Efforts toward Ambergris Biosynthesis

Abstract

Ambergris is a very rare and highly valued fauna natural perfume. Its main component, ambrein, undergoes oxidative degradation to produce ambroxide, forming the unique ambergris fragrance. To meet the market demand while not offending the law of protecting sperm whales, ambrein and ambroxide are chemically synthesized. Recently, the biosynthesis of these compounds has been explored as a green and sustainable production route to ensure the safety of use. The ambrein biosynthesis pathway has been successfully constructed in model microorganisms, leading to de novo biosynthesis of ambrein from glucose and glycerol. In addition, partial biosynthesis of ambroxide has been achieved by modular co-culture of engineered sclareol-producing yeast and a natural fungus converting sclareol to ambradiol, which can be further converted to ambroxide by zeolite. Alternatively, ambroxide can be produced by the chemical transformation of biosynthesized farnesene, followed by enzymatic cyclization. In this paper, the efforts toward biosynthesis of ambrein and ambroxide as representative compounds to substitute the natural ambergris are reviewed, and the challenges and prospects are discussed.

INTRODUCTION

Ambergris is a waxy substance, prized for over a thousand years as medicine, condiment, and perfume. The biological origin of ambergris was claimed in the 19th century, and it is now generally accepted that it is a pathological gut secretion of sperm whales (Physeter microcephalus) arising from undigested horny squid beaks. After the death of the sperm whale, ambergris floats on the sea, and its color becomes lighter while its odor improves under the action of air and sun.¹ The smell of ambergris is described as well-balanced with soft musky, animalic, marine, sweet, earthy, and woody facets.²

Ambergris is a very rare natural product, because it only occurs in about one in a hundred sperm whales, and the ban on commercial whaling by many countries has increasingly limited its sources.³ Studies have shown that the main constituents of ambergris are ambrein, epi-coprosterol, coprosterol, cholesterol, and coprostane-3-one.¹ Ambrein (C₃₀H₅₂O) was first isolated from ambergris in 1820.⁴ Its chemical structure is a tricyclic triterpene alcohol containing one tertiary hydroxyl group and two double bonds. As a major constituent of ambergris, (+)-ambrein was used as a perfume additive for the tenacity or persistence of its odor, and has been reported to possess anti-inflammatory, anti-nociceptive, and aphrodisiac activities.⁵ The first total synthesis of a diastereomeric mixture of ambrein was achieved using farnesylacetic acid and 1-bromomethyl-3,3-dimethyl-1-cyclohexene as raw materials.⁶ Recrystallization of the mixture by seeding with natural ambrein gives (+)-ambrein. In the same year, the total synthesis of (+)-ambrein was accomplished in 0.38% overall yield from geranylacetone through 22 steps or in 1.1% yield from (S)-3-hydroxy-2,2-dimethylcyclohexanone through 21 steps.⁷ (+)-Drimane-8,11-diol has also been used for the synthesis of enantiomerically pure (+)-ambrein.⁸

Ambrein is odorless, while its degradation generates a complex mixture of products, among which (−)-ambroxide is recognized as the prototype of all ambergris odorants. (−)-Ambroxide shows a characteristic and memorable ambery and woody smell historically linked to ambergris, and reminiscent of tobacco and clary sage.⁹ The yearly use of (−)-ambroxide as the most appreciated substitute of ambergris is estimated to be more than 100 t.² (−)-Ambroxide is commercially available from Firmenich SA or Kao Kabushiki Kaisha as Ambroxor Ambroxan and is produced by Givaudan SA as Ambrofix.⁹ The chemical synthesis route of (−)-ambroxide from diterpenoid sclareol was first proposed in 1950.¹⁰ Other natural monoterpenes, sesquiterpenes and diterpenes, e.g., (+)-carvone,¹¹ β-ionone,¹² nerolidol,¹² β-farnesene,¹³ abietic acid¹⁴ and labdanolic acid,¹⁵ were also used for semi-synthesis of (−)-ambroxide. Among all these starting materials, sclareol is the most promising due to its structural similarity to (−)-ambroxide, both having a labdane carbon skeleton. This production route has been industrialized for a long time and is composed of four steps: extraction of (−)-sclareol from the plant Salvia sclarea, oxidation of the (−)-sclareol side chain to yield (+)-sclareolide, reduction of (+)-sclareolide to (−)-ambradiol, and finally cyclodehydration of (−)-ambradiol to give (−)-ambroxide. For this semi-synthesis process, research has been mostly focused on the development of appropriate oxidants and reducing agents.¹⁶⁻¹⁸ However, other challenges remain, for example, the long growth cycle of the plant and its susceptibility to climate changes, the environmental issue associated with the chemical waste, and the chemical residues from the oxidoreduction reactions which would affect the aroma characteristics and the safety of (−)-ambroxide.¹⁹

