Microorganisms inhabiting biofilms form structurally and functionally well-organized communities where the interactions among the participants can be extremely complex1. The in situ model systems arranged in controlled, natural and quasi-stagnant physical–chemical conditions are important to understand the changes that take place in the development of biofilms. Therefore, the present study performed in the RT spring cave highlighted some interesting phenomena on biofilm formation in thermal karst systems.
In the initial stage of biofilm formation, primary colonizing microorganisms generally bind through specific and non-specific interactions to a conditioning film adhering to solid surfaces, consisting of various organic materials19. It was also clearly visible on the SEM images (Fig. 2) made at the beginning of the model experiment. In the next stage of biofilm formation, microbial cells embed into an EPS. The increasing EPS production was observed from the 9–12 weeks during the experiment of biofilm development in the RT spring cave. The EPS consists of polysaccharides, glycoproteins, exoenzymes and nucleic acids that attach directly to the surface, also known as a substrate20,21.
The microscopically observed biofilm maturation correlated well with species richness estimators and diversity indices. The number of OTUs greatly increased during the first nine weeks when the morphological diversity of the biofilm also became more and more complex. Most of the dynamic changes observed at the OTU level happened until the twelfth week, so we could state that nine to twelve weeks were needed for the maturation of the biofilm. The greatest difference between the species richness estimators was among the water and biofilm samples.
In an earlier study, a natural biofilm sample (RTB) developed for years on the rock of the RT spring cave17 was analyzed. The taxonomic composition of bacteria inhabiting the natural and experimental biofilms was similar. The most abundant community members of the RTB as well as the 3–30 weeks and one-year in situ model biofilm samples were also the same, showing no difference or selection between the glass slide and the carbonate rock as a surface substrate. Based on our previous results17, unforeseen taxonomic bacterial diversity was obtained from these highly radioactive environments (600 ± 21 Bq/L radon concentration in the case of Rudas-Török spring cave) based on next-generation sequencing data, containing mainly unclassified bacteria affiliated with low level similarity to cultured bacterial taxa. The surprisingly high diversity suggests that the microorganisms living here are well adapted to this extreme environment.
As regards the taxonomic diversity, Proteobacteria was detected with relative high abundance both in the water and biofilm samples (Fig. 3). It could be presumed based on the previous studies, because members of this phylum were frequently detected as constituents of bacterial communities in different cave samples6,22,23,24,25,26. The water sample of the RT spring cave was dominated almost exclusively by an unclassified Hydrogenophilaceae (OTU3) that showed the highest sequence similarity with chemo-lithotrophic sulfur-oxidizing bacteria (Fig. 3). OTU3 (with its 96.1% relative abundance) could be considered as the potential ‘core OTU’ of the thermal water. To our knowledge, this is the first report about such extraordinary high relative abundance of an OTU in the thermal waters of karst caves. Deja-Sikora et al.27 reported a similar phenomenon, the dominance of one unclassified Betaproteobacteria OTU affiliated with Comamonadaceae (abundance ranging from 1.7 to 57.8%) in sulfide-rich waters of the Carpathian Foredeep. The authors assumed that members of Comamonadaceae were likely to represent archetypal microbial species in those waters. The dominant read of OTU3 originated from the discharging deep thermal waters, however, was almost completely absent from the biofilm samples of the spring cave. This finding was surprising because the role of bacteria arriving with the discharging water sample was hypothesized in the formation of the biofilm. Among Proteobacteria, an unclassified Gammaproteobacteria was the most abundant in the biofilm samples (Fig. 3), nevertheless, no more information is known about these unclassified phylotypes.
The phylum Chloroflexi was dominant in the biofilms throughout the studied period with the members of unclassified Anaerolineaceae (Fig. 3). These Gram-negative, filamentous, thermophilic and strictly anaerobic, chemo-organotrophic organisms28 can serve as the basis of the biofilm formation in the RT spring cave. The non-cultivated members of the class Dehalococcoidi have fermentative metabolism and can use N-acetylglucosamine under anoxic conditions (using nitrate as electron acceptor). N-acetylglucosamine, which forms the backbone of the murein of most bacterial cell walls, is released continuously when cells are destroyed29. Presumably, the representatives of these filamentous Chloroflexi can be the first adherent organisms according to the SEM images (Fig. 2), and the low oxygen level in the spring cave may have favored their reproduction. The OTUs assigned to the phylum Chloroflexi were also frequent not only in the biofilm formed on the glass slides but in the biofilm developed on the rock surface of the RT spring cave as well17.
