Utilization of Protein-Rich Agricultural Residues in the Biotechnological Production of L-Lactic Acid and 1,3-Propanediol for Added Value

Abstract

Due to limited fossil resources and climate change, biotechnological processes converting renewable resources into industrial raw materials are increasingly important. Many of these processes require yeast extract for microorganism growth, a high-cost factor. Therefore, the use of inexpensive, protein-rich agricultural raw materials as a source of nutrients is desirable. However, their usage usually results in lower product concentrations and productivity in the fermentation process. This work investigates the nutritional requirements for the production of L-lactic acid using Lactobacillus casei ATCC 393 and 1,3-propanediol using Clostridium butyricum DSM 25047, aiming to replace complex nutrient sources with hydrolyzed protein-rich agricultural raw materials. In the production of 1,3-propanediol, yeast extract was largely (80%) replaced by rapeseed meal hydrolyzate, achieving the same final product concentration and maximum productivity. In the production of L-lactic acid, rapeseed meal hydrolyzate supplemented with B vitamins, mineral salts, cysteine, and tryptophan replaced yeast and meat extracts, achieving the same final product concentration with comparable maximum productivity.

1. Introduction

Renewable raw materials are a key factor in numerous strategies for restructuring our economic systems towards a climate-neutral bioeconomy. This applies in particular to the production of basic substances and materials for the chemical industry, such as starting materials for plastics, detergents and cleaning agents, paints and coatings, and much more. Consequently, biologically-based components for the polymer and packaging sectors are in competition with cost-effective petrochemical plastics. To offer an ecologically and economically viable alternative to petrochemical plastics, it is essential to optimize yield, productivity, and final concentration in the biotechnological production of monomers. Additionally, production costs, especially those related to the medium, must be reduced. Besides the substrate, nutrient sources such as yeast extract, priced at 9–10 € kg⁻¹, constitute a significant portion of the production costs [1]. Techno-economic evaluations identify yeast extract as a major cost factor in lactic acid production, accounting for up to 38% of total costs [2,3,4]. Yeast extract and other complex nutrient sources are utilized in fermentation processes as they supply organic nitrogen, such as amino acids, along with vitamins, salts, trace elements, and nucleic acids [5]. In numerous technical processes, protein-rich residues are produced as by-products at low cost and could be used as a complex nitrogen source in bioprocesses. The substitution of yeast extract with agricultural residues has been researched for years and is becoming increasingly important due to the growing number of biotechnological processes in industrial production. Using the example of fermentative lactate production, many studies can be found in the literature where various residues were used as sources of carbon and nutrients [6].

Alternative nutrient sources like affordable, protein-rich agricultural by-product can substitute yeast extract [7,8,9]. However, these agricultural residues must first be broken down through enzymatic or chemical hydrolysis due to their complex molecular structure [10]. A significant challenge with common hydrolyzates is the inadequate availability of free amino acids and small peptides, typically due to insufficient hydrolysis, leading to reduced lactic acid productivity [7]. To address this, a previous study developed a hydrolysis method using 3 M H₂SO₄, successfully replacing yeast extract with hydrolyzates derived from low-cost, protein-rich agricultural residues. This was achieved through a systematic analysis of the nutritional requirements for D-lactic acid production using Sporolactobacillus inulinus [7,11].

In the present study, this knowledge will be transferred to two industrially relevant processes, the L-lactic acid production with Lactobacillus casei and the 1,3-propanediol production with Clostridium butyricum, thereby using chemically digested rapeseed meal (RM) as a protein-rich agricultural raw material.

RM is a by-product of the industrial production of rapeseed oil made from rapeseed [12]. The worldwide average annual production of rapeseed oil in the years from 1996 until 2000 was 12.64 × 10⁶ t [13]. In recent years, interest in the processing of rapeseed has increased, as rapeseed oil is used for both food and technical purposes. During the production process, large amounts, up to 48% of the mass of the processed seeds, of RM occurs. The protein content is 38–48%. The amino acid composition of RM is very balanced, which is why it is used as a high-protein component in the production of animal feed. However, due to the presence of antinutritional compounds and high fiber content, their usage as a feed additive is limited and large quantities of rapeseed meal remain unused [13,14].

In 2011, annual lactate production worldwide was about 370,000 t [15]. That amount consisted almost exclusively of L-lactic acid, which is used as a preservative, acidifier, and starting material for bio-based and biodegradable polymers (polylactides, PLA) [16]. Due to high productivity and yields, lactate is preferably produced by fermentation with lactic acid bacteria, especially Lactobacillus spp. [6]. Since these are auxotrophic for various amino acids and vitamins, they require complex nutrient sources to produce L-lactic acid [17].

