Novel Food Information: β-Lactoglobulin protein from yeast strain Komagataella phaffii yRMK-66 

On this page

• Background
• Introduction
• Development of the Production Organism
• Characterization of the Production Organism
• Product Information
• Dietary Exposure
• Nutrition
• Chemistry
• Toxicology
• Allergenicity
• Conclusion
• Regulatory Requirements Beyond Pre-Market Food Safety Assessment: Labelling

Background

Health Canada has notified Remilk that it has no objection to the food use of β-lactoglobulin protein ("r-βLG") from yeast strain Komagataella phaffii yRMK-66. The Department conducted a comprehensive assessment of this protein according to its Guidelines for the Safety Assessment of Novel Foods. These Guidelines are based upon internationally accepted principles for establishing the safety of foods with novel traits.

Depending on the intended end use of the r-βLG ingredient used in making consumer-ready products, additional regulations may apply beyond pre-market food safety considerations. Health Canada has consulted with the Canadian Food Inspection Agency (CFIA) on these matters and a description of these additional requirements is provided to support regulatory compliance.

The following provides a summary of the notification from Remilk and the evaluation by Health Canada. This document contains no confidential business information.

Introduction

Remilk has developed a genetically modified yeast strain, K. phaffii yRMK-66, that expresses β-lactoglobulin that is the same as the major component of whey protein from cow's milk, referred to as r-βLG. The r-βLG ingredient is a soluble protein purified from the fermentation media and spray dried to yield a shelf-stable product that possesses the functionality of milk protein. The intent is for r-βLG to be sold in bulk quantities direct to food manufacturers, as a non-animal source ingredient for use in a variety of foods such as nutritional products (bars, meal replacers, nutrition beverages, instant breakfasts), plant-based beverages, dairy-based products (fluid milk and drink mixes, cheese, dips, yogurt and fermented beverages, ice cream), desserts and confections, baked goods, dry mixes (e.g., for cakes, muffins, pancakes), breakfast cereal, savory snacks, spreads, sauces and condiments, soups, and egg substitute products. r-βLG is not intended for sale direct to consumers or for use in infant formula.

Health Canada has not previously authorized an animal-derived protein produced by microbial fermentation as a novel food.

The safety assessment performed by Food Directorate evaluators was conducted according to Health Canada's Guidelines for the Safety Assessment of Novel Foods. These Guidelines are based on harmonization efforts with other regulatory authorities and reflects international guidance documents in this area (e.g., Codex Alimentarius). The safety assessment considered the development and characterization of the r-βLG production strain; characterization, composition and nutritional quality of the r-βLG compared to whey protein from cow's milk; the potential for the r-βLG to be toxic or cause allergic reactions; and estimation of its level of consumption. Remilk has provided data to support that r-βLG is safe for use as food in Canada.

The Food Directorate has a legislated responsibility for the pre-market assessment of novel foods and novel food ingredients, as detailed in Division 28 of Part B of the Food and Drug Regulations (Novel Foods). r-βLG is considered to be a novel food under the following part of the definition of novel foods: "c) a food that is derived from a plant, animal, or microorganism that has been genetically modified such that

1. the plant, animal or microorganism exhibits characteristics that were not previously observed in that plant, animal or microorganism."

Development of the Production Organism

The production strain, K. phaffii yRMK-66, was developed starting from K. phaffii CBS 7435 (also known as NRRL Y-11430), a reference strain with a refined published genome. K. phaffii strains have a history of safe use in the production of food, feed, and pharmaceutical ingredients. K. phaffii is the current scientific name for the yeast species formerly classified as Pichia pastoris. This yeast species has simple requirements for growth similar to Saccharomyces cerevisiae, capacity to grow to high density, and amenability to expression and secretion of recombinant proteins through insertion of gene expression units for proteins of interest within the loci of the alcohol oxidase genes. In terms of previous novel food assessments, a strain of this species was developed to produce soy leghemoglobin, first authorized in Canada in 2020. There are no safety concerns with host strain.

