M. tundripaludum SV96 and M. rosea SV97 isolated from Arctic wetland soils in Norway (Wartiainen et al., 2006a,b) were used in this study. Methanotrophs were cultivated on nitrate mineral salt (NMS) medium (DSMZ medium 921; initial pH ∼6.80) with the addition of 1 mM lanthanum chloride (LaCl₃). The pre-inoculum was grown in 120-ml serum bottles containing 10-ml NMS medium, 20% CH₄, and 80% air in headspace and incubated statically at 20°C. All experiments were conducted under sterile conditions, and the serum bottles used for methanotroph cultivation were sealed with butyl rubber stoppers and capped with aluminum crimps.

Wild-type and engineered A. baylyi ADP1 strains (Luo et al., 2019) carrying the plasmid pBAV1C-‘tesA-undA (A. baylyi ADP1 ‘tesA-undA) were used to evaluate the growth on the methanotroph spent media. For the pre-inoculum, A. baylyi ADP1 cells were inoculated in 10-ml culture tubes containing LB medium (5 g L^–1 yeast extract, 10 g L^–1 tryptone, and 5 g L^–1 NaCl) supplemented with 0.5% glucose and 25 μg mL^–1 chloramphenicol (for A. baylyi ADP1 ‘tesA-undA). The inoculated tubes were aerobically grown overnight at 30°C and 300 rpm. For 1-undecene synthesis, A. baylyi ADP1 ‘tesA-undA was induced with 0.5 mM cyclohexanone (Luo et al., 2019).

Batch tests were performed in 120-ml airtight serum bottles with a working volume of 15 ml of NMS medium. The tests were conducted in triplicates. The precultures of M. tundripaludum SV96 and M. rosea SV97 were inoculated at an initial optical density at 600 nm (OD_600nm) of 0.02, and the growth, CH₄ utilization, and organic acid profiles were monitored every 2 days for 14 days at 20°C under static conditions. Both methanotrophs were tested under the three different gas supplementation schemes (Figure 2). On day 0, the bottles were filled with 20% CH₄ and 80% air into the headspace, accounting for the initial O₂/CH₄ molar ratio of ∼1.2, and incubated for 7 days. After the CH₄ and O₂ concentrations were depleted on day 7, the batch bottles were supplemented with three gas compositions into the headspace: test I: CH₄ + air (20% CH₄ and 80% air), test II: only CH₄ (20% CH₄ and 80% N₂), and test III: only air (20% air and 80% N₂). Bottles containing only NMS without bacterial cells were used as controls. The bottles were incubated for 7 days (days 8–14). Biomass growth (OD_600mm), gas composition in the headspace, and organic acid accumulation in the liquid medium were monitored every 2 days.

Prior to testing with an engineered A. baylyi ADP1, the capacity of A. baylyi ADP1 cells to utilize the spent media of methanotrophs was tested. To obtain methanotroph spent media, M. tundripaludum SV96 and M. rosea SV97 were cultivated in 500-ml airtight bottles with 60-ml NMS medium, as described previously (see section “Evaluation of Organic Acid Production by M. tundripaludum SV96 and M. rosea SV97”). To maximize the organic acid concentration, the bottle headspace was initially filled with 20% CH₄ and 80% air and cultivated for 28 days under optimal conditions for M. tundripaludum SV96 and M. rosea SV97. The gaseous composition in the headspace and organic acid concentrations in the liquid medium were monitored twice a week.

After 28 days, the cultivation of methanotrophs was stopped, and spent media were used for A. baylyi ADP1 cultivation. The spent media was collected by centrifugation at 12,000 rpm for 10 min. The 50-ml supernatant of both methanotrophs was transferred to 250-ml sterile flasks and used for A. baylyi ADP1 cultivation. All tests were conducted in duplicate and incubated at 30°C and 300 rpm. After 4 h of cultivation, the liquid culture was collected to monitor the cell growth, organic acid concentration, and wax ester production. The spent media of methanotrophs without A. baylyi ADP1 were used as a contamination control (control 1), and A. baylyi ADP1 cells in NMS (fresh medium) were used to determine their background growth (control 2).

