Membrane malfunctions in freeze-dried Escherichia coli

Engineering cascade biocatalysis in whole cells for bottom-up synthesis of cello-oligosaccharides: flux control over three enzymatic steps enables soluble production

BACKGROUND

Soluble cello-oligosaccharides (COS, β-1,4-D-gluco-oligosaccharides with degree of polymerization DP 2-6) have been receiving increased attention in different industrial sectors, from food and feed to cosmetics. Development of large-scale COS applications requires cost-effective technologies for their production. Cascade biocatalysis by the three enzymes sucrose-, cellobiose- and cellodextrin phosphorylase is promising because it enables bottom-up synthesis of COS from expedient substrates such as sucrose and glucose. A whole-cell-derived catalyst that incorporates the required enzyme activities from suitable co-expression would represent an important step towards making the cascade reaction fit for production. Multi-enzyme co-expression to reach distinct activity ratios is challenging in general, but it requires special emphasis for the synthesis of COS. Only a finely tuned balance between formation and elongation of the oligosaccharide precursor cellobiose results in the desired COS.

RESULTS

Here, we show the integration of cellodextrin phosphorylase into a cellobiose-producing whole-cell catalyst. We arranged the co-expression cassettes such that their expression levels were upregulated. The most effective strategy involved a custom vector design that placed the coding sequences for cellobiose phosphorylase (CbP), cellodextrin phosphorylase (CdP) and sucrose phosphorylase (ScP) in a tricistron in the given order. The expression of the tricistron was controlled by the strong T7_lacO promoter and strong ribosome binding sites (RBS) for each open reading frame. The resulting whole-cell catalyst achieved a recombinant protein yield of 46% of total intracellular protein in an optimal ScP:CbP:CdP activity ratio of 10:2.9:0.6, yielding an overall activity of 315 U/g dry cell mass. We demonstrated that bioconversion catalyzed by a semi-permeabilized whole-cell catalyst achieved an industrial relevant COS product titer of 125 g/L and a space-time yield of 20 g/L/h. With CbP as the cellobiose providing enzyme, flux into higher oligosaccharides (DP ≥ 6) was prevented and no insoluble products were formed after 6 h of conversion.

CONCLUSIONS

A whole-cell catalyst for COS biosynthesis was developed. The coordinated co-expression of the three biosynthesis enzymes balanced the activities of the individual enzymes such that COS production was maximized. With the flux control set to minimize the share of insolubles in the product, the whole-cell synthesis shows a performance with respect to yield, productivity, product concentration and quality that is promising for industrial production.

Whole cell-based catalyst for enzymatic production of the osmolyte 2-O-α-glucosylglycerol

Background

Glucosylglycerol (2-O-α-d-glucosyl-sn-glycerol; GG) is a natural osmolyte from bacteria and plants. It has promising applications as cosmetic and food-and-feed ingredient. Due to its natural scarcity, GG must be prepared through dedicated synthesis, and an industrial bioprocess for GG production has been implemented. This process uses sucrose phosphorylase (SucP)-catalyzed glycosylation of glycerol from sucrose, applying the isolated enzyme in immobilized form. A whole cell-based enzyme formulation might constitute an advanced catalyst for GG production. Here, recombinant production in Escherichia coli BL21(DE3) was compared systematically for the SucPs from Leuconostoc mesenteroides (LmSucP) and Bifidobacterium adolescentis (BaSucP) with the purpose of whole cell catalyst development.

