Genomics, Exometabolomics, and Metabolic Probing Reveal Conserved Proteolytic Metabolism of Thermoflexus hugenholtzii and Three Candidate Species From China and Japan

Thermoflexus hugenholtzii JAD2T, the only cultured representative of the Chloroflexota order Thermoflexales, is abundant in Great Boiling Spring (GBS), NV, United States, and close relatives inhabit geothermal systems globally. However, no defined medium exists for T. hugenholtzii JAD2T and no single carbon source is known to support its growth, leaving key knowledge gaps in its metabolism and nutritional needs. Here, we report comparative genomic analysis of the draft genome of T. hugenholtzii JAD2T and eight closely related metagenome-assembled genomes (MAGs) from geothermal sites in China, Japan, and the United States, representing “Candidatus Thermoflexus japonica,” “Candidatus Thermoflexus tengchongensis,” and “Candidatus Thermoflexus sinensis.” Genomics was integrated with targeted exometabolomics and 13C metabolic probing of T. hugenholtzii. The Thermoflexus genomes each code for complete central carbon metabolic pathways and an unusually high abundance and diversity of peptidases, particularly Metallo- and Serine peptidase families, along with ABC transporters for peptides and some amino acids. The T. hugenholtzii JAD2T exometabolome provided evidence of extracellular proteolytic activity based on the accumulation of free amino acids. However, several neutral and polar amino acids appear not to be utilized, based on their accumulation in the medium and the lack of annotated transporters. Adenine and adenosine were scavenged, and thymine and nicotinic acid were released, suggesting interdependency with other organisms in situ. Metabolic probing of T. hugenholtzii JAD2T using 13C-labeled compounds provided evidence of oxidation of glucose, pyruvate, cysteine, and citrate, and functioning glycolytic, tricarboxylic acid (TCA), and oxidative pentose-phosphate pathways (PPPs). However, differential use of position-specific 13C-labeled compounds showed that glycolysis and the TCA cycle were uncoupled. Thus, despite the high abundance of Thermoflexus in sediments of some geothermal systems, they appear to be highly focused on chemoorganotrophy, particularly protein degradation, and may interact extensively with other microorganisms in situ.


Additional information for the cultivation of T. hugenholtzii JADT for 13 C-labeled substrate metabolic probing
To accommodate multiple headspace gas samples 200 mL of GBS salts medium (1.0g/L peptone), prepared anaerobically, was distributed to 500 mL Wheaton bottles and pressurized with 1 atm of overpressure of N2.
Culture bottles were anaerobically prepared at the University of Nevada, Las Vegas, then transported to Northern Arizona University where they were vented to bring the bottles to atmospheric pressure at NAU (0.8 atmospheres). Peptone, phosphate buffer, and vitamin solutions were added anaerobically just before inoculation. Filter-sterilized air was added to each bottle for a final concentration of 1% O2.
To compare the 13 CO2 production rate from T. hugenholtzii cultures with that of sterile controls, the ideal gas law was used to convert the volume of CO2 present in the incubations from various processes and additions, to moles of CO2 present from various processes and additions.

PV=nRT
Where n is the number of micromoles of gas; P is pressure in atm; V is volume of the gas in L; T is temperature (°K); and R is the gas constant (0.08205746 (atm*L) / (moles*K)). We also used the atom fraction equation for 13 C 13 C atom fraction = ((δ13C/1000+1)*0.011237)/((δ13C/1000+1)*0.011237+1) to calculate 13 C atom fraction (x( 13 C)) values from δ 13 C values determined by running samples on the Picarro. Using the calculated atom fraction values, we applied a mass balance equation for isotope mixing to determine the contribution of 13 C-CO2 from different processes and additions in the cultures and abiotic controls.
x( 13 C)tVt = x( 13 C)1V1 + x( 13 C)2V2 … Taken together, using cultures with 13 C-labeled substrate additions, cultures with no 13 C-substrate additions, cultures used for total CO2 production rates, and abiotic 13 C-CO2 controls, we were able to attribute 13 C-CO2 production to either T. hugenholtzii metabolism or abiotic processes.

