Rational attenuation of canine distemper virus (CDV) to develop a morbillivirus animal model that mimics measles in humans

Abstract

Morbilliviruses are members of the family Paramyxoviridae and are known for their ability to cause systemic disease in a variety of mammalian hosts. The prototypic morbillivirus, measles virus (MeV), infects humans and still causes morbidity and mortality in unvaccinated children and young adults. Experimental infection studies in non-human primates have contributed to the understanding of measles pathogenesis. However, ethical restrictions call for the development of new animal models. Canine distemper virus (CDV) infects a wide range of animals, including ferrets, and its pathogenesis shares many features with measles. However, wild-type CDV infection is almost always lethal, while MeV infection is usually self-limiting. Here, we made five recombinant CDVs, predicted to be attenuated, and compared their pathogenesis to the non-attenuated recombinant CDV in a ferret model. Three viruses were insufficiently attenuated based on clinical signs, fatality, and systemic infection, while one virus was too attenuated. The last candidate virus caused a self-limiting infection associated with transient viremia and viral dissemination to all lymphoid tissues, was shed transiently from the upper respiratory tract, and did not result in acute neurological signs. Additionally, an in-depth phenotyping of the infected white blood cells showed lower infection percentages in all lymphocyte subsets when compared to the non-attenuated CDV. In conclusion, infection models using this candidate virus mimic measles and can be used to study pathogenesis-related questions and to test interventions for morbilliviruses in a natural host species.IMPORTANCEMorbilliviruses are transmitted via the respiratory route but cause systemic disease. The viruses use two cellular receptors to infect myeloid, lymphoid, and epithelial cells. Measles virus (MeV) remains an important cause of morbidity and mortality in humans, requiring animal models to study pathogenesis or intervention strategies. Experimental MeV infections in non-human primates are restricted by ethical and practical constraints, and animal morbillivirus infections in natural host species have been considered as alternatives. Inoculation of ferrets with wild-type canine distemper virus (CDV) has been used for this purpose, but in most cases, the virus overwhelms the immune system and causes highly lethal disease. Introduction of an additional transcription unit and an additional attenuating point mutation in the polymerase yielded a candidate virus that caused self-limiting disease with transient viremia and virus shedding. This rationally attenuated CDV strain can be used for experimental morbillivirus infections in ferrets that reflect measles in humans.

Keywords: attenuation; canine distemper; ferret; measles; model.

Fig 1

Schematic overview of rCDV constructs. Attempted rational attenuation of rCDV^RI by introducing an ATU (Venus) in position 1 (brown), or deleting the C protein (red), introducing EGFP into the L gene (green), or mutating position 589 (vertical line) of the L gene (purple) of rCDV^RIVenus(6), or introducing Venus at position 1 and mutating position 589 of the L gene (blue). rCDV^RIVenus(6), previously shown to be pathogenic in ferrets and raccoons, is shown in gray for comparison. The colors for the different rCDVs are consistent throughout the manuscript.

Fig 2

Clinical parameters of ferrets inoculated with different rCDVs. (A) Change in body temperature relative to day 0 for animals inoculated with different rCDVs (in degrees celcius). (B) The severity score, an average score of body temperature, body weight, lymphocyte count, virus isolation from throat swabs, and infection of lymphocytes for all rCDVs. The severity score of MeV-inoculated ferrets is shown as a negative control. Shaded bands depict the standard error of the mean (SEM). (C) Kaplan-Meier curve depicting the percentage survival of animals in each group.

Fig 3

Shedding of rCDVs. Infectious rCDVs were isolated from (A) throat, (B) rectum, and (C) nose swabs on VDS cells. Titers were determined by endpoint titration and are expressed as TCID₅₀/mL. Shaded bands depict the SEM.

Fig 4

rCDV viremia and lymphopenia. Detection of rCDVs in peripheral blood via (A) virus isolation or (B) flow cytometry. (A) WBCs obtained from rCDV-inoculated ferrets were titrated on VDS cells, and viral loads were expressed as the number of infected cells per 10⁶ total cells. (B) Infection percentage of ferret lymphocytes measured by flow cytometry. The percentage of Venus⁺ cells in the singlet lymphocyte gate was determined. (C) Lymphocyte counts are shown relative to day 0 for animals inoculated with the different rCDVs. Shaded bands depict the SEM.

Fig 5

Phenotype of rCDV^RIVenus(1)-L_H589Y-infected WBC in comparison to rCDV^RIVenus(6)-infected WBC. (A and B) Infection percentage of different cell types measured by flow cytometry. The percentage of Venus⁺ cells in the indicated population was determined for (A) rCDV^RIVenus(6) and (B) rCDV^RIVenus(1)-L_H589Y. Per time point, all available samples were evaluated. Shaded bands indicate the SEM. (C and D) Phenotyping of rCDV-infected cells on 6 dpi via inverse gating of Venus⁺ cells within the singlet gate. The relative population sizes are shown as the average population sizes of (C) n = 14 rCDV^RIVenus(6)-infected ferrets and (D) n = 6 rCDV^RIVenus(1)-L_H589Y-infected ferrets. (E and F) Subsets of WBC were identified in peripheral blood via flow cytometry at the indicated time points for (E) rCDV^RIVenus(6) and (F) rCDV^RIVenus(1)-L_H589Y. The relative population size at each time point is shown as the average population size of all available samples per time point. The gating strategy can be found in Fig. S3.

Fig 6

Detection of Venus⁺ cells for rCDV^RIVenus(6)- and rCDV^RIVenus(1)-L_H589Y-infected animals in BAL cells and different lymphoid tissues over time. (A–D) Infection percentages were measured in single-cell suspensions from different tissues via flow cytometry at (A) 4, (B) 6, (C) 8, and (D) 20–23 dpi. The percentage of Venus⁺ cells within the singlet gate is shown. Symbols represent individual animals, bars the average, and error bars depict the SEM. Man LN, mandibular lymph node; Tb LN, tracheobronchial lymph node; BAL, broncho-alveolar lavage; Ax LN, axillary lymph node; Ing LN, Inguinal lymph node; Mes LN, mesenteric lymph node.