Phenotypic adaptation and regulatory response to copper stress in Cupriavidus metallidurans

PhD thesis defended by Laurens MAERTENS (Prof. Jean-Yves MATROULE) - 25/10/2021

Joint PhD between SCK-CEN and UNamur

Promoter

Prof. Jean-Yves MATROULE, UNamur, Research Unit in Biology of Microorganisms (URBM)

Jury
  • Prof. Xavier DE BOLLE (Département de Biologie, UNamur), President
  • Prof. Jean-Yves MATROULE (Département de Biologie, UNamur), Supervisor
  • Dr. Kristel MIJNENDONCKX (SCK CEN)
  • Prof. Ralf MOELLER (Institute of Aerospace Medicine, Köln)
  • Dr. Andrea SASS (Department of Pharmaceutical Analysis, UGent),
  • Prof. Dirk SPRINGAEL (Soil and Water Management, KULeuven)
  • Dr. Ir. Rob VAN HOUDT (SCK CEN), Co-supervisor
Summary

Microbes are becoming more resistant to commonly used antimicrobials. This trend became noticeable soon after the introduction of mass scale production and use of novel antibiotics early in the 20th century, and is predicted to continue with devastating implications for sectors like healthcare and agriculture. Therefore, there is a need for the development of novel, alternative antimicrobials. For instance, the potent antimicrobial properties of metallic copper (Cu) have been investigated for use in biocontrol applications. Other than the aforementioned sectors, Cu-based antimicrobials could also be of interest for spaceflight operations, where decontamination protocols are presently insufficient. Consequently, there is a risk for the health of spacecraft crew, as well as for structural degradation of equipment. The ESA BIOFILMS project aims to study the bacterial response to metallic Cu aboard the International Space Station. This thesis was conceived in the context of ESA BIOFILMS, to provide new insights into the interaction of bacteria with metallic Cu, and to investigate the regulatory response to Cu exposure.

First, we evaluated the inactivation kinetics of Cupriavidus metallidurans CH34, a model organism for metal resistance, in wet contact with metallic Cu. Viability and membrane permeability were examined for 9 days with viable counts and flow cytometry. After an initial drop in viable counts, a significant recovery was observed starting after 48 hours. This behavior could be explained by either a recovery from an injured/viable-but-non-culturable state or regrowth of surviving cells by metabolizing lysed cells. Either hypothesis would necessitate an induction of Cu resistance mechanisms, since no recovery was seen in a CH34 derivative lacking metal resistance mechanisms, while being more pronounced when Cu resistance mechanisms were pre-induced. Interestingly, no biofilms were formed on the Cu surface, while extensive biofilm formation was observed on the stainless steel control plates. When CH34 cells in mineral water were exposed to CuSO4, a similar initial decrease in viable counts was observed, but cells recovered fully after 7 days. In conclusion, we showed that long-term bacterial survival in the presence of a Cu surface is possible upon the induction of metal resistance mechanisms. This observation may have important consequences in the context of the increasing use of Cu as an antimicrobial surface, especially in light of potential co-selection for metal and antimicrobial resistance.

In a second study, we performed a thorough analysis of the transcriptome of C. metallidurans CH34 acutely exposed to Cu by tagRNA-sequencing. Several metabolic pathways were impacted by Cu exposure, and a broad spectrum of metal resistance mechanisms, not limited to Cu-specific clusters, was overexpressed. In addition, several gene clusters involved in the oxidative stress response and the cysteine-sulfur metabolism were induced. In total, 7500 transcription start sites (TSSs) were annotated and classified with respect to their location relative to coding sequences (CDSs). Predicted TSSs were used to re-annotate 182 CDSs. The TSSs of 2422 CDSs were detected, and consensus promoter logos were derived. Interestingly, many leaderless messenger RNAs (mRNAs) were found and many mRNAs were transcribed from multiple alternative TSSs. We observed pervasive intragenic TSSs both in sense and antisense to CDSs. Antisense transcripts were enriched near the 5’ end of mRNAs, indicating a functional role in post-transcriptional regulation. In total, 578 TSSs were detected in intergenic regions, of which 35 were identified as putative small regulatory RNAs. Finally, we provided a detailed analysis of the main Cu resistance clusters in CH34, which were found to include many intragenic and antisense transcripts. These results clearly highlight the ubiquity of noncoding transcripts in the CH34 transcriptome, many of which are putatively involved in the regulation of metal resistance.

Third, we investigated the role of several newly detected putative sRNAs in C. metallidurans CH34 as regulators for Cu resistance. The presence and transcription start sites of these sRNAs was confirmed via 5’RACE, and their expression in several conditions was confirmed via luxCDABE-based promoter probe experiments. On the antisense strand of the silDCBA cluster, which encodes a tripartite efflux pump conveying copper and silver resistance, a previously unknown transcript was detected. The regulatory capacity of this transcript, dubbed silY, was confirmed by assaying its effect on translation of the polycistronic silDCBA mRNA. In addition, repression of SilDCBA translation by silY induced a Cu-sensitive phenotype, both in Cu tolerance and resistance. The biological importance of the silY sRNA requires further confirmation, with questions regarding its regulation and the precise mechanism of repression, but nevertheless our study has provided novel insights into the regulation of the silDCBA cluster.

In a final study, we examined the role of the environment in the bacterial response to Cu ion exposure. We employed a tagRNA-seq approach to elucidate the disparate responses of two morphotypes of Caulobacter crescentus NA1000 to moderate Cu stress in a complex rich (PYE) medium and a defined poor (M2G) medium. The study of the less metal-resistant C. crescentus allows for a comparison to C. metallidurans, while retaining relevance to oligotrophic environments such as those related to the ESA BIOFILMS project. Indeed, much like Cupriavidus strains, Caulobacter strains have been isolated from spacecraft environments. The transcriptome of C. crescentus NA1000 was more responsive to Cu exposure in M2G, where we observed an extensive oxidative stress response and reconfiguration of the proteome, as well as the induction of metal resistance clusters. In PYE, little evidence was found for an oxidative stress response, but several transport systems were differentially expressed, and an increased need for histidine was apparent. These results show that the response to Cu exposure is strongly dependent on the cellular environment. In addition, induction of the extracytoplasmic function sigma factor SigF and its regulon was shared by the Cu stress responses in both media, and its central role was confirmed by the phenotypic screening of a sigF::Tn5 mutant. In both media, stalked cells were more responsive to Cu stress than swarmer cells, and a stronger basal expression of several cell protection systems was noted in the latter, indicating that the swarmer cell is inherently more Cu resistant. Our approach also allowed for detecting several new transcription start sites, indicating small regulatory RNAs and additional levels of Cu-responsive regulation.

In conclusion, we have provided novel insights into the phenotypic and regulatory responses of bacteria to Cu exposure. Our results can be used to optimize Cu-based antimicrobials in a broad context. For instance, it has become clear that the use of viable counts for quantification of microbial load might lead to a strong underestimation. In addition, the hazard of bacterial Cu resistance mechanisms has been charted, with additional research to be performed on the horizontal transfer of these mechanisms. We have detailed the transcriptome of Cu-exposed cells of both C. metallidurans CH34 and C. crescentus NA1000 (the latter in disparate environments), which has enabled the search for novel regulatory features. As a result, the newly detected silY transcript has emerged as a putative regulator of the silDCBA cluster.