Simple glycolipids of microbes: Chemistry, biological activity and metabolic engineering.

Glycosylated lipids (GLs) are added-value lipid derivatives of great potential. Besides their interesting surface activities that qualify many of them to act as excellent ecological detergents, they have diverse biological activities with promising biomedical and cosmeceutical applications. Glycolipids, especially those of microbial origin, have interesting antimicrobial, anticancer, antiparasitic as well as immunomodulatory activities. Nonetheless, GLs are hardly accessing the market because of their high cost of production. We believe that experience of metabolic engineering (ME) of microbial lipids for biofuel production can now be harnessed towards a successful synthesis of microbial GLs for biomedical and other applications. This review presents chemical groups of bacterial and fungal GLs, their biological activities, their general biosynthetic pathways and an insight on ME strategies for their production.

Introduction

Lipid biotechnology research has focused to date on developing sustainable alternatives to depleting fossil fuels. One strategy was plant-derived fuel, biodiesel [1]. A main drawback of this approach is that oil and land allocated for biodiesel production compete with those allocated for human food consumption. Moreover, replacement of natural vegetations with plants used for biodiesel production generates long-term environmental concerns. Another strategy is to use lipids originating from microbes, called ‘‘single cell oil’’ (SCO) as substrates for biodiesel production. We believe that accumulating knowledge and developed biomolecular tools obtained from lipid engineering of oleogenic microbes can now be harnessed for the microbial production of lipid derivatives of added-value.

Glycosylation of organic molecules, including lipids, usually leads to derivatives of new and/or better physicochemical properties and biological activities [2], [3] that reflect in higher market prices. Metabolic engineering of lipid derivatives has previously investigated polyunsaturated fatty acids [4] and fatty acid derivatives that are used as substrates for oleochemical industries, e.g. heterologous production of ricinoleic acids by Y. lipolytica [5]. Other added-value lipid derivatives of commercial interest include wax esters, polyhydroxyalkanoates (bioplastics), hydroxylated fatty acids, carotenoids, polyenic polymers [6] and glycolipids.

This review focuses on simple glycolipids (SGLs) as an important family of glycolipids (GLs) class. The importance of SGLs stems from the fact that this family of GLs comprises a wide range of bioactive molecules with potential biomedical, pharmaceutical and cosmetic applications [7], [8]. Nonetheless, many simple GLs are limited commercially because of their still low yield and high cost of production, particularly of high purity simple GLs aimed for biopharmaceutical purposes.

We present the chemical groups of simple GLs, their microbial producers and their biological activities. Then, we describe the key biosynthetic enzymes and metabolic precursors involved in biosynthesis of simple GLs. Finally, we discuss metabolic engineering strategies for simple GLs production in native and heterologous hosts.

Definition and classification of simple glycolipids

The term glycolipids (GLs), in general, encompasses a wide diversity of structurally heterogeneous biological compounds that are produced by microbes, plants, animals and humans [9]. As their names suggest, they are composed of glycosyl and lipid moieties. The IUPAC uses the term GLs to broadly designate any compound containing one or more monosaccharide residues bound by glycosidic linkage to a hydrophobic moiety [10]. Our definition of GLs is even broader to include glycoside and non-glycoside GLs in which the sugar and lipid residues are linked together via glycosidic (e.g. O- or N-glycosidic linkages) and non-glycosidic linkages (e.g. ester or amide linkages), respectively (Fig. 1). The glycosyl residue can be mono-, di-, oligo or polysaccharides (e.g. glucose, cellobiose or glycan, respectively), alcohol sugars/polyols (like mannitol, erythritol or arabinol, etc.), amino sugars (like desosamine, etc) or sugar acids (like glucuronic acids). The lipid residue of GLs ranges from fatty acids, fatty alcohols, fatty amino alcohols, polyketides, sterols, hopanoids and carotenoids with different substitutions, chain lengths, saturation levels, branching and di-/oligo-/polymerizations.

Fig. 1

Classification of glycolipids and main types of linkages between their glycosyl and lipid residues. (A) Simple glycolipids (SGLs) comprise glycolipids consisting of glycosyl and lipid residues only, whereas, complex glycolipids (CGLs) contain glycerol, ceramide, phenolic, peptide, nucleoside or polysaccharide residues in addition to the glycosyl and lipid residues. (B) Glycosyl and lipid residues are mainly linked via O-glycosidic and/or ester bonds, and less frequently via N-glycosidic and/or amide bonds.

Numerous classifications exist for GLs [10], the most convenient of which is their classification into simple and complex GLs [11], [12], [13] (Fig. 1). Simple GLs (SGLs), sometimes called saccharolipids [14], are two-component (glycosyl and lipid moieties) GLs in which the glycosyl and lipid moieties are directly linked to each other. Complex glycolipids (CGLs) are, however, structurally more heterogeneous, as they contain, in addition to the glycosyl and lipid moieties, other residues like glycerol (glycoglycerolipids), peptide (glycopeptidolipids), acylated-sphingosine (glycosphingolipids), or other residues (Fig. 1). Polysaccharide-containing GLs, although containing no residues other than glycosyl and lipid moieties, are classified under complex glycolipids because of the complex nature of their polysaccharide residues; however, oligosaccharide-containing GLs are classified as simple GLs [13] (Fig. 1). Simple glycolipids addressed in this review are those of natural microbial origin, therefore, SGLs of synthetic or other biological origins are not mentioned.

Surfactant properties of simple glycolipids

Simple glycolipids (SGLs) are amphiphilic molecules as they comprise both the hydrophilic glycosyl and the lipophilic lipid residues. This amphiphilic nature confers surfactant activity to most GLs; those of which with pronounced surfactant activity are called biosurfactant. Compared to petroleum-derived (e.g. alkylbenzene sulfonates) or plant-based (e.g. alkyl polyglycosides) synthetic surfactants [15], microbially-produced SGL biosurfactants are mostly of higher surface activity, higher emulsifying power, lower critical micelle concentrations, higher biodegradability (compared to petroleum-derived surfactants), lower ecotoxicity [16] and lower protein denaturing potency [17], [18], [19]. The advanced properties of microbial SGLs are suggested to be attributed to a peculiar mosaic distribution of regions of polarity over the GL molecule, as well as to their branched or sometimes circular structures compared to synthetic surfactants [18]. Moreover, most SGLs are naturally produced as complex mixtures of congeners or homologues that vary in the number of glycosyl units and extent of their acylation, the number of conjugate lipid chains, their lengths, the extent of unsaturations and substitutions; these factors together contribute to their unique surfactant properties and behaviors [18].

