Imaging of Small-Animal Models of Infectious Diseases

Abstract

Infectious diseases are the second leading cause of death worldwide. Noninvasive small-animal imaging has become an important research tool for preclinical studies of infectious diseases. Imaging studies permit enhanced information through longitudinal studies of the same animal during the infection. Herein, we briefly review recent studies of animal models of infectious disease that have used imaging modalities.

Small-animal imaging has become an important research tool in studies of infectious diseases and has significantly contributed to both our understanding of pathogenesis and preclinical investigations on drug development. Noninvasive imaging research permits enhanced information through longitudinal studies of animal models of human diseases. Infectious diseases are important causes of morbidity and mortality in humans worldwide. During the past decade, several different small-animal imaging modalities have been applied to studies of infectious disease, including magnetic resonance imaging (MRI), computed tomography (CT), positron emission tomography (PET), bioluminescence imaging (BLI), and intravital imaging. Multiple-modality imaging has become even more attractive because it permits evaluation of the same animals by different imaging technologies, thus reporting on alterations in anatomical characteristics, metabolism, function, and the location of infectious agents. The development of imaging applications in animal models of infectious diseases using these modalities can quickly move from basic research to the clinic. Herein, we review some recent applications of small-animal imaging technologies to the study of infectious diseases.

Overview of Imaging Technologies

MRI is a noninvasive imaging modality with high resolution (approximately 50 to 100 μm for small-animal studies) and excellent intrinsic soft tissue contrast. MRI can be used to image anatomical structures, blood flow, and diffusion in the clinic and in experimental animals. Contrast agents (gadolinium or iron-based agents) can be used to specifically label cells or tissues for diagnostic applications. Although micro-CT is the gold standard for imaging bone in mice, contrast agents are required to enhance soft tissues. CT permits longitudinal studies of anatomical characteristics like MRI. It is a high-resolution (>50 μmol/L), fast (minutes) X-ray–based technique. A concern with CT, particularly with longitudinal studies, is the radiation dose, which may be high enough to induce changes in the biological pathways being studied, and the need for contrast agents for soft tissue imaging.

PET is a highly sensitive (pmol/L) molecular imaging technique that can be used to visualize a variety of in vivo biological processes. Although the resolution of micro-PET (1 to 2 mm) is not as high as that of CT or MRI, it is adequate for small-animal imaging. The molecule 2-deoxy-2-[¹⁸F] fluoro-d-glucose (FDG) is routinely used in the clinical setting for the detection of cancer, based on increased uptake of glucose by malignant cells. In addition to ¹⁸F-FDG, many other radiotracers can be used in micro-PET studies.

Single-photon emission CT (SPECT) uses radiotracers that permit visualization of specific physiological information, such as blood flow and perfusion, or to measure biodistribution of a radiolabeled molecule or cell. The radioisotopes used are typically longer lived than those used for PET [eg, technetium-99m (half-life of 6 hours) or indium-111 (half-life of 67 hours)]. SPECT is performed in combination with CT to provide anatomical detail; however, the additional exposure to radiation has to be considered in planning longitudinal studies.

Ultrasonographic (US) imaging has high spatial resolution (approximately 50 μm) and contrast in soft tissue. In addition to being portable, US is a fast and economical technique. US has been extensively used for echocardiographic studies of small animals. Recent advances in technology and in contrast agent development have improved resolution such that US studies of many organ systems of small animals are possible.

In vivo BLI has been applied in many studies of small animals and cells. It can be used to monitor gene expression and to track cells. Bioluminescent and fluorescent probes have been engineered to monitor enzymes and the activity of other biologically important molecules. These probes can be used to follow disease progression or response to treatment. Tumor cell lines and microbial pathogens that express luciferase or fluorescent proteins are commonly used for preclinical studies, as are transgenic animals that stably express bioluminescent or fluorescent proteins.

Noninvasive imaging of animals using these technologies reduces animal numbers by permitting the use of animals as their own controls. In addition, therapeutic agents can be developed and tested using imaging technologies that are directly translatable to the clinic. Longitudinal imaging of chronic diseases permits continuous monitoring of disease progression and response to treatment. Table 1 summarizes the advantages and limitations of imaging technologies commonly used in studies of small-animal models of infectious diseases.

Table 1.