In recent years, biosynthesis of valuable compounds has attracted increasing attention due to its advantages of short production cycles, mild and green production conditions, and the specificity of enzyme-catalyzed reactions. Efforts have been made toward biosynthesis of (+)-ambrein and (−)-ambroxide, paving the way for green and sustainable production of ambergris. This paper aims to provide an overview of the recent research progress in the biological production of these valuable chemicals by describing the different synthetic pathways (Figure 1), summarizing the advances in each synthetic pathway, and discussing the remaining challenges and future perspectives.

Figure 1.

Synthetic pathways of ambrein and ambroxide. ① Engineering of enzymes for converting biosynthesized squalene to (+)-ambrein; ② ambroxide biosynthesis based on sclareol bioconversion; ③ biosynthetic pathway based on (E,E)-homofarnesol. The different biosynthetic routes are marked by orange, green, and blue, respectively. The semi-synthesis route is marked in gray. The pathways from sclareol to sclareolide and ambradiol catalyzed by ATCC 20918 and ATCC 20624, respectively, remain unclear. The chemical or biological catalysts are marked in bold. FPP, farnesyl pyrophosphate; GGPP, geranylgeranyl diphosphate; LDPP, (13E)-8α-hydroxylabden-15-yl diphosphate; AaSHC^D337C, the D337C variant of squalene-hopene cyclase from Alicyclobacillus acidocaldarius; BmeTC, tetraprenyl-β-curcumene cyclase from Bacillus megaterium; SsLPS, LDPP synthase from Salvia sclarea; SsScS, sclareol synthase from S. sclarea.

Biosynthesis of Ambrein

Engineering of Enzymes for Converting Squalene to (+)-Ambrein

Biocatalytic synthesis of (+)-ambrein was first achieved using squalene as the substrate (Figure 2), under the cascade catalysis of the D377C mutant of squalene-hopene cyclase from Alicyclobacillus acidocaldarius (AaSHC) and the tetraprenyl-β-curcumene cyclase from Bacillus megaterium (BmeTC).²⁰ By substitution of the last aspartic acid in the DXDD motif with cysteine, the product profile of AaSHC was altered. Squalene was converted by AaSHC^D377C to the monocyclic 3-deoxyachilleol A instead of the pentacyclic hopene,²¹ and 3-deoxyachilleol A was then further converted to (+)-ambrein by BmeTC. Based on this, de novo biosynthesis of (+)-ambrein was realized for the first time by introducing AaSHC^D377C and BmeTC into a squalene-producing Escherichia coli strain which was constructed by expressing the ScERG9 gene encoding the squalene synthase from Saccharomyces cerevisiae. The final strain produced 2.6 mg·L^–1 (the mass of product per liter of culture) of (+)-ambrein from glycerol.²² However, this yield is too low to meet the needs of industrial production.

Figure 2.

Routes for the biocatalytic conversion of squalene to ambrein. Route 1: Squalene is converted by AaSHC^D377C to the monocyclic 3-deoxyachilleol A and is then further converted to ambrein by BmeTC. BmeTC could also utilize squalene as a substrate to convert it to three byproducts, including 8α-hydroxypolypoda-13,17,21-triene, onoceranoxide, and 14β-hydroxyonocera-8(26)-ene. Route 2: The BmeTC^D373C mutant catalyzes the reactions from squalene to ambrein; meanwhile, 8α-hydroxypolypoda-13,17,21-triene is formed as a byproduct. Route 3: The double mutant BmeTC^Y167A/D373C produces ambrein from squalene via 8α-hydroxypolypoda-13,17,21-triene, forming 3-deoxyachilleol A, onoceranoxide, and 14β-hydroxyonocera-8(26)-ene as byproducts. Route 4: A dual enzyme system composed of wild-type BmeTC and BmeTC^Y167A/D373C produces ambrein from squalene via 8α-hydroxypolypoda-13,17,21-triene and 8α-hydroxypolypoda-13,17,21-triene, with onoceranoxide and 14β-hydroxyonocera-8(26)-ene as byproducts.

