Selecting Animal Models for Infectious Disease Research

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-08-11

Key Takeaways

• The selection of an animal model for infectious disease research must be hypothesis-driven, focusing on reproducing the specific pathogen-host interaction, route of infection, host response, and clinical outcome of interest, rather than seeking a universally "best" model.
• Candidate models must demonstrate actual susceptibility to the pathogen; apparent resistance can often stem from pathogen specialization on its natural host rather than evolved immunity, rendering the model irrelevant for studying natural disease mechanisms.
• Immune system relevance is paramount, requiring a match between the model's cytokine signaling pathways (e.g., TNF, IL-6), Toll-like receptor repertoire, and MHC polymorphism with those of the target species to ensure accurate translation of vaccine or therapeutic efficacy.
• Genetic control within the model species is critical, as heritable variation in susceptibility (e.g., estimated heritability of 0.06-0.09 for Mycobacterium avium subsp. paratuberculosis in cattle) can significantly influence experimental outcomes and obscure treatment effects.
• Practical and ethical considerations, including the 3Rs framework (Replacement, Reduction, Refinement), availability of genetic tools, housing requirements, and cost, must be rigorously assessed alongside scientific criteria to justify the chosen model and procedures.
• Potential failure modes such as pathogen adaptation through serial passage, silent co-infections, physiological mismatches in cytokine kinetics, and genetic drift in outbred stocks necessitate proactive monitoring and validation of the model throughout the research process.

Infectious disease research depends on animal models that reproduce the relevant features of pathogen-host interaction. The choice of model determines whether findings translate across species, whether the study answers the question it poses, and whether the resources invested produce usable knowledge. This article provides a structured framework for that choice, written for veterinary researchers who design infection studies, interpret comparative data, or review protocols. It addresses the scientific logic of model selection, the biological criteria that distinguish useful from misleading models, and the practical constraints that shape final decisions.

The central question is not which animal species is the best general model for infection, but which model best represents the specific pathogen, route, host response, and outcome of interest. A model that reproduces human tuberculosis pathology may fail to reproduce the transmission dynamics of the same pathogen. A model that captures the acute cytokine response to bacterial lipopolysaccharide may say nothing about chronic carriage. The framework below treats model selection as a hypothesis-driven process, not a menu of standard options.

At a Glance

The Logic of Model Selection

Model selection begins with a precise statement of the research objective. The objective may be to define pathogenesis, test a vaccine, evaluate a therapeutic, or study transmission. Each objective places different demands on the model. A pathogenesis study requires a model that reproduces the tissue tropism and lesion spectrum of the natural disease. A vaccine trial requires a model with an immune system close enough to the target species that protective responses are likely to transfer. A transmission study requires a model that supports the pathogen's natural shedding and acquisition routes.

The pathogen itself imposes the first constraint. Most pathogens infect a limited range of host species, and the evolutionary basis of that restriction matters. Resistance in a species outside the normal host range may reflect evolved immunity or may be an incidental by-product of pathogen specialisation on its source host. Cross-inoculation studies show that apparent nonhost resistance is often the product of pathogen specialisation, meaning the pathogen has lost the capacity to infect that species through adaptation to its natural host. A model that resists infection for this reason will not reveal the mechanisms of disease in the target species, because the block occurs at an early step that is irrelevant to the natural host.

Host-Pathogen Compatibility

Compatibility operates at multiple levels. The pathogen must be able to attach, enter, replicate, and persist in the candidate host. It must also elicit a host response that resembles the target disease. The Syrian hamster illustrates this principle. The Syrian hamster immune response to infectious pathogens is similar to humans, and this species has been used to study bacterial, viral, and parasitic infections, including the evaluation of vaccines and therapeutics. The hamster is not universally useful, however. Its value depends on the specific pathogen and the aspect of disease under study. The same review notes challenges in using hamsters as an alternative model, including limited immunological reagents and differences in disease course for some pathogens.

