Skip to main content
Download PDF

Development of a hybrid bio-purification process of lactic acid solutions employing an engineered E. coli strain in a membrane bioreactor

Biotechnology for Biofuels and Bioproducts volume 17, Article number: 48 (2024) Cite this article

Abstract

Background

A potential alternative to lactic acid production through sugar fermentation is its recovery from grass silage leachate. The separation and purification of lactic acid from fermentation broths remain a key issue, as it amounts to up to 80% of its industrial production cost. In this study, a genetically engineered E. coli strain (A1:ldhA), that cannot catabolize lactic acid, has been used to selectively remove impurities from a synthetic medium comprising typical components (i.e., glucose and acetic acid) of green grass silage leachate. A systematic approach has been followed to provide a proof-of-concept for a bio-purification process of lactic acid solutions in a membrane bioreactor operating in semi-continuous mode.

Results

The synthetic medium composition was initially optimized in shake-flasks experiments, followed by scale-up in bench-scale bioreactor. Complete (i.e., 100%) and 60.4% removal for glucose and acetic acid, respectively, has been achieved in batch bioreactor experiments with a synthetic medium comprising 0.5 g/L glucose and 0.5 g/L acetic acid as carbon sources, and 10 g/L lactic acid; no lactic acid catabolism was observed in all batch fermentation tests. Afterwards, a hybrid biotechnological process combining semi-continuous bioreactor fermentation and ultrafiltration membrane separation (membrane bioreactor) was applied to in-situ separate purified medium from the active cells. The process was assessed under different semi-continuous operating conditions, resulting in a bacteria-free effluent and 100% glucose and acetic acid depletion, with no lactic acid catabolism, thus increasing the purity of the synthetic lactic acid solution.

Conclusions

The study clearly demonstrated that a bio-purification process for lactic acid employing the engineered E. coli strain cultivated in a membrane bioreactor is a technically feasible concept, paving the way for further technological advancement.

Background

Lactic acid (LA) is a biomolecule with two optical isomers in nature, D- and L-LA. It is widely used in industry, i.e., in medical and pharmaceutical, cosmetic, food, polymer and textile sector [1, 2]. Although, LA can be produced by a chemical synthesis process, 90% of the LA worldwide demand is produced through sugars’ fermentation [3]. The specifications of LA purity level depend on its industrial use, i.e., pharmaceutical industry requires higher than 90% purity, while food-grade LA has 80–90% purity [2]. Due to the physicochemical properties of LA, such as high water affinity [1], as well as the wide range of impurities in the fermentation broth (e.g., residual sugars, heavy metals and organic acids), downstream purification remains a significant bottleneck of the LA production process through fermentation [4]. Specifically, it has been estimated that LA purification may comprise between 30% and 40% [5] and 80% [6] of the total LA production cost, depending on the targeted purity level. Thus, the efficiency of the downstream purification process is a key issue that research has focused on. [2].

Currently, LA produced from sugars’ fermentation at industrial scale requires a series of downstream processing steps of the fermentation broth and usually include: cells’ separation, precipitation in the form of slightly soluble lactate salts, e.g., by addition of Ca(OH)2, filtration, acidification, carbon adsorption, evaporation, and crystallization process [1]. Precipitation using Ca(OH)2 usually requires an excess of H2SO4 to recover LA, which results in the production of high quantities of CaSO4 as a by-product, and subsequently a lower purity solution of LA. Researchers have studied various alternative purification methods, such as membrane technology (ultrafiltration—UF, nanofiltration—NF, electrodialysis—ED), ion-exchange/adsorption, reactive distillation, and hybrid short path evaporation [2, 4, 7,8,9]. However, in most cases, an integrated process is required to reach high levels of purity. For instance, a downstream purification system may include a combination of different membrane separation processes (such as UF and NF), ion exchange, and vacuum-assisted evaporation [10] to reach a purity level of > 99.5%. Another, hybrid-membrane purification process that led to high-quality LA (> 90.0% purity) combined concentration, ultrasonic liquid–liquid extraction, phase separation, and distillation to recover and purify LA from the fermentation broth [11]. Although, high LA purity was achieved, the process was proven quite complex with low recovery rate, mainly due to the liquid–liquid extraction process.

LA recovery from green grass silage leachate, a by-product of green grass biorefinery, is a potentially alternative production source. Grass silage leachate is produced during grass silage maturation (fermentation), and its yield may reach up to 300–400 L per ton of fresh grass silage. LA is a major component of grass silage leachate at a typical concentration range between 22 and 38 g/L [12]. Since, a consortium of microorganisms is growing during the ensiling process both isomers D- and L-LA are present in the grass silage leachate [13, 14]. Towards recovering LA from grass silage leachate, Ecker et al. developed a LA recovery process that consisted of UF, two-step NF, ED, and reverse osmosis (RO) processes [15]. However, impurities that have a high chemical affinity to LA, such as acetic and butyric acid, remained in this LA enriched stream.

Novel biologically based removal and purification processes have gained increasing attention as potential alternatives to the conventional purification processes. Such processes may take advantage of the intrinsic high selectivity of biological catalysts (enzymes) to convert or catabolize specific impurities present in the process solution. The high selectivity is of significant importance, especially when the impurities have high chemical affinity to the main component, i.e., mixtures of short-chain organic acids or of different oligo-saccharides. Moreover, biologically based purification processes may also purify optical isomers that practically have the same physical and chemical properties; thus, their physicochemical separation can be only achieved using chiral chromatograph, that is an expensive, highly specialized method that cannot be applicable to all types of optical isomers [16]. For instance, a novel recycling process for the polylactic acid polymer (PLA), which is based on an alkaline thermochemical hydrolysis process followed by a biological purification step of the resulting lactate solution, has been suggested. Lactate solution produced by PLA hydrolyzation consists of D- and L-lactate, where L-lactate is dominant. An E. coli strain (DC1001) was genetically modified to selectively catabolize D-lactate without consuming L-lactate. This biologically based purification process was employed to a D- and L-LA solution and resulted in a highly purified L-LA solution (> 99%) [17, 18]. The proof-of-concept of the biological purification process was exhibited in a batch experiment of 1 L flask containing 125 g/L D- and L-LA, that resulted in almost complete (i.e., 100%) removal of D-LA after 30 h of aerobic incubation at 37 °C. The authors did not study the possibility of reusing the E. coli cells or different culture modes (i.e., fed-batch, semi-continuous or continuous) that would facilitate process scale-up.

