Abstract
Sulfate-reducing bacterial (SRB) biofilms are prevalent across diverse
environments, playing key roles in biogeochemical sulfur cycling while
also contributing to industrial challenges such as biofouling and
biocorrosion. Understanding the genetic and physiological adaptations
of SRB biofilms to different surfaces is crucial for developing
mitigation strategies. This study presents a comparative transcriptomic
analysis of Oleidesulfovibrio alaskensis G20 biofilms grown on copper
and polycarbonate surfaces, aimed at elucidating their differential
responses at the molecular level. RNA sequencing revealed 1255
differentially expressed genes, with copper-grown biofilms exhibiting
upregulation of Dde_1570 (flagellin; log2FC 2.31) and Dde_0831
(polysaccharide chain length determinant; log2FC 1.15), highlighting
enhanced motility and extracellular polymeric substance production.
Conversely, downregulated genes on copper included Dde_0132 (Cu/Zn
efflux transporter; log2FC −3.37) and Dde_0369 (methyl-accepting
chemotaxis protein; log2FC −1.19), indicating a metabolic shift and
stress adaptation to metal exposure. Morphological analysis via SEM
revealed denser biofilm clusters with precipitates on copper, whereas
biofilms on polycarbonate were more dispersed. AFM analysis showed a
4.6-fold increase in roughness on copper (44.3 ± 3.1 to
205.89 ± 8.7 nm) and a 3.8-fold increase on polycarbonate
(521.12 ± 15.2 to 1975.64 ± 52.6 nm), indicating surface erosion and
structural modifications. Protein-protein interaction analysis
identified tightly regulated clusters associated with ribosomal
synthesis, folate metabolism, and quorum sensing, underscoring their
role in biofilm resilience. Additionally, functional annotations of
uncharacterized genes revealed potential biofilm regulators, such as
Dde_4025 (cytochrome-like protein; log2FC 4.18) and Dde_3288 (DMT
superfamily permease; log2FC 3.55). These findings provide mechanistic
insights into surface-dependent biofilm formation, with implications
for designing antifouling materials and controlling microbial-induced
corrosion.
Keywords: Biocorrosion, Biofilm, Metallic, Non-metallic, SRB, Surface
roughness, Transcriptomics
Highlights
* •
Revealed novel gene expression patterns on metallic vs.
non-metallic surfaces.
* •
SEM showed direct adhesion on copper vs. EPS-mediated attachment on
polycarbonate.
* •
AFM detected 4.6 × roughness on copper, 3.8 × on polycarbonate
surfaces.
* •
Nanowire linkages observed in copper biofilms; absent on
polycarbonate.
* •
Upregulated flp gene on copper indicates flp-mediated adhesion to
metal surfaces.
1. Introduction
Biofilms are complex assemblages of microorganisms attached to surfaces
and encased in a self-produced extracellular polymeric substance (EPS),
are prevalent in diverse environments ranging from natural ecosystems
to industrial and clinical settings [[31]1]. These resilient microbial
communities exhibit remarkable resistance to antimicrobial agents and
immune responses, posing significant challenges in sectors such as
healthcare, water treatment, and oil industries [[32]2,[33]3].
Moreover, marine infrastructure in particular suffers from
contamination and materials degradation as a result of biofilm
formation, generating research incentive today [[34]4]. The study of
biofilms, therefore, holds immense importance for developing strategies
to mitigate their detrimental effects and harness their beneficial
applications.
Sulfate-reducing bacteria (SRB) biofilms have been extensively studied
in relation to metal exposure and corrosion, with research on
Desulfovibrio desulfuricans G20 (recently renamed as a
Oleidesulfovibrio alaskensis G20) demonstrating metal precipitation,
such as hexavalent uranium and lead, within both the periplasm and the
biofilm matrix [[35]5]. These SRB biofilms are known to associate with
enteric bacteria and aerobes on underground pipelines, contributing to
corrosion [[36]6]. Studies have shown that SRB biofilms primarily
consist of proteins with minimal exopolysaccharides, which aligns with
observations from Desulfovibrio vulgaris biofilms on glass, where the
EPS showed lower carbohydrate content compared to other bacteria
[[37]7]. While these studies provide insights into the corrosion and
metal-reducing capabilities of SRB biofilms, the broader physiological
adaptations between biofilms grown on metallic and non-metallic
surfaces in SRBs remain poorly understood. Oleidesulfovibrio alaskensis
G20, another sulfate-reducing bacterium, is a key organism in biofilm
research due to its ecological significance and industrial implications
[[38]8,[39]9]. This anaerobe thrives in environments like marine
sediments, oil reservoirs, and wastewater systems, playing a crucial
role in the sulfur cycle and biocorrosion [[40]10]. Known for its
robust biofilm-forming capabilities, O. alaskensis G20 serves as an
ideal model for investigating the molecular mechanisms governing
biofilm development on various surfaces, though the impact of different
substrata on its biofilm behavior remains insufficiently explored.