Small-Animal Imaging Modalities

Modality	Optimal use	Advantages: serial studies (all)	Limitations
CT	Anatomical body imaging of bone, lung, and heart	High resolution of anatomical structures <50 μm	Not quantitative
			Ionizing radiation dose
			Cost
			Contrast agents required for soft tissue
			Contrast, tumors, and angiography
PET	Metabolism, perfusion, cell proliferation, apoptosis, hypoxia, and functional imaging	Quantitative function Picomolar sensitivity Molecular targeting Dynamic imaging	Ionizing radiation (requires injection of a radiotracer)
			Low resolution
			Improved anatomical resolution requires hybrid devices (PET/CT or PET/MRI)
			Cost
SPECT	Functional imaging: metabolism, perfusion, hypoxia, and apoptosis	Qualitative function Picomolar sensitivity Molecular targeting	Ionizing radiation (requires injection of a radiotracer)
			Not quantitative
			Low resolution (improved with hybrid CT)
			Cost
MRI	Soft tissues, angiography, perfusion, functional imaging, and tissue oxygenation	Qualitative function	Medical contraindications
		Quantitative function High spatial and temporal resolution with intrinsic soft tissue contrast Nonionizing radiation	Sensitivity limited for some functional measures
			Cost
			Some applications require contrast agents (gadolinium- and iron-based agents are common)
US	Whole body imaging and echocardiography	Nonionizing radiation	Air and bone produce artifacts
		Anatomical features	Not quantitative
		Molecular targeting
Optical imaging (bioluminescence and fluorescence)	Single-cell and single-gene imaging	Nonionizing	Not for humans
		Inexpensive
		Molecular targeting

This table lists the modality, optimal use, advantages, and limitations of commonly used small-animal imaging modalities (CT, PET, SPECT, MRI, US, and optical imaging).

Parasitic Infections

Chagas Disease

Trypanosoma cruzi is the causative agent of Chagas disease, a neglected tropical disease, endemic to Latin America, and is being diagnosed in nonendemic areas as a result of immigration.¹ Cardiac manifestations of Chagas disease include acute myocarditis and chronic dilated cardiomyopathy, accompanied by congestive heart failure, arrhythmias, cardioembolism, and stroke.² Approximately 30% of infected individuals develop chronic manifestations, including cardiomyopathy, megasyndromes of the gastrointestinal tract, or both. MRI and echocardiography have been useful in the diagnosis of patients infected with T. cruzi.^3–5 An extensive review of advances in imaging animals infected with T. cruzi has recently been published.⁶ Herein, we will focus on the mouse model that has been extensively studied using a variety of imaging modalities, including MRI, echocardiography, and PET imaging using [18F]-FDG.^7–14

MRI has been most useful for evaluating the right ventricle of mice (Figure 1), which is difficult to visualize with standard echocardiography,^8,12,14 whereas echocardiography has been effective for evaluating left ventricular function.^7,13,15 SPECT imaging has been applied in human studies¹⁶; however, it has not been reported in animal studies. On the other hand, PET studies have only been reported in the animal model of T. cruzi infection (Figure 1). [¹⁸F]-FDG-PET has detected changes in glucose metabolism, presumably due to inflammation, early during the course of T. cruzi infection and before significant changes in heart structure or function are detected.¹³ Infection can also result in loss of smooth muscle tone and destruction of ganglia throughout the bowel and bladder, resulting in megasyndromes of the esophagus, colon, intestines, and bladder, accompanied by severe constipation, difficulty swallowing, and malnutrition (Figure 1).¹⁷ MRI^18–20 and X-ray methods^21,22 have been useful for studying these organs in mice and for evaluating therapeutic strategies. For example, by using MRI, we demonstrated that the administration of the calcium channel blocker, verapamil, early (but not late) in the murine infection reduces the infection-associated increase in right ventricular internal diameter.^15,23 Treatment of infected mice with an endothelin-converting enzyme inhibitor also reduced right ventricular internal diameter, thus illustrating the role of endothelin in the pathogenesis of T. cruzi–induced cardiomyopathy.²⁴ MRI of the mouse model has also demonstrated a loss of adipose tissue, which is consistent with increased expression of enzymes associated with lipolysis.²⁵

Figure 1.