The organic compound 1,3-propanediol (1,3-PD) can be used as a bio-based component in polyesters, such as polytrimethylene terephthalate and polyurethanes. The properties of these polyesters can be easily modified, allowing for versatile applications [18]. Historically, 1,3-PD was produced chemically, but these methods were completely replaced by a biological process developed by DuPont in 2003. The DuPont process uses recombinant E. coli and glucose as the carbon substrate and has achieved high titers of 135 g L⁻¹ and productivities of 3.5 g L⁻¹ h⁻¹ [19,20]. There is a large number of key players in the production of 1,3-PD, namely CovationBio PD (previously Dupont Tate & Lyle) with a current production capacity of bio-1,3-PD of 77,000 tons per year, Metabolic Explorer, Zhangjiagang Glory Chemical Industry Co. Ltd., and Merck KGaA [20,21]. Wild-type strains, such as Klebsiella spp., Clostridium spp., Citrobacter spp., Enterobacter spp., and Lactobacillus spp., are able to produce 1,3-propanediol from glycerol by anaerobic fermentation. With wild-type strains, however, the production of 1,3-propanediol from glucose is only possible by a two-step process or co-fermentation [18]. Due to the growth-inhibiting effect of glycerol, a fed-batch strategy is usually used for fermentation, which can produce high concentrations of 1,3-propanediol. Bock et al. (2004) achieved with Clostridium sp. a final titer of 103 g L⁻¹ with a productivity of 1.9 g L⁻¹ h⁻¹ [22]. A significantly higher productivity of 3.3 g L⁻¹ h⁻¹ was achieved by Wilkens et al. (2012) with the strain AKR102a, where 93.7 g L⁻¹ 1,3-propanediol with a yield of 0.52 g g⁻¹ was produced [23]. These results show the high potential of 1,3-propanediol production with wild-type microorganisms. However, since 50% of the production costs are attributable to the raw materials used, with glycerol and yeast extract being the major parts, a cheap substitute of yeast extract as a source of nutrients as well as using inexpensive feedstocks makes sense [21,24,25].

The aim of this study is to elucidate the nutrient requirements for the fermentation processes of L-lactic acid and 1,3-propanediol and to replace the complex nutrient source with hydrolyzed protein-rich agricultural raw materials, resulting in equal productivity, yield, and final titer.

2. Materials and Methods

The experiments were performed with a self-isolated strain Clostridium butyricum, which is deposited at the DSMZ (German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany) as DSM 25047 [23]. For long-term storage, glycerol stock cultures were prepared. A solution of the cell culture was mixed with the same amount of pure glycerol (1:1 ratio), shock frozen in liquid nitrogen, and stored at −80 °C. Before initiating the experiments, at least three generations of pre-cultures were cultivated as static cultures at 32 °C. These cultures were grown in 50-mL glass vials, sealed with butyl rubber septa, and filled with 30 mL of mineral salt medium under anaerobic conditions. All media components used were of pro-analysis quality, sourced from either Merck (Darmstad, Germany), Sigma–Aldrich (Saint Louis, MO, USA), or Roth (Karlsruhe, Germany). Per liter of distilled water, the mineral salt medium contained: 9.09 g KH₂PO₄, 0.53 g NH₄Cl, 0.123 g MgSO₄·7H₂O, 0.017 g CaSO₄·2H₂O, 0.01 g FeSO₄·7H₂O, 1 g yeast extract (Merck, granulated, Darmstadt, Germany), 1 mL of a 1 g L⁻¹ resazurin solution, 0.25 g L⁻¹ cysteine·HCl, and 2 mL trace element solution. Per liter of distilled water, the trace element solution (DSMZ medium 144) contained: 64 g nitrilotriacetic acid, 1 g FeCl₃·4H₂O, 0.5 g MnCl₂·4H₂O, 0.85 g CoCl₂·6H₂O, 0.5 g CaCl₂·2H₂O, 0.5 g ZnCl₂, 0.1 g CuCl₂, 0.05 g H₃BO₃, 0.05 g Na₂MoO₄·2H₂O, 0.13 g NiCl₂·6H₂O, 5 g NaCl, and 0.1 g Na₂SeO₃·5H₂O. The pH value of the trace element solution was adjusted to 6.5 with NaOH/HCl. Pure glycerol (≥98%) was added to the medium as a carbon source (25 g L⁻¹). The reference and optimization cultivations took place as static cultures in 50-mL glass vials sealed with butyl rubber septa filled under anaerobic conditions with a working volume of 30 mL modified mineral salt medium at 32 °C. In addition to the mineral salt medium 200 mM MES (2-(N-morpholino)ethanesulfonic acid), 4 g L⁻¹ yeast extract and 50 g L⁻¹ glycerol, as an initial substrate concentration, was used in the modified medium. This modification should ensure sufficient buffering and rule out substrate limitation in order to see the influence of nutrients.

