RNA viruses alter house dust mite physiology and allergen production with no detected consequences for allergenicity

RNA viruses have recently been detected in association with house dust mites, including laboratory cultures, dust samples, and mite‐derived pharmaceuticals used for allergy diagnosis. This study aimed to assess the incidence of viral infection on Dermatophagoides pteronyssinus physiology and on the allergenic performance of extracts derived from its culture. Transcriptional changes between genetically identical control and virus‐infected mite colonies were analysed by RNAseq with the support of a new D. pteronyssinus high‐quality annotated genome (56.8 Mb, 108 scaffolds, N50 = 2.73 Mb, 96.7% BUSCO‐completeness). Extracts of cultures and bodies from both colonies were compared by inspecting major allergen accumulation by enzyme‐linked immunosorbent assay (ELISA), allergen‐related enzymatic activities by specific assays, airway inflammation in a mouse model of allergic asthma, and binding to allergic patient's sera IgE by ImmunoCAP. Viral infection induced a significant transcriptional response, including several immunity and stress‐response genes, and affected the expression of seven allergens, putative isoallergens and allergen orthologs. Major allergens were unaffected except for Der p 23 that was upregulated, increasing ELISA titers up to 29% in infected‐mite extracts. By contrast, serine protease allergens Der p 3, 6 and 9 were downregulated, being trypsin and chymotrypsin enzymatic activities reduced up to 21% in extracts. None of the parameters analysed in our mouse model, nor binding to human IgE were significantly different when comparing control and infected‐mite extracts. Despite the described physiological impact of viral infection on the mites, no significant consequences for the allergenicity of derived extracts or their practical use in allergy diagnosis have been detected.


INTRODUCTION
House dust mites (HDM) of the family Pyroglyphidae (Acari: Astigmata) are the principal elicitors of indoors allergy globally (Sánchez-Borges et al., 2017). HDM-allergy is diagnosed and treated by allergen immunotherapy (AIT) using pharmaceutical products derived from extracts of industrially reared HDM. Mite culture conditions are key determinants of the final potency and consistency of these products, which depend on a complex repertoire of allergens produced by HDM (www.allergen.org) . The physiological status of the mites can thus have an impact over standardization, a primary concern of the industry (Casset et al., 2012;Zimmer et al., 2021).
Viruses infect almost every living species on Earth, establishing relationships with the host that vary from pathogenic to mutualistic, and being essential for evolution (Harris & Hill, 2021). Recent studies have detected the presence of RNA viruses in different allergy-eliciting HDM, including one of the most medically important species, Dermatophagoides pteronyssinus (Trouessart) (Guo et al., 2021;Vidal-Quist et al., 2021) (Dp).
Most of these viruses show relationship with genotypes infecting invertebrates, not mammals, thus their pathogenicity to humans is unlikely. Viral RNA has been detected in house dust samples and HDM laboratory cultures, and proof was provided that viruses can also be found (viable or not) in commercial Dp extracts used for allergy diagnosis and in reference standards from the United States Food and Drug Administration (FDA) .
The possible incidence of these viruses on both mites and atopic human individuals, after exposure to dust particles or commercial extracts from infected mites, is mostly unknown. From the HDM perspective, the identified RNA viruses seem to establish persistent chronic infections that do not induce pathogenic effects compromising survival or development ); yet it remains to be explored to what extent they can alter the mite's physiology.
From the human's viewpoint, such viruses may have the potential to (positively or negatively) influence the sensitization process and/or the development of the allergic reaction, either by the direct effect of viral components (Iwasaki, 2012), or by altering the HDM allergenic profile or immunogenicity, as it has been suggested for other allergenic mites harbouring different microbiomes (Hubert et al., 2019).
Arthropods have evolved a myriad of mechanisms to counter pathogen infections, from behavioural strategies or physical barriers to sophisticated cellular and humoral innate immunity responses (Evans et al., 2006;Liu et al., 2017;Möllmann & Colgan, 2022), yet our knowledge of mite's immune system remains scarce. Comparative genomics has shown that components of the canonic immune system in Drosophila melanogaster (Meigen) are not always conserved in mites (Palmer & Jiggins, 2015). Of note, viral infection in Dp appears to be located in digestive epithelia , which is precisely the site where many of the allergens are expressed, including the three major allergens Der p 1, 2, and 23 (Herman et al., 2014;Park et al., 2000;Weghofer et al., 2013). It is therefore possible that viral infection might affect the production of allergens or other immunoactive factors such as digestive proteases, among others (Jacquet, 2021; Reithofer & Jahn-Schmid, 2017).
The present work takes advantage of two genetically identical Dp colonies generated in a previous work from the same inbreeding line , one virtually free of viruses (as per RT-PCR) and one artificially infected with five different RNA viruses, to study the impact of viral infection over the mite's physiology and allergenicity. Specifically, using a new Dp genome assembly and annotation, we inspect mite transcriptional changes that result from viral infection, with emphasis on viral response and allergen-related genes. We also compare major allergen accumulation and allergen-related enzymatic activities in whole cultures and purified mite bodies from both colonies.
Finally, we investigate whether viral infection alters the allergenicity of mite extracts by assessing airway inflammation in a mouse model of HDM-induced asthma, as well as by detecting IgE levels on sera from Dp-sensitized subjects in ImmunoCAP IgE-binding assays.

