All animal experimentations were conducted at the specific pathogen-free animal facility of the Bernhard Nocht Institute for Tropical Medicine (BNITM) in agreement with the German Animal Welfare Act and the relevant German authority (Behörde für Gesundheit und Verbraucherschutz, Hamburg, approval numbers 84/15, 103/2018. All mice were kept in individually ventilated cages (maximum of 5 mice per cage). Mice were sacrificed by an overdosed CO₂ narcosis followed by cervical dislocation in accordance with the German animal protection law. 8-12 week old C57BL/6 mice were used and either bred in the animal facility of the BNITM or were obtained from Janvier Labs.

The life cycle of L. sigmodontis was maintained in their natural reservoir, the cotton rats (Sigmodon hispidus). Therefore, cotton rats were anesthetized by isoflurane narcosis and blood was collected from the retro-bulbar sinus in order to count the microfilariae (MF, L1). Cotton rats, which were used for further infection of blood-sucking mites (Ornithonyssus bacoti), had an infection rate of 500 – 2000 MF per µl blood. Infected cotton rates were exposed to mites that ingested MF during a blood meal. Infected mites were kept at 29°C and 90% humidity for 14 days to allow maturation of L1 to L3. Experimental mice were anesthetized with ketamine/xylazine (100 mg and 5 mg/kg body weight) and exposed to these infected mites for 16 hours, i.e., naturally infected. To quantify the worm load, infected mice were sacrificed at indicated time points and the thoracic cavity was flushed with 10 mL PBS. Viable worms were defined by movement. Coated worms were defined as worms either moving or immobile that were covered with host cells. By contrast, dead worms were free of visible host cells and did not move anymore. The anthelminthic drug flubendazole (FBZ, Sigma, Germany) was administered by repeated s.c. injection on 5 consecutive days with a dose of 5 mg/kg body weight.

For influenza infection C57BL/6 mice were anesthetized with ketamine/xylazine (100 mg and 5 mg/kg body weight) and i.n. infected with 25 µL 1 x 10³ plaque forming units (PFU) 2009 pH1N1 influenza A Hamburg/NY1580/09. The influenza virus was isolated from pharyngeal swabs of a female patient as described previously (23). Health status of the mice was monitored daily according to the animal protocols approved by the Hamburg authorities until day 3 post challenge infection when mice were sacrificed to quantify viral burden in the lungs.

Mice were vaccinated by i.p. injection of 3.75 µg either non-adjuvanted (Begripal, Seqirus) or adjuvanted (Fluad, Seqirus) vaccine against influenza in 200 µL PBS (Carl Roth, Germany). Blood was collected from the vena fascialis at the indicated time points, 2-4 weeks after vaccination, and allowed to coagulate for 1 h at RT. After centrifugation (10.000 x g for 10 min) serum was transferred into a fresh tube and stored at -20°C until further analysis.

Murine serum samples were thawed and heat inactivated for 30 min at 56°C in a water bath (Memmert, Germany). Serial dilutions of sera in 25 µL PBS (Carl Roth, Germany) were incubated with 25 µl of 2009 pH1N1 influenza A virus in duplicates in 96-well-V-plates (Greiner Bio-one, Austria). The virus solution was standardized to 8 hemagglutination units before. After 30 min incubation at room temperature, 50 µL 1% fresh chicken red blood cell solution (Lohman, Germany) in 0.9% NaCl (Carl Roth, Germany) was added to each well. The dilution that still inhibited agglutination was calculated as a titre after a further 1-hour-incubation at 4°C.

Madin-Darby canine kidney (MDCK) cells were provided by G. Gabriel (Heinrich-Pette Institute, Hamburg, Germany). Cells were stored in liquid nitrogen and thawed 2 weeks prior to the plaque assay. Cells were quickly transferred to pre-warmed RPMI 1640 medium with L-Glutamine (Lonza, Switzerland) supplemented with 10% FCS (Capricorn Scientific, Germany) and washed twice. Cells were grown in T75 flasks (Greiner Bio-one, Austria) for 2 weeks in 15 ml RPMI 1640 medium with L-Glutamine supplemented with 5% FCS, Hepes (Lonza, Switzerland), and Gentamycin (Lonza, Switzerland). Cells were splitted every 3-4 days in a 1:6 ratio. One day prior to the start of plaque assay, cells from one confluent T75 were splitted to 4-5 6-Well plates (Greiner Bio-one, Austria). MDCK cells were grown overnight at 37°C in 3 ml RPMI cell culture medium which results in a semi-confluent cell layer.

