WO2010062197A1

WO2010062197A1 - Dairy product and process

Info

Publication number: WO2010062197A1
Application number: PCT/NZ2009/000258
Authority: WO
Inventors: Eduard Nekrasov; Vasily Svetashev
Original assignee: Eduard Nekrasov; Vasily Svetashev
Priority date: 2008-11-25
Filing date: 2009-11-25
Publication date: 2010-06-03

Abstract

Processes for obtaining a ganglioside-enriched extract or one or more gangliosides from a ganglioside-containing lipid composition (GCLC) are disclosed. The processes comprise: (a) mixing a GCLC with one or more C₄-C₁₀ alcohols and water to produce a composition comprising an aqueous phase and an organic phase, and (b) substantially separating the aqueous phase from the organic phase, and wherein the aqueous phase comprises a ganglioside-enriched extract or one or more gangliosides.

Description

DAIRY PRODUCT AND PROCESS FIELD OF THE INVENTION

The invention generally relates to a gangliosides and ganglioside enriched extracts; process for obtaining gangliosides or ganglioside-enriched extracts from a ganglioside-containing lipid composition (GCLC); gangliosides and ganglioside-enriched extracts produced by the processes; and foods, beverages and pharmaceutical compositions comprising the gangliosides and ganglioside-enriched extracts.

BACKGROUND OF THE INVENTION

Gangliosides are a type of glycosphingolipid commonly found in the outer part of the plasma membrane of cells. Gangliosides are known to modulate cell signalling and transduction events. They have been implicated as being important in immunology, brain function and neurodegenerative diseases.

Gangliosides are one class of complex polar lipids found in relatively high concentrations in the brain of mammals. Glycosphingolipids such as gangliosides and cerebrosides are derivatives of the lipid sphingosine or a related amino alcohol that include a carbohydrate group as a polar head group. While cerebrosides have a monosaccharide polar head group, gangliosides have a complex oligosaccharide that includes the acidic sugar derivative sialic acid. Some non-acidic glycosphingolipids can contain two or more carbohydrate units such as lactosylceramide and more complex neutral oligoglycosylceramides.

Other sphingolipids include sphingomyelins, which are sphingolipids having a polar head group that is a phosphate derivative such as phosphocholine or phosphoethanolamine.

Other types of polar lipids include the glycerophosphoHpids. Both sphingomyelins and glycerophospholipids are phospholipids but in glycerophospholipids the lipid backbone is made of glycerol and fatty acids rather than ceramide, a fatty acid amide of sphingosine.

Due to their recently discovered properties, gangliosides are attractive targets for research. While present only in low concentrations in most tissues, brain tissue is a rich source of gangliosides. Extraction methods generally rely on the highly polar nature of the gangliosides to separate them from less-polar lipids such as triacylglycerides. However, ganglioside-rich tissues such as the brain also tend to be rich in other types of complex polar lipids, making it difficult to extract the gangliosides alone.

In particular, it is difficult to obtain a ganglioside-enriched product that does not also contain high levels of glycerophospholipids (phospholipids). Due to their similar solubility, many purification techniques are ineffective at separating these polar lipids from each other. However, for some applications a ganglioside material that is low in phospholipid content is preferred.

One of the most common methods of ganglioside extraction or concentration is die Folch partitioning method (Folch, et al., 1957). This method involves mixing a source of gangliosides in chloroform/methanol/water mixture (approximately 8:4:3) to form a two-phase system - a lower phase comprising mainly chloroform and an upper phase comprising mainly methanol and water. The polar gangliosides partition into the upper layer, along with other polar lipids such as phospholipids.

A modification of the Folch partitioning method can also be used to obtain a GCLC from a ganglioside-containing source. The modified Folch partitioning process involves mixing the source material with methanol, chloroform, and a small volume of water to extract the lipids. The gangliosides and other polar lipids are extracted into the solvent mixture to form a GCLC. Non- lipid material is not extracted.

The disadvantage of the Folch partitioning method and the modified Folch partitioning methods is that they both use a halogenated solvent and therefore cannot be used to obtain gangliosides for use in food processing or pharmaceutical applications.

Another three component system proposed for obtaining gangliosides is that of diisopropyl ether, butanol and aqueous NaCl (6:4:5, v/v/v) (Ladish and Gillard, 1985). This technique partitions gangliosides between the two phases that result from the mixture of diisopropyl ether, butanol and aqueous NaCl. The gangliosides tend to partition into the lower aqueous phase while the other lipids, including other complex polar lipids such as phospholipids, tend to partition into the upper organic phase.

Ladish and Gillard found that optimal separation of gangliosides from phospholipids was obtained using a diisopropyl ether:butanol ratio of 6:4. A higher proportion of butanol caused both phospholipids and gangliosides to be increasingly partitioned into the organic phase. If the aqueous NaCl solution was replaced with distilled water, a greater proportion of the phospholipids were partitioned into the aqueous phase.

The purified ganglioside fractions obtained by the Ladish process also contain complex neutral glycosphingolipids, i.e., glycosphingoHpids that incorporate polysaccharide chains but do not incorporate sialic acid.

Unfortunately, the Ladish solvent system is unsuitable for many applications as dϋsopropyl ether easily forms highly explosive peroxides.

Therefore, there is a need for a process for extracting gangliosides from ganglioside-containing lipid compositions that overcomes at least one of the disadvantages of the prior art, or at least provides the public with a useful choice.

SUMMARY OF THE INVENTION

In a first aspect the invention provides a process for obtaining a ganglioside-enriched extract from a ganglioside-containing lipid composition (GCLC), the process comprising (a) mixing a GCLC with one or more C₄-C₁₀ alcohols and water to produce a composition comprising an aqueous phase and an organic phase,

(b) substantially separating the aqueous phase from the organic phase, wherein the aqueous phase comprises a ganglioside-enriched extract.

In a second aspect, the invention provides a process for increasing the ratio of one or more gangliosides to one or more other complex lipids in a GCLC, the process comprising

(a) mixing a GCLC with one or more C₄-C₁₀ alcohols and water to produce a composition comprising an aqueous phase and an organic phase

(b) substantially separating the aqueous phase from the organic phase, wherein the wt% ratio of one or more gangliosides to one or more other complex lipids is increased in the aqueous phase compared to the wt% ratio of one or more gangliosides to one or more other complex lipids in the GCLC, and wherein the aqueous phase comprises a ganglioside-enriched extract. In one embodiment the one or more other complex lipids are selected from the group consisting of phospholipids, sphingolipids and neutral glycosphingolipids. Preferably, the one or more other complex lipids are phospholipids.

In another embodiment the wt% ratio of one or more gangliosides to the phospholipids is increased in the aqueous phase compared to the ratio in the GCLC.

In one embodiment of the above aspects the aqueous phase is about 10% to about 2000% enriched, preferably about 100% to about 1000%, more preferably about 200% to about 600% w/w enriched in gangliosides relative to one or more other complex lipids in comparison to the GCLC.

In a third aspect, the invention provides a process for increasing the ratio of one or more gangliosides to the total complex lipids in a GCLC, the process comprising (a) mixing a GCLC with one or more C₄-C₁₀ alcohols and water to produce a composition comprising an aqueous phase and an organic phase

(b) substantially separating the aqueous phase from the organic phase, wherein the wt% ratio of one or more gangliosides to total complex lipids is increased in the aqueous phase compared to the wt% ratio of one or more gangliosides to total complex lipids in the GCLC, and wherein the aqueous phase comprises a ganglioside-enriched extract.

In one embodiment the wt% ratio of one or more gangliosides to the total complex lipids is increased in the aqueous phase compared to the ratio in the GCLC.

In one embodiment of the above aspects the process further comprises recovering the aqueous phase comprising the ganglioside-enriched extract.

In another embodiment of the above aspects the process further comprises recovering one or more gangliosides from the ganglioside-enriched extract.

In one embodiment of the above aspects the one or more gangliosides are recovered from the ganglioside-enriched extract by reversed phase solid phase extraction. In one embodiment of the above aspects the enrichment is with respect to the total complex lipid content. In another embodiment the enrichment is with respect to a particular complex lipid, such as phospholipids.

In another aspect the invention provides a process for obtaining a ganglioside-enriched extract from a GCLC, the process comprising:

(a) mixing at least about 10 kg of a GCLC with butanol and water to produce a composition comprising an aqueous phase and an organic phase,

(b) substantially separating the aqueous phase from the organic phase, wherein the aqueous phase comprises a ganglioside-enriched extract, and wherein the ganglioside-enriched extract is at least about 100% enriched in gangliosides compared to the GCLC from which it was extracted.

Preferably the ganglioside-enriched extract is at least about 100% enriched, more preferably at least about 200% enriched, even more preferably 500% enriched, most preferably 1000% enriched.

In another aspect, the invention provides a process for extracting gangliosides from a GCLC, the process comprising (a) mixing a GCLC with one or more C₄-C₁₀ alcohols and water to produce a composition comprising an aqueous phase and an organic phase,

(b) substantially separating the aqueous phase from the organic phase, and

(c) recovering one or more gangliosides from the aqueous phase.

In one embodiment the one or more gangliosides are recovered from the aqueous phase by reversed phase solid phase extraction.

In one embodiment of the above processes of the invention, sphingomyelin is recovered from the organic phase.

In one embodiment of the above processes of the invention, phospholipids are recovered from the organic phase.

In one embodiment of the above processes of the invention, cerebrosides and diglycosylceramides are recovered from the organic phase. Preferably, the cerebrosides and diglycosylcer amides are selected from the group glucosylceramide, gakctosylceramide and lactosylceramides.

In another aspect the invention provides a ganglioside-enriched extract or one or more gangliosides obtained according to a process of the invention.

In another aspect the present invention provides a ganglioside-enriched extract that is at least about 100%, preferably 200%, more preferably 500%, most preferably 1000% enriched in gangliosides compared to the GCLC from which it was extracted.

In another embodiment the ganglioside-enriched extract is enriched in one or more gangliosides relative to the total complex lipids in the GCLC from which it was extracted.

In another embodiment the ganglioside-enriched extract is enriched in one or more gangliosides relative to the total phospholipids in the GCLC from which it was extracted.

In another aspect, the present invention provides a food or beverage comprising one or more gangliosides or one or more ganglioside-enriched extracts of the invention.

In one embodiment the food or beverage is an infant formula or a maternal formula.

In another aspect, the invention provides a pharmaceutical composition comprising one or more gangliosides or one or more ganglioside-enriched extracts of the invention.

The following embodiments may relate to any of the above aspects.

In one embodiment the GCLC is an extract of brain and other organs, organ fractions; colostrum and colostrum fractions; blood and blood fractions; dairy ingredients such as milk or milk fractions, buttermilk or buttermilk fractions, whey or whey fractions, milk fat globule membranes; dairy powder including dairy powder enriched in milk fat globular membrane proteins; egg, marine materials such as kina and kina fractions, sea urchin roe or sea urchin roe fractions. Preferably, the dairy ingredient is milk or a milk fraction, buttermilk or buttermilk fractions, milk fat globule membranes, dairy powder or a combination thereof.

Preferably, the GCLC is an extract of a dairy ingredient, animal brains, or sea urchin roe. In one embodiment, the GCLC comprises between about 10 to 100% w/w lipid, preferably between about 20 to 90% w/w, more preferably between about 30 to 90% w/w lipid.

In one embodiment, the GCLC comprises between about 10 to 100% w/w complex lipids, preferably between about 20 to 95% w/w more preferably between about 30-95% w/w complex lipids.

In one embodiment, the GCLC comprises about 0.5 to 10% w/w gangliosides, preferably 1 to 7% w/w, more preferably about 1 to 5% w/w gangliosides.

In one embodiment, the GCLC comprises 1 to 90% w/w phospholipids, preferably about 5 to 70% w/w phospholipids, more preferably about 10 to 60% w/w phospholipids.

In one embodiment, the GCLC comprises between 3 to 30% w/w sphingomyelin, preferably about 5 to 25% w/w sphingomyelin, more preferably about 7 to 20% w/w sphingomyelin.

In one embodiment of the above processes, the one or more C₄-C₁₀ alcohols is a C₄-C₈ alcohol, preferably a C₄-C₆ alcohol, more preferably a C₄ alcohol. In a more preferred embodiment the C₄-C₁₀ alcohol is n-butanol.

In one embodiment the C₄-C₁₀ alcohol is selected from the group comprising n-butanol, n- pentanol, n-hexanol, n-heptanol, n-octanol, n-nonanol or n-decanol. Preferably the C₄-C₁₀ alcohol is selected from n-butanol, n-hexanol and n-decanol, more preferably n-butanol.

In one embodiment the concentration of GCLC in the mixture of C₄-C₁₀ alcohol and water is between about 0.01% to about 20% w/w. In another embodiment the concentration of GCLC in the mixture is between about 0.5% to about 10% w/w, preferably between about 1% to about 5% w/w, more preferably between about 1% to about 3% w/w. In another embodiment the amount of GCLC in the mixture is between about 3% to about 5% w/w. In a more preferred embodiment the amount of GCLC in the mixture is about 1% w/w.

In one embodiment, the C₄-C₁₀ alcohol to water ratio is about 95:5 to 5:95, preferably about 70:30 to 30:70, more preferably about 60:40 to 40:60 w/v. In a preferred embodiment the alcohol to water ratio is about 50:50 v/v. In one embodiment the GCLC is mixed with one ot more C₄-C₁₀ alcohols and a buffet to produce a composition comprising an aqueous phase and an organic phase. Preferably the buffer is citric acid-sodium citrate buffer. Preferably the pH of the aqueous phase is between about 5 and 8. More preferably, the pH of the aqueous phase is between about 5 and 6, most preferably the pH of the aqueous phase is about 6.

In one embodiment a chelating agent is added to the water before or after mixing with one or more C₄-C₁₀ alcohols and the GCLC. Preferably the chelating agent is added to the water before mixing with one or more C₄-C₁₀ alcohols and the GCLC. Preferably the chelating agent is EDTA, citric acid or a salt of citric acid. More preferably, the chelating agent is EDTA or trisodium citrate.

In this specification, where reference has been made to patent specifications, other external documents, or other sources of information, this is generally for the purpose or providing a context for discussing the features of the invention. Unless specially stated otherwise, reference to such external documents or such sources of information is not to be construed as an admission that such documents or such sources of information, in any jurisdiction, are prior art or form part of the common general knowledge in the art.

It is intended that reference to a range of numbers disclosed herein (for example, 1 to 10) also incorporates reference to all rational numbers within that range (for example, 1, 1.1, 2, 3, 3.9, 4, 5, 6, 6.5, 7, 8, 9 and 10) and also any range of rational numbers within that range (for example, 2 to 8, 1.5 to 5.5 and 3.1 to 4.7) and, therefore, all sub-ranges of all ranges expressly disclosed herein are hereby expressly disclosed. These are only examples of what is specifically intended and all possible combinations of numerical values between the lowest value and the highest value enumerated are to be considered to be expressly stated in this application in a similar manner.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a graph showing the molar ratio of sialic acid (SA) to phospholipids (PL) in the feed streams A-E after butanol/water partitioning, as outlined in Example IB. S = starting material, B = butanol phase and W = water phase. Figure 2 is a gtaph showing the effect of the butanol to water ratio on the molar ratio of sialic acid (SA) to phospholipids (PL) in feed stream E after butanol/ water partitioning as outlined in Example 3. S = starting material, B = butanol phase and W = water phase.

