key: cord-1044181-35vt4nbk
authors: Spacova, Irina; De Boeck, Ilke; Cauwenberghs, Eline; Delanghe, Lize; Bron, Peter A.; Henkens, Tim; Simons, Alix; Gamgami, Imane; Persoons, Leentje; Claes, Ingmar; van den Broek, Marianne F. L.; Schols, Dominique; Delputte, Peter; Coenen, Samuel; Verhoeven, Veronique; Lebeer, Sarah
title: Live biotherapeutic throat spray for respiratory virus inhibition and interferon pathway induction
date: 2022-01-25
journal: bioRxiv
DOI: 10.1101/2022.01.25.477549
sha: 7c702423872d491c4cc3a3023f193908834c2d30
doc_id: 1044181
cord_uid: 35vt4nbk

Respiratory viruses such as influenza viruses, respiratory syncytial virus (RSV), and coronaviruses initiate infection at the mucosal surfaces of the upper respiratory tract (URT), where the resident respiratory microbiome has an important gatekeeper function. In contrast to gut-targeting administration of beneficial bacteria against respiratory viral disease, topical URT administration of probiotics is currently underexplored, especially for the prevention and/or treatment of viral infections. Here, we report the selection and formulation of a broad-acting throat spray with live lactobacilli which induce interferon regulatory pathways and are able to inhibit respiratory viruses. Rational selection of Lactobacillaceae strains was based on safety, applicability, and potential antiviral and immunostimulatory efficacy in the URT. Three strains, Lacticaseibacillus casei AMBR2, Lacticaseibacillus rhamnosus GG and Lactiplantibacillus plantarum WCFS1 significantly reduced the cytopathogenic effects of RSV, influenza A/H1N1 and B viruses, and HCoV-229E coronavirus in co-culture models with bacteria, virus and host cells. Subsequently, these strains were formulated in a throat spray and human monocytes were employed to confirm the formulation process did not reduce the interferon regulatory pathway-inducing capacity. Administration of the throat spray in healthy volunteers revealed that the lactobacilli were capable of temporary colonization of the throat in a metabolically active form.

the interferon regulatory pathway-inducing capacity. Administration of the throat spray in healthy 27 volunteers revealed that the lactobacilli were capable of temporary colonization of the throat in a 28 metabolically active form. 29 severe inflammation [2,3] and/or subsequent pneumonia [4, 5] . Severe disease caused by viruses is 38 often accompanied by reduced type I and III interferon (IFN) production and/or overproduction of pro-39 inflammatory mediators [3,6,7]. Type I and III IFNs play an important role for innate immunity at 40 mucosal barrier surfaces, such as the respiratory epithelial barrier, where they provide first line 41 antiviral defense mechanisms [8] . 42

The respiratory mucosal surfaces targeted by viruses also harbor a microbiome consisting of resident 43 microorganisms which have important multifactorial gatekeeper functions, including direct inhibition 44 of incoming pathogens, maintaining epithelial barrier function, and immune homeostasis [9] . 45 Importantly, the stability of the core URT microbiome can be compromised during viral RTIs, for 46 instance as observed with influenza [10] . This can facilitate pathobiont overgrowth, as a result of virus-47 induced damage and immune dysfunction [11, 12] . Direct supplementation of beneficial bacteria from 48 the URT microbiome could act on different stages of viral disease, for instance via enhancement of the 49 airway epithelial barrier, direct inhibition of pathogens or stimulation of the immune system, as we 50 have recently reviewed [13] . Stimulation of antiviral immunity is one of the key mechanisms reported 51 lactobacilli was significantly lower compared to the non-probiotic LPS-containing Gram-negative 104 control Escherichia coli DH5α ( Figure 1C , E). To assess whether the bacteria had to be metabolically 105 active to induce immunostimulatory effects, ultraviolet (UV)-inactivated strains were used ( Figure 1E -106 F). Both UV-inactivated L. plantarum WCFS1 and L. rhamnosus GG, as well as the mix of the three 107 strains, still significantly induced the IRF and NF-κB pathways, although to a lesser extent than their 108 viable counterparts ( Figure 1E [27]. HCoV-229E generally causes mild symptoms, but it has several features in common with SARS-117