The expressed protein of interest is the βLG variant B of cow (Bos taurus). As a mainstay animal in domestic food production, there is no concern with this gene donor source from a safety perspective. The codons were optimized for expression in the host yeast strain, but this resulted in no changes to the primary amino acid sequence relative to bovine βLG. Expression regulation elements were derived from the alcohol oxidase (aox1) locus from K. phaffii.

The prokaryote-derived antibiotic resistance markers used during strain development are commonly present in plasmids used for genetic modification. In keeping with biotechnology best practices, the development approach ensured the absence of these genes from the final strain, which was confirmed through strain characterization.

There are no concerns with respect to the donor organisms used in developing the production strain.

The r-βLG production strain was developed using standard and well-described genetic modification methods. The introduced expression unit contained a yeast codon-optimized version of the sequence encoding bovine β-lactoglobulin, under control of endogenous K. phaffii regulatory elements from alcohol oxidase. The amino acid sequence of the expressed β-lactoglobulin is identical to the bovine version of this protein. No concerns were raised with the genetic modification methodologies and construct designs.

Characterization of the Production Organism

Molecular characterization of K. phaffii strain yRMK-66 was performed using whole genome sequencing (WGS), integrating data obtained using two complementary methods in order to produce a complete, high-fidelity sequence. The sequence quality metrics were provided, as well as a description of the sequence processing pipeline for genome assembly. The refined sequence captured the four yeast chromosomes, the mitochondrial genome, and two "yeast killer plasmids". The latter are small linear extrachromosomal DNA that are relatively conserved across different yeast genera. When functional, the genes on these plasmids mediate the production and secretion of an exotoxin that kills other yeasts in proximity (discussed in Sturmberger et al., 2016^{Footnote 1}; reviewed in Schmitt and Breinig, 2006^{Footnote 2}). A study of P. pastoris (K. phaffii) killer plasmids, however, concluded that they were likely non-functional in this species. Given this context, the killer plasmids raise no concerns for human health.

Based on the information provided, the sequencing data quality, coverage breadth and accuracy were suitable for downstream analysis.

Commercial bioinformatics software was used to identify all sequence differences between the assembled genome for the r-βLG production strain yRMK-66 and the parental strain CBS 7435. Sequence analysis showed successful site-specific integration of the constructs for r-βLG expression. The coding regions were all found to be intact and contained no unintended modifications. Additional minor differences were identified outside of the insertion site, and bioinformatics analysis including search against databases of known coding sequences found no hits of concern with respect to safety. Antibiotic resistance markers used in development were absent from the production strain. Stability was assessed by real-time quantitative PCR and showed the number of genomic copies of the gene for r-βLG was not different during fermentation and in different runs.

The potential for the modifications to give rise to unintended and spurious proteins was investigated by searching the sequence for open reading frames (ORFs) and searching these against the NCBI database. Twelve putative ORFs were found. Four returned no hits to known proteins. The remaining matches were to segments of bacterial or yeast proteins, none of which were identified as being of concern to human health. The ORF analysis did not raise any safety concerns.

Overall, no concerns were raised with respect to the characterization of the r-βLG production strain, K. phaffii strain yRMK-66.

Product Information

The manufacturing of r-βLG was described as being in compliance with Good Manufacturing Practices, and includes preventive control plans designed to address post-market requirements. A list of chemicals used in the media for inoculation and fermentation stages and the solutions for downstream processing was provided, as well as their concentrations. Inputs included carbon sources, salts of various mineral and nutrient compounds, biotin, pH adjusting agents, and antifoaming agents. It was stated that these raw materials are of food grade or high-quality chemical or pharmaceutical grade. There are no concerns regarding the sourcing and specifications for these inputs from the microbial safety perspective.

