Intracellular pathogens colonize specific subcellular niches. Some niches may only be transient sites of colonization during pathogen uptake, while others are occupied for most of the pathogen lifecycle. Some pathogens also replicate at multiple subcellular locations. For example, Salmonella enterica serovar Typhimurium localize and replicate mainly in the Salmonella-containing vacuole. However, a subpopulation of Salmonella proliferate in the cytoplasm (Knodler et al., 2010). Viral subcellular tropism enables access to host enzymes essential for productive viral infection and host lipids and proteins to shield the viral particle from the immune system. For example, hepatitis C replicates on viral-induced cytoplasmic structures called the membranous web, which shields the virus from host pattern-recognition receptors (Neufeldt et al., 2016). Leishmania parasites and Coxiella burnetii bacteria multiply in the phagolysosome, a low-pH environment that requires specific metabolic adaptations for successful colonization (Saunders et al., 2014). In contrast, T. cruzi parasites and several bacteria (e.g. Listeria monocytogenes) proliferate in the cell’s cytoplasm where nutrients are abundant (Knodler et al., 2010; Lentini et al., 2018). This localization may also facilitate intercellular pathogen transfer (Knodler et al., 2010; Dowd et al., 2020).

Infection can also result in damage at specific subcellular sites. For example, T. cruzi causes mitochondrial swelling in chronic Chagas disease (Garg et al., 2003; Gupta et al., 2009). Mitochondrial damage is also observed in Herpes Simplex Virus-1 (HSV-1)-infected myenteric neurons (Khoury-Hanold et al., 2016) and during L. monocytogenes epithelial cell infection (Stavru et al., 2011), suggesting that this is a common alteration that may be caused by the metabolic stress of infection, pathogen mechanisms to avoid host cell apoptosis, or avoidance of other host defense mechanisms (Stavru et al., 2011). Examples of other sites of subcellular structural alterations include the endoplasmic reticulum, Golgi and nucleus during dengue virus infection (Win et al., 2019). Subcellular sites of disease tropism may be sites of direct contact with host structures, as in T. cruzi infection, in which cytoplasmic parasites interact with the mitochondria via parasite flagella (Lentini et al., 2018). In contrast, HSV-1 was distal to damaged mitochondria (Khoury-Hanold et al., 2016).

Intracellular pathogens preferentially colonize specific cell types. For viruses, this is first driven by the availability of entry receptors, associated processing factors, and fusion mediators, in the appropriate structural conformation (Wang and Shenk, 2005; Park et al., 2016; Orchard et al., 2018) for initial viral entry, or cell-to-cell interactions between infected and uninfected cells leading to viral spread (Shannon-Lowe et al., 2006). Subsequently, cellular tropism is determined by the ability of the infected cell to degrade internalized viruses and/or prevent their proliferation. For example, the poliovirus receptor is expressed on multiple tissues that are not sites of viral replication; viral tropism restriction is due to pre-existing expression of interferon-stimulated genes (ISGs) (Ida-Hosonuma et al., 2005). Similar processes are observed for influenza virus infection across different lung cell types (Fay et al., 2020). Depending on the receptor used for Human Immunodeficiency virus-1 (HIV-1) internalization, host TRIM5α either restricts the virus via autophagic targeting (Langerhans cells) or does not (subepithelial DC-SIGN⁺ dendritic cells) (Ribeiro et al., 2016). Last, the availability of the necessary resources to enable viral proliferation also regulates viral tropism. Metabolism controls the building blocks necessary to viral replication (Rodríguez-Sánchez et al., 2019), but recent results have also demonstrated that the availability of metabolites such as ceramide can also regulate virus entry receptor conformation and viral uptake (Orchard et al., 2018).

Apicomplexan parasites actively invade the host cells (Guérin et al., 2017). In contrast, many other intracellular bacterial and eukaryotic pathogens induce their own uptake by the host. For example, Leishmania parasites rely on phagocytosis for entry and are thus primarily tropic to neutrophils and then macrophages (Peters et al., 2008). Such host cells may be actively recruited to the sites of initial pathogen colonization by the mechanisms of transmission, for example via bioactive proteins and metabolites found in vector saliva or even egested vector gut bacteria (Peters et al., 2008; Dey et al., 2018). Lack of uptake can also determine pathogen tropism: disseminative Neisseria gonorrhoeae strains are unable to bind neutrophil receptor CEACAM3, enabling them to avoid immune killing and to disseminate beyond mucosal sites (Roth et al., 2013).