Lactobacillus casei ATCC 393 was obtained from the American Type Culture Collection (Manassas, VA, USA). The pre-culture was initiated by inoculating with a cryo stock culture stored at −80 °C in 50% (v/v) glycerol and incubated in 100 mL shake flasks for 17 h at 38 °C without stirring. For both the pre-culture and the subsequent cultivation, a modified MRS (deMan, Rogosa, and Sharpe [26]) medium was utilized, consisting of 20 g L⁻¹ (pre-culture) or 110 g L⁻¹ (cultivation) glucose; 5 g L⁻¹ meat extract (Carl Roth GmbH and Co. KG, Karlsruhe, Germany); 5 g L⁻¹ yeast extract (Merck, Fermtech, Darmstadt, Germany); and 2 g L⁻¹ K₂HPO₄, 2 g L⁻¹ (NH₄)₂H citrate, 5 g L⁻¹ sodium acetate, 0.2 g L⁻¹ MgSO₄, 50 mg L⁻¹ MnSO₄, and 90 g L⁻¹ pulverized CaCO₃ (Carl Roth, Karlsruhe, Germany). A total nitrogen content of 1.11 g L⁻¹ was used for cultivation with hydrolyzates and combinations of yeast extract and hydrolyzate, which corresponded to a total nitrogen content of 5 g L⁻¹ meat extract and 5 g L⁻¹ yeast extract. The cultures were inoculated with 2% (v/v) (pre-culture) and handled in 24-well plates containing 1.5-mL culture volumes (Sarstedt, Nümbrecht, Germany) at 37 °C and 400 rpm, with at least duplicate samples. The average values derived from these duplicate experiments, along with their corresponding standard deviations (represented as error bars), are shown in the figures. The solutions were autoclaved at 121 °C for 20 min to sterilize them.

The B vitamin solution and amino acid solution for the modified MRS medium, containing 5 g L⁻¹ meat extract and 5 g L⁻¹ yeast extract, were calculated on the findings of Klotz et al. [11] and can be found in detail in Krull et al. [26].

Rapeseed meal (RM) was obtained from Archer Daniels Midland AG (Chicago, IL, USA). Its total nitrogen content amounted to 5.39%. A detailed characterization of this agricultural residue can be found in [11].

For hydrolysis, rapeseed meal, an agricultural residue, was milled using an ultra-centrifugal mill (750 µm, ZM 200, Retsch, Haan, Germany) and sieved to achieve a particle size of less than 710 µm. In the chemical hydrolysis process, 400 g of this residue was continuously stirred with 1.2 L of 3 M H₂SO₄ in a 2-L triple-necked round-bottom flask fitted with a reflux condenser. This mixture was maintained at 110 °C for 24 h, without controlling the pH. After cooling to room temperature, the pH of the hydrolyzate was adjusted to 6 using Ca(OH)₂. The resulting slurry was centrifuged at 4600 g for 20 min at 10 °C, and the pellet was washed twice with 400 mL of distilled water. The supernatants were combined and filtered through a pleated filter (MN 615 ¼ Ø 150 mm, Macherey-Nagel, Düren, Germany). The pH of the hydrolyzate was then adjusted to 6.5 with 1 M NaOH, sterilized at 121 °C for 20 min, and analyzed for total nitrogen content [26].

To quantify the substrate concentrations, glycerol and glucose, the product concentrations, lactic acid and 1,3-propanediol (1,3-PD), as well as the side products of high-performance liquid chromatography (HPLC), were used. The HPLC system consisted of a Dionex ICS-5000 from Thermo Scientific (Sunnyvale, CA, USA) fitted with an Aminex HPX-87H column (300 × 7.8 mm) from Bio-Rad (Hercules, CA, USA) and a refractive index detector (RI-101, Shodex, Tokyo, Japan) operating at 60 °C. The mobile phase used was 5 mM H₂SO₄ with a flow rate of 0.7 mL min⁻¹. Samples of fifty microliters were initially dissolved in 950 µL of 50 mM H₂SO₄ and then centrifuged at 20,800 g for 10 min at room temperature (RT). The resulting supernatant was subsequently diluted with 50 mM H₂SO₄.

Total nitrogen (TN) was analyzed using the Kjeldahl method according to Krull et al. [26].

A Phenomenex (Torrance, CA, USA) analytical kit (EZ:faast GC-FID Physiological) was used for the quantification of free amino acids according to Krull et al. [26]. For peptide-bounded amino acids, 50 mg of a dried sample was hydrolyzed with 1 mL 6 M HCl + 0.02% phenol in a head-space vial (cleaned with concentrated HCl and water). The vial was sealed with a septum, deep-frozen in liquid nitrogen, evacuated, and heated to 110 °C for 48 h. The hydrolyzate was diluted to 100 mL and filtered. Free amino acids were analyzed according to the manufacturer’s instructions. A gas chromatograph with a flame ionization detector GC-17A from Shimadzu (Kyoto, Japan) with a Zebron ZB-AAA column (10 m × 0.25 mm × 0.25 μm) from Phenomenex (Torrance, CA, USA) was used for the measurement with H₂ as carrier gas with a flow rate of 2.15 mL min⁻¹. The temperature of 110 °C was heated up to 250 °C with a gradient of 20 °C min⁻¹, further up to 320 °C with 10 °C min⁻¹, and was held for 1 min.

