Microbial conversion of syngas to single cell protein

Syngas has been widely utilized as substrate in microbial processes to produce various high-value products. Nevertheless, its applicability for single cell protein (SCP) production by hydrogen-oxidizing bacteria (HOB) has never been explored. Especially, the role of carbon monoxide in SCP synthesis is unknown. Thus, this study first investigated the effect of different CO and H 2 ratios in syngas on Cupriavidus necator H16 growth and SCP production. The growth of C. necator H16 was significantly restrained with the increase of CO content, and it almost ceased once the CO:H 2 ratios were above 1. In addition, the growth rate of C. necator H16 increased with the increase of shaking frequency and inoculum volume. Furthermore, larger gas – liquid interfacial areas and gas input amount resulted in a higher growth rate and OD value. The raw protein contents were around 50 – 60% regardless of CO concentration, and the amino acid profiles showed no apparent differences. The results showed that although the presence of CO led to the reduction of the growth rate of C. necator H16, it had a limited effect on the SCP quality. H 2 and CO 2 in syngas were the primary substrates for SCP production, while CO served more as a toxic inhibitor. The successful demonstration of SCP synthesis from syngas may provide a superior route for the valorization of waste biomass by turning it into food or feed.


Introduction
Food security is a worldwide challenge in the 21st century due to population growth and climate change.The global population could grow to 9.7 billion in 2050 [1].Conventional cultivation approaches for human food production have the challenges of competition for land, water, and energy, greenhouse gases emission, nitrogen losses, and pesticides pollution [2][3][4].Therefore, innovations in agricultural technology are needed to improve the productivity of the food to avoid triggering irreversible damages or unsustainable development in the ecosystem [5].Moreover, it is necessary to find an alternatively, sustainable, and efficient method to deal with the food crisis.
Single cell protein (SCP), a promising and environmentally friendly substitute for animal feed and human nutrition, has attracted more and more attention in recent years [6,7].SCP is generally produced as biomass through fermentation processes, and it shows impressive characteristics with high protein content and a broad amino acid spectrum [8].Among the available SCPs producing microorganisms (e.g., fungi, yeast, algae, and bacteria), hydrogen-oxidizing bacteria (HOB), also called Knallgas bacteria, is a group of bacteria that could utilize hydrogen as the electron donor and oxygen as the electron acceptor to fix carbon dioxide into protein (up to 75% of cell content) [9].Cupriavidus necator H16 is a model HOB strain that has been intensively studied for SCP production using H 2 and CO 2 [10].In addition to the high protein content compared with other SCP producers, short generation time is also the advantage of HOB for protein production [11].The typical substrate of HOB for SCP production is a mixed gas, including hydrogen, oxygen, and carbon dioxide.The stoichiometry of the autotrophic culture of C. necator is shown as the following equation [12].Thus, the availability of renewable substrates (especially H 2 ) is critical for the practical application of this biological process [13].
Syngas (or synthesis gas) derived from biomass gasification is one of the most inexpensive and flexible substrates (costing ≤ $6 per million Btu) for the production of renewable biofuel (e.g., ethanol) and commodity chemicals through biological fermentation [14][15][16][17].Since syngas is primarily composed of carbon monoxide, hydrogen, and carbon dioxide, it could be an alternative substrate for HOB to produce SCP.The role of CO in the HOB-driving SCP synthesis process needs to be uncovered, as it is known as either a poison to many microbes or a carbon and energy source for the conventional syngas fermenting bacteria.
Thus, this study aimed to explore the feasibility of converting syngas to SCP using HOB and explore the role of CO in this novel syngas fermentation process.The growth of HOB under different syngas compositions was first evaluated.Subsequently, the syngas-to-protein bioconversion was further investigated at varied parameters, including shaking frequency, inoculum volume, and gas:liquid ratio.Finally, the quantity and quality of produced protein were characterized and compared with the conventional process using H 2 , O 2, CO 2 mixture gas as substrates., 0.004 g NH 4 VO 3 , 0.050 g ferric ammonium citrate, 5 mL trace element solution and 5 mL standard vitamin solution in deionized water.The trace element solution (DSMZ_Medium 27) and standard vitamin solution (DSMZ_Medium 81) were prepared by following the recipes given by DSMZ [18,19].Except for the standard vitamin solution, the solutions were autoclaved for 15 min at 121 • C after being sparged with pure N 2.After that, the vitamin solution was added through the sterilized filter (Nylon 0.2 μm, Agilent, USA).The final pH of the medium was approximate 6.8-6.9.

