Introduction
Zika virus (ZIKV) is a mosquito-transmitted infection affecting developing fetuses, children, and adults. It causes sporadic outbreaks, the largest of which occurred in 2015 and was declared a public health emergency of international concern by the World Health Organization (WHO) [
1]. Half of the global population is currently at risk of infection and live in tropical and subtropical regions where
Aedes mosquitoes—the vector responsible for the transmission of ZIKV—are endemic [
2,
3]. With global climate change, rapid population growth in the tropics, and an expected return to high-volume air travel in the post-COVID pandemic era, even more people will be at risk of exposure to such mosquito-borne viruses. Despite the continued public health threat of ZIKV, no therapeutics or vaccines are approved for human use. Various vaccine candidates and antivirals were developed and evaluated in clinical trials (reviewed in [
4,
5]) but met with various challenges in recent years. Vaccine and therapeutic efficacy trials are hampered by the difficulty in patient recruitment due to the low natural ZIKV infection rates and require controlled deliberate infection of trial participants [
6]. Moreover, global efforts in vaccine development have been dampened by refocusing of resources towards the current COVID pandemic.
One way to expedite the bench-to-bedside development of these therapies is by employing non-invasive imaging biomarkers. These facilitate spatial localization of viral disease (or disease severity) and enable deeper in situ molecular insights into disease pathogenesis. This is not attainable with the current diagnostic gold standard of detecting active viral replication in peripheral blood using polymerase chain reaction (PCR) [
7]. Imaging biomarkers that correlate with disease progression have the potential to predict response to therapy and rapidly inform go or no-go decisions in drug evaluations. This is most relevant in the context of developing therapeutics against highly virulent and lethal pathogens—such as Ebola and Nipah viruses—but is nonetheless relevant to ZIKV and other related flavivirus diseases, including the endemic dengue virus (DENV) [
8]. Therefore, there is a real clinical need to identify and characterize infectious disease-related imaging biomarkers [
9].
We previously identified
18F-fluorodeoxyglucose ([
18F]FDG) uptake as a preclinical imaging biomarker for early prediction of response to therapeutic interventions for DENV infection. Lesions of increased [
18F]FDG uptake in the intestines were found to correlate with disease severity, tissue viral burden, and expression of inflammatory cytokines in lethal and non-lethal murine models of DENV infection [
10]. Clinical [
18F]FDG-PET also distinguishes infection-mediated tissue inflammation patterns, namely, in lymphoid tissues, in DENV patients during early disease [
11,
12]. In contrast, there have been no reports of [
18F]FDG-PET imaging in ZIKV disease, in either humans or animal models.
Here we examined [18F]FDG uptake in the lymphoid organs of a murine model of acute lethal ZIKV disease. The spatial and temporal dynamics of metabolic tracer uptake in ZIKV were also compared with DENV to identify common tissue-specific imaging biomarkers between these two related viruses. Tissue uptake in ZIKV was further correlated with pathology and markers of viral disease and inflammation in tissues. This study demonstrates that splenic [18F]FDG uptake is a robust surrogate for interrogating of tissue viral burden and inflammation status in ZIKV.
Discussion
To date, ZIKV studies have mainly focused on neurological complications resulting from maternal infection, as well as male infertility post-infection. Hence, the effect of ZIKV infection on the induction of the inflammatory response remains poorly understood. In this preclinical study, we investigated how [
18F]FDG-PET could serve as a non-invasive imaging biomarker of ZIKV disease and inflammation. The hallmark of murine ZIKV disease in [
18F]FDG-PET was high tracer uptake in the lymphatic tissues, especially the spleen and axillary lymph nodes (Ax.LN). Consistent with previous studies [
24,
25], ZIKV infection in the spleen resulted in gross tissue enlargement and increased viral replication, with the latter trending intimately with tracer tissue uptake. Additionally, this study demonstrated that there are elevated levels of the pro-inflammatory cytokines IL-6 and TNF-α in the spleen, which correlated strongly with tracer in the tissue. In contrast, [
18F]FDG in Ax.LN only correlated with viral burden and not inflammatory response. With marked differences between [
18F]FDG uptake and associated ex vivo biomarkers at mid and late disease, early uptake patterns in lymphoid tissues appear to be important in dissecting acute viral ZIKV pathogenesis and host inflammatory response.