Surface properties are fundamental determinants of biofilm formation,
influencing microbial adhesion, colonization, and biofilm architecture
[[41]11]. Surface characteristics such as roughness, hydrophobicity,
charge, and chemical composition can significantly impact the initial
attachment of bacteria and the subsequent biofilm maturation process
[[42]12]. For instance, hydrophobic surfaces tend to promote bacterial
adhesion through hydrophobic interactions, while charged surfaces may
interact with bacterial cell walls via electrostatic forces
[[43]13,[44]14]. Comparative studies have underscored the need to
understand these interactions at a molecular level to develop effective
biofilm control strategies. In this study, we focus on two contrasting
surfaces: copper and polycarbonate. Copper, a metal with
well-documented antimicrobial properties, is extensively employed in
applications ranging from plumbing systems to medical devices due to
its ability to inhibit microbial growth [[45]15]. copper ions can
disrupt bacterial cell membranes, generate reactive oxygen species, and
interfere with essential cellular processes, making it a potent
material for reducing biofilm formation [[46]16]. Conversely,
polycarbonate, a synthetic polymer known for its durability and
biocompatibility, is widely employed in medical and industrial devices
[[47]17]. Despite its smooth and inert surface, polycarbonate can
support substantial biofilm formation, presenting challenges in medical
device-related infections and biofouling [[48]18].
Biofilm formation on metallic copper presents unique challenges due to
its antimicrobial nature and redox reactivity [[49]19]. The copper
surface not only releases toxic Cu^2+ ions but also participates in
redox cycling (Cu^2+/Cu^+), generating heterogeneous microenvironments
across the biofilm depth [[50]20]. Cells in the outer biofilm layers
may experience oxidative stress and metal toxicity, while those closer
to the surface may interact directly with the solid phase, resulting in
differential gene expression and stress adaptations [[51]21]. Our prior
study (Thakur et al., 2024) demonstrated that O. alaskensis G20
biofilms reach a physiologically stable state by day 7 when exposed to
30 μM copper [[52]9]. This maturation phase is characterized by
increased EPS production, downregulation of cell division genes (e.g.,
ftsZ, ftsA), and the induction of stress response and electron transfer
genes including cyt3 and NiFe-hydrogenases. In parallel, our previous
transcriptomic analysis of O. alaskensis planktonic cultures under
copper exposure (Tripathi et al., 2023) revealed
concentration-dependent regulation of genes linked to metal efflux
(copA, cusA), dissimilatory sulfur metabolism (dsrAB), and redox
balance (trxB, sodB, rubrerythrin) [[53]22]. Given that gene regulatory
programs during initial adhesion differ substantially from those
governing biofilm maturation, our objective was to elucidate long-term,
surface-dependent expression profile. Building on our published prior
observations, we selected day 7 to capture a representative mature
biofilm state. By comparing gene expression profiles between biofilms
formed on copper and polycarbonate, rather than conventional planktonic
cultures, we were able to uncover material-specific transcriptional
adaptations shaped by distinct physicochemical cues. These
surface-resolved comparisons enabled us to identify regulatory pathways
that would remain obscured in standard planktonic analyses, thus
offering a more ecologically and mechanistically relevant view of
biofilm development under abiotic stress.
In particular, we investigate the expression of genes involved in key
processes such as adhesion, EPS production, stress response, and
metabolic activity. For instance, genes encoding surface adhesins,
which facilitate the initial attachment of cells to surfaces, are
likely to exhibit differential expressions in response to the
contrasting properties of copper and polycarbonate. Stress response
genes, which enable biofilms to withstand adverse conditions, and
metabolic genes, which reflect the physiological state of the biofilm,
will also be key focal points of our analysis. These surface-resolved
comparisons enabled us to identify regulatory pathways that would
remain obscured in standard planktonic analyses, thus offering a more
ecologically and mechanistically relevant view of biofilm development
under abiotic stress.
