Characterizing the Effect of Campylobacter jejuni Challenge on Growth Performance, Cecal Microbiota, and Cecal Short-Chain Fatty Acid Concentrations in Broilers

Author details:

1 Department of Poultry Science, The University of Georgia, Athens, GA 3060, USA; 2 Toxicology and Mycotoxin Research Unit, US National Poultry Research Center, Athens, GA 30605, USA; 3 Department of Animal and Dairy Science, The University of Georgia, Athens, GA 30602, USA.

1. Introduction

C. jejuni is the leading cause of gastrointestinal disease in the United States [1,2]. The financial impact attributed to C. jejuni infection is around USD 2 billion [3]. C. jejuni is highly prevalent among poultry species, including broiler breeders, broilers, turkeys, and ducks [4,5]. Among these, broilers are the main reservoir for C. jejuni, establishing a persistent and benign infection that can reach up to 1 × 10⁹ CFU/g in the ceca [6]. The majority of C. jejuni cases are associated with the consumption of contaminated poultry products [7]. Typically, broilers are infected with C. jejuni at 2–3 weeks of age, and C. jejuni persists until slaughter day [8]. During slaughter, the release of cecal contents may result in cross-contamination, ultimately producing contaminated poultry products [9,10].

In humans, C. jejuni infection manifests as a self-limiting enteric disease characterized initially by fever and vomiting, followed by abdominal pain that transitions into watery or bloody diarrhea [11,12]. Conversely, in broilers, C. jejuni is a near-commensal bacterium, leading to asymptomatic infection [13]. The mechanism behind this dichotomy between benign infection in broilers and clinical disease in humans remains poorly understood. Efforts to elucidate these differences involved investigations into the role of avian intestinal mucus in conferring protection against C. jejuni [14–16]. The difference in structure and avian mucus efficacy in trapping pathogens is a major reason behind the benign infection in broilers [16,17]. Additionally, different reports emphasized the commensal nature of C. jejuni infection in broilers, highlighting the absence of an effective immune response despite the high colonization rate in the ceca [13,18]. Nonetheless, other papers have challenged the notion of C. jejuni being merely a commensal bacterium, citing observed reduction in performance parameters following the infection of C. jejuni [19–22].

The progression in DNA-sequencing techniques has enhanced our comprehensive understanding of C. jejuni’s enteric lifestyle within broilers [23]. Profiling the gut microbiota through sequencing has revealed the impact of C. jejuni infection on the microbial composition in the ceca, manifesting as an increase in the microbial diversity and increasing the relative abundance of Firmicutes at the expense of Bacteroidetes [24]. At the same time, other studies have reported an increase in the relative abundance of Clostridium [25], Streptococcus, and Blautia [24] compared to the non-infected groups.

Previous studies have primarily focused on sequencing a small fragment of the 16S rRNA gene, typically the V3-V4 regions. However, in the current study, a more comprehensive approach was utilized, which consisted of sequencing the entire 16S rRNA gene (V1–V9 regions) [26,27]. This approach should provide a more detailed result concerning the microbial composition. We hypothesized that C. jejuni would alter the gut microbial communities and function, leading to impaired bird growth. Therefore, this study was carried out to evaluate the effects of a C. jejuni challenge on performance production, cecal microbial composition, expression of microbial functional pathways, and SCFA concentrations in broilers.

2. Materials and Methods

All animal protocols utilized in this experiment were approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Georgia (AUP: A2021 6-012-Y1-A2).

2.1. Birds and Experimental Setup

The 35-day experiment was conducted with 60 one-day-old Cobb 500 broilers, randomly allocated into two treatments: control and challenge. Each treatment was replicated in six pens with five birds per pen. The birds had ad libitum access to water and feed, and the broilers were fed a basal diet (as shown in Table 1) throughout the experiment. On day 21, the birds in the challenge group were orally gavaged with 1 × 10⁸ C. jejuni/mL, while the control group was mock challenged with PBS [13]. Body weight and feed intake were recorded weekly. The average feed intake and body weight gain were adjusted for mortality to calculate the feed conversion ratio per pen.

Table 1. Basal diet composition fed from day 0 to day 35 of age.