Experimental procedure
Glass serum bottles (total volume of 240 mL) were used in the experiment, with 40 mL medium and 200 mL headspace.After autoclaving, the headspace was replaced with 350 mL aseptic mixed gas (CO, H 2 , CO 2, and O 2 ).N 2 was added to ensure the same pressure before starting the tests.The initial ratio of H 2 , CO 2 , and O 2 was 10: 2: 3 according to Eq. (1) for the batch experiments.Triplicate experiments were conducted under every condition.The initial optical density at 600 nm (OD 600 ) wavelength after inoculation was around 0.08 ± 0.01 unless otherwise stated.
Syngas compositions are varied with the feedstock characteristics [20,21].Since CO and H 2 are the main components in syngas, the effect of different CO and H 2 ratios was first explored.The gas compositions of experimental groups are shown in Table 1.CO:H 2 ratios at 0.5, 1, 1.5, and 2 were set up, named Syngas-0.5,Syngas-1 Syngas-1.5, and Syngas-2, respectively.Without CO, Blank, and Control groups were also operated.The gas atmosphere of Blank group was set the same as Without CO group but no inoculation, while the Control group was with the same inoculation but without gas substrate supply (only N 2 in headspace).All bottles were placed in an incubator (INCU-Line® | VWR, USA) under 30 • C and 150 rpm shaking frequency.The sample taken interval is 12 h or 24 h.

Growth performance of C. necator H16
The growth of C. necator H16 was evaluated by OD 600 value during the cultivation.Liquid samples were taken from bottles and measured by spectrophotometer (Varian Cary 50 Bio, Varian, Australia).The initial and final pH of the medium was measured by a pH meter (PHM92, Radiometer analytical, Denmark).At the end of each batch, the remaining culture medium was collected to measure biomass concentration and amino acid profile, and the biomass concentration was expressed in cell dry weight (CDW, g/L).Firstly, the culture was centrifuged at 8000 rpm for 10 min, and then the supernatant was discarded and rinsed the precipitation with distilled water three times.Finally, the precipitates were lyophilized by a freeze dryer (Coolsafe 100-9 Pro, Denmark) overnight.Weighting the powder to calculate the biomass concentration.

Gas composition measurement
0.2 mL gas sample was taken from headspace for determining the gas composition (CO, H 2 , CO 2, and O 2 ) using the gas chromatograph equipped with a thermal conductivity detector (TCD) (GC-TRACE 1310, Thermo Scientific, USA) [22].

Evaluation of SCP and amino acid profile
The raw SCP was quantified by the Pierce TM BCA Protein Assay Kit (Thermo Fisher Scientific, USA), and the protein concentration was calculated according to the final biomass concentration.For amino acid analysis, 0.01 g lyophilized biomass was put into a glass vial adding 500 μL 6 N HCl.After that, the glass vials were placed into a Teflon liner which contained 20 mL 6 N HCl.The liner was flushed with N 2 and then vacuumed.The liner was then moved to the microwave rotor for hydrolysis.During the hydrolysis, the temperature was increased to 130 • C in 5 min and then held for 30 min, followed by cooling down at 0 • C for 15 min.Remarkably, tryptophan and cysteine cannot be analyzed by this hydrolysis method.The hydrolysate was analyzed by HPLC-MS/MS.

The performance of C. necator H16 under different syngas compositions
The effect of different CO and H 2 ratios on C. necator H16 for SCP production was studied first.The growth curves of C. necator H16 were shown in Fig. 1a, which indicates that C. necator H16 was inhibited remarkably with syngas.Compared to Without CO group, all groups with syngas performanced a lower growth and longer growth cycle.Besides, the OD 600 at the end of the batch decreased with the increasing of CO content.OD 600 value reached 1.49 without CO, but it was approximately 0.80 at Syngas-0.5 and Syngas-1 groups, and hardly OD value increased for Syngas-1.5 and Syngas-2 groups, which suggested that CO was an inhibitor during the bioconversion.Once the ratios were above 1 (~28% CO (v/v)), the growth of C. necator H16 was