Membrane Bioreactor (MBR) technology is a well-established hybrid technology in the field of wastewater treatment, advantageously combining biological processes and membrane separation in a single process step. Other advantages include the small plant footprint, high quality of effluent, efficient bacteria and viruses rejection, and stable operation even in cases of high or shock loadings [19]. MBR technology may be also employed to overcome challenges of single-strain biotechnological production (fermentation) processes, such as cells’ wash-out, removal of toxic metabolic by-products, operation in continuous mode etc. Taleghami et al. [20] studied the production of LA in a laboratory scale MBR employing Lactobacillus bulgaricus grown on lactose as carbon source. Both UF and NF membranes were used to separate the LA solution from the fermentation broth achieving a maximum productivity of 17.1 g L−1 h−1. MBR process has been also used for the production of bioethanol [21] and polyhydroxyalkanoates (PHA) [22] employing organic wastes or by-products as substrate. Bioethanol production through fermentation of wheat straw hydrolysate was studied using an immersed MBR and a recombinant, xylose-utilizing Saccharomyces cerevisiae strain achieving continuous fermentation operation and process intensification [21]. An immersed MBR was also used for the production of PHA through a semi-continuous cultivation and recovery of Cupriavidus necator bacteria grown on food-derived Volatile Fatty Acids (VFA) [22]. The semi-continuous mode resulted in increased bacterial tolerance to VFA concentration and maximum PHA production. Recently, Raveschot et al. [23] used Lactobacillus helveticus strains to produce bioactive peptides (BAP) in a continuous fermentation mode in a 3 L MBR employing a hollow-fiber microfiltration membrane. The bioreactor was continuously fed with fresh substrate at the same flow rate that the permeate was withdrawn. The integrated continuous process exhibited threefold increase of BAP productivity compared to batch fermentation allowing also the production and purification of the BAP in a single process step.

In this study, an engineered strain of E.coli (A1:ldhA), incapable of catabolizing LA [24], has been used to selectively catabolize and remove impurities in a synthetic medium comprising LA, glucose, acetic acid (AA), and salt (NaCl) as typical components of green grass silage leachate. Although, these impurities are not the only one found in real grass silage leachates, they are representative of the main classes of organic impurities (simple sugars and short-chain organic acids) found in such streams. The main goal is to demonstrate the technical feasibility (proof-of-concept) of the novel bio-purification process employing the E.coli (A1:ldhA) strain in an MBR with synthetic LA solutions. Therefore, it was decided that the synthetic LA solution to be rather simple and well-defined, to facilitate the concept’s development. Initially, flask tests were performed to define an optimum synthetic medium composition, and to prove the concept of the bio-purification of LA in bench-scale flask tests. Kinetic data were collected before fermentation scale-up. Then, batch fermentation was validated in bench-scale bioreactor tests, followed by semi-continuous fermentations in an MBR; the membrane performance during the MBR operation was evaluated as well, to validate the integrated MBR bio-purification process.

Materials and methods

Strain and media

E.coli (A1:ldhA), a genetically modified strain from wild type E. coli W3110, was used for this study provided by Dr. Joseph B. Sweeney (School of Biosystems Engineering and Food Science, University College Dublin). The strain is genetically modified to be incapable of utilizing LA, but capable to selectively consume organic acids and simple sugars. The strain stocks were stored in 25% v/v pure glycerol at -80 °C and plated on Tryptic Soy Agar (TSA) with 50 μg/mL kanamycin prior to each broth culture. A single colony was inoculated into Tryptic Soy Broth with 50 μg/mL kanamycin. Minimal M9 medium was used both at the 2nd pre-culture and the main culture. The M9 medium was composed of: 6.8 g Na2HPO4, 3.0 g KH2PO4, 1.0 g (NH4)2SO4, 0.5 g NaCl, 4 mL of 0.1 M MgSO4 × 7 H2O, 100 μL of 1 M CaCl2 × 2 H2O and 1 mL of SL-10 trace elements solution per liter deionized water. SL-10 solution contained 1 mL 25% v/v HCl, 1.5 g FeCl2 × 4 H2O, 70.0 mg ZnCl2, 100 mg MnCl2 × 4 H2O, 3.0 mg H3BO3, 190 mg CoCl2 × 6 H2O, 2.0 mg CuCl2 × 2 H2O, 24 mg NiCl2 × 6 H2O, and 36 mg Na2MoO4 × 2 H2O, per liter deionized water [25]. Culture media were supplemented with glucose, AA and LA (L-LA isomer) to simulate the composition of a LA solution, originating from green grass leachate after some preliminary downstream purification steps, i.e., UF for removal of suspended solids and cation exchange softening. Glucose and AA were selected as typical impurities that may be found in such process streams [12].

Experimental procedure

Shake flasks experiments

A series of preliminary tests were conducted to assess the effect of the reduction to 50% and 25% of the original amount of the supplied macronutrients included in M9 minimal medium (i.e., Na2HPO4, KH2PO4, and (NH4)2SO4). Afterwards, the effect of different concentrations of glucose, AA, LA and NaCl on the E. coli strain performance in flask fermentations was assessed. First, the effect of glucose and AA concentrations was assessed for approx. 0.5, 1.0 and 2.0 g/L of glucose, and 0.5, 1.0 and 2.0 g/L of AA, respectively; the LA concentration was chosen to be approx. 5 g/L. To assess the E. coli strain tolerance to increasing LA concentrations batch experiments were performed, at the following LA concentrations 5, 10, 20, 30, 35 and 40 g/L, and approx. 0.5 g/L of both glucose and AA. The highest LA concentration that did not inhibit bacteria growth was selected to assess the E. coli strain salinity tolerance using three different NaCl concentrations (0.1, 0.5 and 1% w/v), in the presence of approx. 0.5 g/L of both glucose and AA.

In each batch experiment, a negative control flask was also prepared, comprising LA as sole carbon source, to verify that the E. coli strain could not catabolize LA. Each experiment lasted for 48 or 72 h with sampling every 24 h to measure the optical density at 600 nm (OD600), the concentration of LA, glucose and AA, and to monitor the pH. Total nitrogen (TN) and total organic carbon (TOC) were also measured to verify that no nitrogen limitation takes place and to estimate the total organic content of the fermentation medium. To complete the growth optimization in shake-flask tests, kinetic data were collected as a final step before scaling-up the process in a bench-scale bioreactor. A 12-h staggered starting time was applied to two identical fermentation batches, to obtain data during 24 h of sampling at hourly intervals (i.e., Batch 1 for the time period of 0–12 h; Batch 2 for the time period of 12–24 h).