CoV-2, such as homologous epitopes of the spike protein, and is suitable for high throughput screening 118 due to biosafety level 2 [28]. To avoid bacterial metabolites interfering with the assay read-outs, the 119 lactobacilli were UV-inactivated for the assays, retaining the capacity of bacterial cell surface molecules 120 to sequester or block viral particles [29] . First, we established the concentration of bacteria that had 121 to be simultaneously added with the HCoV-229E virus to inhibit the virus-induced cytopathic effect in 122 susceptible human Huh7 cells by 50% (antiviral EC50) (Figure 2A,B) . L. casei AMBR2 and L. rhamnosus 123 GG showed the strongest effect, with respective concentrations of 2.57x10 6 colony-forming units 124 (CFU)/ml and 5.06x10 6 CFU/ml required for 50% viral inhibition, which was lower than the required 125 concentration of 1.26x10 7 CFU/mL for the non-probiotic E. coli DH5α strain. For L. plantarum WCFS1, 126 the required concentration for 50% viral inhibition was 1.17x10 7 CFU/ml. 127

Subsequently, we also evaluated a set-up where the three bacterial strains or their combination in 128 equal ratios were pre-incubated with HCoV-229E, RSV or influenza viruses, allowing trapping and/or 129 inactivation of virus particles, and afterwards unbound virus particles were added to human cells 130 ( Figure 2C ). HCoV-229E ( Figure 2D ) and RSV ( Figure 2E ) pre-incubated with L. casei AMBR2, L. 131 rhamnosus GG, L. plantarum WCFS1 or their combination significantly lost the capacity to reduce 132 human cell viability compared to virus pre-incubation with phosphate-buffered saline (PBS) which 133 served as an inactive control. For example, the percentage of viable Huh7 cells infected with HCoV-134 229E was on average 2.9% for the PBS condition, and 34%, 30%, and 18% for L. casei AMBR2, L. 135 rhamnosus GG, and L. plantarum WCFS1 conditions, respectively. Using the combination of the three 136 strains also resulted in 18% of viable Huh7 cells, similar to L. plantarum WCFS1 only. Pre-incubation 137 with L. rhamnosus GG or a combination of the three lactobacilli also significantly reduced the 138 cytopathic effects induced by the common influenza A/H1N1 virus compared to PBS ( Figure 2F ). A 139 reduction of cytopathic effects of influenza B was also observed after pre-incubation of L. rhamnosus 140 GG ( Figure 2H ), while the effects of the tested lactobacilli on the cytopathic effects of influenza A/H3N2 141 virus were less pronounced ( Figure 2G ). The pre-incubation of the non-probiotic control E. coli DH5α 142 with the tested viruses did not improve viability of human cells, while the cell viability in all lactobacilli 143 conditions was significantly higher compared to E. coli DH5α. The potential antiviral activity of L. casei 144 AMBR2, L. rhamnosus GG, and L. plantarum WCFS1, in addition to their other beneficial properties 145 (Table 1) oil was chosen to increase bacterial retention in the throat and to ensure a sufficient dosage of viable 152 probiotic CFU counts. First, the viability of each bacterial strain in freeze-dried powder form was 153 evaluated at 4°C and 25°C ( Figure 3A ). For all three strains, the viability at 4°C remained stable over 154 time. For L. casei AMBR2 and L. plantarum WCFS1, no log reductions were observed, with 2.79 x 10 10 155 CFU/g powder and 3.56 x 10 11 at start for AMBR2 and WCFS1, respectively, to 2.21 x 10 10 CFU/g powder 156 and 3.15 x 10 11 CFU/g powder after 26 weeks. For L. rhamnosus GG, the viability decreased from 1.18 157

x 10 11 CFU/g powder to 7.15 x 10 10 CFU/g. At room temperature (25°C), L. plantarum WCFS1 seemed 158 most stable, with 1 log reduction at 26 week (5.84 x 10 10 ), while this was 8.57 x 10 9 for L. rhamnosus 159 GG and 2.02 x 10 7 for L. casei AMBR2. 160