The genetic background of the model species also influences outcome. Susceptibility to infection varies within species, and this variation can be heritable. In dairy cattle, the estimated heritability of susceptibility to Mycobacterium avium subsp. paratuberculosis infection was 0.06 in the overall population and 0.09 in vaccinated animals. A model chosen without attention to genetic background may produce results that reflect the genetics of the particular stock instead of the biology of the disease. This is especially important in outbred models, where individual variation in susceptibility can obscure treatment effects or produce false conclusions about pathogen virulence.

Immune System Relevance

The immune response is the central mediator of both protection and pathology in infectious disease. A model must therefore reproduce the immune pathways that are relevant to the clinical question. The major histocompatibility complex (MHC) is the most variable region of the vertebrate genome and directly shapes immune recognition of pathogens. MHC variants influence immune recognition, susceptibility to infectious and autoimmune diseases, and other traits relevant to infection studies. Models with limited MHC diversity may fail to represent the range of immune responses seen in the target population, and results obtained in such models may not predict outcomes in genetically diverse hosts.

Cytokine responses are equally important. Tumor necrosis factor (TNF) illustrates the dual role of inflammatory mediators in infection. TNF contributes directly to the normal immune response to infection, limiting the spread of pathogens, but exaggerated TNF release in response to overwhelming infection produces aberrations in coagulation, cell adhesion, chemotaxis, and vascular integrity. A model that does not reproduce this balance may misrepresent both the protective and the pathological aspects of the disease. The same cytokine network is discussed in relation to the systemic inflammatory response syndrome, where TNF contributes to organ damage and mortality in sepsis, burns, trauma, and pancreatitis. When the research question concerns inflammatory pathology, the model must show a comparable cytokine cascade, also comparable pathogen growth.

Distinguishing Evolved from Nonhost Resistance

The evolutionary origin of resistance in a candidate model species determines what the model can teach. Resistance that results from hosts evolving to resist challenge by their pathogens differs from resistance that is an incidental by-product of pathogen specialisation. These two forms of resistance have different genetic mechanisms and different implications for translational research. A model that resists infection through an evolved, specific immune mechanism may share relevant biology with the target species. A model that resists because the pathogen cannot use its surface receptors or intracellular machinery will not.

This distinction has practical consequences. When a candidate species shows resistance to a pathogen, the researcher should ask whether that resistance reflects a mechanism present in the target species or a species-specific block. Cross-inoculation experiments with related pathogens can help answer this question. If the pathogen infects closely related species but not the candidate, the block is likely specific to the candidate. If the pathogen infects only its natural host and no related species, the resistance may be general and less informative.

Practical Model Selection: A Decision Framework

The selection process begins with a structured assessment of the pathogen, the research question, and the candidate species. A useful starting point is to score each candidate model across five domains: pathogen susceptibility, immune response fidelity, genetic tool availability, practical feasibility, and ethical acceptability. The weighting of each domain shifts with the study objective. A vaccine efficacy trial weights immune response fidelity heavily, whereas a pathogenesis study focused on bacterial dissemination may prioritize susceptibility and route of inoculation.

Step 1: Define the Pathogen-Host Interaction

Characterize the pathogen's natural host range before considering any laboratory species. If the pathogen does not productively infect the candidate species, the model fails at the first hurdle. Cross-inoculation experiments show that apparent resistance in a nonhost species often reflects pathogen specialisation on its source host instead of an active immune defense, and this distinction has direct genetic and mechanistic consequences for the model. A pathogen that has coevolved with its natural host will engage different receptors, evade different immune effectors, and produce different pathology than it will in an incidental host.

Ask three questions at this stage. Does the pathogen complete its lifecycle in the candidate species? Does the resulting disease reproduce the target clinical syndrome? Does the route of inoculation match the natural transmission route? For agents with fastidious growth requirements or narrow host tropism, the Syrian hamster has proven valuable because its immune responses to bacterial, viral, and parasitic pathogens resemble those of humans, making it a strong candidate for pathogenesis studies and therapeutic evaluation.