2. Material and methods
2.1. Bacterial strain and growth condition
The anaerobic batch cultivation of O. alaskensis G20 was carried out in
125 mL serum bottles, which were tightly sealed using rubber septa and
aluminum caps to maintain anaerobic conditions. The growth medium
utilized was lactate sulfate medium (LS4D), comprising 60 mM lactate as
the electron donor and 50 mM sulfate as the electron acceptor [[54]22].
The pH of the medium was carefully adjusted to 7.2 using 5 M sodium
hydroxide (NaOH). To prepare the seed culture, 100 mL of the medium was
placed in a serum bottle and sterilized by autoclaving at 121 °C for
15 min. Post-autoclaving, the medium was deoxygenated by purging with
filter-sterilized ultrapure nitrogen at a pressure of 10 psi for
20 min. The seed culture was obtained from a stock maintained in a 40 %
glycerol solution at −80 °C. Prior to inoculation, the seed culture was
further purged with filter-sterilized ultrapure nitrogen for 1 h to
eliminate any residual hydrogen sulfide (H[2]S) [[55]8]. A 10 % (v/v)
inoculum of O. alaskensis G20 seed culture (OD[600] = 0.178), was used
for the experimental setup as described under section [56]2.2.
2.2. Setup of CDC batch reactor
Two anaerobic CDC biofilm reactors (Biosurface Technologies
Corporation, Bozeman, MT, USA), each filled with the same LS4D medium,
were employed for biofilm cultivation [[57]5]. The reactors were
sterilized via autoclaving, after which 200 mL of the sterilized LS4D
medium was aseptically added to each reactor. Each reactor contained
eight modified coupon holders designed to accommodate polycarbonate
(PC) and copper coupons (Cu). Copper (Alloy 101) Disc Coupon and
polycarbonate Disc coupons (Biosurface Technologies – RD128 copper and
RD128-PC), with dimensions of 12.7 mm diameter × 3.8 mm thickness were
used as-received (AR). These coupons underwent UV sterilization for
30 min before being placed into the holders. Following secure placement
of the coupons, the reactors were sealed, and anaerobic conditions were
established by purging with filter-sterilized ultrapure nitrogen gas at
a pressure of 10 psi for 40 min. Subsequently, the reactors were moved
into an anaerobic chamber (COY lab products) set to 30 °C, where they
were continuously stirred at a speed of 60 rpm. Each reactor was
inoculated with a 10 % (v/v) aliquot of the prepared seed culture.
Biofilm formation and growth were monitored over a period of 7 days.
Abiotic control experiments with copper and polycarbonate coupons were
conducted separately using the same LS4D media to facilitate surface
analysis through AFM and SEM.
2.3. Extraction and quantification of EPS components
EPS extraction was carried out from O. alaskensis G20 biofilms
cultivated on copper and polycarbonate surfaces. Biofilms were
harvested on day 7, to capture the progression of biofilm development.
The surfaces were gently rinsed with sterile, anaerobic
phosphate-buffered saline (50 mM PBS, pH 7.2) inside an anaerobic
chamber to remove loosely attached planktonic cells while preserving
the integrity of the surface-associated biofilm matrix. The remaining
biofilm was then carefully scraped from the surfaces using sterile,
single-use cell scrapers for EPS extraction. The collected biomass was
resuspended in 1 mL of 1.5 M NaCl solution to facilitate the release of
EPS and centrifuged at 5000×g for 15 min at 25 °C. The resulting
cell-free supernatant, containing the extracted EPS, was transferred to
sterile 1.5 mL microcentrifuge tubes for biochemical analysis [[58]9].
Carbohydrate content within the EPS fractions was determined using a
modified phenol-sulfuric acid method, with glucose as the calibration
standard [[59]23,[60]24]. Protein concentration was quantified using
the Pierce 660 nm Protein Assay Kit (Thermo Fisher Scientific, Waltham,
MA, USA), following the manufacturer's protocol for the Protein Broad
Range (BR) Assay. Bovine serum albumin (BSA) was used as the protein
standard in all assays. Nucleic acid was quantified using Nanodrop
UV–Vis spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA).