2.2. C. jejuni Challenge Preparation

Under microaerobic conditions, the C. jejuni ATCC 33650 strain was cultivated on a campy-CEFEX agar and incubated at 42 ◦C for 48 h. After incubation, the bacterial cells were collected using a sterile inoculum loop and suspended in PBS (NaCl, Sigma Chemical Co., St. Louis, MO, USA). Subsequently, the optical density of the bacterial suspension was adjusted at an OD value of 0.2, as measured at 540 nm using a spec-20. The measured OD value of 0.2 corresponded to 1 × 10⁸ CFU/mL of C. jejuni [28]. This equivalence was further confirmed through serial dilution and plating on Campy-CEFEX plates.

2.3. Cecal Sample Collection and DNA Extraction

One bird per pen was euthanized on days 28 and 35 of the experiment, and cecal samples were collected under aseptic conditions into a sterile cryogenic tube. The samples were immediately frozen using liquid nitrogen. DNA extraction was performed using a hybrid protocol [29]. This protocol combines enzymatic and mechanical methods to optimize DNA extraction from the cecal content. Briefly, 0.35 g of the cecal sample was transferred into a 2 mL lysing matrix E tube (MP Biomedicals LLC, Irvine, CA, USA). The mechanical disruption of cecal samples was carried out using a QIAGEN vortex adapter for the Vortex-Genie 2 vortex (QIAGEN, Venlo, The Netherlands) for 10 min at maximum speed. A QIAamp Fast DNA Stool Mini Kit (QIAGEN, Venlo, The Netherlands) was used for the enzymatic extraction. Upon completing the DNA extraction, the concentration and the purity of the DNA were checked via spectrophotometry using a Synergy H4 Hybrid Multi-Mode Microplate Reader along with the Take3 Micro-Volume Plate (BioTek Instruments Inc., Winooski, VT, USA). Samples with concentrations lower than ten ng/µL were disqualified, and the DNA extraction process was repeated.

2.4. DNA Sequencing and Bioinformatics Analysis

Following the DNA extraction, all nine variable regions of the 16S rRNA gene were sequenced (regions V1 to V9), as described earlier [30], at Loop Genomics (San Diego, CA, USA). The whole 16S rRNA gene was synthetically reconstructed from a series of standard Illumina PE150 reads, which were assembled to rebuild the whole 16S rRNA gene, as established previously [26].

The resulting DNA sequences were converted to FASTQ files and were imported into QIIME 2 [31]. The control of sequence quality and filtration of chimeras were achieved using the QIME 2 DADA2 plugin [32]. An amplicon sequence variant (ASV) frequency table and a summary of each ASV length were generated. A phylogenetic tree was generated using the QIME 2 phylogeny plugin [33]. Taxonomic classification was performed using the QIME 2 feature-classifier plugin, which uses the Naïve Bayes classifier trained on the SILVA 138 SSU database [34]; the reads were classified via taxon using the fitted classifier [35]. The ASV table was rarefied to a common sampling depth of 1331 sequences/samples for alpha and beta diversity analyses. The following alpha diversity indices were computed: number of observed features (ASV), Shannon diversity index, Faith’s phylogenetic diversity index, and Pielou’s evenness index. For beta diversity, Bray–Curtis distances were used. The relative bacterial abundance was also quantified at the phylum, family, genus, and species levels. Finally, the Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2) was used to make inferences about the metabolic functions of the microbial community [36], and metagenome metabolic functions were assessed using the MetaCyc pathway database.