Mite cultures
Control and virus-infected colonies were established in our laboratory in a previous investigation. RNA viruses (DerpV1, 3, 4, 5 and 6 as detected by RT-PCR; NCBI Bioproject PRJNA684040) were transmitted to cultures of the inbreeding line J-1-1-1-1 (Vidal-Quist et al., 2019) (control colony) by inoculating virus-containing faecal pellets from an in-house stock colony .
Both colonies have been maintained in laboratory standard growth conditions as described elsewhere (Vidal-Quist et al., 2015). Prior to the experiments, the expected viral load in both colonies was confirmed by RT-PCR following published methods  ( Figure S1). The head-to-head comparison of control and virus-infected colonies throughout different stages of culture maturation was achieved by seeding test cultures at a 1:14 ratio with cultures of both colonies synchronized at the late exponential growth phase (140 mg of a 39 days-old seed culture onto 2 g of fresh growth medium, using cotton-plugged 50 ml Erlenmeyer containers). Replicated cultures were inspected periodically to confirm synchronicity between and within colonies. Whole mite culture (WMC) samples (40-60 mg; exact weights recorded) were pulled after thoroughly mixing the culture and stored at À20C until extraction. Mixed-stages mite bodies were separated from WMC at the exponential growth phase by the paper disk method (Vidal-Quist et al., 2015), and sampled using a sterile needle for RNA or aqueous extraction. All the activities were compliant with the rules of the Bioethics and Biosecurity Committee of Consejo Superior de Investigaciones Científicas (CSIC, Spain).

Genome assembly and annotation
A genome assembly was obtained by Oxford Nanopore Technologies (ONT) sequencing, using high molecular weight (HMW) DNA isolated from eggs of the inbred colony J-1-1-1-1 (obtained in a previous work (Vidal-Quist et al., 2019)) as template. Details on methods for bioinformatic assembly and gene annotation and curation are described in text S1 (section A), together with NCBI accession numbers for raw sequencing data. Briefly, ONT reads (1,145,352 reads, N50 = 27,051 nt, $360x coverage) were assembled followed by several consecutive rounds of polishing using ONT, Illumina gDNA, and RNA sequences (in that order). Then, haplotigs and repeated contigs were identified and purged to obtain the final assembly (NCBI database accession JAMYKS000000000; Bioproject PRJNA843460). Protein coding genes were first automatically predicted, followed by manual review and curation of gene names and/or models after visualizing mapped RNA sequences using GenomeView (https:// genomeview.org). Finally, the completeness of genome assembly was assessed using the annotated proteome and the lineage dataset arthropoda_odb10 or arachnida_odb10 as reference.

RNA extraction, RNAseq and RT-qPCR
Total RNA was extracted from purified mixed-stages mite body samples with the TRIzol Reagent (Life Technologies, Carlsbad, USA) following manufacturer instructions, adding glycogen as carrier and diluting in ultrapure RNAse-free water. RNAseq procedure together with in silico analysis, including differential expression, GO enrichment and KEGG pathway analyses, as well as RT-qPCR methodological details are described in text S1 (section B). Table S1 indicates the set of primer pairs used for RT-qPCR.

