A comprehensive overview of the Chloroflexota community in wastewater treatment plants worldwide

ABSTRACT Filamentous Chloroflexota are abundant in activated sludge wastewater treatment plants (WWTPs) worldwide and are occasionally associated with poor solid-liquid separation or foaming, but most of the abundant lineages remain undescribed. Here, we present a comprehensive overview of Chloroflexota abundant in WWTPs worldwide, using high-quality metagenome-assembled genomes (MAGs) and 16S rRNA amplicon data from 740 Danish and global WWTPs. Many novel taxa were described, encompassing 4 families, 13 genera, and 29 novel species. These were widely distributed across most continents, influenced by factors such as climate zone and WWTP process design. Visualization by fluorescence in situ hybridization (FISH) confirmed their high abundances in many WWTPs based on the amplicon data and showed a filamentous morphology for nearly all species. Most formed thin and short trichomes integrated into the floc structure, unlikely to form the typical inter-floc bridging that hinders activated sludge floc settling. Metabolic reconstruction of 53 high-quality MAGs, representing most of the novel genera, offered further insights into their versatile metabolisms and suggested a primary role in carbon removal and involvement in nitrogen cycling. The presence of glycogen reserves, detected by FISH-Raman microspectroscopy, seemed widespread across the phylum, demonstrating that these bacteria likely utilize glycogen as energy storage to survive periods with limited resources. This study gives a broad overview of the Chloroflexota community in global activated sludge WWTPs and improves our understanding of their roles in these engineered ecosystems. IMPORTANCE Chloroflexota are often abundant members of the biomass in wastewater treatment plants (WWTPs) worldwide, typically with a filamentous morphology, forming the backbones of the activated sludge floc. However, their overgrowth can often cause operational issues connected to poor settling or foaming, impairing effluent quality and increasing operational costs. Despite their importance, few Chloroflexota genera have been characterized so far. Here, we present a comprehensive overview of Chloroflexota abundant in WWTPs worldwide and an in-depth characterization of their morphology, phylogeny, and ecophysiology, obtaining a broad understanding of their ecological role in activated sludge.

Chloroflexota have a versatile facultative anaerobic metabolism, likely providing them with a competitive advantage in treatment plants with nitrogen removal and/or enhanced biological phosphorus removal systems characterized by alternating oxic and anoxic stages (2).The preferred substrates of known Chloroflexota filamentous species are carbohydrates, complex polymers such as cellulose, or amino acids (2).All the genera described have the potential to ferment different products, and some have the genes to carry out dissimilatory nitrate reduction or partial denitrification (11,12), indicating a potential role in nitrogen removal from wastewater.The isolation of a nitrite-oxidiz ing bacterium, Nitrolancea hollandica, belonging to this phylum further supports their possible involvement in several steps in the nitrogen cycle (13).
Historically, filamentous bacteria in AS were identified using morphological features defined by specific staining methods and light microscopy (14)(15)(16), which often resulted in imprecise classification with little phylogenetic resolution (4).The introduction of high-throughput DNA sequencing and bioinformatics tools offered a breakthrough for the profiling of microbial communities.However, incomplete universal reference databases, along with a lack of taxonomy for most uncultured lineages, including abundant Chloroflexota in WWTPs, have hampered our ability to study these key organisms at lower taxonomic ranks (17).To improve the taxonomic resolution in microbial profiling studies, we introduced the Microbial Database for Activated Sludge (MiDAS), which includes a global, ecosystem-specific 16S rRNA gene reference database for wastewater treatment systems (MiDAS 4) (18).It also serves as a powerful tool for the design of genus-or species-specific fluorescence in situ hybridization (FISH) probes, which can subsequently be applied in combination with other techniques (e.g., microautoradiography or Raman microspectroscopy) for physiological characterization (1,19).Furthermore, the recent retrieval of thousands of high-quality (HQ) metagenomeassembled genomes (MAGs) (20,21), together with the retrieval of several MAGs from Chloroflexota abundant in AS (12,(22)(23)(24), will allow us to obtain an improved overview of the phylogeny and role of Chloroflexota in the AS system.
Here, we present a comprehensive overview of the Chloroflexota abundant in WWTPs worldwide, using HQ MAGs and amplicon data from Danish and global WWTPs in combination with the MiDAS 4 database.In total, we described 4 families, 13 genera, and 29 novel species, which appeared to be widely distributed across most continents and influenced by factors such as climate zone and WWTP process design.The design of specific genus-level FISH probes enabled investigation of their morphology, abundance, and spatial arrangement.Most of the novel Chloroflexota presented a typical filamentous morphology and demonstrated the presence of glycogen reserves, as detected by FISH-Raman.Moreover, the annotation of 53 HQ MAGs, recently retrieved from Danish WWTPs (21), provided further insights into Chloroflexota functional potential and their involvement in nutrient cycling.This study represents a fundamental milestone in the understanding of the ecological role of these microorganisms in the activated sludge microbial community.