Although the unique surface properties of some SGLs qualified some of them to be marketed as ecological surfactants [20], [21], yet, their competitiveness in the detergent market is limited because of their higher prices compared to alkyl polyglycosides synthetic surfactants which are at least 50% less expensive. For example, the estimated cost of large-scale production of the SGLs: sophorolipid and rhamnolipid biosurfactants, are about US$ 2.5–3/Kg [21], [22] and US$ 5–20/Kg [23], respectively, compared to US$ 1–3/Kg for the synthetic alkyl polyglycoside surfactants [23].

Aside from their surface activities, nearly all natural SGLs have interesting biological activities, as described later, that let them occupy market niches not approachable by synthetic surfactants [24]. Noteworthy, the biological activities of SGLs are thought to stem from their surface activities [25].

Chemical groups and origins of microbial simple glycolipids

Microbially produced SGLs are classified in chemical groups based on their chemical structures so that every group comprises SGLs members sharing unique glycosyl and/or lipid moieties for SGLs produced by bacteria (Table 1) and fungi (Table 2). In this classification, some SGLs congeners are classified in separate groups when they originate from different microbial origins and vice versa. Under each SGL group, exhaustive list of its members, together with their chemical names, their microbial producers as well as their taxonomic phyla is mentioned (Table 1, Table 2). Furthermore, the confirmed chemical structures of representative or prototypic members of each SGL group are presented (Fig. 3).

Table 1

Chemical groups and members of bacterial simple glycolipids as well as names and phyla of native producers.