Step 2: Score Immune Response Fidelity

The immune system is the principal mediator of both protection and pathology. A model that fails to reproduce the relevant immune axis will generate misleading efficacy and safety data. Tumor necrosis factor illustrates the problem. TNF is central to the normal containment of infection, yet exaggerated TNF release drives the coagulation abnormalities, vascular leak, and tissue injury seen in overwhelming sepsis. A species that does not mount a comparable TNF response will not model the septic syndrome faithfully, and anti-TNF interventions evaluated in that species may show artefactual benefit or harm.

Compare the candidate species with the target species across the following immune parameters:

MHC variability deserves particular attention because it shapes immune recognition and susceptibility to infectious disease across vertebrate species. Inbred laboratory strains with fixed MHC haplotypes may fail to represent the outbred target population. Where the research question concerns vaccine coverage or differential susceptibility, consider using outbred stocks or defined MHC-congenic lines.

Step 3: Assess Genetic Tools and Manipulability

The availability of genetic reagents determines which mechanistic questions can be answered. Inbred strains, gene knockouts, conditional expression systems, and validated antibody panels each expand the experimental repertoire. A species with excellent pathogen susceptibility but no genetic tools will support descriptive studies only. Conversely, a genetically tractable species that resists the pathogen will produce negative results that are difficult to interpret.

For quantitative trait analysis, the target species itself may be the most appropriate model. Heritability estimates for susceptibility to Mycobacterium avium subsp. paratuberculosis in dairy cattle, for example, have been calculated directly from field populations, demonstrating that genetic variation in infection susceptibility exists in the target species and can be exploited without recourse to a surrogate. When the research question concerns host genetics, the target species is often the only valid model.

Step 4: Evaluate Practical and Ethical Feasibility

Housing requirements, cost per animal, generation time, and the availability of specific pathogen-free colonies all constrain model choice. The Guide for the Care and Use of Laboratory Animals sets out the institutional responsibilities for veterinary care, housing, and oversight that apply regardless of species. The NC3Rs resources on replacement, reduction, and refinement provide practical guidance on minimizing animal numbers and improving welfare through refined procedures, anesthesia, and endpoint criteria.

Ethical scoring should consider the severity of the anticipated infection, the duration of the study, the requirement for humane endpoints, and the availability of non-animal alternatives. A model that requires prolonged severe disease to produce a measurable outcome may be less acceptable than one that produces a milder but more informative phenotype. Where the pathogen causes zoonotic disease, additional containment requirements will affect both feasibility and welfare.

Step 5: Build the Decision Matrix

The following matrix summarizes the comparative assessment for common model species. Scores are relative and must be adjusted for the specific pathogen under study.

The target species, when it can be used, offers the highest fidelity for pathogen-host interactions but carries the greatest practical and ethical burden. The mouse offers the deepest genetic toolkit but frequently fails to reproduce human or veterinary disease phenotypes. The Syrian hamster occupies a middle ground, with broad pathogen susceptibility and human-like immune responses that make it particularly suitable for emerging infectious disease research where no established model exists.

When the Correct Choice Changes

The optimal model shifts with the study phase. Early pathogenesis studies may favour a susceptible, tractable species with good immune fidelity. Vaccine efficacy trials require a species that mounts a protective response comparable to the target population. Therapeutic studies need a model where the pharmacokinetics and toxicity profile of the candidate drug are known. Biomarker discovery requires serial sampling, favouring larger species or those with accessible blood volumes.