Sampling and fixation
Sampling of AS was carried out within the Danish MiDAS survey (25) and the global MiDAS project (18).In short, fresh biomass samples from full-scale AS WWTPs were collected and either sent to Aalborg University (Danish MiDAS) or preserved in RNAlater and shipped to Aalborg University with refrigerating elements (Global MiDAS).Upon arrival, samples were stored at −20°C for sequencing workflows and fixed for FISH with 50% ethanol (final volume) or 4% paraformaldehyde (final volume), as previously described (26).

Community profiling using 16S rRNA gene amplicon sequencing
DNA extraction, sample preparation, and amplicon sequencing were performed as previously described (18,25).Briefly, DNA was extracted using a custom plate-based extraction protocol based on the FastDNA spin kit for soil (MP Biomedicals).The protocol is available at https://www.midasfieldguide.org/guide(aau_wwtp_dna_v.8.0).For Global MiDAS samples, RNAlater was removed by centrifugation and resuspension of the sample in 320 µL of PBS.For Danish MiDAS samples, 160 µL of sample was mixed with 160 µL of PBS.All samples were transferred to Lysing Matrix E barcoded tubes, and bead beating was performed in a FastPrep-96 bead beater (MP Biomedicals) (3 × 120 s, 1,800 rpm, 2 min incubation on ice between beatings).Community profiling was performed using 16S rRNA amplicon sequencing.V1-V3 16S rRNA gene regions were amplified using the 27F (AGAGTTTGATCCTGGCTCAG) (27) and 534R (ATTACCGCGGCTGC TGG) (28) primers, and the resulting amplicons were used in all the analyses.The V4 16S rRNA gene region was amplified using the 515F (GTGYCAGCMGCCGCGGTAA) (29) and 806R (GACTACNVGGGTWTCTAAT) (30) primers for comparison with the previous data set.Data were analyzed using R (version 3.5.2) (31), RStudio software (32), and visualized using ampvis2 (version 2.7.5) (33) and ggplot2 (34).The Köppen-Geiger climate zone classification (35) was utilized to categorize the countries participating in the global MiDAS project (18).Details about the classification of climate zones and the countries belonging to them can be found in Table S1.

Phylogenetic analysis based on the 16S rRNA gene, FISH probe design, and evaluation
Phylogenetic analysis of 16S rRNA gene sequences and design of FISH probes for the novel Chloroflexota were performed using the ARB software v.6.0.6 (36).A phylogenetic tree was calculated based on comparative analysis of aligned 16S rRNA gene sequen ces, retrieved from the MiDAS 4 database (18), using the maximum likelihood method and a 1,000-replicates bootstrap analysis.Coverage and specificity were evaluated and validated in silico with the MathFISH web tool for hybridization efficiencies of target and potentially weak non-target matches (37).When needed, unlabeled competitors and helper probes were designed.All probes were purchased from Biomers (Ulm, Germany), labeled with 6-carboxyfluorescein (6-FAM), indocarbocyanine (Cy3), or indodicarbocya nine (Cy5) fluorochromes.

