Factors influencing persistence of Legionella pneumophilaserogroup 1 in laboratory cocultures

Legionella pneumophila is an ubiquitous microorganism present in both natural and artificial water systems that may cause a potential life threatening form of pneumonia called Legionnaires’ disease. Especially when colonizing the water distribution systems of buildings of hospitals or other health facilities, the risk for infections for immunocompromised individuals increases greatly. In such artificial aquatic systems, the biofilm covering the interior of pipelines and, more in general, the various plumbing elements represent relatively nutrient-rich spots where Legionella (and other bacteria) can attach and multiply [1],[2]. Biofilm is a matrix composed by heterogeneous aggregates of bacteria, fungi, protozoa, and algae embedded into extracellular polymeric substances [3]. As a result, many engineered water systems such as air conditioning cooling towers, water boilers, whirlpools and spas, drinking water distribution networks, shower heads, and dental-unit water lines provide an environment conducive to the growth and multiplication of Legionella species ([2] and references therein).

The survival of microorganisms in man-made systems results from the (interactive) effects of several factors which depend themselves from the structure and management of the water network. Those factors include temperature [4], pipe material [5], nutrient levels [6]-[8], frequencies, type and concentration of disinfectants [9],[10], water velocity [11] and hydraulic conditions [12]. Among the micronutrients, lower levels of certain metals enhance growth of L. pneumophila[13] and especially iron has been linked to L. pneumophila extracellular growth, intracellular replication, and virulence [14],[15]. As iron availability in water distribution systems may be linked with the older age and metal constituents of pipes [13], corrosion products are important factors in the survival and growth of L. pneumophila in artificial habitats [13],[16].

Several empirical environmental studies have examined the simultaneous effects of chemical elements, biofilm biomass (also quantified as heterotrophic plate counts, HPC, at 22°C and/or 37°C) and other factors on L. pneumophila presence [17]-[19] with inconsistent results [20],[21]. For example, occurrence of Legionella spp in samples from public and private structures (including hospitals, hotels and private houses) was positively associated with HPC at 22°C (but not with HPC 37°C), manganese (but not iron) and negatively with copper and higher water temperatures [22].

In experimental cocultures the growth and/or persistence of L. pneumophila was positively associated with several isolates of heterotrophic bacteria from potable water [23]-[26] as well as negatively with others, that were able to inhibit Legionella growth [27]-[29]. In particular, P. aeruginosa, another species often present in artificial water systems, seems able to secrete inhibitory substances like lactones [30] and resulted not conductive to L. pneumophila[26].

More generally, Legionella persistence in artificial systems was linked to its capacity of parasitize several species of protozoans and to multiply within them [2], to efficiently exploit available nutrients from the surroundings [2], and to obtain nutrients with necrotrophic feeding [31]. It has been suggested that multiplication of Legionella may dependent upon planktonic growth in alternative to sessile growth (e.g. within biofilm), because the planktonic phase would increase the likelihood of protozoan encounters [32]. If Legionella can persist in a planktonic phase also in absence of protozoans remains unclear, and this information is relevant for adopting the necessary management strategies in water systems where the presence of L. pneumophila increases the infection risk.

The aim of this study is to test the persistence of L. pneumophila in experimental planktonic cocultures subsided with iron, P. aeruginosa and other non Legionella bacteria (quantified as HPC at 37°C) isolated from the drinking water of a large hospital. In this system, despite the risk management interventions with regular disinfections, L. pneumophila has been isolated several times at several water point of use in the past few years [33].

Our results show a positive association between L. pneumophila concentration and the supplementation of HPC and iron and a negative association with P. aeruginosa presence. The main reason of this positive association is probably the fact that bacteria of the genus Legionella depends on other microbes to survive in nutrient poor environments such as the experimental planktonic co-cultures. In water distribution systems and other man-made aquatic systems, Legionella interacts with other microrganisms by adopting several survival strategies [17],[37]. The primary survival strategy is linked to Legionella ability of deriving the necessary nutrients (carbon, nitrogen, and amino acids) from dead microrganisms (via necrophilic exploitation, [31]) or by efficient uptake of nutrients excreted by other bacteria [2]. Second, Legionella is also able to parasitize a wide range of protozoans and to multiply within them [2],[38]. The same capacity is shown when Legionella infects alveolar macrophages of humans, causing pulmonary infections. Third, Legionella seems also able to enter a planktonic growth phase (opposed to the benthonic attached phase), in response to low nutrient conditions within the biofilm [2]. Additionally, biofilm embedded Legionella cells may have enhanced protection from chlorine based biocides that are commonly used to manage the infection risks in private and public buildings [1],[18],[39].