Table 1
The gas composition with different CO:H 2 ratios (Under 30 nearly the same at Syngas-0.5 and Syngas-1, the growth rate was higher under 0.5.It took around 72 h to reach the stationary phase at CO:H 2 ratio of 0.5, while it took 96 h at 1.The corresponding growth rate was 0.0119/h and 0.0082/h (based on OD value) at the ratio of 0.5 and 1, respectively.Small aggregations were observed in the bottles.It was reported that the excretion of extracellular polymeric substances (EPS) could be improved in extreme environments by microorganisms to protect cells, and EPS could also serve as an adhesive agent to reunite cells [25,26].The results indicated that CO was a negative factor for C. necator H16 growth.CO contents were stable in all groups over time (Fig. 1b), which was direct evidence that CO cannot be utilized as a carbon source by C. necator H16 during syngas fermentation.The changes of H 2 , CO 2, and O 2 contents in the headspace over time were also monitored (Fig. 1c, 1d, and Fig. S1).They were continually consumed during the growth of C. necator H16 and were almost completely utilized when the OD 600 reached the maximum.Compared Fig. 1c with 1d, we can notice the consumption of CO 2 was after H 2 consumption, and there was CO 2 remaining in Syngas-1 group at 96 h, when H 2 was used up completely.H 2 acts as an energy source for HOB.In addition to providing reduction equivalent for HOB to fix CO 2 , the generated electrons from oxidized H 2 are also transferred to the respiratory chain and recover energy in the form of ATP.It means that the fixation of CO 2 is based on the oxidation of H 2 , so H 2 consumption precedes CO 2 .It was expected that the most suitable consumption of H 2 and CO 2 for C. necator was according to Eq. ( 1).Recent studies have suggested that the stoichiometric formula of HOB varies in different growth periods [27], the requirement of H 2 increased with the increasing of bacteria density at consumption per unit CO 2 , thus H 2 and CO 2 were not consumed in a fixed ratio all the time.C. necator H16 hardly grew in Syngas-1.5 and Syngas-2, and a small amount of gas was consumed, leaving much residual gas in the bottles at the end of the batch.In the Blank group, all gas contents in the headspace were stable, which stated that gas dissolution in the medium was limited under the tested conditions and would not contribute to the changes of gas contents.
CO 2 is a greenhouse gas and is regarded as one of the major impurities in syngas.Multiple approaches such as membrane technology, catalytic strategy, and adsorption have been intensively studied for CO 2 separation from syngas [28][29][30].In this study, CO 2 and H 2 were assimilated by HOB, while CO was left after fermentation.Thus, the syngas-to-protein process studied here could also be a new strategy for syngas purification (i.g., to obtain pure CO).CO is a widely used reducing agent for various carbonylation reactions in the industry to produce valued products and catalytic processes [31,32].On the other hand, the toxicity of CO on SCP synthesis should be addressed to improve the gas-to-protein conversion rate.
In most cases of actual syngas, CO is the dominant component.According to the tests under different ratios of CO and H 2 , although the growth of C. necator H16 was affected by CO, a comparable OD value can still be obtained in the Syngas-1 group.Hence, the CO:H 2 ratio of 1 was chosen in the following tests of several influence factors.
The biomass concentration, crude protein content, and amino acid profile were compared between the group Syngas-1 and group Without CO (Table 2).The CDWs of the two groups were 0.420 ± 0.038 g/L and 0.379 ± 0.014 g/L, respectively.Even though the OD 600 maximum without CO was almost twice that of Syngas-1, there was not much difference in biomass concentration.It could be due to cell aggregations caused by CO toxicity in group Syngas-1, leading to a lower OD value.The crude protein contents were 52.01 ± 0.88% of group Without CO and 50.03 ± 0.78% of group Syngas-1, which were in the range of 50 %-80% reported by the previous SCP studies [33].
The contents of amino acids in crude protein were 56.08% and 48.12% in groups Without CO and Syngas-1, respectively.CO slightly reduced the quality of the produced protein.The profile of the amino acid was summarized in Fig. 2. According to P-value of the ANOVA analysis at 0.05 level (Table S1), the differences in the content of each amino acid in the two groups were not significant.Eighteen amino acids were detected from the biomass.Histidine, isoleucine, leucine, lysine, methionine, phenylalanine, threonine, and valine are essential amino acids for human beings, which cannot be made by the body and must come from food [34].Tryptophan also belongs to the essential amino acid.Unfortunately, the analytic method we employed cannot detect it.The others are classified as the non-essential amino acids that the human body can produce.Although essential amino acid contents in biomass from our study were lower than soybean and fishmeal, the amino acid composition was balanced.The biomass has the feasibility of being used as animal feed or food (Table S2).
Overall, the results showed that the quantity of crude protein decreased slightly when CO content was high in syngas, but the amino acid quality was almost not affected.Although CO slows down the growth of C. necator H16, syngas could still be an alternative substrate for SCP production.The syngas-to-protein process has excellent potential to separate CO 2 and H 2 from CO for syngas purification.