Common name: Chemical names (Cx: chain length of fatty acid chains)	Producer	Phylum
Bacteria
1- Rhamnolipids
Monorhamnolipids: α-l-rhamnopyranosyl-R,R-3-(3′-hydroxyalkanoyloxy)alkanoate (C8-16)	Spp. of Pseudomonas and Burkholderia[29]	Proteobacteria
Dirhamnolipids: α-l-rhamnopyranosyl-(1–2)-α-l-rhamnopyranosyl-R,R-3-(3′-hydroxyalkanoyloxy)alkanoate (C8-16)	Spp. of Pseudomonas and Burkholderia[29]	Proteobacteria
2- Glucolipids
Rubiwettin RG1: β-d-glucopyranosyl 3-(3′-hydroxytetradecanoyloxy)decanoate	Serratia rubidaea[30].	Proteobacteria
3- Trehalolipids
α,α-(1-1)-Trehalose 6-mono-O-mycolates	Rhodococcus erythropolis[31]	Actinobacteria
α,α-(1-1)-Trehalose 2,3-di-O-mycolates	Tsukamurella sp. [32]	Actinobacteria
Cord factor: α,α-(1-1)-Trehalose 6,6′-di-O-mycolates	Spp. of Mycobacterium, Rhodococcus, Arthrobacter, Nocardia and Gordonia[31], [33]	Actinobacteria
α,α-(1-1)-Trehalose 2,3,6′-tri-O-mycolates	Rhodococcus aurantiacus[34]	Actinobacteria
STL-1, α,α-(1-1)-Trehalose 2,2′-di-O-succinoyl-3,4-di-O-alkanoates	Rhodococcus erythropolis[35]	Actinobacteria
STL-2, α,α-(1-1)-Trehalose 2,3,4-mono-O-succinoyl-di-O-alkanoates	Rhodococcus erythropolis[35]	Actinobacteria
STL-3, α,α-(1-1)-Trehalose 2,3,4,2′-mono-O-succinoyl-tri-O-alkanoates	Spp. of Rhodococcus[36], [37], Arthrobacter[38]	Actinobacteria
4- Other glycosylated (non-trehalose containing) mycolates
Sucrose 6-mono-O-mycolates	Spp. of Arthrobacter, Corynebacterium, Nocardia, Brevibacterium[39], [40]	Actinobacteria
Fructose 6-mono-O-mycolates	Spp. of Arthrobacter, Corynebacterium, Nocardia, Brevibacterium[39], [40]	Actinobacteria
Fructose 1,6-di-O-mycolates	Spp. of Arthrobacter, Corynebacterium, Nocardia, Brevibacterium[39], [40]	Actinobacteria
Glucose-6-β-hydroxy-α-hexadecenoyl-eicosenoate	Brevibacterium thiogenitalis[41]	Actinobacteria
Mannose 6-mono-O-mycolates	Arthrobacter sp. [42]	Actinobacteria
Maltose 6-mono-O-mycolates	Arthrobacter sp. [42]	Actinobacteria
Maltose 6,6′-di-O-mycolates	Arthrobacter sp. [42]	Actinobacteria
Maltotriose 6,6′,6″-tri-O-mycolates	Arthrobacter sp. [42]	Actinobacteria
Cellobiose 6-mono-O-mycolates	Arthrobacter sp. [42]	Actinobacteria
5- Trehalose-containging Oligosaccharide lipids
Lipid Q: β-d-glucose-(1–3)-α,α-(1-1)-trehalose hexanoyl-succinoyl-3-(hexanoyloxy)octanoate-3-(hexanoyloxy)decanoate	Rhodococcus sp. [43], [44]	Actinobacteria
GL2: β-d-glucose-(1″-2′)-α,α-(1-1)-trehalose 4,6,2″,3″ tetra-O-alkanoates (C8-10)	Tsukamurella sp. [32]	Actinobacteria
GL3: β-d-glucose-(1″-2′)-α,α-(1-1)-trehalose-(6′-1‴)-β-d-galactose-4,6,2″,3″-tetra-O-alkanoates (C8-10)	Tsukamurella sp. [32]	Actinobacteria
β-d-glucose-(1–3)-α,α-(1-1)-trehalose-(6-1)-β-d-glucose-(6-1)-β-d-glucose mono-O-succinoyl-hepta-O-alkanoate (C2-8)	Nocardia corynebacteroides[45], [46], [47]	Actinobacteria
4,6-(1-Carboxyethylidene)-3-O-Me-β-d-glucose-(1–3)-4,6-(1-carboxyethylidene)-β-d-glucose-(1–4)-β-d-glucose-(1–6)-α,α-(1″-1′)-trehalose-4″-O-alkanoyl-6′-O-alkenoate	Mycobacterium smegmatis[48], [49]	Actinobacteria
6- Glycosylated fatty alcohols
Alkane 1,2-diol glycoside; Hexose 1-(O-hexose)alk-2-yl alkanoate (Diol = C19-20, alkanoate = C14-16)	Roseiflexus castenholzii[50]	Chloroflexi
1-(O-hexose)-3,25-hexacosanediol and its homologue: 1-(O-hexose)-3,27-octacosanediol	Spp. of cyanobacteria e.g. Anabaena, Nodularia, Calothrix, Synechococcus[51]	Cyanobacteria
1-(O-hexose)-3-keto-25-hexacosanol and its homologue: 1-(O-hexose)-3-keto-27-octacosanol		Cyanobacteria
1-(O-hexose)-3,25,27-octacosanetriol		Cyanobacteria
1-(O-hexose)-3-keto-25,27-octacosanediol OR its isomer: 1-(O-hexose)-27-keto-3,25-octacosanediol		Cyanobacteria
7- Glycosylated macro-lactones/-lactams
Barsilinolide A/B/C: 2-deoxy-α-l-fucopyranoside of C32-membered macrolactone	Nocardia brasiliensis[52], [53]	Actinobacteria
Fluvirucins: amino sugar glycosides of C14-membered macrolactam	Spp. of Actinomadura, Streptomyces, Microtetraspora and Saccharotrix mutabilis	Actinobacteria
Vicenistatin: amino sugar (vicenisamine) glycoside of C20-membered macrolactam	Streptomyces sp. [54], [55]	Actinobacteria
Vicenistatin M: d-mycarose glycoside of C20-membered macrolactam	Streptomyces sp. [54], [55]	Actinobacteria
Erythromycins A, B, D, C, E, F and Erythromycin esters (C14-membered macrolactam glycosides)	Streptomyces erythreus and Nocardia spp, and other Streptomyces spp. [56]	Actinobacteria
Oleandomycin (C14-membered macrolactam glycosides)	Streptomyces antibioticus[57]	Actinobacteria
Pikromycin, Narbomycin, 5-O-mycaminosyl-narbonolide (C14-membered macrolactam glycosides)	Streptomyces felleus and S. narbonensis[56]	Actinobacteria
10,11-Dihydropikromycin, Kayamicin (C14-membered macrolactam glycosides)	Streptomyces narbonensis[56]	Actinobacteria
Spinosyns (Tetracyclic macrolide) containing forosamine (amino sugar) and tri-O-methyl rhamnose	Saccharopolyspora spinosa[56], [58]	Actinobacteria
Lepicidin A	Saccharopolyspora spinosa[56]	Actinobacteria
Leucomycins, Josamycin, Platenomycins, Medicamycin, Espinomycins	Streptomyces kitasatoensis[56]	Actinobacteria
Carbomycin B, platenomycins W1/W2, Niddamycin, Midecamycin A3/A4	Streptomyces platensis[56]	Actinobacteria
Acumycin (cirramycin B), Cirramycin F and derivatives	Streptomyces griseoflavus, S. fradiae, S. flocculus[56]	Actinobacteria
Chalcomycin, Neutramycin	Streptomyces bikiniensis, S. rimosus, S. hirsutus[56]	Actinobacteria
Aldgamycin F, E and Swalpamycin	Streptomyces lavendulae, S. avidinii, S amandii (for swalpamycin) [56]	Actinobacteria
Spiramicins	Streptomyces ambofaciens[56], [59]	Actinobacteria
Tylosins	Streptomyces fradie, S. hygroscopicus[56], [60]	Actinobacteria
Concanamycins	Streptomyces diastatochromogenes[56]	Actinobacteria
Tetrins and related compounds, Maduralide	Streptomyces sp. [56]	Actinobacteria
Pimaricin	Streptomyces natalensis[56]	Actinobacteria
Colubricidin A	Streptomyces sp. [56], [61]	Actinobacteria
Nystatin	Streptomyces noursei[62]	Actinobacteria
Amphotericin B	Streptomyces nodosus[63]	Actinobacteria
Oasomycins, Desertomycins	Streptomyces cinnamoneus (previsouly Streptoverticillium baldacci)	Actinobacteria
Rapamycin	Streptomyces hygroscopicus[64]	Actinobacteria
Avermectins	Streptomyces avermetilis[65]	Actinobacteria
PM100117 and PM100118	Streptomyces caniferus	Actinobacteria
