J . Chem. Tech. Biotechnol. 1996,66, 109-120
`
`Review
`
`Biosurfactants : Recent Advances
`Sung-Chyr Lin
`Department of Chemical Engineering, National Chung Hsing University, Taichung, Taiwan
`(Received 3 April 1995; revised version received 17 November 1995; accepted 1 December 1995)
`
`Abstract: Surfactants find applications in a wide variety of industrial processes.
`Biomolecules that are amphiphilic and partition preferentially at interfaces are
`classified as biosurfactants. In terms of surface activity, heat and pH stability,
`many biosurfactants are comparable to synthetic surfactants. Therefore, as the
`environmental compatibility is becoming an increasingly important factor in
`selecting industrial chemicals, the commercialization of biosurfactant is gaining
`much attention. In this paper, the general properties and functions of bio-
`surfactants are introduced. Strategies for development of biosurfactant assay,
`enhanced biosurfactant production, large scale fermentation, and product
`recovery are discussed. Also discussed are recent advances in the genetic engin-
`eering of biosurfactant production. The potential applications of biosurfactants
`in industrial processes and bioremediation are presented. Finally, comments on
`the application of enzymes for the production of surfactants are also made.
`
`Key words : biosurfactant, glycolipid, lipopeptide, rhamnolipid, surfactin, ELISA.
`
`INTRODUCTION
`
`ester sulfonates or sulfates (anionic) and quaternary
`ammonium salts (cationic).
`Surfactants are amphiphilic molecules that tend to par-
`One of the most widely used indexes for evaluating
`tition preferentially at the interface between fluid phases
`surfactant activity is the critical micelle concentration
`(CMC). The CMC is in effect the solubility of a sur-
`of different degrees of polarity and hydrogen bonding
`(such as oil/water or air/water interfaces). The forma-
`factant within an aqueous phase or the minimum sur-
`tion of such an ordered molecular film at the interface
`factant concentration required for reaching the lowest
`lowers the interfacial energy (interfacial tension) and is
`interfacial or surface tension values. At concentrations
`the CMC, amphiphilic molecules associate
`responsible for the unique properties of surfactant mol-
`above
`ecules. In addition to lowering the interfacial tension,
`readily to form supramolecular structures such as micel-
`the molecular layer can also dominate the interfacial
`les, bilayers and vesicles. The interfacial tension between
`rheological behavior and mass transfer. Because of these
`the aqueous and oleic phases changes very little above
`properties, surfactants find applications in an extremely
`the critical micelle concentration because all additional
`wide variety of industrial processes including emulsifica-
`surfactant molecules form micellar structures since the
`tion for emulsion polymerization, foaming for food pro-
`oil/water interface already has a monomolecular layer
`cessing, detergency
`for household and
`industrial
`of amphiphiles. The forces that hold these structures
`cleaning, wetting and phase dispersion for cosmetics
`together include hydrophobic, van der Waals', electro-
`and textiles, or solubilization for agrochemicals. The
`static and hydrogen bonding interactions. Since no
`total sales volume of specialty surfactants in the USA in
`chemical bonds are formed, these structures are fluid-
`1992 was estimated at $1.7 billion, and was expected to
`like and are easily transformed from one state to
`rise at a rate of 3-5% annually.' Examples of com-
`another as conditions such as electrolyte concentration
`mercially available ionic surfactants include fatty acids,
`and temperature are changed.
`109
`J . Chem. Tech. Biotechnol. 0268-2575/96/$09.00 0 1996 SCI. Printed in Great Britain
`
`PETITIONERS
`
`EXHIBIT NO. 1020 Page 1 of 12
`
`

`

`110
`
`S.-C. Lin
`
`Lipids can form micelles (spherical or cylindrical) or
`bilayers based mainly on the area of the hydrophilic
`head group and the chain length of the hydrophobic
`tail. Molecules with small chain lengths and large head
`groups generally form spherical micelles. Those with
`smaller head groups tend to associate into cylindrical
`micelles, while those with long hydrophobic chains form
`bilayers which, in turn, under certain conditions form
`vesicles.' The formation of micelles can result in the
`solubilization of oil or water in the other phase, giving
`rise to a microemulsion. The unique properties of micel-
`les are being explored for applications such as the
`extraction of proteins from fermentation broth and the
`removal of organics and metal ions from aqueous
`streams for environmental application^.^-^
`Another parameter frequently used for predicting sur-
`factant behavior
`is the hydrophilic and lipophilic
`balance (HLB) value. Generally, surfactants with HLB
`values less than 6 are more soluble in the oil phase;
`those with HLB values between 10 and 18 have the
`opposite characteristics.6
`Many biological molecules are amphlphilic and parti-
`tion preferentially at interfaces. Those compounds
`which exhibit particularly high surface activity are clas-
`sified as biosurfactants. The physicochemical properties,
`such as decreases in interfacial tension, heat and pH sta-
`bility, of many biosurfactants have been shown to be
`comparable to synthetic surf act ant^.^ In addition, the
`chemical diversity of naturally produced amphiphiles
`offers a wider selection of surface active agents with
`properties closely tailored
`to specific applications.