MRI and [¹⁸F]-FDG-PET images of control uninfected (A, C, and E) and infected (B, D, and F) mice. MRI of the gastrointestinal (GI) tract (A and B) and heart (C and D) and PET of the heart (E and F). Large white arrow indicates enlarged GI tract and right ventricle of the infected mice in B, D, and F.

These figures are reproduced from Jelicks et al¹² (C and D), Prado et al¹³ (E and F), and Ny et al¹⁸ (A and B), with permission of The American Journal of Tropical Medicine and Hygiene.

African Trypanosomiasis

Another neglected tropical disease (World Health Organization classification) is human African trypanosomiasis (HAT) or sleeping sickness. HAT is a parasitic disease transmitted by the tsetse fly that continues to be an important cause of human morbidity and mortality in sub-Saharan Africa due, in part, to armed conflicts resulting in population shifts. HAT is caused by infection with Trypanosoma brucei rhodesiense (East Africa) or T. brucei gambiense (West Africa). East African disease is actually a zoonosis among game animals, and humans become the accidental host. Disease progression is rapid, with early invasion of the central nervous system (CNS), and, if untreated, death occurs within 9 months. The Gambian type is a chronic disease, with invasion of the CNS occurring late, with a variety of neuropsychiatric disorders. Rodent and primate models mimic the CNS effects of the human disease. MRI has become an important tool for diagnosis of parasitic diseases of the CNS.²⁶ MRI studies of human patients with African trypanosomiasis report gadolinium enhancement, suggesting that the blood-brain barrier (BBB) is compromised by infection.^27–30 Recently, Rodgers et al³¹ used contrast agent–enhanced MRI to evaluate changes in the BBB integrity associated with the early CNS stage of the disease using a well-established murine model of HAT (Figure 2). They found that T₁- and T₂-weighted MRI with the administration of a gadolinium-based contrast agent (Magnevist; Bayer HealthCare, Uxbridge, Middlesex, UK) could detect significant dysfunction in the BBB of infected mice early in the CNS stage of the disease, when only mild to moderate histopathological changes are apparent. Additional MRI studies designed to evaluate BBB integrity throughout the course of trypanosome infection from the early acute stage, when no histopathological changes are detected, to post-treatment reactive encephalopathy, when the animals exhibit severe meningoencephalitis, are needed to establish the value of MRI for diagnosis and evaluation of therapeutic interventions in humans.

Figure 2.

A–D: MRI scans generated after the administration of contrast agent in an uninfected mouse (A and C) and an animal scanned 28 days after infection with T. brucei (B and D). Arrowheads indicate the presence of clear meningeal enhancement in the infected mouse compared with the uninfected animal imaging to assess BBB damage in murine trypanosomiasis (reproduced from Rodgers et al³¹ with permission of the authors and The American Journal of Tropical Medicine and Hygiene). E: Anatomical images from selected slices in control and infected mice. Slices shown are at the level of the caudate/putamen (CP), thalamus (TH), hippocampus (HC), and entorhinal cortex (ER) (arrows). The corresponding anatomical regions are indicated on each slice for the control animals. Con, control; Inf, infected. The images in panel E were reproduced from Kennan et al,³² with permission of Springer Publishing.

Malaria

Infection with Plasmodium falciparum, a causative agent of malaria in humans, accounts for almost one million deaths per year, and cerebral malaria is one of the most severe complications of this infection.³³ The pathogenesis of cerebral malaria is likely multifactorial and includes a reduction in cerebral blood flow associated with vasospasm, adherence of infected red blood cells to the endothelium, up-regulation of inflammatory mediators to the CNS and to the cerebral microvasculature,^32,34–37 and both systemic and cerebral metabolic disturbances.^38,39 Neuroimaging, particularly T₁- and T₂-weighted MRI and MR angiography, has become an important tool in elucidating the underlying mechanisms of cerebral malaria through longitudinal studies in animal models and humans that do not rely on autopsy material.^36,40–43