The effect of different shaking frequencies on the growth of C. necator H16
C. necator H16 used gas substrates here for SCP synthesis, gas-liquid mass transfer capacity could be one of the key factors affecting its growth.Shaking is one of the methods to promote increased gas and liquid transfer.Thus, bioconversion of syngas to SCP was further studied at different shaking frequencies (0, 80, and 150 rpm) under CO:H 2 ratio of 1, in which the gas substrates include 100 mL CO, 100 mL H 2 , 20 mL CO 2 , 30 mL O 2 and 100 mL N 2 (same as the Syngas-1 group shown in Table 1).As shown in Fig. 3a, it was observed that higher shaking frequency contributed to a higher cell growth rate.The OD reached to the maximum at around 120 h with 150 rpm, while C. necator H16 was still growing after 168 h at 0 and 80 rpm.The OD 600 value only reached 0.56 without shaking while achieving 0.89 at 150 rpm.The contents of H 2 , CO 2 and O 2 were decreased with the time in all groups (Fig. 3b, 3c  and 3d) and all gases were almost used up at 150 rpm when C. necator H16 grew to the stationary phase (120 h), while a large percentage of gases were left at 0 and 80 rpm after 168 h.If the batch cycle was extended for a longer reaction time, C. necator H16 would continue to grow.No or lower shaking frequency obviously limited the gas consumption and restricted the growth of gas-fed bacteria.Increasing shaking frequency could promote gas-liquid mass transfer [35], which means more gas substrates were available for C. necator H16 during the same time.Shaking, stirring, and suppling gas by bubble column are the common methods to increase gas and liquid mass transfer.Specially designed reactors can also be run to meet the requirements in industrial applications.

The effect of different inoculum volumes on the growth of C. necator H16
The inoculum volume is essential to shorten the lag phase of bacteria [36,37], especially in a severe environment such as the existence of carbon monoxide in this study.The growth of C. necator H16 was investigated with different inoculation volumes (0.5, 1 and 2 mL, initial OD 600 were 0.04 ± 0.01, 0.08 ± 0.01, and 0.16 ± 0.03).In Fig. 4a, the larger the inoculum volume, the higher the obtained growth rate.Even though the final OD values were almost the same, the time to reach the stationary phase decreased with the increase of inoculum volume.The lag-phase time was shortened obviously with 2 mL inoculation.C. necator H16 with all inoculum volume ended with almost the same OD value, and nearly all gases were consumed (Fig. 4b, c, and d).Large inoculum volume could shorten the growth cycle, however, it might result in a cost increase [38].Hence, in consideration of economic benefifs, an appropriate inoculation amount should be adopted for industrial applications.

The performance of C. necator H16 with different gas:liquid ratios
In the above tests, the nutrients (e.g., nitrogen source, trace elements and vitamins) in the medium were sufficient to support the growth of HOB, so gas substrate supply was the limiting factor at the end of the batch.Thus, the effect of gas:liquid ratios (5:1, 8:1, and 10:1) on C. necator H16 growth was further studied.The medium volumes were fixed as 40 mL and different sizes of bottles were used to achieve different gas:liquid ratios, the groups were named Syngas-S (small), Syngas-M (medium) and Syngas-L (large), respectively.The ratio of CO and H 2 was 1, and the initial headspace pressure was the same as the above tests.We can see from Fig. 5 that the maximum cell growth increased with the increase of gas:liquid ratio and the corresponding maximum OD 600 value were approximately 0.79, 1.98, and 2.72 of groups Syngas-S, Synags-M and Synags-L.The maximum OD 600 was obtained at around 96 h in all groups at around 96 h, the growth rates based on OD value were 0.0082/h, 0.0206/h, and 0.0283/h, respectively.A larger amount of gas meant more energy and carbon supply, and the cells could reach higher OD 600 until the gas substrate was consumed entirely.Besides, the gas:liquid mass transfer coefficient can be enhanced by increasing the gas-liquid interfacial area [39].All the bottles we used were cylindrical, and the bottom diameters increased with the increase of the bottle working volume, so the relationship of interfacial areas was that A Syngas-S < A Syngas-M < A Syngas-L (not considered shaking), which could explain that the increasing trend of growth rate with bottle size.The results imply that high productivity could be obtained by increasing the input gas volume and gas-liquid interfacial area.