Production system and patient status also matter. A study of paratuberculosis in dairy cattle must account for the genetic background of the target herd, the management system, and the vaccination history, because susceptibility varies with these factors. A study of a respiratory pathogen in pigs will use different models depending on whether the question concerns nursery-age disease, finishing-stage pathology, or sow-to-piglet transmission. Regional disease status and regulatory standards, such as those in the WOAH terrestrial animal health code, may restrict the use of certain species or require additional containment.

Species-specific clinical references, such as the MSD Veterinary Manual, provide baseline physiological and pathological data that inform model selection and interpretation. Professional guidance from bodies such as the American Veterinary Medical Association can clarify institutional expectations for veterinary oversight and welfare assessment. These resources do not replace the investigator's responsibility to justify the model, but they supply the comparative data needed to do so rigorously.

Recognized Complications and Failure Modes

Model failure rarely announces itself as a single dramatic event. More often it emerges as a pattern of discordant data that accumulates across experiments. The most common failure modes are predictable, and each has an early warning sign.

Pathogen adaptation to the host. Serial passage through a model species selects for variants that replicate efficiently in that host, often at the cost of the traits you intended to study. Attenuation of virulence, loss of antigenic epitopes, and altered tissue tropism all occur. Detect this early by comparing the challenge stock's genome or phenotype against the original isolate at defined passage intervals. If you observe a progressive change in lesion character or pathogen load across experiments using the same stock, suspect adaptation before reinterpreting your hypothesis.

Silent co-infection. A research colony that appears healthy can carry endemic agents that modify immune responses and confound infection outcomes. Sendinel screening programs detect these agents, but only if the panel matches the pathogens relevant to your model. A hamster colony, for example, may carry agents that alter cytokine responses to bacterial challenge in ways that mimic or mask treatment effects. The Syrian hamster's immune responses resemble those of humans for many pathogens, but this fidelity cuts both ways: an intercurrent infection can produce human-like pathology for the wrong reason. Early detection requires baseline serology or PCR panels on sentinel animals before the study begins, not after anomalous results appear.

Physiological mismatch. The model's baseline physiology may differ from the target species in ways that distort the readout. TNF biology illustrates the problem. Tumor necrosis factor plays a major role in the systemic inflammatory response to infection, but the kinetics and magnitude of TNF release differ across species and across challenge routes. If your endpoint is a cytokine profile, you must establish the normal range and temporal curve for your model under sham conditions first. A single time point measurement taken without this baseline cannot distinguish a true treatment effect from a species-specific kinetic difference.

Genetic drift in the model population. Outbred stocks change over time. Inbred strains drift more slowly but are not static. The heritability of susceptibility to a given pathogen can be substantial, as demonstrated for Mycobacterium avium subsp. paratuberculosis in dairy cattle, where heritability of infection susceptibility was estimated at 0.06 in the overall population and 0.09 in vaccinated animals. Comparable genetic variation exists in laboratory species. Detect drift by maintaining cryopreserved reference stocks and by periodically recharacterising the colony's response to a standard challenge.

Common Errors in Model Selection

Less experienced investigators often choose a model because it is familiar or because a published protocol exists, instead of because it matches the biological question. The corrective action is to work backwards from the pathogen-host interaction you need to reproduce.

A second frequent error is treating the model species as a generic stand-in for the target species. Cross-inoculation experiments show that resistance to a pathogen is often the product of pathogen specialisation on its source host, and that resistance in a nonhost species may have a completely different genetic basis. A pathogen that fails to infect a candidate model may do so for reasons unrelated to the immune mechanism you wish to study. The corrective action is to test the pathogen's ability to complete its life cycle in the candidate model before committing to the full study.

A third error is ignoring major histocompatibility complex variability. MHC variants influence immune recognition and susceptibility to infectious disease, and the relevant diversity in an outbred model population may not match the target species. The corrective action is to characterize MHC diversity in the model population and to consider whether the study question requires a defined MHC background.

