of spillover. Alternatively, changing environmental • Structural barriers Host susceptibility • Innate immune response and molecular compatibility or social conditions can alleviate these • Replication and dissemination cycles completed bottlenecks, which can cause surges in Probability of infection spillover infections. Our framework provides the foundation for operational models that are required for quantitative evidence-based Figure 1 | Pathways to spillover. The risk of spillover is determined by a series of processes that link Nature Reviews | Microbiology the ecological dynamics of infection in reservoir hosts, the microbiological and vector determinants risk analysis, preparedness, surveillance of survival and dissemination outside of reservoir hosts, the epidemiological and behavioural deter- and control. minants of exposure, and the within-host biological factors that shape the susceptibility of recipient hosts. The distribution and intensity of infection in reservoir hosts, followed by pathogen release, Barriers to spillover movement, survival and possible development to infectious stage, determine the pathogen pressure, The probability of spillover is determined by which is defined as the amount of pathogen available to the recipient host at a given point in space the interactions among the barriers and the and time. Pathogen pressure then interacts with the behaviour of the recipient host (and vector for associated bottlenecks that might prevent vector-borne pathogens) to determine the likelihood, dose and route of exposure. A series of within- cross-species transmission. Many of these host barriers then determine host susceptibility, and, therefore, the probability and severity of interactions are nonlinear and dynamic in infection for a given pathogen dose. space and time. Pathogen pressure. The series of processes The first set is the natural history of infection The mode of pathogen release from that culminate in pathogen pressure (the in hosts, which includes the duration, reservoir hosts determines the major routes amount of a pathogen that is available to intensity and severity of infection and the of transmission. Pathogens may be released in humans at a given point in time and space) level of shedding. Second, the movement host excretions, through slaughter or through includes pathogen dynamics in reservoir and behaviour of hosts affect contact and the an arthropod vector (FIG. 1). The probability hosts, pathogen release from reservoir hosts, likelihood of exposure within and between of a pathogen being released from a reservoir and pathogen survival or dispersal outside of species. These factors interact with the host is affected by its presence and viability in reservoir hosts. abundance, density, demographic turnover, relevant tissues, such as the blood for many Pathogen dynamics in reservoir hosts spatial distribution and physiological state of vector-borne pathogens, tissues contacted can be represented as three variables that hosts to determine the efficiency of spread . or consumed during butchering and eating determine the distribution and intensity Collectively, these processes determine how for some food-borne pathogens, and tissues of infection in time and space: the density of the pathogen is distributed across reservoir through which external shedding occurs for reservoir hosts, the prevalence of infection host populations. Such pathogen distribution direct or environmental routes. For example, among reservoir hosts, and the average can be highly variable (for example, pulses the viral load and excretion rates in the intensity of infection in an infected reservoir of Sin Nombre virus infections in deer salivary glands are key determinants for the host in time and space (Supplementary mice (Peromyscus maniculatus) populations transmission of rabies virus from carnivores, information S1 (box)). Many ecological and in response to climate-driven increases in whereas viral loads in the intestinal and physiological factors influence these variables population density) , or stable (as illustrated respiratory tracts affect the transmission 24–26 in communities of reservoir animals; by Mycobacterium bovis infections in of avian influenza virus from poultry . however, two sets of factors are dominant. populations of livestock and wildlife) . Likewise, the release of pathogenic Leptospira NATURE REVIEWS | MICROBIOLOGY VOLUME 15 | AUGUST 2017 | 503 © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES spp. from animal hosts requires colonization Following the release of a pathogen from information S1 (box)). Spillover of pathogens of the renal tubules . The excreted pathogen its reservoir host, the opportunity for spillover that have short survival times (for example, load depends on the quantity of leptospires transmission is influenced by the duration influenza A virus when transmitted through 28 33,34 that effectively colonize the tubules , the of pathogen survival outside of its host, the the respiratory route) may require close rate of release and the urinary output of the extent of spatial dispersal through passive interactions between reservoir and recipient host . Moreover, the pathogen undergoes transport (for example, through water, on hosts. Consequently, spillover patterns in several changes in its lipopolysaccharide fomites or in the air), and possible pathogen recipient hosts correspond to the prevalence content and proteome during colonization reproduction or obligate developmental stages patterns in reservoir hosts. By contrast, 30,31 and shedding in the urine , which suggests outside of the primary host (for example, if pathogens survive for sufficient periods that priming in the renal milieu is required Yersinia pestis, the causative agent of plague, of time outside of their reservoir hosts, they to adapt for survival and infectivity in the must multiply within flea vectors before it can may be dispersed beyond the home range of external environment. The rate of pathogen be transmitted to humans ). These processes the host through fomites or environmental release is a crucial determinant of spillover can be represented as the probability that the transport. In this case, the release of a risk, and care must be taken to appropriately pathogen (shed, harvested or colonized in a pathogen from its reservoir host and human formulate models that represent the rate of vector) survives and is infectious at a given exposure to the pathogen may become release for each route of transmission (BOX 1; point in time, and is dispersed or transported disconnected in space and time. An example Supplementary information S1 (box)). to a particular location (Supplementary is the spread of aerosolized Coxiella burnetii a b c Reservoir host distribution Animal ecology, population biology, biogeography, behavioural ecology, landscape ecology, agricultural sciences Reservoir host density Pathogen prevalence Disease ecology, animal epidemiology, infectious disease dynamics, immunology, Infection intensity microbiology, veterinary medicine Pathogen release from reservoir host Microbiology, disease ecology, vector ecology, epidemiology, spatial ecology, Pathogen survival and spread infectious disease dynamics Time Scenario 1 Human epidemiology, medical anthropology, Scenario 2 vector ecology, social sciences, behavioural Threshold Human exposure ecology, infectious disease dynamics Structural barriers Microbiology, innate and adaptive immunology, cell biology of pathogen–host Linear interactions, pathology, genetics, Innate immune response and molecular compatibility Sigmoidal evolutionary biology Threshold Replication and dissemination cycles completed Dose Spillover Figure 2 Barriers to spillover and dose–response relationships. both scenarios, the mean dose over the time interval is the same. Bottom Nature Reviews | Microbiology a | Determinants of spillover are being studied by researchers in many disci - panel: the likelihood that this dose will translate into infection depends on the plines. b | A pathogen must overcome a series of barriers to transmit from one functional form of the dose–response relationship. If the dose–response rela - species to another. If any of these barriers is impenetrable, spill over cannot tionship is linear (green line), these two excretion scenarios generate the same occur. Spillover of some pathogens requires that gaps (depicted as holes) in total probability of spillover over the time interval shown. However, for non - all of the barriers align within a narrow window in space and time (indicated linear dose–response relationships, the total probability of spillover differs by the blue arrow, see Supplementary information S2 (movie)). For other between scenarios. If the relationship is sigmoidal (red line), there is some patho gens, protracted survival in the environment (for example, Bacillus probability of spillover whenever the dose exceeds zero (indicated by the anthracis spores ), or wide dissemination (for example, the spread of aero - intensity of the red shading in the top panel), but the total spillover probability solized Coxiella burnetii by wind ), may stagger the alignment of barriers to in scen ario 2 is markedly higher. In the extreme case in which the recipient spillover. c | Top panel: hypothetical dose available over time for a given patho - host can be infected only by a dose that exceeds a sharp threshold, as sus- 67,68,79 gen. In scenario 1 (dashed light blue line), the pathogen is excreted consist - pected for Bacillus anthracis , the pathogen in scenario 2 will spill over ently from infected reservoir hosts. In scenario 2 (solid light blue line), the when the dose peaks above the threshold (blue solid line near peak), but the pathogen is excreted in regular but short high-intensity pulses over time. In pathogen in scenario 1 will never spill over. 504 | AUGUST 2017 | VOLUME 15 w w w.nature.com/nrmicro © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. Dose Probability of infection PERSPECTIVES by wind, which can lead to outbreaks of Box 1 The mathematics of spillover Q fever in humans that live several kilometres The opportunities for cross-species transmission are influenced by processes that occur at scales from the livestock reservoir hosts . from molecules to landscapes (FIG. 1). These processes are subjects of intense study, and their As illustrated by rabies virus, pathogenic characterization is complicated by their variability in space and time, nonlinear responses and Leptospira spp. and E. coli O157 (FIG. 3), interactions with outside factors. Consequently, it is impossible to integrate all of the determinants the bottlenecks that hinder the transfer of spillover transmission — or to assess the effects of gaps in our knowledge about these of pathogens between species depend determinants — without appropriate tools, such as mathematical and computational models . on the ecology of the reservoir host In Supplementary information S1 (box), we present a general mathematical model of the spillover and the pathogen, and the interactions process, which provides a template for integrating our knowledge of processes for specific disease among the determinants of spillover. For systems. This model framework essentially translates FIG. 1 into mathematical expressions. It allows example, the primary driver of pathogen for variation in space and time, and uses different formulations for transmission through pathogen excretion, slaughter or arthropod vectors. pressure for rabies virus is the prevalence The mathematical model reflects the modular nature of the spillover process, as emphasized in of infection in key hosts (such as domestic 36 the main text, while highlighting dependencies among factors in ways such as the following: dogs ). Nonlinearities in rabies transmission • Factors that are linked to disease ecology of the reservoir host and the mode of pathogen release generate a threshold effect in susceptible host determine the amount of pathogen released to the environment or vector. density below which the pathogen cannot • Pathogen survival and transport outside of the animal host, which give rise to pathogen pressure persist. These thresholds can be used to set at a particular place and time, are modelled with simple probability kernels. vaccination targets for disease elimination . • Human risk behaviours determine how this pathogen pressure translates to exposure dose. By contrast, pathogen pressure of L. interrogans is also affected by fluctuations • The probability of infection for a given dose and route of exposure is encapsulated in the dose–response relationship (FIG. 2c). in reservoir host density (such as rodents ), and prevalence and shedding from infected Mathematically, the focal point of this process is the dose to which the recipient host is exposed. animals . However, if human exposure All upstream factors come together, with appropriate functional dependencies, to shape this dose. To a reasonable approximation, which is consistent with current practice in quantitative microbial occurs through mechanisms that aggregate risk assessment , the consequent risk of infection can be modelled independently through the and disperse pathogens shed by many dose–response relationship. individuals (through accumulation in the environment, sustained survival after exiting the host , and dispersal through rain, rivers and flood waters ), the detailed dynamics occurs after human-mediated dispersal to exposure through different routes of in reservoir hosts do not matter because of the pathogen through irrigation, meat transmission . Human behaviours, such 46–48 they get integrated out by the environmental processing and food transportation . as occupational interactions with reservoir reservoir. In this scenario, spillover risk In this instance, outbreaks of E. coli O157 host animals, the consumption of certain is determined by the aggregate pathogen are determined by the pathogen pressure animal products or the use of particular pressure, human behaviours that determine on vegetables or in hamburger meat, environments, may increase the risk exposure and the integrity of within-human potentially derived from many sources. of infection . barriers to infection. For example, when As the dose that is required for E. coli O157 Exposure is often conceptualized as a 49,50 flooding mobilizes Leptospira spp. during spillover is thought to be very low , public simple point of contact. However, nonlinear the wet season in Brazil, human exposures health policies aim to completely eliminate interactions between pathogen pressure, can become widespread and epidemics pathogen pressure in food that is processed human risk behaviour and environmental 40 50 of spillover infection can occur . During for human consumption . To achieve this factors can lead to unexpected complexity, these extreme environmental events, control goal, interventions are focused on creating especially for vector-borne