Figure 3 is a graph showing the effect of the concentration of feed stream E in the butanol/ water mixture on the molar ratio of sialic acid (SA) to phospholipids (PL) in each phase after butanol/water partitioning as outlined in Example 4. S = starting material, B = butanol phase and W = water phase.

Figure 4 is a graph showing the effect of different alcohols and different temperatures on the molar ratio of sialic acid (SA) to phospholipids (PL) in each phase after alcohol/ water partitioning, as outlined in Example 5. S = starting material, B = butanol phase, H = hexanol phase, W = water phase after partitioning with butanol or hexanol, and D-W = water phase after partitioning with n-decanol.

DESCRIPTION OF THE INVENTION

1. Definitions

The term "comprising" as used in this specification means "consisting at least in part of. When interpreting each statement in this specification that includes the term "comprising", features other than that or those prefaced by the term may also be present. Related terms such as "comprise" and "comprises" are to be interpreted in the same manner.

The term "complex lipid" as used in this specification means a lipid selected from the group comprising phospholipids and sphingoHpids including glycosphingolipids (cerebrosides, neutral di- and oligoglycoceramides and gangliosides), ceramides and sphingomyelins. Complex lipids may be found in milk and other dairy ingredients. Other sources of some complex lipids include any animal tissue but especially brain and nervous tissue, eggs, fish, deer velvet, deer antler and plant lipids. Preferably the complex lipids extracted according to the present invention are derived from a dairy ingredient. Suitable dairy ingredients include colostrum, milk, fractions of colostrum or fractions of milk. Preferably the dairy ingredient is derived from cow, buffalo, goat, sheep or human milk. Most preferably the dairy ingredient is cow-derived. All of the complex lipids in, for example, a ganglioside containing source, GCLC or ganglioside-enriched extract are collectively referred to as the "total complex lipids". The term "gangHoside-containing lipid composition" (GCLC) as used in this specification means a composition or material that comprises a significant proportion of lipid and contains one or more gangliosides.

The term "infant formula" as used in this specification means a composition for infants aged between 0 days and 6 months old. The term "follow-on formula" as used in this specification means a composition for infants aged 6 months to 1 year. The term "growing up formula" as used in this specification means a compositions directed to infants and children aged 1 year upwards. Growing-up formula includes growing-up milk powders or GUMPs. It will be appreciated by those skilled in the art that the age ranges for the different compositions: "infant formula", "follow-on formula" and "growing-up formula" can vary from child to child depending on the individual's development. These products may be in liquid form as concentrates or ready-to-drink liquids or provided as powder concentrates.

The term "maternal formula" as used in this specification means a composition for pregnant women to take during pregnancy.

The term "solids" as used in this specification means the dry material that remains once water and other solvents have been removed. The "solids" may include material that is in the solid state or liquid state, for example, oils.

2. Ganglioside-containing lipid compositions

GCLCs are feed materials for the processes of the invention. The GCLC for use in the processes of the invention may comprise or be obtained from any suitable source of gangliosides. Suitable sources of gangliosides include but are not limited to animal tissues such as brain and other organs, organ fractions; colostrum and colostrum fractions; blood and blood fractions; dairy ingredients such as milk or milk fractions, buttermilk or buttermilk fractions, whey or whey fractions, milk fat globule membranes; dairy powder including dairy powder enriched in milk fat globular membrane proteins; egg, marine materials such as kina and kina fractions, sea urchin roe or sea urchin roe fractions. Preferably the dairy ingredient is derived from cow, buffalo, goat, sheep or human milk. Most preferably the dairy ingredient is cow-derived. A GCLC for use in the processes of the invention may be obtained by any process known in the art. In one embodiment, the GCLC is obtained by solvent extraction of a ganglioside-containing source, for example by Folch lipid extraction using methanol and chloroform. If the ganglioside- containing source is a solid or semi-solid, generally, it will first be blended, chopped, pulverised, crushed, comminuted and/ or ground.

In another embodiment, the GCLC is obtained by supercritical or near-supercritical extraction of a ganglioside-containing source. Preferably the ganglioside-containing source is a dairy ingredient, such as a milk fraction. Methods of supercritical or near supercritical extraction of a ganglioside-containing source are discussed in, for example, WO 2004/066744 and WO 2006/041316.

In one embodiment the GCLC is obtained by supercritical or near supercritical extraction of ganglioside-containing dairy ingredients. In one embodiment the GCLC is obtained by near supercritical extraction of dairy ingredients using dimethyl ether. In another embodiment the

GCLC is obtained by supercritical extraction of dairy material using CO₂ and ethanol co-solvent.

The proportion of lipid in the GCLC depends on the nature of the ganglioside-containing source and the method of its preparation.

In addition to one or more gangliosides, GCLCs generally also comprise a mixture of other complex lipids, in addition to fatty acids and other non-lipid components.

In one embodiment the GCLC may also contain one or more other complex lipids including but not limited to glycerophospholipids such as phosphatidyl choline, phosphatidyl inositol, phosphatidyl glycerol, cardiolipin, phosphatidyl serine, phosphatidyl ethanolamine, ethanolamine and their lyso forms; phosphosphingolipids such as sphingomyelin; cerebrosides like lactosylceramide (LacCer), galactosylceramide (GalCer) and glucosylceramide (GluCer); and diglycosylceramides such as lactosylceramide (LacCer).

In one embodiment the GCLC may also contain one or more non-polar lipids such as triacylglycerides; sterols such as cholesterol and cholesterol esters; and beta-carotene.

A GCLC for use in the processes of the invention may also contain non-lipid material such as protein and/or carbohydrates, for example lactose. In one embodiment the GCLC comprises at least about 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95 or 100 % w/w lipid, and useful ranges may be selected between any of these value (for example, about 5 to about 100%, about 10 to about 100%, about 15 to about 100%, about 20 to about 100%, about 25 to about 100%, about 30 to about 100%, about 35 to about 100%, about 40 to about 100%, about 45 to about 100%, about 50 to about 100%, about 20 to about 90%, about 30 to about 90%, about 40 to about 90%, about 50 to about 90%, about 60 to about 90%, about 70 to about 90%, and about 80 to about 90% w/w lipid).

In one embodiment the GCLC comprises at least about 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95 or 99.5% w/w complex lipid, and useful ranges may be selected between any of these values (for example, about 5 to about 95%, about 10 to about 95%, about 15 to about 95%, about 20 to about 95%, about 25 to about 95%, about 30 to about 95%, about 35 to about 95%, about 40 to about 95%, about 45 to about 95%, about 50 to about 95%, about 60 to about 95%, about 65 to about 95%, about 70 to about 95%, about 75 to about 95%, about 80 to about 95%, about 85 to about 95%, about 90 to about 95%, about 10 to about 70%, about 15 to about 70%, about 20 to about 70%, about 25 to about 70%, about 30 to about 70%, about 35 to about 70%, about 40 to about 70%, about 45 to about 70%, about 50 to about 70%, about 55 to about 70%, about 60 to about 70%, about 65 to about 70% w/w complex polar lipid).

Complex lipids present in die GCLC may include but are not limited to phospholipids, such as phosphatidyl inositol, phosphatidyl glycerol, phosphatidyl serine, phosphatidyl choline, phosphatidyl ethanolamine, cardiolipin and their lyso forms; acylalkylphospholipids; sphingolipids such as sphingomyelins, ceramides, ceramide aminoethylphosphonate and sphingosines, and glycosphingoHpids such as gangliosides, cerebrosides and diglycosylceramides.

In one embodiment the GCLC comprises at least about 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, 9.0, 9.5 or 10.0 wt% gangliosides, and useful ranges may be selected between any of these values (for example, 0.5 to about 10%, about 1 to about 10%, about 2 to about 10%, about 3 to about 10%, about 0.5 to about 9%, about 0.5 to about 8%, about 0.5 to about 7%, about 0.5 to about 6%, about 0.5 to about 5%, about 0.5 to about 4%, about 0.5 to about 3%, about 0.5 to about 2%, about 0.5 to about 1%, about 1 to about 5%, about 1.5 to about 4%, about 2 to about 4%, about 2.5 to about 4%, about 3 to about 4%, about 1 to about 7.%, about 2 to about 7, about 3 to about 7%, about 4 to about 7%, about 5 to about 7% and about 6 to about 7%). In one embodiment the GCLC comprises at least about 1, 2, 3, 4, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90 or 99.5% w/w phospholipids and useful ranges may be selected between any of these values (for example, about 1 to about 95%, about 2 to about 95%, about 3 to about 95%, about 4 to about 95%, about 5 to about 95%, about 10 to about 95%, about 15 to about 95%, about 20 to about 95%, about 25 to about 95%, about 30 to about 95%, about 35 to about 95%, about 40 to about 95%, about 45 to about 95%, about 50 to about 95%, about 55 to about 95%, about 60 to about 95%, about 65 to about 95%, about 70 to about 95%, about 80 to about 95%, about 85 to about 95%, about 90 to about 95%, about 10 to about 70%, about 15 to about 70%, about 20 to about 70%, about 25 to about 70%, about 30 to about 70%, about 35 to about 70%, about 40 to about 70%, about 45 to about 70%, and about 50 to about 70% w/w phospholipid).

In one embodiment the GCLC comprises at least about 0.1, 0.2, 0.3, 0.4, 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28 or 30 % w/w sphingomyelin, and useful ranges may be selected between any of these values (for example, about 0.1 to about 30%, about 0.5 to about 30%, about 1 to about 30%, about 2 to about 30%, about 3 to about 30%, about 4 to about 30%, about 5 to about 30%, about 10 to about 30%, about 15 to about 30%, about 20 to about 30%, about 0.1 to about 28%, about 6.1 to about 25%, about 0.1 to about 20%, about 0.1 to about 18%, about 0.1 to about 16%, about 0.1 to about 14%, about 0.1 to about 12%, about 0.1 to about 10%, about 0.1 to about 8%, about 0.1 to about 6%, about 0.1 to about 5%, about 0.1 to about 4%, about 0.1 to about 3%, about 0.1 to about 2%, about 0.1 to about 1%, about 0.1 to about 0.5%, about 2 to about 20%, about 3 to about 20%, about 4 to about 20%, about 5 to about 20%, about 6 to about 20%, about 7 to about 20%, about 8 to about 20%, about 9 to about 20% and about 10 to about 20% w/w sphingomyelin).

3. Alcohols

The processes of the invention include a step of mixing a GCLC with a C₄-C₁₀ alcohol. C₄-C₁₀ alcohols for use in the process may be linear or branched. In one embodiment the C₄-C₁₀ alcohol is a linear alcohol selected from the group consisting of n-butanol, n-pentanol, n-hexanol, n- heptanol, n-octanol, n-nonanol or n-decanol. In another embodiment the C₄-C₁₀ alcohol is a branched C₄-C₁₀ alcohol, such as sec-butanol or 2-methylbutan-2-ol. The C₄-C₁₀ alcohol may be monohydric, dihydric or ttihydtic. In a preferred embodiment the C₄-C₁₀ alcohol is monohydric.

The C₄-C₁₀ alcohol may contain one or more alkene groups. In one embodiment the C₄-C₁₀ alcohol contains one alkene group, for example hex-3-en-l-ol or oct-4-en-2-ol.

The processes of the invention may use one C₄-C₁₀ alcohol or a mixture of more than one C₄-C₁₀ alcohol.

In one embodiment the C₄-C₁₀ alcohol is n-butanol (butanol). Butanol was found to maximise extraction of gangliosides containing complex di, tri, or polysialic acids. However, other C₄-C₁₀ alcohols such as hexanol may provide a better separation for simple monosialic acid gangliosides (see Example 5). Therefore, the selection of C₄-C₁₀ alcohol for use in the processes of the invention may depend on the ganglioside composition in the GCLC to be extracted, and the particular method of separation used. Where GCLC originates from a dairy source, butanol is preferred.

The processes of the invention allow undesirable solvents such as chloroform and dϋsopropyl ether to be avoided.

4. Mixing the GCLC with one or mote C₄-C₁₀ alcohols and water

In the processes of the invention, the GCLC is mixed with one or more C₄-C₁₀ alcohols and water. Buffers and/or salts may be added to the water to promote selective partitioning of lipids between the phases. Their presence may also aid later separation of the aqueous and organic phases into separate fractions. Citric acid-sodium citrate buffer is a preferred additive, as is NaCl.

Chelating agents such as EDTA or trisodium citrate may also be added to chelate ions that may interfere with ganglioside partitioning.

Therefore, the term "water" in the context of the methods of the invention, includes aqueous solutions of water and other additives.

4.1 GCLC concenttation

As shown in Example 4, varying the concentration of GCLC in the alcohol/water mixture may affect a) partitioning of the lipid components of the GCLC into the aqueous and organic phases, b) dissolution of the GCLC and water in butanol, and c) residue and flake (insoluble precipitate) formation.

For example, if the concentration of GCLC is high, more mixing time will be needed to affect partitioning of the lipid components. Preliminary heating of the mixture can improve phase separation where the GCLC is present in high concentration (see section 4.3 below).

A high concentration of GCLC can also reduce the solubility of the lipids in the aqueous phase, causing solid flakes of material to accumulate in the aqueous phase and interphase. This material can be recovered by filtration. Where the concentration of GCLC is high, the proportion of solid product material that is retained on the filter may be up to 18% of the total solids that can ultimately be recovered from the water phase.

In one embodiment the concentration of GCLC in the mixture is at least 0.01, 0.05, 0.1, 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 or 20% w/w, and useful ranges may be selected between any of these values (for example, about 0.01 to about 20%, about 0.05 to about 20%, about 0.1 to about 20%, about 0.5 to about 20%, about 1.0 to about 20%, about 2.0 to about 20%, about 3.0 to about 20%, about 4.0 to about 20%, about 5.0 to about 20%, about 10.0 to about 20%, about 15.0 to about 20%, about 0.01 to about 5.0%, about 0.05 to about 5.0%, about 0.1 to about 5.0%>, about 0.2 to about 5.0%, about 0.5 to about 5.0%, about 1.0 to about 5.0%, about 2.0 to about 5.0%, about 3.0 to about 5.0%, about 4.0. to about 5.0%, about 0.01 to about 10.0%, about 0.05 to about 10.0%, about 0.1 to about 10.0%, about 1.0 to about 10.0%, about 2.0 to about 10.0%, and about 5.0 to about 10.0% w/w).