To obtain the methanotroph spent media, both M. tundripaludum SV96 and M. rosea SV97 were cultivated in 500-ml airtight bottles with 60-ml NMS medium, as described in section “Cultivation of Wild-Type A. baylyi ADP1 in Spent Media of Different Methanotrophs.” After 28 days, the spent media were collected and directly used for cultivation to identify any potential inhibitory effects on the growth of A. baylyi ADP1 cells imparted by the spent media supernatant (see section “Cultivation of Wild-Type A. baylyi ADP1 in Spent Media of Different Methanotrophs”). Thus, the original spent media, containing the methanotrophs and organic acids, devoid of any nutrient supplementation were employed as the growth medium for A. baylyi ADP1 cultivations in this experiment. The tests were conducted in quadruplicate in a sealed 20-ml glass tube containing 5 ml of the spent medium. Subsequently, A. baylyi ADP1 ‘tesA-undA cells (initial OD_600nm, 0.02) were added to vials. After 1 h of incubation at 30°C and 300 rpm, 0.5 mM of cyclohexanone was added to the vials to induce 1-undecene synthesis and the vials were further incubated for 23 h at 30°C and 300 rpm. Liquid culture samples from M. tundripaludum SV96 and M. rosea SV97 cultivation without A. baylyi ADP1 ‘tesA-undA were included as experimental controls. The growth of A. baylyi ADP1 ‘tesA-undA was estimated based on the difference in OD_600nm between the tests and controls after incubation for 24 h (ΔOD_600nm).

Liquid samples were filtered through a 0.2 μm membrane (Chromafil^® Xtra PET-20/25, Macherey-Nagel, Düren, Germany) prior to liquid metabolite analysis using a Shimadzu high-performance liquid chromatograph equipped with a Rezex RHM-Monosaccharide H⁺ column (Phenomenex, Torrance, CA, United States), as described by Okonkwo et al. (2018). The gas samples in the headspace (CH₄, CO₂, O₂, and N₂) were measured using a Shimadzu gas chromatograph GC-2014 equipped with a thermal conductivity detector and a Carboxen-1000 60/80 column (Agilent Technologies, Santa Clara, CA, United States). The oven temperature was held at 35°C for 3.75 min and then increased with the rate of 30°C min^–1 until 150°C for 3 min. The injector and detector were 155 and 160°C, respectively. Helium was used as the carrier gas at 30 ml min^–1.

1-undecene in the headspace was detected using solid phase microextraction gas chromatography–mass spectrometry, as previously described by Luo et al. (2019). Compounds were identified using the NIST/EPA/NIH Mass Spectral Library (NIST 05). Bacterial growth was measured as OD_600nm using an Ultrospec 500 pro spectrophotometer (Amersham Biosciences, United Kingdom) and as cell dry weight (CDW) using the gravimetric method. The conversion factor between CDW and OD_600nm obtained from the experimental measurements was 0.2914 g L^–1 OD^–1 (R² = 0.9948) and 0.3015 g L^–1 OD^–1 (R² = 0.9892) for M. tundripaludum SV96 and M. rosea SV97, respectively.

Statistical analysis was performed using Minitab 16.0 (United States). Significant differences in the obtained data sets (e.g., growth of the tested strains, gas concentrations and utilization, and concentrations and yields of the products produced by each strain) within the varied treatments were analyzed using one-way analysis of variance with Tukey’s multiple comparison tests at the 95% confidence interval, where P-value ≤ 0.05 was considered statistically significant.