We next evaluated different concentrations of the strains in the mixture. L. casei AMBR2 at 50%, L. 161 plantarum WCFS1 at 33.3%, and L. rhamnosus GG at 16.7% were found the optimal ratio upon long-162 term storage at room temperature, which reflected the intended storage conditions. Next, the viability 163 of the combined bacterial strains in powder form ( Figure 3B ) and in the throat spray formulation 164 ( Figure 3C ) was evaluated at 4°C, 15°C and 25°C. For the mixed powders ( Figure 4B ), viability decreased 165 slightly from 2.09 × 10 11 CFU/g at the start, to 1.11 x 10 11 , 4.51 x 10 10 and 8.01 x 10 9 CFU/g at 26 weeks 166 of storage at 4°C, 15°C and 25°C, respectively. For the spray formulation ( Figure 3C ), viability starting 167 with 3.78 x 10 9 CFU/g spray remained stable at 4°C and 15°C at 26 weeks. At 25°C, a 2 log reduction 168 was observed (3.3 x 10 7 CFU/g) at 26 weeks. 169

We subsequently confirmed the retention of immunostimulatory activity in human monocytes of the 170 strains and their combination in powder form ( Figure 3D -E), and in the spray formulation in oil ( Figure  171 3F-G). All single strains in powder form and their combination were still capable of significant IRF and 172 NF-κB induction ( Figure 4D -E) at a dose of 10 8 CFU/ml, which corresponds to the L. casei AMBR2 173 concentration per puff previously tested in healthy volunteers [19] . The immunostimulatory action of 174 the throat spray formulation with the three strains in an oil suspension was also compared to a placebo 175 oil formulation without lactobacilli. The throat spray formulation with L. casei AMBR2 at 50%, L. 176 plantarum WCFS1 at 33.3%, and L. rhamnosus GG at 16.7% in oil significantly induced IRF and NF-κB in 177 human monocytes ( Figure 4F -G). While the placebo formulation also induced NF-κB, albeit to a lower 178 degree compared to the spray formulation with lactobacilli, it did not significantly affect IRF. 179

180 Finally, we evaluated the retention of the lactobacilli in the formulated spray in the throat of 12 healthy 181 volunteers via cultivation, quantitative polymerase chain reaction (qPCR) and 16S rRNA amplicon 182 sequencing using a longitudinal placebo-controlled sampling set-up ( Figure 4A ). The volunteers used 183 the verum spray at the start of the study by spraying two puffs containing approximately 9.5x10 8 CFU 184 of lactobacilli, or the placebo spray not containing lactobacilli, and throat swabs were collected at 185 baseline, after 30 min, and after 2 hours ( Figure 4A) . plantarum) were detected in 5 out of 6 participants. In the placebo group, these Lactobacillus ASV 196

were not detected, except for one participant that had low endogenous relative abundances of the 197

Lactobacillus casei ASV after 30 minutes. After 2 hours, 5 out of 6 participants in the verum group still 198 had detectable Lactobacillus ASVs. 199

To confirm and quantify the high abundances of the administered strains observed by sequencing DNA 200 derived from samples in the verum group after bacterial administration, we aimed to estimate the 201 CFU/ml counts based on targeted qPCR ( Figure 4D ). In line with the sequencing data, after 30 minutes, 202 the estimated CFU counts for L. rhamnosus GG in the verum group were between 1.26 x 10 4 -9.24 x 10 5 203 CFU/ml. For L. casei AMBR2, estimated CFU/ml counts ranged from 1.72 x 10 5 to 1.8 x 10 7 CFU/ml and 204 for L. plantarum WCFS1 from 4.63 x 10 4 CFU -3.36 x 10 6 CFU. After 2 hours, the amount of detected 205 lactobacilli decreased. L. rhamnosus GG and L. plantarum WCFS1 were not detected anymore except 206 in one participant. L. casei AMBR2 on the other hand, which was administered in the highest ratio of 207 50% in the spray, was still detected in 5 of the 6 participants, with a median CFU/ml count of 4x10 3 208 CFU/ml. 209