4.2 Alcohol to water ratio

Varying the C₄-C₁₀ alcohol/water ratio in the mixture of GCLC, C₄-C₁₀ alcohol and water may significantly affect (a) the level of ganglioside enrichment in the product, and (b) the ratio of gangliosides to phospholipids partitioned into the water phase.

A higher C₄-C₁₀ alcohol/water ratio provides a higher ratio of gangliosides to phospholipids in the aqueous phase. However, a lower C₁-C₄ alcohol/ water ratio leads to higher partitioning of gangliosides into the aqueous phase providing a higher yield of gangliosides. The C₄-C₁₀ alcohol/water ratio may be varied to obtain the desired results, provided that the alcohol forms a biphasic system with water at the ratio selected.

In one embodiment, the C₄-C₁₀ alcohol/water ratio is at least 5:95, 10:90, 15:85, 20:80, 25:75, 30:70, 35:65, 40:60, 45:55, 50:50, 55:45, 60:40, 65:35, 70:30, 75:25, 80:20, 85:15, 90:10 or 95:5 w/w and useful ranges may be selected between any of these values (for example, about 5:95 to about 95:5, about 10:90 to about 95:5, about 15:85 to about 95:5, about 20:80 to about 95:5, about 25:75 to about 95:5, about 30:70 to about 95:5, about 35:65 to about 95:5, about 40:60 to about 95:5, about 45:55 to about 95:5, about 50:50 to about 95:5, about 55:45 to about 95:5, about 60:40 to about 95:5, about 65:5 to about 95:5, about 70:30 to about 95:5, about 75:5 to about 95:5, about 80:20 to about 95:5, about 85:15 to about 95:5, about 80:20 to about 95:5, about 75:25 to about 95:5, about 90:10 to about 95:5, about 5:95 to about 90:10, about 5:95 to about 85:15, about 5:95 to about 80:20, about 5:95 to about 75:25, about 5:95 to about 70:30, about 5:95 to about 65:35, about 5:95 to about 60:40, about 5:95 to about 55:45, about 5:95 to about 50:50, about 5:95 to about 45:50, about 5:95 to about 40:60, about 5:95 to about 35:65, about 5:95 to about 30:70, about 5:95 to about 25:75, about 5:95 to about 20:80, about 5:95 to about 15:85, about 5:95 to about 10:90 w/w).

In one embodiment the C₄-C₁₀ alcohol and the water constitute over 95% of the solvent mixed with the GCLC. Small amounts of other organic solvents may be present, provided that they do not significantly interfere with partitioning of the ganglioside into the aqueous phase. Preferred extraction solvents are butanol/water mixtures, preferably between 30/70 and 70/30, more preferably between 40/60 and 60/40, and most preferably 50/50.

4.3 Mixing

In the processes of the invention the GCLC is mixed with one or more C₄-C₁₀ alcohols and water to produce a composition comprising an aqueous phase and an organic phase. During mixing, the components of the GCLC partition into one of the two phases. The "aqueous phase" may also be referred to as the "water phase".

The GCLC, C₄-C₁₀ alcohol and water components may be mixed in any order. In one embodiment the GCLC is added to a mixture of one or more C₄-C_n, alcohols and water. In a preferred embodiment the GCLC is mixed in the one or more C₄-C_1n alcohols before the addition of water. In one embodiment of the processes of the invention the GCLC is mixed with one or more C₄- C₁₀ alcohols and water at room temperature. In another embodiment the GCLC is mixed with one or more C₄-C₁₀ alcohols and water at between about 5°C and room temperature. In another embodiment the GCLC is mixed with one or more C₄-C₁₀ alcohols and water at a temperature between room temperature and about 95°C. In a preferred embodiment the GCLC is mixed with one or more C₄-C₁₀ alcohols and water at a temperature between room temperature and about 70°C, preferably between about 5O⁰C and 70⁰C.

In one embodiment, the GCLC is mixed with one or more C₄-C₁₀ alcohols at a temperature between 50⁰C and 70⁰C before the addition of water.

The GCLC, one or more C₄-C₁₀ alcohols and water may be mixed using any method known in the art, such as agitation, shaking, stirring or by inverting the vessel containing the GCLC, C₄-C₁₀ alcohol and water.

The intensity of mixing may vary from vigorous to gentle. Without wishing to be bound by theory, higher intensity mixing will generally facilitate the partitioning of gangliosides into the aqueous phase better than lower intensity mixing. However, high intensity mixing can lead to the formation of emulsions from which it is difficult to separate the organic and aqueous phases.

In one embodiment, the mixing is carried out at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90 or 95°C and useful ranges may be selected between any of these values (for example, about 5 to about 95⁰C, about 10 to about 95°C, about 20 to about 95°, about 30 to about 95°C, about 40 to about 95°C, about 50 to about 95°C, about 60 to about 95°C, about 70 to about 95°C, about 80 to about 95°C, about 90 to about 95°C, about 5 to about 30⁰C, about 10 to about 30⁰C, about 20 to about 30⁰C, about 5 to about 60⁰C, about 10 to about 60⁰C, about 20 to about 6O⁰C, about 30 to about 60⁰C, about 40 to about 60⁰C and about 50 to about 60⁰C).

In one embodiment, the mixing is carried out for between 1 second and 48 h, preferably for between 5 minutes and 4 hours, more preferably, between 20 minutes and 2 hours. The period of mixing necessary to partition the one or more gangliosides into the aqueous phase of the composition may vary depending on the nature of the GCLC, the temperature at which the composition is mixed and the intensity of mixing. High intensity mixing may need to be followed by a larger time period to achieve satisfactory separation of the aqueous and organic phases. Persons skilled in the art will be able to modify the mixing process to achieve the desired, separation.

5. Separating the aqueous and organic phases

The composition produced by mixing the GCLC, C₄-C₁₀ alcohol and water in the processes of the invention forms two substantially distinct layers: an organic phase and an aqueous phase. These phases are substantially separated — i.e. the phases form two discernable layers, although one phase may contain solvent present in the other phase. In some embodiments the demarcation between the two phases is clear, while in other embodiments the interface between the solvents is less clear (for example, when mixing of the solvents results in the formation of an interphase). Preferably the demarcation between the two phases is clear. In one embodiment the separation process comprises separation of the composition into distinct organic and aqueous phases. Separation into distinct phases can be achieved by, for example, allowing the phases to settle so the lighter organic phase rises to the top of the mixture, with the heavier aqueous phase on the bottom. Other phase separation techniques are known in the art. In another embodiment the separation process comprises physical separation of the organic and aqueous phases into distinct fractions.

In one embodiment the process further comprises recovering one or more gangliosides from the aqueous phase.

Separation of the organic and aqueous phases may be carried out using any suitable technique known in the art, such as, by allowing the mixture to stand, centrifuging the mixture, or by gently agitating the mixture, for example by slowly stirring the mixture. Various factors, such as temperature, agitation and the intensity of mixing, influence the time required for the organic phase and aqueous phase of the composition to separate.

In one embodiment where separation does not comprise physical separation of the two phases into distinct fractions, the process of the invention may further comprise recovery of the aqueous phase which comprises a ganglioside-enriched extract.

Different buffers and pHs may aid separation depending on the types of gangliosides present in the GCLC. In one embodiment the pH of the aqueous phase is between about 5-9. Preferably, the pH of the aqueous phase is about 5 to 8. Mote preferably, the pH of the aqueous phase is about 5 to 6.

In one embodiment, the composition is separated at a temperature between 5°C and room temperature. In another embodiment the mixture is separated at a temperature between room temperature and 95°C. In a preferred embodiment the mixture is separated at a temperature between room temperature and 70°C. In another preferred embodiment the mixture is separated at a temperature between 50°C and 70°C.

In one embodiment, the mixture is heated and cooled alternately to effect the separation. In a preferred embodiment the mixture is heated to a temperature between 50⁰C and 70⁰C and then cooled to room temperature to effect the separation.

In one embodiment, the mixture is separated over a period of time between 1 second and 48 h. In a preferred embodiment the mixture is separated over a period of time between 1 second and 4 h. The period of time required for the mixture to separate depends on the intensity of mixing. Without willing to be bound by theory, the time taken for separation is generally longer when high intensity mixing has been used.

The organic and aqueous phases can be separated into distinct fractions from one another using any method known in the art. In one embodiment the upper layer is decanted from the lower layer. In another embodiment the lower layer is drained from the upper layer. In another embodiment the upper layer is removed from the lower layer by batch or continuous centrifugation, by batch or continuous clarification, by batch or continuous decanting or by any other means known in the art.

6. Recovering the gangliosides

In one embodiment, one or more gangliosides may be recovered from the ganglioside-enriched extract. The ganglioside-enriched extract may be enriched in one or more gangliosides compared to the GCLC feed material. The term "one or more gangliosides" may refer to all of the gangliosides present in, for example, a ganglioside containing source, GCLC or ganglioside- enriched extract, collectively or may refer to a particular ganglioside, for example GD₃, present in a sample. In the processes of the invention one or more gangliosides may be recovered from the ganglioside-enriched extract using any method known in the art. In one embodiment the gangliosides are recovered from the aqueous phase by removing the water. Water may be removed using any method known in the art, for example by distillation or evaporation under vacuum, freeze drying, or spray drying.

Gangliosides may also be recovered from the ganglioside-enriched extract by standard separation processes such as dialysis followed by freeze drying, solid-phase extraction (SPE) in particular, reversed phase SPE, and different types of chromatography including preparative thin layer chromatography (TLC) and column chromatography such as anion-exchange chromatography on DEAE-Sephadex™.

7. Optional processes

In one embodiment, the organic phase obtained in step (b) of the processes of die invention is mixed with further C₄-C₁₀ alcohol and water, then step (b) is repeated.

Repeating the mixing and separation steps may increase the degree of partitioning and recovery of the complex lipids present in the GCLC.

In some circumstances, when the GCLC is mixed with C₄-C₁₀ alcohol and water, some gangliosides and other complex lipids may form an insoluble solid material in the aqueous phase or at the organic/aqueous interface. This will occur when the complex lipids are not completely soluble in the solvent system selected under the conditions used. This solid material can be recovered by filtration. To ensure that the lipid component of this solid material is recovered, the solid material may be eluted with chloroform/methanol (2:1, v/v) and filtered. Non-lipid material will not dissolve and will be retained on the filter. lipid material will dissolve. The solvent can be removed from the dissolved lipid, which can then be mixed with C₄-C₁₀ alcohol and water under conditions under which it remains soluble.

In one embodiment, the processes of the invention additionally comprise removing contaminants from die aqueous fraction obtained in step (b). Contaminants may be removed by one or more processes selected from the group comprising dialysis, ultrafiltration, reversed-phase solid phase extraction (RP-SPE) and different types of chromatography including anion-exchange chromatography, gel filtration, TLC and others. Contaminants may include oil-soluble flavours, pigments, sugars such as lactose, amino acids, short chain peptides, inorganic salts and urea. Contaminants may also include small amounts of other complex lipids that were recovered along with the gangliosides in the aqueous phase.

The fraction eluted with methanol and chloroform/methanol can be evaporated to provide a ganglioside-enriched extract.

8. Scale

A process of the invention may be carried out on any reasonable scale. In one embodiment, die process may be carried out on laboratory scale. In another embodiment the process may be carried out on pilot plant scale. In another embodiment the process may be carried out on an industrial scale. Preferably, a process of the invention is carried out on an at least a pilot plant scale.

A process of the invention may be carried out as a batch process or a continuous process.

The scale of the process may be defined by, for example, the mass of GCLC used; and/ or the rate of production of ganglioside-enriched extract.

In one embodiment, the scale of the process is defined by die mass of GCLC used. Preferably, the mass of GCLC used is at least about 10 mg, 100 mg, 1 g, 10 g, 100 g, 1 kg, 10 kg, 100 kg, 1 T, 10 T, 100 T, or 1000 T. Preferably, the mass of GCLC used is at least about 100 g. More preferably, the mass of the GCLC used is at least about 1 kg, most preferably 10 kg.

In another embodiment the scale of die process is defined by the rate of production of ganglioside-enriched extract.

9. The ganglioside-enriched extract

In one embodiment, die invention provides a ganglioside-enriched extract produced by a process of the invention. The degree of enrichment can be determined by comparing die relative weight of these gangliosides in the GCLC with the relative weight of the ganglioside in the solids present in the ganglioside-enriched extract. Where the GCLC comprises solvent material, the ganglioside content is measured with respect to the solids present in the GCLC, not the total weight of the GCLC sample.

In one embodiment the ganglioside-enriched extract may be enriched in one or more gangliosides compared to the GCLC from which it was obtained. In other words, the wt% ratio of one or more gangliosides in the ganglioside-enriched extract is increased compared to the GCLC from which the extract was obtained.

The enrichment in one or more gangliosides may be relative to the total lipid content, the complex lipid content, or the content of a particular complex lipid, when compared to the content of the GCLC. Increasing the ratio of one or more gangliosides to other components in the GCLC may produce a ganglioside-enriched extract that can more easily be purified to provide one or more gangliosides.

As would be understood by a person of skill in the art, depending on the levels of water or other solvent present, the ganglioside-enriched extract may in fact contain less wt% gangliosides then the GCLC feed material. However, the wt% of one or more gangliosides relative to one or more of (a) the total lipid content, (b) the complex lipid content, or (c) the content of a particular complex lipid will be increased in the ganglioside-enriched extract when compared to the GCLC.

For example, if a GCLC with a ganglioside content of 3% w/w of total solids is subjected to a process of the invention to provide a ganglioside-enriched extract comprising 6% w/w gangliosides (as a wt % of total lipids), the ganglioside-enriched extract is enriched by 100% compared to the GCLC. Similarly, a 10% enrichment is a 1.1 fold increase and a 1000% enrichment is a 11 fold increase. Preferably, the ganglioside-enriched extract is enriched by about 10-2000%, more preferably by about 100-1000%, most preferably by about 200-600%.

In one embodiment the ganglioside-enriched extract is enriched by at least 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 125, 150, 175, 200, 225, 250, 275, 300, 325, 350, 375, 400, 425, 450, 475, 500, 525, 550, 575, 600, 625, 650, 675, 700, 725, 750, 775, 800, 825, 850, 875, 900, 925, 950, 975 or 1000% when compared to the GCLC, and useful ranges may be selected between any of these values (for example, about 10 to about 1000, about 20 to about 1000, about 30 to about 1000, about 40 to about 1000, about 50 to about 1000, about 100 to about 1000, about 200 to about 1000, about 300 to about 1000, about 400 to about 1000, about 500 to about 1000, about 600 to about 1000, about 700 to about 1000, about 800 to about 1000, about 900 to about 1000, about 20 to about 300, about 30 to about 300, about 40 to about 300, about 50 to about 300, about 60 to about 300, about 70 to about 300, about 80 to about 300, about 90 to about 300, about 100 to about 300, about 150 to about 300, about 200 to about 300, about 250 to about 300, about 20 to about 600, about 30 to about 600, about 40 to about 600, about 50 to about 600, about 60 to about 600, about 70 to about 600, about 80 to about 600, about 90 to about 600, about 100 to about 600, about 200 to about 600, about 300 to about 600, about 400 to about 600 and about 500 to about 600% compared to the GCLC).