Viral loads were determined in lung homogenates by MDCK plaque assay. Lungs were removed at indicated time points and stored at -70°C in 0.1% BSA (Serva, Germany) in PBS. Lungs were thawed and homogenized. The supernatant was collected after centrifugation (10 min, 950 x g, 4°C). Serial dilutions of lung homogenates (10-¹ to 10^-6) were added to a confluent MDCK cell culture in 6-well-plates. After 30 min at 37°C, inoculum was removed, cells washed one time with 1x PBS and overlaid or plates were directly overlaid with 3 mL 1.25% Avicel (FMC, USA) in Minimum Essentiel Medium (Lonza, Switzerland) containing 1µg/mL N-tosyl-L-phenylalanine chloromethyl ketone (TPCK) treated -Trypsin (Sigma, Germany). After a further incubation for 72 h at 37°C and 5% CO₂, the overlay was removed and the plates were washed with PBS. The cells were fixed with 0.5 mL 4% paraformaldehyde (PFA, Carl Roth, Germany) for 30 min at 4°C. PFA was removed and plates were incubated for 10 min with 1 mL/well 1% crystal violet (Merk, Germany). The staining was stopped by removing crystal violet and washing the plates with tap water.

The thoracic cavity was flushed with 10 ml cold PBS. Cells were sedimented by centrifugation (1200 rpm, 4°C, 6 min) and the number of cells determined by counting with trypan blue (Sigma, Germany). Spleens were removed and transferred to a 50 ml tube with 10 ml PBS and kept on ice until further processing. Spleens were pounded between two glass slides (Engelbrecht, Germany) in a petri dish (Sarstedt, Germany) and transferred back to the 50 ml tube. After centrifugation (1200 rpm, 4°C, 6 min), the supernatant was discarded and erythrocytes lysed by addition of 5 ml ACK lysis buffer (150 mM NH₄Cl, 1 mM KHCO₃, 0.1 mM Na₂EDTA) for 2 min at room temperature. Lysis was stopped by addition of 30 ml cold PBS. Cells were centrifuged again and the supernatant discarded. The cell pellet was resuspended in 10 ml PBS and filtered over a 70 µm Falcon™ cell strainer (Corning). The cell strainer was washed with 10 ml PBS and the cell suspension centrifuged. The obtained cell pellet was resuspended in 10 ml PBS and counted in a 1:10 dilution with trypan blue. To obtain single cell suspension cells were vortexed during washing steps. Single cells were confirmed by trypan blue (Sigma, Germany) cell counting under a microscope.

Single cells (3 x 10⁶) from the spleen or thoracic cavity (TC) were stained with 1 µL Zombie UV™ Fixable Viability Kit or Zombie Aqua Fixable Viability Kit in 1 mL PBS for 30 min at 4°C. For surface staining, cells were first stained for 15 min at 37°C with APC-labelled anti-mouse LAG-3 (clone: C9b7W, Lot: B215463) and PE/Cy7-labeled anti-mouse CD49b (clone: Hma2, Lot: B210811) Ab. After an additional incubation for 15 min at RT, cells were washed and stained for 30 min at 4°C with BV510-labeled anti-mouse CD4 (clone: RM4-5, Lot: B277537). For subsequent intracellular staining with anti-mouse AF700-labeled anti-mouse Foxp3 (clone: FJK-16S, Lot: 4307313), cells were fixed and permeabilized with Thermofisher Scientific Foxp3/Transcription factor staining buffer set (Thermofisher, Germany) according to the manufacturer’s protocol. Samples were analysed on a LSRII (Becton Dickinson) using FlowJo software (TreeStar).

Data were analysed using Graph Pad Prism, tested for normality distribution and further tested either with unpaired t test (parametric) or Mann-Whitney test (non-parametric). Asterisks for all analyses * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, **** p ≤ 0.0001.

