Marianne Stage Strickertsson:
|
Probiotics, which are bacteria that are ingested for their health benefits, were initially described by Elie Metchnikoff in 1908 and today have an established place in the market in dietary supplements and functional foods. Probiotics are mainly distributed as freeze-dried products or in foods, with fermented milks being an especially common food vehicle. Most of the common probiotics on the market belong to the group of lactic acid bacteria, and Lacticaseibacillus rhamnosus GG and Lactobacillus acidophilus LA-5 are among the major marketed strains.
Before a probiotic strain reaches the market, it is subjected to a massive upscaling and several downstream processes such as cooling, concentration, addition of cryoprotective reagents, freezing and freeze-drying which may influence the characteristics that was originally used in the evaluation of the strain as a potential probiotic. During these processes, evaluation of the effect of industrial production on probiotics may help ensure that the final product possesses the same potential as the initially characterized novel strain.
For the strains Lacticaseibacillus rhamnosus GG and Lactobacillus acidophilus LA-5, the effects of industrial production on their probiotic potential has received attention. Industrial production processes have been associated with genomic instability of L. rhamnosus GG, leading to reported genomic deletions which in turn resulted in loss of the adhesive pili that have been associated with several probiotic properties of L. rhamnosus GG. In addition, the production of the strain L. acidophilus LA-5 with high stability during storage has been shown to be challenging.
In vitro methods developed to characterize probiotic functionality are mostly conducted on overnight cultures of laboratory grown strains and not on bacteria that have been through an industrial production process. As previous studies have demonstrated that changes in bacterial surface molecules induced by various growth conditions can affect the probiotic properties, in vitro assessment of probiotic functionality would ideally be performed on probiotics in the physiological state and matrix in which they are produced and ingested.
The aim of the present thesis was to explore the preservation of probiotic functionality of L. rhamnosus GG from the original stock to the final products as frozen and freeze-dried cultures and when co-fermented in a yogurt. Investigation of industrially-produced L. rhamnosus GG encompassed both genomic integrity, to ensure retention of genes with probiotic potential, and phenotypic properties. In addition, we wanted to investigate the possible effects of variations in harvest point and holding times on the transcriptome and phenotypes of industrially-produced L. acidophilus LA-5. Preliminary experiments included transcriptomic characterization of smallscale produced L. acidophilus LA-5 and assessment of adhesion properties on the strain subjected to various production setups.
In Paper I, genomic integrity was examined by next generation sequencing of samples from the original stock of L. rhamnosus GG to the final freeze-dried product including intermediate fermentations. Our findings demonstrate genomic integrity of L. rhamnosus GG in the investigated production process. In particular, preservation of the spaCBA-srtC1 gene cluster encoding the adhesive pili was of interest and was confirmed to be both genomically preserved and phenotypically expressed. In addition, in vitro tolerance to acid and bile, modulation of intestinal barrier function and modulation of cytokine secretion in dendritic cells were evaluated and showed consistent phenotypes across seven production batches.
In Paper II, we further characterized the strain as part of a probiotic yogurt that was produced in an application technology center using a downscaled version of industrial yogurt production. Samples of frozen L. rhamnosus GG used for inoculation, yogurt from the day of production and yogurt at end of the 28-day shelf life were examined for genomic stability and phenotypic traits using the same in vitro assays as in Paper I. Additionally, 60 single colony isolates were purified from the product populations and included in the genomic analysis. Genomic integrity was preserved in all products, but in one single colony isolate out of 60, gene loss had occurred that included the pili-encoding spaCBA-srtC1 gene cluster. Genomic analysis of the yogurt, from which the isolate exhibiting gene deletions was purified, did not reveal any fraction of reads missing in the coding genes, and it was reasoned that the gene loss could have emerged during the purification of single colony isolates or reflect a minor non-detectable subpopulation. Evaluation of phenotypic behavior revealed some variations in phenotypic behavior between samples delivered as frozen and freeze-dried product compared to L. rhamnosus GG in yogurt.
In Preliminary manuscript I, L. acidophilus LA-5 was small-scale produced mimicking the industrial production process including downstream processing. The transcriptomic profile was assessed on samples harvested from different points on the growth curve, which demonstrated that the point of harvesting was still reflected in the product after downstream processing. From the transcriptomic profile, seven genes that could distinguish growth phase upon harvesting were selected. The ability to distinguish growth phase in this setup was confirmed on an individual fermentation using qPCR. In order to determine if the reflection of harvest point still existed after prolonged holding, another set of small-scale fermentations were produced with different holding times and assessed by the 7 genes. The samples, harvested in stationary phase and subjected to prolonged holding, retained a high resemblance with the gene expression of stationary phase samples without holding. Similar was observed for three large-scale industrial production batches. The preliminary study demonstrated that harvest point can still be reflected in the transcriptomic profile after downstream processing in small-scale fermentations. Adhesion assessment indicated similar adhesion properties within the groups examining either harvest points, prolonged holding or the industrial batches. However, samples subjected to different holding times and the three industrial batches showed a significant higher adhesion capacity than the stationary phase sample from the samples harvested at different points on the growth curve.
In conclusion, data presented in this thesis demonstrates genomic integrity of industriallyproduced L. rhamnosus GG both as a freeze-dried product and when delivered in yogurt. Moreover, consistency in phenotypic behavior was demonstrated among production batches, whereas variation in phenotypic characteristics existed when L. rhamnosus GG was grown differently and in different matrixes. Finally, the growth condition upon harvesting is reflected in the transcriptome after downstream processing in small-scale fermented L. acidophilus LA-5, but no difference in adhesion capacity in these samples could be observed.