Cellular-level disease tropism can reflect direct damage by the pathogen, for example due to cell rupture by lytic viruses, or to disrupted cellular physiology. However, cellular-level pathogen tropism and disease tropism are not necessarily identical. For example, Zika virus infects multiple cell types, including microglia-like cells, astrocytes and neural progenitor cells, but only caused significant death of the latter cell type (Muffat et al., 2018). Disease tropism may also be due to collateral damage from pathogen invasion or pathogen-mediated effects on adjacent cells. For example, in the absence of myeloperoxidase, superoxide production in Salmonella-infected neutrophils damages the surrounding cells (Schürmann et al., 2017). Co-culture of neurons with T. cruzi and IFNγ-activated macrophages led to neuronal death, which was abrogated by inhibitors of nitric oxide production (Almeida-Leite et al., 2007). Zika virus increases lipid droplet levels not only in infected cells but also in adjacent uninfected cells (Chen et al., 2020). Many pathogens also lead to immune cell exhaustion. Thus, Mycobacterium tuberculosis leads to CD8⁺ T cell exhaustion, even though it does not directly infect these cells (Russell et al., 2019).

Organ and tissue tropism are relevant to both intracellular and extracellular pathogens. Some pathogens may show uniform burden throughout a given organ [e.g. Pseudomonas aeruginosa in the gallbladder (Bachta et al., 2020)], while others may show preferential tropism to select organ regions [e.g. higher T. cruzi load at the heart base in strain CL Brener infection of C3H/HeJ mice (McCall et al., 2017); higher norovirus load in the distal small intestine in wild-type C57BL/6 mice (Grau et al., 2020)]. In the case of intracellular pathogens, cellular tropism will strongly influence tissue tropism, for example if cells expressing the necessary receptor are more common in a given organ or tissue. In the case of SARS-CoV-2, viral tropism to the respiratory and gastrointestinal tract has been linked to co-expression of the ACE2 receptor and TMPRSS2 protease in these tissues (Sungnak et al., 2020; Ziegler et al., 2020). Tropism to olfactory epithelium may be potentiated by its expression of neuropilin-1, which enables entry of SARS-CoV-2 with furin-cleaved spike protein (Cantuti-Castelvetri et al., 2020). Such determinants of tissue tropism should thus be apparent in cell culture models, where the pathogen can only infect cells from the target organ. For example, respiratory enterovirus strains are restricted to the lung in vivo and show much more restrictive in vitro tropism than disseminative enteric enterovirus strains (Royston et al., 2018). In contrast, many other intracellular pathogens are promiscuous in cell culture [e.g. T. cruzi (Franco et al., 2019) or Toxoplasma gondii (Cabral et al., 2016)]. T. cruzi shows initial broad tropism in vivo during acute infection, followed by more restricted pathogen and disease tropism in chronic disease stages [mainly heart, stomach and colon, with some exceptions (Lewis et al., 2014; Lewis et al., 2016; Hossain et al., 2020)]. In the case of T. gondii, broad spatial distribution can be observed following infection with a virulent bioluminescent strain, including localization to the intestines, lung, liver, brain, heart, and kidney (Saeij et al., 2005). However, fine study of infected cell types in the brain indicates that parasites preferentially infect neurons over astrocytes in vivo, and this preference is only partially abrogated by immunosuppression (Cabral et al., 2016)

Sites of initial pathogen entry, infection of highly migratory cells such as macrophages, or pathogen motility, will also determine whether a pathogen remains restricted to its initial site of invasion or can disseminate to other organs (see below). Pathogen and disease tropism are thus also influenced by circulatory patterns, leading for example to the accumulation of Leishmania donovani-infected macrophages (McCall et al., 2013) and Plasmodium-infected red blood cells (Brugat et al., 2014) in the spleen. Direct binding to microbiota bacteria may also facilitate viral invasion (Erickson et al., 2018).