2.4. Biofilm visualization and surface characterization using scanning
electron microscope (SEM) and atomic force microscopy (AFM)
The biofilm architecture on polycarbonate and copper surfaces were
examined on day 7 using a field-emission scanning electron microscope
(Helios 5 CX FIB-SEM, Thermo Fisher Scientific, Waltham, MA, USA).
Briefly, 5 % glutaraldehyde fixative solution was prepared by mixing
0.4 mL of glutaraldehyde with 0.6 mL of cacodylate buffer. Each biofilm
coupon was gently rinsed with filter-sterilized 50 mM
phosphate-buffered saline (PBS) of pH 7.2 to eliminate planktonic cells
and debris. Using sterile forceps, the coupons were then transferred to
24-well plates, each containing 3 mL of the prepared fixative solution.
Biofilm samples were fixed by incubating at 4 °C overnight. After
fixation, the biofilms were washed twice with 3 mL of sterile distilled
water, allowing each wash to last for 20 min to remove any residual
fixative [[61]9]. Coupons were dehydrated by immersing them
sequentially in increasing concentrations of ethanol (50 %, 60 %, 70 %,
80 %, 90 %, and 100 %) for 15 min each. Once dehydration was complete,
the biofilm samples were air-dried thoroughly and stored in sterile
petri dishes until ready for SEM imaging. The SEM images were obtained
using secondary electron imaging operated at an accelerated voltage of
3 kV. The as-received coupons and the biofilm grown on copper and
polycarbonate coupons were coated with 10 nm platinum nanoparticles to
enhance conductivity and improve visualization (voltage at 3 kV).
However, the as-received and biofilm-removed copper and polycarbonate
surfaces were left uncoated with nanoparticles to avoid interference
with subsequent AFM analysis.
To analyze the surface profile of copper and polycarbonate coupons (AR,
abiotic control, post-biofilm), atomic force microscopy (AFM) was
performed using the Anasys nanoIR3 system (Bruker, MA, USA) at the
South Dakota School of Mines and Technology (SDSM&T) facility in Rapid
City, SD, USA. Biofilms were carefully removed from the surfaces prior
to analysis. The coupons were gently rinsed with autoclaved water and
treated with 30 μg ml^−1 proteinase K for 1 h at room temperature
(30 °C) to degrade the EPS matrix [[62]23]. Subsequently, the samples
underwent ultrasonication in an ultrasonic bath for 20 min to ensure
thorough biofilm removal. Furthermore, to ensure complete removal of
cell from the surface, the coupons were gently scraped using scratch
free nylon scrapper. The coupons were rinsed with autoclaved water
followed by 80 % (v/v) ethanol and dried in a biohazard cabinet. The
cleaned and dried coupons were mounted on a double-sided sticky black
carbon tape (5 mm in diameter) affixed to a mica substrate, providing
stable adhesion and minimizing contamination for high-resolution
imaging. The scanning was performed in tapping mode using a
PR-EX-TNIR-D-10 probe with specifications included a radius of 20 nm, a
spring constant of 40 N/m, a width of 40 μm, and an operational
frequency of 300 kHz. The scan rate was maintained at 0.2 Hz to ensure
high-resolution imaging, with areas of 5 × 5 μm and 30 × 30 μm analyzed
at an X–Y resolution of 512 × 512 pixels. Before scanning, the probe
was carefully engaged with the sample surface to achieve optimal
interaction. During the scanning process, the setpoint amplitude was
maintained at 5.281 V (target: 9 V) with a drive strength of 25.8 % at
a resonance frequency of 167 kHz, allowing for stable imaging
conditions without damaging the sample. The biofilm removal from both
the copper and polycarbonate surfaces was validated using SEM imaging
prior to proceeding with AFM characterization. All samples were
analyzed in biological triplicates, with three distinct surface sites
evaluated per sample to ensure spatial reproducibility and statistical
robustness.
The Anasys software suite was employed for all stages of the AFM
analysis, including data acquisition, parameter optimization, and
post-scan processing. The software enabled the generation of
topographical 3D surface views. Additionally, quantitative surface
characterization was performed, including roughness calculations such
as roughness (Ra), skewness (Rsk), and kurtosis (Rku). Profile analysis
was also conducted to examine surface features and variations in pits
and valleys, providing an assessment of the microstructural properties
of the samples.