Although mouse models do not perfectly recapitulate human cerebral malaria, mice infected with P. berghei-ANKA (Antwerpen-Kasapa) are an excellent model for imaging. ANKA is a strain of P. berghei which confers experimental cerebral malaria in certain rodent strains. These rodents have been invaluable in helping to uncover early markers of the disease that are apparent even before any evidence of swelling in the brain, using T₁- and T₂-weighted MRI and MR angiography.^41,43 Flow-alternating arterial inversion spin-labeling MRI studies and single-voxel proton spectroscopy demonstrated decreased cerebral blood flow, as well as neuronal and axonal injury in mice infected with P. berghei-ANKA.^36,43 Saggu et al⁴¹ recently reported the first detection of damage to the optic and trigeminal nerves using T₂-weighted MRI. They previously characterized the disease in mice and observed several characteristic features, including BBB breakdown, hemorrhage, reduced brain perfusion, ischemia, hemodynamic dysfunction, and brain edema.^42,43 In the more recent report, using high-field (11.75-T) MRI, they identified damage to the cranial nerves as the earliest hallmark of the disease before any detectable brain swelling.⁴¹ They observed a hypointense signal in the trigeminal nerves, with significantly reduced dimensions. In addition, they demonstrated that the optic nerves were either hypointense or not visible in images of infected mice.⁴¹ These markers may be clinically relevant and could assist in the early detection of cerebral malaria.

By using MRI and MR angiography to study infected mice deficient in IL-12 receptor β2, Fauconnier et al⁴⁴ demonstrated that this molecule is essential for the development of cerebral malaria. Although wild-type mice developed the microvascular pathological characteristics associated with cerebral malaria, the mice deficient in IL-12 receptor β2 developed no neurological signs of the disease.⁴⁴

Findings of the mouse studies demonstrating the diagnostic value of MRI and the potential for therapeutic applications have been an important basis for the increased use of MRI in patients with cerebral malaria in endemic areas.⁴⁰

[¹⁸F]-FDG-PET imaging has been applied in studies of cerebral malaria in a nonhuman primate model and demonstrated decreased cerebral metabolic activity.³⁸ A diffuse and heterogeneous reduction of metabolic activity in the frontal and temporal lobes before any evidence of neuropathological findings was observed, suggesting that cerebral metabolic changes occur before parenchymal damage in primate cerebral malaria models.³⁸ A diffuse reduction in activity was postulated to result from decreased blood flow due to sequestration of parasitized red blood cells to the cerebral microcirculation.³⁸

Intravital microscopy can examine the living brain through a cranial window. This technique allows for long-term imaging of a single area in the brain for comparison of histopathological alterations and behavioral performances with microvascular changes.⁴⁵ Intravital microscopy has emerged as an important tool in determining the underlying pathological features contributing to cerebral malaria. Cabrales and coworkers^37,45,46 have taken advantage of this useful tool to explore the cerebral microvasculature during disease progression. Intravital microcopy was used to visualize the microvasculature after administration of a calcium channel blocker, nimodipine, to mice infected with P. berghei-ANKA and demonstrated reversal of the cerebral vascular disturbances during infection associated with improved survival and motor coordination.³⁷ In other studies, the benefits of treating cerebral malaria in a mouse model with exogenous nitric oxide were demonstrated.^46,47 Intravital imaging will contribute to a greater understanding of microcirculatory hemodynamics and vascular pathological characteristics during the pathogenesis of cerebral malaria in the mouse and will allow researchers to visually assess the function of specific vascular genes by infecting mice deficient in various genes to determine the efficacy of various therapeutic treatments.

Schistosomiasis

Schistosomiasis is a disease caused by several species of the trematode, Schistosoma, most notably S. mansoni, S. japonicum, and S. hematobium. The adults are intravascular. The first two species predominantly cause diseases of the liver and mesentery, whereas the latter one causes diseases of the urogenital tract. All three species have invaded the CNS as well. The major lesion is the formation of a granuloma surrounding the egg. This disease is diagnosed by the detection of ova in the feces and urine of infected individuals and by biopsy material. Salem et al⁴⁸ used fluorescence molecular tomography, MRI, and [¹⁸F]-FDG-PET imaging to evaluate the worm burden in mice infected with S. mansoni. [¹⁸F]-FDG uptake was correlated with worm burden and was useful for monitoring response to praziquantel treatment. Their results demonstrate the potential for using PET imaging in evaluation of therapeutics for this infection.