We use mice that are infected with the filarial parasite L. sigmodontis mice as an established model for chronic human filarial infections (10–12) to study the impact of concurrent helminth infection on the efficacy of vaccinations in the mouse system. We reported previously that the efficacy of vaccination against seasonal influenza using the trivalent split subunit vaccine Begripal that is licensed for humans was impaired in mice with underlying L. sigmodontis infection (21). Here, we aimed at the restoration of vaccination efficacy by drug-induced deworming. To this end, mice were naturally infected with L. sigmodontis by exposure to infected mites. 4 weeks later, when immature adults had established an infection in the thoracic cavity, mice were treated by s.c. injection of FBZ on 5 consecutive days ( Figure 1A ). In our first approach, the anti-influenza vaccination was applied simultaneously with the deworming regimen and neutralizing Ab titres were quantified 3 weeks after vaccination by haemagglutination inhibition (HI) assay. L. sigmodontis-infected mice responded to the vaccination with drastically reduced HI titres compared to naïve control mice (dark blue symbols to black symbols, Figure 1B ) as we had shown before (21). FBZ-treatment did not restore the Ab response in these helminth-infected mice (light blue symbols to grey symbols, Figure 1B ). All mice were subsequently challenged by infection with the human pathogenic 2009 pH1N1 Influenza A strain that replicates in mice without further adaptation (22–24). Viral load in the lung was quantified three days post influenza virus infection ( Figure 1C ). Vaccinated control mice displayed reduced viral load compared to both, non-vaccinated mice and vaccinated helminth-infected mice, as expected (21). The FBZ treatment of vaccinated helminth-infected mice did not improve protection from the influenza virus challenge infection. Vaccinated FBZ-treated and helminth-infected mice displayed the same elevated viral load as the vaccinated non-treated helminth-infected mice (light blue to dark blue symbols, Figure 1C ). As additional control, we show that FBZ-treatment of the non-helminth-infected control group did neither modulate Ab response nor 2009 pH1N1 viral load in the lung (black symbols to grey symbols p = 0.43, Figure 1B and p = 0.27 Figure 1C ).

As the FBZ-treatment had no effect on responsiveness to vaccination we next controlled the efficacy of FBZ-induced deworming. To quantify the worm burden at the moment of vaccination, we recorded the number of worms in the thoracic cavity of helminth-infected FBZ-treated and untreated mice after the 3^rd FBZ application ( Supplementary Figure S1A ). Additionally, we recorded the worm burden 16 days after the final FBZ application to analyse the endpoint of the applied deworming regimen ( Supplementary Figure S1B ). While the worm burden was not statistically significantly reduced (p=0.07) after the third FBZ injection ( Supplementary Figure S1A ), the FBZ treatment resulted in complete clearance of viable worms and a significant reduction of killed and coated worms two weeks later ( Supplementary Figure S1B ). It should be noted that we had to shift the onset of FBZ-treatment in our second approach from day 28 to day 14 post L. sigmodontis infection, to allow analysis of the worm burden in FBZ-treated and untreated mice at days 30-35. An analysis at later time points would not report FBZ-mediated clearance correctly, since C57BL/6 mice start to naturally clear the L. sigmodontis infection via granuloma formation after day 35 (10, 13, 14).

In summary these results show that FBZ-treatment successfully eliminated L. sigmodontis parasites in the thoracic cavity of infected mice 2 weeks after the onset of therapy. Application of the vaccination against influenza to helminth-infected mice during the FBZ-treatment, when viable worms were still present, failed to induce a neutralizing Ab response. Subsequent protection from a 2009 pH1N1 influenza A virus challenge infection was not established.

Consequently, we next applied the vaccination against influenza two weeks after the final FBZ treatment, a time point when viable worms are eradicated due to drug treatment ( Supplementary Figure S1B and Figure 2 ). The HI titre of helminth-infected and FBZ-treated mice recorded 3 weeks after the vaccination was still significantly lower than in the naive FBZ-treated control group ( Figure 2B : p = 0.047). However, in direct comparison the helminth-infected and non-treated mice displayed a more drastic reduction in the HI titre (p < 0.0001) compared to non-helminth-infected control mice. Nevertheless, FBZ-treated helminth-infected mice were not protected against challenge infection with 2009 pH1N1 influenza A virus ( Figure 2C ). All mice that were vaccinated after FBZ-mediated clearance of the helminth infection displayed high viral loads in the lung, while the vaccinated control mice displayed sterile immunity in 7 out of 11 mice (non-helminth-infected control) and 6 out of 10 mice (non-helminth-infected and FBZ-treated control, Figure 2C ). Again, FBZ-treatment of vaccinated, non-helminth-infected mice did not modulate HI response or viral load compared to non-treated, vaccinated non-helminth-infected mice ( Figure 2B , black symbols to grey symbols p = 0.48; Figure 2C : p = 0.57).