The mean values calculated from the duplicate experiments and the corresponding standard deviations (error bars) are presented in the figures.

3. Results and Discussion

3.1. Cultivation with Complex Nutrient Source

In order to see the influence of the complex nutrient source on the production of L-lactic acid and 1,3-PD, a series of tests were carried out with sufficient buffering and excluding substrate limitation. Under these conditions, the concentrations of the complex components were varied in order to validate the resulting changes in yield, productivity, and final product titer (Figure 1).

As a first approach to reduce the complex media components, the concentration of peptone (P), yeast extract (YE), and meat extract (ME) in the MRS medium was varied in 24-well plates to produce L-lactic acid with L. casei. Combinations of 0, 3, 5, and 10 g L⁻¹ of the various complex media components were tested [26]. Comparable L-lactic acid titers and productivities were already obtained with a total concentration of complex components of 10 g L⁻¹, instead of the original 25 g L⁻¹. The best result was achieved with the combination of 5 g L⁻¹ yeast extract and 5 g L⁻¹ meat extract. The use of peptone and an increase in the concentration of complex components seems to be inefficient for the cultivation of L. casei. As a reference for further optimization experiments, the modified MRS medium with 5 g L⁻¹ yeast extract and 5 g L⁻¹ meat extract was used as a complex nutrient source (Figure 1a). To clarify the nutrient requirements, a nutrient limitation was created by reducing the yeast extract to 3 g L⁻¹ and meat extract to 0 g L⁻¹, which drastically reduced L-lactic acid production. Using the modified medium, a final L-lactic acid titer of approximately 99 g L⁻¹, a yield of 0.91 g g⁻¹ (w/w) ± 0.02 (w/w), and a total glucose consumption occurred after 50 h of cultivation. Under nutrient limitation (red, open squares limited MRS-medium with 3 g L⁻¹ yeast extract), after a cultivation time of 70 h, still 65 g L⁻¹ glucose was measured as well as approximately 40 g L⁻¹ L-lactic acid. This deficiency in L-lactic acid production and glucose consumption should be compensated by adding different nutrients such as ammonium sulfate and urea as N-sources, trace elements, phosphate, and MgSO₄. However, no improvement could be achieved through these additions.

In the production of 1,3-PD using C. butyricum, an increase in yeast extract concentration was accompanied by increased 1,3-PD product formation and higher glycerol conversion (Figure 1b). Accordingly, with 1 g L⁻¹ yeast extract, the lowest final concentration of 1,3-PD (22.3 g L⁻¹) was obtained, whereas the highest final concentration (29 g L⁻¹) was achieved with 10 g L⁻¹ yeast extract. Using 5 and 10 g L⁻¹ yeast extract, the glycerol was almost completely consumed and the highest yields of 0.57 g·g⁻¹ were achieved, whereby the yields of all preparations did not differ significantly from one another (0.53–0.57 g g⁻¹). The maximum productivity increased accordingly with the yeast extract concentration applied. Using 1 g L⁻¹ yeast extract, the maximum productivity was 2.6 g L⁻¹ h⁻¹, using 4 g L⁻¹ yeast extract the maximum productivity was 6 g L⁻¹ h⁻¹, and it was highest with the starting concentration of 10 g L⁻¹ yeast extract with 7.3 g L⁻¹ h⁻¹. Compared to the literature, the achieved yields, as well as the overall productivities after 12 h cultivation time (increasing from 1.4 g L⁻¹·h⁻¹ using 1 g L⁻¹ yeast extract till 2.5 g L⁻¹ h⁻¹ using 10 g L⁻¹ yeast extract), fit very well [23]. Very good final product titer (28.4 g L⁻¹) and maximum productivities (6 g L⁻¹ h⁻¹) were achieved with 4 g L⁻¹ yeast extract, which explains why this was chosen as the concentration for future cultivation experiments.

3.2. Analysis and Replacement of Complex Nutrient Source

A major problem with the use of unpretreated/untreated nitrogen-rich agricultural residues as a nutrient source is the insufficient degree of hydrolysis or availability of free amino acids and small peptides. Although enzymatic hydrolysis is still the predominant hydrolysis method, the study of Brock et al. shows that this problem can be overcome by chemical hydrolysis using 3 M H₂SO₄. Furthermore, it has been shown that it is not the type of agricultural residue but only its degree of hydrolysis that determines the fermentation efficiency of S. inulinus and the production of D-lactic acid so that the yeast extract could be successfully replaced by a variety of hydrolyzed agricultural residues [7]. The aim of this study is to elucidate the nutritional requirements for the fermentation processes of L-lactic acid and 1,3-propanediol, as well as the replacement of the complex nutrient source by hydrolyzed rapeseed meal. It is, therefore, important to analyze the free amino acids of yeast extract, meat extract, rapeseed meal, and hydrolyzed yeast extract, hydrolyzed meat extract, and hydrolyzed rapeseed meal (Figure 2).