Results

Expression from pQE30 and pET21 plasmids in E. coli BL21(DE3) gave recombinant protein at 40–50% share of total intracellular protein, with the monomeric LmSucP mostly soluble (≥ 80%) and the homodimeric BaSucP more prominently insoluble (~ 40%). The cell lysate specific activity of LmSucP was 2.8-fold (pET21; 70 ± 24 U/mg; N = 5) and 1.4-fold (pQE30; 54 ± 9 U/mg, N = 5) higher than that of BaSucP. Synthesis reactions revealed LmSucP was more regio-selective for glycerol glycosylation (~ 88%; position O2 compared to O1) than BaSucP (~ 66%), thus identifying LmSucP as the enzyme of choice for GG production. Fed-batch bioreactor cultivations at controlled low specific growth rate (µ = 0.05 h⁻¹; 28 °C) for LmSucP production (pET21) yielded ~ 40 g cell dry mass (CDM)/L with an activity of 2.0 × 10⁴ U/g CDM, corresponding to 39 U/mg protein. The same production from the pQE30 plasmid gave a lower yield of 6.5 × 10³ U/g CDM, equivalent to 13 U/mg. A single freeze–thaw cycle exposed ~ 70% of the intracellular enzyme activity for GG production (~ 65 g/L, ~ 90% yield from sucrose), without releasing it from the cells during the reaction.

Conclusions

Compared to BaSucP, LmSucP is preferred for regio-selective GG production. Expression from pET21 and pQE30 plasmids enables high-yield bioreactor production of the enzyme as a whole cell catalyst. The freeze–thaw treated cells represent a highly active, solid formulation of the LmSucP for practical synthesis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12934-021-01569-4.

Browning of freeze-dried probiotic bacteria cultures in relation to loss of viability during storage

Freeze-dried cultures of Lactobacillus acidophilus (La-5) showed visible brown discoloration even after a short storage at relatively mild conditions (a(w) = 0.22 and 30 degrees C), and the browning processes were found to coincide with bacteria inactivation. It was demonstrated, by using high-pressure treatment for obtaining bacteria samples with different ratios of live/dead bacteria, that death of bacteria is not a prerequisite for the browning processes. Furthermore, it was shown that hydroxymethylfurfural (HMF) (or condensation products of HMF) introduces accelerated viability loss when HMF is added to the freeze-drying medium. Discoloration of bacteria cultures containing only sucrose/maltodextrin or lactose/maltodextrin in the freeze-drying matrices is suggested to be related to various types of nonenzymatic browning reactions, including carbonyl-protein (or carbonyl-DNA) interactions and carbohydrate condensation/polymerization (without involvement of proteins), the latter proceeding at low a(w) following hydrolysis of the peptidoglycan layer in the bacteria cell wall. More than one single type of browning reaction is accordingly concluded to be related to bacteria death, and the loss of viability in freeze-dried bacteria seems to be influenced by oxidation reactions, browning reactions, and the physical instability of the bacteria membrane/cell wall.

Oxidative stress and its effects during dehydration

Water is usually thought to be required for the living state, but several organisms are capable of surviving complete dehydration (anhydrobiotes). Elucidation of the mechanisms of tolerance against dehydration may lead to development of new methods for preserving biological materials that do not normally support drying, which is of enormous practical importance in industry, in clinical medicine as well as in agriculture. One of the molecular mechanisms of damage leading to death in desiccation-sensitive cells upon drying is free-radical attack to phospholipids, DNA and proteins. This review aims to summarize the strategies used by anhydrobiotes to cope with the danger of oxygen toxicity and to present our recent results about the importance of some antioxidant defense systems in the dehydration tolerance of Saccharomyces cerevisiae, a usual model in the study of stress response.