Bacterial Infections

Mycobacteria

Tuberculosis is a major public health problem worldwide. The mouse model of Mycobacterium tuberculosis has been extensively investigated. Infected mice have been used in a study of superparamagnetic iron oxide nanoparticles conjugated with a surface antibody, developed to improve diagnosis of extrapulmonary M. tuberculosis.⁴⁹ These M. tuberculosis nanoparticles resulted in a 14-fold increase in signal intensity of granulomas on T₂-weighted MRI images and provided a novel noninvasive method for diagnosing extrapulmonary M. tuberculosis infections. Although promising, more basic research must be performed to evaluate biodistribution and binding/endocytosis of the particles. In a PET study, Harper et al⁵⁰ used copper-64(II)-diacetyl-bis(N4-methyl-thiosemicarbazone), a tracer used to detect hypoxia, to evaluate hypoxia in tuberculosis lesions in mice. During acute infection or in control mice, there was no accumulation of copper-64(II)-diacetyl-bis(N4-methyl-thiosemicarbazone), whereas in chronically infected mice, the tracer accumulated in the lesions in a progressive, time-dependent manner. The accumulated copper-64(II)-diacetyl-bis(N4-methyl-thiosemicarbazone) colocalized with the lesion by CT imaging. [¹⁸F]-FDG-PET has been used to evaluate bactericidal activity of drug therapy in mice that were aerosol infected with M. tuberculosis.⁵¹ Lesion-specific [¹⁸F]-FDG-PET activity correlated with treatment in mice that develop caseating lesions. In another study, Davis et al⁵² used SPECT imaging of exogenously labeled M. tuberculosis. Mice were infected with wild-type M. tuberculosis or M. tuberculosis P_hsp60 thymidine kinase (TK) strains. The M. tuberculosis P_hsp60 TK strain was engineered to express TK using Phsp60, a highly active constitutive mycobacterial promoter. At specific time points, the mice were injected with 1-(2′deoxy-2′-fluoro-β-d-arabinofuranosyl)-5-[¹²⁵I]-iodouracil ([¹²⁵I]-FIAU), a nucleoside analog substrate for bacterial TK, and SPECT and CT imaging studies were performed at 3 and 24 hours after injection with the mice in a biocontainment device. SPECT imaging detected and localized the M. tuberculosis P_hsp60 TK strain, but not the wild-type M. tuberculosis. Their data suggest that as few as 5 to 10 million M. tuberculosis P_hsp60 TKs inside a granuloma could be detected using SPECT. These studies demonstrate the application of noninvasive imaging to monitor treatment response in small-animal models. The methods can be used to test therapeutics and develop regimens that can be extended to humans, who can then be monitored using the same type of imaging procedures.

Other Bacteria

Listeria monocytogenes is a Gram-positive bacillus that causes a variety of human and animal infections, including gastroenteritis, sepsis, miscarriage, stillbirth, and neonatal meningitis. Hardy et al⁵³ used US (for viability), MRI (for morphological characteristics), and BLI (for tracking L. monocytogenes) to study Listeria-induced miscarriage, using a pregnant mouse model. Ultrasonography revealed sustained bradycardia in the infected fetuses, although MRI detected no malformations of the fetuses, even when BLI indicated a high degree of infection. The study demonstrates the potential of using the multimodality approach to study the causes of miscarriage.

Bettegowda et al⁵⁴ used SPECT and CT imaging to localize bacterial infections with [¹²⁵I]-FIAU in mice. Escherichia coli, Staphylococcus aureus, Streptococcus pneumonia, Enterococcus faecalis, and Staphylococcus epidermidis were used to generate localized experimental infections in the thigh of mice; these mice were subsequently injected via the tail vein with [¹²⁵I]-FIAU imaged by SPECT at specific time points. All five strains of bacteria could be imaged with robust uptake at 4 hours after injection, and a high signal/noise ratio was observed 48 hours after injection. The authors speculate that [¹²⁴I]-FIAU, a positron emitter, or other positron-emitting nucleosides would be even more sensitive and that novel diagnostics and therapeutics could be developed based on the widespread presence and substrate specificities of bacterial TKs.