Fluorescence in situ hybridization, quantitative FISH, and Raman microspec troscopy
FISH was performed as described by Daims et al. (38).The optimal formamide concentration for each novel FISH probe was determined after performing hybrid ization at different formamide concentrations in the range of 0%-70% (with 5% increments).The intensity of at least 50 cells was measured using ImageJ (39) software.Optimal hybridization conditions are described in Table S2.EUBmix (40,41) was used to target all bacteria, and NON-EUB (42) was used as a negative control for sequence-independent probe binding.Quantitative FISH (qFISH) biovolume fractions of individual genera were calculated as a percentage area of the total biovolume, hybridizing with both EUBmix probes and specific probes.In the case of a specific probe not overlapping with EUBmix, a mix of EUBmix and CFXmix (43,44), both labeled in Cy5, was used as a universal probe for total biomass coverage.qFISH analyses, performed using the Daime image analysis software (45), were based on 30 fields of view taken at 630× magnification.Microscopic analysis was performed with an Axioskop epifluorescence microscope (Carl Zeiss, Germany) equipped with a LEICA DFC7000 T CCD camera or a white-light laser confocal microscope (Leica TCS SP8 X).Raman microspectroscopy was applied in combination with FISH to look for the storage polymers polyphosphate (poly-P), glycogen, and polyhydroxyalkanoates (PHAs) as previously described (19).
Genomes were annotated as previously described (51).Briefly, the EnrichM v5.0 "annotate" pipeline (github.com/geronimp/enrichM) was used to annotate the protein sequences of the genomes against the EnrichM v10 database, which included the Kyoto Encyclopedia of Genes and Genomes (KEGG) (52) orthology (KO) number annotated Uniref100 database.Enrichm "classify" --cutoff 1 was used to determine the presence of 100% complete KEGG modules, such as for transporters and glycolysis (SData 2, available at 10.6084/m9.figshare.23586210,and SData 3, available at 10.6084/m9.figshare.23586219).Additionally, the MAGs were uploaded to the "MicroScope Microbial Genome Annotation & Analysis Platform" (MAGE) (53) for manual inspection and cross-validation of KO annotations found using EnrichM.The KEGG and Microcyc pathways annotations in MAGE were used for the investigation of specific pathways, such as cobalamin production and degradation of aromatic compounds.The Blast search option in MAGE was used for homolog searches of reductive dehalogenase genes.

Phylogenetic evaluation of the Chloroflexota members abundant in WWTPs
The phylogenetic diversity of novel and well-known Chloroflexota abundant in global WWTPs was evaluated using a comparison of genome-based and 16S rRNA genebased phylogenies to obtain a robust taxonomic assignment and to resolve potential discrepancies between 16S rRNA gene-and genome-based classification methods.The phylogenomic analysis (Fig. 1, File S1) revealed clustering into different novel families and genera, largely supported by 16S rRNA gene-based classification using the MiDAS 4 reference database (Fig. 2, File S1).
Despite the absence of genomic information, 16S rRNA gene-based phylogeny (Fig. 2), FISH probe design, and experimental analysis (see below) were possible for additional genera abundant in both global and Danish activated sludge samples.Therefore, we propose to rename the genera with placeholder names midas_g_391, midas_g_550, and midas_g_9648 as Ca.Amarofilum, Ca.Pachofilum, and Ca.Tricholinea, respectively.16S rRNA gene-based analysis using the MiDAS 4 database showed that midas_g_169 corresponded to Ca. Defluviifilum, as defined by Speirs et al. ( 2), and we suggest adopting this name in future studies.

