Auto and co-aggregation of L. pneumophila and H. pylori with drinking water bacteria
In a previous study several bacterial strains were isolated from heterotrophic biofilms formed on uPVC coupons in a two-stage chemostat system . For the present work, the selection of the bacteria used was based on the prevalence of these isolated strains in biofilms, i.e., the strains that were always present in biofilm samples when detected by culture were used rather than those only found intermittently.
In the aggregation studies it was observed that there was no auto-aggregation of any of the bacteria tested in this study, as demonstrated previously for Brevundimonas
vesicularis, Acidovorax delafieldii and V. paradoxus [34, 38]. No co-aggregation of L. pneumophila or H. pylori was observed with any of the bacteria isolated from drinking water biofilms, demonstrating that while all of the bacteria used in this study have the ability to form biofilms they are attaching to the uPVC surfaces without aggregating in the planktonic phase with the other microorganisms .
L. pneumophila in biofilms
The L. pneumophila cells from the inocula prepared for the biofilm experiments were quantified for total, PNA-positive and cultivable cells. Results showed that cultivable and PNA numbers were similar but were only 50% of the numbers obtained by SYTO 9 staining. It is still controversial whether PNA probes detect dead cells or if they just produce a detectable signal with viable cells. PNA probes have been used to detect pathogens in mixed biofilms but it has not been well established if this technique can also detect non-viable cells [23, 29, 39]. However the similarity in the cultivable and PNA-positive numbers, and the difference between PNA-labelled and total cells (stained by SYTO 9), strongly indicates that the PNA probe fails to detect dead cells. PNA probes bind specifically to rRNA molecules emitting a signal that can be visualized under microscopy. The intensity of that signal is related to the rRNA content, i.e., the higher the rRNA content the brighter the signal is . A very low content of rRNA would result in insufficient brightness and cells would not be visualized. After cellular death the content of rRNA decreases significantly and therefore some authors have suggested that the emission of a bright signal is a good indication of cell viability [39, 41, 42]. The results obtained in the present study (Figure 1a) support that there is a correlation between the number of viable cells and the number of cells that bind to the PNA probe but further studies should be performed to correlate the PNA-positive cells and their metabolic state.
As it has been demonstrated before by other authors [43, 44], the attachment of L. pneumophila cells to the uPVC surface occurred on the first day of biofilm formation and the numbers of total and PNA stained cells, from mono-species biofilms, did not change significantly (P > 0.05). Nevertheless, the numbers of cultivable cells increased in the first two weeks and decreased during the rest of the experiment. It has been demonstrated that L. pneumophila can survive in tap water for long periods without losing cultivability [45, 46], but is not able to replicate in axenic cultures in tap water or in low nutrient media, except when associated with biofilms or parasitizing amoebal species [29, 47, 48]. After two weeks the cultivability decreased but was not completely lost for the 32 days of the experiment which indicates that biofilms are a protective niche for L. pneumophila, even in axenic culture. Conversely, PNA-positive numbers with a high fluorescence intensity remained constant and, for the same reason explained before, this suggests that cells are still viable. Moreover, the fact that total L. pneumophila and L. pneumophila PNA-positive cells remained constant with time indicates that there is no damage to DNA and rRNA, respectively. Conversely, the variation of PNA-positive numbers in dual-species biofilms was used as an indicator of the variation of viable L. pneumophila cells inside of those biofilms.
The results of dual-species biofilms showed that when biofilms were formed in the presence of M. chelonae the percentage of cultivable L. pneumophila in relation to L. pneumophila PNA-positive cells was slightly superior compared to mono-species biofilms or dual-species biofilms with the other strains isolated from drinking water. Although the difference is not statistically significant this result indicates that this strain has a small positive effect on L. pneumophila cultivability. In contrast, the numbers of cultivable L. pneumophila decreased when this pathogen was associated with Acidovorax sp. indicating that this species has a negative impact on L. pneumophila cultivability. It was also observed that the numbers of cultivable L. pneumophila when co-cultivated with Sphingomonas sp. decreased and, although the statistical analysis showed that the difference is not significant, the fact that the cultivability was almost four-fold lower appears to reveal an antagonistic effect. Conversely, it appears that both strains affect negatively sessile L. pneumophila cultivability, either by competition for nutrients or production of a metabolite toxic to L. pneumophila. The fact that these two species were isolated on R2A reveals that they have low nutritional requirements to grow and might even be able to grow in water, contrary to L. pneumophila which is not able to grow in absence of, for example, L-cysteine and high iron concentrations [49, 50]. This is corroborated by the values shown in Table 1, where cultivable Acidovorax sp. and Sphingomonas sp. numbers are 6.55 × 106 and 1.06 × 106 CFU cm-2 suggesting that these two microorganisms could be metabolically active in the biofilm despite the poor nutrient concentration of the medium (filtered tap water). Another possible explanation for the lower numbers of cultivable L. pneumophila when biofilms were formed in co-culture with Sphingomonas sp., can be related to the structure of the biofilm. Figure 2 shows a 32 days-old biofilm formed by L. pneumophila and L. pneumophila associated with Sphingomonas sp. The biofilm formed in the presence of Sphingomonas sp. had a different morphology, and although the thickness of the biofilm has not been measured, the presence of microcolonies suggests the presence of thicker structures where anaerobic zones might occur. Wadowsky et al.  have demonstrated that in anaerobic conditions L. pneumophila loses cultivability and if biofilms formed by L. pneumophila and Sphingomonas sp. have indeed anaerobic zones, then it is possible that L. pneumophila located in those places has become uncultivable. It would therefore be interesting to undertake further research to measure the thickness of different parts of the biofilm and the respective concentration of oxygen and relate those results to the cultivability of cells from those regions. However, the fact that the numbers quantified by the use of a PNA probe remained constant, might indicate that these cells may still be viable and can probably recover cultivability in favorable conditions.