In addition to analyzing the DNA of the bacteria, we also cultivated throat swabs to evaluate whether 210 the administered lactobacilli were still viable. Cultured throat swabs from the verum group 211 demonstrated colony morphologies typical for the three administered Lactobacilaceae strains, and the 212 identity was confirmed via colony PCR and sequencing of the 16S rRNA gene, confirming that the 213 administered strains are not only detected in the throat via their DNA, but also remain viable. 214 rhamnosus GG were designed (Table 2) 

SARS-CoV-2 Drives Development of COVID-19

Switch from protective to adverse inflammation 431 during influenza: viral determinants and hemostasis are caught as culprits

Clinical course and risk factors for mortality of adult 434 inpatients with COVID-19 in Wuhan, China: a retrospective cohort study

Viral pneumonia

Imbalanced Host 440 Response to SARS-CoV-2 Drives Development of COVID-19

Impaired type I interferon activity 443 and inflammatory responses in severe COVID-19 patients

American 444 Association for the Advancement of Science

Importance of Type I and III Interferons at Respiratory and 447 Intestinal Barrier Surfaces

The microbiota of the respiratory tract: 450 gatekeeper to respiratory health

Microbiome disturbance and 453 resilience dynamics of the upper respiratory tract during influenza A virus infection

Dynamics of Bacterial Colonization With 456

Haemophilus influenzae, and Moraxella catarrhalis During Symptomatic 457 and Asymptomatic Viral Upper Respiratory Tract Infection

Secondary Bacterial Infections Associated with Influenza Pandemics

Topical Microbial Therapeutics against

Respiratory Viral Infections

Lactobacillus rhamnosus GG protects mice from H1N1 influenza virus infection by regulating 466 respiratory immune responses

Lung transcriptional 469 unresponsiveness and loss of early influenza virus control in infected neonates is prevented by 470 intranasal Lactobacillus rhamnosus GG

Microbes, metabolites, and the gut-lung axis

Scale Phylogenomics 476 of the Lactobacillus casei Group Highlights Taxonomic Inconsistencies and Reveals Novel Clade-477

Control Microbiome Study of Chronic Otitis Media with Effusion in Children Points at Streptococcus 480 salivarius as a Pathobiont-Inhibiting Species. mSystems

Have a Niche in the Human Nose

Lacticaseibacillus casei AMBR2 modulates the epithelial barrier function and immune response in a 486 donor-derived nasal microbiota manner

Lacticaseibacillus 489 casei AMBR2 Restores Airway Epithelial Integrity in Chronic Rhinosinusitis With Nasal Polyps

Intranasal Administration of Lactobacillus paracasei NCC2461 Modulates Allergic Airway Inflammation 554 in Mice. Cong Y, editor

Intranasal 557 administration of probiotic Lactobacillus rhamnosus GG prevents birch pollen-induced allergic asthma 558 in a murine model

Protection of chickens against H9N2 561 avian influenza virus challenge with recombinant Lactobacillus plantarum expressing conserved 562 antigens

Lactobacillus species facilitates protection against influenza virus infection in mice

Antiviral Potential of Selected

Enterococcus faecium on the Propagation of Swine Influenza A Virus In Vitro

Lactobacillus plantarum WCFS1 and 575 its host interaction: a dozen years after the genome