2.5. Total RNA extraction from biofilm
Total RNA extraction of biofilm cells grown on polycarbonate (control)
and copper (test) samples was performed on day 7. To remove planktonic
cells, coupons were gently rinsed with anoxic 50 mM phosphate-buffered
saline (PBS, pH 7.2). The adherent cells on the polycarbonate and
copper surfaces were then carefully scraped into 1 mL of the same PBS
buffer. The cell pellets were collected by centrifugation at 10,000×g
for 10 min at 4 °C, followed by three washes with anoxic PBS to remove
any remaining debris. All the steps were performed under anaerobic
conditions using anaerobic chamber (COY lab products). Cell pellets
were transferred into sterile 2 mL RNase-free microcentrifuge tubes and
stored at −80 °C until further processing. RNA extraction was conducted
using the TRIzol reagent kit (Invitrogen, Carlsbad, CA, USA) according
to the manufacturer's protocol [[63]25,[64]26]. Subsequent RNA
purification was carried out using the RNA Clean & Concentrator Kit
(Zymo Research, Irvine, CA, USA, Catalog #R1019), with the purified RNA
eluted in 10 μL of nuclease-free water. RNA concentrations were
determined using a Qubit RNA Assay Kit in conjunction with a Qubit 3.0
fluorometer (Thermo Fisher Scientific, Waltham, MA, USA). RNA integrity
was assessed using the Bioanalyzer 2100 system (Agilent Technologies,
Santa Clara, CA, USA), confirming that all samples had an RNA integrity
number (RIN) greater than 7. All experiments were performed in 4
biological replicates.
2.6. Complementary DNA (cDNA) library preparation and sequencing
Library preparation and RNA sequencing were carried out at the Oklahoma
Medical Research Foundation NGS Core Facility (Oklahoma City, OK, USA).
The process involved several key steps: initially, ribosomal RNA (rRNA)
was removed using the RiboCop rRNA Depletion Kit (Lexogen, Greenland,
NH, USA). Following this, cDNA libraries were constructed using the
Swift Rapid RNA Library Kit, adhering to the manufacturer's
instructions (Swift Biosciences, Ann Arbor, MI, USA). The quality of
the prepared libraries was assessed through Kapa qPCR and an Agilent
Tapestation 4150 system (Agilent Technologies, Santa Clara, CA, USA).
Sequencing was conducted on the Illumina NovaSeq platform, utilizing an
S4 flow cell with a 150-base pair paired-end configuration. This setup
generated approximately 20 million reads per sample, with 10 million
reads per direction. All experiments were performed in 4 biological
replicates to ensure statistical reliability [[65]22].
2.7. QC of raw RNA sequencing reads and data analysis
We receive the raw file in FASTQ.gz format and analysis were conducted
on the Galaxy web-based scientific platform, ensuring reproducibility
and accessibility of the bioinformatics workflow. To ensure
high-quality data for subsequent analysis, raw RNA-seq reads underwent
rigorous quality control using the FASTQC tool, which assessed the
quality of the sequences by generating average Q scores across the full
length of each sequence file. Following this initial QC, the reads were
processed using Trimmomatic (version 0.38), with an Illumina-specific
clipping step employed to remove adapter sequences (TruSeq3
paired-end). Next, a sliding window approach was applied to further
filter the reads, truncating those segments where the 4-base average Q
score dropped below 20. Reads longer than 20 bp after this filtration
step were retained for downstream analysis. The quality-controlled
reads were then mapped to the Oleidesulfovibrio alaskensis G20
reference genome (NCBI assembly accession# ASM1266v1) using the HISAT2
alignment program, ensuring high precision in aligning reads to the
genomic framework. Read counts corresponding to each gene were
calculated using the O. alaskensis G20 genome annotation file (GFF3,
Ensembl), in conjunction with the read summarization tool
featureCounts. For differential gene expression analysis, the DESeq2
package was utilized to estimate logarithmic base-2 fold changes
(log2FC) in gene expression between the biofilm grown on polycarbonate
(control) and copper surfaces. DESeq2 employs the median-of-ratios
normalization method to account for differences in sequencing depth
between samples, providing robust differential expression results.