Geographical distribution of Chloroflexota in global full-scale WWTPs
We analyzed the occurrence and diversity of both novel and well-known Chloroflexota genera abundant in global AS ecosystems, using data from the global MiDAS survey (18).On a global scale, the phylum Chloroflexota was the fourth most abundant phylum, making up 6.3% of the total reads (Fig. 3a), similar to previous observations in Denmark (1), Spain (55), and Australia (56).Among the most abundant genera worldwide (Fig. 3b), several well-known microorganisms appeared to be widespread, such as Ca.Villigracilis (1), Ca.Promineofilum (11), Ca.Sarcinithrix (1), Ca.Amarolinea (12), and Kouleothrix (57).
All other abundant Chloroflexota were mainly undescribed but potentially important for the process, such as the genus Ca.Defluviilinea (former UTCFX1), observed as part of the heterotrophic bacteria in anammox bioreactors (58,59), or Ca.Flexifilum, belonging to the former family A4b, first identified in nitrifying-denitrifying industrial WWTPs (60).Interestingly, few novel genera were only found locally in individual countries, such as the genus midas_g_75442, present only in Malaysia (Fig. 3b).
Ca. Defluviifilum and Ca.Caldilinea saccharophila (former Ca.Amarithrix), together with the novel Ca.Amarofilum and Ca.Pachofilum, were commonly found in high abundance in Danish WWTPs, ranging from 0.1 to 1.1% but reaching up to 8% in some samples (Fig. S1).
Examining global WWTP community composition enabled a deeper insight into factors affecting the occurrence of the different genera.We expected that the previously demonstrated slow-growth and facultative anaerobic metabolism of Chloroflexota species would favor their prevalence in long-sludge-age WWTPs and WWTPs with biological N and P removal (2).Ca.Villigracilis, Ca.Promineofilum, and Ca.Amarobacter occurred in higher abundance in WWTPs with a complex process design involving both N and P removal, indicating a potential role in these processes (Fig. 4a).The same genera also appeared to be influenced by the fraction of industrial wastewater (shown as the chemical oxygen demand fraction in the influent), preferring low to medium content (<50%) of industrial wastewater (Fig. 4b).This observation confirms previous findings, where filamentous Chloroflexota were detected in low abundance by FISH in industrial sludge (61).The differences in the Chloroflexota communities were more accentuated when considering the different climate zones (Fig. 4c), with the highest abundances of all genera observed in dry and temperate climates.Interestingly, some genera seemed to be specific to areas with hot temperatures, such as Ca.Defluviilinea and Ca.Tricholinea, dominant in dry and arid climates, or Ca.Brachythrix, which seemed predominant in countries with polar climates (Fig. 4c and d).