Host metabolism and thus nutrient and immunomodulatory metabolite availability also differ strongly between organs (Quinn et al., 2020). For example, levels of purines, aspartate and histidine are lower in the skin than the liver (Murakami et al., 2014). Variable nutrient availability restricts the tropism of pathogens with strict nutritional requirements. Dermotropic Leishmania have higher transporter expression, which may enable them to address skin-associated nutrient limitations, in contrast to viscerotropic Leishmania (McCall et al., 2015). Streptococcus pneumoniae adaptations to ear vs. circulatory environments has been tied to enhanced ability of blood S. pneumoniae strains to utilize raffinose as a carbon source (Minhas et al., 2019). Ear-tropic and lung-tropic S. pneumoniae strains also express different nutrient transporters (Minhas et al., 2020). Nutrient competition is a key mechanism by which the microbiota restricts pathogen colonization (Lam and Monack, 2014; Brugiroux et al., 2016). Depletion of such commensals enhances the risk of colonization by pathogens such as Clostridioides difficile (Buffie et al., 2015). The balance between microbiota-derived butyrate and acetate also shapes colonization-associated gene expression in Campylobacter jejuni and may explain its preferential tropism for the colon over the small intestine (Luethy et al., 2017). Metabolites also regulate immune responses, influencing pathogen organ tropism. For example, the intersection between secondary bile acid production by the microbiota and host bile acid receptor FXR levels regulate cellular production of IFN-λ and consequently norovirus tropism in the intestines (Grau et al., 2020). Bile acids also regulate norovirus endocytosis and virus release from endosomes (Murakami et al., 2020).

Immune responses strongly regulate pathogen tropism, as evidenced by the many cases of atypical disease presentations in immunocompromised individuals: central nervous system rather than cardiac involvement in AIDS-Chagas disease patients (Pinazo et al., 2013), invasive fungal dermatophyte infections in primary immunodeficiency patients (Pilmis et al., 2016), or broad Human Cytomegalovirus (HCMV) tropism in systemic lupus erythematosus (SLE) patients receiving immunosuppressive therapy (Arai et al., 2012). In experimental models, immunosuppression abolishes the select chronic-stage tropism of pathogens such as T. cruzi ( Lewis et al., 2016 ). Some organs are less accessible to cells of the adaptive immune response; such immune privilege may explain Trypanosoma brucei (McCall and McKerrow, 2014) or T. gondii (Saeij et al., 2005) brain tropism. However, beyond the well-characterized case of immune privilege, immune responses can also differ between organs that are freely accessible to immune cells. For example, CD4⁺ T cells in the liver adopt an IL-10-producing phenotype during Salmonella Typhimurium infection, leading to M2 macrophage polarization and long-term bacterial persistence in the liver; in contrast, CD4⁺ T cells in the spleen produce high IFNγ and low IL-10, leading to bacterial clearance (Kurtz et al., 2020). Immune restriction of pathogen tropism is not limited to adaptive responses: for example, West Nile virus is restricted from the liver due to rapid induction of type I interferons (Suthar et al., 2013).

As with pathogen tropism, disease tropism may also occur throughout a given organ, or be restricted to select organ regions. For example, T. cruzi leads to pathognomonic apical cardiac aneurysms, even though cardiac and cardiac apex parasite load are low (Marin-Neto et al., 2007). Genital HSV-2 infection leads to gastrointestinal and urinary manifestations (Khoury-Hanold et al., 2016). Strikingly, in the case of HSV-1 experimental infection, while viral colonization was observed in the dorsal root ganglia and large intestine, tissue damage was strongest in the latter, driven by excessive neutrophil recruitment and destruction of intestinal ganglia (Khoury-Hanold et al., 2016). Disease tropism to select organs and tissues is influenced by preferential sites of pathogen tropism, the ability of different organs to function even in the presence of pathogens (tolerance) and repair ability after pathogen clearance (see below).

Subcellular, cellular, tissue and organ tropism are most relevant to an individual’s experience of disease. However, pathogens and disease do not show equal distribution across the globe. Thus, geography strongly influences disease tropism, from a continental scale to differential disease incidence between neighborhoods within a given city.

On a broad geographic scale, external factors such as climate influence vector and vector-borne disease distribution [e.g. (Li et al., 2019)]. Likewise, changes in reservoir animal host tropism and reservoir behavior will influence the risk of human exposure (Zhang et al., 2018). Population movements shape global disease tropism, by introducing infected individuals in new locations (Alawieh et al., 2014). Large-scale population movements are often tied to political instability, which in turn influences the availability of disease control measures and population health (Du et al., 2018). Such factors may also play out on a smaller scale, within a city or even a building (Chng et al., 2020).

