We used a tree shrew (TupaiaBelangeri, family Tupaiidae) model to study clinical signs, virus shedding, pathology of influenza virus A H1N1 and sialic acid receptor type distribution. Our results demonstrated that influenza H1N1 virus replicated efficiently in respiratory tract of tree shrews, and showed mild or moderate clinical signs and pathological changes. These findings in tree shrews seemed in accord with related manifestations in human influenza infections . It also revealed that upper respiratory tract of tree shrew may be more permissive to human clinical isolates when inoculated by intranasal route. Patterns of influenza virus receptor distribution in the upper and lower respiratory tract are also similar in tree shrews and humans [29–32]. Taken together, our results suggested that tree shrews could be a promising alternative animal model for the study of influenza pathogenesis.
Main clinical signs in this model include slightly increased body temperature and nasal secretion, but anorexia and lethargy were not obvious. Although sneezing occurred occasionally, it was more frequently associated with the common cold than with influenza. Tree shrews usually had fever after inoculation with high challenge dose of influenza virus, and dropped to undetectable levels by around one week, which was similar to human infection timeline . Because both of systemic and respiratory symptoms were characteristically observed in tree shrews infected, indeed the symptomatology found in tree shrews was partly similar to human influenza infections without complication . Additionally, the disease manifestations of influenza virus infection in tree shrews also partly resembled those in an excellent mammalian animal model (eg. ferret) . However, loss appetite, congested eyes and otologic manifestation were not observed in tree shrews, but in ferrets [33, 34]. The mouse model can manifest no obvious clinical signs of influenza-like illness, but develop severe pneumonia. Thus far, the tree shrews seem not superior to the ferret model, but have the advantage over the rodent model in the clinical similarity. The serological data showed that tree shrews readily seroconvert in response to intranasal inoculation of virus, and serum neutralizing antibody titers of infected animals range from 80 to 640. Considering the clinical symptoms and antibody immune response observed in tree shrews, it is reasonable to presume this model is potential to evaluate the efficacy of antiviral agents and vaccine for the prevention of influenza infection. However, the tree shrew infected with human influenza model didn’t result in lethality, thereby detection of increasing nasal wash titers can serve as endpoints for determination of vaccine efficacy.
In particular, in our study the tested influenza H1N1 viruses (including classic strain, seasonal isolate and novel swine H1N1) could infect the tree shrews without prior adaptation. Thus, tree shrew model could be one of the attractive options for the study of pathogenesis and antiviral agent shortly once new influenza viruses emerge.
Like humans, tree shrews inoculated with human influenza viruses demonstrated a primarily upper respiratory tract infection. Influenza viruses could be isolated at high titers from nasal washes, but no virus could be detected in tracheas and lungs. In addition, the infections in tree shrew were only induced by the high virus challenge dose (up to105 TCID50), without mortality. Nevertheless, it is likely that the tested human influenza infections are self-limited in tree shrew model like humans. Histopathological analysis indicated that clinical isolate influenza H1N1 virus infection caused exudative and interstitial pneumonia, moderate bronchitis, mild bronchiolitis, interstitial edema and inflammatory infiltrates, which showed considerable similarities to influenza virus pneumonia in human  and this also suggested that damage to the ciliated epithelium may be caused by the inflammatory response.
It is important to note that receptors play a crucial role in determining the host specificity and tissue tropism of virus . The hemagglutinin of influenza viruses initiates infection by binding sialic acid (SA) that is bound to glycans through SA α2,3 Gal or SAα2,6 Gal linkage . Therefore, the lectinhistochemistry data are important to evaluate the tree shrew as a model for influenza virus. We established that both human influenza (SA a2,6-Gal, SNA) and avian (SA a2,3-Gal, MAA II specific) receptor types were present in tree shrew respiratory tract, with each tissue showing distinctive anatomical distribution of the two receptors. This suggested that the respiratory tracts may be permissive to viral entry or infection. In tree shrews, the SA α2,6 Gal receptor that is more frequently associated with human influenza viruses was restricted primarily to the trachea and some bronchus, whereas the SAα2,3 Gal receptor preferentially bound by avian viruses was more abundantly present in the pulmonary alveoli and respiratory epithelium, which was similar to humans and ferrets [36, 37]. Considering the anatomical receptor distribution in upper and lower respiratory tract in tree shrew, it is reasonable to hypothesize that the tree shrew can be infected with human influenza viruses, but also potentially be infected with high pathogenic avian influenza virus, even resulting in a lethal pneumonia model. However, this requires further investigation.
Currently, tree shrews are widely used in medical and biological research, especially in virology. However, tree shrews, used as experimental animals, need a suitable source for this study. Many countries now carry out laboratory research on tree shrews, such as in the German Primate Center , and tree shrews have also been bred successfully. In China, Kunming Medical University has already achieved local standards for experimental tree shrews, which provides comprehensive quality assurance for artificial propagation and experimental studies of tree shrews [39, 40]. In addition, although ferret is possibly the best model for influenza in humans, the use of this species for routine research purposes is prohibitively expensive [12, 37]. However, compared to ferrets, tree shrews can be bred more easily and are relatively inexpensive to maintain for study.
In conclusion, our study aimed to determine whether the tree shrew provides a useful small alternative model for the study of influenza H1N1 virus infection. The main advantages of tree shrews as an experimental model are phylogenetically close to primates , susceptible to influenza without prior adaptation, small size, low expense and husbandry, which make the model more accessible to the researchers. Although the tree shrews mimic well the disease in humans, the absence of severe infection and possible differences in drug pharmacokinetics in tree shrews and humans may limit the study of antiviral treatment. Further efforts are needed to determine its pathogenesis in tree shrew model such as viral replication in the extra pulmonary organs as well as its application of assessments of antiviral agents and vaccine. To this end, the tree shrew model will be useful for assessment of circulating strains that could challenge human public health.