In one embodiment, die ganglioside-enriched extract may be enriched in one or more gangliosides relative to the amount of complex lipids when compared to the GCLC.

For example, if gangliosides make up 5% w/w of complex lipids in a GCLC and the process of the invention provides a ganglioside-enriched extract where gangliosides comprise 20% w/w the complex lipid content, the ganglioside-enriched extract is enriched by 400% relative to the complex lipid content when compared to the GCLC.

In one embodiment the ganglioside-enriched extract is enriched by at least 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 125, 150, 175, 200, 225, 250, 275, 300, 325, 350, 375, 400, 425, 450, 475, 500, 525, 550, 575, 600, 625, 650, 675, 700, 725, 750, 775, 800, 825, 850, 875, 900, 925, 950, 975 or 1000% relative to the complex lipid content when compared to the GCLC, and useful ranges may be selected between any of these values (for example, about 10 to about 1000, about 20 to about 1000, about 30 to about 1000, about 40 to about 1000, about 50 to about 1000, about 100 to about 1000, about 200 to about 1000, about 300 to about 1000, about 400 to about 1000, about 500 to about 1000, about 600 to about 1000, about 700 to about 1000, about 800 to about 1000, about 900 to about 1000, about 20 to about 300, about 30 to about 300, about 40 to about 300, about 50 to about 300, about 60 to about 300, about 70 to about 300, about 80 to about 300, about 90 to about 300, about 100 to about 300, about 150 to about 300, about 200 to about 300, about 250 to about 300, about 20 to about 600, about 30 to about 600, about 40 to about 600, about 50 to about 600, about 60 to about 600, about 70 to about 600, about 80 to about 600, about 90 to about 600, about 100 to about 600, about 200 to about 600, about 300 to about 600, about 400 to about 600 and about 500 to about 600% relative to the total lipid content when compared to the GCLC).

In another embodiment, the ganglioside-enriched extract may be enriched in one or more gangliosides relative to the amount of phospholipids when compared to the GCLC. In one embodiment the ganglioside-enriched extract is enriched by at least 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 125, 150, 175, 200, 225, 250, 275, 300, 325, 350, 375, 400, 425, 450, 475, 500, 525, 550, 575, 600, 625, 650, 675, 700, 725, 750, 775, 800, 825, 850, 875, 900, 925, 950, 975 or 1000% relative to the phospholipid content when compared to the GCLC, and useful ranges may be selected between any of these values (for example, about 10 to about 1000, about 20 to about 1000, about 30 to about 1000, about 40 to about 1000, about 50 to about 1000, about 100 to about 1000, about 200 to about 1000, about 300 to about 1000, about 400 to about 1000, about 500 to about 1000, about 600 to about 1000, about 700 to about 1000, about 800 to about 1000, about 900 to about 1000, about 20 to about 300, about 30 to about 300, about 40 to about 300, about 50 to about 300, about 60 to about 300, about 70 to about 300, about 80 to about 300, about 90 to about 300, about 100 to about 300, about 150 to about 300, about 200 to about 300, about 250 to about 300, about 20 to about 600, about 30 to about 600, about 40 to about 600, about 50 to about 600, about 60 to about 600, about 70 to about 600, about 80 to about 600, about 90 to about 600, about 100 to about 600, about 200 to about 600, about 300 to about 600, about 400 to about 600 and about 500 to about 600% relative to the phospholipid content when compared to the GCLC).

In another embodiment, the ganglioside-enriched extract may be enriched in one or more gangliosides relative to the amount of one or more other complex lipids when compared to the GCLC.

In one embodiment the ganglioside-enriched extract is enriched by at least 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 125, 150, 175, 200, 225, 250, 275, 300, 325, 350, 375, 400, 425, 450, 475, 500, 525, 550, 575, 600, 625, 650, 675, 700, 725, 750, 775, 800, 825, 850, 875, 900, 925, 950, 975 or 1000% relative to one or more other complex lipids when compared to the GCLC, and useful ranges may be selected between any of these values (for example, about 10 to about 1000, about 20 to about 1000, about 30 to about 1000, about 40 to about 1000, about 50 to about 1000, about 100 to about 1000, about 200 to about 1000, about 300 to about 1000, about 400 to about 1000, about 500 to about 1000, about 600 to about 1000, about 700 to about 1000, about 800 to about 1000, about 900 to about 1000, about 20 to about 300, about 30 to about 300, about 40 to about 300, about 50 to about 300, about 60 to about 300, about 70 to about 300, about 80 to about 300, about 90 to about 300, about 100 to about 300, about 150 to about 300, about 200 to about 300, about 250 to about 300, about 20 to about 600, about 30 to about 600, about 40 to about 600, about 50 to about 600, about 60 to about 600, about 70 to about 600, about 80 to about 600, about 90 to about 600, about 100 to about 600, about 200 to about 600, about 300 to about 600, about 400 to about 600 and about 500 to about 600% relative to one or more complex lipids when compared to the GCLC).

10. Uses of the ganglioside-enriched extract or gangliosides produced by a process of the invention

The one or more gangliosides and ganglioside-enriched extracts of the invention can be incorporated into food and beverages, including any food or beverage composition that is able to carry lipid material

In one embodiment, the ganglioside-enriched extract and/or gangliosides can be incorporated into infant formulas, or maternal formulas.

The processes of the invention provide a convenient means of concentrating gangliosides present in a composition. In particular, the concentration of gangliosides relative to other complex lipids can be increased. The resulting ganglioside-enriched extract may be further purified to provide a highly pure source of gangliosides that may find application in the food, research and pharmaceutical industries. For example, phospholipid impurities may be removed using saponification with NaOH, followed by acidification with HCl. The resulting fatty acids are extracted with hexane and reversed-phase SPE used to obtain a highly pure sample of gangliosides, for example 96% gangliosides.

The invention consists in the foregoing and also envisages constructions of which the following gives examples only.

EXAMPLES

Materials and Methods

All solvents and reagents were of analytical or higher grade. For alcohol - water partitioning neat alcohols were used: 1-butanol (HPLC grade, Scharlau, Spain), 1-hexanol (99% by GLC, BDH, England), 1-decanol (specially pure, BDH, England). Cartridges for solid phase extraction (SPE) were from Phenomenex (USA). The following standards were used: N-acetylneuraminic acid (NANA, Sigma, USA), L-lysine (Sigma, USA), ganglioside GM3 (Matreya, USA), ganglioside GD3 (IRL), mixed gangliosides from bovine brain (Matreya, USA), glucosylceramide (IRL), lactosylceramide (IRL), lactose (BDH, England). TLC aluminium sheets were from Merck (Germany). Filter paper was 595 from Schleicher & Schuell GmbH (Germany).

Determination of phosphorus content Total phosphorus was determined in the phases after phase separation as well as in the starting material (crude lipid extracts in die cases of sea urchin roe) using the standard procedure described by Vaskovsky et al. (1975). Aliquots of the solutions corresponding to about 100 μg of solids were transferred into Pyrex test-tubes (7 ml). The solvents were evaporated to dryness under vacuum in the Speed Vac® Plus evaporator SC210A (Savant Instruments, Inc.). The samples were digested with 72% perchloric acid (0.2 ml) by heating in a sand badi at 180-200° C for 20 min. After cooling, working reagent I (4.8 ml) was added into each test-tube. Working reagent I was prepared by addition of 1 N sulphuric acid (48 ml) to the molybdate stock reagent (4 ml) and adjusting the volume to 100 ml with water. The samples were mixed dioroughly with a Vortex mixer and heated in a boiling water-bath for 15 min. After cooling, the absorbance was measured at 815 nm against a blank using a UV- Vis spectrometer (Unicam Helios Beta, Spectronic Unicam, UK). The analysis of each sample was performed in triplicate.

The phosphorus content was calculated from a linear calibration curve created using sodium dihydrophosphate as the standard. The content of phospholipids in the samples was calculated using an average phospholipid molecular weight of 750.0.

Determination of sialic acid content

Total sialic acid was determined in the phases after phase separation as well as in each starting material using a modified version of the procedure described by Zanetta et al. (1999) Aliquots of the solutions corresponding to about 2 mg of solids were transferred into Pyrex test-tubes (7 ml) with Teflon-lined screw caps. A solution of lysine (the internal standard) was added into each test-tube in an amount of 50 μg. Control samples were run in parallel containing lysine (50 μg); lysine (50 μg) and N-acetykieuraminic acid (10 μg); and lysine (50 μg) and N-acetylneuraminic acid (100 μg). The solvents were evaporated to dryness under vacuum using a Speed Vac® Plus evaporator. Benzene (100 μl), methanol (240 μl) and the methanolysis reagent (160 μl) were added, and the closed test-tubes were incubated at 80°C for 20 h. The methanolysis reagent was prepared by the dropwise addition of distilled acetyl chloride to cold methanol to a final concentration of 12.5% (by volume) (Lepage and Roy, 1986). After methanolysis, the samples were evaporated to dryness under a light stream of argon, and anhydrous acetonitrile (0.5 ml) and heptafluorobutyric anhydride (60 μl) added. The test-tubes were preincubated at 50°C for about 15 min and then heated at 100°C for 30 min in a sand bath. The samples were left in the acylation reagent until GLC analysis. Before analysis the samples were evaporated under a gentle stream of argon and dissolved in 25 μl of benzene and 25 μl of acetonitrile.

For example 1, GLC analysis was performed on a Hewlett Packard 5890 Series II gas chromatograph equipped with a 30 m capillary column (Solgel IMS; 0.25 μm film phase; SGE International Pty Ltd., Australia). Injector and flame ionization detector temperatures were 280°C and the temperature program was 1.2°C/min between 100°C and 14O⁰C, followed by 4°C/min from 14O⁰C to 260⁰C then maintaining this temperature for 10 min. The carrier gas (helium) pressure was 100 kPa. The chromatograph worked under HP GC ChemStation Rev.A.06.03[509] software. The sialic acid content was calculated using a calibration curve created using N- acetylneuraminic acid (NANA, Type VI, from Escherichia colt, Prod. # A-2388, Sigma, USA) as the standard. L-Lysine (50 μg) was added to each calibration mixture as an internal standard. Ratio of peak area of NANA to one of lysine was multiplied by 50 and the obtained value was plotted against amounts of NANA used for calibration. The calibration curve was not linear so second- order polynomial models were adopted to fit the experimental data. The polynomial models were dependent on the NANA range. For the range 5-50 μg of NANA the model was y = 0.0044x² + 0.0768x (R²=0.9994); for the range 50-100 μg of NANA the model was y = 0.0033x² + 0.1207x (K²- 0.9995); for the range 100-500 μg of NANA the model was y = 0.0003x² + 0.3969x (R^2=:0.9986), where 'y' is the ratio of peak area of sialic acid to one of lysine (50 μg) multiplied by 50 and 'x' is the amount of sialic acid in μg. For examples 2-4 GLC analysis was performed on an Agilent Technologies 6890N Network GC system equipped with a 30 m capillary column (Solgel IMS; 025 μm film phase; SGE International Pty Ltd., Australia). Injector and flame ionization detector temperatures were 280⁰C and the temperature program was 1.2°C/min between 100 and 14O⁰C, followed by 4°C/min from 140 to 260⁰C then maintaining this temperature for 10 min. The carrier gas (hydrogen) pressure was 8 psi with flow rate of 1.0 ml/min. For the detector hydrogen flow rate was 40.0 ml/min, air flow rate was 450 ml/min, and makeup flow rate (N₂) was 20.0 ml/min. The chromatograph worked under Agilent Technologies Enhanced ChemStation software. The sialic acid content was calculated as described above. For the range 5-100 μg of NANA the model was y=0.002x² + 0.3127x (R²=0.996), where y is the ratio of the peak area of sialic acid to one of lysine (50 μg) multiplied and V is the amount of sialic acid in μg-

Determination of ganglioside content from sialic add content Gangliosides are characterised by the presence of sialic acid units. Gangliosides may be mono, di, tri or tetrasialogangliosides. The sialic acid content of a material can be used as an estimate of the amount of gangliosides present. Non-lipid bound sialic acid (non-LBSA) is water soluble, and therefore partitions to the water fraction in the processes of the invention. The lipid bound sialic acid (LBSA) will be found in both the aqueous and organic fractions.

In determining the degree of enrichment of gangliosides, one must consider several factors that may affect the results. These are discussed below.

Not all sialic acid is associated with gangliosides. Sialic acid can be present as lipid bound sialic acid (LBSA) or non-LBSA. All LBSA is generally ganglioside-associated. Non— LBSA includes free sialic acid and sialic acid bound to non-lipid components such as carbohydrates. The presence of non-LBSA must be accounted for.

The sialic acid content of the feed streams shown in Table 1 represents the total sialic acid present in the material. For Feed Stream E, LBSA makes up 76.0% of this, as determined in Example 6. Using this information the amount of LBSA in an aqueous fraction can be determined.

LBSA can be converted to ganglioside content by estimating the average mol sialic acid per mol of ganglioside in a given material and using the relative molecular weights of sialic acid and gangliosides to produce a conversion factor for each GCLC.

The feed streams used in the processes of the invention may differ significantly in lipid composition because of (a) their biological origin (for example, sea urchin roe, bovine brain, bovine milk) or (b) the method used for lipid extraction (for example, Folch extraction for sea urchin roe and supercritical fluid extraction for bovine brain powder and bovine dairy powder).

The gangliosides of the feed materials may differ in their structure and as a result, in their solubility in organic solvents and aqueous solutions. Gangliosides of sea urchin roe are monosialic or disialic containing one or two glucose units (Ledeen & Yu 1982). Some of sea urchin gangliosides can be sulfated by a sialic acid residue (Prokazova et al. 1981; Kubo et al. 1990). Bovine milk gangliosides include predominately disialogangliosides (GD3 and its O- acetykted forms), monosialogangliosides (GM3 and a ganglioside with a branching structure), and some trisialogangliosides (GT3 and its O-acetylated forms) (Takamizawa et al. 1986; Ren et al. 1992). The carbohydrate parts of bovine milk gangliosides, in addition to sialic acid residues, contain 2 units of glucose and galactose (GM3, GD3 and GT3) or 6 units of glucose, galactose and N-acetylglucosamine (the ganglioside with a branching structure). Bovine brain gangliosides are the most complex in composition and consist of 1-4 residues of sialic acids and 4 residues of neutral monosaccharides (glucose, galactose and N-acetylgalactosamine) (Ledeen & Yu 1982).

Depending on the ganglioside make up of the feed streams, the amount of ganglioside present in a material (% w/w) is about 2.5-4 x the amount of sialic acid present in % w/w.