Overnight pre-culture was performed at 37 °C and 200 rpm, in a shaking incubator (Lab Companion SI-600R). The 2nd pre-culture was incubated at 37 °C for 22 h with shaking at 200 rpm. The medium and the glassware used were sterilized in an autoclave (Raypa Steam Sterilizer) at 121 °C for 20 min. Minimal M9 media were used for shake-flask fermentation, supplemented with glucose, AA, LA, and NaCl according to the aforementioned experimental methodology. 250 mL flasks, containing 100 mL of fermentation medium, were inoculated with 4 mL (4%) inoculum from the 2nd pre-culture and were incubated at 37 °C and 200 rpm. The experiments were performed in biological triplicates.

Batch bioreactor fermentation tests

To gain insight in the scale-up potential of the biologically based purification process, batch cultivations were performed in a 3 L bench-scale bioreactor (BioFlo® 120, Eppendorf S.E.) with a working volume of 1.8 L. The bioreactor was inoculated with 4% v/v of the 2nd pre-culture after 22 h cultivation. Culture media contained minimal M9 medium supplemented with 0.5 g/L glucose, 0.5 g/L AA, 10 g/L LA, and 0.1% w/v NaCl. Fermentations lasted either 24 h or 30 h to achieve complete AA consumption. Temperature was set to 37 °C, agitation to 450 rpm and air flow rate to 1.7 vvm. The pH was controlled at 7.0 by automatic addition of 2.5 N NaOH and 2.5 N H2SO4 solution. Dissolved oxygen (DO) and OD600 were continuously monitored by the respective probes (Eppendorf Optical DO Sensor, ISM® and Hamilton Dencytee Unit 225). An hourly sampling frequency during 0–6 h and 20–28 h time periods was applied; the collected data were used to calculate growth kinetics and the efficiency of glucose and AA consumption rates, which in turn were used to design the MBR semi-continuous process.

Semi-continuous MBR fermentation tests

The semi-continuous MBR fermentation process comprised two stages: initially, the bioreactor was operated in batch mode that lasted from 24 to 34 h, employing the operating parameters that were tested in the batch bioreactor fermentations ("Batch bioreactor fermentation tests" section). The aim of this process stage was to increase the concentration of the active biomass in the bioreactor, while achieving almost complete catabolism/removal of both glucose and AA. Then, an intermittent removal of the treated (purified) synthetic LA solution started, followed by feeding of fresh synthetic LA solution (i.e., permeate/feed cycles). A custom-made UF membrane module [26] comprising commercial PVDF hollow fiber (0.03 μm pore size) membranes (PURONR MBR, Koch Separation Solutions) and a peristaltic pump were used to withdraw the treated (purified) synthetic solution from the bioreactor. The membrane module was submerged in the bioreactor vessel and autoclaved in-situ, since it has been certified that the membrane is not affected by the sterilization process [26]. The volumetric flow of the peristaltic pump was regulated to maintain a membrane flux of approx. 30 L/m2/h (LMH). The pump operated until the targeted volume of treated permeate was withdrawn through the membrane. Then, an equal volume of fresh synthetic medium was fed and fermented for different time intervals. The permeate/feed cycles were repeated achieving a semi-continuous operation of the MBR. The permeate/feed cycle time intervals and the operating conditions are summarized in Table 1.

Table 1 Experimental design of semi-continuous MBR experiments
Full size table

Analytical methods

High-performance liquid chromatography (HPLC)

Glucose concentrations were assessed using an HPLC, Shimadzu LC-10AT VP Liquid Chromatograph (Shimadzu, Germany), using a ShodexTM Sugar SH1011 (8.0 mm I.D. × 300 mm) column and a Shimadzu Refractive Index Detector RID-10A. The same method and column were used for estimating the concentration of organic acids in the sample using a Shimadzu SPD-M20A UV detector. Fermentation broth HPLC sample preparation consisted of centrifugation at 4500 rpm for 7 min, filtration through 0.45 μm filters, and appropriate dilution. 5 mM H2SO4 was used as eluent. The flow rate was set to 0.4 mL/min, and the oven temperature to 60 °C.

Total organic carbon (TOC) and total nitrogen (TN)

TOC and TN were measured in a Shimadzu TOC-5000A TOC/TN Analyzer. A calibration curve of 10–100 ppm and 0.5–5 ppm were used for the TOC and TN measurements, respectively. Each fermentation broth sample was centrifuged at 4500 rpm for 7 min and diluted accordingly.

Optical density (OD600) & dry biomass

Optical density was measured for cell growth determination at 600 nm by a photometer (UV1700 Pharmaspec Shimadzu UV–Vis Spectrophotometer) during fermentation experiments in flasks. OD600 in bioreactor experiments were measured online through a Dencytee Arc (Hamilton Company) cell density sensor. OD600 data were recorded every 6 or 12 min. A correlation between OD600 and dry biomass was performed by regular analysis of dry biomass concentrations. This was performed using culture samples of 10 mL that were centrifuged at 4500 rpm for 7 min, the supernatant was disposed and the remaining cell-pellet was washed twice with 1 mL H2O. The pellet was dried at 60 °C for at least 48 h, until recording constant weight.

Results and discussion

Batch fermentations in shake flasks experiments

The first step towards developing the LA bio-purification process was the determination of an optimum synthetic medium composition, which will facilitate the growth of the E. coli strain and the subsequent experimental steps. At the same time, the synthetic medium should be representative of the composition of a typical grass silage leachate after some preliminary downstream processing. To minimize the use of macronutrients in the bio-purification process, which would reduce operational costs in large-scale applications, M9 minimal medium macronutrients concentrations were reduced to 50% and 25% of the original quantities. Results from these preliminary tests showed that the E. coli strain was capable of consuming both glucose and AA within 24 h under all macronutrient concentrations tested (see Additional file 1: Fig. S1). Therefore, it was decided to use 25% of the typical quantity of macronutrients included in M9 minimal media, for the subsequent experimental steps.

Effect of glucose and AA concentration

The E. coli strain growth and ability to catabolize different concentrations of glucose and AA, while leaving LA intact in the medium, was initially assessed. Results from three different shake-flask tests (Test 1, Test 2 and Test 3) with different initial concentrations of glucose and AA are presented in Table 2. The parameter of initial LA concentration was kept constant for all tests, i.e., 5 g/L, while glucose and AA concentrations were approx. 0.5 g/L in Test 1, they were doubled for Test 2 (i.e., approx. 1.0 g/L), and further increased for Test 3 (i.e., approx. 2.0 g/L). Indeed, LA concentration remained constant, roughly 5 g/L, regardless of the initial glucose and AA concentrations tested (Table 2). According to Table 2, the highest growth was recorded at 24 h for Tests 1 and 3, and at 48 h for Test 2, reaching an OD600 value of 0.93 ± 0.04, 0.67 ± 0.02, and 1.12 ± 0.02, respectively. At 72 h, it is obvious that all fermentations had entered stationary phase. The lowest TOC concentration within the culture medium was achieved at the end of the exponential growth phase; 2136 ppm for Test 1, 2343 ppm for Test 2, and 3228 ppm for Test 3. TN measurements showed (Table 2) that there is no nitrogen limitation until 72 h when TN was 24 ppm, 18 ppm, and 39 ppm for Tests 1, 2 and 3, respectively.