Yeast strains can produce squalene or 2,3-oxidized squalene through the intrinsic mevalonate (MVA) pathway and sterol biosynthesis pathway. In Pichia pastoris strain engineered by increasing intracellular squalene level and introducing AaSHC^D377C and BmeTC, 2 mg·L^–1 of (+)-ambrein was produced.²³ Meanwhile, large amounts of the intermediate 3-deoxyachilleol A and the byproduct 8α-hydroxypolypoda-13,17,21-triene were also detected, which were generated from squalene by the bifunctional cyclase BmeTC. To further increase the production of (+)-ambrein, BmeTC was engineered by mutating the last aspartic acid in the DXDD motif to cysteine.²³ Surprisingly, the BmeTC^D373C mutant efficiently catalyzed the whole reaction from squalene to (+)-ambrein at a titer of 15 mg·L^–1 [0.8 mg·g^–1 dry cell weight (DCW)] in shake flasks and 105 mg·L^–1 in fed-batch fermentation.²³ When S. cerevisiae was used as the chassis to produce (+)-ambrein by the above strategy, a (+)-ambrein titer of 2.9 mg·L^–1 (0.7 mg·g^–1 DCW) was achieved in shake flasks.²⁴ The difference in (+)-ambrein titers between these two yeast chasses was mainly derived from cell density, with approximately 5-fold higher OD₆₀₀ for P. pastoris than that for S. cerevisiae.

To further improve the reaction efficiency of mutant BmeTC^D373C, alanine screening was conducted for the six residues presumed to be located near the intermediates during the reaction based on the modeling structure of BmeTC. The resulting double mutant BmeTC^Y167A/D373C produced (+)-ambrein from squalene via 8α-hydroxypolypoda-13,17,21-triene with a yield of 21.5%, approximately 10-fold higher than that of BmeTC^D373C which produced (+)-ambrein from squalene via 3-deoxyachilleol A (2.2%).²⁵ However, 3-deoxyachilleol A was accumulated as a byproduct during the biocatalysis of BmeTC^Y167A/D373C. To accelerate squalene conversion to 8α-hydroxypolypoda-13,17,21-triene and meanwhile to reduce 3-deoxyachilleol A accumulation, a dual enzyme system was constructed by coupling the wild-type BmeTC and BmeTC^Y167A/D373C (Figure 2), showing 20 times higher yield than using BmeTC^D373C alone (46.0% vs. 2.2%).²⁵ The authors anticipated the replacement of BmeTC^D373C with this dual enzyme system in the above-mentioned engineered P. pastoris strain may lead to an (+)-ambrein titer of 2 g·L^–1.²⁵ However, the performance of the dual enzyme system was evaluated using purified enzymes in vitro, and a large amount of the intermediates 3-deoxyachilleol A and 8α-hydroxypolypoda-13,17,21-triene as well as the byproducts onoceranoxide and 14β-hydroxyonocera-8(26)-ene were accumulated. How efficient it could be when constructed in P. pastoris remains a question.

Biosynthesis of Squalene as the Precursor of Ambrein

Besides engineering the two key enzymes (AaSHC and BmeTC) in the (+)-ambrein biosynthesis pathway, strengthening the supply of squalene as the key precursor is another premise for efficient ambrein biosynthesis. Squalene is a triterpenic hydrocarbon synthesized in all types of cells as a key intermediate in the formation of eukaryotic sterols and bacterial hopanoids. Currently, the primary commercial source of squalene is the liver of deep-sea sharks, which is challenged by rising cost because of diminishing supply and concerns about environmental sustainability.²⁶ Due to the high value of squalene, many efforts have been made to construct efficient squalene cell factories for low-cost and sustainable production.