Survival curves for microbial species stored by freeze-drying

The survival of a variety of species of microorganism following storage for up to 20 years has been analyzed. The organisms were freeze-dried, sealed in ampoules under vacuum (<1 Pa) and stored in the dark at 5 degrees C. The yeast that was tested, Saccharomyces cerevisiae, showed only 8% survival when recovered shortly after freeze-drying, but subsequent loss during storage was the least among all the tested microorganisms. The decrease in the logarithm of survival per year (log survival) was -0.010, which corresponds to a survival rate of 97.7% per year. The Gram-negative bacteria tested, Escherichia coli, Pseudomonas putida, and Enterobacter cloacae, showed 42.6, 33.5, and 50.8% survival shortly after freeze-drying, which was higher than the corresponding survival of S. cerevisiae, but the subsequent loss during storage was greater than S. cerevisiae, the log survival figures being -0.041, -0.058, and -0.073 per year. These values correspond to survival rates of 91.0, 87.5, and 84.5% each year. The Gram-positive bacteria tested, Lactobacillus acidophilus and Enteroccoccus faecium, showed 62.5 and 85.2% survival shortly after freeze-drying, which was even higher than that of the Gram-negative species, and these organisms also showed better survival during storage than Gram-negative bacteria; their log survival rates were -0.018 and -0.016 per year, which corresponded to survival rates of almost 96% per year. Comparison of these results with other published data for different drying conditions suggests that survival during storage is strongly influenced by the degree of vacuum under which the ampoules were sealed. The excellent survival after freeze-drying of each species might be attributable to the high level of desiccation and to sealing under vacuum.

Desiccation tolerance of prokaryotes

The removal of cell-bound water through air drying and the addition of water to air-dried cells are forces that have played a pivotal role in the evolution of the prokaryotes. In bacterial cells that have been subjected to air drying, the evaporation of free cytoplasmic water (Vf) can be instantaneous, and an equilibrium between cell-bound water (Vb) and the environmental water (vapor) potential (psi wv) may be achieved rapidly. In the air-dried state some bacteria survive only for seconds whereas others can tolerate desiccation for thousands, perhaps millions, of years. The desiccated (anhydrobiotic) cell is characterized by its singular lack of water--with contents as low as 0.02 g of H2O g (dry weight)-1. At these levels the monolayer coverage by water of macromolecules, including DNA and proteins, is disturbed. As a consequence the mechanisms that confer desiccation tolerance upon air-dried bacteria are markedly different from those, such as the mechanism of preferential exclusion of compatible solutes, that preserve the integrity of salt-, osmotically, and freeze-thaw-stressed cells. Desiccation tolerance reflects a complex array of interactions at the structural, physiological, and molecular levels. Many of the mechanisms remain cryptic, but it is clear that they involve interactions, such as those between proteins and co-solvents, that derive from the unique properties of the water molecule. A water replacement hypothesis accounts for how the nonreducing disaccharides trehalose and sucrose preserve the integrity of membranes and proteins. Nevertheless, we have virtually no insight into the state of the cytoplasm of an air-dried cell. There is no evidence for any obvious adaptations of proteins that can counter the effects of air drying or for the occurrence of any proteins that provide a direct and a tangible contribution to cell stability. Among the prokaryotes that can exist as anhydrobiotic cells, the cyanobacteria have a marked capacity to do so. One form, Nostoc commune, encompasses a number of the features that appear to be critical to the withstanding of a long-term water deficit, including the elaboration of a conspicuous extracellular glycan, synthesis of abundant UV-absorbing pigments, and maintenance of protein stability and structural integrity. There are indications of a growing technology for air-dried cells and enzymes. Paradoxically, desiccation tolerance of bacteria has virtually been ignored for the past quarter century. The present review considers what is known, and what is not known, about desiccation, a phenomenon that impinges upon every facet of the distributions and activities of prokaryotic cells.

Survival differences among freeze-dried genetically engineered and wild-type bacteria

Because the death mechanisms of freeze-dried and air-dried bacteria are thought to be similar, freeze-drying was used to investigate the survival differences between potentially airborne genetically engineered microorganisms and their wild types. To this end, engineered strains of Escherichia coli and Pseudomonas syringae were freeze-dried and exposed to air, visible light, or both. The death rates of all engineered strains were significantly higher than those of their parental strains. Light and air exposure were found to increase the death rates of all strains. Application of death rate models to freeze-dried engineered bacteria to be released into the environment is discussed.