2.8. Bioinformatics analysis
2.8.1. Mapping of genomic data to protein and pathway database
The dataset generated from DeSeq2 on the Galaxy web-platform was
carefully post-processed to interpret the fold changes associated with
differentially expressed genes (DEGs). Initially, gene IDs from the
table were mapped to corresponding gene and protein names by retrieving
the O. alaskensis G20 datasets from UniProt, ensuring accurate
alignment of identifiers across databases. This step provided a
structured mapping of gene information to facilitate downstream
analysis. To link gene IDs with associated pathways, a custom Python
web-scraping script was developed. This script automates the retrieval
of pathway annotations for each gene ID from the KEGG webserver,
streamlining data integration and enhancing the functional
interpretation of DEGs. Following the mapping, DEGs were further
filtered based on statistical significance, applying a p-value cutoff
of <0.05 to retain only the most relevant changes. Certain sections of
the study applied an additional threshold, considering only genes with
an absolute log2 fold change (|log2FC|) greater than 1. This criterion
ensured a focus on genes with substantial expression changes, providing
a clearer distinction between significantly upregulated and
downregulated genes.
2.8.2. Gene ontology and Rich Factor analysis
DEGs were classified into GO terms based on their associated biological
processes (BP), molecular functions (MF), and cellular components (CC),
using the Gene Ontology Consortium's database. The z-score was employed
to measure the enrichment level of each GO term, reflecting the extent
to which observed gene counts for each term deviated from expected
counts. The z-score was calculated using the statistical equation:
z-score =
[MATH:
X−μσ :MATH]
, where, X represents the observed count of genes associated with a
specific GO term in the dataset, μ is the mean count of genes
associated with that GO term under a null hypothesis or background
distribution, σ is the standard deviation of gene counts for that GO
term in the background distribution [[66]27,[67]28]. The z-score in
this study was calculated and plotted using a custom python script.
The Rich Factor was used as a metric to quantify the level of pathway
enrichment [[68]29]. For each pathway, the Rich Factor was calculated
in MS-Excel using the following formulae:
[MATH: RichFactor=Tot<
mi>alnumberofDEGsinaspecificpathwayTota
lnumberofgeneswithinthatpathway :MATH]
2.8.3. Network analysis using cytoscape
The STRING database that integrates both known and predicted
protein-protein interactions (PPIs) was used to predict functional
interactions of proteins [[69]30,[70]31]. The PPI network between the
significant DEGs (p-value <0.05) was visualized using the PPI
visualization software Cytoscape (version-3.9.0) [[71]32]. O.
alaskensis G20 was selected as the organism, and a confidence score
cutoff of greater than 0.40 was used to find top interactors. Highly
connected regions of the network were detected using ClusterONE
(version 1.0) algorithm with the following criteria: minimum size = 5,
minimum density = 0.05, and edge weights = combined score. StringApp
(Cytoscape plugin) was used to perform pathway enrichment analysis and
import PPI networks from the STRING database to Cytoscape. The most
enriched gene set was selected based on a false discovery rate of
1.0 × 10^−6. The most enriched pathway was selected for analyzing the
interaction network, and a doughnut graph was assigned to each of the
nodes with upregulated and downregulated log2FC values. In the PPI
network, the nodes correspond to the proteins, and the edges represent
the interactions.
2.8.4. Putative annotation of hypothetical proteins
The online webserver MOTIFSearch
([72]https://www.genome.jp/tools/motif/) was employed to determine the
motifs present in each hypothetical protein [[73]33]. This webserver
offers a powerful platform for motif analysis by leveraging various
databases and pattern repositories. This webserver leverages
information from various sources including Pfam, NCBI-CDD (TIGRFAM,
COG, and SMART), and PROSITE pattern to identify motifs within the
amino acid sequences of the proteins. To ensure reliable results and
minimize the false discovery rate, a statistical analysis was conducted
using a low e-value threshold. The cutoff value of 1 × 10^−5 was chosen
as the criterion for significance. This threshold helps identify motifs
with a high level of confidence, enhancing the accuracy and reliability
of the motif predictions. Therefore, the putative functionality of the
proteins was determined through a motif-based search, which helped
assign potential functional annotations based on identified motifs. To
ensure enhanced reliability and accuracy in determining the
functionality of the identified genes, individual amino acid FASTA
files were subjected to ProteInfer, a deep network for protein
functional inference [[74]34]. By using this server, we obtained GO
terms along with associated confidence scores ranging from 0 to 1,
providing a more comprehensive understanding of the functional
annotations. The NCBI-VAST server was also used to perform
structure-similarity search to uncover the functions of the folding
domains.