Squalene is biosynthesized via the head-to-head condensation of two molecules of farnesyl diphosphate (FPP) under the catalysis of squalene synthase. In E. coli, the precursor FPP is produced through the methylerythritol 4-phosphate (MEP) pathway, where 1-deoxyxylulose-5-phosphate synthase (DXS) and isopentenyl diphosphate isomerase (IDI) are known as rate-limiting enzymes.²⁷ The overexpression of E. coli dxs and idi, in the presence of hopA and hopB (encoding squalene synthase) together with hopD (encoding farnesyl diphosphate synthases) from Streptomyces peucetius, elevated squalene yield by 2-fold.²⁸ To establish an efficient E. coli-based system for squalene production, two different squalene synthases in combination with precursor pathways was tested. By overexpressing the endogenous IDI (idi) and farnesyl diphosphate synthase (ispA) genes in the MEP pathway and the endogenous acetyl-CoA acetyltransferase (atoB) together with the heterologous mevalonate (MVA) pathway from S. cerevisiae and the squalene synthase from human or Thermosynechococcus elongatus in E. coli, up to 230 mg·L^–1 (54 mg·g^–1 DCW) or 150 mg·L^–1 (55 mg·g^–1 DCW) of squalene was produced in shake-flask cultures.²⁹ Recently, extension of the cell membrane of a squalene-producing E. coli strain to provide more storage room for this lipophilic product by overexpression of Tsr, a protein induced invagination of the inner membrane to form multilayered structure, resulted in squalene production of 612 mg·L^–1.³⁰

In yeast, squalene is an intrinsic intermediate in the ergosterol pathway and is produced through the MVA pathway. Strengthening the yeast MVA pathway by overexpressing the truncated form of HMG-CoA reductase (tHMG1) as the well-recognized rate-limiting enzyme increased the squalene production by 10-fold.³¹ By upregulation of the structural genes, addition of the squalene monooxygenase inhibitor, and downregulation of the ethanol production pathway, a squalene titer of 304.49 mg·L^–1 was achieved.³² Other metabolic engineering strategies include improving NADPH and acetyl-CoA supply, pathway compartmentalization, promoter engineering, and optimization of fermentation conditions, as reviewed recently by Paramasivan and Mutturi.³³ At present, the highest squalene titer in yeast is 21.1 g·L^–1 (437.1 mg·g^–1 DCW) in a two-stage fermentation process, achieved by a combinatorial strategy of cytoplasmic and mitochondrial engineering to alleviate the metabolic burden and improve cell growth.³⁴

In recent years, Thraustochytriaceae has emerged as another promising sustainable source of squalene due to their fast growth and high squalene contents (10–317.74 mg·g^–1 DCW).³⁵ However, genetic manipulation of the native biochemical pathways in these marine protists for further improvement of squalene production remains challenging.

The high production of squalene in the model microorganisms implies the potential of (+)-ambrein bioproduction from squalene. By screening novel enzymes capable of efficiently converting squalene to (+)-ambrein and enzyme modification to improve their activities, the (+)-ambrein titer may be enhanced to the g·L^–1 level or even higher.

Biosynthesis of Ambroxide

As a substitute for natural ambergris, ambroxide has received extensive attention. Recently, two major approaches were proposed to achieve green and sustainable production of ambroxide using sclareol and (E,E)-homofarnesol as the key precursors, respectively (Figure 1).

Ambroxide Biosynthesis Based on Sclareol Bioconversion

Some natural microbial strains have been reported to metabolize sclareol to ambroxide precursors, although the metabolic pathways remain unknown (Table 1). Cryptococcus albidus ATCC 20918 and Filobasidium magnum JD1025 can transform sclareol to sclareolide, while Bensingtonia ciliata ATCC 20919 and Hyphozyma roseonigra ATCC 20624 can convert sclareol to ambradiol.³⁶⁻³⁸ This lays the foundation for development of an ambroxide production route by microbial conversion of sclareol to replace the current semi-synthesis route.

Table 1. Microorganisms Naturally Metabolizing Sclareol.

