In this study, we hypothesized that the concentration of mite nitrogenous waste products and mite population density influence microbial community structure inside mites and in the environmental culture of D. pteronyssinus. We previously observed differences in the microbiome composition of mites from young and old cultures [14] and the time changes of microbiome during colony culture growth of D. farinae [26]. Thus, we expected that the mites contaminate their environment by through their nitrogenous waste product guanine [18–20], which can serve as nutrient source for microorganisms. In contrast, increasing mite density increases the mite grazing effect on some microorganisms due to the selective feeding [27, 28]. In our experiments, the composition of mite microbial communities was remarkably consistent between replicates and composed of relatively few dominant taxa – 11 bacterial OTU and 3 fungal OTU. By plating of D. pteronyssinus cultures, we also identified bacterial taxa that were not detected by barcode sequencing in this study (i.e., Micrococcus aloeverae and Paenibacillus glucanolyticus) and fungal taxa (i.e., Penicillium griseofulvum, Penicillium brasilianum, and Hyphopichia pseudoburtonii) [29] which are the result of different methodical approach (e.g., PCR bias and depth of sequencing can contribute to distinct observed microbial communities in cultivation-dependent and cultivation-independent methods). The external bulk microbiome (i.e., the bulk medium containing food and mite excretions) presented a unidirectional shift in the structure of both bacterial and fungal communities, with increasing diversity over time, with the exception of fungal diversity in the mite internal culture, which showed a slight decrease. The microbial taxa observed in the environmental samples at the start of the experiment were nearly exclusively L. fermentum and S. cerevisiae, and these taxa were almost entirely absent by the end of the experiment.
The study confirmed the above described pattern of culture growth in D. pteronyssinus; however, we identified different durations of culture growth phases [5, 16, 17]. We were not able to distinguish the latency phase because our cultures were started with a high number of mites from the beginning; this also resulted in faster culture cycle, i.e. decline after 84 days. Similarly, shortening of the culture time and absence of latency phase was observed for D. farinae previously [26]. During culturing, the mites produce guanine, a nitrogenous waste metabolite [18–20]. The dynamic of concentration and microcosm respiration showed similar patterns to D. farinae [26]. The nitrogenous waste dynamics is connected to the density of mite population, decreasing in the decline phase of D. pteronyssinus for D. farinae [26]. The microbial respiration in microcosms showed two small peaks after 30 and 70 days in the both species of mites [26]. However, the two peaks of respiration detected here were not related to the microbial biomass evaluated here as the numbers of DNA copies amplified by universal primers. Previously we discussed the relationship between decreases of respiration and explained the first decrease of respiration by the decrease of S. cerevisiae and the second one connected to the decrease microbial biomass expressed as the numbers of copies of 18S rDNA in the samples of culture in the decline phase [26]. One explanation is that there were exchanges in the microbial community after both peaks of respiration. However, we did not find any effect of respiration on the composition of microbial community.
Culture growth time was the most important factor influencing the composition of microbes. This result agrees well with our previous findings for house dust mite cultures [14, 26]. The general explanation for saprophagous arthropods is that the arthropods influence the microorganisms by vectoring of microorganisms, selective feeding, opening the free spaces to microbial growth by diet fragmentation and waste production which modifies the culture environment [30]. The effect of these mite activities are suggested to increase with increasing mite population density [31]. In this study, mite population density increased during culture growth up until the decline phase. Although we did not measure vectoring and feeding directly, these are connected to mite density.
The ability of mites to change their culture environment through feeding activity and feces production was reported for stored product mites T. putrescentiae and fungal cultures [32, 33]. Feeding on fungi was well documented for D. pteronyssinus [34, 35]. The selective feeding of D. pteronyssynus on S. cerevisiae was expected, and earlier experiments have shown that the addition of yeasts to wheat germ flakes accelerated the growth of mites, though the population density was lower than on dried fish meal [36]. Mite feeding and digestion of S. cerevisiae explains the decrease of the numbers of yeasts in the culture; and rapid digestion can reduce the number of yeast DNA copies in the mite gut. We hypothesize that mite feeding on the yeasts leads to overgrazing and replacement by A. penicillioides. A previous study showed that this fungus was introduced to the culture via mites [37]. The microbiome analyses in the current study supports this finding, because the mites introduced into the culture contained 50% of A. penicillioides reads. The density of the fungus (observed as the numbers of its spores) increased in the culture with increasing time and mite density [37]. Here we detected a positive relationship between guanine and the relative abundance of amplicons derived from A. penicillioides in the environmental culture, and this may indicate that this fungus uses guanine as a source of nitrogen for mycelial growth. In prior studies, mites were able to feed on this fungus, but their fitness decreased in terms of survival which led to reduced population growth [38]. However, the mites were not able to produce second generation on an axenic diet without A. penicillioides fungus [38]. In this study, the relative abundance of C. mucifera increased with culture time in both inside mite and environmental culture samples. Prior studies have shown that the addition of a similar taxon Trichomonascus ciferrii (Candida ciferrii) to the diets of mites accelerates population growth of D. pteronyssinus in comparison to the rearing diet [29]. This demonstrates that both A. penicillioides and C. mucifera can serve as a food source for mites in the later stages of mite culture growth.