In situ characterization of Chloroflexota abundant in Danish and global WWTPs
We designed new FISH probes to target and characterize the abundant novel Chloroflexota genera in situ and re-evaluated the coverage and specificity of existing FISH probes (Fig. 2; Table 1).In silico evaluation of the widely-applied CFXmix (43, 44) using the MiDAS 4 database showed good coverage of the phylum in the AS ecosystem, and it is recommended to be used in combination with EUBmix (40,41) for better coverage of the Chloroflexota (Table 1).All but two of the genera investigated, Ca.Epilinea and Ca.Promineofilum, did hybridize with the EUBmix.In silico evaluation of previously published genus-level FISH targeting Ca.Amarolinea, Ca.Sarcinithrix, Ca.Promineofilum, and Ca.Villigracilis probes showed high specificity and hybridized with filaments of variable length and thickness (1, 3, 11, 61, 62) (Fig. 2; Table 1).The majority of the novel Chloroflexota appeared to have the same conventional morphology (Fig. 5), with filaments of different length and thickness often found in bundles inside the flocs or sometimes creating inter-floc bridges (Fig. 5; Table 1).Epiphytic bacteria were found on protruding filaments belonging to Ca. Flexifilum and other bacteria of the Flexifilaceae family, Ca.Trichofilum, Ca.Defluviifilum, Ca.Amarofilum, and Ca.Pachofilum (Fig. 5).Interestingly, the short filaments of Ca.Epilinea were found sometimes to be themselves attached to other Chloroflexota filaments (Fig. 5).The surface adhesion mechanism of these microorganisms is still unclear, but the presence of pili has been previously reported for several isolates (63)(64)(65), and these appendages could mediate the adhesion process (66).The two genera from the order Dehalococcoidia, Ca.Amarobacter and Ca.Amarobacillus, were small rods.We screened the MAGs for the presence of genes involved in the assembly of bacterial adhesins, such as flagella, pili, lectins, or functional amyloids, known to be involved in cell-to-cell interaction in activated sludge (67,68).The majority of the MAGs encoded potential for assembly of type IV pili and tight adherence protein (Tad) (SData 2), which have previously been shown to produce an adhesive matrix for cell-to-cell aggregation in bacteria from the Anaerolinae class (69).Few MAGs also encoded the potential for flagellar-mediated motility (SData 2), which is an atypical feature for Chloroflexota bacteria and has only been shown for the Tepidiforma isolates so far (70,71).
The FISH probes designed for the novel Chloroflexota genera were applied to Danish and, when possible, global activated sludge samples for FISH-based quantification (Table S3).Amplicon sequencing relative read abundances were in most cases very similar, while a few genera (Ca.Defluviifilum and Ca.Leptofilum) had slightly lower and some slightly higher (Ca.Amaribacillus, Ca.Amaribacter, and Ca.Epilinea) read abundances than qFISH results.These differences are likely due to variations in cell size, extraction efficiency, and 16S rRNA gene copy number variation (73).The 16S rRNA gene copy numbers in the MAGs analyzed in this study varied from one to five (SData 1), which could lead to overestimation of some genera when using amplicon-based quantification (73).Primers are also known to introduce bias in amplicon sequencing (73).Therefore, the composition of the Chloroflexota community was compared using two commonly applied primer sets (the V1-V3 and V4 regions of the 16S rRNA gene sequences) (Fig. S2).The overall relative abundance of the phylum Chloroflexota was comparable, but significant differences appeared in the relative abundances of specific genera (Fig. S2).The V4 data set showed lower relative abundances of some genera, such as Ca.Promineofilum and Ca.Leptofilum, and the almost complete disappearance of Ca.Villigracilis and Ca.Defluviilinea, which were likely not targeted by the V4 primer set.These findings and the similarity of the abundances calculated by qFISH and V1-V3 amplicon sequencing confirmed that the V1-V3 primer set is more suited to encompass the diversity of the phylum Chloroflexota in AS systems.This is of particular importance if the potential effect on the settling properties of these filamentous bacteria is evaluated using amplicon sequencing data.
To investigate the in situ physiology of the novel Chloroflexota genera, we performed Raman microspectroscopy in combination with the new FISH probes.This approach allows the detection of general cellular components, such as nucleic acids, membrane lipids, or proteins (74), as well as storage polymers important for the physiology of microorganisms involved in nitrogen or phosphorus removal in activated sludge (19,51).Most of the Chloroflexota genera showed the presence of peaks characterizing common biological components, such as phenylalanine, nucleic acids, and lipids, as well as a peak for glycogen, which most likely serves as a storage compound to survive periods with low energy sources (Fig. S3).The presence of glycogen, which appears to be a conserved feature of the phylum, confirms the potential for glycogen accumulation proposed by our metabolic predictions (see below) and studies of other known Chloroflexota (11,12,63).Although granules of PHAs and poly-P have been identified previously in Chloroflexota isolates (63), these storage polymers were not detected in situ.