Table 2 Data from batch fermentation tests in shake flasks with different initial concentrations of glucose and AA
Full size table

The concentration of glucose and AA in the synthetic LA solution was analyzed to assess their removal efficiency. Initial glucose concentrations were slightly higher than the targeted values, which did not seem to affect the experimental procedure. Complete glucose consumption was observed at 48 h for Test 3 (2.33 ± 0.04 g/L initial glucose concentration), while glucose was consumed within 24 h for Test 1 (0.61 ± 0.03 initial glucose concentration) and Test 2 (1.24 ± 0.03 initial glucose concentration). Due to glucose consumption, AA was excreted in the culture media as described previously [27]; in fact, the higher the initial glucose concentration the higher the AA production. Similar behavior was observed in Test 3; the E. coli cells first consumed glucose inhibiting the consumption of less preferred carbon sources such as AA, following a classic diauxic growth pattern. In addition, acetate overflow was also observed, meaning that part of the uptaken glucose into E. coli cells was first catabolized into AA and excreted into the medium. According to the literature, the excreted AA can be used by E. coli for cell growth after glucose consumption, named “acetate switch” [28]. However, in case of Test 3, the production of AA reduced media pH, possibly hindering further bacteria growth (Fig. 1). In particular, during Test 3, AA increased from 2.00 ± 0.04 g/L to 2.72 ± 0.10 g/L and pH decreased from 7.08 ± 0.02 down to 5.79 ± 0.04 within 72 h, whereas for Test 2, the concentration of AA increased roughly by 0.50 g/L, while pH decreased to 6.59 ± 0.01 at t = 24 h. Afterwards, AA started to be consumed, which resulted to a pH increase from 6.59 ± 0.01 to 8.89 ± 0.06 at t = 72 h [29]. As growth under higher glucose concentrations was hindered likely due to the overflow metabolism of glucose, which results in AA excretion and the corresponding pH reduction, the initial concentration of 0.5 g/L for both glucose and AA was chosen as the optimum for further experimental set ups.

Fig. 1
figure 1

Results from batch fermentation tests in shake flasks with different concentrations of glucose and AA; AA concentration at different fermentation time points compared to fermentation broth pH profile; Test 1: 0.5 g/L glucose + 0.5 g/L AA; Test 2: 1.0 g/L glucose + 1.0 g/L AA; Test 3: 2.0 g/L glucose + 2.0 g/L AA

Full size image

Effect of LA concentration

The second parameter that was assessed, for its effect on the E. coli strain growth, was LA concentration. Six different LA concentrations of 5, 10, 20, 30, 35 and 40 g/L were tested. As illustrated in Fig. 2a, cell growth was lower for concentrations above 20 g/L. In particular, for 30 g/L LA, cells reached an OD600 of 0.62 ± 0.01, while for 35 and 40 g/L LA, a value of 0.54 ± 0.02 and 0.53 ± 0.02 was achieved, respectively. However, for 10 g/L LA, a maximum OD600 value of OD600 = 1.17 ± 0.01 was observed at 24 h. The maximum OD600 measurement that was observed at 48 h for the E. coli strain grown on 20 g/L LA, was comparable to that at 24 h for 5 and 10 g/L of LA. Surprisingly, glucose was consumed within 24 h in all fermentations. On the other hand, the AA consumption seems to be affected, as depicted in Fig. 2b. In the presence of 5 and 10 g/L LA, AA was consumed within 24 h. In the case of 20 g/L LA, AA consumption was delayed and was totally consumed within 48 h, while for all LA concentrations above 20 g/L, AA was not totally consumed, even at longer fermentation times (i.e., 72 h). LA concentration remained constant during all the fermentations, which is very encouraging as it reflects the selective catabolism of other organic compounds under all experimental conditions. Considering that the rate of glucose and AA consumption is significant for the process performance and that the E. coli strain performance declined for LA concentrations above 10 g/L, an LA concentration of 10 g/L was selected for all future experiments.

Fig. 2
figure 2

Results from batch fermentation tests of 72 h in shake flasks with different concentrations of LA and NaCl. a Growth of E.coli cells in the presence of different LA concentrations; b OD600 of E.coli cells in the presence of 10 g/L LA and for different concentrations of LA; c AA concentration in the fermentation broth compared to broth’s pH profile different NaCl concentrations; d AA concentration in the fermentation broth compared to broth’s pH profile in the presence of 10 g/L LA and different NaCl concentrations

Full size image

Effect of NaCl concentration

The last set of experiments focused on the impact that salinity has on the E. coli strain performance. Salinity tolerance (i.e., NaCl concentration) was assessed by testing three different concentrations of salt in the synthetic medium, i.e., 0.1, 0.5 and 1% w/v NaCl. Typically, M9 media and accordingly the synthetic media in the previous experiments were characterized by 0.01% w/v salinity, while in grass silage NaCl concentration may reach up to 10 g/L (i.e., approx. 1.0% w/v) NaCl [30]. The flask fermentation process was negatively affected for increased NaCl concentrations (i.e., 0.1, 0.5 and 1.0% w/v), but the effect was more prominent for 0.5% and 1% w/v NaCl (Fig. 2c). For 0.1% w/v NaCl, an OD600 of 0.94 ± 0.05 at t = 48 h was observed, while for 1% w/v NaCl, an OD600 of 0.75 ± 0.00 at t = 48 h. In addition, Fig. 2d shows that 100% removal of AA was achieved in the presence of 0.1% w/v NaCl, whereas only 60.7% and 47% of AA was consumed within 72 h, when 0.5% w/v and 1% w/v NaCl was added to the synthetic media, respectively. Glucose consumption was not affected as NaCl concentration increased; glucose was totally depleted within 24 h. On the contrary, pH values were affected mainly due to the different AA consumption profiles. At 0.01 and 0.1% w/v NaCl, the AA was totally consumed at t = 72 h and the pH values rose around 8.0. At higher NaCl concentrations, due to the incomplete AA consumption the pH was maintained around 7.0 which is the optimum value for the E. coli cells to grow.