A higher degree of hydrolysis and a higher free amino acid content of yeast extract led to the efficient utilization of organic nitrogen needed for protein biosynthesis. The free amino acid concentration is crucial for the high D-lactic acid concentrations obtained by the cultivation of S. inulinus [11]. Since the transport mechanisms in S. inulinus were found to be insufficient for the uptake of peptides, the free amino acid concentration is crucial for the high D-lactic acid concentrations obtained by the cultivation of S. inulinus. This has been shown to be the limiting factor in this process. Therefore, experiments were carried out to determine the effect of using RM-hydrolyzate (RMH) and YE-hydrolyzate (YEH) instead of yeast and meat extracts (Figure 3).

In all experiments with hydrolyzates, the total nitrogen content was comparable to 5 g L⁻¹ yeast extract and 5 g L⁻¹ meat extract during the cultivation of L.casei, and comparable to 4 g L⁻¹ yeast extract during the cultivation of C. butyricum.

Figure 3 shows that a slight increase in L-lactic acid production was observed when using only the chemical hydrolyzates of rapeseed meal (RMH), which were produced by acid-catalysis, compared to the 3 g L⁻¹ yeast extract. When hydrolyzates were used without other supplements such as YE or ME, the average L-lactic acid concentration reached 51.4 g L⁻¹ after 68.5 h, which was half the concentration of the reference cultivation that included 5 g L⁻¹ YE and 5 g L⁻¹ ME [26]. B vitamins are essential for the metabolism of Lactobacilli and have already been studied in relation to their growth, productivity, and final concentration of L-lactic acid [11,26]. B vitamins are included in yeast and meat extracts, and it is assumed that the B vitamins will be destroyed during the hydrolysis process [27].

By totally replacing the YE with hydrolyzed YE and hydrolyzed RM, a strong influence on the whole production process of C. butyricum was observed. The experiments with the hydrolyzates exhibited a lag phase that was prolonged by approximately 4 h compared to the reference. A comparison of the optical densities (OD_605nm) showed that the exclusive use of RMH resulted in about 73% less biomass after 12 h than the reference with yeast extract. In order to investigate the effect of the selected hydrolysis method on the hydrolyzate with regard to destroyed nutrients and how the use of the hydrolyzate influences the cultivation process, hydrolyzed yeast extract (YEH) was used. The use of YEH resulted in about 36% less biomass after 12 h than the reference, with otherwise the same slope of the growth curve. In accordance with the growth, the prolonged lag phase when using the hydrolyzates can also be recognized in the production of 1,3-PD (Figure 3b). Interestingly, after 24 h of cultivation, titers of 1,3-PD comparable to the reference were achieved. It is, therefore, possible to replace the yeast extract completely with RMH, with a loss of productivity.

For the production of L-lactic acid using L. casei, the hydrolyzate appears to lack an essential substance, e.g., vitamins, which are destroyed by the hydrolysis, inhibited, and are no longer sufficiently available. In the production of 1,3-PD, the growth of C. butyricum and the productivity are strongly influenced, but after hydrolysis, no substances seem to be essentially missing or completely inhibiting the process, so that after 24 h, comparable titers of 1,3-PD as the reference using 4 g L⁻¹ yeast extract are achieved.

3.3. Nutrient Requirements

In contrast to D-lactate [7], the free amino acids (AAs) produced by hydrolysis do not appear to be sufficient, so further investigations into the nutrient requirements of L. casei and C. butyricum are necessary. Target supplementation based on the nutrient requirements could increase the L-lactic acid and 1,3-PD production, as well as improve understanding of the production processes to increase the use of alternative raw materials and thus reduce the production costs, which is an important point for the industrial implementation of these processes.

3.3.1. B Vitamin Requirements

To clarify the B vitamin requirements of L. casei, a nutrient limitation was generated by reducing the yeast extract to 3 g L⁻¹, which drastically reduced lactate production (Figure 3a). This limitation should be compensated for by adding B vitamins at the concentrations contained in 5 g L⁻¹ YE and 5 g L⁻¹ ME (1.25 mg L⁻¹ B1, 500 µg L⁻¹ B2, 10 mg L⁻¹ B3, 2.5 mg L⁻¹ B5, 1 mg L⁻¹ B6, 20 µg L⁻¹ B7, 750 µg L⁻¹ B9, 1 µg L⁻¹ B12, and 75 mg L⁻¹ inositol). The cultivation time was prolonged from 68 h to 143 h (Figure 4a). The addition of the entire vitamin solution increased L-lactic acid production from 65 g L⁻¹ in the limited MRS medium to 88 g L⁻¹ due to the positive effect of niacin (vitamin B3). By adding niacin alone, a L-lactic acid concentration of 89 g L⁻¹ was achieved, which corresponds to 90% of the reference with 5 g L⁻¹ yeast extract and 5 g L⁻¹ meat extract.