diseases. For efforts must focus on preventing exposure successive bottlenecks in several barriers to example, in rats, both a high prevalence to contaminated sources (for example, by spillover, including decreasing cattle density, of Y. pestis and high mortality may be wearing protective clothing and boots ) preventing faecal contamination during necessary to drive outbreaks of bubonic and reducing the infectious inoculum meat processing and increasing cooking plague in humans. Widespread exposure rather than reducing the source of pathogen temperatures to reduce exposure dose in of humans through flea bites occurs only 43,47,51 shedding, as the release of Leptospira spp. ground beef . Cumulatively, these efforts after a decrease in the abundance of rats, into the environment by animal reservoirs are usually successful, but high levels of which are the primary hosts of Y. pestis in occurs before the extreme precipitation. shedding from cattle during summer can peridomestic settings . Indeed, historically, Similarly, pathogen pressure of E. coli O157 sometimes overwhelm interventions . high rat mortality (‘rat-fall’) was an is affected by the density of its cattle host indication of an imminent human plague 42 32 population , by variation in shedding Exposure. The next phase of spillover — epidemic . Thus, killing rodents in response among individuals and by prevalence in exposure — bridges the upstream processes to cases of bubonic plague in humans could herds . Each of these factors can be highly that generate pathogen pressure and the inadvertently increase the severity of the 44,45 54 skewed and seasonal . If spillover events within-host processes in the recipient that epidemic . Conversely, and controversially, are driven by contact between humans and determine whether a given dose generates zooprophylaxis, which involves diverting cattle, then variation in pathogen load among a spillover infection (see below). The vector bites from humans by increasing animals would interact with nonlinear dose– interaction between recipient hosts and the local population density of another response functions to determine spillover pathogen pressure determines both the animal host, may decrease the risk of human risk (see below). However, this individual dose and the route of exposure. Different exposure . For example, the presence of variation matters less if human exposure behaviours of the recipient host are relevant chickens and dogs in rural areas of Argentina NATURE REVIEWS | MICROBIOLOGY VOLUME 15 | AUGUST 2017 | 505 © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES decreased the rate at which Triatoma infestans in the dose to which a host is exposed at individual receptivity . Physical barriers transmitted Trypanosoma cruzi, the causative a given location and time (the integral include the skin, mucous membranes, agent of Chagas disease, to humans . of the pathogen pressure in space and mucus, stomach acid or the absence of However, increasing the population density of time to which the host has been exposed functional receptors that enable the pathogen reservoir hosts may also affect vector survival, (Supplementary information S1 (box)). to enter its target cells or tissues . Interferon- vector abundance and pathogen prevalence induced and other innate immune responses in reservoir hosts, which, in turn, increases Probability of infection. Following may be triggered after the initial infection pathogen pressure and offsets reductions cross-species exposure of a recipient host, the of a cell, resulting in protective mechanisms 56,57 in human–vector contact rates . These within-host barriers and their interactions such as apoptosis or the induction of complexities highlight the need to understand with the strain of pathogen determine the interferon-induced resistance in surrounding the mechanisms that contribute to particular functional relationship between the pathogen cells . In addition, interfering defensive routes of spillover. dose and the likelihood that an infection will proteins in the host cell cytoplasm may block All of the factors that precede human establish. Within-host barriers to infection the replication of intracellular pathogens. In exposure, mediated by human behaviour vary widely and depend on the specific other cases, cells lack functional host factors and environmental factors (FIG. 1), cumulate combinations of pathogen, host species and that are required for the replication of the Rabies virus Leptospira interrogans Escherichia coli O157 Toxoplasma gondii Ebola virus Reservoir distribution Reservoir density Pathogen prevalence Infection intensity ? ? Pathogen release ? ? Pathogen survival and spread ? ? Human exposure ? ? Within-host barriers 43,44 Figure 3 Bottlenecks to spillover. Different barriers permit or constrain the heterogeneous shedding from cattle (although it is still unknown whether Nature Reviews | Microbiology flow of pathogens from one species to another. The figure is illustrative, super-shedding is a characteristic of particular individuals or is a transient owing to the lack of sufficient data for more than one or two barriers for any phase that occurs in most cattle ). In some contexts, exposure is an impor- given system. The width of the gaps in barriers represents the ease with tant bottleneck; for example, when the pathogen is eliminated from food which a pathogen can flow through the barriers and will vary depending on through cooking. Widespread dispersal leads to uncertainties about the 46,47 context. The question marks represent points at which the barriers are source of many outbreaks , and weak within-human barriers enable low 49,50 especi ally poorly understood and highlight gaps in our knowledge of some doses of E. coli to cause infection . Humans are frequently exposed to patho gens that are of global concern (for example, the lack of information Toxoplasma gondii carried by domestic cats and intermediate hosts, but the on disease dynamics in reservoir hosts of Ebola virus). Many rabies virus reser - parasite rarely causes disease because most humans have strong within-host voirs, such as domestic dogs, are widely distributed. The prevalence of rabies immunological barriers. Cats are widely and densely distributed, but the virus is generally low and the incidence of spillover closely tracks the preva - prevalence of T. gondii is low and cats shed oocysts only once in their life- lence of infection in the reservoir host. Rabies virus is almost always fatal to time . However, sporulated oocysts survive in the environment for long 25 112 spillover hosts . Interventions are usually aimed at reducing the preva lence periods of time . Limiting exposure to oocysts may prevent spillover; how- in reservoir hosts through vaccination . Leptospira interrogans survives in ever, this is challenging when it is unclear whether cats or the environment 111,113 water and soil after being shed in the urine of a wide range of rodents and are the major sources of infection in humans . Ebola virus has not been 29 114 other reservoir hosts . Key bottlenecks to the zoonotic spillover of this isolated from bats and the definitive reservoir bat species is unknown ; 114,115 pathoge n are exposure and within-host barriers. For example, during floods therefore, characteristics of infection in bats are unknown . The patho- in Brazil, many humans that are exposed do not become infected, probably gen is released through excretion or slaughter, then survives for up to a week, 41 116 because the initial within-host barrier, the skin, is not penetrated . However, depending on the environmental conditions . The most tractable bottle- once L. interrogans penetrates the skin (for example, through skin wounds), necks for intervention may be the zoonotic exposure of humans through 110 97,117,118 1–10 leptospires may be sufficient to cause systemic infection . Therefore, interaction with bats, bushmeat or the carcasses of other species , wearing protective clothing and boots is an effective control measure . because once exposed, the within-host barriers to Ebola virus may be Important bottlenecks to Escherichia coli O157 spillover include extremely low . 506 | AUGUST 2017 | VOLUME 15 w w w.nature.com/nrmicro © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES 60,61 pathogen . Even when pathogens can to a low but constant dose may generate the Once a pathogen has penetrated the replicate within cells, several barriers same probability of infection as intermittent within-host barriers to replicate and can prevent their transmission to other high-intensity exposures (FIG. 2c). disseminate in the new host, the outcome 62,63 cells and thus the establishment of an The genetic, immunological and of the infection may range from subclinical infection. For example, avian influenza virus physiological state of the host also can elimination of the microorganism to must pass through a series of within-host modulate the dose–response relationship. the death of the new host, and from barriers to infect a human, including mucins Immunosuppression (for example, due dead-end spillover infection to sustained in respiratory tract excretions, specific to AIDS, immunosuppressive drugs, human-to-human transmission. For many receptor molecules that constrain virus entry co-infections or malnutrition) increases important zoonotic pathogens, such as into cells and have different distributions gaps in within-host barriers, which shifts HIV or Zika virus, the transmission that in the respiratory tracts of different host dose–response curves and increases drives the current public health crisis is 69,70 81,82 species, suboptimal viral polymerase that susceptibility . For example, in human-to-human and the events that restricts the ability of the virus to replicate immunosuppressed hosts, the decreased led to spillover are long past. Although in cells of the human respiratory tract, viral number or activity of lymphocytes can understanding disease severity and onward neuraminidase that is inefficient in its role in reduce the dose that is required to establish transmission is essential for understanding the release of influenza viruses from infected an infection with the widespread pathogen the consequences of emerging infectious cells, and innate immune responses that are Toxoplasma gondii, or cause the loss of control diseases, these processes are beyond the initiated early and that block infection in of T. gondii infections that are usually kept in scope of this article. Our current knowledge 63,64 71 both infected and neighbouring cells . check by sustained immune pressure (FIG. 3). of the biological features of pathogens and From an epidemiological perspective, Seasonality in human immune function (for characteristics of host–pathogen interactions these within-host interactions between example, enhanced baseline inflammation and that determine these outcomes are described zoonotic pathogens and hosts can be altered cellular composition of the immune elsewhere (for example, see REFS 83,84). encapsulated by the functional relationship system in winter compared with summer) between pathogen dose and the probability may also alter the permeability of within-host Assessing zoonotic risk of an infection. Although there is much to barriers by altering the magnitude and speed When gaps in barriers to spillover are highly learn about dose–response relationships, of immune responses . Finally, the probability dynamic in time and space, they may vary they are expected to be nonlinear as, at and severity of infection at a given dose are asynchronously, so that the alignment of minimum, they must saturate at high doses shaped by host genetics ; triathletes with gaps in all barriers may be fleeting and because the probability of infection cannot a particular gene polymorphism were at spillover may seem random (Supplementary exceed one . This nonlinearity imposes a increased risk of leptospirosis after swallowing information S2 (movie)). Research methods filter on the dynamics of pathogen pressure lake water compared with athletes who lacked that group multiple barriers or integrate and exposure (FIG. 2c). If the dose–response this polymorphism . data over space and time may not capture relationship is highly nonlinear, such that Many of the interactions at the crossroads these dynamics. For example, ecological small changes in dose lead to large changes of exposure, inoculum dose and host niche models are often used to study in the probability of an infection, then response are poorly understood. Therefore, zoonotic risk by assessing the distribution of variation in any of the upstream factors that very little is known about the interactions reservoir hosts or vectors , but this approach culminate in an exposure dose (including between dose, timing of exposure and overlooks variation in downstream barriers released dose, pathogen survival and human probability of infection. The current that might drive risk. Alternatively, niche behaviour) may have disproportionate effects dose–response paradigm is based on models that are based on the documented on the probability of spillover. Such effects discrete transient exposures, but the effects occurrence of spillover may capture the could generate opportunities for targeted of protracted or cumulative exposure to accumulated distribution of all conditions control measures. Moreover, nonlinear environmental pathogens (for example, that enabled barriers to be breached over dose–response relationships may imply to low concentrations of Leptospira spp. in time (FIG. 1), but they cannot isolate the that infrequent high-intensity exposures floodwater) are unclear . Repeated low-dose precise barriers that affect spillover risk are more likely to cause spillover infections exposure can increase host immunity (for example, see REF. 86). Therefore, niche than continuous low-intensity excretion. to infection (for example, as postulated models tend to overestimate the spatial range This phenomenon has been reported for for poultry handlers who are exposed to of spillover risk and do not readily enable 76 87 occupational exposure to Bacillus anthracis avian influenza , dairy farmers who are extrapolation to novel conditions . Examples aerosols; tannery workers who were exposed exposed to E. coli O157 (REF. 77) and mice of this include Hendra virus and Marburg to infrequent high doses of B. anthracis that are exposed to continuous infections of virus, which can be excreted in discrete spores in imported goat hair were more parasites ). However, increases in immunity temporal and spatial pulses from their bat 20,88,89 likely to die of anthrax than those who were are not always observed; for example, such reservoir hosts . However, for spillover, exposed to frequent low doses of B. anthracis effects on immunity were not observed shedding must align with environmental and 66–68 spores . Conversely, if doses are far below in tannery workers who were exposed bat population conditions that generate levels 67,68,79 the inflection point on the dose–response to B. anthracis . Moreover, it may be of pathogen pressure that are sufficient to curve (FIG. 2c), then the system may be difficult to differentiate between a cumulative produce an infectious dose (FIG. 2), and with insensitive to changes in dose. If the dose– dose effect and the increasing opportunity exposure behaviours and susceptibility of the response function is close to linear, the total to initiate an infection with each additional recipient hosts. As some of these conditions exposure dose over time is equal and host low-dose exposure (if each infectious unit vary among seasons and years, the pattern responses do not change as a consequence of has a probability of causing an infection that of outbreaks in livestock or humans has high 20,80 20,89 early exposures, then longer-term exposure is above zero) . spatial and temporal variability . However, NATURE REVIEWS | MICROBIOLOGY VOLUME 15 | AUGUST 2017 | 507 © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES as niche models often summarize risk across (for example, common exposure to infected Outlook large areas and long durations, they overlook animal hosts and tsetse fly vectors, and low The framework presented in this Opinion important heterogeneities and they lack the resistance in humans due to the ability of article highlights that an important frontier specificity that is required for public health trypanosomes to neutralize or avoid human in research on zoonotic spillover is to 98,99 intervention. Although niche models can innate immune activity ). In all scenarios, understand the functional and quantitative help to identify regional-to-continental irrespective of the frequency with which gaps links among the determinants of spillover. 