The manufacturing process begins with starter culture stocks that are maintained in a working cell banking system to preserve batch-to-batch purity and uniformity. Growth scale-up proceeds in stages, where the biomass is increased in a series of vessels. The pH and temperature are tightly controlled and monitored. Methanol and glycerol are added to induce r-βLG expression and secretion into the media supernatant. Downstream processing to recover the protein at the end of fermentation proceeds by separation from the production cell biomass, followed by purification using filtration steps to concentrate the protein and remove low molecular weight impurities, germ filtration, and finally spray drying. The manufacturing equipment design, construction, operation, monitoring, cleaning, and maintenance described, including steps to prevent microbial contamination of the final product, present no concerns.

The sequence for mature βLG fused to a signal peptide for secretion was introduced into the production strain. Post-translational removal of the signal peptide by the cell's machinery yields the mature 162 amino acid protein with a molecular weight of 18.4 kDa. The protein is monomeric at pH lower than 3, exists as a dimer in milk, and can associate into higher order (multimeric) structures and aggregates, depending on the conditions. The submission presented data supporting protein characterization in order to assess the identity, structure, and purity of the protein, as well as to show equivalence to βLG found in cow's milk.

The coding sequence was optimized for expression in yeast, however the amino acid sequence was preserved, as confirmed by deduction from DNA sequencing. No DNA level differences were noted within the six introduced copies of the r-βLG coding sequence. SDS-PAGE results showed that r-βLG protein was of the same apparent size as a bovine βLG reference sample, and was of high purity. Proteomic analysis using liquid chromatography and mass spectrometry (LC-MS/MS) confirmed the amino acid sequence with high confidence and showed that r-βLG is the most abundant protein in all three test samples (estimated > 99.4 % of total protein). Aside from the signal peptide removal, no post-translational modifications were detected, either from the reference or the test samples. It was stated that the parental line secretes very low levels of endogenous yeast proteins, and removal of the cellular biomass during purification is expected to minimize the potential for host proteins to be present in the final product. Residual yeast proteins detected with this method presented no safety concerns.

In terms of the higher order structure, data was obtained using circular dichroism to assess the overall secondary structure. Protein folding and stability data was also provided, generated using an intrinsic fluorescence method and protein melting experiments. Results from these analyses all demonstrated a high degree of structural similarity between r-βLG and reference βLG protein from cow.

To demonstrate protein purity, HPLC traces from separate batches of r-βLG were provided showing a single major peak. Size exclusion chromatography (SEC) analysis performed under non-denaturing conditions similarly produced a single major peak, identified as the r-βLG (or cow βLG) dimer. In both assays, retention times were nearly identical between the reference and test samples.

Taken together, the protein characterization data support the equivalency of Remilk's r-βLG is to βLG from cow's milk. It also demonstrates that a final product of high purity can be manufactured consistently.

Dietary exposure

The petitioner estimated intakes of r-βLG based on the conservative assumption that r-βLG is added to all foods in each proposed food category at the maximum level (e.g., 5 to 35 % depending on the food). The estimates were determined using food consumption data from the Canadian Community Health Survey (CCHS) 2015.

Based on the proposed uses, the petitioner estimated all-ages mean and 90th percentile r-βLG intakes of 31.9 and 80.0 g/day.

r-βLG is intended as a partial replacement for milk-protein ingredients in foods. Based on the proposed uses, it is not expected to change the overall consumption, by the general population, of any foods to which it is added.

Nutrition

Under physiological conditions and in milk, bovine β-lactoglobulin exists as a dimer which dissociates into two monomers at a pH below 3.5. The molecular weight of the r-βLG is consistent with monomeric, bovine β-lactoglobulin. Given that the pH of gastric fluid is about 2, β-lactoglobulin dimers in food would mostly dissociate in the stomach into monomers that are substantially equivalent to the notified r-βLG.

Overall, the proximate (protein, fat, carbohydrate, moisture, and ash) composition of r-βLG is comparable to commercial whey protein isolate. However, the ash content (1 % w/w) is somewhat lower than the minimum level reported for conventional whey protein isolates (2 % w/w). Nonetheless, commercial whey protein isolate (WPI) is considered the most relevant conventional comparator for r-βLG based on overall similarities in proximate composition and possible food uses.