We have shown previously that L. sigmodontis infection induced a systemic and sustained expansion of type 1 regulatory T cells (Tr1) that was correlated to the observed interference with vaccination efficacy (21). To understand why the vaccination against influenza failed in the FBZ-treated previously helminth-infected mice despite the successful deworming at the time point of vaccination, we next analysed the composition of regulatory T cell populations. To this end, mice were naturally infected with L. sigmodontis and FBZ-treated as described in Figure 2 . Instead of vaccinating the mice two weeks after the final FBZ treatment they were sacrificed for analysis ( Figure 3A ). Thereby, we analysed the thoracic cavity that is the site of worm infection as an indicator of local changes ( Figures 3B, D ) and the spleen as an indicator of systemic changes ( Figures 3C, E ) in the regulatory T cell populations. Foxp3⁺ regulatory T cells (Treg) were identified as CD4⁺Foxp3⁺ cells and Tr1 cells were identified as CD4⁺Foxp3^-LAG-3⁺CD49b⁺ cells ( Supplementary Figure S2 ). Helminth-infected mice displayed an elevation in Tr1 cells in the thoracic cavity ( Figure 3B ) and in the spleen ( Figure 3C ), while Treg expanded only locally in the thoracic cavity ( Figures 3D, E ), as we have shown before (21). FBZ treatment of helminth-infected mice abrogated the Treg expansion in the thoracic cavity two weeks after treatment ( Figure 3D ). A significant expansion of Tr1 cells, however, was still observed in the thoracic cavity and in the spleen ( Figures 3B, C ). As additional control we show that FBZ-treatment of naïve control mice did not modulate Tr1 and Treg populations (white to grey symbols Figure 3B : p = 0.55; Figure 3C : p = 0.36; Figure 3D : p = 0.58; Figure 3E : p = 0.48)

In summary these experiments show that vaccination after efficient deworming ( Supplementary Figure S1B ) did not induce a statistically significant elevation of neutralizing anti-HA Ab ( Figure 2B ), did not revert the helminth infection-induced Tr1 cell expansion ( Figures 3B, C ) and did not restore vaccination-induced protection against influenza ( Figure 2C ).

From the translational point of view the ultimate goal of vaccination is the establishment of sterile immunity to influenza virus infection despite pre-existing helminth infections. In a previous study, we observed that neither introduction of a prime-boost immunization nor application of the MF59-adjuvanted anti-influenza vaccine Fluad restored vaccination efficacy in helminth-infected mice (unpublished observation). As also drug-induced deworming alone did not restore vaccination efficacy, we next asked if a combination of both approaches would outcompete helminth-induced immunomodulation. To this end, mice were naturally infected with L. sigmodontis, FBZ-treated and vaccinated 2 weeks after the final treatment. Instead of a single vaccination that is usually applied for influenza vaccination of humans, mice received 2 consecutive injections ( Figure 4A ). This prime-boost vaccination regimen with the non-adjuvanted influenza vaccine Begripal did not elevate the Ab response to vaccination, nor restore the protection in helminth-infected but untreated mice ( Figures 4B, C ) as we observed before (unpublished data). FBZ-treatment of helminth-infected mice led to a partial restoration of the prime-boost vaccination efficacy. Although the HI titre was still significantly lower in FBZ-treated and helminth-infected mice compared to the FBZ-treated non-helminth infected control group, sterile protection against challenge infection with 2009 pH1N1 was established in 57% of the mice ( Figure 4C ). Likewise, a prime-boost vaccination with the MF59-adjuvated influenza vaccine Fluad (25) did not restore the Ab titre to normal levels ( Figure 4D ) nor establish sterile protection from challenge infection with 2009 pH1N1 influenza A virus in helminth-infected mice ( Figure 4E ). However, FBZ-treatment 2 weeks before the prime-boost vaccination with Fluad resulted in sterile immunity in 100% of the formerly helminth-infected mice ( Figure 4E ). Again FBZ-treatment as such did not modulate the vaccination-response of non-helminth infected mice ( Figures 4B, C , E : p > 0.99, Figure 4D : p = 0.078).