2.5. Short-Chain Fatty Acid Analysis

Short-chain fatty acid analysis was carried out as described previously [37]. Briefly, the cecal contents were collected on days 28 and 35 of the study and stored in a cryogenic tube. The samples were immediately frozen using liquid nitrogen and stored at −80 ◦C until further analysis. On the day of analysis, 1 g of each sample was diluted in 3 mL of distilled water and placed into 15 mL conical tubes. The samples were then homogenized via vortexing for 30 s, and 1.6 mL of each sample was transferred into a new tube and centrifuged at 10,000× g for 10 min. Following, 1 mL of supernatant was transferred to a new tube and mixed with 200 µL of a metaphosphoric acid solution (25% w/v). Each sample was vortexed for 30 s to ensure proper mixture and stored at −20 ◦C overnight. The next day, the samples were thawed and centrifuged at 10,000× g for 10 min. The supernatant was then placed into polypropylene tubes with ethyl acetate in a 2:1 ratio of ethyl acetate to the supernatant. The samples were vortexed for 10 s and then left to settle for 5 min. Next, 600 µL of the top layer was transferred into screw-thread vials to analyze the SCFA concentrations. A Shimadzu GC-2010 Plus gas chromatograph (Shimadzu Corporation, Kyoto, Japan) with a flame ionization detector and a capillary column (Zebron ZB-FFAP; 30 m × 0.32 mm × 0.25 µm; Phenomenex Inx., Torrance, CA, USA) was used for SCFA analysis. This equipment utilized helium as the carrier gas. The sample injection volume was 1.0 µL. The column temperature started at 110 ◦C and gradually increased to 200 ◦C. The injector and detector temperatures were set to 250 ◦C and 350 ◦C, respectively. The samples’ peak heights were compared to standards to determine the concentrations of SCFAs in the samples.

2.6. Statistical Analysis

A t-test was used to analyze the effect of C. jejuni on the performance production and concentrations of SCFAs, with the pen being considered the experimental unit. The individual microbial taxa and alpha diversity indices were analyzed using the Kruskal– Wallis H test. Differences in beta diversity were assessed via permutational multivariate analysis of variance (MANOVA—Adonis). The p-values were statistically significant when p ≤ 0.01 and considered a trend between 0.01 and 0.05.

3. Results and Discussion

3.1. Performance Parameters

Broilers serve as the principal reservoir for thermotolerant Campylobacter species, particularly C. jejuni, which establish colonization within the avian gastrointestinal tract between two and three weeks of age [38]. The considerable prevalence of C. jejuni in broilers implies its ability to outcompete different commensal bacteria, establishing a persistent infection that persists until slaughter day [39]. C. jejuni is usually detected in broilers around two to four weeks of age and is rarely detected at younger ages [38]. The absence of C. jejuni in newly hatched chicks can be partially attributed to maternal antibodies [40]. In our experiment, broilers were challenged with C. jejuni at three weeks of age to mimic the spread of C. jejuni infection within poultry farms. The C. jejuni challenge did not affect the performance production, feed intake, and feed conversion ratio (Table 2) compared to the control group (p > 0.05). These findings agree with results reported in different studies, as the C. jejuni challenge on day 14 [13] or day 21 [18] did not alter the performance parameters compared to the control group (p > 0.05). Such results highlight the near-commensal nature of C. jejuni and its relatively harmless infection, whose effects are typically limited to the gut. However, it should be noted that other studies have reported a decrease in body weight gain in C. jejuni-challenged broilers on days 35 [20] and 49 [41] of age. In these studies, the difference in production performance was observed four weeks post-inoculation, a time not covered in our study

Table 2. The effect of C. jejuni challenge on bird performance.

3.2. Microbial Diversity

After completing all filtering steps, the resulting number of sequences across the samples ranged from 1331 to 11,268 and were rarified to a sequencing depth of 1331 sequences per sample for the calculation of diversity indexes. A total of 7 phyla, 99 genera, and 83 species were identified in the ceca, in addition to 35% of other species that were unclassified by the used classifier. Table 3 summarizes the different computed alpha diversity indices. The gut microbiota play a central role in mediating colonization resistance against foreign microorganisms [42]. However, microorganisms have developed a set of virulence factors that enable them to bypass colonization resistance and persist in the gut [43]. In our study, C. jejuni decreased (p ≤ 0.05) alpha diversity, namely, the number of observed features and the Shannon diversity index, compared to the control group on day 28. The decrease in these alpha diversity indices reflects a reduction in microbial diversity and richness compared to the control group. Our results align with previously reported results as the C. jejuni challenge decreased the microbial diversity in the ceca compared to the control group [23]. Broilers were infected with C. jejuni at day 21 of age, and it is possible that C. jejuni established a persistent colonization load that led to a decrease in diversity as it dominated the ceca. However, the microbial diversity in the ceca was restored at day 35 of age, as no difference was recorded between the control and the C. jejuni-challenged birds. C. jejuni’s ability to persist in infection without causing a chronic shift in the microbial communities highlights its near-commensal nature. The beta diversity (Figure 1) between the samples was calculated via the Bray–Curtis dissimilarity, which showed a clear separation between the control and challenged groups (p < 0.01). The differences observed in the alpha diversity indices were mirrored in the beta diversity, indicating a possible shift in the overall structure of the microbial communities following C. jejuni challenge. It was also noted that C. jejuni-infected broilers samples were not as clustered as the control group samples, indicating a higher variability within C. jejuni-infected broilers compared to the control group.