Organism	Product	Description	Reference
Cryptococcus albidus ATCC 20918	sclareolide	60 g·L–1 sclareol was converted to 43 g·L–1 sclareolide within 5 d in fermenter	(36)
Filobasidium magnum JD1025	sclareolide	30 g·L–1 of sclareol was converted to 21.62 g·L–1 sclareolide within 72 h in flasks	(37)
Bensingtonia ciliata ATCC 20919	ambradiol	8.42 g·L–1 sclareol was converted to 5.35 g·L–1 ambradiol within 15 d in fermentor	(36)
Hyphozyma roseonigra ATCC 20624	ambradiol	50 g·L–1 of sclareol was converted to 35.5 g·L–1 ambradiol with within 13 d in flasks	(38)
		Proteomics was performed to preliminarily elucidate the molecular mechanism of sclareol metabolism	(39)
		Metabolite profiling was performed to analyze the biocatalytic conversion of sclareol to ambradiol by ATCC 20624	(40,41)
		The dynamic time course of sclareol biotransformation was explored by ATCC 20624 resting cell assays and several intermediates produced during biotransformation were identified	(42)

Among the above natural producers, H. roseonigra ATCC 20624 has been the most extensively studied. Proteomics of the cells cultivated using sclareol as the sole carbon source, compared to those cultivated without sclareol, revealed up-regulation of 18 aldehyde dehydrogenases. This result suggested possible roles of aldehyde dehydrogenases in the metabolism of sclareol,³⁹ in accordance with the semi-synthesis route, which involves chemical oxidation and reduction reactions. Metabolite profiling showed that sclareolide may be an intermediate during the conversion of sclareol to ambradiol.^40,41 However, a recent study explored the dynamic time course of sclareol biotransformation by resting cells of H. roseonigra and proposed a different metabolic pathway where ambradiol and sclareolide are not interconverted. The authors suggested that ambradiol and sclareolide may derive from different metabolic pathways. In addition, several putative intermediates featured with the same labdane carbon backbone, including isomerization and oxidation in the side chain, were found.⁴² However, the metabolic pathway has still not been fully elucidated, and the genes responsible for this process remain unrevealed. The lack of knowledge on sclareol metabolism hinders the reconstruction of the ambradiol formation pathway in a model microorganism. In contrast, the well-studied sclareol synthetic pathway enabled the construction of sclareol-producing microbial cell factories (Table 2), which provides the possibility of changing the sclareol supply method from plant extraction to fermentation.

Table 2. Metabolic Engineering of Microorganisms for Sclareol Biosynthesis.

Organism	Yield	Description	Fermentation type	Reference
E. coli	1.5 g·L–1	overexpression of CrtE from Pantoea agglomerans, atoB from E. coli, mvaA, and mvaS from Staphylococcus aureus, mvaK1, mvaK2, mvaD, and fni from Streptococcus neumoniae, ERG20 from S. cerevisiae, SsLPS and SsScS from S. sclarea	fed-batch in 3.7L bioreactor	(43)
S. cerevisiae	8.96 mg·L–1	fusion expression of BTS1 and ERG20 as well as LPPS and TPS, overexpression of tHMG1	shake flask	(48)
S. cerevisiae	408 mg·L–1	BY4741 with fusion expression of BTS1 and ERG20 as well as LPPS and TPS, overexpression of tHMG1, crossed with BY4742 to form diploids	fed-batch in 3L bioreactor	(44)
S. cerevisiae	750 mg·L–1	fusion expression of CcCLS and ERG20 (F96C), overexpression of the catalytic domain of HMG2(CD-HMG2) and SsScS, deletion of rox1, dos2, yer134c, vba5, ynr063w, and ygr259c	shake flask	(45)
S. cerevisiae	357 mg·L–1	genomic integration of tHMG1, ERG20 (F96C), and CrtE03M (encoding the C81T variant of GGPP synthase from Xanthophyllomyces dendrorhous), two copies of SsLPS and SsScS, deletion of rox1, downregulation of ERG9 using PHXT1	shake flask	(46)
S. cerevisiae	11.4 g·L–1	MBP (maltose-binding protein from E. coli) fused with SsLPS and SsScS, overexpression of ERG20 (F96C), tHMG1, and HMGR from Silicibacter pomeroyi, ERG10, LAC1, OYE3, three copies of HMG2(K6R), BTS1-PaGGPPS (from Phomopsis amygdali) fusion protein, downregulation of ERG9 using PHXT1, deletion of rox1, dos2, yer134c, vba5, ynr063w, and ygr259c in the central metabolism engineered strain Y&Z036	fed-batch in 1L bioreactor	(47)