2.8.5. Visualization
Graphical representations of datasets for analyzing DEGs and pathways
were created using custom Python scripts and GraphPad Prism
[[75]35,[76]36]. Matplotlib and Seaborn libraries were used in Python
to generate the plots [[77]37].
3. Results
3.1. Differential gene expression analysis
The global transcriptomic landscape of O. alaskensis G20 biofilms,
grown on two distinct surface conditions (copper and polycarbonate),
was analyzed using high-throughput Illumina sequencing. The sequencing
yielded comprehensive transcriptome data for four biological replicates
from each condition. The resultant reads were aligned to the reference
genome of O. alaskensis G20, as obtained from the NCBI database,
containing a total of 3376 genes [[78]38]. Pairwise comparisons between
the transcriptomes of biofilms formed on polycarbonate and copper
surfaces were conducted to identify DEGs between these two conditions.
The analysis revealed that out of the total 3376 genes in O.
alaskensis, 3251 genes exhibited differential expression between the
biofilms grown on the two surfaces (shown in [79]Fig. 1A). To refine
the analysis and ensure statistical robustness, the identified DEGs
were filtered based on a stringent significance threshold (p-value
<0.05). This yielded 1255 DEGs that were considered statistically
significant, suggesting that these genes may play a crucial role in
modulating the biofilm's physiological response to different surface
environments. To further categorize the molecular functions of the
significant DEGs, these genes were mapped to the Kyoto Encyclopedia of
Genes and Genomes (KEGG) database. A total of 383 DEGs were
successfully mapped to known KEGG pathways, implicating some of these
genes in pathways associated with biofilm formation, stress response,
and surface interaction. In contrast, 872 genes did not map to any
known pathways, indicating that these may represent novel or
less-characterized molecular functions in the biofilm response, or
other pathways, potentially specific to the unique stresses posed by
copper or polycarbonate surfaces. The Venn diagram in [80]Fig. 1A
visually represents the distribution of the differentially expressed
genes, highlighting the total number of DEGs, the significant subset of
DEGs (1255), and the genes mapped to KEGG pathways (383). Additionally,
it distinguishes the DEGs that remain unmapped to any pathway (872).
This comprehensive depiction underscores the transcriptional plasticity
of O. alaskensis biofilms in response to surface composition. The focus
of the study was directed toward the 1255 DEGs that demonstrated
statistical significance. The top 20 upregulated and downregulated
genes are shown in [81]Table 1.
Fig. 1.
[82]Fig. 1
[83]Open in a new tab
(A) Venn diagram of the total gene distribution and differentially
expressed genes (DEGs) in the O. alaskensis G20 genome. The figure
illustrates the entire set of genes in the O. alaskensis G20 genome as
the superset. The subsets represent the total number of DEGs
(0 ≤ p-value ≤1), DEGs with a p-value ≤0.05, and those with and without
associated metabolic pathways. (B) The volcano plot displays the
overall upregulated and downregulated DEGs with a p-value ≤0.05. The
top 20 most significant genes are highlighted by their gene identifiers
(Dde).
Table 1.
Top 20 upregulated and downregulated genes with log2FC and standard
error.
Gene ID Gene Name Protein Name log2FC Standard Error (±)
Dde_0404[84]^a Dde_0404 PEP motif putative anchor domain protein 4.84
0.17
Dde_1615[85]^a NA Amino acid transporter, AAT family 4.08 0.23
Dde_1616[86]^a Dde_1616 Metallo-beta-lactamase family protein 3.62 0.12
Dde_0147[87]^a NA PEP motif putative anchor domain protein 3.14 0.13
Dde_3028[88]^a Dde_3028 Carbon monoxide dehydrogenase (EC 1.2.7.4) 3.07
0.10
Dde_2562[89]^a NA Pyridine nucleotide-disulfide oxidoreductase family
protein 2.80 0.11
Dde_2509[90]^a NA PEP motif putative anchor domain protein 2.55 0.11
Dde_0882[91]^a NA Superoxide dismutase (EC 1.15.1.1) 2.42 0.13
Dde_1570[92]^a NA Flagellin 2.31 0.11
Dde_3160[93]^a rimP Ribosome maturation factor RimP 2.23 0.11
Dde_0823[94]^a NA PEP motif putative anchor domain protein 2.09 0.09
Dde_0594[95]^a NA Type 1 secretion target domain-containing protein
1.96 0.10
Dde_3511[96]^b NA Putative transcriptional regulator, ModE family −4.95
0.28
Dde_0657[97]^b NA 4Fe–4S ferredoxin iron-sulfur binding
domain-containing protein −3.57 0.10
Dde_0132[98]^b NA Permease −3.37 0.16
Dde_0408[99]^b NA Glycerol-3-phosphate dehydrogenase (EC 1.1.5.3) −3.23
0.16
Dde_0407[100]^b NA Transcriptional regulator, DeoR family −2.92 0.13
Dde_0409[101]^b NA Major intrinsic protein −2.29 0.12
Dde_0751[102]^b NA Probable membrane transporter protein −2.09 0.10
Dde_0989[103]^b atpF ATP synthase subunit b −1.50 0.07
[104]Open in a new tab
NA: Not applicable.