In summary, the synthetic medium simulating the composition of grass silage leachate used in all the subsequent experiments comprised 0.5 g/L glucose and 0.5 g/L AA as carbon sources, 10 g/L LA and 0.1% w/v NaCl. Furthermore, the series of experiments for defining the composition of the synthetic medium that does not affect the E. coli strain’s growth, offered preliminary evidence that the strain can selectively catabolize organic impurities (i.e., glucose and AA) in the presence of different concentrations of LA, without catabolizing any of the LA present. Kinetic data collection under these conditions followed to design the bench-scale batch bioreactor experiments.

Kinetic data collection in shake flask experiments

The last step in the flask scale experiments was the collection of kinetic data on the E. coli growth and the removal of glucose and AA. These results would provide valuable data for the design of the up-scaled bioreactor fermentation process. Two different batch tests were performed by 12-h time difference, and samples were collected on an hourly base to cover a total period of 24 h. Figure 3a shows that glucose was consumed within 7 h, while LA remained constant at 11.69 ± 0.33 g/L. AA was produced by the cells in the first 7 h reaching 0.70 ± 0.01 g/L, and was totally consumed within 24 h, afterwards. The cells started using AA as carbon source only after complete consumption of the glucose [28]. This is a typical diauxic growth behavior, which has been widely reported in the relevant literature [31, 32].

Fig. 3
figure 3

Comparison of results from batch fermentation test in shake flasks and bench scale bioreactor for kinetic data collection. The synthetic solution comprises: 0.5 g/L AA, 0.5 g/L glucose, 10 g/L LA and 0.1% w/v NaCl. Temporal profile of glucose, AA and cells’ growth (left y-axis) and LA concentration (right y-axis); a Batch fermentation in flask-scale; b Batch fermentation in 3 L bench-scale bioreactor

Full size image

According to Fig. 3a, E. coli cells present a short lag phase and grow exponentially from 3 to 7 h of fermentation while glucose was catabolized by the cells, reaching 0.29 ± 0.02 g/L concentration of dry biomass. The cells then grew at a slower rate when AA was present in the culture as sole carbon source. The final biomass concentration at 24 h of fermentation reached up to 0.43 g/L concentration of dry biomass. pH was also measured during the kinetic study. At t = 0, pH was 7.11 ± 0.00, and subsequently dropped down to 6.80 ± 0.02 during glucose consumption and AA excretion. From 8 to 24 h, pH increased up to 7.10 ± 0.02, which is near the optimum value for E. coli growth [33].

Batch bioreactor fermentation tests

Batch fermentation experiments in the lab-scale bioreactor were designed according to the aforementioned kinetic data results. The duration in batch bioreactor fermentation tests was prolonged for 6 h, since AA consumption was not complete after 24 h (Fig. 3b). Biomass growth was similar for flasks and bioreactor experiments, reaching at 24 h of fermentation up to 0.43 ± 0.01 g/L and 0.39 ± 0.03 g/L, respectively. However, comparing the growth curves in flasks and bioreactor fermentations, a slight difference is observed concerning the lag- and the exponential-phase. Fermentation in flasks exhibited a longer lag-phase which lasted for 3 h, while in the bioreactor cells grew faster, probably due to better aeration and mixing conditions of the bioreactor. Subsequently, cells enter a stationary phase with 2-h difference, at t = 7 h for flasks and at t = 5 h for bioreactor tests, reaching similar concentrations of biomass (Fig. 3b). Exponential growth ended, when glucose has been totally depleted in both cases. LA concentration remains constant in bioreactor experiments just like shake flask tests. Regarding AA concentration in bioreactor tests, a similar trend to flasks tests is observed, i.e., excretion of AA while glucose is consumed. However, AA is not totally consumed in bioreactor tests, given that at 20 h of fermentation its concentration is 0.30 g/L ± 0.01. AA consumption rate was calculated to assess the efficiency of the process in bioreactor tests compared to flask tests. AA consumption rate is rather linear both in flasks and bioreactor tests, as depicted in Fig. 3. However, the AA consumption rate in flasks is higher (0.035 g/L/h) compared to the bioreactor (0.022 g/L/h) resulting in 60.4% removal in the bioreactor within 30 h compared to 100% AA removal in flask within 24 h. These results could be explained due to the challenges associated to fermentation scale-up, such as non-uniforms conditions in the bioreactor that may cause negative effects on the strain or different physiological responses to high shear-flow dynamics, gas–liquid mass transfer limitations, substrate concentration gradients etc. [34]. The results from the batch fermentations in the bench-scale bioreactor were quite promising since the E. coli strain could successfully catabolize the organic impurities (i.e., glucose and AA), while LA concentration remained constant. Summarizing all the above results, it can be stated that the development of an LA bio-purification process using the E. coli strain seems to be feasible and provide a valid proof-of-concept that could be further developed to a semi-continuous MBR process.

Semi-continuous MBR fermentation tests

The concept of the LA bio-purification was further studied by testing different operating parameters in a semi-continuous MBR process aiming to provide a clear proof-of-concept of the integrated bio-purification system. The semi-continuous process was divided in two stages, the initial fermentation process (1st stage), which was similar to the batch fermentation described in "Batch bioreactor fermentation tests" section, and the 2nd stage when cycles of membrane filtration operation and biological fermentation process (i.e., the MBR concept) are repeated to provide a purified LA effluent. The 1st stage aims to develop a mature cell culture in addition to the total depletion of glucose and AA contained in the initial fermentation broth, before adding fresh fermentation medium for further cells’ growth.

The growth behavior for all the MBR experiments (6 in total), during the 1st stage of the process was similar, while the cells’ growth differs depending on the applied conditions of each MBR test, when the process entered in the semi-continuous MBR mode (i.e., 2nd stage), (Additional file 1: Fig. S2). A typical OD600 profile is depicted in Fig. 4, concerning MBR 6 tests. An exponential growth is observed for the first 5 h which concurs with the glucose removal by the cells. By the end of the exponential growth, OD600 is approx. 2.0. Afterwards, AA is used as the sole carbon source and the cells kept growing at a slower rate and by the end of the batch operating stage the cells have reached stationary phase. When the periodic permeate/feeding cycles start, distinctive peaks appear during OD600 monitoring, which correspond to the intermittent removal of the treated (purified) synthetic LA solution through the UF membrane module. This is clear evidence that cells are retained by the UF membrane module, while the permeate (i.e., treated LA solution) is withdrawn. Since the volume of the fermentation broth is reduced, while the cells are retained within the bioreactor, the cells concentration increases. When fresh synthetic medium is added in the bioreactor, the fermentation broth volume is restored to its initial value and a rather sharp drop of OD600 is observed. Afterwards, cells continue growing using the carbon sources of the fresh media.