Both M. tundripaludum SV96 and M. rosea SV97 used CH₄ as the sole carbon and energy source and O₂ as an electron acceptor. The biomass production of M. tundripaludum SV96 was greater than that of M. rosea SV97 in all tests (Figure 3). Highest biomass production of both methanotrophs was observed in the supplementation of both CH₄ + air (test I) on day 14 at concentrations of 0.60 ± 0.03 and 0.48 ± 0.05 g CDW L^–1 for M. tundripaludum SV96 and M. rosea SV97, respectively. In all tests, the growth yield of M. rosea SV97 was lower than that of M. tundripaludum SV96 (P < 0.05) because of the typical carbon assimilation pathway of Type I methanotrophs (RuMP), showing more efficient channeling of carbon from CH₄ (C-CH₄) to biomass than that of Type II methanotrophs (serine cycle) (Kalyuzhnaya et al., 2015).

The spent media of both methanotrophs contained similar organic acid compounds, including formate, acetate, and succinate, whereas malate was only present in the spent medium of M. tundripaludum SV96. However, the results suggest that M. tundripaludum SV96 is more efficient and promising for CH₄ conversion into organic acids than M. rosea SV97. Furthermore, the concentration and production yield of total organic acids present in the spent medium of M. tundripaludum SV96 were higher than those of M. rosea SV97 in all gas supplementation tests (P < 0.05) (Figure 3). The efficiency of C-CH₄ conversion into organic acids was 4.8–7.0% and 0.7–1.8% (of consumed C-CH₄) for M. tundripaludum SV96 and M. rosea SV97, respectively (Supplementary Table 1). This could be due to the different carbon assimilation pathways between the RuMP and serine pathways in Type I and Type II methanotrophs. The RuMP pathway efficiently links to the glycolytic pathway, where pyruvate is converted to organic acids, whereas the serine cycle of Type II methanotrophs has a high flux through acetyl-CoA to yield an intracellular storage compound such as polyhydroxybutyrate (PHB) under nutrient-deficient conditions (Kalyuzhnaya et al., 2015; Nguyen et al., 2021). PHB is also a native product of M. rosea SV97 (Wartiainen et al., 2006b). Regarding the production of organic acids in Type II methanotrophs, which has not been widely reported, Costa et al. (2000) observed methanotroph-driven CH₄ conversion into acetate, which was subsequently consumed by heterotrophs in a denitrification bioreactor using CH₄ as an electron donor. Vecherskaya et al. (2009) also observed succinate, acetate, and 2,3-butanediol excreted in the culture medium of the Type II methanotroph Methylocystis parvus during CH₄ oxidation under microaerobic and anaerobic conditions. Based on the ¹³C analysis in their study, these organic acids were likely from PHB degradation under microaerobic conditions (5–10% O₂) (Vecherskaya et al., 2009).

During the 14-day incubation period, organic acid concentrations gradually accumulated in the liquid culture of both methanotrophs in all gas supplementation tests, corresponding with the increase in their biomass production (Figure 3). On days 10–14 onward, three gas supplementation tests induced three different headspace conditions: microaerobic (O₂-limited), anaerobic, and aerobic conditions. These conditions resulted in O₂/CH₄ molar ratios in the headspace of 0.2–0.3, <0.01, and 2–10 for the addition of CH₄ + air (test I), CH₄ (test II), and air (test III), respectively (Figure 4). Interestingly, the three gas supplementation tests in our study showed similar total organic acid production yields (Figures 5A,B). Lee et al. (2021) compared the production of organic acids in M. capsulatus Bath strain under various conditions, including aerobic, oxygen-limited, sulfur-limited, and nitrogen-limited conditions. The authors observed acetate production in all studied conditions, and nitrate-nitrogen limitation induced the highest acetate production (approximately 1.9 mmol-acetate g^–1 CDW). Furthermore, previous studies on organic acid production by type I methanotrophs (Lee et al., 2021; Takeuchi and Yoshioka, 2021) have reported that acetate and formate are also produced under aerobic conditions but at lower concentrations than under O₂-limited conditions. In our study, however, the three gas supplementation tests likely varied in the distribution of organic acids contained in the spent medium, particularly in M. tundripaludum SV96, a Type I methanotroph. Regarding organic acid production yields per consumed C-CH₄, both formate (3.5%) and acetate (2.5%) were dominant in the test with CH₄ + air supplementation (test I), whereas only acetate (5.1%) was dominant in the test with only air supplementation (test III) (Figures 5C,D). These results suggest that it is possible to target dominant organic acid compounds by the controlled feeding of CH₄ and O₂. The excretion of high formate concentration by M. tundripaludum SV96 during CH₄ + air supplementation (test I) might be due to imbalanced growth during O₂-limited conditions, which was previously reported in type I methanotrophs by Gilman et al. (2017). This also corresponded with the pH reduction observed in the test with CH₄ + air supplementation (test I) (Supplementary Figure 1).