^a
upregulated genes.
^b
downregulated genes.
These DEGs likely participate in adaptive responses critical for
biofilm survival, attachment, and maintenance on the two contrasting
surfaces. The volcano plot shown in [105]Fig. 1B illustrates the
differential expression pattern of genes based on their log2FC (x-axis)
and their statistical significance (-log10 p-value, y-axis). Each point
represents a single gene, where red points indicate genes that are
significantly upregulated in copper-grown biofilms compared to
polycarbonate-grown biofilms, and blue points represent downregulated
genes. Grey points correspond to genes that did not show significant
changes in expression between the two conditions. Notably, several
genes such as Dde_3028, Dde_1616, Dde_0404, Dde_2509, and Dde_2562
exhibited a significant positive fold change and a corresponding high
statistical significance, indicating their critical role in the biofilm
development and growth on copper surface. The upregulated genes may be
involved in copper detoxification, biofilm matrix reinforcement, or
oxidative stress mitigation. The presence of copper likely triggers
metal stress, necessitating the upregulation of genes associated with
metal ion efflux, stress response, or redox balance. In contrast the
genes such as Dde_3511, Dde_0657, Dde_0407, and Dde_0408 exhibited
significant downregulation, suggesting their reduced relevance or
repression in copper-induced stress conditions. The downregulated genes
may be associated with processes that are more active on inert surfaces
like polycarbonate, possibly related to attachment, matrix formation,
or non-stress-related metabolic activities.
Overall, the volcano plot emphasizes the genes with the most
significant fold changes, highlighting the molecular foundations of
biofilm adaptation to surface-induced stress. Specifically, the genes
with the highest log2FC and most significant p-values are labeled with
their gene identifiers to indicate their prominent role in differential
gene expression. These findings suggest that the biofilm's
transcriptional response is finely tuned to counteract the toxic
effects of copper while maintaining biofilm integrity on both surfaces.
3.2. Rich Factor and pathway enrichment
A comprehensive analysis of gene expression revealed that, without
applying a log2FC cutoff, 383 genes were mapped to 70 unique KEGG
pathways. In contrast, imposing a cutoff of log2FC ≥ 1 significantly
narrowed the focus, resulting in 119 genes being mapped to 47 unique
pathways. To evaluate the enrichment levels of specific pathways in the
differential expression analysis between biofilms grown on
polycarbonate (control) and copper surfaces, we employed the Rich
Factor ([106]Fig. 2) as a key metric. This evaluation was conducted
across all 70 pathways, both with and without the application of the
log2FC cutoff. The Rich Factor was calculated as the ratio of the
number of DEGs within a specific pathway to the total number of genes
associated with that pathway. A high Rich Factor value indicates that a
substantial proportion of genes within the pathway are differentially
expressed, suggesting that the pathway is significantly affected by the
experimental conditions being studied. Conversely, a low Rich Factor
suggests that only a small fraction of the pathway's genes is
differentially expressed, implying minimal impact of the condition on
that pathway.
Fig. 2.
[107]Fig. 2
[108]Open in a new tab
The heatmap highlights the Rich Factor values across various pathways
before and after filtering with the log2FC cutoff. Each column
represents a distinct metabolic or biosynthesis pathway, with color
intensity indicating the magnitude of the Rich Factor, where darker
shades represent higher enrichment values. (For interpretation of the
references to color in this figure legend, the reader is referred to