8- Glycomacrodiolides (glycosylated macrocyclic dilactones)
Glucolipsin A, B: dilactone of two glucosides of 3-hydroxy fatty acids C19/C19	Streptomyces purpurogeniscleroticus, Nocardia vaccinii[66]	Actinobacteria
Fattiviracin A1: dilactone of two glucosides of 3-,17-, ω-1-trihydroxy fatty acids C24/C33	Kibdelosporangium albatum[67]	Actinobacteria
Cycloviracin B1 and B2: dilactones glucosides of 3-,19-, ω-1-trihydroxy fatty acids (C22–28) and of 3-,17-, ω-1-trihydroxy fatty acids (C22–24)	Streptomyces microflavus[67]	Actinobacteria
Elaiophylins, Efomycin G	Streptomyces spp. [68]	Actinobacteria
Halichoblelides A, B, C	Streptomyces spp. [69], [70], [71]	Actinobacteria
Bispolides A1, A2, A3, B1, B2a, B2b and B3	Microbispora species[72]	Actinobacteria
Macroviracins A-D: related to fattiviracin and cycolviracins	Streptomyces sp. [73]	Actinobacteria
9- Glyco-carotenoids/-terpenoids:
9.1-Acyclic glycocarotenoids
Rhodopsin glucoside	Halorhodospira abdelmalekii, H. halochloris[74]	Proteobacteria
Dihydroxylycopene mono-/di-glucosides and their acyl (C12:0 or C14:1) derivatives	Halorhodospira abdelmalekii, H. halochloris[74]	Proteobacteria
d-Glucosyl 4,4″-diapocarotene-6,6′-dioic acid	Pseudomonas rhodos[75], Rhizobium lupini[76], [77]	Proteobacteria
1′-glucosyloxy-3′,4′-didehydro1′,2′-dihydro-ψ,ψ-carotene monoester	Chondromyces apiculatus[78], Myxococcus fulvus[79]	Proteobacteria
Staphyloxanthin: α-d-glucopyranosyl 1-O-(4,4′-diaponeurosporen-4-oate) 6-O-(12-methyltetradecanoate)	Staphylococcus spp. [80]	Firmicutes
4-D-glucopyranosyloxy-4,4′-diaponeurosporene	Streptococcus faecium[81]	Firmicutes
Hydroxy-diaponeurosporene glucoside esters	Heliorestis sp. [82]	Firmicutes
Rhodopin β-D-glucoside, Rhodopinal β-D-glucoside	Rhodopseudomonas acidophila, Rhodospirillum tenue and Rhodocyclus purpureus[83]	Proteobacteria
Oscillaxanthin: 1,1′-dihydroxy-2,2′-di-β-l-rhanmosyl- 1,2,1′,2′-tetrahydro-3,4,3′,4′-tetradehydrolycopene	Oscillatoria rubescens[84]	Cyanobacteria
Bacterioruberin mono- and di-glycosides	Unidentified Halophilic bacterium[85]	Proteobacteria
Diapolycopenedioic acid xylosyl esters A, B, and C	Rubritalea squalenifaciens[86]	Verrucomicrobia
Methyl 5-glucosyl-5,6-dihydro-apo-4,4′-lycopenoate	Planococcus maritimus[87]	Firmicutes
Vancoresmycin	Amycolatopsis[88]	Actinobacteria
9.2-Monocyclic glycocarotenoids
Salinixanthin	Salinibacter ruber[89], Rhodothermus marinus[90]	Bacteroidetes
Phleixanthophyll, 4-ketophleixanthophyll	Mycobacetrium phlei[91]	Actinobacteria
Phleixanthophyll palmitate: (2′-S)-1′-[(6-O-palmityl-β-d-glucopyranosyl)oxy]-3′,4′-didehydro-1′,2′-dihydro-β,ψ-caroten-2′-ol	Nocardia sp. [92]	Actinobacteria
1′-[(6-O-acyl-β-d-glucopyranosyl)oxy]-1′,2′-dihydro-β,ψ-caroten-4-one	Rhodococcus rhodochrous[93], [94]	Actinobacteria
Myxobactone	Myxococcus fulvus[79], [95]	Proteobacteria
Myxobactin	Myxococcus fulvus[96]	Proteobacteria
Keto-myxocoxanthin glucoside ester (Myxobactone ester)	Roseiflexus castenholzii[97]	Chloroflexi
OH-γ-carotene glucoside laurate: 1'-[(6-O-lauryl-β-d-glucopyranosyl)oxy]-1′,2′-dihydro-β,ψ-carotene	Chlorobium tepidum[98]	Chlorobi
OH-chlorobactene glucoside laurate; 1'-[(6-O-lauryl-β-d-glucopyranosyl)oxy]-1′,2′-dihydro-ɸ,ψ-carotene	Chlorobium tepidum[98]	Chlorobi
OH-γ-carotene glucoside ester derivative	Chloroflexus aurantiacus[99]	Chloroflexi
1′-β-glucopyranosyl-3,4,3′,4′-tetradehydro-1′,2′-dihydro-β,ψ-caroten-2-one	Meiothermus ruber[100]	Deinococcus-Thermus
Myxoxanthophyll like glycocarotenoid: (3R,2′S)-myxol-2′-(2,4-di-O-methyl-α-L-fucoside)	Synechocystis sp. [101]	Cyanobacteria
Sioxanthin; (2′S)-1′-(β-D-glucopyranosyloxy)-3′,4′didehydro-1′,2′-dihydro-φ,ψ-caroten-2′-ol	Salinispora sp. [102]	Actinobacteria
9.3-Bicyclic glycocarotenoids
Corynexanthin monoglycoside	Corynebacterium sp. [103]	Actinobacteria
Corynexanthin diglycoside	Arthrobacter sp [104]	Actinobacteria
Sarcixanthin monoglucosides	Curtobacterium flaccumfaciens[105], Micrococcus luteus[106], M. yunnanensis[107].	Actinobacteria
Sarcixanthin diglucosides	Micrococcus luteus[106], M. yunnanensis[107].	Actinobacteria
Zeaxanthin mono- and di-glucosides	Erwinia herbicola, Rhodobacter sphaeroides[108]	Proteobacteria
Zeaxanthin mono- and di-rhamnosides (mainly Z-isomers), Zeaxanthin di-glucoside	Sulfolobus shibatae[109]	Archeabacteria
Zeaxanthin mono- and di-rhamnosides	Corynebacterium autotrophicum (Xanthobacter autotrophicus) [110]	Proteobacteria
Aastaxanthin dirhamnoside	Sphingomonas astaxanthinifaciens[111]	Proteobacteria
Myxocoxanthin rhamnoside	Sorangium compositum[112]	Proteobacteria
Thermozeaxanthin-13, -15, and -17 (Zeaxanthin mono-β-D-glucoside-branched fatty acid esters)	Thermus thermophilis[113]	Deinococcus-Thermus
Thermobiszeaxanthin-13-13, -13-15, and -15-15 (Zeaxanthin di-β-D-glucoside-branched fatty acid esters)	Thermus thermophilis[113]	Deinococcus-Thermus
Adonixanthin and astaxanthin glucosides	Agrobacterium aurantiacum[114], [115]	Proteobacteria
Decaprenoxanthin mono- and diglucoside (2R,6S,2′R,6′S)-(2,2′-bis(4-hydroxy-3-methyl-2-butenyl)ε,ε-carotene) di-β-D-glucoside	Corynebacterium glutamicum[116]	Actinobacteria
10- Glycosylated hopanoids/sterols
Bacteriohopanetetrol cyclitol ether	Chloracidobacterium thermophilum[117], [118]	Acidobacteria
BHT cyclitol	Burkholderia cenocepacia[119], Zymomonas mobilis[120][121]	Proteobacteria
BHT glucosamine	Burkholderia cenocepacia[119], Zymomonas mobilis[120], [121]	Proteobacteria
O-α-d-Glucuronopyranosyl BHT	Rhodospirillum rubrum[122]	Proteobacteria
Cholesteryl-α-d-glucopyranoside	Helicobacter pylori[123]	Proteobacteria
Cholesteryl-6-O-tetradecanoyl-α-d-glucopyranoside, Cholesteryl-6-O-dodecanoyl-α-d-glucopyranoside	Helicobacter pylori[123], H. felis, H. muridarum, H. mustelae, H. fennelliae, and H. cinaedi[124]	Proteobacteria
Cholesteryl-6-O-acyl-β-d-galactopyranoside	Borrelia burgdorferi, B. garinii and B. afzelii[125]	Spirochaetes
Cholesteryl-6-O-acyl-β-d-glucopyranoside	Borrelia hermsii[125]	Spirochaetes
Cholesteryl-β-d-glucopyranoside and its 3,4,6-triacyl derivatives	Mycoplasma gallinarum[126]	Tenericutes
Cholesteryl-α,α′-di-D-glucoside; α-d-glucopyranosyl-(1 → 3)-(O-acyl)-α-d-glucopyranosyl-(1 → 3)-cholesterol	Acholeplasma axanthum[127]	Tenericutes