EXAMPLE 1

A Preparation of GCLCs from different sources of gangliosides

Five separate GCLCs were prepared as feed streams for the processes of the invention as set out below.

Feed stream Λ (sea urchin roe not washed with water)

Roe of sea urchin Evechinus chloroticus was purchased in the local market. The content of a pack was freeze-dried without washing. Crude lipids were extracted from the freeze-dried roe (32.8g) using the Folch lipid extraction process (Folch et al., 1957) as described below.

The freeze-dried material was ground in a Waring blender for 2 min at high speed. To the ground material (32.19 g), water, 16.1 ml (0.5 v/w), methanol, 215 ml (6.7 v/w), and chloroform, 429 ml (13.3 v/w), were added. The lipids were extracted under continuous stirring on the magnetic stirrer overnight at room temperature. The extract was filtered through a sintered glass funnel with filter paper (595, Ref. No. 10311612, Schleicher & Schuell GmbH, Germany) on the top. The total volume of filtrate recovered was 698 ml. A portion of filtrate (498 ml) was evaporated on a rotary evaporator to dryness yielding 7.341 g of GCLC. The extract was high in salts. The phospholipid content of the extract was 17.0% and sialic acid content was 1.6%.

Feed stream B (sea urchin roe washed with water)

Crude lipids were extracted as described for Feed stream A except that the roe was washed with cooled distilled water (0.5 L) three times before being freeze-dried to give 51.6g of material. The resulting GCLC obtained after Folch lipid extraction contained 19.4 % phospholipids, 1.7% sialic acid and significantly less salts than the GCLC produced from unwashed sea urchin roe (Feed stream A).

Feed stream C (bovine brain lipid extract)

This GCLC was produced by near-critical dimethyl ether extraction of a beef brain powder. The

GCLC contained 43% phospholipids and 0.7% sialic acid.

Feed stream D (total dairy lipid extract)

Feed stream D was produced by near-critical dimethyl ether extraction of a dairy powder enriched in milk fat globular membrane proteins, and reduced in lactose. The GCLC comprised over 90% total lipid by mass, approximately 40% complex lipid by mass, 34% phospholipids and 1.6% sialic acid.

Feed stream E (polar dairy lipid extract)

Feed stream E was produced by supercritical CO₂ extraction of feed stream D. The GCLC comprised approximately 80% complex lipid by mass, 64% phospholipids and 2.1% sialic acid.

The content of phospholipids (by % phosphorus) and gangliosides (by % sialic acid) in the five feed streams is summarized in Table 1 below. Phospholipid and sialic acid content (w/w) were calculated in accordance with the methods set out in the Materials and Methods section.

Table 1 Phospholipid (PL) and ganglioside (as sialic acid) content of Feed stteams A-E

B Partitioning of GCLC (Feed streams A-E)

The components of the GCLCs were partitioned into aqueous and organic phases using the processes of the invention.

Each feed stream was dissolved or dispersed in butanol (50 parts, v/w) under incubation at 50°C for about 1-1.5 h with sporadic shaking. Water was added (50 parts, v/w) and the mixture was shaken thoroughly before incubating at 50°C for 0.5 h. The separation into aqueous and organic phases was achieved at room temperature.

Details of partitioning for each feed stream are provided below.

Feed stream A. The crude lipid extract was dissolved in 32 ml chloroform — methanol — water (2:1 :0.2, v/v/v) under heating at 6O⁰C for 0.5 h. An aliquot of the solution, 13.4 ml was transferred into a 250 ml Schott bottle. The solvents were evaporated under a stream of argon, then under vacuum to dryness. To the recovered crude lipids (2.563 g) butanol was added, 128 ml, and the mixture was incubated at 50⁰C for 1.5 h with sporadic shaking. The solids only partially dissolved in butanol. Water (128 ml) was added to the mixture. The final concentration of crude lipids in butanol — water (1:1, v/v) mixture was 1% (w/v). The mixture was shaken thoroughly and incubated at 50⁰C for 0.5 h, resulting in the complete dissolution of solids. The mixture was transferred into a separating funnel for phase partitioning.

After phase partitioning overnight, the phases were separated. The upper butanol phase, enriched in lipids, was 143 ml in volume with a lipid concentration of 15.46 mg/ml that corresponded to 2210.8 mg or 86.3% of the initial crude lipids present before partitioning. The lower water phase enriched in water-soluble compounds including gangliosides, was 114 ml in volume with a lipid concentration of 2.78 mg/ml that corresponded to 316.9 mg or 12.4% of the initial crude lipids present before partitioning.

Feed stream B

The lipid extract of feed stream B was partitioned in accordance with the process outlined for

Feed stream A. Feed stream C

2.5 g of Feed stream C was placed in a 250 ml Schott bottle. Butanol, 125 ml, was added and the mixture was incubated at 50⁰C for 1 h. Complete dissolution of the lipid extract was achieved. Water (125 ml) was added, the mixture was shaken thoroughly and incubated at 50⁰C for 0.5 h. The mixture was transferred into a separating funnel for phase partitioning.

After phase partitioning the phases were separated. The upper butanol phase, enriched in lipids, was 143 ml in volume with a lipid concentration of 16.38 mg/ml that corresponded to 2342.3 mg or 93.7% of the initial crude lipids present before partitioning. The lower water phase enriched in water-soluble compounds including gangliosides, was 110 ml in volume with a lipid concentration of 1.6 mg/ml that corresponded to 176 mg or 7.7% of the initial crude lipids present before partitioning.

Feed stream D

The lipid extract of Feed stream D was partitioned in accordance with the process outlined for Feed stream A.

Feed stream E 2.636 g of Feed stream E, was placed in a 250 ml Schott bottle. Butanol, 132 ml, was added and the mixture was incubated at 5O⁰C for 1.8 h. with sporadic shaking. Near-complete dissolution of the lipid extract was achieved. Water (132 ml) was added, the mixture was shaken thoroughly and incubated at 50⁰C for 0.5 h with sporadic shaking. The mixture was transferred into a separating funnel for phase partitioning.

After phase partitioning the phases were separated. The upper butanol phase, enriched in lipids, was 150 ml in volume with a lipid concentration of 15.86 mg/ml that corresponded to 2379.0 mg or 90.3 % of the initial crude lipids present before partitioning. The lower water phase enriched in water-soluble compounds including gangliosides, was 116 ml in volume with a lipid concentration of 1.78 mg/ml that corresponded to 206.5 mg or 7.8% of the initial crude lipids present before partitioning.

Results and discussion

The ganglioside and phospholipid content of the two phases after butanol/ water partitioning are presented in the Tables 2 and 3 and Figure 1. Table 2 shows the initial content of gangEosides (expressed as sialic acid content) and phospholipids as well as their distribution between the butanol and aqueous phases after separation.

Table 2 Distribution of gangliosides (as sialic acid) and phospholipids between phases after butanol/ water partitioning for Feed streams A-E

Abbreviations: B — butanol phase, n.d. — not detected, PL — phospholipids, SA — sialic acid, W — water (aqueous) phase.

The proportion of sialic acid distributed in the aqueous phase during butanol/water partitioning varied widely from 7 to 62% of total sialic acid recovered (Table 2). For Feed stream C, sialic acid was detected only in the water phase and that portion corresponded to 52% of sialic acid detected in the starting feed material. Different factors could be responsible for the rate of ganglioside distribution into the aqueous phase. Salts present in the starting material are likely to have a pronounced effect on ganglioside distribution between the water and butanol phases, as the experiment with sea urchin roe demonstrated. Unwashed sea urchin roe contained a significant amount of salts because it had been prepared in salt water. Sea urchin roe that had been washed with water prior to extraction (Feed stream B) gave a significantly higher yield of gangliosides in the water phase compared to sea urchin roe that had not been washed (Feed stream A).

Another factor which may affect the degree of partitioning is die nature of gangliosides present. More polar gangliosides are more readily partitioned into the aqueous phase. Sea urchin roe is rich in monosialic gangliosides with chromatographic behaviour similar to the ganglioside GM3. These gangliosides partition predominanύy in the organic phase. Beef brain and milk contain principally more polar gangliosides which partition mainly, or completely, into die water phase. It seems that the initial content of gangliosides in the starting material has litde effect on the proportion of gangliosides that partition into the water phase. The Feed streams D and E produced from milk, containing presumably the same gangliosides but in different quantities, differed litde in the proportion of sialic acid in the water phases.

Phospholipids are considered to be impurities diat are difficult to remove from a final preparation of gangliosides without using special techniques. Phospholipids were concentrated predominantly in die butanol phase after butanol/ water partitioning. Their content in the water phase varied from 1.5 to 12.7% of the total phospholipids (Table 2).

The yield of solids and the content of ganglioside (as sialic acid) and phospholipid in die solids after butanol/water partitioning are presented in die Table 3.

Table 3 Yield of gangliosides (as sialic acid) and phospholipids in phases aftet butanol/ water partitioning for Feed streams A-E

* Solids in the starting material used for partitioning. Abbreviations: B — butanol phase, n.d. - not detected, PL — phospholipids, SA — sialic acid, W — water phase.

The yield of solids in the water phase was highest for Feed streams A and B which had the lowest content of gangliosides and phospholipids. However the molar ratio of sialic acid to phospholipids was especially high for Feed stream B (Figure 1). Butanol/ water partitioning of the feed streams D and E gave a good yield of solids with a high content of gangliosides (Table 3) and a molar ratio of sialic acid to phospholipids close to or above 1.

The content of sialic acid in the water phase of these two feed streams probably reflects its content in the starting material (Tables 1 and 3). Feed stream E contained more gangliosides and yielded a higher content of these compounds in the water phase compared to Feed stream D. Feed stream C yielded a high proportion of phospholipids in the water phase compared to ganglioside content (Table 2, Figure 1), possibly due to the low content of gangliosides in the starting material (Table 1).

EXAMPLE 2 - Preparation of highly concentrated ganglioside-eήriched extract

A dairy lipid extract (DLE) produced by ethanol extraction of low lactose beta-serum powder was used in the process of the invention. The DLE comprised approximately 2-3% w/w ganglioside.

Butanol (5 L) was added to 100 g of DLE and left for 2-3 h at 50⁰C with occasional shaking. Water (5 L) was added, shaken vigorously to dissolve the residue and left for 2 h at 50⁰C. The mixture was allowed to cool at room temperature until complete phase separation was achieved. The upper organic phase was removed and ammonium acetate was added to the lower aqueous phase to a final concentration of up to 0.2 M. Then methanol was added to a final concentration of 3% (by v/v) and butanol - to a final concentration of 4-5.5% (by v/v). After the phase separation the organic upper phase was discarded and the resulting lower aqueous phase was further purified by reversed-phase solid phase extraction (SPE) to obtain a composition comprising 95%+ w/w gangliosides. In one variant, glycerolipids in the water phase were hydrolysed under alkaline conditions and gangliosides were recovered by reversed-phase SPE to obtain a composition comprising 96% w/w gangliosides. EXAMPLE 3 Effect of butanol to water ratio on partitioning

The optimal ratio of butanol to water for ganglioside and phospholipid partitioning was investigated. Feed stream E (prepared as per Example 1) was placed in 100 ml graduated cylinders with stoppers, approximately 1.01 g per cylinder. Butanol, in varying volumes from 30 to 70 ml per cylinder, was added and the mixture was incubated at 5O⁰C for 2.5 hrs with sporadic shaking. Water was added to adjust the total volume to 100 ml. The mixtures were shaken thoroughly and stored at room temperature for phase partitioning. Once separated, the volumes of the phases were measured and aliquots were taken for the determination of solid concentration and analysis of phosphorus and sialic acid content, in accordance with procedures provided in the Materials and Methods section.

Increasing the butanol portion from 30 to 70 volume parts resulted in quicker phase partitioning and a thinner interphase. These effects are related to the content of solids in the water phases, as calculated on the basis of the solids concentration and measured volumes of the phases (Table 5, column 4). The larger the volume of the butanol phase, the less solids were present in the water phase. However, the decrease in the solids content in the water phase as the volume of butanol was increased was more significant going from 30% to 40% v/v of butanol, where the difference in content of solids was 3%, compared to going from 50% to 70% v/v butanol where the difference was only 2%.

Varying the butanol to water ratio while keeping the total concentration of the starting material in the mixture at 1% had a pronounced effect on the recovery of gangliosides in the water phase, phospholipid content in the water phase, and as a result, on the molar ratio of sialic acid to phospholipids in those phases (Tables 4 and 5, Figure 2). The higher the water content in the mixture, the more complete was the recovery of gangliosides into the water phase (Table 4, columns 3 and 4). The same was true for phospholipids, i.e. less butanol resulted in more phospholipids being distributed into the water phase (Table 4, column 5).

Table 4 Effect of butanol to water ratio on the distribution of gangliosides (as sialic acid) and phospholipids between phases for Feed stream E

Butanol to SA content, % of total SA content, % of PL content, % of water ratio Phase SA recovered in two SA in starting total PL recovered in

(v/v) phases material two phases

Abbreviations: B — butanol phase, PL — phospholipids, SA - sialic acid, W - water phase.

The content of gangliosides (as sialic acid) in the solids of the water phases varied, with the highest concentrations being observed when the butanol content was 40-60% w/w (the highest concentration was at 40% butanol) (Table 5, column 5). Phospholipid content in the solids of the water phases decreased with increasing butanol content in the mixture (Table 5, column 6). However, this decrease was smaller at a butanol concentration above 50% compared to 40% butanol and, especially, 30% butanol. As a result, the molar ratios of sialic acid to phospholipids varied less at butanol concentrations above 50%, compared to concentrations less than 50% (Figure 2). Nevertheless, the recovery of gangliosides in the water phases was significantly higher at 50% butanol as compared with 60% butanol (Table 4).

Therefore, the optimal ratio of butanol to water may vary depending on the objective of the process. The optimal ratio for ganglioside enrichment is from 40:60 to about 60:40 (v/v). The highest recovery of gangliosides is at 40:60 v/v. To achieve a high ratio of gangliosides to phospholipids while still obtaining a high recovery of gangliosides, 50:50 v/v is optimal.

Table 5 Effect of the butanol to water ratio on the yield of solids, gangliosides (as sialic acid) and phospholipids in the phases after butanol/water partitioning for Feed stream E

* Solids in the starting material used for partitioning. Abbreviations: B — butanol phase, PL — phospholipids, SA - sialic acid, W - water phase.

EXAMPLE 4 Effect of feed stream concentration in the butanol/ water mixture

The optimal concentration of the GCLC in the butanol/water mixture was investigated.

Feed stream E was placed in 100 ml Schott bottles in amounts of 1.01, 2.00, 3.01, 5.00, and 10.01 g per bottle (1%, 2%, 3%, 5% and 10% samples, respectively), followed by the addition of butanol in volumes of 50.0, 49.0, 48.5, 47.5, and 45.0 ml, respectively. The mixtures were incubated at 50°C for 2.75 h with sporadic shaking. Water was added equal to the volumes of butanol added, so that the total volume was approximately 100 ml for each mixture. The mixtures were shaken thoroughly and stored at room temperature for phase partitioning. The volume of each phase was measured and aliquots taken for the determination of solids concentration and analysis of phosphorus and sialic acid content in accordance with the Methods and Materials section.