ELISA and enzymatic assays
Extracts from mite bodies or whole mite cultures (WMC) were obtained by homogenization in ice-cold 0.15 M NaCl followed by estimation of soluble protein using the Lowry method with the DC Protein Assay kit (Bio-Rad, Hercules, CA, USA) and BSA as standard. Der p 1 and Der p 23 allergens were quantified by Sandwich ELISA using the Der p 1 ELISA kit (EL-DP1A) and the Der p 23 ELISA kit 2.0 (EPC-DP23-1), respectively (Indoor Biotechnologies, Charlottesville, USA).
Der p 2 allergen was quantified using the ALK's in-house Der 2 ELISA kit (Barber et al., 2012). It is to be noted that preliminary assays showed that Indoor Biotechnologies Mite Group 2 ELISA kit (EL-D2) was not compatible with the Der p 2 variant produced by the inbred colonies used in this study (variant Der p 2.0101 in the J-1-1-1-1 colony; Vidal-Quist et al., 2019). Six enzymatic activities were estimated on mite extracts following previously described methods . Briefly, activities related to allergens Der p 1 (cysteine protease), Der p 3 (trypsin), Der p 6 (chymotrypsin), and Cathepsin D (body-associated aspartic protease used as marker of exponential growth) were analysed using fluorometric assays; whereas Der p 4 (alpha-amylase) and Der p 8 (glutathione S-transferase, GST) were estimated by colorimetric assays. Reaction conditions, including substrates, additives (such as protease inhibitors or activators), and buffers are summarized in Table S2; detailed protocols are available in the cited reference.

Murine asthma model with mite-derived extracts
Mice of the strain C57BL/6 were exposed to Dp extracts to induce experimental allergic asthma, as previously described (Debeuf et al., 2016). In brief, mice were sensitized intratracheally on day 1 with either 1 μg Dp extract or with PBS as control, followed by

IgE-binding assay
Binding of IgE from sera of Dp-sensitized patients was quantified by Fluoroenzyme Immunoassay in the Phadia ImmunoCAP system using the four types of mite extract described above as allergen source. Informed consent was obtained from all patients to conduct the analysis, which was approved by the Institutional Ethics during 1 h at room temperature and IgE-binding assays were performed in a Phadia 250 ImmunoCAP system using undiluted patient's sera.

Data analysis
Specific enzymatic activity, allergen content, RT-qPCR gene transcription data (Cq' values, obtained by log2 transformation of normalized relative quantities, as described in Vidal-Quist et al., 2015), and human serum IgE concentration were compared using unpaired t-test or 2-way ANOVA coupled with post-tests, as indicated for each assay.
Prior to parametric tests, normality was confirmed by the Shapiro-

Genome assembly
A high-quality assembly was obtained by nanopore sequencing of HMW DNA from eggs of the previously obtained inbred colony J-1-1-1-1 (Vidal-Quist et al., 2019). Quality parameters were on par with the best assemblies obtained so far on allergy-eliciting Astigmata mites (Table S3). The assembled genome size was 56.8 Mb in 108 scaffolds, with N50 of 2.73 Mb, and the largest contig being 5.93 Mb.
More than 3600 genes were manually curated, paying special attention to genes encoding allergens and related proteins, peptidases, heat shock proteins and immunity/stress-related proteins; as well as   Table S6). GO enrichment analysis was conducted separately on both upregulated and downregulated gene sets to reveal general F I G U R E 1 Comparison of enzymatic activity and allergen profiles during different phases of culture growth of the control and virus-infected D. pteronyssinus colonies. Whole culture samples were extracted at days 28, 31, 36, 29 and 42, and analysed for specific enzymatic activities (panels a-d) and major allergen content by ELISA (panels e-g). Data are means and SE of six biological replicates. Estimates (dependent variable) were compared by two-way ANOVA (virus content and culture age as independent variables) considering repeated measures (for culture age factor), followed by Sidak's post-test for comparisons at each time point. 'ns' denotes 'not significant'; * p < 0.05; **** p < 0.0001.

Overall impact of viral infection on gene transcription
T A B L E 2 Differential expression of antiviral immunity-related genes upon viral infection    39 KO terms identified in the 'Biosynthesis of amino acids' pathway (map01230) were related to downregulated genes ( Figure S5).