90,91 concentrations of risk , risk assessments align, the concept of hierarchical barriers To our knowledge, all of the processes that that are more quantitative and more precise can be used to organize and quantify the are necessary to achieve spillover have not with regard to space, time and which conditions that enable spillover. been connected, compared and quantified barriers they address are needed to guide The influence of particular barriers may for any single zoonotic pathogen. We concrete action. vary in space and time, and this variation address this gap, in part, by introducing Epidemiological investigations of spillover — coupled with data on realized spillover a conceptual and quantitative model that also need to account for conditions that are events — can help elucidate factors that can be used to integrate existing data, highly dynamic in space and time. If the shape infection risk, even in the absence identify high-priority data gaps, investigate alignment of gaps in all barriers is fleeting, of information on other barriers. In the conditions that widen or align gaps in delayed diagnoses or inconsistent case westernmost province of the Democratic barriers to spillover, and identify the best detection may delay outbreak investigations Republic of Congo, the observed lack gaps on which to focus intervention efforts. until the conditions that enabled spillover of monkeypox spillover, despite high We suggest that future research focuses have changed. Similarly, investigations are seroprevalence in the suspected reservoir on developing case studies that contribute sometimes triggered once the case count hosts (Heliosciurus spp. and Funisciuris to fully quantifying the determinants of becomes high. These challenges differ among spp.), was attributed to cultural norms that spillover and their linkages, with the goal pathogens with different values of R (the forbade the consumption of small rodents . of making operational contributions to risk basic reproductive number or expected The inconsistency between ecological data assessment. We provide a mathematical number of secondary infections caused by that suggested high pathogen pressure and framework that formalizes the ideas a typical infected individual in a susceptible epidemiological data that indicated a lack presented here to guide the formulation of population). For supercritical pathogens of spillover, focused attention on human mechanistic spillover models for particular with R >1, which can cause major epidemics behaviours that affect the probability of zoonotic pathogens (BOX 1; Supplementary through sustained transmission in human exposure. Research approaches that integrate information S1 (box)). We anticipate that populations (for example, Ebola virus, Zika data on multiple barriers are more likely to this synthetic framework will provide a virus and the pandemic strain of severe acute discern such behavioural effects. foundation for cross-scale data integration, respiratory syndrome coronavirus (SARS- Broad-scale discovery of novel transdisciplinary investigation, and a 4,81,92 CoV) ), spillover becomes challenging to microorganisms has the potential to new body of theory on spillover that is study because a given human case is likely characterize the pool of possible zoonotic necessary for risk assessment and public to be far removed in time or space from the pathogens and provide valuable baseline health planning. 101,102 spillover event that triggered an outbreak. information . However, each of the Raina K. Plowright is at the Department of Microbiology Subcritical pathogens with 0 <R <1, which ~63,000 species of mammals, birds, and Immunology, Montana State University, Bozeman, cause self-limiting outbreaks or ‘stuttering reptiles, amphibians and fish contains a Montana 59717, USA. chains’ in human populations (for example, multitude of infectious viruses, bacteria and Colin R. Parrish is at the Baker Institute for Animal 93,94 101,102,104–106 monkeypox or avian influenza viruses ), parasites . Although each of these Health, College of Veterinary Medicine, raise distinct challenges because any given microorganisms and parasites can be viewed Cornell University, Ithaca, New York 14853, USA. individual could have been infected by either as a potential pathogen, the vast majority Hamish McCallum is at the Griffith School of an animal or a human source . It is easiest to may not cause disease in their natural Environment, Griffith University, Brisbane, study the spillover of pathogens with R = 0 hosts, and the extent to which they infect or Queensland 4111, Australia. that are not transmitted between humans cause pathology in other species, including Peter J. Hudson is at the Center for Infectious 7,9,10 (for example, rabies virus or West Nile humans, is unknown . Therefore, Disease Dynamics, Pennsylvania State University, 25,95 virus ), in which every case is an instance discovery alone cannot address the potential State College, Pennsylvania 16802, USA. of spillover. The 2014–2015 Ebola virus risk of spillover. The translation of new Albert I. Ko is at the Department of Epidemiology of epidemic in West Africa is a prime example discoveries of microorganisms into guidance Microbial Diseases, Yale School of Public Health, whereby delayed response and investigation for public health practitioners requires the New Haven, Connecticut 06520–8034, USA. prevented researchers from reconstructing identification of the barriers to microbial Andrea L. Graham is at the Department of Ecology the conditions that initiated the human infection of humans, the conditions that & Evolutionary Biology, Princeton University, 96,97 epidemic of a supercritical pathogen . facilitate the breaching of these barriers, Princeton, New Jersey 08544, USA. Ebola virus infection is an extreme example and, therefore, the microbiological and James O. Lloyd-Smith is at the Department of Ecology & of spillover infection that only occurs during environmental contexts that pose the Evolutionary Biology, University of California, the rare alignment of gaps in barriers, and, greatest risk to human populations. For the Los Angeles, Los Angeles, California 90095-7239, USA; and at Fogarty International Center, National Institutes accordingly, the precise determinants of risk foreseeable future, the greatest practical of Health, Bethesda, Maryland 20892–2220, USA. are poorly understood (FIG. 3). By contrast, contribution of pathogen discovery for other zoonoses, such as trypanosomiasis and sequence characterization to the Correspondence to R.K.P. [email protected] in some parts of Africa, incidence is high epidemiology of emerging pathogens is likely because the pathogen flows through to be in the rapid post hoc identification of doi:10.1038/nrmicro.2017.45 consistently wide gaps in barriers to infection novel pathogens after spillover. Published online 30 May 2017 508 | AUGUST 2017 | VOLUME 15 w w w.nature.com/nrmicro © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES 1. Christou, L. The global burden of bacterial and viral 26. Webster, R. in Viral Zoonoses and Food of Animal 49. Teunis, P., Ogden, I. & Strachan, N. Hierarchical dose Origin (eds Kaaden, O.‑ R., Czerny, C.‑ P. & zoonotic infections. Clin. Microbiol. Infect. 17, response of E. coli O157: H7 from human outbreaks 326–330 (2011). Eichhorn, W.) 105–113 (Springer, 1997). incorporating heterogeneity in exposure. Epidemiol. 2. Morens, D. M., Folkers, G. K. & Fauci, A. S. The 27. Ko, A. I., Goarant, C. & Picardeau, M. Leptospira: the Infect. 136, 761–770 (2008). challenge of emerging and re‑emerging infectious dawn of the molecular genetics era for an emerging 50. Tuttle, J. et al. Lessons from a large outbreak of diseases. Nature 430, 242–249 (2004). zoonotic pathogen. Nat. Rev. Microbiol. 7, 736–747 Escherichia coli O157:H7 infections: insights into (2009). 3. Jones, K. E. et al. Global trends in emerging infectious the infectious dose and method of widespread diseases. Nature 451, 990–993 (2008). 28. Costa, F. et al. Influence of household rat infestation contamination of hamburger patties. Epidemiol. Infect. This study analyses the general phylogenetic and on Leptospira transmission in the urban slum 122, 185–192 (1999). environment. PLoS Negl. Trop. Dis. 8, e3338 (2014). geographical risk factors for many different 51. Cobbold, R. N. et al. Rectoanal junction colonization of emerging diseases, as well as temporal and spatial 29. Costa, F. et al. Patterns in Leptospira shedding in feedlot cattle by Escherichia coli O157: H7 and its trends in emerging infections. Norway rats (Rattus norvegicus) from Brazilian slum association with supershedders and excretion communities at high risk of disease transmission. 4. Briand, S. et al. The international Ebola emergency. dynamics. Appl. Environ. Microbiol. 73, 1563–1568 N. Engl. J. Med. 371, 1180–1183 (2014). PLoS Negl. Trop. Dis. 9, e0003819 (2015). (2007). 5. Smith, G. J. et al. Origins and evolutionary genomics 30. Monahan, A. M., Callanan, J. J. & Nally, J. E. 52. Cascio, A., Bosilkovski, M., Rodriguez‑ Morales, A. & Proteomic analysis of Leptospira interrogans shed in of the 2009 swine‑ origin H1N1 influenza A epidemic. Pappas, G. The socio‑ ecology of zoonotic infections. Nature 459, 1122–1125 (2009). urine of chronically infected hosts. Infect. Immun. 76, Clin. Microbiol. Infect. 17, 336–342 (2011). 6. Fevre, E. M., Wissmann, B. V., Welburn, S. C. 4952–4958 (2008). 53. Macpherson, C. N. Human behaviour and the 31. Nally, J. E., Chow, E., Fishbein, M. C., Blanco, D. R. & Lutumba, P. The burden of human African epidemiology of parasitic zoonoses. Int. J. Parasitol. trypanosomiasis. PLoS Negl. Trop. Dis. 2, e333 & Lovett, M. A. Changes in lipopolysaccharide O 35, 1319–1331 (2005). (2008). antigen distinguish acute versus chronic Leptospira 54. Keeling, M. J. & Gilligan, C. A. Metapopulation 7. Grice, E. A. & Segre, J. A. The skin microbiome. interrogans infections. Infect. Immun. 73, 3251–3260 dynamics of bubonic plague. Nature 407, 903–906 Nat. Rev. Microbiol. 9, 244–253 (2011). (2005). (2000). 32. Smego, R., Frean, J. & Koornhof, H. Yersiniosis I: 8. Guarner, F. & Malagelada, J.‑R. Gut flora in health and This study uses dynamic models to explain disease. Lancet 361, 512–519 (2003). microbiological and clinicoepidemiological aspects of historical patterns of bubonic plague, and shows 9. Gilbert, S. F., Sapp, J. & Tauber, A. I. A symbiotic view plague and non‑ plague Yersinia infections. Eur. J. Clin. that, counterintuitively, culling rats may exacerbate Microbiol. Infect. Dis. 18, 1–15 (1999). of life: we have never been individuals. Q. Rev. Biol. plague. 87, 325–341 (2012). 33. Weber, T. P. & Stilianakis, N. I. Inactivation of 55. Hess, A. & Hayes, R. O. Relative potentials of domestic 10. Parrish, C. R. et al. Cross‑ species virus transmission influenza A viruses in the environment and modes of animals for zooprophylaxis against mosquito vectors transmission: a critical review. J. Infect. 57, 361–373 and the emergence of new epidemic diseases. of encephalitis. Am. J. Trop. Med. Hyg. 19, 327–334 Microbiol. Mol. Biol. Rev. 72, 457–470 (2008). (2008). (1970). This article reviews the general features that are 34. Koopmans, M. et al. Transmission of H7N7 avian 56. Gürtler, R. E. et al. Domestic animal hosts strongly influenza A virus to human beings during a large associated with the emergence of viruses in new influence human‑ feeding rates of the Chagas disease hosts to cause epidemics or pandemics. outbreak in commercial poultry farms in the vector Triatoma infestans in Argentina. PLoS Negl. 11. Woolhouse, M. E. & Gowtage‑ Sequeria, S. Host range Netherlands. Lancet 363, 587–593 (2004). Trop. Dis. 8, e2894 (2014). 35. Tissot‑ Dupont, H., Amadei, M.‑A., Nezri, M. & and emerging and reemerging pathogens. 57. Kilpatrick, A. M. & Randolph, S. E. Drivers, dynamics, Emerg. Infect. Dis. 11, 1842–1847 (2005). Raoult, D. Wind in November, Q fever in December. and control of emerging vector‑ borne zoonotic 12. Taylor, L. H., Latham, S. M. & Woolhouse, M. E. J. Emerg. Infect. Dis. 10, 1264 (2004). diseases. Lancet 380, 1946–1955 (2012). Risk factors for human disease emergence. 36. Hampson, K. et al. Synchronous cycles of domestic 58. Schmid‑ Hempel, P. Variation in immune defence as a Phil. Trans. R. Soc. Lond. B Biol. Sci. 356, 983–989 dog rabies in sub‑ Saharan Africa and the impact of question of evolutionary ecology. Proc. Biol. Sci. 270, control efforts. Proc. Natl Acad. Sci. USA 104, (2001). 357–366 (2003). 13. Morse, S. S. Factors in the emergence of infectious 7717–7722 (2007). 59. Akira, S., Uematsu, S. & Takeuchi, O. Pathogen diseases. Emerg. Infect. Dis. 1, 7–15 (1995). 37. Brochier, B. et al. Large‑ scale eradication of rabies recognition and innate immunity. Cell 124, 783–801 using recombinant vaccinia–rabies vaccine. Nature 14. Lloyd‑ Smith, J. O. et al. Epidemic dynamics at the (2006). human–animal interface. Science 326, 1362–1367 354, 520–522 (1991). 60. Trobaugh, D. W. & Klimstra, W. B. MicroRNA (2009). 