Metabolic potential of the abundant Chloroflexota genera
Functional analysis of the 53 Chloroflexota MAGs revealed similar metabolisms to previously published models (11,12).They exhibited a very versatile metabolism, revealing heterotrophic lifestyles with possible involvement in the degradation of complex organic compounds and the utilization of a wide selection of different sugars and amino acids as carbon sources (Fig. 6; SData 2 and 3).Manual inspection of the MAGs confirmed that all the MAGs encoded pathways for full central carbon processing through glycolysis, the pentose phosphate pathway, and the TCA cycle (Fig. 6; SData 2 and 3).
The acetate transporter gene (actP) was encoded only in two MAGs, belonging to the genera Kouleothrix and Ca.Brachythrix.The uptake of acetate and short-chain fatty acids has also not been experimentally confirmed in some of the abundant genera (1,61,75).Nonetheless, the presence of the gene acs was widespread in the MAGs, suggesting the potential use of acetate or fatty acids as carbon sources by these organisms, maybe deriving from internal pools (Fig. 6; SData 2 and 3).The glyoxylate cycle was complete in part of the genomes (15/53), but only one MAG belonged to Ca. Brachythrix, which encoded the full potential for aerobic pyruvate oxidation to acetyl-CoA (aceE and aceF) (Fig. 6; SData 2 and 3).Additionally, genes encoding for biosynthesis (fadD) or beta-oxida tion of fatty acids (acd, fadJ, and fadA) were widespread in the genera and identified in nearly all MAGs (Fig. 6; SData 2 and 3).Interestingly, oxidation of long-chain fatty acids appeared to be one of the preferred energy sources for the bacteria belonging to the two non-filamentous genera in the order Dehalococcoidia, Ca.Amarobacter and Ca.Amarobacillus.They had a high copy number of genes involved in this pathway (fadD, acd, paaF, fadJ, fadA), comparable to those found in the well-known lipid users in AS systems Ca.Microthrix (76), indicating their potential specialization as lipid consumers (Fig. 6; SData 2 and 3).Acetate use as a carbon source is common in Dehalococcoidia isolates (71,77), and beta-oxidation has also been proposed as a potential metabolic route to obtain carbon and reducing equivalents by Dehaloccoidia found in marine sediments.Further analysis using transcriptomics and/or proteomics could help clarify the activity and expression levels of these enzymes (78).Additionally, we screened the Dehalococcoidia MAGs for the presence of reductive dehalogenase genes, a typical feature of some members of this class.However, no homologs of any component of the reductive dehalogenase complexes were identified, as already observed for the closely related Tepidiformales isolates (70).
Anaerobic sugar uptake has previously been demonstrated in situ in several Chloroflexota genera, and the potential for fermentation of substrates to acetoin or lactate was indicated in the genomes of Ca.Amarolinea and Ca.Promineofilum (1,11,12,61).The potential for fermentation was encoded by most of the MAGs (ALDH, adh, and adhE), with ethanol as a possible by-product (Fig. 6; SData 2 and 3).Part of the MAGs (15/53) encoded genes for formate dehydrogenase (fdh and/or fdoGHI), potentially used to reduce formate produced during anaerobic fermentation, as suggested for Ca.Promineofilum breve (11).
Alternative electron acceptors under anaerobic conditions included nitrate, with potential dissimilatory nitrate reduction to nitrite (narGHI) identified in the four MAGs associated with the genus Ca.Amarolinea, as previously reported for Ca.Amarolinea aalborgensis (12), and in the two MAGs representing the new genus Ca.Defluviilinea (Fig. 6; SData 2 and 3).Genes for nitrite reduction to nitric oxide (nirK) were present in 10/53 MAGs, while no MAGs encoded potential for nitric oxide reduction to nitrous oxide (norBC).Twenty-five MAGs encoded genes (nosZ) for potential reduction of the latter to gaseous nitrogen (Fig. 6; SData 2 and 3).The potential for dissimilatory nitrite reduction to ammonia (nrfAH) was widespread across the MAGs, similar to previous findings for Ca.Amarolinea aalborgensis (12) or Ca.Promineofilum breve (11), while no potential for nitrification was detected (Fig. 6; SData 2 and 3).The potential for sulfate reduction to H 2 S through the assimilatory pathway (sat, cysH, and sir) was also widespread across the MAGs (Fig. 6; SData 2 and 3).Some Chloroflexota bacteria are described as being involved in biogeochemical sulfur cycling (78)(79)(80).
Many bacteria that live under alternating oxic-anoxic conditions produce storage compounds, such as poly-P, glycogen, and PHAs, which can be used in dynamic systems when environmental carbon or energy reserves are scarce.Examples are polyphosphateaccumulating organisms involved in biological P removal in activated sludge (81).Genes indicating the potential for polyphosphate accumulation, such as the phosphate transporters (pit, pstSCAB) and the polyphosphate kinase (ppk), were widespread across the MAGs (Fig. 6; SData 2 and 3).However, this storage compound was not detected in any Chloroflexota genus in situ.Only MAGs belonging to Ca. Epilinea and Ca.Avedoeria encoded the full potential for PHA accumulation (phaABC) (Fig. 6; SData 2 and 3), but this intracellular polymer was not detected experimentally.Previous genome studies indicated the presence of glycogen as a storage compound in Ca.Amarolinea aalborgen sis and Ca.Promineofilum breve (11,12).The potential for glycogen biosynthesis was confirmed by the identification of the genes involved in the pathway (glgABC, galU, glgP, and glgY), as well as for its degradation (malQ) (Fig. 6; SData 2 and 3), and experimentally validated in most of the genera.
The potential for carbon fixation through the enzyme RuBisCO and the Calvin-Ben son-Bassham cycle was also present in several MAGs (9/53), as previously observed for Kouleothrix and other Chloroflexota genera (2) (SData 2 and 3).Interestingly, the potential for bacteriochlorophyll biosynthesis and photorespiration was encoded by Ca.Ribeiella danica and Ca.Kouleothrix ribensis, similar to the isolates of the class Chloroflexia Roseiflexus castenholzii and Chloroflexus aurantiacus (SData 2 and 3) (23).As for the latter, Ca.Ribeiella and Kouleothrix present a fused form of the genes encoding for type II photosystem reaction centers (pufLM), recognized by manual inspection of the genomes (23).However, it is unclear if these pathways play a role in the activated sludge environment, where light is not easily accessible and the Calvin-Benson-Bassham pathway would be an energetically expensive alternative to the available organic carbon.Further analysis of the activity and expression level of these enzymes could help to clarify their role in situ.
Chloroflexota bacteria are also known to play a role in the degradation of recalcitrant organic matter, for example, in deep oceans (70,82,83).Therefore, we screened the MAGs for potential catabolism of aromatic compounds (SData 2).Although a complete pathway could not be predicted, the MAGs harbored several genes encoding enzymes involved in the degradation of different compounds, such as catechol or toluene, which could indicate the potential for aromatic biodegradation in consortia with other microorganisms (70).A potential syntrophic metabolism could also be hypothesized for essential enzyme cofactors, such as cobalamin and other vitamins of the B group (84)(85)(86).None of the MAGs harbored a completed pathway for de novo biosynthesis of cobalamin, but a few encoded some of the enzymes involved in the first reactions of the process, as well as potential for salvage and remodeling from cobyrinate (SData 2).