For the investigation of flakes located in the interphase and lower aqueous phase after butanol/water partitioning, the lower fractions of the 1%, 3% and 5% samples were treated as follows. After the collection of aliquots for analysis, the volume of the rest of the aqueous fraction was measured and filtered through a coarse sintered glass funnel with filter paper (595, Ref. No. 10311612, Schleicher & Schuell GmbH, Germany) on the top. The solids on the filter were washed with 5 ml of water saturated with butanol and eluted with methanol, 10 ml, and chloroform - methanol (2:1, v/v), 20 ml for the 1% and 3% samples, and 30 ml for the 5% sample. The fractions eluted with methanol and chloroform - methanol were combined and evaporated on a rotary evaporator to dryness. Dispersion of Feed stream E in butanol at 50⁰C required more time as the concentration of the feed stream increased. Feed stream concentrations of 1% and 2% were completely dispersed ■within 1.25-1.75 h at 50⁰C with sporadic shaking, while higher concentrations of feed stream (3% and more) required 2.5-2.75 h under the same conditions. However, the time required for the feed stream to disperse in butanol can be significantly decreased by incubating with continuous agitation as described in Example 5.

Increasing the feed stream concentration in the butanol/water mixture resulted in the accelerated formation of residue and flakes that located chiefly in the interphase and impeded phase separation. This process is significantly influenced by temperature. Preliminary incubation of the mixture at 50⁰C for approximately 0.5-1 h with subsequent short-term cooling to room temperature can improve phase separation, making the butanol phase mostly clear. However, incubation did not prevent interphase formation or effect complete flake dissolution in the water phase. The mixture with a GCLC of 10% was the most complex, with the upper organic phase being turbid and the lower aqueous phase being heterogeneous in consistency.

The accumulation of flakes in the water phase may partly reflect the distribution of solids between phases. Increasing the feed stream concentration decreases the ratio of solids in the butanol phase to solids in the water phase. These changes are more pronounced when the feed stream concentration increases from 1% to 3% and especially up to 10% (Table 8, column 4).

The limited solubility of the solids that were distributed into the water phase with increasing feed stream concentration was demonstrated in an experiment in which the water phases were filtered through a coarse sintered glass funnel covered with filter paper. The yields of solids recovered from the filter are presented in Table 6.

Table 6 Recovery of solids by filtration of the water phase after butanol/ water partitioning at different feed concentrations for Feed stream E

The weight of solids recovered by filtration of the water phase increases with increasing feed stream concentration in the water/butanol mixture. Although this increase is not directly proportional to the concentration of solids in the phase, the proportion of solids retained on the filter could be significant: up to 18% of the solids distributed in the water phase.

With an increase in feed stream concentration, the proportion of total sialic acid distributed into the water phase increased from 77% at a feed stream concentration of 1% w/w, up to 93% at feed stream concentrations of 3-5% (Table 7, column 3). Further increase in the feed stream concentration up to 10% w/w resulted in lower recovery of gangliosides in the water phase so that only 87% of the sialic acid partitioned in the water phase. Although recovery of gangliosides into the water phase is higher at higher feed stream concentrations, the percentage content of gangliosides (measured as sialic acid content) in the solids of the water phase decreased with increasing feed stream concentration. The decrease is more significant when going from 1% to 2% w/w feed stream concentration and, especially, when going from 5% to 10% w/w of feed stream concentration (Table 8, column 5). There was only slight decrease in ganglioside concentration with an increase of the feed stream concentration from 2% to 5% w/w.

Table 7 Effect of feed concentration in butanol/water mixture on distribution of gangliosides (as sialic acid) and phospholipids between phases for Feed stream E

Abbreviations: B — butanol phase, PL - phospholipids, SA - sialic acid, W — water phase. The GCLC concentration has a pronounced effect on the phospholipid distribution between the two phases. Increasing the feed stream concentration in the butanol/water mixture led to an increased proportion of phospholipids being distributed into the water phase (Table 7, column

5).

However, the percentage content of phospholipids in the water phase solids was around 30% (w/w) at feed stream concentrations of 3% w/w, and increased significantly at feed stream concentrations of 5% and, especially, 10% w/w (Table 8, column 6).

Table 8 Effect of feed concentration in butanol/water mixture on the yield of solids, gangliosides (as sialic acid) and phospholipids in the phases after butanol/water partitioning for Feed stream E

* Percentage of sum of solids recovered in both phases. Abbreviations: B — butanol phase, PL — phospholipids, SA — sialic acid, W - water phase.

The molar ratio of sialic acid to phospholipids decreased with an increase in the feed stream concentration (Figure 3). The decrease was more pronounced at feed concentrations of 5% and 10% w/w, reflecting an increase in phospholipid content in the water phases. For one pot ganglioside partitioning the feed concentration in the butanol/water mixture can be varied in the range of 1-3% with the highest molar ratio of gangliosides to phospholipids being achieved at 1% w/w. Higher feed stream concentrations (3-5%) result in a more complete recovery of gangliosides in the water phase and can be used for repeated partitioning of the water phase with butanol to remove excess phospholipids and other non-gangliosides. The highest feed stream concentration tested (10%) results in abundant flakes, residue formation, improper phase separation, and a high proportion of phospholipids being distributed into the water phase.

EXAMPLE 5 Effect of alcohol and temperature

A Incubation at 50⁰C - emulsion formation

Feed stream E was placed in 100 ml Schott bottles in amounts of approximately 1 g per bottle followed by the addition of an alcohol (1 -butanol, 1-hexanol or 1 -decanol) in volumes of 50.0 ml. The bottles were placed in jackets heated by water and incubated at 50°C under continuous agitation with magnetic stirrers. Complete dissolution of material -was achieved within 45 min. Water, 50 ml, was added into each bottle and the mixtures were incubated at 50⁰C under continuous agitation for another 30 min. During this incubation period emulsions were formed, so agitation was stopped to allow phase partitioning. The samples -were incubated at 50⁰C for 4 h. While the mixture with butanol completed phase partitioning within 4 h, the mixtures with hexanol and decanol formed very stable emulsions in the upper phases. To achieve phase partitioning in those samples the mixtures were alternately cooled to room temperature and warmed to 50⁰C for several hours. This treatment led to complete phase partitioning of the mixture with hexanol, but incomplete phase partitioning of the mixture with decanol. Finally, the mixtures were centrifuged at 2500 rpm for 10 min. This resulted in phase partitioning of the mixture with decanol, though an interphase of about 5ml still remained. All samples were treated in the same manner. The volumes of the phases were measured and aKquots were taken for the determination of solids concentration and analysis of phosphorus and sialic acid content. For determination of the solids concentration, the aliquots of the water fractions and the butanol fractions were evaporated in a SpeedVac concentrator to constant weight. The aliquot of the hexanol fraction was mixed with an excess of water and butanol and evaporated on a rotary evaporator under vacuum at 73°C. The concentration of solids and the sialic acid content were not determined for the aliquot of the decanol phase, due to the very slow rate of decanol evaporation.

B Incubation at 70⁰C - no emulsion Feed stream E was placed in 100 ml Schott bottles in amounts of approximately 1 g per botde followed by the addition of an alcohol (1-butanol, 1-hexanol or 1-decanol) in volumes of 50.0 ml. The bottles were placed in jackets heated by water and incubated at 70°C under continuous agitation with magnetic stirrers for 1 h. The solids dissolved completely within 20 min. The solutions were cooled at room temperature for 1 h and water, 50 ml per botde added. The mixtures were gently shaken, by carefully inverting, to prevent emulsion formation. The botdes were placed in the jackets heated by water and incubated at 70°C under gende continuous agitation (about 100 rpm) with magnetic stirrers for 4 h. Fractions of the upper and lower phases were collected from the heated mixtures and cooled to room temperature. The volumes of phases were measured after cooling and aliquots were taken from the phases for the determination of solids concentration and analysis of phosphorus and sialic acid content. The aliquots were dried as described in Example 5A.

C Incubation at 50⁰C — no emulsion The starting material and procedure of incubation were as described in Example 5B above except that all of the incubations were carried out at 50°C.

Results and discussion

Butanol is the lowest member of the homologous series of aliphatic alcohols which forms a biphasic system with water. The higher alcohols used in this study are less miscible (1-hexanol) or immiscible (1-decanol) with water. The lipid extract (Feed stream B) was soluble in the alcohols tested under continuous agitation, at temperature of 50°C or above and at feed stream concentrations of 2%.

The differences between Examples 5A, 5B and 5C were: the temperature used; the time of phase partitioning; and the degree of the mixture agitation. The experiment described in Example 5A was performed similarly to the experiments described in foregoing examples. Namely, after the feed stream was dissolved in alcohol and water, the mixture was shaken thoroughly followed by incubation at 5O⁰C and phase partitioning at room temperature.

In Tables 9 and 10 the temperature for Example 5A is described as '50⁰C, then room T⁰C. In these experiments, the alcohol/water mixtures were allowed more time to reach equilibrium. Unlike Example 5A, the phase separations of Examples 5B and 5C were performed at either 5O⁰C or 70⁰C, giving a shorter time (4h) before fractions were taken for cooling and further analysis. Intensive agitation of the feed stream in an alcohol solution with water caused emulsion formation (Example 5A). The emulsion was more stable with increasing hydrocarbon chain length, from C₄ to C₆ and further to C₁₀. It took approximately 4 h to achieve phase separation of the butanol/water mixture at 50°C without additional treatment. However, the mixtures with hexanol and, especially, with decanol formed far more stable emulsions. Additional treatments were required to achieve phase separation, including repeated cooling and heating (for hexanol) and centrifugation (for decanol). In subsequent experiments (Examples 5B and 5C) precautions were taken during mixing to prevent emulsion formation (see above).

The nature of the alcohol had a pronounced effect on the distribution of gangliosides and phospholipids between the phases. In all three experiments the proportion of gangliosides distributed into the aqueous phase was highest in the hexanol/watet mixtures (up to 93% w/w of sialic acid in the starting material), followed by the butanol/water mixtures (up to 66% w/w of sialic acid content in the starting material, Table 9, column 5).

The decanol/water mixtures were least effective for ganglioside enrichment into the aqueous phase, with the highest recovery being 52% w/w sialic acid, obtained only after intensive agitation and prolonged partitioning (Example 5a). Examples 5B and 5C produced small proportions of gangliosides from partitioning of the decanol/water phases (Table 9, column 5).

The distribution of phospholipids into the aqueous phase followed, in general, the same trend established for the distribution of gangliosides. The proportion of phospholipids distributed into the aqueous phase was higher in the hexanol/ water mixtures than the butanol/water or decanol/water mixtures of the same experiment (Table 9, column 6). Butanol/water mixtures resulted in more phospholipids in the aqueous phase compared to decanol/water mixtures, except in Example 5A which gave a contrary result.

The ratios of sialic acid to phospholipids in the aqueous fraction were higher in the mixtures with butanol, except the decanol/water mixture incubated at 7O⁰C which had an unusually high ratio (Figure 4). However, as mentioned above, only a small portion of gangliosides, as well as phospholipids, partitioned from the decanol phase into the aqueous phase in this case (Table 9, columns 4-6). Table 9 Effect of alcohol and temperature on distribution of gangliosides (as sialic acid) and phospholipids between phases for Feed stream E

Abbreviations: B — butanol phase, D — decanol phase, H — hexanol phase, n.d. — not determined, PL — phospholipids, SA — sialic acid, W — water phase.

The yields of solids in the aqueous phases, calculated on die basis of the fraction volume and concentration of solids, mainly confirmed die pattern of phospholipid and ganglioside distribution established fot the different alcohol mixtures (Table 10, column 5). However, the yields of solids varied significandy among the examples, resulting in large variations in sialic acid and phospholipid content in the solids of the water phases (Table 10, columns 6 and 7). Those variations may reflect the heterogeneity of the Feed stream E used in this experiment, for example, in lactose content. The effect of temperature on the distribution of gangliosides and phospholipids appeared to depend on the alcohol used. For butanol and hexanol, more gangliosides distributed into the aqueous phase at 50⁰C than at 7O⁰C (Table 9). These results may be due to the alcohol fraction becoming more saturated with water at a higher temperature. Prolonged incubation of the mixtures at room temperature (Example 5A) had little effect on the recovery of gangliosides in the aqueous phase, though the proportions of gangliosides present were higher in the both cases. For decanol the best recovery of gangliosides into the aqueous phase was obtained in Example 5A where decanol was mixed with water thoroughly and the mixture was allowed to phase partition for a prolonged period of time. In the case of decanol the main factors blocking the distribution of gangliosides into aqueous phase may be viscosity or the interfacial tension/pressure of the alcohol. This proposition is supported by the higher recovery of gangliosides into aqueous phase at higher temperature (Table 9).

Table 10 Effect of alcohol and temperature on the yield of gangliosides (as sialic acid) and phospholipids in phases after alcohol/ water partitioning for Feed stream E

*Solids in the starting material used for partitioning. Abbreviations: B — butanol phase, D — decanol phase, H — hexanol phase, n.d. — not determined, PL — phospholipids, SA — sialic acid, W — water phase.

The distribution of phospholipids in the butanol/water mixtures varied little with different temperatures and treatments (50°C or 70°C, short-term or prolonged partitioning). In these cases the proportions of phospholipids recovered in the aqueous phases were in the range 3.3-4.2 % w/w of total phospholipids (Table 9, column 6). Very close values for the proportion of phospholipids distributed into the aqueous phases in the hexanol/ water mixtures at 50°C and 70°C were obtained, 6.3 and 6.5% w/w, respectively (Table 9, column 6). However, thorough mixing of the phases and prolonged partitioning at room temperature gave a higher value (10.7%) w/w. The distribution of phospholipids in the decanol/ water mixtures was similar to that observed for the distribution of gangliosides: very low proportions of phospholipids in the aqueous phases after short-term partitioning at 50°C or at 70⁰C; and significantly higher values after thoroughly mixing the decanol with water and partitioning over a prolonged period (Table 9, column 6).

It seems that the temperature of partitioning has little effect on the ratio of sialic acid to phospholipids except in the decanol/water mixture (Figure 4, incubations at 50°C and 70°C). The thoroughly mixed butanol/water mixture (Figure 4, incubation at 50°C then at room temperature), which allowed prolonged partitioning, gave a higher ratio in the water phase than mixtures partitioned for a shorter time. However, the reverse was observed for the hexanol/water and decanol/water mixtures (Figure 4).