For the production of 1,3-PD, the effect of vitamin solutions in combination with the hydrolyzates as a yeast extract substitute was investigated in different variants (Figure 4b). The vitamins were divided into three solutions: Vit1 = B1, B2, B3, B5, and B6; Vit2 = B9; and Vit3 = B7 and B12. The vitamin solutions had higher concentrations according to the concentrations in 4 g L⁻¹ YE (0.856 mg L⁻¹ B1, 0.832 mg L⁻¹ B2, 11.164 mg L⁻¹ B3, 4.936 mg L⁻¹ B5, 1.04 mg L⁻¹ B6, 0.367 mg L⁻¹ B7, 2.888 mg L⁻¹ B9, 0.019 mg L⁻¹ B12) [27]. The addition of vitamin solutions to the RMH does not affect growth, substrate consumption, or 1,3-PD production (Figure 4b). Compared to cultivation using the RMH, there is no increase in productivity, even when the vitamins present in the yeast extract are supplemented. With an initial glycerol concentration of 50 g L⁻¹, this strain does not require any vitamins in the batch process.

3.3.2. Amino Acids Requirements

Similar to the investigation of the B vitamin requirement, the requirement for amino acids (AA) was determined by adding an AA mix to the medium. Similar to the B vitamin mix, the AA mix is based on the composition and concentration of the AA profile of 5 g L⁻¹ YE and 5 g L⁻¹ ME for L-lactic acid production with L. casei (Figure 5a). Without using a complex media mixture and by supplementing the medium with the amounts of amino acids present in 5 g L⁻¹ YE and 5 g L⁻¹ ME (Figure 5a: all AA), a final titer of 7 g L⁻¹ L-lactic acid was achieved after 65.5 h, while the reference cultivation with 5 g L⁻¹ YE and 5 g L⁻¹ ME resulted in approximately 99 g L⁻¹ L-lactic acid, a total glucose consumption, and a yield of 0.91 g g⁻¹ (w/w) ± 0.02 (w/w). As shown in Figure 5a, the sole addition of free AA to the RMH or the limiting MRS medium does not lead to an increase in productivity or comparable L-lactic acid final titers, as achieved under reference conditions with 5 g L⁻¹ YE and g L⁻¹ ME. In prior research, the titer of free AAs in the medium was analyzed post-cultivation [26]. The concentration of tryptophan was found to be under the detection limit. Additionally, the analytical method employed could not detect the AAs arginine and cysteine. To assess the impact of these three AAs, a solution containing 150 mg L⁻¹ arginine, 50 mg L⁻¹ cysteine, and 80 µg L⁻¹ tryptophan (Arg/Cys/Trp) was supplemented to the medium without any addition of complex media sources. With this approach, the achieved concentration of 30.9 g L⁻¹ L-lactic acid was still lower than using the limited MRS medium with 3 g L⁻¹ YE, which reached approximately 40 g L⁻¹ L-lactic acid.

The influence of the addition of amino acid solutions together with the produced hydrolyzates as a yeast extract substitute was investigated in different variants. Based on analyses of the yeast extract used [27], the amino acids to be tested were used according to the total nitrogen content of 4 g L⁻¹ yeast extract. As the solubility and heat stability of the individual amino acids differs, they were divided into three solutions according to their properties: amino acid solution 1 (AA1) = alanine, glycine, valine, leucine, isoleucine, threonine, serine, proline, methionine, phenylalanine, lysine, histidine, cysteine; amino acid solution 2 (AA2) = Tryptophan, Tyrosine; Cysteine was added to AA1, which was not found in the yeast extract due to the analytical method but is an important amino acid for many bioprocesses. Due to its sensitivity, AA2 was sterile-filtered. The utilization of RMH and the utilization of the three amino acid solutions resulted in around 73% and 45% less biomass production after 12 h compared to the reference with 4 g L⁻¹ yeast extract. The cultivations with the RMH showed a lag phase extended by approx. 4 h and OD_605nm values lower than 4, whereas the reference had an OD_605nm of 11. The addition of AA1 to the RMH showed a strong inhibition of growth and only low OD_605nm values up to OD_605nm 2 were measured in the first 12 h. During cultivation using RMH supplemented with all AA and the cultivation using RMH and AA2 achieved the highest OD_605nm values of 15 and 13. Biomass formation during the first 8 to 10 h was comparable to the reference. As the 1,3-propanediol production of this strain is growth-coupled [23], a similar behavior can be seen in the 1,3-PD production (Figure 5b). In the reference, with the cultivations with RMH, the cultivations with RMH supplemented with all AA solutions, and the cultivations with RMH supplemented with AA2, product formation already began after 4 h and reached the highest concentrations (27–28 g L⁻¹) within the first 12 h. The use of all amino acids without hydrolyzate led to an early but reduced formation of 1,3-propanediol (1,3-PD) and reached concentrations of 17 g L⁻¹ after 12 h. In accordance with the growth, significantly lower concentrations (up to 7 g L⁻¹) were formed using RMH, and cultivation occurred using RMH supplemented with AA1. The yields of all variations are between 0.55 and 0.58 g g⁻¹ and are in similarly high ranges of the theoretical maximum yield. However, the maximum productivity differs and varies between 2.5 g L⁻¹ h⁻¹ (use of all AAs, Figure 5b: upside down open green triangle) and 6.6 g L⁻¹ h⁻¹ (reference 4 g L⁻¹ YE, black closed square), whereby a maximum productivity of 6.5 g L⁻¹ h⁻¹ was achieved with the combination of RMH and AA2. The cultivation processes show that the fixed amount of 4 g L⁻¹ YE can be replaced equally well by the simultaneous use of RMH and the amino acids contained in the yeast extract. When comparing the successfully tested variants, it becomes clear that the addition of AA2 was crucial. Prior analysis of the hydrolyzates revealed that the amino acid tryptophan is destroyed by the selected acid digestion method and is therefore missing in the hydrolyzates (Figure 2). With the addition of this amino acid in AA2 and RMH, the results of the references could be reproduced, even without the main proportion of the other amino acids. However, if tryptophan was completely absent, this had a strong effect on growth and the course of product formation. With the hydrolyzates alone and the variation of hydrolyzate with AA1, the lag phase was greatly delayed and extended by about twice as much. Growth was delayed even more by the addition of the easily soluble amino acids. Overall, tryptophan proved to be essential for C. butyricum AKR102a. Compared to the reference, it was also noticeable that the cultivation with hydrolyzed YE took a similar but temporally strongly delayed course, which also illustrates the influence of the digestion method due to the absence of tryptophan (Figure 3b, open black squares).