38. Andre‑ Fontaine, G., Aviat, F. & Thorin, C. Waterborne regulation of RNA virus replication and pathogenesis. leptospirosis: survival and preservation of the This study delineates stages of zoonoses on the Trends Mol. Med. 23, 80–93 (2017). basis of changes in transmissibility, as reflected virulence of pathogenic Leptospira spp. in fresh water. 61. Duggal, N. K. & Emerman, M. Evolutionary conflicts in R . It also reviews the literature on modelling Curr. Microbiol. 71, 136–142 (2015). between viruses and restriction factors shape 39. Lau, C. L., Smythe, L. D., Craig, S. B. & Weinstein, P. transmission dynamics of zoonoses and identifies immunity. Nat. Rev. Immunol. 12, 687–695 gaps in our knowledge. Climate change, flooding, urbanisation and (2012). 15. Johnson, C. K. et al. Spillover and pandemic leptospirosis: fuelling the fire? Trans. R. Soc. Trop. 62. Air, G. M. & Laver, W. G. The neuraminidase of Med. Hyg. 104, 631–638 (2010). properties of zoonotic viruses with high host plasticity. influenza virus. Proteins 6, 341–356 (1989). Sci. Rep. 5, 14830 (2015). 40. Reis, R. B. et al. Impact of environment and social 63. Kuiken, T. et al. Host species barriers to influenza virus 16. Lloyd‑ Smith, J. O., Funk, S., McLean, A. R., Riley, S. gradient on Leptospira infection in urban slums. infections. Science 312, 394–397 (2006). & Wood, J. L. Nine challenges in modelling the PLoS Negl. Trop. Dis. 2, e228 (2008). 64. Lipsitch, M. et al. Viral factors in influenza pandemic emergence of novel pathogens. Epidemics 10, 35–39 41. Phraisuwan, P. et al. Leptospirosis: skin wounds and risk assessment. eLife 5, e18491 (2016). control strategies, Thailand, 1999. Emerg. Infect. Dis. (2015). 65. Schmid‑ Hempel, P. & Frank, S. A. Pathogenesis, 17. Gortazar, C. et al. Crossing the interspecies barrier: 8, 1455–1459 (2002). virulence, and infective dose. PLoS Pathog. 3, e147 opening the door to zoonotic pathogens. PLoS Pathog. 42. Spencer, S. E., Besser, T. E., Cobbold, R. N. & (2007). French, N. P. ‘Super’or just ‘above average’? 10, e1004129 (2014). 66. Brachman, P. S. & Fekety, F. R. Industrial anthrax. 18. Wolfe, N. D., Dunavan, C. P. & Diamond, J. Origins Supershedders and the transmission of Escherichia Ann. NY Acad. Sci. 70, 574–584 (1958). of major human infectious diseases. Nature 447, coli O157: H7 among feedlot cattle. J. R. Soc. 67. Brachman, P. S., Kaufman, A. & Dalldorf, F. G. Interface 12, 0446 (2015). 279–283 (2007). Industrial inhalation anthrax. Bacteriol. Rev. 30, 646 19. Pepin, K. M., Lass, S., Pulliam, J. R., Read, A. F. & This study examines the dynamics of E. coli (1966). Lloyd‑ Smith, J. O. Identifying genetic markers of transmission and the roles of super-shedder This study is one of the only studies to calculate the individuals in those processes. adaptation for surveillance of viral host jumps. risk of spillover infection using comparable doses Nat. Rev. Microbiol. 8, 802–813 (2010). 43. Matthews, L. et al. Heterogeneous shedding of administered over time and provides evidence for 20. Plowright, R. K. et al. Ecological dynamics of emerging Escherichia coli O157 in cattle and its implications for the outcome of repeated low-dose versus single control. Proc. Natl Acad. Sci. USA 103, 547–552 bat virus spillover. Proc. R. Soc. B Biol. Sci. 282, high-dose exposure. 20142124 (2015). (2006). 68. Coleman, M. E., Thran, B., Morse, S. S., Hugh‑ This study outlines the conditions that enable 44. Hancock, D., Besser, T., Rice, D., Herriott, D. & Tarr, P. Jones, M. & Massulik, S. Inhalation anthrax: dose spillover of bat viruses into other hosts and A longitudinal study of Escherichia coli O157 in response and risk analysis. Biosecur. Bioterror. 6, provides an example of the infections that are fourteen cattle herds. Epidemiol. Infect. 118, 147–160 (2008). 193–195 (1997). the subject of this review. 69. Bollaerts, K. et al. Human salmonellosis: estimation 21. Hudson, P. J., Rizzoli, A. R., Grenfell, B. T., 45. Besser, T. E., Davis, M. A. & Walk, S. T. in Population of dose–illness from outbreak data. Risk Anal. 28, Heesterbeek, H. & Dobson, A. P. The Ecology of Genetics of Bacteria: A Tribute to Thomas S. Whittam 427–440 (2008). (eds Walk, S. T. & Feng, P. C. H.) 303–324 (2011). Wildlife Diseases (Oxford Univ. Press, 2002). 70. Kau, A. L., Ahern, P. P., Griffin, N. W., Goodman, A. L. 22. Hjelle, B. & Glass, G. E. Outbreak of hantavirus 46. Gerba, C. P. & Smith, J. E. Sources of pathogenic & Gordon, J. I. Human nutrition, the gut microbiome infection in the Four Corners region of the United microorganisms and their fate during land application and the immune system. Nature 474, 327–336 of wastes. J. Environ. Qual. 34, 42–48 (2005). States in the wake of the 1997–1998 El Nino— (2011). Southern Oscillation. J. Infect. Dis. 181, 1569–1573 47. Elder, R. O. et al. Correlation of enterohemorrhagic 71. Greene, C. E. Infectious Diseases of the Dog and Cat (2000). Escherichia coli O157 prevalence in feces, hides, and (Elsevier Health Sciences, 2013). carcasses of beef cattle during processing. Proc. Natl 23. Thoen, C. O., Steele, J. H. & Kaneene, J. B. Zoonotic 72. Dopico, X. C. et al. Widespread seasonal gene Tuberculosis: Mycobacterium bovis and Other Acad. Sci. USA 97, 2999–3003 (2000). expression reveals annual differences in human Pathogenic Mycobacteria (John Wiley & Sons, 2014). This study calculates the decreasing pathogen immunity and physiology. Nat. Commun. 6, 7000 pressure (availability for human exposure) of E. coli 24. Ducatez, M., Webster, R. & Webby, R. Animal influenza (2015). epidemiology. Vaccine 26, D67–D69 (2008). O157 as carcasses progress through the various 73. Gingles, N. A. et al. Role of genetic resistance in 25. Rupprecht, C. E., Hanlon, C. A. & Hemachudha, T. stages of processing at meat processing plants. invasive pneumococcal infection: identification and Rabies re‑examined. Lancet Infect. Dis. 2, 327–343 48. Pennington, H. Escherichia coli O157. Lancet 376, study of susceptibility and resistance in inbred mouse (2002). 1428–1435 (2010). strains. Infect. Immun. 69, 426–434 (2001). NATURE REVIEWS | MICROBIOLOGY VOLUME 15 | AUGUST 2017 | 509 © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES viruses: a processbased per ‑ spective. Am. Nat. 187, 111. Elmore, S. A. et al. Toxoplasma gondii: epidemiology, 74. Lingappa, J. et al. HLA‑DQ6 and ingestion of contaminated water: possible gene–environment E53–E64 (2016). feline clinical aspects, and prevention. Trends interaction in an outbreak of leptospirosis. 92. Ksiazek, T. G. et al. A novel coronavirus associated Parasitol. 26, 190–196 (2010). with severe acute respiratory syndrome. 112. Dubey, J. Toxoplasma gondii oocyst survival under Genes Immun. 5, 197–202 (2004). 75. Pujol, J. M., Eisenberg, J. E., Haas, C. N. & N. Engl. J. Med. 348, 1953–1966 (2003). defined temperatures. J. Parasitol. 84, 862–865 Koopman, J. S. The effect of ongoing exposure 93. Hutin, Y. et al. Outbreak of human monkeypox, (1998). dynamics in dose response relationships. Democratic Republic of Congo, 1996 to 1997. 113. Jones, J. & Dubey, J. Waterborne toxoplasmosis — PLoS Comput. Biol. 5, e1000399 (2009). Emerg. Infect. Dis. 7, 434 (2001). recent developments. Exp. Parasitol. 124, 10–25 94. Li, Q. et al. Epidemiology of human infections with (2010). 76. Yang, W. & Shaman, J. Does exposure to poultry and wild fowl confer immunity to H5N1? Chin. Med. J. avian influenza A (H7N9) virus in China. 114. Leroy, E. M. et al. Fruit bats as reservoirs of Ebola 127, 3335 (2014). N. Engl. J. Med. 370, 520–532 (2014). virus. Nature 438, 575–576 (2005). 95. Hayes, E. B. et al. Epidemiology and transmission 115. Pourrut, X. et al. Spatial and temporal patterns of 77. Reymond, D. et al. Neutralizing antibodies to Escherichia coli vero cytotoxin 1 and antibodies to dynamics of West Nile virus disease. Emerg. Infect. Zaire ebolavirus antibody prevalence in the possible O157 lipopolysaccharide in healthy farm family Dis. 11, 1167–1173 (2005). reservoir bat species. J. Infect. Dis. 196, S176–S183 96. Baize, S. et al. Emergence of Zaire Ebola virus disease (2007). members and urban residents. J. Clin. Microbiol. 34, 2053–2057 (1996). in Guinea — preliminary report. N. Engl. J. Med. 371, 116. Prescott, J. et al. Postmortem stability of Ebola virus. 78. Scott, M. High transmission rates restore expression 1418–1425 (2014). Emerg. Infect. Dis. 21, 856 (2015). 97. Saéz, A. M. et al. Investigating the zoonotic origin of 117. Leroy, E. M. et al. Human Ebola outbreak resulting of genetically determined susceptibility of mice to nematode infections. Parasitology 132, 669–679 the West African Ebola epidemic. EMBO Mol. Med. 7, from direct exposure to fruit bats in Luebo, (2006). 17–23 (2015). Democratic Republic of Congo, 2007. Vector Borne 98. Kieft, R. et al. Mechanism of Trypanosoma brucei Zoonotic Dis. 9, 723–728 (2009). 79. Cohen, M. L. & Whalen, T. Implications of low level human exposure to respirable B. anthracis. gambiense (group 1) resistance to human 118. Leroy, E. M. et al. Multiple Ebola virus transmission Appl. Biosafety 12, 109 (2007). trypanosome lytic factor. Proc. Natl Acad. Sci. USA events and rapid decline of central African wildlife. 80. French, N., Kelly, L., Jones, R. & Clancy, D. Dose– 107, 16137–16141 (2010). Science 303, 387–390 (2004). response relationships for foot and mouth disease in 99. Simarro, P. P. et al. Estimating and mapping the 119. Judson, S., Prescott, J. & Munster, V. Understanding population at risk of sleeping sickness. PLoS Negl. Ebola virus transmission. Viruses 7, 511–521 (2015). cattle and sheep. Epidemiol. Infect. 128, 325–332 (2002). Trop. Dis. 6, e1859 (2012). 81. Faria, N. R. et al. Zika virus in the Americas: early 100. Jezek, Z. & Fenner, F. in Monographs in Virology Acknowledgements Vol. 17 (ed. Melnick, J. L.) 119–121 (Karger, 1988). The authors thank J. Wood and E. Fleishman for helpful epidemiological and genetic findings. Science 352, 345–349 (2016). 101. Anthony, S. J. et al. A strategy to estimate unknown comments and conversations. R.K.P. and H.M. are sup‑ 82. Hahn, B. H., Shaw, G. M., De Cock, K. M. & viral diversity in mammals. mBio 4, e00598‑13 ported by the Commonwealth of Australia, the State of New (2013). South Wales and the State of Queensland under the Sharp, P. M. AIDS as a zoonosis: scientific and public health implications. Science 287, 607–614 (2000). This study estimates the number of viruses from National Hendra Virus Research Program, awarded through 83. Geoghegan, J. L., Senior, A. M., Di Giallonardo, F. nine viral families in one bat host, and uses that the Rural Industries Research and Development to extrapolate and estimate that there would be Corporation. R.K.P. is supported by the US National & Holmes, E. C. Virological factors that increase the transmissibility of emerging human viruses. 320,000 viruses from those families in mammals. Institutes of General Medical Sciences IDeA Program (grants Proc. Natl Acad. Sci. USA 113, 4170–4175 (2016). 102. Temmam, S., Davoust, B., Berenger, J.‑M., Raoult, D. P20GM103474 and P30GM110732), P. Thye, the Morris & Desnues, C. Viral metagenomics on animals as a Animal Foundation, Montana University System Research This study identifies and quantifies biological features of viruses that best determine human tool for the detection of zoonoses prior to human Initiative (grant 51040‑MUSRI2015‑03), a Defense infection and transmissibility between humans. infection? Int. J. Mol. Sci. 15, 10377–10397 Advanced Research Projects Agency (DARPA) Young Faculty 84. Casadevall, A. & Pirofski, L. Host–pathogen (2014). Award and the US Department of Defense Strategic interactions: the attributes of virulence. J. Infect. Dis. 103. Hoffmann, M. et al. The impact of conservation on Environmental Research and Development Program the status of the world’s vertebrates. Science 330, (SERDP; grant RC‑2633). J.O.L.‑S. is supported by the US 184, 337–344 (2001). 85. Miller, R. H. et al. Ecological niche modeling to 1503–1509 (2010). National Science Foundation (NSF; grants OCE‑1335657 estimate the distribution of Japanese encephalitis 104. Ley, R. E. et al. Evolution of mammals and their gut and DEB‑1557022) and the US Department of Defense microbes. Science 320, 1647–1651 (2008). SERDP (grant RC‑2635). J.O.L.‑S., A.L.G. and P.J.H. are virus in Asia. PLoS Negl. Trop. Dis. 6, e1678 (2012). 86. Levine, R. S. et al. Ecological niche and geographic 105. Turnbaugh, P. J. et al. The human microbiome project: supported by the RAPIDD program of the Science & distribution of human monkeypox in Africa. PLoS ONE exploring the microbial part of ourselves in a changing Technology Directorate of the Department of Homeland world. Nature 449, 804 (2007). Security, the Fogarty International Center (part of the US 2, e176 (2007). 87. Kearney, M., Simpson, S. J., Raubenheimer, D. & 106. Dobson, A., Lafferty, K. D., Kuris, A. M., National Institutes of Health), and by IDEAS (Infectious Helmuth, B. Modelling the ecological niche from Hechinger, R. F. & Jetz, W. Homage to Linnaeus: how Disease Evolution Across Scales), which is a Research many parasites? How many hosts? Proc. Natl Acad. Coordination Network (DEB‑1354890) funded by the US functional traits. Phil. Trans. R. Soc. B Biol. Sci. 365, 3469–3483 (2010). Sci. USA 105, 11482–11489 (2008). National Science Foundation. 88. Plowright, R. K. et al. Transmission or withinhost ‑ 107. Heesterbeek, H. et al. Modeling infectious disease dynamics in the complex landscape of global health. Competing interests statement dynamics driving pulses of zoonotic viruses in reservoirhost populations ‑ . PLoS Negl. Trop. Dis. 10, Science 347, aaa4339 (2015). The authors declare no comparing interests. e0004796 (2016). 108. Haas, C. N., Rose, J. B. & Gerba, C. P. Quantitative 89. Amman, B. R. et al. Seasonal pulses of Marburg virus Microbial Risk Assessment (John Wiley & Sons, Publisher’s note circulation in juvenile Rousettus aegyptiacus bats 2014). Springer Nature remains neutral with regard to jurisdictional 109. Mitscherlich, E. & Marth, E. H. Microbial Survival in claims in published maps and institutional affiliations. coincide with periods of increased risk of human infection. PLoS Pathog. 8, e1002877 (2012). the Environment: Bacteria and Rickettsiae Important 90. Pigott, D. M. et al. Mapping the zoonotic niche of in Human and Animal Health (Springer Science & Business Media, 2012). SUPPLEMENTARY INFORMATION Marburg virus disease in Africa. Trans. R. Soc. Trop. See online article: S1 (box) | S2 (movie) Med. Hyg. 109, 366–378 (2015). 110. Silva, É. F. et al. Characterization of virulence of 91. Brierley, L., Vonhof, M., Olival, K., Daszak, P. & Leptospira isolates in a hamster model. Vaccine 26, ALL LINKS ARE ACTIVE IN THE ONLINE PDF 3892–3896 (2008). Jones, K. Quantifying global drivers of zoonotic bat 510 | AUGUST 2017 | VOLUME 15 w w w.nature.com/nrmicro © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nature Reviews Microbiology Pubmed Central

Pathways to zoonotic spillover

Download PDF
9 pages

Loading next page...