Fig. 4
figure 4

Cells’ growth profile in MBR 6 test obtained by online monitoring of OD600; shaded area corresponds to the batch operation of the MBR (1st stage) while arrows indicate the time points of the permeate/feeding cycles during the semi-continuous MBR operation

Full size image

As in flask and batch-bioreactor experiments, the AA removal determines the process design of the MBR system, according to obtained data. E. coli cells initially consume glucose quite fast and then AA at a slower rate, following the diauxic growth pattern, as it has already been mentioned. Glucose was consumed in less than 1 h, after each feeding step with fresh synthetic medium; however, AA removal was slower and, in some cases, not consumed at all. In Table 3, data of critical metrics referring to the semi-continuous MBR tests are provided. It is important to mention that concerning LA, it remained constant at the initial concentration with minor variation.

Table 3 Overview of MBR tests and calculated metrics for the efficiency of the MBR process
Full size table

Table 3 shows that AA maximum consumption rate was extremely low in experiment MBR 1, which actually means that AA was practically not removed. AA maximum consumption rate corresponds to the highest value of the AA consumption rate calculated separately for each filtration/feed cycle of the respective MBR test. In MBR 2 experiment, more time (i.e., 30 h instead of 24 h) was provided to the cell culture for the 1st stage of the process to achieve total depletion of AA, before starting the 2nd stage of the semi-continuous process. It was assumed that this would improve the 2nd stage of the process, and as a result the efficiency of the entire MBR process, would be improved. Indeed, AA consumption rate increased from 0.0061 g/L/h to 0.0206 g/L/h, but still remained low in comparison with the previous flask and batch experimental steps. The total removal of AA in MBR 2 was 68.8% and the permeate recovery rate was increased from 11.30 mL/h to 25.27 mL/h; however, the process duration remained quite long.

A possible explanation may be the insufficient N concentration that was observed both at the start, as well as at the end of each feeding cycle (i.e., 7.0 ± 3.0. and 6.0 ± 0.4 in average for MBR 1 and 2 tests, respectively). To solve this issue, the M9 quantity, for the 2nd stage of the process, was doubled in MBR 3. However, the efficiency of the process was still rather low. Although, AA depletion reached 100%, the AA maximum consumption rate remained almost the same, possibly also due to the long-time intervals for each membrane filtration/fermentation cycle. When nitrogen concentration was doubled for the whole process in MBR 4 test, the AA maximum consumption rate reached 0.0376 g/L/h and the net permeate recovery was more than 3 times higher (i.e., 66.67 mL/h) compared to MBR 3. However, the AA was not totally catabolized during the batch process (1st stage) within 24 h; possibly, more time was necessary for the cells to remove the initial AA quantity.

Therefore, in MBR 5 test, the batch operation (1st stage) of the MBR process was increased by 10 h (34 h in total), to achieve complete depletion of the available carbon sources (both glucose and AA) by the cells, before starting the membrane filtration. Thereafter, an effluent, free of impurities (i.e., AA, glucose and cells) was achieved in MBR 5 within a reasonable timeframe. i.e., 3 h permeate/feeding cycle for a permeate/feeding volume of 230 ml, which is approx. 12.5% of the MBR working volume. The final optimized operating protocol was repeated in MBR 6 for validation purpose. Permeate recovery rate was identical for MBR 5 and 6 (i.e., 76.67 mL/h) and it was increased by 10 mL/h compared to MBR 4. It was also obvious that, the hindering of cell’s catabolism due to lack of nitrogen availability had been solved; the mean TN concentration at the end of each feeding cycle was approx. 50–60 ppm. Finally, the max consumption rate of AA was 0.0764 g/L/h for the optimum MBR working conditions (MBR 6).

The transmembrane pressure (TMP) of the UF filtration in the MBR is another operating parameter of high importance, which was monitored and assessed in this study. The membrane filtration performance was evaluated by measuring TMP which indicates the level of the filtration resistance due to membrane fouling. Table 3 summarizes critical parameters related to the TMP measurements for the semi-continuous MBR process. It is shown that, the maximum TMP value varied between − 29 and − 72 mbar. The overall TMP drop is rather low and fouling does not seem to be an issue for this process, possibly due to the low cell culture density and the limited production of extracellular polymers by this E. coli strain.

Figure 5 provides a comparison between the 1st semi-continuous MBR experiment (MBR 1) with the final optimized process (MBR 6). The presented data refer to the 2nd stage of the process (i.e., filtration/feeding cycles). Thus, timepoint 0 corresponds to the point of the 1st fresh synthetic medium feed. It is evident that the duration of the process regarding the MBR operation (2nd stage), has been significantly reduced from 35 to 12 h. TN measurements also depicted in Fig. 5 clearly show that the low N concentration in MBR 1 (i.e., below 13 ppm) affects the efficiency of the process compared to the MBR 6 in which sufficient N (above 53 ppm) is provided. In both cases, glucose was consumed and LA remained constant; however, AA removal is 12.5 times higher for the optimized process (MBR 6) when compared to the initial trial (MBR 1). The 100% AA depletion before each cycle of the membrane filtration coupled with 6.8-fold higher permeate recovery rate, reflects positively on the performance of the integrated MBR process.

Fig. 5
figure 5

Concentration profiles of AA, Glucose, TN and LA of the initial semi-continuous MBR test (MBR 1) and the final optimized MBR test (MBR 6); Time refers to the 2nd stage of the integrated MBR process; a AA concentration profile; b Glucose concentration profile; c TN concentration profile; d LA concentration profile

Full size image

Table 4 provides data regarding the membrane performance for the filtration cycles of two MBR experiments, namely MBR 1 (initial) and MBR 6 (optimized). It is observed that the maximum reduction in TMP occurs in the 3rd filtration cycle for both MBR 1 (-70.8 mbar) and MBR 6 (-36.0 mbar). According to the membrane specifications, the maximum operating filtration TMP drop is 600 mbar [35]. Therefore, even if the maximum TMP reduction is double for MBR 1, compared to MBR 6, the fouling of the membrane is minimum in both experiments. Lower TMP drop rate has been also achieved after process optimization, e.g., -48.6 mbar/h compared to -15.1 mbar/h at the 3rd membrane filtration cycle of MBR 1 and 6, respectively. The membrane fouling tendency was further evaluated through the TMP temporal variation, during the filtration cycles of MBR 1 and MBR 6. The overall efficient membrane filtration performance for both is evident from Fig. 6.