Similar to the previous findings for the D. farinae microbiome [14], no known acaropathogenic bacteria were identified here. This finding is consistent with the known microbiome of Korean strains of house dust mites [7, 13]. In general, bacteria have been suggested to be of lower importance than fungi for D. pteronyssinus population growth in comparison to the role in the growth of D. farinae [39]. This was observed in a study in which Micrococcus lysodeikticus was added to the mite diet, and the acceleration of D. farinae population growth in the comparison to control diet was measured [39]. However, in more recent laboratory experiments, addition of Bacillus cereus, Micrococcus aloeverae and Kocuria rhizophila into the diet decreased population growth rate of D. pteronyssinus mites in comparison to control diet, while addition of Staphylococcus nepalensis and Paenibacillus glucanolyticus, in turn, increased mite population growth rate [29]. In another study, a B. cereus strain isolated from T. putrescentiae feces was added to two types of diets, leading to decreased T. putrescentiae growth. The effect was much higher when the basis of the mite diet was dry dog food (rich in protein and fat) as compared to whole-meal spelt flour [40]. Thus, the diet used does affect bacterial growth, and although bacterial communities appear to be strongly associated with mites (mutualistic), some biotests indicated suppressive action by some bacterial taxa [40]. Mite-microbe-diet interactions are complex, and microorganisms can serve as commensals, antagonists, and as food. In addition, the type of diet can influence the mite environmental community composition. Mite bodies and feces could serve as microbial vectors and/or their reservoirs in mite cultures much in the same way as they do for fungi. Furthermore, some bacteria such as B. cereus can grow on mite bodies, exuviae and feces, providing a highly interactive relationship [40]. Similarly, the data presented in this study suggest that bacteria from the genera Kocuria, Virgibacillus, and Staphylococcus are introduced to the bulk culture media from the mites themselves, and this introduction is visible by 14 days of cultivation. Staphyloccus and Kocuria belong to the most frequently identified bacteria from the isolates obtained by plating D. farinae mites [41]. Thus, these bacteria appear to be carried by D. pteronyssinus from environment to environment through mite bodies and feces. Mites can change their feeding preference to consume diets with more Staphyloccocus, which could increase in the mite internal profiles of 40- and 84-day old samples. Guanine concentration in the bulk medium can inhibit growth of the bacteria; we observed a negative correlation between guanine and the bacterial taxa L. fermentum, L. massiliensis, and O. arenosus. However, these data are based on sequencing results and therefore represent relative abundance. Thus, in relative numbers, a shift in absolute abundance that is masked by changes in the proportions of other taxa in a sample.
It has been suggested that feeding of mites on A. penicillioides does not contribute to the allergenicity of D. pteronyssinus [38]. The allergen profiles of experimentally-derived fungus-free and fungus-fed mites were shown to be identical [38]. Thus, it is likely that there are other factors that influence allergen production besides the fungi. Other studies have shown that there are differences in allergen profiles among extracts from exponentially growing and declining mite cultures [5, 16, 17]. These differences could be associated to the population demographic parameters (juveniles/adult proportion) and/or by adaptation to the diet. Standard mite diets include S. cerevisiae, but the addition of other components, including proteins, lipids, and carbohydrates, modulates the major allergens Der p 1 or Der p 2 content in the mite bodies [42]. Although previous observation did not confirm the effect A. penicillioides on allergen production, the feeding of mites on other microorganisms can modulate the production, because they contains different nutrients than standardized diets.
We did not detect any Gram-negative bacteria that increased in relative abundance during growth of the mite D. pteronyssinus in this study. The data observed here indicate that our mite cultures contain almost no reads of Gram-negative bacteria, e.g., Bartonella-like bacteria [6] and Gram-negative parasitic Cardinium [43] observed in D. farinae. The observed absence of Gram-negative bacteria is important from a medical perspective because Gram-negative bacteria can produce endotoxins (i.e., heat stable lipopolysaccharides), contaminating commercially produced allergen extracts [6, 7]. However, we cannot exclude the possibility that Gram-negative bacteria could contaminate cultures under certain circumstances and become problematic. To confirm these findings more strains of D. pteronyssinus should be studied, and highly sensitive quantitative PCR assays can be employed to target bacterial taxa of interest.
One important finding of our study is the presence of Malassezia in the ingested community, although at low relative abundance (e.g. median values of 0.16%). Based on the proportion of the reads and the presence of the taxon in D. farinae [26], we believe that it is not a contaminant or other artifact of this study. Malassezia yeasts represent lipid-dependent and lipophilic organisms that are part of the skin microflora, but can be involved in skin disorders such as dermatitis or otitis eczema in humans and pet animals [44–46]. Our rearing diet did not contain any human or animal skin (a food source for Malassezia), and one possible explanation is that the Malassezia yeast detected in this study are part of a long-term symbiotic association with the mite D. pteronyssinus. We therefore propose the following hypotheses: (i) mites can be reservoirs and vectors of these yeasts, and (ii) in sensitive individuals, mite allergens can interact with the yeast in the development of cutaneous allergy diseases. Future research, including nutritional-based studies and better identification of Malassezia strains, will be necessary to confirm these hypotheses.