The microbiota is strongly influenced by these extrinsic factors: urbanization, loss of traditional foods and living practices, and exposure to processed food, consumer chemicals and antimicrobials (Sankaranarayanan et al., 2015; Vangay et al., 2018; McCall et al., 2020). In turn, microbiota composition affects susceptibility to infectious diseases [e.g. (Buffie et al., 2015; Villarino et al., 2016)]. Thus, the global variation in microbiota composition may be responsible for some of the observed global variations in infectious disease prevalence. In addition to this “second genome” (Grice and Segre, 2012), variations in genetic background between populations may also lead to variable pathogen and disease tropism between geographic regions, although this is challenging to deconvolute from healthcare and service access, behavioral patterns, vector and reservoir tropism. Well-characterized genetic factors that influence tropism in the context of malaria and show differential prevalence between geographic regions include Duffy receptor presence/absence vs. susceptibility to P. vivax malaria (Twohig et al., 2019).

Overall, pathogens and infectious diseases show preferential tropism at multiple levels, from the microscopic to the planetary scale. In the following sections, we will examine in more detail the specific factors that influence tropism.

In the case of pathogens newly-introduced into a mammalian body, initial tropism will be determined by the route of administration: vector-borne transmission will lead to pathogen deposition in the skin and/or vasculature, while food-borne parasites will initially colonize the gastrointestinal tract, and so on. The initial method of entry can also shape final tropism, with transmission by the tick vector leading to broader tissue tropism for several Borrelia burgdorferi sensu lato complex strains than needle injection (Sertour et al., 2018). In the case of pathobionts, members of the microbiota that can be pathogenic under certain circumstances, initial tropism will depend on which regional microbiota they came from, and the factors that led to them becoming pathogenic, such as disruption of local mucosal surfaces leading to invasion [e.g. (Ayres et al., 2012)].

Having entered the mammalian body, the ability of the pathogen to remain will be influenced by the availability of appropriate niches at the entry site. An intracellular pathogen that does not encounter the necessary invasion receptors at the site of colonization and is unable to disseminate will lead to an abortive infection. pH, temperature and nutritional requirements also determine whether the invading pathogen can establish itself. For example, MERS is resistant to fed-state simulated gastric fluid and partially resistant to fed-state simulated intestinal fluid; however, fasted-state simulated gastric fluid kills the virus. Thus, timing of viral exposure vs. meals may influence the success of oral MERS infection (Zhou et al., 2017). Attachment is also relevant to extracellular pathogens; for example, enteroaggregative E. coli showed differential patterns of adhesion on enteroids from each intestinal segment (Rajan et al., 2018). Temperature effects may be direct, due to differential pathogen proliferation at temperatures associated with visceral organs vs. skin or mucosal surfaces (McCall and Matlashewski, 2010; McCall and Matlashewski, 2012), or indirect via effects on the host. For example, airway epithelial cells produce more interferon at 37°C than 33°C, thus restricting rhinovirus to the cooler upper airways (Foxman et al., 2015).

Host- or microbiota-derived metabolites can either promote pathogen persistence (production of a key nutrient) or inhibit it (lack of a key nutrient; production of toxic host metabolites). As an example of the former mechanism, gastrointestinal Salmonella retain genes that enable utilization of alternative carbon sources, compared to extraintestinal Salmonella which are not able to access these resources. Extraintestinal strains also show loss of functions associated with anaerobic metabolism (Nuccio and Bäumler, 2014). For the latter mechanism, high levels of D-serine in the bladder prevent enterohemorrhagic E. coli (EHEC) bladder colonization by reducing type 3 secretion system virulence factor expression, host cell attachment and pedestal formation (Connolly et al., 2015). The microbiota can also directly interfere with or promote invasion by the pathogen. For example, successful influenza A virus transmission is inhibited by nasal carriage of Streptococcus pneumoniae, which cleaves host nasal sialic acid, the receptor for influenza (Ortigoza et al., 2018).

Lastly, colonization will also be determined by whether a pathogen can withstand or co-opt the initial immune responses. For example, Leishmania parasites infect neutrophils recruited to the initial site of the sandfly vector bite (Peters et al., 2008). In contrast, presence of influenza-specific resident memory T cells in the upper respiratory tract protected against nasal colonization and viral dissemination to the lung (Pizzolla et al., 2017).

Host regulation of pathogen tropism can even apply to prions (PrP^Sc): PrP^Sc strain differential tropism is related to the expression levels of endogenous healthy PrP^C proteins (Béringue et al., 2020). Initial steps of pathogen tropism can already be active processes. For example, pathogen proteins will subvert host cell functions to determine subcellular localization. Indeed, the Shigella type III secretion system translocon is sufficient to induce cytoplasmic localization when expressed ectopically in Escherichia coli (Du et al., 2016).