3.4. Replacement of Complex Nutrient Source

Based on the previous results, a final experiment (Figure 6) was carried out in order to see if it is possible to replace the complex nutrient sources completely.

In the case of L-lactic acid production, yeast extract, peptone, and meat extract were completely replaced by rapeseed meal hydrolyzate (RMH) with a total nitrogen content of 1 g L⁻¹ (equivalent to the total nitrogen content of 10 g L⁻¹ YE), B vitamins (based on the concentrations of 5 g L⁻¹ YE and 5 g L⁻¹ ME, see subchapter 3.3.1), and mineral salts (25 mg L⁻¹ FeSO₄·7H₂O, 10 mg L⁻¹ ZnSO₄·7H₂O, 50 mg L⁻¹ MnCl₂, 500 mg L⁻¹ MgSO₄, 500 mg L⁻¹ MgCl₂, 500 mg L⁻¹ CaCl₂), 1 g L⁻¹ cysteine, and 0.2 g L⁻¹ tryptophan.

In all approaches the initial 116 g L⁻¹ glucose was completely metabolized within 70 h. In the production of L-lactic acid, the complex nitrogen sources could be completely replaced by rapeseed meal hydrolyzate (Figure 6a). The lag phases of the approaches differ, but after about 24 h the glucose was converted to a final concentration of 105 g L⁻¹ ± 1.8 L-lactic acid with a comparable maximum productivity, even without any addition of yeast extract or meat extract.

Under glucose limitation, homofermentative lactic acid bacteria tend to produce by-products such as formate or acetic acid via alternative pyruvate catabolic pathways [28,29]. According to Slavica et al., the production of acetic acid increased under glucose limitation in a MRS medium [30]. Under reference conditions, 5.2 g L⁻¹ acetate was produced, in a limited MRS medium with 3 g L⁻¹ YE, and the addition of RMH 4.5 g L⁻¹ acetate was produced after glucose was completely consumed, whereas 0.8 g L⁻¹ acetate were produced when YE was completely replaced. The yield of 0.93 g g⁻¹ is, therefore, slightly higher than the yield of 0.91 g g⁻¹ under reference conditions. According to Wang et al., YE is an efficient nutrient for high lactic acid productions, which results in high costs; therefore, the replacement with an economical alternative nitrogen source would reduce the costs. In addition, vitamins and minerals have significant effects on lactic acid fermentation. Among others, sodium sulfate, sodium acetate, K₂HPO₄, MnSO₄·4H₂O, and FeSO₄·7H₂O were found to be significant [31]. Hujanen et al. also studied the composition of the medium in order to replace expensive yeast extract with cheaper sources of nitrogen and amino acids using two homofermentative strains, Lactobacillus casei NRRL B-441 and Lactobacillus casei subsp. rhamnosus NRRL B-445. The results showed that lactic acid production was significantly influenced by the type and initial concentration of the nitrogen source, and barley malt sprouts and grass extract were the best economic alternatives. The final lactic acid concentration was even higher when sprout extract (100 g L⁻¹ lactic acid) or grass extract (103 g L⁻¹ lactic acid) was used together with 0.4% yeast extract, compared to 2.2% yeast extract used as the sole source of nitrogen and growth factors (95 g L⁻¹ lactic acid) [32]. These findings match very well with the results presented in this paper.