Table 3. Effect of C. jejuni on alpha indices

Figure 1. The effect of C. jejuni challenge on the Bray–Curtis dissimilarity index. Birds in the challenge group were orally gavaged with 1 × 10⁸ CFU of C. jejuni/bird on day 21, and cecal samples were collected on day 28 and day 35 from control (n = 6; red) and challenged (n = 6; blue) groups. Each dot represents a replicate for each group (n = 6) on each day

Figure 1. The effect of C. jejuni challenge on the Bray–Curtis dissimilarity index. Birds in the challenge group were orally gavaged with 1 × 108 CFU of C. jejuni/bird on day 21, and cecal samples were collected on day 28 and day 35 from control (n = 6; red) and challenged (n = 6; blue) groups. Each dot represents a replicate for each group (n = 6) on each day

3.3. Bacterial Diversity at the Phylum Level

Table 4 shows the relative abundance of the phyla on days 28 and 35, in which Firmicutes and Bacteroidota dominated the cecal microbiota, constituting more than 93% of the total classified phyla in the ceca. These results align with previously reported numbers, as Firmicutes dominated the cecal, ileal, and jejunal microbiota found in broilers [42]. Concerning the changes in gut microbial composition, on day 28, the C. jejuni challenge increased the relative abundance of Cyanobacteria compared to the control group. Cyanobacteria constitute a diverse bacterial phylum that produces an extensive array of toxins known as cyanotoxins [44]. The exposure of birds and other animals to cyanotoxins, originating from blooms of aquatic Cyanobacteria, is well documented as a cause of neurologic and hepatic disease [45]. However, the role of Cyanobacteria in broiler gut microbiota remains insufficiently understood. A study has hinted at a potential correlation between the increased relative abundance of Cyanobacteria and increased production of IL-10 in broilers [46]. Notably, C. jejuni is considered a near-commensal organism within broilers. Establishing such commensal colonization necessitates a swift regulation of the pro-inflammatory immune response [47]. The increase in the relative abundance of Cyanobacteria, subsequently leading to an increased production of IL-10, can partially elucidate the tolerogenic immune response exhibited by broilers [46]. This response facilitates C. jejuni’s asymptomatic colonization and aids its persistence until the slaughter age.

Table 4. Effect of C. jejuni on the phyla relative abundance.

Similarly, the relative abundance of Desulfobacterota increased following C. jejuni challenge compared to the control group. Desulfobacterota is a phylum mainly present in marine environments and the mammalian intestinal tract; it consists of many sulfurreducing bacteria [48]. The main sources of sulfur in the diet are cysteine and methionine, while the main source of sulfur supplied by the host is the sulfated mucins [49]. In our experiment, both groups received the same diet. Hence, the dietary source cannot explain the increase in Desulfobacterota. One possible explanation is that C. jejuni infection increases the presence of mucin-degrading bacteria that can liberate L-fucose, which C. jejuni can use [50]. The degradation of mucus results in the liberation of sulfur, thus increasing the relative abundance of Desulfobacterota. Future experiments should investigate different metals such as iron, copper, and sulfur, which play a vital role in bacterial metabolism and function, to gain a broader understanding of the several changes occurring in the gut.

3.4. Bacterial Diversity at the Genera and Species Level

The top 10 most abundant genera are shown in Figure 2, and the significantly different genera on days 28 and 35 are shown in Table 5. Similarly, the 10 most abundant species are illustrated in Figure 3, and the significantly different species on days 28 and 35 are presented in Table 6. The cecal microbiota were mainly composed of Alistipes_A_871400, Romboutsia_B, and Faecalibacterium on day 28, while Alistipes_A_871400, Romboutsia_B, and Gemmiger_A_73129 dominated the cecal microbiota on day 35. These genera were identified at the species level, with Alistipes excrementavium, Romboutsia ilealis, and Faecalibacterium sp002160895 dominating the cecal microbiota on days 28 and 35. Alistipes excrementavium and Romboutsia ilealis are newly characterized bacteria that were recently detected in the ceca of broilers [51], explaining the lack of studies concerning the role of both species in broilers. On the other hand, Faecalibacterium sp002160895 belongs to the Faecalibacterium genera, widely recognized as beneficial inhabitants of the gut ecosystem, exerting a range of beneficial effects such as actively competing with pathogenic bacteria for nutrients and adhesion sites [52].