In another study of S. aureus, an engineered analog of prothrombin was used to study the bacteria in endocarditic vegetations with noninvasive fluorescence or PET imaging.⁵⁵ Fluorescent-labeled prothrombin (AF680-ProT) was injected into a mouse model of endocarditis, and imaging was performed using fluorescence molecular tomography fused to X-ray CT. High local concentrations of AF680-ProT were observed in S. aureus–induced vegetations 24 hours after injection of the probe. For the PET-CT studies, ⁶⁴Cu-diethylenetriaminepentaacetic acid-ProT was used, along with a genetically engineered S. aureus strain that expressed luciferase at sites of infection. The multimodality approach permitted confirmation of the presence of the bacteria using BLI, which was correlated with the PET imaging of ⁶⁴Cu-diethylenetriaminepentaacetic acid-ProT and demonstrated that PET imaging could be used to evaluate bacterial load. ⁶⁴Cu-PET has also been used to study the dissemination of Francisella tularensis, the cause of tularemia, when administered intranasally, intratracheally, intragastrically, intradermally, i.p., or i.v. in mice.⁵⁶ The results demonstrated that Francisella rapidly disseminates within hours to multiple tissues via most routes of administration, although different trafficking patterns were observed. Infection via the pulmonary routes resulted in rapid spread to the lung and gastrointestinal tract.

⁶⁷Gallium-citrate scintigraphy has been used in routine diagnostics of infections in the clinical setting; however, it has a long (3-day) half-life and is expensive, in addition to having safety concerns. ⁶⁸Ga, a PET tracer, has a short (68-minute) half-life, a lower cost, and fewer safety concerns. Nanni et al⁵⁷ tested ⁶⁸Ga-chloride as a PET tracer in mice infected with Chlamydia muridarum. Although the tracer demonstrated some promise for assessing genital infection, ⁶⁸Ga uptake was high in control mice with aseptic inflammation caused by the sham procedure and in the infected mice. In another study, Streptococcus pyrogenes infection and lipopolysaccharide inflammation in mice were investigated using fluorescent and SPECT/CT imaging with the tracer, ¹¹¹In-labeled tetraazacyclododecanete tetraacetic acid-biotin, linked to zinc-dipicolylamine-biotin with streptavidin.⁵⁸ There was significantly higher accumulation of this tracer in the live bacterial infection in one thigh compared with the sterile inflammation in the other thigh, suggesting that zinc-dipicolylamine may be useful for distinguishing between infection and inflammation. MRI has also been used in longitudinal studies of meningitis in mice.

Viral and Fungal Infections

BLI has been a powerful technique for tracking bacteria, fungi, parasites, and viruses that cause infectious diseases.^59,60 Kang et al⁶¹ have used BLI to monitor virus progression after CNS infection of mice with murine γ-herpesvirus. Murine γ-herpesvirus is similar to human γ-herpesvirus, Epstein-Barr virus, and Kaposi’s sarcoma–associated herpesvirus and provides a mouse model for studying the involvement in neurological diseases. After CNS infection, the virus spreads to the spleen, and latent virus could be activated from both the brain and the spleen. Their results suggest a role for the brain as a site for viral persistency after CNS infection. Rift Valley fever is another viral infection that has been studied by BLI in mice.⁶² The infection is typically asymptomatic or mild, although a few patients exhibit complications and death is associated with high viral load in blood. Real-time dissemination of the virus in immune-deficient mice was tracked and demonstrated that the thymus, spleen, and liver were infected first, with the liver being the main location for viral replication. BLI can also be used to track fungal infections. The fungus Aspergillus terreus is a life-threatening complication in immune-compromised patients. Slesiona et al⁶³ have used bioluminescent A. terreus in mice and demonstrated long-term persistence of the A. terreus conidia using BLI. Studies, using the luminescent bacteria strain, S. aureus Xen 29, have demonstrated the use of BLI to evaluate therapeutics in mice.^64,65 These studies represent a few recent examples of the potential value of BLI for tracking infection.

MRI has also been valuable for studying viral infections. Diffusion-tensor MRI has been applied to study neuronal loss during HIV-1 infection in a humanized mouse model.⁶⁶ Structural changes in gray matter were revealed by imaging and confirmed by immunohistochemistry. To our knowledge, this study was the first to demonstrate such associations and underscored the potential of humanized mouse models for research in infectious diseases.

Concluding Remarks

Much progress has been made in tracking parasites, bacteria, fungi, and viruses using multimodality small-animal imaging approaches.^59,60 The future development of multimodality molecular imaging studies for investigating the pathogenesis of infection and for evaluating the therapeutics and translation to the clinical setting for diagnostic and theranostic applications will provide researchers and clinicians the tools to have a positive impact on patient care.