 
/lp/pubmed-central/pathways-to-zoonotic-spillover-wIss3nDMqY

References (130)

Publisher
Pubmed Central
Copyright
© Nature Publishing Group, a division of Macmillan Publishers Limited. All Rights Reserved. 2017
ISSN
1740-1526
eISSN
1740-1534
DOI
10.1038/nrmicro.2017.45
Publisher site
See Article on Publisher Site

Abstract

PERSPECTIVES Although many recent articles have OPINION examined the fields of zoonoses or emerging 2,3,10–15 pathogens , a synthetic mechanistic understanding of animal-to-human 14,16 transmission is lacking . Much attention has been dedicated to the characterization Raina K. Plowright, Colin R. Parrish, Hamish McCallum, Peter J. Hudson, 3,11,12,15 of emerging infections ; for example, Albert I. Ko, Andrea L. Graham and James O. Lloyd-Smith the high frequency of zoonoses among 3,12 emerging infections , their socio-economic, Abstract Zoonotic spillover, which is the transmission of a pathogen from environmental and ecological 2,13,17,18 a vertebrate animal to a human, presents a global public health burden but is a drivers , and their phylogenetic and poorly understood phenomenon. Zoonotic spillover requires several factors to geographical distribution . Similarly, the phases of zoonotic emergence in the align, including the ecological, epidemiological and behavioural determinants of 11,14,18 human population , adaptation and pathogen exposure, and the within-human factors that affect susceptibility to 10,11,19 compatibility of zoonoses in humans , infection. In this Opinion article, we propose a synthetic framework for and approaches to modelling the animal-to-human transmission that integrates the relevant mechanisms. This 14,16 transmission of zoonoses , have also framework reveals that all zoonotic pathogens must overcome a hierarchical series been addressed in the literature. However, a comprehensive understanding of the of barriers to cause spillover infections in humans. Understanding how these processes that enable a pathogen from a barriers are functionally and quantitatively linked, and how they interact in space vertebrate animal to establish infection and time, will substantially improve our ability to predict or prevent spillover in a human, and how these processes are events. This work provides a foundation for transdisciplinary investigation of hierarchically, functionally and quantitatively spillover and synthetic theory on zoonotic transmission. linked, remains a fundamental deficit in 14,16 research on zoonoses . In this Opinion The phenomenon of cross-species spillover host distribution, pathogen prevalence and article, we present a mechanistic structure is the defining characteristic of pathogens pathogen release from the reservoir host, that integrates the determinants of spillover that transmit from vertebrate animals to followed by pathogen survival, development and the interactions among them (FIG. 1). humans (zoonoses). The public health and dissemination outside of the reservoir However, we do not address broader burden that is presented by zoonoses hosts. Second, human and vector determinants of pathogen emergence or includes outbreaks of pathogens such as behaviour determine pathogen exposure; factors that affect disease severity or onward Ebola virus, influenza A virus (H1N1) specifically, the likelihood, route and dose of transmission in humans. pdm09 and Middle East respiratory exposure. Third, genetic, physiological and Although many of the individual syndrome coronavirus (MERS-CoV), immunological attributes of the recipient determinants of spillover are subjects of as well as the ongoing transmission of human host, together with the dose and intensive study, each is usually addressed in endemic pathogens, such as Salmonella route of exposure, affect the probability and isolation in a specialized discipline (FIG. 2). spp., Leptospira spp., Trypanosoma spp., severity of infection. Accordingly, the better-characterized 1–6 Mycobacterium spp. and West Nile virus . Each phase presents multiple barriers factors become the focus of public health Spillover transmission is promoted by to the flow of a pathogen from a reservoir interventions. For example, reservoir hosts successive processes that enable an animal host to a recipient host. Spillover requires or vectors are often targeted for control pathogen to establish infection in a human. the pathogen to pass every barrier and before the concatenation and relative The probability of zoonotic spillover is thus can only occur when gaps align influence of processes that lead to spillover determined by interactions among several in each successive barrier within an are understood, which sometimes leads factors, including disease dynamics in appropriate window in space and time to inefficient or even counterproductive the reservoir host, pathogen exposure (FIG. 2). Consequently, zoonotic spillover is a interventions . In other cases, multiple and the within-human factors that affect relatively rare event, and although humans mechanisms are aggregated in analyses that susceptibility to infections. These factors are continually exposed to many potentially obscure the interactions or heterogeneities can be partitioned into three functional infectious pathogens that are derived from that drive risk. Although the aggregation of phases that describe all major routes of other species, most of these microorganisms mechanisms may be appropriate at times, 7–10 transmission (FIG. 1). In the first phase, cannot infect or cause disease in humans . identifying discrete mechanisms and how the amount of pathogen available to the This Opinion article focuses on spillover they interact to drive spillover is essential to human host at a given point in space and transmission, strictly defined as the processes recognize the assumptions that are implicit time, known as the pathogen pressure, is that enable a pathogen from a vertebrate in simpler models, and to clarify which determined by interactions among reservoir animal to establish infection in a human. processes must be modelled explicitly and 502 | AUGUST 2017 | VOLUME 15 w w w.nature.com/nrmicro © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES which can be combined. For example, • Reservoir host distribution does assessment of the risk of acquiring a Distribution and • Reservoir host density intensity of infection zoonotic infection require the measurement • Prevalence of infection in reservoir hosts • Intensity of infection of the pathogen burden carried by individual reservoir hosts, or is it sufficient to estimate the cumulative abundance of a pathogen Excretion Slaughter Vector borne in the environment over time? This is a key question for pathogens such as Leptospira Pathogen release Shedding rate Harvest rate Biting rate (vector– from reservoir host reservoir host) interrogans, Giardia spp., and Escherichia coli O157, and the answer may depend on modes Pathogen survival, Pathogen survival Pathogen survival Vector survival of contact and dose–response relationships development and and movement and transport of and movement dissemination meat in humans (see below). Models that integrate data from experiments, the field Pathogen pressure and epidemiological studies, even if only Human exposure Human behavior Butchering, Biting rate partially parameterized, may be necessary to pathogen that leads to contact preparation and (vector–human) to make such determinations. with pathogen eating We describe how pathogens overcome a series of barriers to pass from reservoir hosts to humans. Crucially, nonlinear interactions Dose and route of exposure among the barriers create bottlenecks in the flow of a pathogen between species. Such bottlenecks provide opportunities for public health interventions that could lead to substantial reductions in the risk of spillover. Alternatively, changing environmental • Structural barriers Host susceptibility • Innate immune response and molecular compatibility or social conditions can alleviate these • Replication and dissemination cycles completed bottlenecks, which can cause surges in Probability of infection spillover infections. Our framework provides the foundation for operational models that are required for quantitative evidence-based Figure 1 | Pathways to spillover. The risk of spillover is determined by a series of processes that link Nature Reviews | Microbiology the ecological dynamics of infection in reservoir hosts, the microbiological and vector determinants risk analysis, preparedness, surveillance of survival and dissemination outside of reservoir hosts, the epidemiological and behavioural deter- and control. minants of exposure, and the within-host biological factors that shape the susceptibility of recipient hosts. The distribution and intensity of infection in reservoir hosts, followed by pathogen release, Barriers to spillover movement, survival and possible development to infectious stage, determine the pathogen pressure, The probability of spillover is determined by which is defined as the amount of pathogen available to the recipient host at a given point in space the interactions among the barriers and the and time. Pathogen pressure then interacts with the behaviour of the recipient host (and vector for associated bottlenecks that might prevent vector-borne pathogens) to determine the likelihood, dose and route of exposure. A series of within- cross-species transmission. Many of these host barriers then determine host susceptibility, and, therefore, the probability and severity of interactions are nonlinear and dynamic in infection for a given pathogen dose. space and time. Pathogen pressure. The series of processes The first set is the natural history of infection The mode of pathogen release from that culminate in pathogen pressure (the in hosts, which includes the duration, reservoir hosts determines the major routes amount of a pathogen that is available to intensity and severity of infection and the of transmission. Pathogens may be released in humans at a given point in time and space) level of shedding. Second, the movement host excretions, through slaughter or through includes pathogen dynamics in reservoir and behaviour of hosts affect contact and the an arthropod vector (FIG. 1). The probability hosts, pathogen release from reservoir hosts, likelihood of exposure within and between of a pathogen being released from a reservoir and pathogen survival or dispersal outside of species. These factors interact with the host is affected by its presence and viability in reservoir hosts. abundance, density, demographic turnover, relevant tissues, such as the blood for many Pathogen dynamics in reservoir hosts spatial distribution and physiological state of vector-borne pathogens, tissues contacted can be represented as three variables that hosts to determine the efficiency of spread . or consumed during butchering and eating determine the distribution and intensity Collectively, these processes determine how for some food-borne pathogens, and tissues of infection in time and space: the density of the pathogen is distributed across reservoir through which external shedding occurs for reservoir hosts, the prevalence of infection host populations. Such pathogen distribution direct or environmental routes. For example, among reservoir hosts, and the average can be highly variable (for example, pulses the viral load and excretion rates in the intensity of infection in an infected reservoir of Sin Nombre virus infections in deer salivary glands are key determinants for the host in time and space (Supplementary mice (Peromyscus maniculatus) populations transmission of rabies virus from carnivores, information S1 (box)). Many ecological and in response to climate-driven increases in whereas viral loads in the intestinal and physiological factors influence these variables population density) , or stable (as illustrated respiratory tracts affect the transmission 24–26 in communities of reservoir animals; by Mycobacterium bovis infections in of avian influenza virus from poultry . however, two sets of factors are dominant. populations of livestock and wildlife) . Likewise, the release of pathogenic Leptospira NATURE REVIEWS | MICROBIOLOGY VOLUME 15 | AUGUST 2017 | 503 © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES spp. from animal hosts requires colonization Following the release of a pathogen from information S1 (box)). Spillover of pathogens of the renal tubules . The excreted pathogen its reservoir host, the opportunity for spillover that have short survival times (for example, load depends on the quantity of leptospires transmission is influenced by the duration influenza A virus when transmitted through 28 33,34 that effectively colonize the tubules , the of pathogen survival outside of its host, the the respiratory route) may require close rate of release and the urinary output of the extent of spatial dispersal through passive interactions between reservoir and recipient host . Moreover, the pathogen undergoes transport (for example, through water, on hosts. Consequently, spillover patterns in several changes in its lipopolysaccharide fomites or in the air), and possible pathogen recipient hosts correspond to the prevalence content and proteome during colonization reproduction or obligate developmental stages patterns in reservoir hosts. By contrast, 30,31 and shedding in the urine , which suggests outside of the primary host (for example, if pathogens survive for sufficient periods that priming in the renal milieu is required Yersinia pestis, the causative agent of plague, of time outside of their reservoir hosts, they to adapt for survival and infectivity in the must multiply within flea vectors before it can may be dispersed beyond the home range of external environment. The rate of pathogen be transmitted to humans ). These processes the host through fomites or environmental release is a crucial determinant of spillover can be represented as the probability that the transport. In this case, the release of a risk, and care must be taken to appropriately pathogen (shed, harvested or colonized in a pathogen from its reservoir host and human formulate models that represent the rate of vector) survives and is infectious at a given exposure to the pathogen may become release for each route of transmission (BOX 1; point in time, and is dispersed or transported disconnected in space and time. An example Supplementary information S1 (box)). to a particular location (Supplementary is the spread of aerosolized Coxiella burnetii a b c Reservoir host distribution Animal ecology, population biology, biogeography, behavioural ecology, landscape ecology, agricultural sciences Reservoir host density Pathogen prevalence Disease ecology, animal epidemiology, infectious disease dynamics, immunology, Infection intensity microbiology, veterinary medicine Pathogen release from reservoir host Microbiology, disease ecology, vector ecology, epidemiology, spatial ecology, Pathogen survival and spread infectious disease dynamics Time Scenario 1 Human epidemiology, medical anthropology, Scenario 2 vector ecology, social sciences, behavioural Threshold Human exposure ecology, infectious disease dynamics Structural barriers Microbiology, innate and adaptive immunology, cell biology of pathogen–host Linear interactions, pathology, genetics, Innate immune response and molecular compatibility Sigmoidal evolutionary biology Threshold Replication and dissemination cycles completed Dose Spillover Figure 2 Barriers to spillover and dose–response relationships. both scenarios, the mean dose over the time interval is the same. Bottom Nature Reviews | Microbiology a | Determinants of spillover are being studied by researchers in many disci - panel: the likelihood that this dose will translate into infection depends on the plines. b | A pathogen must overcome a series of barriers to transmit from one functional form of the dose–response relationship. If the dose–response rela - species to another. If any of these barriers is impenetrable, spill over cannot tionship is linear (green line), these two excretion scenarios generate the same occur. Spillover of some pathogens requires that gaps (depicted as holes) in total probability of spillover over the time interval shown. However, for non - all of the barriers align within a narrow window in space and time (indicated linear dose–response relationships, the total probability of spillover differs by the blue arrow, see Supplementary information S2 (movie)). For other between scenarios. If the relationship is sigmoidal (red line), there is some patho gens, protracted survival in the environment (for example, Bacillus probability of spillover whenever the dose exceeds zero (indicated by the anthracis spores ), or wide dissemination (for example, the spread of aero - intensity of the red shading in the top panel), but the total spillover probability solized Coxiella burnetii by wind ), may