Compared to organic acid production in different Type I methanotroph species under O₂-limited conditions, the M. tundripaludum SV96 used in our study showed comparable and higher organic acid yields per biomass production (CDW), that is, formate, acetate, and succinate (Table 1). These studies cultivated methanotrophs with initial headspace CH₄ and O₂ concentrations of 20 and 5%, respectively, to evaluate the organic acid production during microaerobic CH₄ oxidation (Kalyuzhnaya et al., 2013; Gilman et al., 2015, 2017). However, in our study, M. tundripaludum SV96 was cultivated aerobically before allowing O₂-limiting condition to occur. Further studies are required to determine whether prior cultivation under aerobic conditions enhances organic acid production under microaerobic conditions. In addition, the organic acid yields of M. tundripaludum SV96 obtained in our study (4.8–7.0% of consumed C-CH₄) were higher than those of M. capsulatus Bath (<5% of consumed CH₄) (Lee et al., 2021), whereas M. alcaliphilum 20Z enabled the convert 40–50% of the consumed CH₄ into mostly acetate and formate under O₂-limited conditions (20% CH₄:5% O₂) (Kalyuzhnaya et al., 2013).

Wild-type A. baylyi ADP1 cells grew in methanotroph spent media containing formate, acetate, and succinate (and malate for M. tundripaludum SV96). During the 4-h incubation period, A. baylyi ADP1 cells incubated in the spent media from M. tundripaludum SV96 and M. rosea SV97 grew to an OD_600nm of 0.14 ± 0.01 and 0.12 ± 0.03, respectively (Supplementary Figure 2). In this test, A. baylyi ADP1 cells likely utilized acetate, succinate, and malate as carbon sources for biomass assimilation (Supplementary Figure 2). However, formate is not utilized by cells as a carbon source, but it is used to maintain the cellular redox balance (Kannisto et al., 2015). The presence of fatty acid fractions derived from A. baylyi ADP1 biomass on thin layer chromatography plates also indicated the growth of A. baylyi ADP1 (Supplementary Figure 3). Typically, A. baylyi ADP1 produces wax esters as carbon storage compounds that are often associated with growth (Mangayil et al., 2019). However, they were not detected in this study (Supplementary Figure 3). This may have been due to the rapid degradation of accumulated wax ester under carbon-limiting conditions. Under such conditions, A. baylyi ADP1 cells utilize stored carbon to maintain cellular activities (Fixter et al., 1986; Salmela et al., 2018). The use of the spent media of methanotrophs for wax production is a possible direction for further study, but this approach likely requires a higher quantity of organic acids.

