Brugia malayi infection in ferrets - A small mammal model of lymphatic filariasis

Abstract

Background: The lack of effective short-course therapies for treatment of the adult stage of filarial worms is a major limitation in the global effort to eliminate lymphatic filariasis. Studies using current small mammal models of lymphatic filariasis are limited by difficulties in quantifying adult worm numbers and in assessing lymphatic anatomy and function.

Methodology/principal findings: Here, we re-established Brugia malayi infection of ferrets as a model for lymphatic filariasis and demonstrated parasitological, immunological, and histological parallels with human infection. Subcutaneous injection of L3 larvae into a hind-footpad resulted in a mean of 18 adult worms recovered 16 weeks post-infection, primarily from the draining inguinal and femoral lymphatics of the injected limb. Infected ferrets developed microfilaremia, with patency lasting from 12-26 weeks post-infection. Quantitative PCR assessing cytokine transcription by antigen-stimulated lymph node cells demonstrated a mixed Th1/Th2 response occurring during early infection. Immunoregulation with production of down-regulatory cytokine IL-10 occurred just prior to peak microfilaremia. Histological analysis revealed progressive inflammation of the lymphatic vessel walls, with intimal thickening and disorganization of collagen fibers. Inflammation was observed as early as 8 weeks post-infection and extended into the perivascular and subcutaneous tissues by 16 weeks post-infection. Finally, we developed a novel ferret PET/CT lymphoscintigraphy method demonstrating substantial changes in lymphatic anatomy and function as early as 3 weeks post-infection, with progression over the course of infection.

Conclusions/significance: B. malayi infection of ferrets is a robust model of human lymphatic filariasis that can be utilized to study efficacy of novel antifilarial agents against adult worms residing within lymphatic vessels. In conjunction with PET/CT lymphoscintigraphy, this model can also be used to investigate pathogenesis of lymphatic dysfunction in lymphatic filariasis and efficacy of medications aimed at reversing lymphatic dysfunction after clearance of adult worms.

Fig 1. Adult B. malayi worm recovery from infected ferrets.

A) Number of adult worms recovered from animals euthanized at 8 (n = 8), 16 (n = 14), and 28 (n = 4) weeks’ post-infection. Not statistically significant by Kruskal-Wallis test, followed by Dunn post hoc multiple comparisons. B) Number of adult worms recovered from the lymphatic vessels of the leg into which larvae were injected versus number of adult worms recovered from the PBS injected leg; n = 26 per group; ***p < 0.001 (Mann-Whitney test). Blue and pink circles indicate male and female ferrets, respectively.

Fig 2. Timecourse of microfilaremia, eosinophilia and plasma levels of BmAg-specific IgG in B. malayi-infected ferrets.

The mean and SEM values of (A) microfilaria per milliliter of blood, (B) eosinophil numbers per microliter of blood (shaded box indicates the normal cell range for ferrets), and (C) BmAg-specific IgG levels produced following B. malayi infection. Weeks 0 to 8 PI, n = 12; weeks 10 to 16 PI, n = 8; weeks 18–28, n = 4. Baseline values of Mfs/ml, eosinophil numbers, and antibody ODs were compared to corresponding values at the indicated post-infection timepoints for statistical significance; *p< 0.05, **p<0.01, ***p<0.001 (Kruskal-Wallis test, followed by Dunn post hoc multiple comparisons).

Fig 3. Ex vivo stimulation of lymphocytes with B. malayi antigen induces proliferation.

Proliferation of (A) Splenocytes and (B) draining lymph node cells in response to stimulation with microfilariae antigen (Mf-Ag), gravid female antigen (BmAg), or adult male antigen (Male-Ag). Shown here are the mean and SEM values. nd = no data. Not statistically significant by Kruskal-Wallis test, by Dunn’s multiple comparisons (Mf-Ag and BmAg comparisons), nor by Mann-Whitney (Male-Ag comparisons).

Fig 4. Ex vivo stimulation of splenocytes with BmAg reveals that modulation of cytokine gene transcription is dependent on duration of B. malayi infection.

Cytokine transcription profiles, IFNγ (A), IL-4 (B), and IL-10 (C) of splenocytes stimulated for 72 hours with or without 20 μg/ml of BmAg (antigen derived from gravid female worms). M = Media alone; ND = Not Detected; n = 4 for each data set. Differences in transcript amounts produced by stimulated versus unstimulated lymphocytes were analyzed for statistical significance; *p< 0.05 (Mann-Whitney test).

Fig 5. Histology of lymphatic vessels and surrounding tissue.

(A) un-infected controls (20x), (B) ferrets infected for 8-weeks (40x), (C) 16-weeks (20x), and (D) 28-weeks (20x). Formalin treated tissue samples were stained with hematoxylin and eosin.

Fig 6. Histology of lymphatic vessels and surrounding tissue from lymphatic vessels without (A, B - 20x) and with (C -20x, D - 10x) resident adult worms.

Formalin treated tissue samples were stained with hematoxylin and eosin.

Fig 7. Visualization of lymphatic anatomy using PET imaging.

PET and CT images were acquired from ferret F814 pre-infection and 3-, 8-, 20, and 28-weeks post B. malayi infection. The portrayed maximum intensity projection (MIP) images were constructed by collapsing data from 60 individual PET files (each consisting of a one-minute scan) into a single static frame. The resulting PET image was aligned to the corresponding CT image. Green bars show the designated Region-of-Interest used for kinetic analysis of tracer flow depicted in Fig 8.

Fig 8. Kinetic analysis of PET imaging demonstrates lymphatic dysfunction at all timepoints post-infection.

Detection over time of tracer-containing lymph at the mid-femur ROI for ferrets F814 (A) and M433 (B)—signal is reported as % of total injected ¹⁸F-FDG. (C-D) Elapsed time for tracer to reach 0.025% of the total injected at the ROI. Dashed line indicates 0.025% threshold used for elapsed time analysis.