Table 2

Chemical groups and members of fungal simple glycolipids as well as names and phyla of their native producers.

Common name: Chemical names (Cx: chain length of fatty acid chains)	Producer	Phylum
Fungi
1- Mannosyl-erythritol lipids (MEL, Ustilipids) and MEL congeners
MEL	Ustilago maydis, Pseudozyma (Candida) antarctica[128], [129], Kurtzmanomyces[130]	Basidiomycota
MEL	Geotrichum candidum[131]	Ascomycota
Mannosylmannitol lipids (MML), mannosylribitol lipids (MRL) and mannosylarabitol lipids (MAL)	Pseudozyma parantarctica[132]	Basidiomycota
2- Cellobiose lipids (CL, Ustilagic acids)
Cellobiose (β-D-Glc-(1 → 4)-β-D-Glc) 2″-O-hexanoic acid 1-O-16-ω,ω-1-dihydroxyhexadecanoate or 1-O-16-ω,ω-1,α-trihydroxy hexadecanoate	Ustilago maydis[128]	Basidiomycota
Cellobiose 6′-O-acetyl-2″-O-β-hydroxyalkanoyl-1-O-16-ω,ω-1-dihydroxyhexadecanoate or 1-O-16-ω,ω-1,α-trihydroxy hexadecanoate methyl ester	Ustilago maydis[128], Pseudozyma fusiformata[133]	Basidiomycota
Cellobiose 16-O-ω,ω-1-dihydroxyhexadecanoate or 1-O-16-ω,ω-1,α-trihydroxy hexadecanoate	Sympodiomycopsis paphiopedili[134]	Basidiomycota
Microcin: cellobiose 2″,3″,4″,6″,6′-penta-O-acetyl-1-O-16-ω,α-dihydroxyhexadecanoate	Cryptococcus humicola[135]	Basidiomycota
Flocculosin: 2-(2′,4′-diacetoxy-5′-carboxy-pentanoyl) octadecyl cellobioside	Anthracocystis (Pseudozyma) flocculosa[136]	Basidiomycota
3- Sophorolipids
Sophorose (β-D-Glc-(1 → 2)-β-D-Glc) -1-O-16-ω,ω-1-dihydroxyalkanoate or 1-O-16-ω,ω-1,α-trihydroxy alkanoate (C16-C18:0-2)	Starmerella (Candida) bombicola, Candida apicola and other spp.[137]	Ascomycota
Sophorose (β-D-Glc-(1 → 2)-β-D-Glc) -1-O-16-ω,ω-1-dihydroxyalkanoate or 1-O-16-ω,ω-1,α-trihydroxy alkanoate (C16-C18:0-2)	Cryptococcus curvatus[137]	Basidiomycota
Sophorose 6′-mono-O-acetyl or 6′,6″-di-O-acetyl -1-O-16-ω,ω-1-dihydroxyalkanoate or 1-O-16-ω,ω-1,α-trihydroxy alkanoate (C16-C20:1)	Starmerella (Candida) bombicola, Candida apicola and other spp.[137], Wickerhamiella domercqiae [138], [139]	Ascomycota
Sophorose lipid lactonic/ring form, lactonization of free carboxyl group with C-4″ or C-6'' (intramolecular ester bonds)	Starmerella (Candida) bombicola, Candida apicola and other spp.[137]	Ascomycota
Dimeric and trimeric sophorolipids (intermolecular ester bonds between carboxyl of one molecule to C-4″ of another molecule)	Candida spp. [140]	Ascomycota
4- Glucosyl-di-xylosyl lipids (Glykenins)
Glykenins A, B, C: O-β-d-glucose-(1 → 2)-O-β-d-xylose-(1 → 2)-O-β-d-xylose tetrahydroxyhexacosanoic acids, mono-di or tri-acetylated	Basidiomycetous sp. [141]	Basidiomycota
5- Polyol fatty acid esters (Liamocins and their congeners)
Liamocins	Aureobasidium pullulans	Ascomycota
Mannitol and pentitol esters of 3-D-hydroxypalmitic and 3-D-hydroxystearic acids	Rhodotorula glutinis and Rhodotorula graminis	Basidiomycota
6- Glucosyl and mannosyl lipids
Monoglucosyloxyoctadecenoic acid	Aspergillus niger[142]	Ascomycota
Halymecin B: mannosylated tetramer of 3,5-dihydroxydecanoic acid	Fusarium sp. [143]	Ascomycota
Halymecins F: acetylated halemycin B, halymecin G: mannosylated trimer of 3,5-dihydroxydecanoic acid	Simplicillium lamellicola[144]	Ascomycota
(3R,5R)-3-O-β-d-mannosyl-3,5-dihydroxydecanoic acid	Simplicillium lamellicola[144]	Ascomycota
7- Glycosylated polyketides
Roselipin 1, 2: 2,4,6,8,10,12,14,16,18-nonamethyl-5,9,13-trihydroxy-2E,6E,10E-icosenoic acid mannosylated (+-acetylated) at C-13, D-arabitol ester [145], [146]	Gliocladium[145], [146]	Ascomycota
TMC-151 A ∼ F: 2,4,6,8,10,12,14,16,18-nonamethyl-5,9,13-trihydroxy-2E,6E,10E-icosenoic acid mannosylated (+-acetylated) at C-13, D-mannitol ester [145], [146]	Gliocladium catenulatum[147]	Ascomycota
TMC-154: isolmeric form of roselipin 1 and TMC-171 A ∼ C: as roselipin 3 but esterified to mannitol	Gliocladium[148]	Ascomycota
Roselipins 3A to 3E: 14,15-dehydro derivatives of roselipin 1A/B	Clonostachys candelabrum[149]	Ascomycota
Cladionol A: 15-mannosyl-2,4,6,8,10,12,14,16,18,20-decamethyl-3,7,11,15-tetrahydroxy-4E,8E,12E-docosenoic acid arabitol ester	Gliocladium[150]	Ascomycota
8- Glucosyl-galactosyl lipids
Emmyguyacin 1A: α-d-glucopyranosyl-α-d-galactopyranose 3′-O-hydroxydocosanoate with 17-((carboxycarbonyl)oxy) group of oxalate ester at OH of C-17	Fungal species[151]	NA
Emmyguyacin 1B: Trehalose 3′-O-docosanoate with 17-((carboxycarbonyl)oxy) group of oxalate ester at OH of C-17	Fungal species[151]	NA
Emmyguyacin 2: as emmyguyacin 1A without the oxalate ester	Fungal species[151]	NA
9- Glycosylated sterols
Ergosterol-β-d-glucopyranoside	Pichia pastoris, Sordaria macrospora, Rhynchosporium secalis[152]	Ascomycota
10- Glycosylated paraconic acids
Gobienines A/B/C (non-confirmed structure [28])	Acarospora gobiensis (Lichen) [27]	Ascomycota

Fig. 3

Structures of prototypic members of bacterial and fungal simple glycolipid (SGL) groups. The glycosyl and lipid residues are colored in red and blue, respectively. Bacterial and fungal SGLs are represented in the upper and lower halves (separated by a line) of the figure, respectively. The representative structure of fungally produced glycosylated paraconic acids (20th group of SGLs) is not given as their structures have been debated [27], [28].

Based on our survey of microorganisms producing SGLs, we found that 50% of all known microbial SGLs are produced by microbes belonging to the phylum Actinobacteria (Fig. 2). Second in rank to Actinobacteria, comes phylum Proteobacteria followed by the two major fungal phyla, Ascomycota and Basidiomycota, consecutively (Fig. 2).