Sclareol is naturally derived from the flowers and leaves of S. sclarea (known as Clary sage). Since the elucidation of the enzymes (13E)-8α-hydroxylabden-15-yl diphosphate (LDPP) synthase (SsLPS) and sclareol synthase (SsScS) responsible for sclareol biosynthesis in 2012,⁴³ heterologous sclareol production has been realized in E. coli and S. cerevisiae. By introducing the exogenous MVA pathway and sclareol synthetic genes into E. coli, 1.5 g·L^–1 of sclareol was produced in two-phase high-density fermentation with dodecane as the organic phase.⁴³ More efforts have been made to construct a sclareol-producing yeast. By strengthening the precursor supply and fusion expression of BTS1 encoding geranylgeranyl diphosphate (GGPP) synthase and ERG20 encoding FPP synthase as well as LPPS encoding SsLPS and TPS encoding SsScS for substrate channeling, 408 mg·L^–1 of sclareol was produced by the diploid yeast in fed-batch fermentation.⁴⁴ Episomal expression of Cistus creticus 8-hydroxy copalyl diphosphate synthase (CcCLS) and SsScS in S. cerevisiae together with the deletion of six endogenous genes (rox1, dos2, yer134c, vba5, ynr063w and ygr259c) resulted in sclareol production of up to 750 mg·L^–1 in shake-flask cultivation.⁴⁵ Genomic integration of SsLPS and SsScS together with enhancing the precursor supply, weakening the competitive pathway and balancing the metabolic flux, led to the production of 357 mg·L^–1 sclareol.⁴⁶ Recently, a modular approach, where the global metabolism was divided into three modules: central metabolism for the supply of acetyl-CoA, isoprenoid biosynthesis pathway, and regulation factor modules, was used to improve sclareol production in S. cerevisiae, resulting in the highest ever reported sclareol production of 11.4 g·L^–1.⁴⁷

The progress in sclareol biosynthesis provides the possibility for fermentative production of ambroxide using a mixed culture of a sclareol-producing engineered strain and a sclareol-converting natural microorganism. By dividing the biosynthetic pathway of ambradiol into a sclareol-producing engineered yeast as the upstream pathway module and the sclareol-converting H. roseonigra ATCC 20624 strain as the downstream pathway module, a modular co-culture system was established to enable ambradiol biosynthesis (Figure 3). Subsequent optimization of inoculation ratio, carbon source supplementation, and sclareol secretion promotion led to production of 493.1 mg·L^–1 of ambradiol in an induction-free co-culture system.⁴⁶ This is the first example of the complete biosynthesis of ambradiol from glucose, which can be easily converted to ambroxide by a cyclodehydration process,^49,50 leading to a new green route of (−)-ambroxide production.

Figure 3.

Flowchart of the artificial microbial consortium. The solid-line arrows represent biological reactions, and the dotted-line arrow represents the non-biological cyclodehydration reaction.

Biosynthetic Pathway Based on (E,E)-Homofarnesol

Enzymatic synthesis of (−)-ambroxide from homofarnesol is another attractive approach, where a SHC cyclizes the C16 terpene (E,E)-homofarnesol to (−)-ambroxide in a single step.⁴⁹ In 1986, it was first reported that AaSHC could yield (−)-ambroxide from (E,E)-homofarnesol by enzymatic cyclization. However, the conversion rate was very low (3%).⁴⁹ Later, a novel SHC from Zymomonas mobilis (ZmoSHC1) was found, which exhibited over 50-fold higher conversion of homofarnesol in comparison to squalene conversion.⁵⁰ The highest conversion of the enzyme toward homofarnesol (10 mM) reached approximately 40%,⁵⁰ which is, however, still too low to meet the industrial requirement. By random mutagenesis of AaSHC, the best variant 215G2 (M132R/A224V/I432T) was obtained, which allowed the conversion of 125 g·L^–1 (E,E)-homofarnesol with a cells-to-substrate ratio of 2:1 in 72 h.⁹ The volumetric productivity increased when the cells-to-substrate ratio was changed to 1:1, fully converting up to 300 g·L^–1 (E,E)-homofarnesol in 3 days.² Based on this, Givaudan has developed a cost-effective process for the production of Ambrofix^TM (Figure 4).²

Figure 4.