Genome sequence of

Lactobacillus pentosus KCA1: vaginal isolate from a healthy premenopausal woman

Topical cream containing live lactobacilli decreases malodor-581 producing bacteria and downregulates genes encoding PLP-dependent enzymes on the axillary skin 582 microbiome of healthy adult Nigerians

Oral use of

Lactobacillus rhamnosus GR-1 and L. fermentum RC-14 significantly alters vaginal flora: randomized, 585 placebo-controlled trial in 64 healthy women

Use of a fermented dairy 588 probiotic drink containing Lactobacillus casei (DN-114 001) to decrease the rate of illness in kids: the 589 DRINK study. A patient-oriented, double-blind, cluster-randomized

Consumption of a fermented dairy product 592 containing the probiotic Lactobacillus casei DN-114001 reduces the duration of respiratory infections 593 in the elderly in a randomised controlled trial

Augmentation of cellular immunity and reduction of influenza 595 virus titer in aged mice fed Lactobacillus casei strain Shirota

Lactobacillus bulgaricus, Lactobacillus rhamnosus and 599

Lactobacillus paracasei Attenuate Salmonella Enteritidis, Salmonella Heidelberg and Salmonella 600

Typhimurium Colonization and Virulence Gene Expression In Vitro

Application of pulsed electric field in production of 602 ice cream enriched with probiotic bacteria (L. rhamnosus B 442) containing intracellular calcium ions

Persistence of 606 probiotic strains in the gastrointestinal tract when administered as capsules, yoghurt, or cheese

Probiotic attributes of Lactobacillus strains isolated from food 609 and of human origin

Identification and 611 classification of Lactobacillus acidophilus, L. gasseri and L. johnsonii strains by SDS-PAGE and rRNA-612 targeted oligonucleotide probe hybridization

Laboratories BR. BRL pUC host: E. coli DH5α competent cells

WCFS1 and their combination induce (C) nuclear factor (NF)-κB 624 and (D) interferon regulatory factors (IRFs) in human THP-1 Dual monocytes upon co-incubation

WCFS1 and their combination with L. casei AMBR2 also induce (E) NF-κB and 626 (F) IRFs in human THP-1 Dual monocytes. The medium condition represents the cells as such and serves as a 627 baseline, while Poly(I:C) at 50 µg/ml with Lipofectamine (Poly(I:C) 50) serves as control IRF

Data is depicted as mean±SD per 629 condition. *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001 as determined by a One-way ANOVA test followed 630 by Dunnett's multiple comparisons test compared to the medium condition. decrease cytopathogenic effects of HCoV-229E, RSV and influenza A/H1N1, A/H3N2 and B viruses in vitro

A/H3N2 (G) or influenza B (FH could 637 inhibit the cytopathic effects induced in human cells. EC50: 50% Effective concentration producing 50% inhibition 638 of virus-induced cytopathic effects, as determined by measuring the cell viability with the colorimetric MTS cell 639 viability assay. CFU: colony-forming units. PBS serves as control that does not affect viral infection, E. coli DH5α 640 represents a non-probiotic laboratory strain. Data depicted as mean±SD per condition

0001 as determined by a One-way ANOVA test followed by Dunnett's multiple 642 comparisons test compared to the medium or PBS conditions. baseline (T0), and 30 660 minutes (T1) and 2 hours (T2) after the throat spray was used. The presence of L. casei AMBR2, L. rhamnosus GG, 661 and L. plantarum WCFS1 was evaluated via 16S rRNA amplicon sequencing (relative abundances) in panel C. At 662 30 minutes and 2 hours, qPCR with species-specific primers was used to estimate the CFU/ml counts in the verum 663 group in panel D. Based on the standard curve, the detection limit was estimated to be at 10 3 CFU/ml. disease. (B) Documented mechanisms from this and previous studies through which select beneficial 667 lactobacilli formulated as an URT spray can promote the antiviral activity of the URT microbiome. The modes 668 of action

IFN: interferon

AECs: airway epithelial cells; APC: antigen-presenting cell; Teff: T effector cell