Summarizing the results obtained in Examples 5A, 5B and 5C, the shorter-chain alcohols (butanol and hexanol) produced better recovery of gangliosides into aqueous fraction after alcohol/ water partitioning than the longer-chain alcohol (decanol). In the case of hexanol the distribution of gangliosides into the aqueous phase was even more complete than in the case of butanol. However, the completeness of ganglioside recovery was accompanied by increased contamination of the aqueous phase with phospholipids, so the ratio of sialic acid to phospholipids was higher in the aqueous phase of the butanol/water mixture than in the hexanol/water mixtures. Taking into account the better phase separation of butanol/water mixtures and significantly lower boiling point of 1-butanol (117-118⁰C versus 137.5°C for 1- pentanol and 157°C for 1-hexanol) butanol appears to be the most attractive alcohol for ganglioside recovery by partitioning with water. The butanol/ water partitioning can be performed at different temperatures so that it achieves complete dissolution of the feed material, though better results were obtained at 5O⁰C. For the distribution of gangliosides into the aqueous fraction, there does not appear to be a significant difference between mixing the butanol and water thoroughly or not and partitioning the mixtures for a short or prolonged period of time. However, better results were obtained after thorough mixing of the water and butanol and prolonged incubation.

EXAMPLE 6 Determination of lipid-bound sialic acid in Feed stream E

Iipid-bound sialic acid (LBSA) content was determined in two samples of Feed stream E after chloroform/methanol/ water partitioning and recovery of gangliosides from water fractions on reversed-phase cartridges. Taking into account the distribution of gangliosides between chloroform and water phases, the total LBSA content in a sample was calculated as the sum of sialic acid content in the chloroform phase and the gangliosides recovered from the water phase.

The chloroform/methanol/water partitioning was performed according to Folch et al. (1957), and Svennerholm & Fredman (1980). 200 mg of Feed stream E was placed into a vial with a screw cap. For Folch partitioning the solids were dissolved in 4 ml of chloroform — methanol (2:1, v/v) followed by addition of 0.8 ml of 0.1 M NaCl. The mixture was shaken and left for phase partitioning. After phase partitioning the water phase was transferred into another vial. The chloroform phase was mixed with 1.92 ml of Folch theoretical upper phase consisting of chloroform - methanol — 0.1 M NaCl (3:48:47, v/v/v). The phase partitioning was achieved by centrifugation at 3500 rpm for 5 min. The water fractions were combined and treated for gangliosides recovery on reversed-phase cartridges as described below. The chloroform fraction was filtered through a cotton wool wad into a round-bottomed evaporative flask and evaporated to dryness on a rotary evaporator. The solids were dissolved in chloroform — methanol (2:1, v/v) in a concentration of 20 mg/ml.

For Svennerholm & Fredman partitioning the sample was dissolved in 4 ml of chloroform — methanol - water (4:8:3, v/v) followed by addition of 0.7 ml of water. The mixture was shaken and left for phase partitioning. After phase partitioning the water phase was transferred into another vial. The chloroform phase was mixed with 0.4 ml of methanol and 0.27 ml of 0.01 M KCl. The phases were separated and processed as described for Folch partitioning.

Gangliosides in the water phases were recovered on the reversed-phase cartridges for solid phase extraction (Strata C18-E, 55 μm, 70 A, 500 mg/6 ml, Part. No 8B-S001-HCH) according to the procedure of Williams & McCluer (1980) with some modifications. The cartridges were washed in sequence with 10 ml of methanol, 15 ml of chloroform - methanol (2:1, v/v), and 10 ml of methanol. The procedure was repeated twice. Finally, the cartridges were preconditioned with 20 ml of methanol - water (1:1, v/v). The Water phases obtained after the Folch or Svennerholm & Fredman partitioning procedures were mixed with ammonium acetate up to a concentration of 0.2 M. After the first application of the fractions onto the cartridges water was added to the effluents to a final ratio of methanol to water 1:1 (v/v). The mixtures were applied onto the cartridges repeatedly. Salts were washed from the cartridges with 20 ml of methanol — water (1:1, v/v). Gangliosides were eluted with 10 ml methanol and 15 ml of chloroform — methanol (2:1, v/v). The combined ganglioside fractions were evaporated to dryness on a rotary evaporator and dissolved in 0.5 ml methanol — water (1:1, v/v) for further analysis.

Sialic acid content was determined in (a) the starting material after suspension of the feed in methanol — water (1:1, v/v), (b) the chloroform fractions and (c) the ganglioside fractions recovered from the aqueous phase. Determination of sialic acid content was carried out as described in the Materials and Methods section with automatic injection (1 μl) using an autosampler Agilent Technologies 7683 Series Injector. The fractions were analysed in three runs with a calibration curve of sialic acid determined for each run separately. LBSA content was calculated as a sum of sialic acid content in the chloroform fraction and the ganglioside fraction recovered from the water fraction.

LBSA content in the water fraction of Feed stream E after butanol — water partitioning was calculated on the basis of LBSA content in Feed stream E, total sialic acid content in the water phase (Tables 1 and 2), and an assumption that all non-lipid-bound sialic acid (non-LBSA) partitioned into the water phase. The percentage of LBSA and its yield in the water phase is discussed below.

Results

Sialic acid can be a component of different compounds, including gangliosides, oligosaccharides and glycoproteins. Moreover, it may also be present in the free form. In our analytical approach we measured total sialic acid that could reflect not only ganglioside content but other sialic acid- containing compounds also. Analyses of dairy lipid extracts (Feed streams D and E) by the TLC- resorcinol method (Christie 2003) showed the presence of resorcinol-positive compounds on die start position of TLC plates (not shown). The location of these compounds on die TLC plate indicates diat diey are highly polar. These polar compounds containing sialic acid might contribute to the evaluated extent of ganglioside enrichment found after butanol - water partitioning. In order to distinguish sialic acid content of gangliosides from other compounds containing sialic acid the LBSA was recovered from two Feed stream E samples (A and B). Feed stream E sample A comprised about 87% w/w lipid and no protein. Feed stream E sample B comprised about 90% w/w lipid and 10% protein. The content of total sialic acid and LBSA in the two samples of Feed Stream E is presented in Table 11.

Table 11 Content of total sialic acid and lipid-bound sialic acid (LBSA) in the two samples of Feed stream E

* Average of three analytical determination ± standard deviation.

** Average for the two procedures of partitioning ± standard deviation.

Taking into account the percentage of LBSA in sample B as equal to 76.0% we can estimate the content of LBSA in the solids of the water phase after butanol/water partitioning. For the data presented in Table 3 for the partitioning of Feed stream E there is 17.2% w/w solids of sialic acid in the aqueous phase. The concentration of LBSA in the solids of the water phase is 10.8% (w/w) and the yield of LBSA in the water phase is 53% (w/w), assuming that the sialic acid-bearing polar compounds (non-LBSA) partition totally into the water phase (see Materials and Methods). Assuming GD3 is the major ganglioside of bovine milk with a molecular weight of 1473 (containing stearic acid) the ganglioside content makes up 26% w/w of solids in the water phase. Removal of lactose and other low-molecular contaminants by any means as listed in Example 1 can further increase the ganglioside content in the solids recovered from the water phase.

TLC analyses of the other feed streams (Le. the non-dairy lipid extracts) did not reveal any highly polar compounds containing sialic acid besides gangliosides.

EXAMPLE 7 Effect of buffets with different pH

Feed stream E was placed in five 100 ml graduated cylinders with stoppers, 1.00-1.01 g per cylinder. Butanol in the volume of 50 ml per cylinder was added and the mixtures incubated at 55°C for 1.25 h. with sporadic shaking, resulting in almost complete dissolution of the solids. A range of buffers of pH 5-9 at a concentration of 0.025 M were added in the volume of 50 ml per cylinder. Weak acidic conditions (pH 5 and 6) were created using citric acid — sodium citrate buffers. Weak alkaline conditions (pH 8 and 8.9) were provided with Tris-HCl buffers. Citric acid is used as a food ingredient and Tris is used as a biological buffer and in the synthesis of pharmaceuticals. The effects of the buffers were compared with a neutral salt, ammonium acetate, at the same concentration (0.025 M) and pH 6.8.

The mixtures were shaken thoroughly and stored at room temperature for phase partitioning. The volumes of the separated phases were measured and aliquots were taken for the determination of solid concentration and analysis of phosphorus and sialic acid content.

The gangliosides present in the water phases were recovered using reversed-phase cartridges for solid phase extraction (Strata C18-E, 55 μm, 70 A, 500 mg/6 ml, Part. No 8B-S001-HCH). The cartridges were washed in sequence with 10 ml of methanol, 15 ml of chloroform — methanol (2:1, v/v), and 10 ml of methanol. The procedure was repeated twice. Finally, the cartridges were preconditioned with 20 ml of methanol - water (1:1, v/v) . A portion (1 ml) of each water fraction obtained after butanol partitioning with each buffer was mixed with 0.47 ml of 0.8 M aqueous ammonium acetate, 0.2 ml of methanol and 0.33 ml of water. The solutions were applied onto the cartridges and the break-through fractions applied onto the cartridge repeatedly. Water soluble compounds were washed from the cartridges, with 15 ml of methanol - water (1:1, v/v). Gangliosides were eluted with 10 ml of methanol and 15 ml of chloroform — methanol (2:1, v/v). The combined ganglioside fractions were evaporated to dryness under vacuum in the Speed

Vac® Plus evaporator SC210A (Savant Instruments, Inc.) and dissolved in 1.0 ml of chloroform - methanol — water (1:1:0.3, v/v/v) for further analysis by TLC and GLC.

The addition of low concentration buffers to the butanol mixtures of Feed stream E did not facilitate phase separation or significantly prevent interphase formation. The water phases of pH 6.8 or higher were opalescent. The butanol — water mixture with Tris-HCl pH 8.9 had a thicker interphase than other compositions. The water phases containing citric acid — sodium citrate buffers were clear.

Increasing the pH from 6.8 (ammonium acetate) up to 8.9 (Tris-HCl) increased the proportion of gangliosides in the aqueous phases (Table 12, columns 3 and 4), as expected. Increased ganglioside partitioning into the aqueous phase was accompanied by a small increase in phospholipid partitioning (Table 12, column 5). The phospholipid content in the water phase increased by 2% of total phospholipids from pH 6.8 to pH 8.9. The sialic acid content in the water phase increased by 7% of total sialic acid in the starting material from pH 6.8 to pH 8.9.

Table 12 Effect of pH in the butanol /water mixture on distribution of gangliosides (as sialic acid) and phospholipids between phases

Abbreviations: B — butanol phase, PL — phospholipids, SA — sialic acid, W — water phase.

Surprisingly, the use of citric acid — sodium citrate buffers increased recovery in the aqueous phase of gangliosides from the GCLC. The ganglioside content was high in the aqueous phases at bodi pH 5.0 and pH 6.0. The sialic acid content in the aqueous phase and its overall recovery was highest at pH 6.0 (Table 12, columns 3 and 4). The phospholipid content was lowest at pH 6.0 (Table 13, column 5), but comparable to the values obtained at other pH.

Recovery of gangliosides (as LBSA) on reversed-phase cartridges for solid phase extraction from the water fractions produced after butanol - water partitioning at different pH is presented in Table 13. The residual LBSA content in the butanol fractions is also shown. The latter values were obtained on the basis of sialic acid concentrations in the butanol phases and the assumption that all sialic acid in the butanol fractions is lipid bound (i.e. included in gangliosides) and that the total LBSA content in Feed stream E is 1.59% (w/w) (see Example 6, Table 11).

Table 13 Effect of pH in the butanol/ water mixture on recovery of gangliosides (as LBSA) from the aqueous fraction on reversed phase

Abbreviations: B — butanol fraction, LBSA - lipid bound sialic acid, W — water fraction.

In general, the recovery of LBSA from the aqueous fractions increased with decreasing pH . The highest yield (69%) was obtained at pH 6.0 and the lowest yield (42 %) at pH 8.0 (Table 13). The sum of LBSA in both the aqueous and butanol fractions was higher at acidic pH and lower at alkaline pH. Again we observed excess of LBSA content at pH 6.0 and, especially, at pH 5.0 and incomplete recovery of LBSA at alkaline pH with the lowest value at pH 8.0.

The results obtained in Example 7 demonstrated that the buffer used for butanol - water partitioning and its pH may have a pronounced effect on gangHoside and phospholipid partitioning between butanol and water phases. The best results were obtained with citric acid — sodium citrate buffer pH 6.0.

EXAMPLE 8 Effect of a calcium salt and chelating agents

Peed stream E was placed in 100 ml Schott bottles in amounts of 1.00-1.01 g per bottle followed by the addition of butanol in a volume of 50.0 ml per bottle. The mixtures were incubated at 5O⁰C for 1.25 h. with sporadic shaking. Water and a 0.1 M solution Of CaCl₂ or 0.002 M solution of EDTA (tetrasodium salt) were added in the proportions required to obtain the desired concentration of CaCl₂ or chelating agent and a total volume of 50.0 ml per bottle. The mixtures were shaken thoroughly and stored at room temperature for phase partitioning. The phases were separated, the volumes of the separate fractions measured, and aliquots taken for the determination of solid concentration and analysis of phosphorus and sialic acid content.

For the analysis of flocculi formed in the presence of CaCl₂, the water fraction containing 50 mM CaCl₂ and the water fraction without any additives were centrifuged at 3500 rpm (Megafuge 1.0, Heraeus Sepatech) for 35 min in total. An aliquot (1.5 ml) of each supernatant was collected for further analysis (supernatant-1) and the residual supernatants discarded. Residues in the centrifuge tubes were suspended in 20 ml of water saturated with butanol. The solutions were centrifuged repeatedly at 3500 rpm for 15 min. Aliquots (1.5 ml) of the resulting supernatants were collected (supernatant-2) and the supernatants discarded. The resulting precipitates were dispersed in 2 ml of chloroform — methanol (1:1, v/v). For TLC analysis the initial water phases and supernatant-1 fractions were spotted in the volume of 16 μl, the supernatant-2 fractions were spotted in the volume of 32 μl, and the suspensions of residues were spotted in the volume of 10 μl. After lipid separation, the gangliosides and phospholipids were detected by TLC using resorcinol (Christie, 2003) and molybdate (Vaskovsky et al., 1975), respectively.

This example demonstrates the effects of a calcium salt or chelating agent (TEDTA) on the partitioning of gangliosides between the aqueous and butanol phases in a butanol — water mixture.

The amount of EDTA added to the butanol - water mixtures in this experiment correspond to approximately one (0.5 mM) and two (0.1 mM) molar equivalents relative to the amount of LBSA in the butanol phase. Calcium chloride was added in concentrations of 1.0, 5.0 and 50.0 mM.

After butanol — water partitioning of the mixtures the aqueous phases were opalescent or slightly turbid in the absence of the calcium salt and turbid with flocculi formation in its presence. Preincubation of the mixtures at 50°C reduced the amount of the residue. The pH of the aqueous phases were close to neutral.