After setting up a “beachhead” at the site of initial invasion, pathogens will stay at that site or disseminate. Local persistence vs. dissemination is determined by the site of initial colonization, obstacles encountered during dissemination, and whether other locations are suitable for pathogen persistence and proliferation. Indeed, simian immunodeficiency virus (SIV) disseminates from the female genital tract into the bloodstream only when local viral loads exceed a certain threshold (Feder et al., 2017). Depending on whether fibroblasts or endothelial cells are infected with HCMV, released viral progeny is either tropic to both cell types, or restricted to fibroblasts (Scrivano et al., 2011). A similar process is observed in EBV infection, where viral progeny from B cells is better at infecting epithelial cells than B cells, and vice versa (Borza and Hutt-Fletcher, 2002). The quality and strength of local immune also determines whether systemic dissemination can occur. For example, stimulating innate immune responses in the skin prevented arboviral dissemination (Bryden et al., 2020). Conversely, immune responses can promote dissemination: for example, varicella zoster virus infection of tonsillar T cells promoted trafficking to tissues (Sen et al., 2014).

The ability to leave the site of colonization is also linked to pathogen abilities to actively penetrate tissues, for example via induction of parasite and host proteases in Entamoeba histolytica infection, enabling invasion out of the gastrointestinal tract (Thibeaux et al., 2014; Siqueira-Neto et al., 2018). Zika virus also co-opts host proteases to cross the blood-testis barrier (Hui et al., 2020). S. aureus may invade bones via forces generated by cell division (Masters et al., 2020). Inflammasome induction by Plasmodium falciparum histidine-rich protein II leads to loss of blood-brain barrier integrity (Pal et al., 2016) and influenza H5N1 infection of endothelial cells promotes vascular leakage and dissemination to extrapulmonary sites (Tundup et al., 2017). Invasiveness does not have to be constitutive: for example, Shigella turn on type 3 secretion system (T3SS) needle production in anaerobic environments such as the gut lumen; approaching mucosal tissues leads to induction of effector secretion (Marteyn et al., 2010).

Such active penetration may not be necessary if the host presents with pre-existing damage. For example, loosened intestinal barrier in hyperglycemic mice increased ability of Citrobacter rodentium to colonize spleen and liver (Thaiss et al., 2018). Similarly, immune-mediated disruptions to lung blood vessels during lung transplant led to increased iron availability that promoted Aspergillus fumigatus invasiveness (Hsu et al., 2018). Dissemination and invasiveness is also facilitated by pathogen motility [e.g. bacterial flagella (Cullender et al., 2013)], access to the lymphatic system [e.g. group A Streptococci, mediated by capsule hyaluronan interaction with lymphatic vessel endothelial receptor 1 (Lynskey et al., 2015)], infection of mobile host cells [e.g. Chlamydia trachomatis infection of dendritic cells (Howe et al., 2019)] or by taking advantage of cell-cell connections [e.g. HSV-1 dissemination from initial sites of vaginal colonization to the dorsal root ganglia, then spine, and then colon enteric nervous system via peripheral nociceptors (Khoury-Hanold et al., 2016)].

One caveat is that it is often challenging to differentiate between a pathogen that remains at local sites of initial invasion without disseminating to other tissues, and a pathogen that does disseminate but is unable to establish itself anywhere but at the initial site. As an example, we observed high T. cruzi load at the heart base during acute experimental T. cruzi infection (McCall et al., 2017); is this due to a lack of parasite dissemination beyond this cardiac region, or to rapid parasite killing at the heart apex? The observation of higher antiparasitic IFNγ at the heart apex argues for the latter scenario (McCall et al., 2017). Conversely, phagocytic cells infected with L. major parasites are less likely to migrate out of the site of intradermal infections than L. donovani-infected cells, indicating that the dermotropism of L. major is due to the first scenario (Zhang et al., 2003). Many immune mechanisms are designed to capture and kill pathogens during the dissemination process. As an example, liver Kupffer cells phagocytose circulating pathogens (McDonald et al., 2020) and circulating antibodies neutralize key pathogen surface molecules or target them for degradation (Engstler et al., 2007). These mechanisms can be subverted by pathogens adapted to live in macrophages such as Leishmania ( Beattie et al., 2013 ), by loss of long-range interactions between gut microbiota and immune function during dysbiosis (McDonald et al., 2020), or by pathogen surface antigenic variation (Engstler et al., 2007).