Fig. 2

Approximate distribution of microbial producers of simple glycolipids in different bacterial and fungal phyla. Incidences of microbial production of chemically unique SGLs in every microbial phylum were counted. Homologues or stereoisomers of the same chemically unique SGLs did not add into these calculations to avoid false overestimations. As an example, rhamnolipids (RLs) exist in two unique structures known so far, mono-rhamnolipids and di-rhamnolipids containing one and two rhamnose moieties, respectively, and are produced by proteobacterial species. Although these two RL congeners has several homologues varying in chain length of their lipid moiety, they were counted as two chemically unique SGLs in our calculations. Only the phylum of the microbial producer and not its genus and species identity that was taken into account; for example, although di-rhamnolipids are produced by different species of the genus Pseudomonas and Burkholderia, all di-rhamnolipids scored one hit in our calculations because all these di-rhamnolipids producers belong to the same phylum, Proteobacteria.

Bacterial simple glycolipids

Overall, bacterially produced simple glycolipids (SGLs) outnumbers fungally-produced ones (Fig. 2). A previous survey of about 16000 pooled natural bacterial metabolites revealed that about 20% of them are glycosylated, about 30% of these glycosylated metabolites are glycosylated lipids of which glycosylated macro-lactones/-lactams take a share of about 20% and other glycosylated lipids (including SGLs) take a share of 10% [26]. Nearly all glycosylated macro-lactones/-lactams are produced by members of the phylum Actinobacteria. We classified baterially produced SGLs in 10 groups (Table 1).

Fungal simple glycolipids

Fungally produced simple glycolipids (SGLs) are less numerous than bacterially-produced ones (Fig. 2). Fungal SGLs are classified in 10 groups that are mainly produced by members of the phyla Ascomycota and Basidiomycota (Table 2).

Physiological roles of simple glycolipids

For most SGLs, the exact physiological roles to their native producers are not clearly known. Generally, SGLs are secondary metabolites that are not essential for cell viability. Nonetheless, given their antimicrobial properties, SGLs are suggested to help producing organism dominate environmental niches by inhibiting the growth of other organisms [153]. In addition, SGLs are required to coordinate multicellular or group behaviors (biofilm formation and swarming) and enhance growth of producing organisms on hydrophobic carbon sources [154], [155], [156]. Some additional roles are assigned to specific SGLs like rhamnolipids, which are considered as virulence factors that modulate host immune response [29]. Similarly to their unglycosylated counterparts, glycosylated carotenoids are postulated to act as photoprotectants and antioxidants to protect organisms from injuries caused by free radicals and active oxygen species [106]. In thermophiles, glycocarotenoids are thought to stabilize and reinforce cell membranes [113]. Hopanoids are sterol analogues in bacteria. Similarly to sterol in eukaryotes, hopanoids and their glycosylated derivatives are thought to help stabilize and regulate membrane fluidity and permeability particularly during shifts in pH and other physicochemical conditions [117], [157]. Sophorolipids are suggested to act as extracellular forms of carbon storage that can be recycled later under starvation conditions [154].

Bioactivities of simple glycolipids

Simple glycolipids (SGLs) have very interesting biological activities on other organisms ranging from viruses to human cells. Although the mechanism of these bioactivities is not definitively known, it is suggested that most of SGLs bioactivities arise from their surface activities. Collectively, many of them have antiviral, antimicrobial, anti-inflammatory and anticancer activities (Table 3). Many reviews are found in literature detailing the potential biomedical and cosmeceutical applications of biosurfactants in general, many of which are simple glycolipids [8], [158], [159], [160].

Table 3

Biological activities of different chemical groups of microbial simple glycolipids.

	Chemical group of simple glycolipids (SGLs)	A	B	C	D	E	F	G	H	I	J	K	L
		Antibacterial	Antifungal	Antiviral	Antiparasitic	Anticancer	↑ Cell differentiationa	Immunomodulatory	Antioxidant	Antiadherent (Biofilm, wounds)	Neuronal activity	Sperm immobilizing activity	↓ Diacylglycerol acyl transferase 2a
	Bacterial SGLs
1	Rhamnolipids	A1	B1	C1	D1			G1		I1
2	Glycolipids (Rubiwettin)
3	Trehalolipids			C3		E3	F3	G3
4	Other glycosylated mycolates
5	Oligosaccharide lipids					E5
6	Glycosylated fatty alcohols
7	Glycosylated macro-lactones/-lactams	A7	B7	C7	D7	E7		G7
8	Glycomacrodiolides	A8	B8	C8	D8	E8		G8
9	Glyco-carotenoids/-terpenoids							G9	H9
10	Glycosylated hopanoids
	Fungal SGLs
11	Mannosyl-erythritol lipids	A11				E11	F11		H11		J11
12	Cellobiose lipids	A12	B12
13	Sophorolipids			C13			F13	G13		I13		K13
14	Glucosyl-di-xylosyl lipids (Glykenins)	A14
15	Polyol fatty acid esters	A15				E15
16	Glucosyl and mannosyl lipids	A16				E16
17	Glycosylated polyketides			C17	D17								L17
18	Glucosyl-galactosyl lipids			C18
19	Glycosylated sterols
20	Glycosylated paraconic acids

The signs ↑ and ↓ denotes for stimulation and inhibition, respectively.

Biosynthesis of simple glycolipids

With few exceptions, the exact biosynthetic steps of majority of simple glycolipids (SGLs) are not yet fully understood. Generally however, biosynthesis of SGLs implicates the supply and linking of glycosyl and lipid precursors. Pathways supplying glycolipid precursors are depicted later (Fig. 5) and are thought to play an important role in regulation of SGLs biosynthesis. Linking of glycosyl and lipid precursors is mostly via O-glycosidic or ester bonds (Fig. 1B) that are formed by glycosyltransferases (GT) (Fig. 4 B1, B2) [215] or acyltransferases (AT) [216] (Fig. 4 A1, A2), respectively. Glycosyltransferases catalyze the transfer of the sugar moiety from an activated glycosyl donor, usually sugar-nucleotide (Leloir GTs) or –phosphate (non-Leloir GTs), to a lipid acceptor (or a sugar acceptor for extending the sugar backbone of glycolipids), by making glycosidic bonds between the hydroxyl groups (nucleophile) of the acceptor and the anomeric carbon of the sugar donor (Fig. 4 B1, B2) [215]. Acyltransferases (AT) catalyze the transfer of the lipid moiety from an activated acyl donor, mostly acyl-CoA or -ACP, to a glycosyl acceptor (or a lipid acceptor for extending the lipid backbone of the glycolipid) by making an ester bond between the hydroxyl group (nucleophile) of the acceptor and the acyl donor's carbonyl group [216] (Fig. 4 A1, A2).