Ecological significance of Chloroflexota in activated sludge
This study provides a broad overview of the Chloroflexota abundant in AS WWTPs and contributes to recognizing their role in these systems.The phylum Chloroflexota encompasses a variety of different metabolisms, ranging from haloalkane reducers (87)(88)(89) to anoxygenic photosynthetic microorganisms (23,64,90,91), sponge holobionts (92-94), and extremophiles (95), all with roles in carbon, sulfur, and nitrogen cycling.It is therefore unsurprising that members of the Chloroflexota are present in the activated sludge ecosystem worldwide, although most lineages remain undescribed.
Our comprehensive approach, which includes the utilization of genome-and 16S rRNA gene-resolved phylogeny, allowed the identification of 4 novel families, 14 novel genera, and 29 novel species, most of which are widely distributed across the continents and are seemingly influenced by factors such as climate zones, temperature, and WWTP process design.However, the variation in presence and abundance between countries and even between plants in the same country may also be largely influenced by the immigration of microbial populations and seasonality (96,97).
Generally, filamentous bacteria act as the backbone of activated sludge flocs, to which floc-forming bacteria attach and grow, typically as microcolonies.However, this beneficial role can change if the filaments are proliferating when the environmental conditions promote their growth, resulting in foaming, bulking, and poor sludge-water separation (98).The Chloroflexota have often been associated with bulking episodes; consequently, the morphological characterization of the novel genera is essential (2,4).Most of the novel genera are, however, characterized by thin and short trichomes; some were short and attached to other filaments, so they did not form filamentous bridging, which prevents the flocs from clustering, as is known from Ca. Amarolinea and a few others, stressing that most are likely good for the floc formation.Additionally, most of the MAGs encoded the potential for the production of an adhesive matrix for cell-tocell aggregation, which supports their beneficial role in floc formation.All filamentous members of Chloroflexota belonged to the orders Anaerolineae and Chloroflexia, while the members from Dehalococcoidia, Ca.Amarobacter and Ca.Amarobacillus, were rod-shaped, showing that not all AS Chloroflexota are filamentous.
Confirming previous findings, Chloroflexota abundant in AS likely have a heterotro phic and facultative anaerobic lifestyle, which may explain their higher abundance in plants with more complex designs with oxic and anoxic conditions (1,2,11,99).The different genera have versatile metabolisms, but with a seemingly important role in carbon cycling, with the potential ability to use various sugars and amino acids, but in some cases also lipids and acetate.They are assumed to have high hydrolytic activity, and the production of exo-enzymes for polysaccharide degradation has been previously shown in situ, which suggests their importance in degrading the exopolymeric matrix rather than soluble substrates (99,100).Fermentation seems to be widespread across all the genera and likely sustains these organisms during anoxic conditions, as shown for some of the related isolates (7,8,10,101).Despite being fermentative and facultative anaerobic microorganisms, Chloroflexota abundant in AS systems generally appear to die off when the biomass is fed to anaerobic digesters, likely due to the substantial differences in environmental conditions, such as higher temperatures or salinity, while different Chloroflexota filaments seem to be predominant in the anaerobic digester environment (102)(103)(104).Their involvement in nitrogen, sulfur, and phosphorus cycling seems to be more limited.However, these organisms still play an important role in nutrient cycling as well as in the degradation of aromatic compounds, if considered active members of the bacterial consortia typical of the activated sludge floc, producing essential substrates that can be used by other microorganisms.Our study provides the foundation for future in-depth characterization of their physiology and relationships with other microorganisms.Following this, we believe that gene expression and regulation studies are the next important steps forward in understanding the ecology of these process-critical bacteria.