stagger the alignment of barriers to in scen ario 2 is markedly higher. In the extreme case in which the recipient spillover. c | Top panel: hypothetical dose available over time for a given patho - host can be infected only by a dose that exceeds a sharp threshold, as sus- 67,68,79 gen. In scenario 1 (dashed light blue line), the pathogen is excreted consist - pected for Bacillus anthracis , the pathogen in scenario 2 will spill over ently from infected reservoir hosts. In scenario 2 (solid light blue line), the when the dose peaks above the threshold (blue solid line near peak), but the pathogen is excreted in regular but short high-intensity pulses over time. In pathogen in scenario 1 will never spill over. 504 | AUGUST 2017 | VOLUME 15 w w w.nature.com/nrmicro © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. Dose Probability of infection PERSPECTIVES by wind, which can lead to outbreaks of Box 1 The mathematics of spillover Q fever in humans that live several kilometres The opportunities for cross-species transmission are influenced by processes that occur at scales from the livestock reservoir hosts . from molecules to landscapes (FIG. 1). These processes are subjects of intense study, and their As illustrated by rabies virus, pathogenic characterization is complicated by their variability in space and time, nonlinear responses and Leptospira spp. and E. coli O157 (FIG. 3), interactions with outside factors. Consequently, it is impossible to integrate all of the determinants the bottlenecks that hinder the transfer of spillover transmission — or to assess the effects of gaps in our knowledge about these of pathogens between species depend determinants — without appropriate tools, such as mathematical and computational models . on the ecology of the reservoir host In Supplementary information S1 (box), we present a general mathematical model of the spillover and the pathogen, and the interactions process, which provides a template for integrating our knowledge of processes for specific disease among the determinants of spillover. For systems. This model framework essentially translates FIG. 1 into mathematical expressions. It allows example, the primary driver of pathogen for variation in space and time, and uses different formulations for transmission through pathogen excretion, slaughter or arthropod vectors. pressure for rabies virus is the prevalence The mathematical model reflects the modular nature of the spillover process, as emphasized in of infection in key hosts (such as domestic 36 the main text, while highlighting dependencies among factors in ways such as the following: dogs ). Nonlinearities in rabies transmission • Factors that are linked to disease ecology of the reservoir host and the mode of pathogen release generate a threshold effect in susceptible host determine the amount of pathogen released to the environment or vector. density below which the pathogen cannot • Pathogen survival and transport outside of the animal host, which give rise to pathogen pressure persist. These thresholds can be used to set at a particular place and time, are modelled with simple probability kernels. vaccination targets for disease elimination . • Human risk behaviours determine how this pathogen pressure translates to exposure dose. By contrast, pathogen pressure of L. interrogans is also affected by fluctuations • The probability of infection for a given dose and route of exposure is encapsulated in the dose–response relationship (FIG. 2c). in reservoir host density (such as rodents ), and prevalence and shedding from infected Mathematically, the focal point of this process is the dose to which the recipient host is exposed. animals . However, if human exposure All upstream factors come together, with appropriate functional dependencies, to shape this dose. To a reasonable approximation, which is consistent with current practice in quantitative microbial occurs through mechanisms that aggregate risk assessment , the consequent risk of infection can be modelled independently through the and disperse pathogens shed by many dose–response relationship. individuals (through accumulation in the environment, sustained survival after exiting the host , and dispersal through rain, rivers and flood waters ), the detailed dynamics occurs after human-mediated dispersal to exposure through different routes of in reservoir hosts do not matter because of the pathogen through irrigation, meat transmission . Human behaviours, such 46–48 they get integrated out by the environmental processing and food transportation . as occupational interactions with reservoir reservoir. In this scenario, spillover risk In this instance, outbreaks of E. coli O157 host animals, the consumption of certain is determined by the aggregate pathogen are determined by the pathogen pressure animal products or the use of particular pressure, human behaviours that determine on vegetables or in hamburger meat, environments, may increase the risk exposure and the integrity of within-human potentially derived from many sources. of infection . barriers to infection. For example, when As the dose that is required for E. coli O157 Exposure is often conceptualized as a 49,50 flooding mobilizes Leptospira spp. during spillover is thought to be very low , public simple point of contact. However, nonlinear the wet season in Brazil, human exposures health policies aim to completely eliminate interactions between pathogen pressure, can become widespread and epidemics pathogen pressure in food that is processed human risk behaviour and environmental 40 50 of spillover infection can occur . During for human consumption . To achieve this factors can lead to unexpected complexity, these extreme environmental events, control goal, interventions are focused on creating especially for vector-borne diseases. For efforts must focus on preventing exposure successive bottlenecks in several barriers to example, in rats, both a high prevalence to contaminated sources (for example, by spillover, including decreasing cattle density, of Y. pestis and high mortality may be wearing protective clothing and boots ) preventing faecal contamination during necessary to drive outbreaks of bubonic and reducing the infectious inoculum meat processing and increasing cooking plague in humans. Widespread exposure rather than reducing the source of pathogen temperatures to reduce exposure dose in of humans through flea bites occurs only 43,47,51 shedding, as the release of Leptospira spp. ground beef . Cumulatively, these efforts after a decrease in the abundance of rats, into the environment by animal reservoirs are usually successful, but high levels of which are the primary hosts of Y. pestis in occurs before the extreme precipitation. shedding from cattle during summer can peridomestic settings . Indeed, historically, Similarly, pathogen pressure of E. coli O157 sometimes overwhelm interventions . high rat mortality (‘rat-fall’) was an is affected by the density of its cattle host indication of an imminent human plague 42 32 population , by variation in shedding Exposure. The next phase of spillover — epidemic . Thus, killing rodents in response among individuals and by prevalence in exposure — bridges the upstream processes to cases of bubonic plague in humans could herds . Each of these factors can be highly that generate pathogen pressure and the inadvertently increase the severity of the 44,45 54 skewed and seasonal . If spillover events within-host processes in the recipient that epidemic . Conversely, and controversially, are driven by contact between humans and determine whether a given dose generates zooprophylaxis, which involves diverting cattle, then variation in pathogen load among a spillover infection (see below). The vector bites from humans by increasing animals would interact with nonlinear dose– interaction between recipient hosts and the local population density of another response functions to determine spillover pathogen pressure determines both the animal host, may decrease the risk of human risk (see below). However, this individual dose and the route of exposure. Different exposure . For example, the presence of variation matters less if human exposure behaviours of the recipient host are relevant chickens and dogs in rural areas of Argentina NATURE REVIEWS | MICROBIOLOGY VOLUME 15 | AUGUST 2017 | 505 © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES decreased the rate at which Triatoma infestans in the dose to which a host is exposed at individual receptivity . Physical barriers transmitted Trypanosoma cruzi, the causative a given location and time (the integral include the skin, mucous membranes, agent of Chagas disease, to humans . of the pathogen pressure in space and mucus, stomach acid or the absence of However, increasing the population density of time to which the host has been exposed functional receptors that enable the pathogen reservoir hosts may also affect vector survival, (Supplementary information S1 (box)). to enter its target cells or tissues . Interferon- vector abundance and pathogen prevalence induced and other innate immune responses in reservoir hosts, which, in turn, increases Probability of infection. Following may be triggered after the initial infection pathogen pressure and offsets reductions cross-species exposure of a recipient host, the of a cell, resulting in protective mechanisms 56,57 in human–vector contact rates . These within-host barriers and their interactions such as apoptosis or the induction of complexities highlight the need to understand with the strain of pathogen determine the interferon-induced resistance in surrounding the mechanisms that contribute to particular functional relationship between the pathogen cells . In addition, interfering defensive routes of spillover. dose and the likelihood that an infection will proteins in the host cell cytoplasm may block All of the factors that precede human establish. Within-host barriers to infection the replication of intracellular pathogens. In exposure, mediated by human behaviour vary widely and depend on the specific other cases, cells lack functional host factors and environmental factors (FIG. 1), cumulate combinations of pathogen, host species and that are required for the replication of the Rabies virus Leptospira interrogans Escherichia coli O157 Toxoplasma gondii Ebola virus Reservoir distribution Reservoir density Pathogen prevalence Infection intensity ? ? Pathogen release ? ? Pathogen survival and spread ? ? Human exposure ? ? Within-host barriers 43,44 Figure 3 Bottlenecks to spillover. Different barriers permit or constrain the heterogeneous shedding from cattle (although it is still unknown whether Nature Reviews | Microbiology flow of pathogens from one species to another. The figure is illustrative, super-shedding is a characteristic of particular individuals or is a transient owing to the lack of sufficient data for more than one or two barriers for any phase that occurs in most cattle ). In some contexts, exposure is an impor- given system. The width of the gaps in barriers represents the ease with tant bottleneck; for example, when the pathogen is eliminated from food which a pathogen can flow through the barriers and will vary depending on through cooking. Widespread dispersal leads to uncertainties about the 46,47 context. The question marks represent points at which the barriers are source of many outbreaks , and weak within-human barriers enable low 49,50 especi ally poorly understood and highlight gaps in our knowledge of some doses of E. coli to cause infection . Humans are frequently exposed to patho gens that are of global concern (for example, the lack of information Toxoplasma gondii carried by domestic cats and intermediate hosts, but the on disease dynamics in reservoir hosts of Ebola virus). Many rabies virus reser - parasite rarely causes disease because most humans have strong within-host voirs, such as domestic dogs, are widely distributed. The prevalence of rabies immunological barriers. Cats are widely and densely distributed, but the virus is generally low and the incidence of spillover closely tracks the preva - prevalence of T. gondii is low and cats shed oocysts only once in their life- lence of infection in the reservoir host. Rabies virus is almost always fatal to time . However, sporulated oocysts survive in the environment for long 25 112 spillover hosts . Interventions are usually aimed at reducing the preva lence periods of time . Limiting exposure to oocysts may prevent spillover; how- in reservoir hosts through vaccination . Leptospira interrogans survives in ever, this is challenging when it is unclear whether cats or the environment 111,113 water and soil after being shed in the urine of a wide range of rodents and are the major sources of infection in humans . Ebola virus has not been 29 114 other reservoir hosts . Key bottlenecks to the zoonotic spillover of this isolated from bats and the definitive reservoir bat species is unknown ; 114,115 pathoge n are exposure and within-host barriers. For example, during floods therefore, characteristics of infection in bats are unknown . The patho- in Brazil, many humans that are exposed do not become infected, probably gen is released through excretion or slaughter, then survives for up to a week, 41 116 because the initial within-host barrier, the skin, is not penetrated . However, depending on the environmental conditions . The most tractable bottle- once L. interrogans penetrates the skin (for example, through skin wounds), necks for intervention may be the zoonotic exposure of humans through 110 97,117,118 1–10 leptospires may be sufficient to cause systemic infection . Therefore, interaction with bats, bushmeat or the carcasses of other species , wearing protective clothing and boots is an effective control measure . because once exposed, the within-host barriers to Ebola virus may be Important bottlenecks to Escherichia coli O157 spillover include extremely low . 506 | AUGUST 2017 | VOLUME 15 w w w.nature.com/nrmicro © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES 60,61 pathogen . Even when pathogens can to a low but constant dose may generate the Once a pathogen has penetrated the replicate within cells, several barriers same probability of infection as intermittent within-host barriers to replicate and can prevent their transmission to other high-intensity exposures (FIG. 2c). disseminate in the new host, the outcome 62,63 cells and thus the establishment of an The genetic, immunological and of the infection may range from subclinical infection. For example, avian influenza virus physiological state of the host also can elimination of the microorganism to must pass through a series of within-host modulate the dose–response relationship. the death of the new host, and from barriers to infect a human, including mucins Immunosuppression (for example, due dead-end spillover infection to sustained in respiratory tract excretions, specific to AIDS, immunosuppressive drugs, human-to-human transmission. For many receptor molecules that constrain virus entry co-infections or malnutrition) increases important zoonotic pathogens, such as into cells and have different distributions gaps in within-host barriers, which shifts HIV or Zika virus, the transmission that in the respiratory tracts of different host dose–response curves and increases drives the current public health crisis is 69,70 81,82 species, suboptimal viral polymerase that susceptibility . For example, in human-to-human and the events that restricts the ability of the virus to replicate immunosuppressed hosts, the decreased led to spillover are long past. Although in cells of the human respiratory tract, viral number or activity of lymphocytes can understanding disease severity and onward neuraminidase that is inefficient in its role in reduce the dose that is required to establish transmission is essential for understanding the release of influenza viruses from infected an infection with the widespread pathogen the consequences of emerging infectious cells, and innate immune responses that are Toxoplasma gondii, or cause the loss of control diseases, these processes are beyond the initiated early and that block infection in of T. gondii infections that are usually kept in scope of this article. Our current knowledge 63,64 71 both infected and neighbouring cells . check by sustained immune pressure (FIG. 3). of the biological features of pathogens and From an epidemiological perspective, Seasonality in human immune function (for characteristics of host–pathogen interactions these within-host interactions between example, enhanced baseline inflammation and that determine these outcomes are described zoonotic pathogens and hosts can be altered cellular composition of the immune elsewhere (for example, see REFS 83,84). encapsulated by the functional relationship system in winter