The dynamic changes in the immune landscape of ZIKV-infected spleens with increasing disease severity and which immune cell subset/s may be driving the FDG uptake within the tissue have not been studied previously. Here we demonstrate that during ZIKV disease progression, total CD45
+ immune cells in the spleen were unchanged while total CD45
¯ non-immune cells increased. Lymphocytes, specifically B and T cells, in the spleen declined during disease, and this is consistent with the known permissiveness of lymphocytes to ZIKV infection [
26]. This decline was counterbalanced by increase in myeloid cells (mainly granulocytes, monocytes, monocyte-derived macrophages (Mo-MAC), and dendritic cells) infiltrating the tissue, as supplied by the bone marrow. In addition, immune cell proliferation characterized by increased ki67 expression was found to be significantly elevated in the spleen and strongly correlated with spleen [
18F]FDG uptake. This implies that the increased energetic demands of highly proliferating cells may be driving the tracer tissue uptake during disease. ki67 expression is reported to positively correlate with [
18F]FDG uptake in various cancers [
27,
28], but this relationship has not been concretely established in the context of infectious disease. A more direct way of resolving which immune cells in the spleen drive [
18F]FDG uptake is by measuring the glucose transporter expression. The glucose transporter (Glut-3) specifically directs glucose and [
18F]FDG uptake in immune cells and is further upregulated upon immune-cell activation [
29,
30]. Subsequent gamma counting of sorted immune cell subsets gated on high Glut-3 expression collected from animals injected with [
18F]FDG would then directly identify the specific immune cells driving increased FDG uptake. This method had been described recently in a study aimed at identifying specific cells that take up a radiotracer in the brain [
31,
32].
In addition to uptake in lymphoid tissues, we also expected increased [
18F]FDG uptake in ZIKV brains and gonads—the tissues known from other mouse ZIKV infection models to be highly permissive to infection [
20,
24,
33]. PET/CT could not discriminate uptake in the testes from the bladder, despite seeing trends towards elevated tracer uptake in ZIKV-infected testes through DAR and gamma counting. Therefore, [
18F]FDG may not be a reliable imaging biomarker to interrogate disease and inflammation that had been well characterized in ZIKV-infected gonads [
34,
35]. Another known complication of ZIKV disease is neuroinflammation resulting from infection of the central nervous system [
33,
36]. Neither PET/CT, DAR, nor gamma counting registered increased [
18F]FDG uptake in ZIKV brains despite strong evidence of neuroinflammation observed in harvested ZIKV brains at late disease (
data not shown). The tracer lacks sensitivity in detecting neuroinflammation due to high basal uptake in normal brain tissues. Imaging approaches targeted to molecular markers of neuroinflammation, such as the translocator protein (TSPO), may be more appropriate tools for interrogating ZIKV infection and host response in the brain [
37]. In addition, larger preclinical cohort studies are required to confirm the significance of the [
18F]FDG uptake patterns in brain and testes to show their predictive value.
To date, there are no virus-specific molecular imaging biomarkers. Previous attempts to develop DENV-targeted molecular imaging probes using labeled antisense oligonucleotides that bind to DENV genome sequences resulted in disappointing in vitro results and failed to reach preclinical evaluations [
38]. To identify PET imaging biomarkers applicable to various viral diseases, we compared the [
18F]FDG-PET profile of the ZIKV model with the closely related DENV model using the same type I/II-interferon receptor-deficient AG129 mouse infection model [
18,
19]. Both models exhibited similar phenotypes of high lethality (Fig.
S1a), severe weight loss (Fig.
S1b), and spleen enlargement seen in mid and late ZIKV and in late DENV disease (Fig.
S1c). [
18F]FDG uptake in DENV spleen is not as robust as intestinal FDG uptake as previously reported in preclinical models [
10]. However, this current study confirmed that in DENV, spleen [
18F]FDG uptake is still sufficiently sensitive as a possible imaging biomarker for disease and more closely recapitulates the clinical PET/CT observations in DENV patients [
11,
12]. Aside from DENV and ZIKV, lymphoid tissue [
18F]FDG uptake has also been reported in other viral infections, seen in [
18F]FDG-PET images of patients infected with chikungunya virus [
39], reactivated Epstein-Barr virus [
40‐
42], and reactivated varicella zoster virus or shingles [
43]. Lymphoid tissue FDG uptake was also captured in non-human primate models of monkeypox virus infection [
44,
45]. Thus, lymphoid tissue [
18F]FDG uptake may be reflective of the general host immune response to invading viruses, and the dynamic changes in uptake observed in these tissues could be exploited for further development of [
18F]FDG as an imaging biomarker to support screening of antivirals or vaccine treatments for ZIKV and other viral infections.
In conclusion, [18F]FDG uptake in the spleen is a useful surrogate for interrogation of in situ tissue viral burden and inflammation status in this ZIKV murine infection model. Moreover, it can also be used to monitor in situ tissue viral replication and IL-6 expression in the DENV murine model. Studies evaluating these imaging biomarkers in human infections and whether experimental therapeutics that ameliorate disease also dampen [18F]FDG uptake in the spleen are also worth pursuing.
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.