After confirming the growth of wild-type A. baylyi ADP1 on the methanotroph spent media, organic acid-rich spent media were used as carbon sources for 1-alkene production by engineered A. baylyi ADP1 ‘tesA-undA. The results showed that A. baylyi ADP1 ‘tesA-undA completely utilized acetate (0.3 mM), succinate (0.05 mM), and malate (0.2 mM) present in the spent mxsedia of methanotrophs, except formate, similar to the wild-type ADP1 cells (Figure 6A). The production of 1-undecene from M. tundripaludum SV96 spent medium (14.1 ± 2.7 μg L^–1) was higher than that from M. rosea SV97 cultivations (1.0 ± 0.5 μg L^–1) (Figure 6C), corroborating with the organic acid concentrations (Figures 6A,B). Likewise, the growth of A. baylyi ADP1 ‘tesA-undA was higher in the spent medium from M. tundripaludum SV96 (0.115 ± 0.06 OD_600nm) than from M. rosea SV97 (0.070 ± 0.054 OD_600nm) (Figure 6C). Production of 1-undecene was not detected in the control cultures (Supplementary Figure 4). In addition, the biomass growth detected in the methanotroph spent media was from solely A. baylyi ADP1 ‘tesA-undA as both methanotrophs could not grow without the presence of CH₄.

The obtained 1-undecene concentrations in this study (1.0 and 14.1 μg L^–1) were lower than that previously reported for A. baylyi ADP1 ‘tesA-undA (Luo et al., 2019; Salmela et al., 2020). This phenomenon could be attributed to the higher substrate concentrations used in these studies. For example, Luo et al. (2019) developed a highly ferulate-tolerant A. baylyi ADP1 strain for 1-undecene production using adaptive laboratory evolution. The authors tested the use of a high concentration of ferulate (100 mM) as the sole carbon source and obtained a 1-undecene production titer of 72 ± 7.5 μg L^–1, corresponding to a production yield of 1.0 μmol mol C_substrate^–1. In another study, Salmela et al. (2020) obtained 1-undecene concentration of up to ∼107 ± 8 μg L^–1 from a two-stage system for 1-undecene production from cellulose which was converted into organic metabolites by Clostridium cellulolyticum (containing 5.2 mM glucose, 4.9 mM acetate, and 6.8 mM lactate). The authors reported the highest 1-undecene production yield of ∼35 μmol mol C_substrate^–1 using lactate as the substrate. Considering the 1-undecene production yield, the use of spent media of methanotrophs as a growth medium in our study was promising and comparable to those in previous studies, resulting in 68.9 ± 11.6 and 40.6 ± 19.8 μmol mol C_substrate^–1 for M. tundripaludum SV96 and M. rosea SV97, respectively. The carbon recovery obtained for 1-undecene production accounted for 0.065 and 0.045% of the total organic acids-carbon consumed by M. tundripaludum SV96 and M. rosea SV97, respectively (Supplementary Table 2). In our study, the spent media of methanotrophs likely did not contain intermediate compounds that were toxic to the growth and 1-undecene production of A. baylyi ADP1 ‘tesA-undA. Furthermore, the media can be directly used for cultivation without purification or additional downstream processing. However, in the future studies, the addition of some key macro/micronutrients should be evaluated, as it would benefit long-term process performance to increase 1-undecene concentration.

These results indicate that the spent media from microaerobic fermentation by methanotrophs are an excellent carbon source for heterotrophs. This two-stage bacterial process extends the range of CH₄-derived products to the C11 compound (1-undecene). In further studies, the two-stage bacterial system will be scaled up and the process parameters will be optimized to be useful and competent for practical applications. In particular, the results obtained from our study show that different O₂ and CH₄ supplementation schemes in the headspace could lead to different concentrations and types of organic acids produced in the system. Furthermore, the contact time between methanotrophs and CH₄ is important for maintaining active cells in fermenters and bioreactors (Guerrero-Cruz et al., 2021). In scale-up methanotroph cultivation, operational parameters such as O₂ and CH₄ inlet concentrations, dilution rates, and nitrogen sources should be optimized. The production of organic acids, particularly acetate and succinate, by methanotrophs has also been observed under aerobic conditions, suggesting potential strategies to develop a bioprocess system to co-cultivate methanotroph and A. baylyi ADP1 in a single system. In addition, the long-term effect of methanotroph cells in the spent media on the A. baylyi ADP1 growth and the 1-alkene production should be evaluated. For example, the spent media directly used as the A. baylyi ADP1 growth medium should be compared with the filtered and the centrifuged spent media prior to application.