High temporal and spatial resolution series of luminescent animal models may conclusively resolve this issue (see below). Importantly, disease resolution or lack of apparent disease at any time at the site of initial pathogen invasion does not mean lack of low-level pathogen persistence. For example, T. cruzi persist at low levels in the skin in mammalian models, facilitating pathogen transmission, in the absence of visible skin lesions (Ward et al., 2020).

Colonizing additional sites is influenced by many of the same factors that influence initial colonization: nutrient and host cell availability (McCall et al., 2015; Quinn et al., 2020; Sungnak et al., 2020; Ziegler et al., 2020), competition (Lam and Monack, 2014), thermal and stress tolerance (Scull et al., 2009; McCall and Matlashewski, 2010; McCall and Matlashewski, 2012), surviving immune responses, for example through colonization of immune privileged organs (see above, as for T. gondii or T. brucei) (Saeij et al., 2005; McCall and McKerrow, 2014) or co-opting defense pathways [e.g. (Mandal et al., 2017; Howe et al., 2019; Ziegler et al., 2020)]. For the multiple pathogens that initially show broad tissue tropism followed by more restrictive tropism [e.g. T. cruzi, T. gondii (Saeij et al., 2005; Lewis et al., 2014)], basal nutrient availability may be less restrictive; in contrast, tissue metabolic remodeling or induction of immunomodulatory metabolites may be more important. For example, we observed infection-induced increases of the metabolite kynurenine mainly in the large intestine and to a lesser extent in the stomach following T. cruzi infection. Kynurenine induces regulatory T cells and may thus contribute to parasite persistence in the colon (Hossain et al., 2020).

Disease tolerance is the ability to withstand the deleterious effects of infection (McCarville and Ayres, 2018). Resilience is the ability to return to health following the clearance of a pathogenic insult (Torres et al., 2016). While this has been predominantly studied at the level of the whole organism, our findings of localized metabolic perturbations and localized metabolic restoration by pharmacological interventions with constant parasite burden argue for spatial disease tolerance (Hossain et al., 2020). Thus, we find that the heart apex appears to be less tolerant to acute T. cruzi infection than the heart base (McCall et al., 2017), while select intestinal regions have a higher capacity to return to normal metabolism in the chronic stage of experimental infection (Hossain et al., 2020).

Some of these characteristics are intrinsic to a tissue, for example with regards to responses to initial infection (Rajan et al., 2020), or ability to regenerate; thus, some locations are more prone to persistent infectious disease damage than others. Cardiac damage, for example, is challenging to reverse, even after the pathogen is cleared [e.g. (Morillo et al., 2015)]. However, the location of damage (disease tropism) is naturally also influenced by pathogen localization and its intersection with organ, tissue and cellular characteristics. For example, lung infection with SARS-CoV-2 leads to lung lesions and respiratory impairment, related to the high viral receptor expression in lung pneumocytes (Finch et al., 2020; Ziegler et al., 2020). These lesions are highly localized; however, virus can also be detected in the kidney, liver, heart, and periodically in the brain (Puelles et al., 2020). SARS-CoV-2 infection is associated with elevated interferon production, but only in the lower respiratory airways; IFNλ impairs the lung barrier and may thus promote increased disease severity and superinfection via those sites (Broggi et al., 2020). This differential interferon production may be related to the temperature sensitivity of the interferon response reported in rhinovirus infection (Foxman et al., 2015). Neurological symptoms of COVID-19 have been particularly mysterious. Recent work demonstrating SARS-CoV-2 tropism to choroid plexus cells and causing breakdown of the blood-central nervous system barrier integrity (Pellegrini et al., 2020), illustrates how pathogen tropism informs the study of disease tropism.

The ability of a pathogen to cause collateral damage to uninfected cells, tissues and organs will also determine disease tropism. This damage may be due to mediators produced by the pathogen [e.g. (Thibeaux et al., 2014)], the activation of non-specific immune responses leading to damage of adjacent cells [e.g. (Almeida-Leite et al., 2007; Schürmann et al., 2017)], or compensatory mechanisms in response to direct tissue damage. An example of the latter mechanism would be hypertrophy of surviving cardiomyocytes in chronic T. cruzi infection, as part of an effort to maintain heart function (Garg et al., 2003). Although collateral damage is assumed to neighbor pathogen localization, this may not always be true. For example, in an immunocompetent mouse model of Zika virus infection, fetal abnormalities were observed even in animals where no placental or embryonic viral RNA were detected, caused by type I interferon-mediated placental damage (Szaba et al., 2018).