The iron and sodium composition of r-βLG was comparable to levels found in some commercial WPIs. Consistent with its relatively low ash content, however, the potassium (405 mg/kg) and calcium (418 mg/kg) content of r-βLG was below the minimum level reported for commercial WPI (potassium: 414 mg/kg; calcium: 566 mg/kg)^{Footnote 3}.

As the potassium content of r-βLG was only two percent lower than the minimum value reported for WPI, it is considered that use of r-βLG would not translate into a nutritionally significant difference in the potassium content of final foods compared to use of some conventional WPIs. Thus, adverse effects on potassium status are considered unlikely.

With the exception of dairy and dairy-substitutes, r-βLG is intended for use in food groups that are, overall, relatively minor contributors to total calcium intake of Canadians. Furthermore, some of these non-dairy foods may be fortified with calcium on a voluntary (e.g., plant-based beverages, flour) or mandatory (e.g., egg substitutes, meal replacements) basis. In the case of dairy products, use of either r- βLG or conventional WPI is expected to contribute a similarly insignificant amount of calcium compared to the much higher amounts intrinsic to the intended foods. For example, it is estimated that the addition of r-βLG at the maximum level to milk, yogurt, or cheese would result in a formulated product with a calcium content only 1 % lower than those made with an equivalent amount of some conventional WPIs. Thus, adverse effects on calcium status are considered unlikely.

Although similar, the amino acid composition of r-βLG is not considered equivalent to whey protein isolate. Notably, the average leucine (13.5 %) and threonine (4.7 %) content of r-βLG differed from the average levels reported in some commercial WPIs^{Footnote 4} expressed as a percentage of protein (leucine: 10.8 %; threonine: 7.2 %). These differences should be expected because r-βLG does not contain the other proteins found in bovine whey. Accordingly, the identified differences in amino acid composition are consistent with the different amino acid composition of bovine βLG compared to bovine α-lactalbumin— the next most abundant bovine whey protein after βLG.

Based on their similarities in amino acid sequence and protein structure, it is expected that the notified r-βLG and bovine βLG would be digested and absorbed in a similar manner. Bovine βLG is known to be relatively resistant to digestion in the stomach, but rapidly digested and absorbed in the small intestine. The second major bovine whey protein, α-lactalbumin, has also been reported to be resistant to gastric digestion which suggests that the digestibility of WPI may not greatly differ from its primary proteins. Given that whey protein is mostly composed of βLG, it is considered that the true digestibility of r-βLG is within the range of common food proteins (72-99 %)^{Footnote 5}, and may be approximated by the value for WPI.

Differences in r-βLG's amino acid composition and digestibility compared to a conventional comparator would be reflected by indicators of protein quality. However, the petitioner did not directly determine either of the indicators accepted by Health Canada to assess compliance with the protein quality provisions of the Food and Drug Regulations. Based on its directly measured amino acid composition and assuming equivalent true fecal digestibility^{Footnote 6}, it is estimated that the Protein Digestibility-Corrected Amino Acid Score (PDCAAS) of r-βLG would be the same (0.72) as some commercial WPIs^{Footnote 7}. This finding is supported by findings that ingestion of bovine βLG or whey protein elicits similar plasma amino acid responses and rates of muscle protein synthesis^{Footnote 8}.

r-βLG is proposed for use as an alternative protein source as part of a mixed diet. In this context, individuals are expected to consume a variety of food proteins. Differences in the digestibility and amino acid composition of these proteins will, naturally, lead to differences in protein quality, but it is possible to meet metabolic demands for protein by including a variety of foods which complement each other in terms of their amino acid profile. Only in an extreme scenario, in which high consumers exclude nearly all other sources of protein from their diet, is potential inadequacy possible. However, such a worst case scenario is unrealistic and implies the consumption of an unbalanced diet which is generally not recommended. It remains the manufacturer's responsibility to ensure that any formulated product complies with the applicable protein quality provisions of the Food and Drug Regulations. Thus, adverse health effects associated with inadequate total protein or essential amino acid intake are considered unlikely given the expected similarities between r-βLG and WPI.