Fig. 5

Sugar and lipid precursors of prominent members of simple glycolipid groups and their furnishing pathways. Biosynthesis of simple glycolipids that harbor glucoside units, like glucosides of astaxanthin [115] and zeaxanthin [223] as well as glucosylated sterols [224], cellobiose and sophorose lipids [225], require UDP-glucose as glycosyl donor. Peculiar glycosyl donors that are activated in other ways than UDP are required in case of rhamnolipids [226], trehalolipids [227], vicenistatin [228] and elaiophylin [68]. All glycosyl donors are derived from glucose -1-phosphate except that of glycosylated hopanoids whose glycosyl donors is derived from β-d-fructofuranose-6-phosphate and ribose [119]. Mannosylerythritol lipids are expected to derive the glycosyl unit, mannosylerythritol, from erythritol and GDP-mannose which originate from d-erythrose-4-phosphate and β-d-fructofuranose-6-phosphate intermediates of the pentose phosphate pathway as described in the yeast Yarrowia lipolytica[229], [230]. The lipid moiety originates mostly from fatty acid synthesis and/or β-oxidation except glycosylated macrolides, carotenoids and sterols/hopanoids whose lipid moiety is furnished from the polyketide for the former and from mevalonate/isoprenoid pathway for the latter two groups. 6PGL: 6-phosphogluconolactone; 6PG: 6-phosphogluconic acid; Ru5P: d-Ribulose-5-phosphate; R5P: d-Ribose-5-phosphate; X5P: d-Xylulose-5-phosphate; S7P: d-Sedoheptulose-7-phosphate; Gly3P: D-Glyceraldehyde-3-phosphate; DHAP: Dihydroxyacetone phosphate; Ery4P: d-Erythrose-4-phosphate; Fru6P: β-d-Fructofuranose-6-phosphate; Fru1,6BP: β-d-Fructofuranose-1,6-bisphosphate; dTDP-Rha: dTDP-rhamnose; Tre6P: Trehalose-6-phosphate; Glc: d-Glucopyranose; Glc1P: Glucose-1-phosphate; Glc6P: Glucose-6-phosphate; UDP-glc: Uridine diphosphate glucose; UDP-AcGln: Uridine diphosphate N-acetylglucosamine; Man1P: Mannose-1-phosphate; GDP-Man: GDP-mannose; : R-3-hydroxy acyl-X1 (X1 = -CoA/-ACP); X2 = Mycolyl Carrier Protein. PPP: Pentose Phosphate Pathway. Dashed lines means that multiple biosynthetic steps are involved.

Fig. 4

Key enzymes of glycolipid biosynthesis and hydrolysis. Last steps of glycolipid biosynthesis involves linking of sugar and lipid moieties via either or both Acyl Transferases (AT) (A1 and A2, forward reactions) and Glycosyl Transferases (GT) (B1 and B2, forward reactions) which catalyze the ester and glycosidic bonds formation, respectively. Glycolipids are catabolized or broken down by Lipid Esterase (LE), Carbohydrate Esterases (CE) and Glycoside Hydrolases (GH) that hydrolyze the bond between alkyl-alkanoate ester, acyl-sugar ester and glycosidic bonds, respectively (reverse reactions). L1: Coenzyme A (CoA-S-) or Acyl Carrier Protein (ACP-S-) activating groups on acyl donors; L2: Nucleotides or phosphates activating groups on glycosyl donors. R: any substitution that could be glycosyl, lipid, or glycolipid units. Notes: β-glucose and R-3-hydroxyalkanoate are used as examples of any sugar and hydroxyl fatty acid of any chain length (n), respectively. Hydrolysis reactions do not generate activated products.

Concerning the fate of SGLs, one report showed that the flocculosin GL can be degraded by its producing yeast, Pseudozyma flocculosa, which feeds on it under nutrient limitations [153]. Glycolipids could theoretically be hydrolyzed by one or more of the following enzymes. First, glycoside hydrolases (GH) that hydrolyze the sugar-sugar or sugar-lipid glycosidic bonds (Fig. 4 B1, B2) [153], [217]. Second, carbohydrate esterases (CE) hydrolyze the sugar-lipid ester bonds (Fig. 4 A2). Lipid esterases (LE), also known as lipases, hydrolyze lipid-lipid ester bonds (Fig. 4 A1) in glycolipids with multimeric hydroxyalkanoate lipid moieties e.g. rhamnolipids (Fig. 3). This hypothesis is corroborated by reports showing the hydrolysis of polymeric hydroxyalkanoates, which share the same lipid moieties as rhamnolipids, by the action of microbial lipases/esterases [218], [219], [220]. Nonetheless, the metabolic fate of SGLs is one of the subjects that require thorough investigations.

Among the poorly studied aspects in SGLs metabolism also are the transport SGLs across microbial membranes. Some SGLs require active transport for their exportation out of the cell, like cellobiose lipids, mannosylerythritol lipids [221] and sophorolipids [222], whereas, many other SGLs are thought to passively diffuse out of the cell.

To give a general overview, we present the general biosynthetic map of SGLs showing the diversity of the immediate glycosyl and lipid precursors of SGLs and the pathways furnishing them (Fig. 5).

Metabolic engineering of simple glycolipids

Metabolic engineering can be employed to satisfy demand for simple glycolipids (SGLs) by offering solutions to the main challenges facing their production and commercialization. The most important challenge is the high cost of production of SGLs at high purities to qualify for medical or cosmeceutical applications (usually >90–95% purities are required) [21], [22]. This high cost stems from a multiplicity of factors including the inherent low yield/productivity of microbial SGLs, costly raw nutritive materials, expensive biosafety containment measures when using pathogenic SGLs producers, expensive/laborious foam control and expensive downstream processing and purification. Futhermore, SGLs are in many cases naturally produced as mixture of homologues/congeners that are difficult to separate; this makes the study and attribution of a specific activity to a specific SGL homologue/congener unattainable. Lastly, there is accumulating evidence that SGLs biosynthesis is tightly regulated in native producers e.g. rhamnolipids production in Pseudomonas aeruginosa [226] and sophorolipids production in Starmerella bombicola [222]. These tight genetic and metabolic regulations possibly explain the limited improvement in SGL yields using simple optimization media components and process conditions in native SGL producers. One should not be misled, however, by the extraordinarily high GL yields reported in literature that are obtained through media optimization, particularly for rhamnolipids [231]. Such reports are questionable due to different quantification methods used that vary in their specificity and/or sensitivity. Standardized protocols for SGLs quantification were made recently available [232], [233] and are expected to profoundly minimize discrepancies in quantification values in glycolipid research.