Impact of virus infection on allergen expression
Seven recognized Dp allergen genes were differentially expressed as a result of viral infection (Table 1)

Regulation of immunity-related genes
RNAseq analysis detected up to 34 immune-response genes significantly regulated under viral infection. The identified genes were related to: RNAi pathway; recognition receptors; components of the Toll, JAK/STAT and Imd signalling pathways; Nup98 antiviral response; and, effector proteins (putative antimicrobial peptides, including defensin and ixodidin homologues) ( Table 2).

Regulation of other stress-related genes
A number genes involved in protection of intracellular proteins, autophagy, cell death and apoptosis, peritrophic membrane function and integrity, and gut epithelial barrier homeostasis were differentially expressed after viral infection. Results are summarized in Table 3; extended results for each group are provided in Text S2 (Section A). In addition, our transcriptomic survey detected other significantly regulated genes for which a mechanistic link to infection could be

Transcriptional changes in proteolysis-related genes
As indicated earlier, the GO biological process 'proteolysis' was significantly enriched among the group of downregulated genes in the infected colony ( Figure S3). Further screening of the genome assembly detected 317 peptidase-encoding genes (EC 3.4 hydrolases; Table S6), 34 of which were significantly regulated under viral infection (Table S8). Notably, none of the peptidase genes showing high overall (e), serum HDM-specific IgG1 levels (OD value). Data were pooled from two independent experiments (n = 6 for PBS/control WMC, 12 for control WMC/control WMC, 6 for PBS/virus-infected WMC, and 11 for virus-infected WMC/ virus-infected WMC). Mann-Whitney statistical test was performed between groups sensitized with WMC. Data are shown as means ± SD; 'ns' denotes 'not significant'. expression (54 genes above the 90th percentile of most expressed genes, as estimated by the Transcripts Per Million RNAseq metric, TPM) was upregulated in infected mites, whereas 11 were significantly downregulated: 2 metallocarboxypeptidases, 4 serine endopeptidases, and 5 cysteine endopeptidases. In the case of serine endopeptidases, the top 3 more expressed genes encode Der p 3, 9 and 6 allergens, which were all significantly downregulated in the infected colony ( Figure 2a). This result is consistent with the decrease of Der p 3 and Der p 6-like enzymatic activities indicated earlier (Figure 1). By contrast, for cysteine endopeptidase-encoding genes, the top 3 more expressed genes, including Der p 1, were not differentially expressed (Figure 2b), which is also consistent with the lack of significant differences for Der p 1 content and its enzymatic activity (Figure 1).

Impact of virus infection on the allergenicity of mite extracts: Murine asthma model
To assess the effect of virus infection on the allergenicity, we tested two types of mite-derived extracts (WMC or purified bodies) from both the control and infected colonies in a model of HDM-driven allergic asthma (Figure 3a, Figure S7A). Upon treatment with WMC or purified bodies, mice from the sensitized group exhibited clear Th2-driven airway inflammation compared to mock-sensitized mites, as defined by airway eosinophilia, presence of Th2-skewed CD4 T cells (ST2 + CD4+) and production of Th2 cytokines in mLN restimulation culture (Figure 3b-d, Figure S7B-D). In addition, increased levels of allergen-specific IgG1 immunoglobulins could be detected in the serum of allergen-sensitized compared to mock-sensitized mites ( Figure 3e, Figure S7E). However, we could not detect any significant differences in these inflammatory parameters comparing the groups that were exposed to extracts from control versus virus-infected colonies (p > 0.05). Additional readouts are available in Figure S8.

Impact of virus infection on the allergenicity of mite extracts: Binding of human IgE
Binding of IgE from sera of Dp-sensitized human subjects (Table S9)

Virus infection, general considerations
In a previous study, we showed that the association of RNA viruses with HDM is remarkably widespread, being found in laboratory or commercial F I G U R E 4 Binding of IgE from D. pteronyssinus sensitized patient's sera. IgE-binding was assessed using custom-made ImmunoCAPs from whole mite culture (WMC) or purified mite body (PMB) extracts from both the control (blue) and virus-infected (red) colonies. Estimates of IgE concentration in sera are shown for each extract source as box-and-whiskers plots (n = 10 patients).
Concentrations (dependent variable) were compared by two-way ANOVA (virus content and extract origin as independent variables) considering matched values per patient, followed by Sidak's post-test for comparisons between each extract type; 'ns' denotes 'not significant'. p values and statistics for each independent factor, and their interaction are shown in the upper box.