Production route of Ambrofix^TM by Givaudan. Byproduct 1 is (3aS,5aS,9aS,9bS)-3a,6,6,9a-tetramethyldodecahydronaphtho[2,1-b]furan, byproduct 2 is (7aS,11aS,Z)-5,8,8,11a-tetramethyl-2,3,6,7,7a,8,9,10,11,11a-decahydrobenzo[b]oxonine, and byproduct 3 is (−)-9b-epi-ambrox.

For massive preparation of the key intermediate (E,E)-homofarnesol, biobased (E)-β-farnesene commercially produced by fermentation can serve as a sustainable feedstock. (E)-β-Farnesene can be chemically converted to (E,E)-homofarnesol in two steps by selective cyclopropanation at the terminal double bond with in situ generated diazomethane followed by rearrangement under carbocationic conditions, leading to hydrolysis into (E,E)-homofarnesol with good 3,4-E,Z-ratios (around 80:20).²

CONCLUSION AND FUTURE PERSPECTIVES

For thousands of years, the only source of ambergris was the rare and mysterious secretion of sperm whales. Since the last century, the chemical synthesis of ambrein and ambroxide has largely expanded the application of ambergris, which, however, raises environmental issues as well as safety risk and unfavorable aromatic characteristics due to impurities. In recent years, the rapid development of biotechnology has enabled the development of biosynthetic routes for these ambergris-related compounds.

Using squalene as the key intermediate, de novo biosynthesis of ambrein has been realized in E. coli, S. cerevisiae, and P. pastoris cell factories by introducing and engineering of TC and SHC. However, the highest ambrein production of only 105 mg·L^–1 is still too low to meet the industrial requirement. Considering that the highest fermentative production of precursor squalene reached 21.1 g·L^–1, the low yield of ambrein should be caused by the insufficient catalytic efficiency of the enzymes converting squalene to ambrein. Screening and engineering of TC and SHC for improved substrate specificity and activity, followed by introduction of their encoding genes into high-squalene producing strains as well as balancing the metabolic flux, is expected to further increase ambrein production.

As the prototype of all ambergris odorants, ambroxide is naturally derived from ambrein degradation via complex reactions in the presence of air and sun. To date, de novo biosynthesis of ambroxide has not been accomplished, although two partial bioproduction routes have been developed, using sclareol and (E,E)-homofarnesol as key intermediates, respectively. In the former route, plant-extracted or fermentatively produced sclareol could be converted to ambradiol by a natural fungus, followed by chemical cyclodehydration to yield ambroxide using catalysts, such as zeolite. If the molecular mechanism of sclareol metabolism in the fungal strain could be revealed using comparative transcriptomics and metabolomics in the presence and absence of sclareol, it would be possible to introduce the encoding genes into previously constructed sclareol-producing cell factories for de novo biosynthesis of ambradiol in a fermentative process. In the latter route, the bio-based farnesene should be converted to (E,E)-homofarnesol via a chemical process, which could then be catalyzed to ambroxide enzymatically. This route is efficient, sustainable, and environmentally friendly. However, if not carefully separated, the mixtures of E/Z homofarnesol isomers generated during the chemical process would give complex mixtures of cyclization products that harm the product aromatic properties. Future attempts may be focused on improving the ability of SHC to distinguish the geometric isomers of the substrate or developing a green route with better 3,4-E,Z-ratios.

Author Contributions

N.H.: methodology, investigation, writing original draft. H.Y.: funding acquisition; project administration. L.Y.: conceptualization, funding acquisition, project administration, supervision, writing—review and editing.

The authors declare no competing financial interest.