Representatives of the genus Nitrospira (Nitrospirota) were also present throughout the experiment. Their highest proportion were observed in the sixth week of the biofilm formation (Fig. 3). The characteristic cell shape typical for the genus Nitrospira has been observed in the three-week biofilm sample on the SEM images, as well (Fig. 2). Nitrogen is frequently a limited nutrient source in caves; therefore, the importance of the nitrogen cycle has been emphasized in other studies25,30. Chemolithotrophic autotrophic prokaryotes, including nitrifiers, play a key role in the primary production of cave environments31. The presence of ammonia-oxidizing Nitrosospira and nitrite-oxidizing Nitrospira and Nitrobacter were revealed previously from the deposits of the cave wall of the western Loess Plateau of China32 and the presence of these organisms were detected in the caves and spring caves of the BTKS5,6,7,8, as well. Through their activity, nitrite-oxidizing aerobic chemolithotrophic bacteria may contribute to the low nitrite concentration values, which were also measured in the cave waters of the BTKS.
A possible reason for the low ammonia content in the BTKS is the oxidation of ammonia, in which members of both the Archaea and Bacteria may be involved. The ammonia-oxidizing archaea (AOA) organisms belonging to the phylum Thaumarchaeota appeared in high proportion in the archaeal clone libraries created from biofilms originated from the caves and spring caves of the BTKS6,7. For the members of the Archaea, an increasing temporal trend was observed in the biofilms from the fifteenth weeks, although the primer-pair which was used for amplification is rather Bacteria-specific33. The diversity and importance of Archaea in karst cave environments, in contrast to the Bacteria, is largely unexplored2,3,34. In the study of the speleothems of the Weebubbie Cave (Nullarbor karst, Australia) and Kartchner Caverns (Arizona, USA), the authors also demonstrated the importance of members of Archaea, especially the ammonia-oxidizing Thaumarchaeota2,3,34. Our findings may confirm the hypothesis that AOA organisms could have an important role in the nitrification process in the RT spring cave as well.
The members of the phylum Planctomycetota (Candidatus Brocadia) proved to be dominant in the biofilm samples (Fig. 3). Representatives of the ‘Candidatus Brocadia’ may participate also in the local nitrogen cycle by the anaerobic oxidation of ammonia (anammox) combined with nitrite reduction that results in the formation of elemental nitrogen35. The anaerobic ammonia-oxidizing bacteria grow very slowly, the fastest growing species also have a 10-day generation time36, which may be associated with the fact that the relative abundance of the phylum showed a significant increase only from the sixth week of biofilm formation.
Representatives of the phylum Patescibacteria (unclassified Parcubacteria) were found in high proportions in the biofilm samples (Fig. 3). These organisms were mostly observed in anoxic environments37, their presence can be associated with the low dissolved oxygen values in the RT spring cave. The members of the Parcubacteria have small genome size (< 1.1 Mbp) and based on this, it can be assumed that they form symbiotic relationships in the biofilm38.
Unclassified bacterial reads were also occurred in a high proportion in the biofilm samples and their relative abundance showed a significant increase during a year. In a metagenomics study of carbonate caves in the Kartchner Caverns, pyrosequencing resulted nearly 400,000 partial 16S rRNA sequence data in the case of the 10 examined samples. Unfortunately, most of the taxa obtained from the cave could not be identified, whereas the vast majority of prokaryotic taxa in the absence of a cultured representative is still unknown to science2.
In conclusion, the in situ experiment performed in this study allowed us to examine the development of biofilm over a year under controlled but natural and near constant physical–chemical conditions. From the adhesion of the first microbial cells, biofilm differed fundamentally from the pool water of the spring cave. At least nine weeks were needed for the development of a mature biofilm regarding the morphological complexity and taxonomic diversity. The prokaryotes involved in the aerobic and anaerobic nitrification processes were characteristic in the biofilm samples, in addition to the anaerobic fermentative and filamentous Chloroflexi, in contrast to the water sample where the dominance of an uncultivated member of the family Hydrogenophilaceae was observed.
Source: Ecology - nature.com