`However, biosurfactants have not yet been employed
`extensively in industry because of technical and/or eco-
`nomic reasons. This is beginning to change as environ-
`mental compatibility
`is becoming an
`increasingly
`important factor for the selection of industrial chemi-
`cals. Unlike synthetic surfactants, microbially-produced
`compounds are easily biodegradable and thus particu-
`larly suited for environmental applications such as bio-
`remediation and the dispersion of oil spills.8-'
`This work is intended to be a general review of bio-
`surfactants with recent progress in biosurfactant assay,
`biosurfactant production, and genetic engineering of
`biosurfactant-producing microorganisms. The advan-
`tages and disadvantages of employing microorganisms
`and enzymes for the production of surfactants are also
`compared.
`
`2 PROPERTIES AND FUNCTIONS OF
`BIOSURFACTANTS
`
`Microbial biosurfactants include a wide variety of
`chemical structures, such as glycolipids, lipopeptides,
`polysaccharide-protein complexes, phospholipids, fatty
`acids and neutral lipid^.^ It is, therefore, reasonable to
`expect diverse properties and physiological functions for
`
`different families of biosurfactants. The structures of
`biosurfactants have been extensively r e v i e ~ e d ~ * ~ ~ '
`'-l
`and will not be covered in this review. It is enough to
`mention that most biosurfactants consist of distinct
`hydrophilic and hydrophobic moieties. The former can
`be either ionic or non-ionic and consist of mono-, di-, or
`polysaccharides, carboxylic acids, amino acids, or pep-
`tides. The hydrophobic moieties are usually saturated,
`unsaturated or hydroxylated fatty acids. For some high
`molecular weight surfactant molecules, such as protein-
`polysaccharide complexes, the hydrophobic and hydro-
`philic moieties are contributed by different molecules.
`
`2.1 Properties
`
`A comprehensive list of biosurfactants which reduce the
`surface tension of the fermentation broth to less than
`30 mN m-l and the interfacial tension against n-
`alkanes to values below 1 mN m- has been compiled.'
`The interfacial properties of surfactants depend on the
`ionic composition of the aqueous phase. For example,
`high NaCl concentrations inactivate the glycolipids of
`Torulopsis apicola. On the other hand, the interfacial
`tension of the fermentation broth of Bacillus Iicheni-
`formis JF-2 decreases by more than an order of magni-
`tude in the presence of 10% (w/v) NaCl but is not
`affected by calcium salt^.'^"^ Interestingly, this micro-
`organism grows and produces biosurfactant under both
`aerobic and anaerobic conditions and in the presence of
`up to 8% NaCl."
`The glycolipids produced by Rhodococcus sp. H1319
`and the biosurfactant from Bacillus licheni,formis JF-
`216.20-22 ha
`ve been shown to reduce the surface tension
`of aqueous solutions to 26-27 mN m-l and the inter-
`facial
`tension
`against
`decane
`or octane
`to
`lo-' mN m-'. These values compare favorably with
`those obtained with commercial synthetic surfactants.
`Some biosurfactants also exhibit good thermal and
`chemical stability characteristics. For example, the lipo-
`peptides from B. licheniformis JF-2 are stable at a tem-
`perature up to 75°C for at least 140 h.'7923 The
`biosurfactant is stable at pH values between 5.5 and 12
`but slowly loses activity under more acidic conditions.
`
`2.2 Physiological functions
`
`The physiological functions of biosurfactants are not
`clear. Although most biosurfactants are considered as
`secondary metabolites, some may play essential roles
`for the survival of the producing-microorganisms either
`through facilitating nutrient transport or microbe-host
`interactions, or as biocides. Almost all of these bio-
`logical in-oioo functions are related to the amphipathic
`properties of the biosurfactants.
`It has been suggested that the production of bio-
`surfactants can enhance the emulsification and solu-
`
`PETITIONERS
`
`EXHIBIT NO. 1020 Page 2 of 12
`
`

`

`Biosurfactants
`
`111
`
`amphiphathic properties. Most of the antibiotic bio-
`surfactants, such as rhamnolipids produced by P .
`aeruginosa3’ and surfactin produced by B. s ~ b t i l i s , ~ ~
`function as antibiotics by solubilizing the major com-
`ponents of cell membranes. By releasing antibiotics into
`the culture medium, microorganisms have a better
`chance of survival in an altered environment.
`
`3 BIOSYNTHESIS AND GENETICS OF
`BIOSURFACTANTS
`
`bilization of hydrocarbon substrates, and therefore
`facilitate the growth of microorganisms on hydrocar-
`bons. Considering
`the kinetics of microorganism
`growth, it is easy to understand
`the relationship
`between microbial growth rate and nutrient concentra-
`tions. By secreting biosurfactants into the growth
`medium, microorganisms relying on non-polar sub-
`strates as sole carbon sources ensure the timely supply
`of carbon source to maintain their survival and growth.