Table 5. Effect of C. jejuni challenge on the genera relative abundance. Only genera significantly affected by the challenge are shown.

Table 6. Effect of C. jejuni challenge on the species relative abundance. Only species significantly affected by the challenge are shown.

Figure 2. The effect of C. jejuni challenge on the top 10 most abundant genera on days 28 and 35. Birds in the challenge group were orally gavaged with 1 × 10⁸ CFU of C. jejuni/bird on day 21, and cecal samples were collected on day 28 and day 35 from control (n = 6) and challenged (n = 6) birds. Following the DNA extraction from the cecal samples, all nine variable regions of the 16S rRNA gene were sequenced (regions V1 to V9). On days 28 and 35, the taxonomic classification was performed using the QIME 2 feature-classifier plugin, which uses the Naïve Bayes classifier trained on the SILVA 138 SSU database

Figure 3. The effect of C. jejuni challenge on top 10 species’ relative abundance on days 28 and 35. Birds were orally gavaged with 1 × 10⁸ CFU of C. jejuni/bird on day 21, and cecal samples were collected on day 28 and day 35 from control (n = 6) and challenged (n = 6) birds. Following the DNA extraction from the cecal samples, all nine variable regions of the 16S rRNA gene were sequenced (regions V1 to V9). On day 28 and 35, the taxonomic classification was performed using the QIME 2 feature-classifier plugin, which uses the Naïve Bayes classifier trained on the SILVA 138 SSU database.

Figure 3. The effect of C. jejuni challenge on top 10 species’ relative abundance on days 28 and 35. Birds were orally gavaged with 1 × 108 CFU of C. jejuni/bird on day 21, and cecal samples were collected on day 28 and day 35 from control (n = 6) and challenged (n = 6) birds. Following the DNA extraction from the cecal samples, all nine variable regions of the 16S rRNA gene were sequenced (regions V1 to V9). On day 28 and 35, the taxonomic classification was performed using the QIME 2 feature-classifier plugin, which uses the Naïve Bayes classifier trained on the SILVA 138 SSU database.

Interestingly, in our experiment, the C. jejuni challenge increased the relative abundance of Faecalibacterium sp002160895 compared to the control group. This result aligns with previous experiments, as the C. jejuni challenge also led to an increase in Faecalibacterium’s abundance [23]. The changes in Faecalibacterium sp002160895 did not have an impact on the bird’s performance as the performance remained similar in both groups (p > 0.05). Faecalibacterium spp. are well-characterized producers of bioactive anti-inflammatory molecules such as shikimic and salicylic acids [53]. These anti-inflammatory molecules regulate the NF-κB/MAPK signaling pathway to reduce inflammation in the gut [53]. As previously mentioned, C. jejuni infection results in a tolerogenic immune response in broilers, and it appears that the increased relative abundance of Faecalibacterium sp002160895 enhances such a response through the production of anti-inflammatory compounds.

The C. jejuni challenge decreased the relative abundance of several genera, such as Sellimonas on day 28 and Acutalibacter and Oribacterium on days 28 and 35, compared to the control group. Acutalibacter is a genus of bacteria from the family of Acutalibacteraceae with one known species (Acutalibacter muris) [51]. Similarly, Oribacterium is a genus of bacteria from the family of Lachnospiraceae [51]. However, little information is known concerning the role of Acutalibacter and Oribacterium in the broiler gut microbiome.

Sellimonas genera were classified on the species levels as Sellimonas intestinalis. C. jejuni challenge decreased the relative abundance of Sellimonas intestinalis compared to the control group. Little information is presented regarding the role of Sellimonas intestinalis in broilers. However, Sellimonas intestinalis was suggested as a potential biomarker for gut hemostasis, as its relative abundance increased in several patients recovering from different dysbiosis events [54]. The decrease in the relative abundance of Sellimonas intestinalis on day 28 could potentially help C. jejuni in colonizing the ceca.