Limitations of the Evidence

The evidence base for model selection is uneven. For some pathogens and some model species, the literature is extensive and mechanistically detailed. For others, the available data consist of single reports or studies with small sample sizes. Expert opinion differs on how much weight to give to phylogenetic relatedness versus functional similarity of the immune response. The Syrian hamster literature illustrates both the strengths and the gaps: the model has proven valuable for studying pathogenesis and for drug and vaccine evaluation across bacterial, viral, and parasitic pathogens, but the available immunological reagents and genetic tools remain more limited than those for mice.

A further limitation is that much of the published work derives from laboratory-adapted pathogen strains and specific-pathogen-free animals. Field isolates and naturally exposed animals may behave differently. Where the evidence base is thin, state the uncertainty explicitly in the study design and consider a pilot experiment to establish feasibility before committing resources.

When to Escalate

Referral or consultation is warranted when the study question exceeds local expertise or resources. Specific circumstances include:

• The pathogen requires biosafety containment beyond the facility's current certification.
• The model species is not one for which the institution has established husbandry and veterinary care protocols. The Guide for the Care and Use of Laboratory Animals provides the framework for institutional animal care programs, and the NC3Rs offers practical guidance on refinement and reduction that should inform protocol design before the study begins.
• The endpoint requires specialised pathology, imaging, or immunology services not available locally.
• The study involves a notifiable pathogen. The World Organization for Animal Health terrestrial animal health standards define reporting obligations for listed diseases, and these obligations apply regardless of whether the infection is experimental or natural.

Regulatory reporting is required when a notifiable agent is detected, when an experimental infection escapes containment, or when an unexpected disease outbreak occurs in the research colony. The responsible institutional animal care and use committee, attending veterinarian, and relevant national authority should be notified without delay. For production species, the attending veterinarian should also consider whether the study findings have implications for herd health or trade status under applicable animal health standards.

Frequently Asked Questions

How do I choose a model when funding or housing constraints rule out the ideal species?

Revisit the decision matrix and identify which selection criteria are most critical to your specific hypothesis. If the ideal species is unaffordable, consider whether a smaller or more tractable species can still answer the primary question. For example, the Syrian hamster offers immune responses similar to humans for several pathogens and may serve as a practical alternative to nonhuman primates for certain viral and bacterial studies, though its limitations must be acknowledged Syrian hamster as an animal model for infectious disease research. If no alternative species preserves the key host-pathogen interaction, consider whether the study question can be reframed or whether a collaborative arrangement with another institution is feasible. Document the compromise and its potential impact on translational validity in the protocol.

What should I do when the reagents or genetic tools for the preferred species are unavailable?

Confirm whether commercial antibodies, knockout lines, or validated assays genuinely do not exist, or whether they are simply difficult to source. Check whether cross-reactive reagents validated for a closely related species are acceptable for your readout. If genetic manipulation is essential and unavailable, consider an alternative species with established transgenic resources, even if its infection phenotype is less faithful. The Syrian hamster, for instance, has historically lagged behind mice in genetic tools, which limits mechanistic studies despite its advantages in modeling human-like pathogenesis Syrian hamster as an animal model for infectious disease research. If the gap cannot be bridged, the study may need to be redesigned around observational endpoints instead of mechanistic manipulation.

How does model selection differ when I am studying a livestock pathogen instead of a human-relevant one?

For livestock pathogens, the target species is often the natural host, which eliminates the extrapolation problem inherent in human infectious disease research. Cattle, swine, and poultry models can directly assess transmission, production impact, and vaccine efficacy under conditions relevant to field application. However, natural-host models are frequently outbred, genetically heterogeneous, and expensive to house at appropriate biosecurity levels. Genetic variation in susceptibility, such as the heritable component documented for Mycobacterium avium subsp. paratuberculosis infection in dairy cattle, can introduce substantial variability into experimental outcomes genetic variation in susceptibility to paratuberculosis in dairy cattle. Account for this by increasing group sizes, using pedigree information, or stratifying animals by known risk factors. Consult species-specific clinical references for baseline physiological and pathological parameters MSD Veterinary Manual professional resources.