This work clearly demonstrates that L. pneumophila can be negatively or positively influenced by other microorganisms present in drinking water. It is important to note that this study was carried out under particular conditions and it will be important to perform more experiments in the future, in particular to study the effect of other drinking water bacteria, the formation of biofilms under dynamic conditions and the incorporation of a disinfectant, such as chlorine.
It is known that other bacteria can influence the growth of L. pneumophila either in nutrient-poor environments, such as drinking water, or in rich artificial media. Toze et al.  have demonstrated that some bacteria commonly present in heterotrophic biofilms, such as Pseudomonas sp. and Aeromonas sp., can inhibit the growth of L. pneumophila while Wadowsky and Yee  demonstrated that Flavobacterium breve can support the satellite growth of this pathogen on BCYE agar without L-cysteine. A curious result was obtained by Temmerman et al.  who demonstrated that dead cells can also support the growth of this pathogen. Although the mechanisms responsible for the influence of different microorganisms on L. pneumophila survival are unknown there is one aspect of L. pneumophila microbial ecology that has been already well-established: L. pneumophila is not able to grow in drinking water unless associated with biofilms or amoebal species [46, 47, 53]. Hence, the knowledge of how microorganisms affect L. pneumophila cultivability is a key factor for the effective control of this pathogen in drinking water and associated biofilms, and requires further investigation.
H. pylori in biofilms
In this study the cells recovered from mono-species H. pylori biofilms were always uncultivable, for all the time points, which is in contrast to the Azevedo et al.  study, where it was demonstrated that after 24 hours sessile H. pylori cells were still cultivable. This might be due to the differences in the method of cell removal from the coupons, the quality of water or the type of uPVC substratum. When the biofilm was formed in the presence of Brevundimonas sp. no cultivable H. pylori cells were ever recovered either. However, for this case, care should be taken in the interpretation of the results. In fact, Brevundimonas was able to grow on CBA medium in a faster and more abundant way then H. pylori. As such, it is impossible to determine whether H. pylori is indeed uncultivable in the presence of this microorganism, or whether it could not be detected because it was overgrown by Brevundimonas. We have attempted to solve this issue by using CBA medium supplemented with antibiotics but, as shown by other authors , available selective medium for H. pylori allows the growth of other species, including Brevundimonas sp. The fact that there were no differences in the results for the PNA-positive cell numbers obtained for H. pylori in mono-species biofilms and in dual-species biofilms with Brevundimonas sp. suggests that this bacterium has little or no effect on the inclusion of H. pylori in biofilms.
Cultivable H. pylori was never recovered from dual-species biofilms at any time point, independently of the second species used, except when H. pylori formed dual-species biofilms in the presence of M. chelonae and Sphingomonas sp. For these two microorganisms, it was observed that H. pylori was able to retain cultivability for a period of between 24 and 48 hours. This suggests that both microorganisms might have a positive effect on the inclusion and survival of this pathogen in drinking water biofilms. The ability of H. pylori to adapt to different physico-chemical parameters has been studied by several authors [30, 55–58], however no studies about the influence of other microorganisms on the survival of this pathogen have been found in the literature except the coculture of H. pylori with the protozoan, Acanthamoeba castellanii . The interaction of microorganisms in biofilms has been widely studied and in this particular case could be the key for the survival of this microorganism in drinking water systems, even if in a VBNC state. More investigations should therefore be performed concerned with the influence of drinking water microorganisms on H. pylori metabolism and survival.