Table 4 Calculated parameters for the comparison of the membrane performance of the initial test in MBR (MBR 1) and the final optimized MBR test (MBR 6)
Full size table
Fig. 6
figure 6

Temporal variation of TMP during membrane filtration cycles for the initial MBR test (MBR 1) and the final optimized MBR test (MBR 6)

Full size image

Conclusions

An optimized synthetic medium was determined, for the growth and study of an engineered E. coli strain capable of selectively catabolizing different carbon sources present in a synthetic LA solution. Its cultivation and the kinetics of glucose and AA were validated in bench-scale bioreactor followed by the semi-continuous MBR process experiments. Complete cell retention without significant membrane fouling and a cell-free effluent with higher purity of LA product-solution were obtained, due to 100% removal of typical impurities (i.e., glucose and AA) in short to medium operating time.

To conclude, it was clearly demonstrated that a bio-purification process for LA employing the engineered E. coli strain cultivated in an MBR process is a technically feasible concept, while the advantages of an efficient semi-continuous MBR bio-purification process are obvious and the technology is worth to be further investigated and scaled-up. Further research towards the validation of the bio-purification process should be performed with real LA fermentation solutions. Furthermore, improved E. coli strains that could tolerate higher LA and/or salts concentrations can be developed either through classic adaptive laboratory evolution (ALE) or genetic engineering approaches, to expand the operating conditions of the bio-purification process.

Availability of data and materials

All data generated during this study are included in this published article and its supplementary information files.

References

  1. Komesu A, Martins PF, Lunelli BH, Oliveira J, Maciel Filho R, Wolf Maciel MR. Evaluation of lactic acid purification from fermentation broth by hybrid short path evaporation using factorial experimental design. Sep Purif Technol. 2014;136:233–40.

    Article  CAS  Google Scholar 

  2. Ahmad A, Banat F, Taher H. A review on the lactic acid fermentation from low-cost renewable materials: recent developments and challenges. Environ Technol Innov. 2020;20.

  3. Oliveira RAD, Komesu A, Vaz Rossell CE, Wolf Maciel MR, Maciel Filho R. A study of the residual fermentation sugars influence on an alternative downstream process for first and second-generation lactic acid. Sustain Chem Pharm. 2020. https://doi.org/10.1016/j.scp.2019.100206.

    Article  Google Scholar 

  4. González-Navarrete C, Sánchez-Ramírez E, Ramírez-Márquez C, Hernández S, Cossío-Vargas E, Segovia-Hernández JG. Innovative reactive distillation process for the sustainable purification of lactic acid. Ind Eng Chem Res. 2021;61(1):621–37.

    Article  Google Scholar 

  5. Din NAS, Lim SJ, Maskat MY, Mutalib SA, Zaini NAM. Lactic acid separation and recovery from fermentation broth by ion-exchange resin: a review. Bioresour Bioprocess. 2021. https://doi.org/10.1186/s40643-021-00384-4.

    Article  Google Scholar 

  6. Hulse JH. Biotechnologies: past history, present state and future prospects. Trends Food Sci Technol. 2004;15(1):3–18.

    Article  CAS  Google Scholar 

  7. Olszewska-Widdrat A, Alexandri M, López-Gómez JP, Schneider R, Mandl M, Venus J. Production and purification of L-lactic acid in lab and pilot scales using sweet sorghum juice. Fermentation. 2019;5(2):36.

    Article  CAS  Google Scholar 

  8. Jantasee S, Kienberger M, Mungma N, Siebenhofer M. Potential and assessment of lactic acid production and isolation–a review. J Chem Technol Biotechnol. 2017;92(12):2885–93.

    Article  CAS  Google Scholar 

  9. de Oliveira RA, Komesu A, Rossell CEV, Maciel FR. Challenges and opportunities in lactic acid bioprocess design—From economic to production aspects. Biochem Eng J. 2018;133:219–39.

    Article  Google Scholar 

  10. Lee HD, Lee MY, Hwang YS, Cho YH, Kim HW, Park HB. Separation and purification of lactic acid from fermentation broth using membrane-integrated separation processes. Ind Eng Chem Res. 2017;56(29):8301–10.

    Article  CAS  Google Scholar 

  11. Anagnostopoulou C, Kontogiannopoulos KN, Gaspari M, Morlino MS, Assimopoulou AN, Kougias PG. Valorization of household food wastes to lactic acid production: a response surface methodology approach to optimize fermentation process. Chemosphere. 2022;296: 133871.

    Article  CAS  PubMed  Google Scholar 

  12. Schaffenberger MS. Green biorefinery—the production, isolation and polishing of amino acids from grass silage juice at pilot and-annotated. Austria: Graz University of Technology; 2013.

    Google Scholar 

  13. Hou M, Wang Z, Sun L, Jia Y, Wang S, Cai Y. Characteristics of lactic acid bacteria, microbial community and fermentation dynamics of native grass silage prepared in Inner Mongolian Plateau. Front Microbiol. 2023;13:5155.

    Article  Google Scholar 

  14. Zhang JG, Tanaka O, Uegaki R, Cai Y, Kobayashi R. The effect of inoculation and additives on d (−)-and l (+)-lactic acid production and fermentation quality of guineagrass (Panicum maximum Jacq) silage. J Sci Food Agric. 2000;80(15):2186–9.

    Article  CAS  Google Scholar 

  15. Ecker J, Schaffenberger M, Koschuh W, Mandl M, Böchzelt H, Schnitzer H, et al. Green biorefinery upper Austria–pilot plant operation. Sep Purif Technol. 2012;96:237–47.

    Article  CAS  Google Scholar 

  16. Faigl F, Mátravölgyi B, Thurner A, editors. Novel methods for the separation of optical isomers. new methodologies and techniques for a sustainable organic chemistry; 2008: Springer.

  17. Chauliac D, Pullammanappallil P, Ingram L, Shanmugam K. Removing chiral contamination of lactate solutions by selective metabolism of the D-enantiomer. Biotech Lett. 2015;37:2411–8.

    Article  CAS  Google Scholar 

  18. Chauliac D, Pullammanappallil PC, Ingram LO, Shanmugam K. A combined thermochemical and microbial process for recycling polylactic acid polymer to optically pure l-lactic acid for reuse. J Polym Environ. 2020;28:1503–12.

    Article  CAS  Google Scholar 

  19. Patsios SI, Karabelas AJ. An investigation of the long-term filtration performance of a membrane bioreactor (MBR): the role of specific organic fractions. J Membr Sci. 2011;372(1–2):102–15.

    Article  CAS  Google Scholar 

  20. Taleghani HG, Ghoreyshi AA, Najafpour G. Lactic acid production with in situ extraction in membrane bioreactor. Appl Food Biotechnol. 2017;4(1):27–34.