Significance and perspective
This study demonstrated the potential of syngas as the substrate for C. necator H16 to produce SCP for the first time.The effect of CO:H 2 ratio on the growth of C. necator H16 was revealed.CO was functioning as an inhibitor hampered the cell metabolism and growth.The study of the variables including shaking frequency, inoculum volume and gas:liquid ratio was necessary for the optimization of the SCP biosynthesis process using this new substrate (i.e., syngas).These parameters are common and basic for the gas-fed processes, but the performance of C. necator H16 under a certain shaking frequency or inoculum volume was unknown, not to mention with a new substrate for SCP production.The explorations of them are important for a new substrate/concept and the later scaling-up of the process for industrial application.The outcomes provided a sustainable integration of renewable gas and hydrogen autotrophic biosynthesis process, which could be supplementary to animal feed and human food supply and a potential method of CO purification.
Though promising, several challenges still need to be addressed in future studies.The crude protein yield was around 50.03%, while it can reach 71% in previous studies using a continuous reactor [12].One of the possible limitations of the batch reactor used in our study is that the insufficient gas supply limited the growth of C. necator H16.A continuous gas flow could be adopted in future studies to improve bioconversion.Several studies related to SCP production integrated nitrogen recovery and assimilation from wastewater using membrane technology  1.
Y. Jiang et al.
or electrochemical system [13,40,41].It is a sustainable and efficient alternative to recycling nutrients in waste streams as feed and food and provides some new ideas.Moreover, an efficient process should be developed to further use the residual CO for SCP synthesis.Co-culture of conventional syngas fermenting bacteria and HOB could be a possible solution for fully syngas-to-protein bioconversion.Furthermore, SCP as a new protein alternative needs time to be accepted for animal feed or human food consumption.The toxicological and nutritional assessments are very important for any food or feed product, meanwhile, the establishment of regulations is beneficial to market operation in the future [33,42,43].

Conclusions
This research demonstrated the microbial conversion of syngas into SCP using a model HOB species, Cupriavidus necator H16.In addition, the effects of different factors, including shaking frequency, inoculum volume, and gas:liquid ratio, were explored.The major conclusions are as follows.
• C. necator H16 can use H 2 and CO 2 but cannot metabolize CO.
• Higher CO content can inhibit cell growth, but it did not affect the amino acid profiles.1. • High shaking frequency and big inoculum volume can facilitate SCP synthesis.• High biomass productivity was obtained with a high gas:liquid ratio and improved mass transfer capacity due to large interfacial areas.
Although utilizing CO as the carbon source by HOB for SCP production was not realized in this study, syngas shows the potential to be used as an alternative substrate for C. necator H16.Furthermore, the syngas-to-protein bioconversion process could also be a potential method for selective recovery of CO from syngas.The concept of integrating syngas and HOB biosynthesis maybe be a crucial addition to food and fodder production in the world.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 1 .
Fig. 1.The performance of C. necator H16 under different syngas compositions: (a) the change of OD 600 over time; the change of (b) CO, (c) H 2 , and (d) CO 2 content in the headspace over time.

Fig. 3 .
Fig. 3.The performance of C. necator H16 under different shaking frequencies.(a) The change of OD 600 over time; the change of (b) H 2 , (c) CO 2, and (d) O 2 contentin the headspace over time.The initial gas composition was the same as the Syngas-1 group shown in Table1.

Fig. 4 .
Fig. 4. The performance of C. necator H16 under different inoculum volumes.(a) The change of OD 600 over time; the change of (b) H 2 , (c) CO 2, and (d) O 2 content in the headspace over time.The initial gas composition the same as the Syngas-1 group shown in Table1.

Table 2
Comparison of SCP production under the gas compositions of Without CO and Syngas-1.
Y.Jiang et al.