Formal taxonomic proposal
Etymologies and protologs for the novel proposed species are provided in File S2.

FIG 1 FIG 2
FIG 1Phylogenetic genome tree of abundant Chloroflexota representatives.The maximum likelihood genome tree was created from the concatenated alignment of 120 single-copy marker gene proteins trimmed to 5,000 amino acids using the WAG + G model and 1,000× UFBoot bootstrapping iterations.Bootstrap support >95% is shown by the solid black circles.Three Cyanobacterota genomes (NCBI accession numbers: GCA_000317655, GCA_002813895, GCA_003566215) were used as an outgroup to root the tree.For NCBI GenBank genome accession numbers, see SData 1. MAGs belonging to novel families are marked with colored boxes, while MAGs clustering with validly published families are marked with gray boxes.MAGs from species representatives are used to construct the tree and are indicated between brackets, as well as the number of available MAGs for each lineage.The scale bar represents substitutions per amino acid base.

FIG 2 (FIG 3
FIG 2 (Continued)of novel species.The alignment used for the tree applied a 20% conservational filter to remove hypervariable positions, giving 1159 aligned positions.Coverage of the existing and designed FISH probes in this study is indicated with black brackets and is based on the MiDAS 4 database(18).Bootstrap values from 1,000 re-samplings are indicated for branches with >70% (gray dot) and >90% (black) support.Species of the phylum Cyanobacteria were used as the outgroup.The scale bar represents substitutions per nucleotide base.

a 12 FIG 5
FIG 5 FISH micrographs of novel Chloroflexota genera in full-scale activated sludge.All bacteria were targeted with EUBmix (blue), and in some micrographs, CFXmix (green) was also applied to target most bacteria belonging to the Chloroflexota phylum.Specific probe targets: (a) Ca.Epilinea (yellow); (b) Ca. (Continued on next page)

TABLE 1
Summary table of morphology and ecophysiology of known Chloroflexota genera (Continued)