Although their cost-effectiveness is still unclear, chemical synthesis of GL could overcome many of the problems of SGLs production. Nonetheless, chemical synthesis of SGLs is confronted also by many other limitations and concerns. First, the difficult stereoselective synthesis of glycolipids which are mostly chiral molecules. An attempt to chemically synthesize mono-rhamnolipid, that is naturally produced as α-l-rhamnopyranosyl-R-β-hydroxydecanoyl-R-β-hydroxydecanoate, resulted in the inevitable co-production of three other diastereomers with different configurations of the β-hydroxyl groups (R,R; R,S: S,S; S,R) and different surface activities [234], [235], [236]. Second, certain ecological/health issues are associated with synthetic approaches that most probably involve the use of non-sustainable petrochemical substrates and generate toxic waste products [237]. Thirdly, the biodegradability and toxicity issues of co-produced new-to-nature SGLs diastereomers require attention and investigation.

Genetic engineering and synthetic biology could offer promising ecological solutions to current challenges facing SGLs production, particularly after recent advances in metabolic engineering and tools for cloning and heterologous expression of large biosynthetic pathways. The following sections discuss some of the metabolic engineering strategies for SGLs production.

Engineering heterotrophic carbon source utilization

Raw nutritive materials accounts for approximately more than 85% of the total estimated production/operation costs of SGLs [21]. A wide range of low-cost renewable raw materials were suggested for SGLs production [238], [239], yet, the capacity of GL producers to utilize these raw materials should be investigated or genetically engineered in the selected production host. A successful example of the latter is the engineering of P. aeruginosa strain to utilize whey waste for RLs production via heterologous expression of E. coli lac genes [240]. Likewise, bacterial and fungal GL producers could be engineered to utilize cheap waste lignocellulosic wastes [241], [242], [243]. Although enhancing the utilization of waste oils by expression of lipid esterases seems a good strategy given the low cost and high GL yields [244], [245] associated with these oily carbon sources, these carbon sources are, however, cumbersome during recovery of glycolipids as they necessitate extra steps for their removal adding to the net cost of glycolipids recovery [246].

Heterologous expression of GL biosynthetic pathway

Containment of biosafety level 2 organisms contribute remarkably in the operational costs of simple glycolipids (SGLs) production. Moreover, working with pathogenic or opportunistic pathogens presents a health risk to manufacturing personnel as well as to public and environment. Heterologous expression of GL biosynthetic genes in hosts that are Generally Recognized As Safe (GRAS) is, therefore, a promising solution as it would require a less costly biosafety level 1 manufacturing facility.

Heterologous expression of rhamnolipids (RLs) in non-pathogenic hosts has received much attention because of the large commercial potential of RLs and because the main and best RLs producer is the opportunistic pathogenic bacterium Pseudomonas aeruginosa [247]. One example is the successful expression of rhamnolipids biosynthetic genes of P. aeruginosa in non-pathogenic bacteria, namely P. putida [246] and P. fluorescens [248] as well as in E. coli [249], [250], the best of which was recombinant P. putida [246], though, all recombinant strains produced RLs at much lower yields than the native producer. Interestingly, the non-pathogenic strain P. chlororaphis is naturally producing mono-rhamnolipids and not di-rhamnolipids as it lacks the gene coding for the second rhamnosyltransferase, rhlC [251]. Heterologous expression of rhlC from P. aeruginosa in P. chlororaphis resulted in production of di-RL at concentration more than twice that of mono-RL [251]. Burkholderia kururiensis is another nonpathogenic heterologous host that was successfully engineered for RLs production [252]. Further engineering strategies were reviewed for overproduction of RL [253].

Several glycosylated carotenoids (C40 and C50), e.g. glucosides of decaprenoxanthin and sarcinaxanthin as well as zeaxanthin, were successfully engineered in Corynebacterium glutamicum by overexpression of genes coding for lycopene cyclization, hydroxylation and glycosylation [254].

Interestingly, 17 genes were heterologously expressed in an E. coli strain that is already producing 6-deoxyerythronolide B precursor to produce the glycosylated macrolide, erythromycin C [255]. The cloned genes encoded the deoxysugar, desosamine, biosynthetic enzymes and the enzymes converting 6-deoxyerythronolide B to erythromycin C [255], [256].

Selection criteria for candidate hosts for heterologous glycolipid production should include, in addition to being non-pathogenic, to be natively tolerant to high concentrations of the target SGLs if high productivities are sought [246]. This is particularly important for SGLs which mostly demonstrate antimicrobial activities.

Moreover, the candidate host should, preferably, abundantly produce the precursors required for SGLs biosynthesis. A good approach would be starting with analysis of the intracellular concentration of lipid and glycosyl precursors. One example is the evaluation of the R-specific enoyl-CoA hydratase-2 (ECH-2) activity in crude cell lysate of target host organism as this predicts the potential of this host to synthesize R-3-hydroxyalkanoate precursors [257] that form the lipid part of many R-3-hydroxyfatty acid-containing glycolipids, e.g. rhamnolipids and rubiwettins. The ECH-2 activity was recently reported to be significantly implicated in rhamnolipids biosynthesis [258].

Oleaginous yeasts, like Yarrowia lipolytica and Rhodosporidium toruloides, are potential candidates in view of their already high lipid flux [259], [260]; therefore, they are supposed to have abundant lipid precursors for GLs biosynthesis.

Blocking competing pathways

Blocking competitive pathways is an important strategy that is expected to enhance GLs biosynthesis. One example is blocking polyhydroxyalkanoates (PHA) synthesis in P. aeruginosa that changed the distribution of produced rhamnolipids (RLs) congeners by doubling the amount of produced mono-rhamnolipids relative to di-rhamnolipids [258]. Also, PHA mutant of P. putida was used to enhance heterologous production of RLs [246].

Tailoring the GL pool composition

Most simple glycolipids (SGLs) are naturally produced in mixtures of congeners and homologues like rhamnolipids [29] and sophorolipids [222]. For studying the functions and properties of each GL species in these natural mixtures, engineered production of purified GL should be sought. This can be achieved by selectively knocking out the genes coding for biosynthesis of specific congeners or forms. A recent example is the production of sophorolipids pool enriched to 88% in the acidic non-lactonized free form by using a mutant strain of Starmerella bombicola [261] that is defective in lactone esterase [262].

Summary and perspectives

More than 5 decades of glycolipids research has led to the discovery of a huge number of simple microbial glycolipids, around 140 of which are cited in this review. In spite of the many publications demonstrating their great biomedical potential, the majority of discovered simple glycolipids are still unable to translate into commercial products because of their high cost of production mainly stemming from low biological yields. Metabolic engineering has the potential to overcome this cost problem particularly after the revolutionary developments in genetic engineering and synthetic biology techniques that were witnessed in the last 5 years.

This review is an attempt to structure the literature available on simple glycolipids aiming at providing metabolic engineers with an outlook on glycolipids and their biosynthesis. It highlights some of aspects and details that are still missing in the biosynthesis, transport and catabolism of glycolipids that need to be pursued and applied profitably in engineering cost-effective microbial glycolipid producers.