The addition of calcium salts affected the distribution of gangliosides between the aqueous and butanol phases and the overall recovery of gangliosides (Table 14). At low concentrations of CaCl₂ (1 and 5 mM) there was a slight increase in the ganglioside content in die butanol phase, corresponding to 2-4% of the total sialic acid.iα the starting material. Sialic acid content in the aqueous phase was significandy lower than in the butanol — water mixture without any additives. At a higher concentration of CaCl₂ (50 mM) the bulk of the sialic acid containing compounds distributed into the butanol phase.

The sum of gangliosides recovered in the butanol and aqueous phases was always lower in the presence of CaCl₂ and decreased with increasing CaCl₂ concentration. This observation indicates that gangliosides are pardy precipitated or adsorbed in the presence of CaCl₂, resulting in incomplete recovery. Flocculi formation in the water phase containing 50 mM CaCl₂ is discussed above.

The aqueous fractions after partitioning with 50 mM CaCl₂ or without any additives were centrifuged. The precipitate obtained was washed with water saturated with butanol and centrifuged as described above. Analysis of the fractions by TLC (not shown) indicated that the precipitate formed in the presence of CaCl₂ contains gangliosides and phospholipids. The precipitate is only partially soluble in water saturated with butanol (supernatant-2). Phospholipids are major constituents of the precipitate, but gangliosides are only present in significant amounts in the presence of 50 mM CaCl₂.

Table 14 Effect of calcium chloride and a chelating agent (EDTA) in the butanol/watet mixture on the distribution of gangliosides (as sialic acid) and phospholipids between phases

This precipitation is likely to be responsible for some loss of gangliosides in the aqueous phase.

The precipitate does not readily settle. Even after prolonged centrifugation (35 min in total at 3500 rpm) the supernatant was still turbid.

Addition of the chelating agent EDTA even in a small concentration (0.5 mM) significantly reduced ganglioside content in the butanol phase and increased sialic acid content in the water phase (Table 14). Using a higher concentration of EDTA (1.0 mM) did not any further improvement in the distribution of gangliosides between the phases. Unlike the citric acid — sodium citrate buffer, EDTA did not improve the yield of gangliosides in the phases. The total recovery of sialic acid was around 88% of its initial content, which was even lower than in the absence of any additives (Table 14).

Phospholipid content increased in the aqueous phase in the presence of EDTA and decreased as the concentration of CaCl₂ was raised to 50 mM, compared to the butanol — water mixture without any additives (Table 14, column 5). At the low concentrations of CaCl₂ (1-5 mM) phospholipid content was close to the control but the profile of phospholipids was slightly different (as found by TLC, not shown). The difference between the content of phospholipids in the water phases in the absence of any additives and in the presence of EDTA was around 20% — only 1.6% of total phospholipids recovered.

This example shows the effects of calcium ions and the chelator EDTA on the distribution of ganglioside between the aqueous and butanol phases. The experiment with a low concentration of EDTA indicates that gangliosides distributed in the butanol phase may be at least partly attributed to their corresponding calcium salts. Some losses of gangliosides in the butanol — water mixtures may also be caused by calcium ions.

EXAMPLE 9 Effect of different salts

The effect of different salts that can maintain definite pH and ion strength in water solutions on ganglioside and phospholipid distribution between butanol and aqueous phases were tested. The salts included sodium chloride (a water solution is neutral), sodium acetate (a water solution is alkaline), ammonium chloride (a water solution is acidic), ammonium acetate (a water solution is neutral) and, a potential chelator for calcium ions, trisodium citrate (a water solution is alkaline). Calcium acetate (a water solution is neutral-slightiy alkaline) was also tested to confirm the effect of calcium as a cation of a salt formed by a weak acid.

Feed stream E was placed in 100 ml Schott bottles in amounts of 1.00-1.01 g per bottle followed by the addition of butanol in a volume of 50.0 ml per bottle. The mixtures were incubated at 50°C for 1.25 h. with sporadic shaking. 0.1 M Solutions of the salts in water were added to the mixtures in a volume of 50.0 ml per bottle. The final concentration of the salts was 0.05 M. The mixtures were shaken thoroughly and stored at room temperature for phase partitioning. The volumes of phases were measured and aliquots were taken for determination of solid concentration and analysis of phosphorus and sialic acid content. Gangliosides (LBSA) were recovered on reversed phase as described in Example 7, except that a smaller volume of the 0.8 M ammonium acetate solution (0.4 ml) and a larger volume of water (0.4 ml) were added to the aqueous fractions (1.0 ml), taking into account the presence of the 0.05 M salt.

The salts tested in the concentration of 0.05 M provided faster phase separation than in the absence of any additives. Trisodium citrate was most effective, providing phase separation within 1 h at the room temperature. However, the salts in this concentration did not prevent completely the formation of an interphase, which was about 1 ml in volume. The aqueous fractions containing the salts generally had neutral pH. The pH of the sodium acetate (pH 7.5) and trisodium citrate (pH 7.9) containing fractions was weakly alkaline (Table 15, column 3).

Lipids of the water fractions were recovered on reversed phase and the sialic acid and phosphorus content of the fractions was measured (Table 15). In general, the total LBSA content in both the butanol and water phases was higher than previously indicated (see Example 6, Table 11). The sums of LBSA in the butanol and water phases exceeded 100% (Table 15, column 5). Phase partitioning with sodium chloride, sodium acetate and trisodium citrate resulted in better recovery of gangliosides into the water phases, as sialic acid content in both the butanol and water phase indicated. The lowest content of LBSA in the butanol phase was observed with sodium acetate and the highest content in the water phase was found in the presence of trisodium citrate.

This is in a good agreement with the expected effect of the slightly alkaline pH provided by both salts in the water fractions. Once again there was a significant excess of LBSA recovered in the water and butanol phases in the case of trisodium citrate (compare with citric acid — sodium citrate buffers, see Example T). Nevertheless, the differences of LBSA content in the butanol and water phases among these three salts were small, if the percentages of total LBSA recovered in the two phases are compared (Table 15, column 4). Phase partitioning with ammonium chloride also provided a high content of LBSA in the water phase, but sialic acid content in the butanol phase was also high with an overall excess of LBSA content in the both phases. Calcium acetate led to the highest content of sialic acid in the butanol phase, the lowest content in the water phase, and the lowest overall recovery of LBSA in both phases. The trend observed was similar to that for ganglioside distribution in the presence of calcium chloride (see Example 8). Phase partitioning with ammonium acetate resulted in the distribution of gangliosides in close proportions between the two phases. Table 15 Effect of different salts in the butanol/ water mixture on recovery of gangliosides (as LBSA) and phospholipids from the water fraction on reversed phase

Abbreviations: B - butanol phase, LBSA — lipid bound sialic acid, PL - phospholipids, W — water phase.

The content of phospholipids recovered after reversed phase solid phase extraction along with gangliosides was higher in the presence of sodium acetate and lower in the presence of trisodium citrate, ammonium acetate and, especially, calcium acetate (Table 15, column 6). The phosphorus content determined direcdy in the water phase was highest (6.1% of total phospholipids in the starting material) in the presence of calcium acetate.

The molar ratio of LBSA to phospholipids was highest in the fraction obtained in the presence of trisodium citrate, followed by the fractions containing ammonium chloride and sodium chloride. The lowest molar ratio of these compounds was obtained in the presence of calcium acetate (Table 15, column 7). Taking into account the molar ratios and recovery of gangliosides into the water phases tπsodium citrate appears to be the best salt for butanol — water partitioning of ganglioside-containing lipid material, followed by sodium chloride and ammonium chloride.

It is not the intention to limit the scope of the invention to the abovementioned examples only. As would be appreciated by a skilled person in the art, many variations are possible without departing from the scope of the invention (as set out in the accompanying claims).

INDUSTRIAL APPLICABILITY

The processes of the invention provide a means of concentrating gangliosides present in a ganglioside-containing lipid composition. The resulting gangliosides and ganglioside-enriched extracts have application in many areas of research and health and can be added to foods and cosmetics.

In addition, the processes of the invention allow separation of the gangliosides from other polar lipids that are generally difficult to separate, such as phospholipids. A ganglioside-enriched extract of the invention that is substantially free of other polar lipids can be further purified to produce a highly pure sample of ganglioside, for example, up to 96% pure. Such highly purified material may be used in the pharmaceutical industry.

REFERENCES

Christie W. W. lipid Analysis; third edition; The Oily Press, Bridgwater, England, 2003, 190-192.

Folch J., Lees M., Stanley G. H. S. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 1957, 226, 497-509.

Kubo H., Irie A., Inagaki F., Hoshi M. Gangliosides from the eggs of the sea urchin, A.nthoάdaris crassispina. }. Biochem. 1990, 108, 185-192.

Ladish S. and Gϋlard B. A solvent partition method for microscale ganglioside purification.

Analyt. Biochem. 1985, 146, 220-231. Ledeen R. W., Yu R. K. Gangliosides: structure, isolation, and analysis. Methods in Enzymology.

1982, 83, 139-191. Lepage G., Roy C. C. Direct transesterification of all classes of lipids in a one-step reaction. J.

Lipid Res. 1986, 27, 114-120.

Prokazova N. V., Mikhailov A. T., Kocharov S. L., Malchenko L. A., Zvezdina N. D., Buznikov G; Bergelson L. D. Unusual gangliosides of eggs and embryos of the sea urchin Strongylocentrotus intermedins. Structure and density-dependence of surface localization. Eur. J. Biochem. 1981, 115,

671-677.

Ren S., Scarsdale J. N., Ariga T., Zhang Y., Klein R. A., Hartmann R., Kushi Y., Egge H., Yu R.

K. 0-Acetylated gangliosides in bovine buttermilk. Characterization of 7-O-acetyl, 9-O-acetyl, and 7,9-di-O-acetyl G_D3. J. Biol. Chem. 1992, 267, 12632-12638.

Svennerholm L., Fredman P. A procedure for the quantitative isolation of brain gangliosides.

Biochim. Biophys. Acta 1980, 617, 97-109.

Takamizawa K., Iwamori M., Mutai M., Nagai Y. Gangliosides of bovine buttermilk. Isolation and characterization of a novel monosialoganglioside with a new branching structure. J. Biol. Chem. 1986, 261, 5625-5630.

Vaskovsky V. E., Kostetsky E. Y., Vasendin I. M. A universal reagent for phospholipid analysis.

J. Chromat. 1975, 114, 129-141.

Williams M. A., McCluer R. H. The use of Sep-Pak™ C₁₈ cartridges during the isolation of gangliosides. J. Neurochem. 1980, 35, 266-269. Zanetta J. -P., Timmerman P., Leroy Y. Gas-liquid chromatography of the heptafiuorobutyrate derivatives of the O-methyl-glycosides on capillary columns: a method for the quantitative determination of the monosaccharide composition of glycoproteins and glycolipids.

Glycobiology. 1999. V. 9. P. 255-266.

Claims

THE CLAIMS DEFINING THE INVENTION ARE AS FOLLOWS:

1. A process for obtaining a ganglioside-enriched extract from a ganglioside-containing lipid composition (GCLC), the process comprising (a) mixing a GCLC with one or more C₄-C₁₀ alcohols and water to produce a composition comprising an aqueous phase and an organic phase, (b) substantially separating the aqueous phase from the organic phase, wherein the aqueous phase comprises a ganglioside-enriched extract.

2. A process for increasing the ratio of one or more gangliosides to one or more other complex lipids in a GCLC, the process comprising

(a) mixing a GCLC with one or more C₄-C₁₀ alcohols and water to produce a composition comprising an aqueous phase and an organic phase,

(b) substantially separating the aqueoμs phase from the organic phase, wherein the wt% ratio of one or more gangliosides to one or more other complex lipids is increased in the aqueous phase compared to the wt% ratio of one or more gangliosides to one or more other complex lipids in the GCLC, and wherein the aqueous phase comprises a ganglioside-enriched extract.

3. A process of claim 2 wherein the wt% ratio of one or more gangliosides to the total phospholipids is increased in the aqueous phase compared to the wt% ratio of one or more gangliosides to the total phospholipids in the GCLC.

4. A process of claim 2 wherein one or more other complex lipids are selected from die group consisting of phospholipids, sphingolipids and neutral glycosphingolipids.

5. A process for increasing the ratio of one or more gangliosides to the total complex lipids in a GCLC, the process comprising

6. A process according to any one of claims 1-5 further comprising recovering the aqueous phase comprising the ganglioside-enriched extract.

7. A process according to any one of claims 1-6 further comprising recovering one or more gangliosides from the ganglioside-enriched extract.

8. A process according to claim 6 or 7 wherein the one or more gangliosides are recovered from the ganglioside-enriched extract by reversed phase solid phase extraction.

9. A process for extracting gangliosides from a GCLC, the process comprising

(b) substantially separating the aqueous phase from the organic phase, and

(c) recovering one or more gangliosides from the aqueous phase.

10. A process according to claim 9 wherein the one or more gangliosides are recovered from the aqueous phase by reversed phase solid phase extraction.

11. A process according to any one of claims 1-10 wherein the C₄-C₁₀ alcohol is n-butanol.

12. A process according to any one of claims 1-11 wherein citric acid-sodium citrate buffer is added to the water before mixing to form an aqueous solution of pH 5 to 6.

13. A process according to any one of claims 1-11 wherein EDTA or trisodium citrate is added to the water before mixing.

14. A process according to any one of claims 1-13 wherein the GCLC is an extract of a dairy ingredient, animal brains, or sea urchin roe.

15. A process according to any one of claims 1-14 wherein the GCLC is obtained by supercritical or near supercritical extraction of a dairy ingredient.

16. A ganglioside-enriched extract or one or more gangliosides obtained according to the process of any one of claims 1-15.

17. A ganglioside-entiched extract that is at least about 100% enriched in gangliosides compared to the GCLC from which it was extracted.

18. A ganglioside-enriched extract according to claim 17 wherein the extract is at least about 200% enriched in gangliosides compared to the GCLC from which it was extracted.

19. A ganglioside-enriched extract according to claim 17 or 18 wherein the extract is at least about 500% enriched in gangliosides compared to the GCLC from which it was extracted.

20. A ganglioside-enriched extract according to any one of claims 17-19 wherein the extract is at least about 1000% enriched in gangliosides compared to the GCLC from which it was extracted.

21. A ganglioside-enriched extract according to any one of claims 17-20 wherein the extract is enriched in one or more gangliosides relative to the total complex lipids in the GCLC from which it was -extracted.

22. A ganglioside-enriched extract according to any one of claims 17-20 wherein the extract is enriched in one or more gangliosides relative to the total phospholipids in the GCLC from which it was extracted.

23. A ganglioside-enriched extract according to any one of claims 17-22 obtained according to the process of any one of claims 1-7 and 11-15.

24. A food or beverage comprising one or more gangliosides or one or more ganglioside- enriched extracts according to any one of claims 16-23.

25. A food or beverage according to claim 24 where in the food or beverage is an infant formula or a maternal formula.

26. A pharmaceutical composition comprising one or more gangliosides or one or more ganglioside-enriched extracts according to any one of claims 16-23.