    CAS  Google Scholar 

  21. Mahboubi A, Uwineza C, Doyen W, De Wever H, Taherzadeh MJ. Intensification of lignocellulosic bioethanol production process using continuous double-staged immersed membrane bioreactors. Biores Technol. 2020;296: 122314.

    Article  CAS  Google Scholar 

  22. Vu DH, Mahboubi A, Root A, Heinmaa I, Taherzadeh MJ, Åkesson D. Application of immersed membrane bioreactor for semi-continuous production of polyhydroxyalkanoates from organic waste-based volatile fatty acids. Membranes. 2023;13(6):569.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Raveschot C, Deracinois B, Bertrand E, Flahaut C, Frémont M, Drider D, et al. Integrated continuous bioprocess development for ACE-inhibitory peptide production by Lactobacillus helveticus strains in membrane bioreactor. Front Bioeng Biotechnol. 2020;8: 585815.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Sweeney J, Murphy CD, McDonnell K. Towards an effective biosensor for monitoring AD leachate: a knockout E. coli mutant that cannot catabolise lactate. Appl Microbiol Biotechnol. 2015;99(23):10209–14.

    Article  CAS  PubMed  Google Scholar 

  25. Sweeney JB, McDonnell K, Murphy CD. Improving the specificity of E. coli acetate/propionate exclusion biosensors via iterative engineering. Enzyme Microbial Technol. 2022;160:110091.

    Article  CAS  Google Scholar 

  26. Nastouli A, Tsirigka A, Harasek M, Karabelas AJ, Patsios SI. The effect of heat sterilization on key filtration performance parameters of a commercial polymeric (PVDF) hollow-fiber ultrafiltration membrane. Membranes. 2022;12(8):725.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Chang D-E, Shin S, Rhee J-S, Pan J-G. Acetate metabolism in a pta mutant of Escherichia coli W3110: importance of maintaining acetyl coenzyme A flux for growth and survival. J Bacteriol. 1999;181(21):6656–63.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Shimada T, Nakazawa K, Tachikawa T, Saito N, Niwa T, Taguchi H, et al. Acetate overflow metabolism regulates a major metabolic shift after glucose depletion in Escherichia coli. FEBS Lett. 2021;595(15):2047–56.

    Article  CAS  PubMed  Google Scholar 

  29. Kirkpatrick C, Maurer LM, Oyelakin NE, Yoncheva YN, Maurer R, Slonczewski JL. Acetate and formate stress: opposite responses in the proteome of Escherichia coli. J Bacteriol. 2001;183(21):6466–77.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Franco M, Tapio I, Pirttiniemi J, Stefański T, Jalava T, Huuskonen A, et al. Fermentation quality and bacterial ecology of grass silage modulated by additive treatments, extent of compaction and soil contamination. Fermentation. 2022;8(4):156.

    Article  CAS  Google Scholar 

  31. Egli T, Lendenmann U, Snozzi M. Kinetics of microbial growth with mixtures of carbon sources. Antonie Van Leeuwenhoek. 1993;63(3):289–98.

    Article  CAS  PubMed  Google Scholar 

  32. Kovárová-Kovar K, Egli T. Growth kinetics of suspended microbial cells: from single-substrate-controlled growth to mixed-substrate kinetics. Microbiol Mol Biol Rev. 1998;62(3):646–66.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Davey K. Modelling the combined effect of temperature and pH on the rate coefficient for bacterial growth. Int J Food Microbiol. 1994;23(3–4):295–303.

    Article  CAS  PubMed  Google Scholar 

  34. Xia J, Wang G, Fan M, Chen M, Wang Z, Zhuang Y. Understanding the scale-up of fermentation processes from the viewpoint of the flow field in bioreactors and the physiological response of strains. Chin J Chem Eng. 2021;30:178–84.

    Article  CAS  Google Scholar 

  35. Solutions KS. Koch Separation Solutions. https://www.kovalus.com/wp-content/uploads/2020/10/puron-mbr-psh-rows.pdf. Accessed 05 Dec 2023.

Download references

Acknowledgements

The authors sincerely thank the excellent technical support provided by Mr. Asterios Lekkas and Mr. Michalis Lekkas in upgrading the experimental setups employed herein.

Funding

This work was funded by the EUROPEAN UNION’S HORIZON 2020 RESEARCH AND INNOVATION PROGRAMME under the Marie Skłodowska-Curie grant agreement No 860477 “AgRefine” (H2020-MSCA-ITN-2019).

Author information

Authors and Affiliations

  1. Laboratory of Natural Resources and Renewable Energies, Chemical Process and Energy Resources Institute (CPERI), Centre for Research and Technology-Hellas (CERTH), Thermi, Thessaloniki, Greece

    Alexandra Nastouli, Anastasios J. Karabelas & Sotiris I. Patsios

  2. Institute of Chemical, Environmental and Bioscience Engineering, TU Wien, Vienna, Austria

    Alexandra Nastouli & Michael Harasek

  3. School of Biosystems and Food Engineering, University College Dublin (UCD), Belfield, Dublin, Ireland

    Joseph Sweeney

Authors
  1. Alexandra Nastouli
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Joseph Sweeney
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Michael Harasek
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Anastasios J. Karabelas
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Sotiris I. Patsios
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Conceptualization, S.I.P.; methodology, A.N. S.I.P. and J.S.; validation, A.N. and S.I.P.; investigation, A.N.; writing—original draft preparation, A.N. and S.I.P.; writing—review and editing, M.H., A.J.K., J.S. and S.I.P.; visualization, A.N. and S.I.P.; supervision, M.H., A.J.K. and S.I.P.; project administration, S.I.P.; funding acquisition, A.J.K. and S.I.P. All authors have read and agreed to the published version of the manuscript.

Corresponding author

Correspondence to Sotiris I. Patsios.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

A patent application to the Hellenic Industrial Property Organization (Greece) relating to the content of the manuscript is pending by A.N. and S.I.P.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file1: Figure S1

. Results from batch fermentation tests (t = 72 h) in shake flasks with 0.5 g/L glucose, 0.5 g/L AA, 5 g/L LA and different quantities of nutrients; a OD600 of E. coli cells; b AA concentration; and c Glucose concentration at different time points. Figure S2. Cells’ growth profile in semi-continuous MBR fermentation tests MBR 1 to MBR 6 obtained by online monitoring of OD600.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Nastouli, A., Sweeney, J., Harasek, M. et al. Development of a hybrid bio-purification process of lactic acid solutions employing an engineered E. coli strain in a membrane bioreactor. Biotechnol Biofuels 17, 48 (2024). https://doi.org/10.1186/s13068-024-02497-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13068-024-02497-2

Keywords

Biotechnology for Biofuels and Bioproducts

ISSN: 2731-3654

Contact us