Mite-virus interaction
The overall transcriptional pattern reported here was in agreement with other observations on pathogenic viral infections in the digestive tract of arthropods, in which defence-related processes are commonly upregulated at the expense of reducing general functions related to primary metabolism, such as digestion (Chtarbanova et al., 2014;Li-Byarlay et al., 2020). However, in a previous study using the same colonies, we demonstrated that the fitness of the population was not compromised by viral infection , despite the repression of primary metabolism. This result could be related to the generally mild intensity of the observed transcriptional change, but also to the type of genes being regulated. As a matter of fact, regarding digestion, previous studies using protease inhibitors have pointed out that proteolytic gut digestion in Dp is probably ruled by Der p 1 and other cysteine proteases (which were not affected by infection in this study), whilst serine proteases (which were repressed in this study) appear to play a minor role on gut digestion (Erban et al., 2017;Vidal-Quist, Ortego, & Hernández-Crespo, 2021).
Despite the apparently low virulence of these RNA viruses on Dp, extensive screening of immunity and stress-related genes in our study revealed a number of defence/adaptation mechanisms that were regulated after viral challenge, and that could have contributed to the observed tolerance to infection. Immunity to viruses on the model species D. melanogaster appears to rely on two main arms: the degradation of viral RNA by the short interfering (siRNA) RNAi pathway (Palmer & Jiggins, 2015;Wynant et al., 2017), and an inducible immune response mediated by the Toll and JAK/STAT signalling pathways (Lemaitre & Hoffmann, 2007;Sparks et al., 2008), and possibly the Imd pathway (Rückert et al., 2014). In D. pteronyssinus, many of the genes regulated upon viral infection were involved in these pathways, as well as pathogen recognition receptors that activate the immune system, and antimicrobial peptides (AMPs) that act as effector proteins (Palmer & Jiggins, 2015). Beyond the canonical immunity set, transcriptomic surveys on D. melanogaster have also found several noncanonical genes also contributing to defence against pathogens (Keehnen et al., 2017). Our work suggests a similar response in D. pteronyssinus, in which some differentially expressed genes in the infected colony indicated stimulation of autophagy, a mechanism that allows the elimination of cytoplasmic components via autophagosomes in infected cells (Shelly et al., 2009), or apoptosis, an antiviral mechanism that can suppress virus-infected cells and for which some viral genomes have evolved proteins interfering its regulation . Other regulated genes included those related to the integrity of the peritrophic membrane (PM), an important host defence factor associated to the susceptibility to oral pathogenic viruses and for which some viruses are able to secrete PM-degrading chitinases (Liu et al., 2017), as well as genes promoting homeostasis (Royet, 2011) and cell-to-cell adhesion in the gut epithelial barrier (Güiza et al., 2018;Izumi et al., 2019), the site of viral infection detected in D. pteronyssinus . In addition, five different heat shock proteins (Hsp) were induced by viral infection, including the two top upregulated genes in our study. Interestingly, these proteins are regarded as a 'double-edged sword': on the one hand, they protect intracellular proteins from degradation under stress conditions maintaining homeostasis in the infected cell, but, on the other, they can be highjacked by the virus for its own benefit to complete different stages of its cycle (Shang et al., 2020;Wan et al., 2020). For the sake of conciseness, further discussion on additional defence-related genes has been addressed as supplementary material (Text S3). However, prevalence and serodominance (contribution to the total IgE binding) for Der p 23 have been very variable across studies, being substantially lower in some cases (Batard et al., 2016;Eder et al., 2020;Thomas, 2016). Der p 23 gets accumulated in the spent culture after being excreted by the mite (Weghofer et al., 2013), most probably for this reason its increase after infection was detected only on WMC samples and not on purified bodies.