Therefore, based on the information provided, the Bureau of Nutritional Sciences has not identified any nutritional concerns related to the proposed food use of r-βLG in Canada.

Chemistry

The petitioner provided specifications for r-βLG, including limits on trace elements, which they have established in the absence of specifications for this protein isolate in the Food Chemicals Codex.

The petitioner analyzed r-βLG for trace elements (arsenic, cadmium, lead, and mercury), 8 mycotoxins, and 17 pesticides deregistered in Canada and therefore considered to be environmental contaminants. The petitioner analyzed multiple batches of the r-βLG for these contaminants and reported either very low, or as was the case for the majority of results, 'not detected' concentrations. The analytical limits of detection (LOD) employed for all contaminants evaluated were deemed to be suitably low. Using either the quantified analytical results or assuming the LOD values, the Bureau of Chemical Safety (BCS) determined that addition of r-βLG to the specified foods is not expected to result in increased concentrations of chemical contaminants in the commodities to which it is added or would have a negligible impact on total dietary exposure to these contaminants. Based on the above information, the BCS is of the opinion that r-βLG does not pose a concern to human health from a chemical contaminants perspective.

BCS notes that the analytical results for arsenic, cadmium, lead and mercury are orders of magnitude lower than the specifications provided by the petitioner for these trace elements. In keeping with the principles of ALARA (as low as reasonably achievable), the BCS would encourage the petitioner to lower the specifications for the trace elements to more accurately reflect the concentrations of arsenic, cadmium, lead and mercury found in r-βLG protein.

A list of substances used during the manufacturing of r-βLG by fermentation of K. phaffii yRMK-66 was included in the submission. Ingredients used in the growth and fermentation media for cultures such as the r-βLG production organism are not typically regulated as food additives and usually there is no requirement under the Food and Drug Regulations for their premarket review. However, as with all substances used in food manufacture, their use must not result in a violation of section 4 of the Food and Drugs Act (FDA).

The BCS has not identified any chemical food safety concerns with the proposed uses of r-βLG produced by K. phaffii yRMK-66.

Toxicology

The petitioner demonstrated that r-βLG is identical to βLG found in bovine milk. Similarity was based on amino acid content, and comparison with a bovine βLG reference in SDS-PAGE and HPLC analyses. The Bureau of Microbial Hazards confirmed equivalency of r-βLG to βLG found in bovine milk. Therefore, βLG has a history of safe food use through the consumption of milk.

EFSA (2022)^{Footnote 9} recently evaluated βLG as a novel food and concluded it is safe for use in a variety of foods, based on the history of consumption of milk and absence of adverse effects in toxicological testing (genotoxicity and 90-day repeated dose oral toxicity in rats).

Although the βLG evaluated was extracted from milk rather than produced from fermentation of a production organism, both proteins were demonstrated to be identical to a reference standard for βLG from bovine milk. The NOAEL obtained from the 90-day study was the highest dose tested, 1000 mg/kg bw per day. The BCS agrees with the EFSA opinion.

The petitioner notes that the use of r-βLG is substitutional, being a replacement for milk protein ingredients in a variety of foods; as such, a similar dietary exposure for the protein is expected to occur.

The petitioner provided genotoxicity testing for r-βLG, which demonstrated the protein is not mutagenic (Ames assay OECD 471) or clastogenic (in vitro mammalian micronucleus test OECD 487) under the conditions tested.

The production organism, K. phaffii (previously known as P. pastoris), is classified as a Biosafety Level 1 (BSL-1) and has a long history of safe use as a production organism. In 2020, Health Canada authorized soy leghemoglobin (LegH) produced from a strain of P. pastoris as a novel food. No toxicological concerns were identified with the use of P. pastoris based upon bioinformatic analyses, in vitro digestion assays, and a 90-day oral toxicity study in rats.

Based on the available information, no toxicological concerns were identified with the proposed use of the novel protein r-βLG.