compared with summer) between pathogen dose and the probability may also alter the permeability of within-host Assessing zoonotic risk of an infection. Although there is much to barriers by altering the magnitude and speed When gaps in barriers to spillover are highly learn about dose–response relationships, of immune responses . Finally, the probability dynamic in time and space, they may vary they are expected to be nonlinear as, at and severity of infection at a given dose are asynchronously, so that the alignment of minimum, they must saturate at high doses shaped by host genetics ; triathletes with gaps in all barriers may be fleeting and because the probability of infection cannot a particular gene polymorphism were at spillover may seem random (Supplementary exceed one . This nonlinearity imposes a increased risk of leptospirosis after swallowing information S2 (movie)). Research methods filter on the dynamics of pathogen pressure lake water compared with athletes who lacked that group multiple barriers or integrate and exposure (FIG. 2c). If the dose–response this polymorphism . data over space and time may not capture relationship is highly nonlinear, such that Many of the interactions at the crossroads these dynamics. For example, ecological small changes in dose lead to large changes of exposure, inoculum dose and host niche models are often used to study in the probability of an infection, then response are poorly understood. Therefore, zoonotic risk by assessing the distribution of variation in any of the upstream factors that very little is known about the interactions reservoir hosts or vectors , but this approach culminate in an exposure dose (including between dose, timing of exposure and overlooks variation in downstream barriers released dose, pathogen survival and human probability of infection. The current that might drive risk. Alternatively, niche behaviour) may have disproportionate effects dose–response paradigm is based on models that are based on the documented on the probability of spillover. Such effects discrete transient exposures, but the effects occurrence of spillover may capture the could generate opportunities for targeted of protracted or cumulative exposure to accumulated distribution of all conditions control measures. Moreover, nonlinear environmental pathogens (for example, that enabled barriers to be breached over dose–response relationships may imply to low concentrations of Leptospira spp. in time (FIG. 1), but they cannot isolate the that infrequent high-intensity exposures floodwater) are unclear . Repeated low-dose precise barriers that affect spillover risk are more likely to cause spillover infections exposure can increase host immunity (for example, see REF. 86). Therefore, niche than continuous low-intensity excretion. to infection (for example, as postulated models tend to overestimate the spatial range This phenomenon has been reported for for poultry handlers who are exposed to of spillover risk and do not readily enable 76 87 occupational exposure to Bacillus anthracis avian influenza , dairy farmers who are extrapolation to novel conditions . Examples aerosols; tannery workers who were exposed exposed to E. coli O157 (REF. 77) and mice of this include Hendra virus and Marburg to infrequent high doses of B. anthracis that are exposed to continuous infections of virus, which can be excreted in discrete spores in imported goat hair were more parasites ). However, increases in immunity temporal and spatial pulses from their bat 20,88,89 likely to die of anthrax than those who were are not always observed; for example, such reservoir hosts . However, for spillover, exposed to frequent low doses of B. anthracis effects on immunity were not observed shedding must align with environmental and 66–68 spores . Conversely, if doses are far below in tannery workers who were exposed bat population conditions that generate levels 67,68,79 the inflection point on the dose–response to B. anthracis . Moreover, it may be of pathogen pressure that are sufficient to curve (FIG. 2c), then the system may be difficult to differentiate between a cumulative produce an infectious dose (FIG. 2), and with insensitive to changes in dose. If the dose– dose effect and the increasing opportunity exposure behaviours and susceptibility of the response function is close to linear, the total to initiate an infection with each additional recipient hosts. As some of these conditions exposure dose over time is equal and host low-dose exposure (if each infectious unit vary among seasons and years, the pattern responses do not change as a consequence of has a probability of causing an infection that of outbreaks in livestock or humans has high 20,80 20,89 early exposures, then longer-term exposure is above zero) . spatial and temporal variability . However, NATURE REVIEWS | MICROBIOLOGY VOLUME 15 | AUGUST 2017 | 507 © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES as niche models often summarize risk across (for example, common exposure to infected Outlook large areas and long durations, they overlook animal hosts and tsetse fly vectors, and low The framework presented in this Opinion important heterogeneities and they lack the resistance in humans due to the ability of article highlights that an important frontier specificity that is required for public health trypanosomes to neutralize or avoid human in research on zoonotic spillover is to 98,99 intervention. Although niche models can innate immune activity ). In all scenarios, understand the functional and quantitative help to identify regional-to-continental irrespective of the frequency with which gaps links among the determinants of spillover. 90,91 concentrations of risk , risk assessments align, the concept of hierarchical barriers To our knowledge, all of the processes that that are more quantitative and more precise can be used to organize and quantify the are necessary to achieve spillover have not with regard to space, time and which conditions that enable spillover. been connected, compared and quantified barriers they address are needed to guide The influence of particular barriers may for any single zoonotic pathogen. We concrete action. vary in space and time, and this variation address this gap, in part, by introducing Epidemiological investigations of spillover — coupled with data on realized spillover a conceptual and quantitative model that also need to account for conditions that are events — can help elucidate factors that can be used to integrate existing data, highly dynamic in space and time. If the shape infection risk, even in the absence identify high-priority data gaps, investigate alignment of gaps in all barriers is fleeting, of information on other barriers. In the conditions that widen or align gaps in delayed diagnoses or inconsistent case westernmost province of the Democratic barriers to spillover, and identify the best detection may delay outbreak investigations Republic of Congo, the observed lack gaps on which to focus intervention efforts. until the conditions that enabled spillover of monkeypox spillover, despite high We suggest that future research focuses have changed. Similarly, investigations are seroprevalence in the suspected reservoir on developing case studies that contribute sometimes triggered once the case count hosts (Heliosciurus spp. and Funisciuris to fully quantifying the determinants of becomes high. These challenges differ among spp.), was attributed to cultural norms that spillover and their linkages, with the goal pathogens with different values of R (the forbade the consumption of small rodents . of making operational contributions to risk basic reproductive number or expected The inconsistency between ecological data assessment. We provide a mathematical number of secondary infections caused by that suggested high pathogen pressure and framework that formalizes the ideas a typical infected individual in a susceptible epidemiological data that indicated a lack presented here to guide the formulation of population). For supercritical pathogens of spillover, focused attention on human mechanistic spillover models for particular with R >1, which can cause major epidemics behaviours that affect the probability of zoonotic pathogens (BOX 1; Supplementary through sustained transmission in human exposure. Research approaches that integrate information S1 (box)). We anticipate that populations (for example, Ebola virus, Zika data on multiple barriers are more likely to this synthetic framework will provide a virus and the pandemic strain of severe acute discern such behavioural effects. foundation for cross-scale data integration, respiratory syndrome coronavirus (SARS- Broad-scale discovery of novel transdisciplinary investigation, and a 4,81,92 CoV) ), spillover becomes challenging to microorganisms has the potential to new body of theory on spillover that is study because a given human case is likely characterize the pool of possible zoonotic necessary for risk assessment and public to be far removed in time or space from the pathogens and provide valuable baseline health planning. 101,102 spillover event that triggered an outbreak. information . However, each of the Raina K. Plowright is at the Department of Microbiology Subcritical pathogens with 0 <R <1, which ~63,000 species of mammals, birds, and Immunology, Montana State University, Bozeman, cause self-limiting outbreaks or ‘stuttering reptiles, amphibians and fish contains a Montana 59717, USA. chains’ in human populations (for example, multitude of infectious viruses, bacteria and Colin R. Parrish is at the Baker Institute for Animal 93,94 101,102,104–106 monkeypox or avian influenza viruses ), parasites . Although each of these Health, College of Veterinary Medicine, raise distinct challenges because any given microorganisms and parasites can be viewed Cornell University, Ithaca, New York 14853, USA. individual could have been infected by either as a potential pathogen, the vast majority Hamish McCallum is at the Griffith School of an animal or a human source . It is easiest to may not cause disease in their natural Environment, Griffith University, Brisbane, study the spillover of pathogens with R = 0 hosts, and the extent to which they infect or Queensland 4111, Australia. that are not transmitted between humans cause pathology in other species, including Peter J. Hudson is at the Center for Infectious 7,9,10 (for example, rabies virus or West Nile humans, is unknown . Therefore, Disease Dynamics, Pennsylvania State University, 25,95 virus ), in which every case is an instance discovery alone cannot address the potential State College, Pennsylvania 16802, USA. of spillover. The 2014–2015 Ebola virus risk of spillover. The translation of new Albert I. Ko is at the Department of Epidemiology of epidemic in West Africa is a prime example discoveries of microorganisms into guidance Microbial Diseases, Yale School of Public Health, whereby delayed response and investigation for public health practitioners requires the New Haven, Connecticut 06520–8034, USA. prevented researchers from reconstructing identification of the barriers to microbial Andrea L. Graham is at the Department of Ecology the conditions that initiated the human infection of humans, the conditions that & Evolutionary Biology, Princeton University, 96,97 epidemic of a supercritical pathogen . facilitate the breaching of these barriers, Princeton, New Jersey 08544, USA. Ebola virus infection is an extreme example and, therefore, the microbiological and James O. Lloyd-Smith is at the Department of Ecology & of spillover infection that only occurs during environmental contexts that pose the Evolutionary Biology, University of California, the rare alignment of gaps in barriers, and, greatest risk to human populations. For the Los Angeles, Los Angeles, California 90095-7239, USA; and at Fogarty International Center, National Institutes accordingly, the precise determinants of risk foreseeable future, the greatest practical of Health, Bethesda, Maryland 20892–2220, USA. are poorly understood (FIG. 3). By contrast, contribution of pathogen discovery for other zoonoses, such as trypanosomiasis and sequence characterization to the Correspondence to R.K.P. [email protected] in some parts of Africa, incidence is high epidemiology of emerging pathogens is likely because the pathogen flows through to be in the rapid post hoc identification of doi:10.1038/nrmicro.2017.45 consistently wide gaps in barriers to infection novel pathogens after spillover. Published online 30 May 2017 508 | AUGUST 2017 | VOLUME 15 w w w.nature.com/nrmicro © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES 1. Christou, L. The global burden of bacterial and viral 26. Webster, R. in Viral Zoonoses and Food of Animal 49. Teunis, P., Ogden, I. & Strachan, N. Hierarchical dose Origin (eds Kaaden, O.‑ R., Czerny, C.‑ P. & zoonotic infections. Clin. Microbiol. Infect. 17, response of E. coli O157: H7 from human outbreaks 326–330 (2011). Eichhorn, W.) 105–113 (Springer, 1997). incorporating heterogeneity in exposure. Epidemiol. 2. Morens, D. M., Folkers, G. K. & Fauci, A. S. The 27. Ko, A. I., Goarant, C. & Picardeau, M. Leptospira: the Infect. 136, 761–770 (2008). challenge of emerging and re‑emerging infectious dawn of the molecular genetics era for an emerging 50. Tuttle, J. et al. Lessons from a large outbreak of diseases. Nature 430, 242–249 (2004). zoonotic pathogen. Nat. Rev. Microbiol. 7, 736–747 Escherichia coli O157:H7 infections: insights into (2009). 3. Jones, K. E. et al. Global trends in emerging infectious the infectious dose and method of widespread diseases. Nature 451, 990–993 (2008). 28. Costa, F. et al. Influence of household rat infestation contamination of hamburger patties. Epidemiol. Infect. This study analyses the general phylogenetic and on Leptospira transmission in the urban slum 122, 185–192 (1999). environment. PLoS Negl. Trop. Dis. 8, e3338 (2014). geographical risk factors for many different 51. Cobbold, R. N. et al. Rectoanal junction colonization of emerging diseases, as well as temporal and spatial 29. Costa, F. et al. Patterns in Leptospira shedding in feedlot cattle by Escherichia coli O157: H7 and its trends in emerging infections. Norway rats (Rattus norvegicus) from Brazilian slum association with supershedders and excretion communities at high risk of disease transmission. 4. Briand, S. et al. The international Ebola emergency. dynamics. Appl. Environ. Microbiol. 73, 1563–1568 N. Engl. J. Med. 371, 1180–1183 (2014). PLoS Negl. Trop. Dis. 9, e0003819 (2015). (2007). 5. Smith, G. J. et al. Origins and evolutionary genomics 30. Monahan, A. M., Callanan, J. J. & Nally, J. E. 52. Cascio, A., Bosilkovski, M., Rodriguez‑ Morales, A. & Proteomic analysis of Leptospira interrogans shed in of the 2009 swine‑ origin H1N1 influenza A epidemic. Pappas, G. The socio‑ ecology of zoonotic infections. Nature 459, 1122–1125 (2009). urine of chronically infected hosts. Infect. Immun. 76, Clin. Microbiol. Infect. 17, 336–342 (2011). 6. Fevre, E. M., Wissmann, B. V., Welburn, S. C. 4952–4958 (2008). 53. Macpherson, C. N. Human behaviour and the 31. Nally, J. E., Chow, E., Fishbein, M. C., Blanco, D. R. & Lutumba, P. The burden of human African epidemiology of parasitic zoonoses. Int. J. Parasitol. trypanosomiasis. PLoS Negl. Trop. Dis. 2, e333 & Lovett, M. A. Changes in lipopolysaccharide O 35, 1319–1331 (2005). (2008). antigen distinguish acute versus chronic Leptospira 54. Keeling, M. J. & Gilligan, C. A. Metapopulation 7. Grice, E. A. & Segre, J. A. The skin microbiome. interrogans infections. Infect. Immun. 73, 3251–3260 dynamics of bubonic plague. Nature 407, 903–906 Nat. Rev. Microbiol. 9, 244–253 (2011). (2005). (2000). 32. Smego, R., Frean, J. & Koornhof, H. Yersiniosis I: 8. Guarner, F. & Malagelada, J.‑R. Gut flora in health and This study uses dynamic models to explain disease. Lancet 361, 512–519 (2003). microbiological and clinicoepidemiological aspects of historical patterns of bubonic plague, and shows 9. Gilbert, S. F., Sapp, J. & Tauber, A. I. A symbiotic view plague and non‑ plague Yersinia infections. Eur. J. Clin. that, counterintuitively, culling rats may exacerbate Microbiol. Infect. Dis. 18, 1–15 (1999). of life: we have never been individuals. Q. Rev. Biol. plague. 87, 325–341 (2012). 33. Weber, T. P. & Stilianakis, N. I. Inactivation of 55. Hess, A. & Hayes, R. O. Relative potentials of domestic 10. Parrish, C. R. et al. Cross‑ species virus transmission influenza A viruses in the environment and modes of animals for zooprophylaxis against mosquito vectors transmission: a critical review. J. Infect. 