In conclusion, it could be shown that the complex components of yeast extract, peptone, and meat extract could be completely replaced by chemically digested rapeseed meal and supplementation with B vitamins, mineral salts, cysteine, and tryptophan to achieve comparable productivities and final titers of 105 g L⁻¹ L-lactic acid.

The 1,3-PD production was transferred to reactor scale (Figure 6b). Based on the previous results, in this final experiment, a cultivation with 5 g L⁻¹ yeast extract was used as a reference, compared to 1 g L⁻¹ yeast extract and RMH (with the total nitrogen content of 4 g L⁻¹ YE), and with complete replacement of the YE by RMH (with the total nitrogen content of 5 g L⁻¹ YE), 0.0472 g L⁻¹ tyrosine and 0.0316 g L⁻¹ tryptophan.

The reduction to 1 g L⁻¹ YE with the addition of RMH (open, green circle) shows that the maximum productivity with 3.9 g L⁻¹ h⁻¹ almost corresponds to the maximum productivity of the reference and 96 g L ¹ 1,3-PD could be produced after 72 h of cultivation. The transfer to reactor scale shows good agreement with the results obtained by Wilkens with a maximum titer of 93.7 g L⁻¹ 1,3-propanediol [23,33].

Wang et al. obtained an average titer of 84.62 g L⁻¹ 1,3-PD and a yield of 0.52 g g⁻¹ during sequential fed-batch cultivation using C. butyricum DL07 and a medium containing 2 g L⁻¹ YE and a feeding solution consisting of 80% glycerol and 40 g L⁻¹ YE. After a stable sequential fed-batch cultivation, YE was replaced by corn steep liquor (CSL). The results showed that corn steep liquor powder can be used as an organic nitrogen source without the loss of 1,3-PD concentration, yield, and productivity [34]. This was also found by Wischral et al., who aim to optimize 1,3-PD production from crude glycerol by C. beijerinckii DSM 791 through fermentation process optimization and metabolic engineering [35].

Oh et al. also pointed out that the nitrogen source is a key factor for the commercial production of 1,3-PD. Klebsiella pneumoniae requires expensive complex nitrogen sources such as YE for 1,3-PD production and cell growth, and the effect of CSL supplementation instead of YE was conducted. The use of 1% CSL supplementation resulted in similar titers of 1,3-PD using 1 g L⁻¹ YE as a control [36].

Maina et al. used soybean cake hydrolyzates as nutrient-rich fermentation supplements in order to replace the complex nutrient supplements required by C. freundii strains for the production of 1,3-PD. The enzymatically produced hydrolyzates were studied in batch and fed-batch cultivations using crude glycerol as substrate. The highest 1,3-PD concentration of 47.4 g L⁻¹ was obtained with a yield and productivity of 0.49 g g⁻¹ and 1.01 g L⁻¹ h⁻¹ [37].

However, the experiment with complete replacement of the yeast extract (Figure 6b: RMH + TYR + TRP, closed, blue asterisk) with RMH shows that the growth phase under the supplementation almost corresponds to the growth phase of the reference using 5 g L⁻¹ YE as a nutrient source, but in the further course of production, apparently unknown substances limit the productivity. The maximum productivity is significantly reduced, so that after 48 h of cultivation only 63 g L⁻¹ 1,3-PD could be achieved. Even continued cultivation did not lead to any further product formation. After 48 h 63 g L⁻¹ 1,3-PD with a productivity of 1.31 g L⁻¹ h⁻¹ and 4.5 g L⁻¹ acetate, 13 g L⁻¹ butyric acids as side products were produced.

4. Conclusions

Yeast extract is an important cost factor in many bioprocesses. It provides the micro-organisms with the nutrients they need for their growth and production. The aim of this project was, therefore, to replace yeast extract with alternative nutrient sources in order to reduce production costs. Using the example of the biotechnological production of L-lactic acid and 1,3-propanediol, which are used, among other things, as starting materials for bio-based plastics and chemicals, or as preservatives in food, cosmetics, and pharmaceuticals, the nutrient requirements of the microorganisms used could be partially clarified. In the production of 1,3-propanediol, it has been shown that yeast extract can be replaced for the most part (80%), with the same final product concentration and maximum productivity. In the production of L-lactic acid, the expensive complex nitrogen sources (yeast extract and meat extract) were completely replaced by the inexpensive, chemically digested rapeseed meal, supplemented with B vitamins, mineral salts, cysteine, and tryptophan. The same final product concentration of L-lactic acid was achieved with comparable maximum productivity. By replacing yeast extract and meat extract with inexpensive, chemically digested rapeseed meal, the costs of biotechnological production can be reduced. This helps to increase the competitiveness of biotechnological production processes compared to production processes based on fossil raw materials.