Despite the changes in the microbial composition and the decrease in the gut diversity following C. jejuni challenge, the predicted microbial function shown in Figure 4 remained unchanged compared to the control group. Similarly, the C. jejuni challenge did not affect the production of SCFA in the ceca shown in Table 7. The SCFAs are important in maintaining gut integrity and function. A higher concentration of SCFA compared to the small intestine activates acetogenesis-dependent genes in C. jejuni, enabling the bacteria to adhere and colonize the ceca [55]. Furthermore, acetate is considered a major energy carbon/energy source for C. jejuni. Although only numerically different in the current study, a high concentration of acetate in the ceca can support C. jejuni’s high load and allow C. jejuni to persist until slaughter age [56].

Table 7. Short-chain fatty acid concentrations on days 28 and 35.

Figure 4. The effect of C. jejuni challenge on the microbial function analysis. Birds were orally gavaged with 1 × 10⁸ CFU of C. jejuni/bird on day 21, and cecal samples were collected on day 28 and day 35 from control (n = 6) and challenged (n = 6) birds. On days 28 and 35, the Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2) was used to make inferences about the metabolic functions of the microbial community, and metagenome metabolic functions were assessed using the MetaCyc pathway database. ARGSYNBSUB-PWY: L-arginine biosynthesis II (acetyl cycle), ARGSYN-PWY: L-arginine biosynthesis I (via L-ornithine), ARGORNPROST-PWY: L-arginine degradation, ARGDEG-PWY: superpathway of L-arginine, putrescine, and 4-aminobutanoate degradation, ARG+POLYAMINE-SYN: superpathway of arginine and polyamine biosynthesis, ANAGLYCOLYSISPWY: glycolysis III (from glucose), ANAEROFRUCAT-PWY: homolactic fermentation, ALLCHORISMATE-PWY: superpathway of horismite metabolism, AEROBACTINSYN-PWY: aerobactin biosynthesis, 3-HYDROXYPHENYLACETATE-DEGRADATION-PWY: 4-hydroxyphenylacetate degradation, 1CMET2-PWY: folate transformations III (E. coli). Mean metabolic pathways indices were analyzed using the Kruskal–Wallis H test

Figure 4. The effect of C. jejuni challenge on the microbial function analysis. Birds were orally gavaged with 1 × 108 CFU of C. jejuni/bird on day 21, and cecal samples were collected on day 28 and day 35 from control (n = 6) and challenged (n = 6) birds. On days 28 and 35, the Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2) was used to make inferences about the metabolic functions of the microbial community, and metagenome metabolic functions were assessed using the MetaCyc pathway database. ARGSYNBSUB-PWY: L-arginine biosynthesis II (acetyl cycle), ARGSYN-PWY: L-arginine biosynthesis I (via L-ornithine), ARGORNPROST-PWY: L-arginine degradation, ARGDEG-PWY: superpathway of L-arginine, putrescine, and 4-aminobutanoate degradation, ARG+POLYAMINE-SYN: superpathway of arginine and polyamine biosynthesis, ANAGLYCOLYSISPWY: glycolysis III (from glucose), ANAEROFRUCAT-PWY: homolactic fermentation, ALLCHORISMATE-PWY: superpathway of horismite metabolism, AEROBACTINSYN-PWY: aerobactin biosynthesis, 3-HYDROXYPHENYLACETATE-DEGRADATION-PWY: 4-hydroxyphenylacetate degradation, 1CMET2-PWY: folate transformations III (E. coli). Mean metabolic pathways indices were analyzed using the Kruskal–Wallis H test

4. Conclusions

In conclusion, this study highlights C. jejuni’s ability to establish near-commensal colonization in broilers, transiently altering the gut microbiota’s composition and diversity without significantly compromising microbial function or the broiler’s performance production. Our findings shed light on the host–pathogen interaction, which is an important element for improving poultry health and food safety. However, further studies remain imperative to fully understand the C. jejuni–host interaction.

This article was originally published in Animals 2024, 14, 473. https://doi.org/ 10.3390/ani14030473. This is an Open Access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).