Inferred amino acid interconversion and biosynthetic capability
Ornithine (M00763) and lysine biosynthesis (M00031) appeared possible (Table S4), yet both biosynthesis pathways were missing genes connecting them to the TCA cycle (Table S3).
Histidine degradation to glutamate through N-formiminoglutamate appears possible (M00045) ( Table S4). The absence of one gene coding for a homoserine acetyltransferase (EC2.3.1.31) may prevent the synthesis of homocysteine and methionine from aspartate or homoserine (Table S3).

Accumulation and degradation of other compounds
Adenine and adenosine were substrates for T. hugenholtzii JAD2 T . This observation was supported by the presence of nucleoside transporters and complete pathways for the degradation of these compounds. Biosynthetic pathways for this compound were also complete in T.
hugenholtzii and all MAGs, indicating Thermoflexus may synthesize them de novo when necessary. In contrast, thymine accumulated in the medium, which seems paradoxical given that thymine pathways are incomplete. This was unexpected given that biosynthesis pathways for this compound are incomplete in T. hugenholtzii JAD2 T and all Thermoflexus MAGs.
A slight increase in riboflavin in the presence of growth was observed in the exometabolomic data despite a gene coding for an ECF-type riboflavin transporter, S component being found. More significant is the thermal degradation of riboflavin, which is thought to be stable at higher temperatures over short periods of time. The long incubation times necessary to grow T. hugenholtzii JAD2 T provide ample opportunity for thermal degradation and production of products, providing deviations in the chemical makeup of the medium over time solely due to temperature. This, along with the demonstrated production and degradation of other compounds due to temperature, highlights the importance of running sterile controls in parallel with cultures during exometabolomic analyses but also illuminates potential challenges when growing thermophiles with long doubling times, due to chemical changes in the medium strictly from temperature.

An unidentified GBS organic extract stimulated Thermoflexus growth
The addition of an organic extract derived from Great Boiling Spring, the source of T.
hugenholtzii JAD2 T , significantly enhanced growth, suggesting that organic extracts commonly used for microbiological media are limited in some beneficial nutrients. A thiamine transport system was found in T. hugenholtzii JAD2 T , and an ascorbate phosphotransferase system was conserved across the genus, which is consistent with our observations that T. hugenholtzii JAD2 T is stimulated by high concentrations of vitamins (see below). Complete carbohydrate metabolic pathways and a variety of transporters for oligo-and monosaccharides suggest these compounds may be utilized by T. hugenholtzii JAD2 T and other Thermoflexus sp., yet these types of compounds are not capable of serving as sole carbon and energy sources for growth (Dodsworth et al., 2014). These data suggest that yet-to-be-determined key nutrient limitations may contribute to the low cell density observed in T. hugenholtzii JAD2 T cultures when grown on peptide-based complex media as a carbon and energy source.

Preparation of Hot Spring Organic Extract
Bulk spring water was collected from GBS by tangential flow filtration and stored in sterile plastic carboys at room temperature until processed. GBS water was flowed through Diaion HP-20 (Supelco Analytical) at ~150 mL per min by a peristaltic pump. Approximately 40 L of hot spring water was passed through the resin. The resin was then rinsed with nanopure water to remove residual salts, and 60% acetonitrile was used to collect the organic extract. The organic extract was dried in 2 mL Eppendorf microtubes using a Speed Vac SC100 (Savant Instruments, Inc., Farmingdale, NY, Model: RT100A) on medium drying rate. The dried extract was solubilized in anaerobic nanopure water in an anaerobic chamber at a concentration of 800 µg/mL. The organic extract was filter-sterilized using a 0.2 µm filter (VWR, N.A. PN: 28145-501) and was stored anaerobically at 4°C in the dark. Growth experiments using the organic extract were carried out as described above for exometabolomics but with the addition of 32 µg/mL (final concentration) to the base medium, these results were not analyzed in this manuscript. Attempts at chemically describing the organic extract using HPLC-MS/MS failed.

Cell counts for exometabolomic experiments and hot spring organic extract additions.
Replicate 1 for exometabolomics demonstrated much higher growth than the other replicates within the same treatment.