In our experiment, Legionella was able to persist in planktonic cultures supplemented with HPC but not in those supplemented with iron alone or with P. aeruginosa (regardless co-presence of iron and/or HPC). Therefore, although iron has been linked to L. pneumophila extracellular growth, intracellular replication, and virulence [14],[15], it was unable to sustain L. pneumophila without the contemporary presence of other microrganisms, which were probably the source of other essential nutrients. This result is also coherent with the observation that supplemented iron increases Legionella persistence in HPC supplemented co-cultures beyond the levels reached in co-cultures with HPC only (Table 3). It has been previously shown that iron (but also other elements necessary for Legionella growth like other metals or nutrients like phosphorus) in bioavailable form can be released by certain pipe constituents via biofilm mediated corrosion or from deposits of materials in stagnant portions of the water distribution system [40],[41]. Additionally, certain chemical compounds released from pipes can interfere with the chlorine based disinfectant decreasing its effects on biofilm cells [42],[43]. Therefore, when managing Legionella risk, the effect of pipe material and the physical structure of the water distribution network should be taken into account.

In our co-cultures the presence of P. aeruginosa, regardless of the level of supplements HPC and iron inhibited L. pneumophila growth from the early beginning of the experiments. This inhibitory effect on Legionella growth has been shown for several microrganisms residing in the same biofilm matrix. For example, [29] suggested that up to 32% of HPC bacteria (including Aeromonas, Vibrio, Pseudomonas and Pseudomonas-like strains) isolated from chlorinated drinking water were able of inhibiting or slowing the growth of Legionella species. Another study showed that L. pneumophila cannot attach in biofilms with pre-existent P. aeruginosa[26]. Although many mechanisms involved in the processes of microbial interference are only partially known [17], the production of inhibitory bacteriocins was linked to Legionella growth suppression in several earlier studies. For example, [28] showed that a bacteriocin producer like P. fluorescens, can inhibit both the formation and the stability of L. pneumophila biofilms. Similarly, it has been shown that Legionella can attach to P. aeruginosa monospecies biofilms but cannot persist for more than 2 days of incubation [26], possibly because of the effect of homoserine lactones produced by P. aeruginosa[30]. It remains to be clarified if Legionella can persist in viable but not cultivable form (VBNC, a form still able to cause human infections, [44]) even in presence P. aeruginosa. This may explain the reason why we found positive samples to Legionella VBNC in a previous cross sectional survey of at point of use water sources in the same hospital from where the biofilm of this experiment was obtained.

Many studies examined so far the relative effect of several factors that are reported empirically to increase the likelihood of Legionella presence in water distribution systems in settings like hospitals or other health care facilities where a formal risk assessment and management is a priority [1]. The results of our experiments support those studies reporting a positive association with HPC counts and iron subsidy and a strong negative association with P. aeruginosa, also in static flow conditions. Notably, all the microbial colonies used in this study were previously isolated from the hospital drinking water at point of use, and our data provide specific insights for the management of Legionella risk in this hospital setting. Although a confirmatory study should be designed in microcosms with dynamic flow conditions, our results highlight the need of controlling for both HPC and metal constituents into water systems of buildings at particular risk for the effects of Legionella exposure on humans such as hospitals and care homes for the elderly. Future studies should also elucidate the effect of protozoans in modifying the persistence of L. pneumophila in co-cultures as well as in water distribution systems. Moreover, additional effort should also investigate what environmental factors of the drinking water systems trigger the entrance to and the exit from VBNC state (including escape strategy from repeated chlorination cycles [39]). This understanding could lead to improved control measures for L. pneumophila in hospital settings.

AS and LM conceived the study, AC and LM carried out the lab experiments, AS carried out the statistical analysis and wrote the first draft of the manuscript, AC, LM and MDG reviewed the manuscript, MDG supervised the whole study. All authors read and approved the final manuscript.