What records should I maintain to support reproducibility and regulatory compliance?

Maintain a complete record of strain or breed, source, health status, age, sex, housing conditions, and any prior experimental exposure. Document the rationale for species selection, including the decision matrix scores and any compromises made. Record all procedures, including anesthesia, sample collection, and humane endpoints, in sufficient detail that another laboratory could replicate the study. Institutional animal care and use documentation should follow the standards described in the Guide for the Care and Use of Laboratory Animals, which covers housing, veterinary care, and oversight requirements Guide for the Care and Use of Laboratory Animals. Apply the principles of replacement, reduction, and refinement when designing the protocol and when reporting the number of animals used NC3Rs guidance on the 3Rs. Archive all records for at least the duration required by your institution or funding body.

How do I explain the choice of animal model to an institutional animal care and use committee?

Present the selection as a structured decision instead of a preference. State the research question, list the candidate species, and explain why each was accepted or rejected using the criteria of host-pathogen compatibility, immune relevance, genetic tools, and feasibility. Acknowledge the limitations of the chosen model explicitly, including where nonhost resistance or evolutionary divergence may affect interpretation evolutionary origins of resistance in host-pathogen interactions. Describe the minimum number of animals needed to answer the question and justify that number statistically. Explain the refinements incorporated into the protocol, such as analgesia, humane endpoints, and group housing where compatible with the study design. The committee will respond more favourably to a transparent, evidence-based rationale than to an assertion that a particular species is standard practice.

When should I consult a specialist or change my model altogether?

Seek specialist input when the infection does not reproduce the expected clinical signs, when pathogen shedding or transmission dynamics differ from field observations, or when the immune response diverges from what the literature predicts. Consult a veterinary pathologist early if lesions are ambiguous, and a biostatistician if variability is unexpectedly high. If the model fails to support the hypothesis after reasonable troubleshooting, consider whether the pathogen-host interaction is fundamentally incompatible. Cross-inoculation studies show that many species are resistant to pathogens outside their normal host range, and this resistance may reflect pathogen specialisation instead of a robust immune response nonhost resistance and pathogen specialisation. Changing models is not a failure if it is driven by evidence. Document the reasons for the change and the data that informed it.

Related Clinical & Scientific Guides

• Refining IACUC Protocols to Minimize Animal Pain and Distress
• Health Monitoring Programs for Laboratory Animal Facilities
• Anesthetic Risk Assessment in Laboratory Animals: Preoperative Evaluation

References and Further Reading

• Syrian Hamster as an Animal Model for the Study on Infectious Diseases.. 2019.
• Role of tumor necrosis factor-alpha in disease states and inflammation.. 1993.
• Tumor necrosis factor in the pathogenesis of infectious diseases.. 1993.
• The origin of specificity by means of natural selection: evolved and nonhost resistance in host-pathogen interactions.. 2013.
• Genetic variation of susceptibility to Mycobacterium avium subsp. paratuberculosis infection in dairy cattle.. 2000.
• The importance of immune gene variability (MHC) in evolutionary ecology and conservation.. 2005.
• Guide for the Care and Use of Laboratory Animals, 8th Edition. National Academies Press, 2011.
• NC3Rs Resources on Replacement, Reduction and Refinement. NC3Rs.
• MSD Veterinary Manual, Professional Edition. MSD Veterinary Manual.

Related Articles

• Animal Model Selection for Neurological Research
• Selecting Animal Models for Neurological Research
• Selecting Appropriate Animal Models for Pain Research
• Selecting Bedding Materials for Laboratory Rodent Welfare
• Scoring Severity of Procedures in Animal Research Protocols

This article is educational professional reference material for veterinary audiences. It is not a substitute for veterinary diagnosis, individual clinical judgment, current product labeling, or applicable regulatory requirements.