Allergenicity

The petitioner acknowledges that given the equivalency of r-βLG to βLG, individuals allergic to milk protein will also be allergic to r-βLG. The petitioner states that all products containing r-βLG will have labelling to indicate that the product contains a milk allergen and will comply with food allergen labelling requirements as provided under section B.01.010.1 of the Food and Drug Regulations (FDR).

To evaluate potential allergenicity of residual K. phaffii proteins (most are present in only trace amounts), the petitioner conducted bioinformatic analyses. Using Version 21 of the www.AllergenOnline.org database (February 7, 2022) and following Codex guidelines, no relevant matches were found. A literature search conducted by the petitioner did not identify reports of allergenicity associated with proteins from the host organism (K. phaffii, or P. pastoris).

Based on the available information, no allergenic concerns, other than to milk protein, were identified with the proposed use of the novel protein r-βLG.

Conclusion

Health Canada's review of the information presented in support of the use of r-βLG produced from K. phaffii yRMK-66 does not raise concerns related to food safety.

Health Canada's opinion refers only to the food use of r-βLG produced from K. phaffii yRMK-66.

This Novel Food Information document has been prepared to summarize the opinion regarding the subject product provided by the Food Directorate, Health Products and Food Branch, Health Canada. This opinion is based upon the comprehensive review of information submitted by the petitioner according to the Guidelines for the Safety Assessment of Novel Foods.

For further information about the safety assessment of this product, please contact:

Novel Foods Section
Food Directorate
Health Products and Food Branch
Health Canada, PL2204A1
251 Frederick Banting Driveway
Ottawa, Ontario K1A 0K9
bmh-bdm@hc-sc.gc.ca

Regulatory Requirements Beyond Pre-Market Food Safety Assessment: Labelling

In addition to the above-mentioned labelling requirements regarding food allergens, r-βLG and food products containing r-βLG must also comply with the labelling requirements as provided under the FDR and the Safe Food for Canadians Regulations (SFCR). The Canadian Food Inspection Agency (CFIA) enforces all these requirements.

For instance, paragraph 218(1)(a) of the SFCR requires that the label of a prepackaged food product bear the common name on the principal display panel. Further, as indicated in the CFIA's Industry Labelling Tool, the common name must be specific and must accurately identify or describe the food, in as simple and direct terms as possible, to allow a person to make an informed purchasing decision. In the case of r-βLG, for example, this could be achieved by including the name of the cell source material.

The Food and Drugs Act (FDA) section 6(1) and subsection 9(1) of the SFCR prohibit selling foods that are likely to be mistaken for standardized foods, unless they comply with the standard. Food with a standard of identity under the SFCR or food compositional standard under the FDR may only contain ingredients that are permitted to be added to the food via the standard. For instance, r-βLG would not meet the definition of "milk" "milk product", or "milk solid" as defined in Canadian Standards of Identity: Volume 1 – Dairy Products and thus would not be permitted to be added to standardized dairy products.

It would also not be permitted to use dairy ingredient common names or the collective class names "milk ingredients" or "modified milk ingredients" as outlined in the Common Names for Ingredients and Components document, which is a document incorporated by reference under the FDR. Further, as noted in item 9, Table 1 – Common names for specific ingredients or components of this document, the common name of "any protein isolate", including r-βLG, is the name of the source of the protein plus "protein" (e.g., "yeast-derived whey protein") or the common name of the protein isolate (e.g., "yeast-derived r-β-lactoglobulin").

In addition, food products containing r-βLG are also subject to the general prohibitions provided under section 5 of the Food and Drugs Act (FDA) and section 6 of the Safe Food for Canadians Act (SFCA), against information and claims that are false or misleading, or likely to create an erroneous impression. It must be noted that these prohibitions also apply to any images or claims appearing on the label of a food product. For example, claims must be truthful and should not give the impression that food products containing r-βLG are free from dairy without qualifying that there may be a risk to consumers allergic to milk.

For further information under CFIA responsibility of this product, including labelling requirements, please contact the CFIA.