Impact of virus infection on allergenicity
Genes encoding minor allergens, based on reported allergenicity hierarchy (Thomas, 2018), were also regulated in the virus-infected colony. The highest upregulation in our study corresponded to the Hsp allergen Der p 28. In addition, the three serine protease allergens, Der p 3, 6 and 9 were among the top downregulated genes in the infected colony. Consistently, trypsin (Der p 3) and chymotrypsin (Der p 6) enzymatic activities were reduced in virus-infected cultures. As for Der p 23, these two allergens are concentrated in faecal pellets (Herman et al., 2014;Zhan et al., 2010), thus their activity is much higher in WMC than in purified bodies .
Despite that the prevalence and serodominance (i.e., potency) of these minor allergens is expected to be low in human cohorts (Thomas, 2018), a clinically relevant contribution of Der p 3, 6 and 9 to allergy cannot be discarded since their serine protease activity exhibits additional IgE-independent immunoactivity (Jacquet, 2021; Reithofer & Jahn-Schmid, 2017).
And, third, electrophoresis of Dermatophagoides spp. proteins and immunoblotting with allergic patient's sera have revealed IgE reactivity to more than one Hsp70 , but also to Hsp20  and Hsp60 (Liao et al., 2018). infer HDM allergy (Thomas, 2016). However, at the same time these effects might be compensated by the repression of other allergens, some of them also eliciting IgE-independent immune responses, as discussed above. On top of that, although yet unknown, direct effects on the human immune system of components from the viruses themselves are also possible (Iwasaki, 2012). Hence, in vivo and in vitro experiments were needed to confirm potential effects on allergenicity. The allergenicity of control and infected-mite extracts did not differ based on the results from our murine model of HDM-driven allergic asthma.
In addition, IgE-binding tests to sensitized human's sera showed no differences between control and infected-mite extracts. These results suggest that HDM-infecting viruses are of minor clinical importance, and that current clinical practice is not to be affected by RNA viruses likely present in the manufacture HDM-based diagnostic products.
However, HDM-derived extracts are also clinically used to induce allergen-specific desensitization, and we cannot fully rule out the possibility that mite-infecting viruses could influence outcomes of AIT. In addition, since our IgE-binding assays do not contemplate human innate immune responses and in vivo mouse models may not be fully extrapolated to humans, the possibility of having missed other immune effects attributable to viruses cannot be discarded.

Concluding remarks
Based on the presented results, despite significant transcriptomic effects, the presence in our Dp colony of two dominant picornaviruses infecting digestive cells (DerpV3 and 1) and three minority RNA viruses from other families (DerpV4, 5 and 6) is not expected to substantially compromise the allergenicity of mite-derived extracts.
However, we demonstrate that viruses are able to alter allergen expression, and particularly major allergen Der p 23. From a pharmaceutical perspective, it remains to be studied if viral infections in industrial mite cultures might have consequences on product consistency within and between companies. Some manufacturers have assessed Der p 23 in their AIT products but, to our knowledge, not for standardization (Batard et al., 2016;Stranzl et al., 2021). We cannot predict how other virus species potentially infecting mite cultures may affect their performance or allergenicity, as this should be studied case-by-case. In this regard, refreshing mite stock cultures by introducing naturally-derived mites should be taken with caution to avoid the introduction of viruses or other pathogens. It is worth noting that the effects of one virus on the host will depend on the rest of the virome; for example, persistent mild infections could prevent proliferation of more pathogenic ones (Escobedo-Bonilla, 2021).
Finally, the high-quality assembly and thorough annotation of the Dp genome described here will be publicly available through the Online Resource for Community Annotation of Eukaryotes (ORCAE), where each gene has a specific entry including protein domains, GO terms, or Blast homology information; and where continuous community curation of the annotation will be available (Sterck et al., 2012).
This release, together with that of other recently annotated mite genomes (https://bioinformatics.psb.ugent.be/orcae/), represents an outstanding tool to assist future molecular biology research on this and other allergy-causing mites.

CONFLICT OF INTEREST
BNL has received research grants or consulting fees from GSK, OncoArendi, ArgenX and ALK. PHC leads a R&D support contract with ALK. In all cases, these relationships were conducted outside the submitted work, thus they are not directly relevant or directly related to the work described in this manuscript. The rest of authors have no potential conflict of interest to declare.

DATA AVAILABILITY STATEMENT
The sequencing, genome assembly and annotation data that support the findings of this study will be openly available in NCBI and ORCAE repositories, accession numbers and URLs are indicated in the text and supplementary information. Any other data that is not available as supplementary information will be available from the corresponding author upon reasonable request.