57, 361–373 and the emergence of new epidemic diseases. of encephalitis. Am. J. Trop. Med. Hyg. 19, 327–334 Microbiol. Mol. Biol. Rev. 72, 457–470 (2008). (2008). (1970). This article reviews the general features that are 34. Koopmans, M. et al. Transmission of H7N7 avian 56. Gürtler, R. E. et al. Domestic animal hosts strongly influenza A virus to human beings during a large associated with the emergence of viruses in new influence human‑ feeding rates of the Chagas disease hosts to cause epidemics or pandemics. outbreak in commercial poultry farms in the vector Triatoma infestans in Argentina. PLoS Negl. 11. Woolhouse, M. E. & Gowtage‑ Sequeria, S. Host range Netherlands. Lancet 363, 587–593 (2004). Trop. Dis. 8, e2894 (2014). 35. Tissot‑ Dupont, H., Amadei, M.‑A., Nezri, M. & and emerging and reemerging pathogens. 57. Kilpatrick, A. M. & Randolph, S. E. Drivers, dynamics, Emerg. Infect. Dis. 11, 1842–1847 (2005). Raoult, D. Wind in November, Q fever in December. and control of emerging vector‑ borne zoonotic 12. Taylor, L. H., Latham, S. M. & Woolhouse, M. E. J. Emerg. Infect. Dis. 10, 1264 (2004). diseases. Lancet 380, 1946–1955 (2012). Risk factors for human disease emergence. 36. Hampson, K. et al. Synchronous cycles of domestic 58. Schmid‑ Hempel, P. Variation in immune defence as a Phil. Trans. R. Soc. Lond. B Biol. Sci. 356, 983–989 dog rabies in sub‑ Saharan Africa and the impact of question of evolutionary ecology. Proc. Biol. Sci. 270, control efforts. Proc. Natl Acad. Sci. USA 104, (2001). 357–366 (2003). 13. Morse, S. S. Factors in the emergence of infectious 7717–7722 (2007). 59. Akira, S., Uematsu, S. & Takeuchi, O. Pathogen diseases. Emerg. Infect. Dis. 1, 7–15 (1995). 37. Brochier, B. et al. Large‑ scale eradication of rabies recognition and innate immunity. Cell 124, 783–801 using recombinant vaccinia–rabies vaccine. Nature 14. Lloyd‑ Smith, J. O. et al. Epidemic dynamics at the (2006). human–animal interface. Science 326, 1362–1367 354, 520–522 (1991). 60. Trobaugh, D. W. & Klimstra, W. B. MicroRNA (2009). 38. Andre‑ Fontaine, G., Aviat, F. & Thorin, C. Waterborne regulation of RNA virus replication and pathogenesis. leptospirosis: survival and preservation of the This study delineates stages of zoonoses on the Trends Mol. Med. 23, 80–93 (2017). basis of changes in transmissibility, as reflected virulence of pathogenic Leptospira spp. in fresh water. 61. Duggal, N. K. & Emerman, M. Evolutionary conflicts in R . It also reviews the literature on modelling Curr. Microbiol. 71, 136–142 (2015). between viruses and restriction factors shape 39. Lau, C. L., Smythe, L. D., Craig, S. B. & Weinstein, P. transmission dynamics of zoonoses and identifies immunity. Nat. Rev. Immunol. 12, 687–695 gaps in our knowledge. Climate change, flooding, urbanisation and (2012). 15. Johnson, C. K. et al. Spillover and pandemic leptospirosis: fuelling the fire? Trans. R. Soc. Trop. 62. Air, G. M. & Laver, W. G. The neuraminidase of Med. Hyg. 104, 631–638 (2010). properties of zoonotic viruses with high host plasticity. influenza virus. Proteins 6, 341–356 (1989). Sci. Rep. 5, 14830 (2015). 40. Reis, R. B. et al. Impact of environment and social 63. Kuiken, T. et al. Host species barriers to influenza virus 16. Lloyd‑ Smith, J. O., Funk, S., McLean, A. R., Riley, S. gradient on Leptospira infection in urban slums. infections. Science 312, 394–397 (2006). & Wood, J. L. Nine challenges in modelling the PLoS Negl. Trop. Dis. 2, e228 (2008). 64. Lipsitch, M. et al. Viral factors in influenza pandemic emergence of novel pathogens. Epidemics 10, 35–39 41. Phraisuwan, P. et al. Leptospirosis: skin wounds and risk assessment. eLife 5, e18491 (2016). control strategies, Thailand, 1999. Emerg. Infect. Dis. (2015). 65. Schmid‑ Hempel, P. & Frank, S. A. Pathogenesis, 17. Gortazar, C. et al. Crossing the interspecies barrier: 8, 1455–1459 (2002). virulence, and infective dose. PLoS Pathog. 3, e147 opening the door to zoonotic pathogens. PLoS Pathog. 42. Spencer, S. E., Besser, T. E., Cobbold, R. N. & (2007). French, N. P. ‘Super’or just ‘above average’? 10, e1004129 (2014). 66. Brachman, P. S. & Fekety, F. R. Industrial anthrax. 18. Wolfe, N. D., Dunavan, C. P. & Diamond, J. Origins Supershedders and the transmission of Escherichia Ann. NY Acad. Sci. 70, 574–584 (1958). of major human infectious diseases. Nature 447, coli O157: H7 among feedlot cattle. J. R. Soc. 67. Brachman, P. S., Kaufman, A. & Dalldorf, F. G. Interface 12, 0446 (2015). 279–283 (2007). Industrial inhalation anthrax. Bacteriol. Rev. 30, 646 19. Pepin, K. M., Lass, S., Pulliam, J. R., Read, A. F. & This study examines the dynamics of E. coli (1966). Lloyd‑ Smith, J. O. Identifying genetic markers of transmission and the roles of super-shedder This study is one of the only studies to calculate the individuals in those processes. adaptation for surveillance of viral host jumps. risk of spillover infection using comparable doses Nat. Rev. Microbiol. 8, 802–813 (2010). 43. Matthews, L. et al. Heterogeneous shedding of administered over time and provides evidence for 20. Plowright, R. K. et al. Ecological dynamics of emerging Escherichia coli O157 in cattle and its implications for the outcome of repeated low-dose versus single control. Proc. Natl Acad. Sci. USA 103, 547–552 bat virus spillover. Proc. R. Soc. B Biol. Sci. 282, high-dose exposure. 20142124 (2015). (2006). 68. Coleman, M. E., Thran, B., Morse, S. S., Hugh‑ This study outlines the conditions that enable 44. Hancock, D., Besser, T., Rice, D., Herriott, D. & Tarr, P. Jones, M. & Massulik, S. Inhalation anthrax: dose spillover of bat viruses into other hosts and A longitudinal study of Escherichia coli O157 in response and risk analysis. Biosecur. Bioterror. 6, provides an example of the infections that are fourteen cattle herds. Epidemiol. Infect. 118, 147–160 (2008). 193–195 (1997). the subject of this review. 69. Bollaerts, K. et al. Human salmonellosis: estimation 21. Hudson, P. J., Rizzoli, A. R., Grenfell, B. T., 45. Besser, T. E., Davis, M. A. & Walk, S. T. in Population of dose–illness from outbreak data. Risk Anal. 28, Heesterbeek, H. & Dobson, A. P. The Ecology of Genetics of Bacteria: A Tribute to Thomas S. Whittam 427–440 (2008). (eds Walk, S. T. & Feng, P. C. H.) 303–324 (2011). Wildlife Diseases (Oxford Univ. Press, 2002). 70. Kau, A. L., Ahern, P. P., Griffin, N. W., Goodman, A. L. 22. Hjelle, B. & Glass, G. E. Outbreak of hantavirus 46. Gerba, C. P. & Smith, J. E. Sources of pathogenic & Gordon, J. I. Human nutrition, the gut microbiome infection in the Four Corners region of the United microorganisms and their fate during land application and the immune system. Nature 474, 327–336 of wastes. J. Environ. Qual. 34, 42–48 (2005). States in the wake of the 1997–1998 El Nino— (2011). Southern Oscillation. J. Infect. Dis. 181, 1569–1573 47. Elder, R. O. et al. Correlation of enterohemorrhagic 71. Greene, C. E. Infectious Diseases of the Dog and Cat (2000). Escherichia coli O157 prevalence in feces, hides, and (Elsevier Health Sciences, 2013). carcasses of beef cattle during processing. Proc. Natl 23. Thoen, C. O., Steele, J. H. & Kaneene, J. B. Zoonotic 72. Dopico, X. C. et al. Widespread seasonal gene Tuberculosis: Mycobacterium bovis and Other Acad. Sci. USA 97, 2999–3003 (2000). expression reveals annual differences in human Pathogenic Mycobacteria (John Wiley & Sons, 2014). This study calculates the decreasing pathogen immunity and physiology. Nat. Commun. 6, 7000 pressure (availability for human exposure) of E. coli 24. Ducatez, M., Webster, R. & Webby, R. Animal influenza (2015). epidemiology. Vaccine 26, D67–D69 (2008). O157 as carcasses progress through the various 73. Gingles, N. A. et al. Role of genetic resistance in 25. Rupprecht, C. E., Hanlon, C. A. & Hemachudha, T. stages of processing at meat processing plants. invasive pneumococcal infection: identification and Rabies re‑examined. Lancet Infect. Dis. 2, 327–343 48. Pennington, H. Escherichia coli O157. Lancet 376, study of susceptibility and resistance in inbred mouse (2002). 1428–1435 (2010). strains. Infect. Immun. 69, 426–434 (2001). NATURE REVIEWS | MICROBIOLOGY VOLUME 15 | AUGUST 2017 | 509 © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved. PERSPECTIVES viruses: a processbased per ‑ spective. Am. Nat. 187, 111. Elmore, S. A. et al. Toxoplasma gondii: epidemiology, 74. Lingappa, J. et al. HLA‑DQ6 and ingestion of contaminated water: possible gene–environment E53–E64 (2016). feline clinical aspects, and prevention. Trends interaction in an outbreak of leptospirosis. 92. Ksiazek, T. G. et al. A novel coronavirus associated Parasitol. 26, 190–196 (2010). with severe acute respiratory syndrome. 112. Dubey, J. Toxoplasma gondii oocyst survival under Genes Immun. 5, 197–202 (2004). 75. Pujol, J. M., Eisenberg, J. E., Haas, C. N. & N. Engl. J. Med. 348, 1953–1966 (2003). defined temperatures. J. Parasitol. 84, 862–865 Koopman, J. S. The effect of ongoing exposure 93. Hutin, Y. et al. Outbreak of human monkeypox, (1998). dynamics in dose response relationships. Democratic Republic of Congo, 1996 to 1997. 113. Jones, J. & Dubey, J. Waterborne toxoplasmosis — PLoS Comput. Biol. 5, e1000399 (2009). Emerg. Infect. Dis. 7, 434 (2001). recent developments. Exp. Parasitol. 124, 10–25 94. Li, Q. et al. Epidemiology of human infections with (2010). 76. Yang, W. & Shaman, J. Does exposure to poultry and wild fowl confer immunity to H5N1? Chin. Med. J. avian influenza A (H7N9) virus in China. 114. Leroy, E. M. et al. Fruit bats as reservoirs of Ebola 127, 3335 (2014). N. Engl. J. Med. 370, 520–532 (2014). virus. Nature 438, 575–576 (2005). 95. Hayes, E. B. et al. Epidemiology and transmission 115. Pourrut, X. et al. Spatial and temporal patterns of 77. Reymond, D. et al. Neutralizing antibodies to Escherichia coli vero cytotoxin 1 and antibodies to dynamics of West Nile virus disease. Emerg. Infect. Zaire ebolavirus antibody prevalence in the possible O157 lipopolysaccharide in healthy farm family Dis. 11, 1167–1173 (2005). reservoir bat species. J. Infect. Dis. 196, S176–S183 96. Baize, S. et al. Emergence of Zaire Ebola virus disease (2007). members and urban residents. J. Clin. Microbiol. 34, 2053–2057 (1996). in Guinea — preliminary report. N. Engl. J. Med. 371, 116. Prescott, J. et al. Postmortem stability of Ebola virus. 78. Scott, M. High transmission rates restore expression 1418–1425 (2014). Emerg. Infect. Dis. 21, 856 (2015). 97. Saéz, A. M. et al. Investigating the zoonotic origin of 117. Leroy, E. M. et al. Human Ebola outbreak resulting of genetically determined susceptibility of mice to nematode infections. Parasitology 132, 669–679 the West African Ebola epidemic. EMBO Mol. Med. 7, from direct exposure to fruit bats in Luebo, (2006). 17–23 (2015). Democratic Republic of Congo, 2007. Vector Borne 98. Kieft, R. et al. Mechanism of Trypanosoma brucei Zoonotic Dis. 9, 723–728 (2009). 79. Cohen, M. L. & Whalen, T. Implications of low level human exposure to respirable B. anthracis. gambiense (group 1) resistance to human 118. Leroy, E. M. et al. Multiple Ebola virus transmission Appl. Biosafety 12, 109 (2007). trypanosome lytic factor. Proc. Natl Acad. Sci. USA events and rapid decline of central African wildlife. 80. French, N., Kelly, L., Jones, R. & Clancy, D. Dose– 107, 16137–16141 (2010). Science 303, 387–390 (2004). response relationships for foot and mouth disease in 99. Simarro, P. P. et al. Estimating and mapping the 119. Judson, S., Prescott, J. & Munster, V. Understanding population at risk of sleeping sickness. PLoS Negl. Ebola virus transmission. Viruses 7, 511–521 (2015). cattle and sheep. Epidemiol. Infect. 128, 325–332 (2002). Trop. Dis. 6, e1859 (2012). 81. Faria, N. R. et al. Zika virus in the Americas: early 100. Jezek, Z. & Fenner, F. in Monographs in Virology Acknowledgements Vol. 17 (ed. Melnick, J. L.) 119–121 (Karger, 1988). The authors thank J. Wood and E. Fleishman for helpful epidemiological and genetic findings. Science 352, 345–349 (2016). 101. Anthony, S. J. et al. A strategy to estimate unknown comments and conversations. R.K.P. and H.M. are sup‑ 82. Hahn, B. H., Shaw, G. M., De Cock, K. M. & viral diversity in mammals. mBio 4, e00598‑13 ported by the Commonwealth of Australia, the State of New (2013). South Wales and the State of Queensland under the Sharp, P. M. AIDS as a zoonosis: scientific and public health implications. Science 287, 607–614 (2000). This study estimates the number of viruses from National Hendra Virus Research Program, awarded through 83. Geoghegan, J. L., Senior, A. M., Di Giallonardo, F. nine viral families in one bat host, and uses that the Rural Industries Research and Development to extrapolate and estimate that there would be Corporation. R.K.P. is supported by the US National & Holmes, E. C. Virological factors that increase the transmissibility of emerging human viruses. 320,000 viruses from those families in mammals. Institutes of General Medical Sciences IDeA Program (grants Proc. Natl Acad. Sci. USA 113, 4170–4175 (2016). 102. Temmam, S., Davoust, B., Berenger, J.‑M., Raoult, D. P20GM103474 and P30GM110732), P. Thye, the Morris & Desnues, C. Viral metagenomics on animals as a Animal Foundation, Montana University System Research This study identifies and quantifies biological features of viruses that best determine human tool for the detection of zoonoses prior to human Initiative (grant 51040‑MUSRI2015‑03), a Defense infection and transmissibility between humans. infection? Int. J. Mol. Sci. 15, 10377–10397 Advanced Research Projects Agency (DARPA) Young Faculty 84. Casadevall, A. & Pirofski, L. Host–pathogen (2014). Award and the US Department of Defense Strategic interactions: the attributes of virulence. J. Infect. Dis. 103. Hoffmann, M. et al. The impact of conservation on Environmental Research and Development Program the status of the world’s vertebrates. Science 330, (SERDP; grant RC‑2633). J.O.L.‑S. is supported by the US 184, 337–344 (2001). 85. Miller, R. H. et al. Ecological niche modeling to 1503–1509 (2010). National Science Foundation (NSF; grants OCE‑1335657 estimate the distribution of Japanese encephalitis 104. Ley, R. E. et al. Evolution of mammals and their gut and DEB‑1557022) and the US Department of Defense microbes. Science 320, 1647–1651 (2008). SERDP (grant RC‑2635). J.O.L.‑S., A.L.G. and P.J.H. are virus in Asia. PLoS Negl. Trop. Dis. 6, e1678 (2012). 86. Levine, R. S. et al. Ecological niche and geographic 105. Turnbaugh, P. J. et al. The human microbiome project: supported by the RAPIDD program of the Science & distribution of human monkeypox in Africa. PLoS ONE exploring the microbial part of ourselves in a changing Technology Directorate of the Department of Homeland world. Nature 449, 804 (2007). Security, the Fogarty International Center (part of the US 2, e176 (2007). 87. Kearney, M., Simpson, S. J., Raubenheimer, D. & 106. Dobson, A., Lafferty, K. D., Kuris, A. M., National Institutes of Health), and by IDEAS (Infectious Helmuth, B. Modelling the ecological niche from Hechinger, R. F. & Jetz, W. Homage to Linnaeus: how Disease Evolution Across Scales), which is a Research many parasites? How many hosts? Proc. Natl Acad. Coordination Network (DEB‑1354890) funded by the US functional traits. Phil. Trans. R. Soc. B Biol. Sci. 365, 3469–3483 (2010). Sci. USA 105, 11482–11489 (2008). National Science Foundation. 88. Plowright, R. K. et al. Transmission or withinhost ‑ 107. Heesterbeek, H. et al. Modeling infectious disease dynamics in the complex landscape of global health. Competing interests statement dynamics driving pulses of zoonotic viruses in reservoirhost populations ‑ . PLoS Negl. Trop. Dis. 10, Science 347, aaa4339 (2015). The authors declare no comparing interests. e0004796 (2016). 108. Haas, C. N., Rose, J. B. & Gerba, C. P. Quantitative 89. Amman, B. R. et al. Seasonal pulses of Marburg virus Microbial Risk Assessment (John Wiley & Sons, Publisher’s note circulation in juvenile Rousettus aegyptiacus bats 2014). Springer Nature remains neutral with regard to jurisdictional 109. Mitscherlich, E. & Marth, E. H. Microbial Survival in claims in published maps and institutional affiliations. coincide with periods of increased risk of human infection. PLoS Pathog. 8, e1002877 (2012). the Environment: Bacteria and Rickettsiae Important 90. Pigott, D. M. et al. Mapping the zoonotic niche of in Human and Animal Health (Springer Science & Business Media, 2012). SUPPLEMENTARY INFORMATION Marburg virus disease in Africa. Trans. R. Soc. Trop. See online article: S1 (box) | S2 (movie) Med. Hyg. 109, 366–378 (2015). 110. Silva, É. F. et al. Characterization of virulence of 91. Brierley, L., Vonhof, M., Olival, K., Daszak, P. & Leptospira isolates in a hamster model. Vaccine 26, ALL LINKS ARE ACTIVE IN THE ONLINE PDF 3892–3896 (2008). Jones, K. Quantifying global drivers of zoonotic bat 510 | AUGUST 2017 | VOLUME 15 w w w.nature.com/nrmicro © 201 7 Mac m ill an Publi shers Li m it ed, part of Spri nger Nat ur e. A ll ri ghts r eser ved.

Journal

Nature Reviews Microbiology – Pubmed Central

Published: May 30, 2017

There are no references for this article.