`This hypothesis is supported by the fact that some
`hydrocarbon-utilization microorganisms
`produce
`reduced amounts of biosurfactants when grown on
`The biosynthesis and genetics of secondary metabolites,
`water-soluble substrate^.^^ It has been demonstrated
`such as biosurfactants, are generally complex and not as
`that the growth of Pseudomonas aeruginosa on n-alkane
`well characterized as those of proteins for several
`could be accelerated by adding a very small amount of
`reasons, such as the diverse structures of biosurfactants,
`a growth stimulant, a rhamnolipid, into the growth
`the possible involvement of various biosynthetic path-
`medium.” The growth of a P. aeruginosa mutant,
`ways, and the poor understanding of microbial genetics
`which produced a reduced amount of rhamnolipid and
`for industrial microbes except for B. subtilis. Only some
`was deficient in utilizing n-parafin,26 on hydrocarbon
`biosynthetic pathways involved in the synthesis of
`was enhanced significantly with added r h a m n ~ l i p i d . ~ ~
`hydrophobic and hydrophilic domains of biosurfactants
`However, a mutant unable to grow on hexadecane pro-
`and a few developments in the genetics of surfactin pro-
`duced twice as much rhamnolipid as the wide-type
`duction by B. subtilis have been reported.
`strain when grown in glucose-containing media, where
`the emulsification of hydrocarbon is not needed. T
`apicola also produces glycolipids, which do not stimu-
`
`late its own growth on h y d r ~ c a r b o n . ~ ~ The hypothesis
`is also contrasted by the facts that Bacillus subtilis pro-
`duces biosurfactant only with water-soluble substrates28
`and that some mutant microorganisms produce ele-
`vated levels of biosurfactants with water-soluble sub-
`strates.26 Some Streptococcus thermophilus strains were
`shown to produce biosurfactants as anti-adhesives with
`glucose as the main carbon source.29
`Some cell-bound biosurfactants may be responsible
`for hydrocarbon transport and the attachment of the
`cells to interface^.^'-^^ This mechanism is supported by
`the observation that 2.5% fatty acid was isolated in the
`polysaccharide moiety from the cell surface of Candida
`tropicola grown on alkanes, while only a trace amount
`of fatty acid was detected in the corresponding poly-
`saccharide fraction from the cells grown on glucose.
`This indicated that the cell-bound polysaccharide-fatty
`acid complex might be involved in the direct transpor-
`tation of hydrocarbon
`substrates
`into
`the cells.
`Biosurfactants produced by Serratia marcescens pre-
`sumably modulate
`the hydrophobicity of
`the cell
`surface, which appears to be an important factor for cell
` interface^.^^
`adhesion and colonization of various
`Anionic phospholipids are believed to play a critical
`role in the membrane insertion of proteins. It has been
`recently demonstrated that anionic phospholipids might
`be responsible for mediating the membrane insertion of
`protein toxin.33
`Various biosurfactants, mainly lipopeptides and gly-
`colipids, have been shown to have biocidic activ-
`As mentioned above, the biocidic activities of
`i t i e ~ . ~ ~ . ~ ’
`biosurfactants may have a direct connection with their
`
`3.1 Biosynthesis
`
`As may be expected from the wide variety of bio-
`surfactant structures that have been determined so far,7
`their formation involves an equally diverse range of bio-
`synthetic pathways. For simplicity, three classes of path-
`ways can be distinguished depending on whether the
`hydrophobic domain, the hydrophilic domain, or both,
`are synthesized de ~ o u o . ~ ~ Obviously, this classification
`
`does not reflect the many different biosynthetic routes
`that are involved in the formation of the lipid and
`hydrophilic domains. Those components that are not
`synthesized de nouo are produced by modification of the
`carbon source, i.e. sugars, alkanes, etc. Often, a variety
`of different carbon substrates can be incorporated into
`the biosurfactant, giving rise to a family of related mol-
`ecules.
`In lipopeptides such as herbicolin A and surfactin,
`both the lipid and the peptide domains are directly syn-
`thesized from carbohydrates. Addition of amino acids
`or fatty acids in the growth medium can affect the yield
`but not the structure of the product.36 The trehalose
`lipids formed by Rhodococcus erythropolis are typical of
`compounds in which the hydrophilic component, in this
`case the disaccharide trehalose, is not affected by the
`carbon substrate, whereas
`the fatty acid domain
`depends on the chain length of the alkane feed.37 In an
`enzymatic step that is probably rate limiting, the fatty
`acid (corynomycolic acid) is esterified to trehalose-6-
`phosphate which is subsequently subject to dephosp-
`horylation and further modification. Finally, the sur-
`factants produced by Arthrobacter paraffineus represent
`an example of the class of compounds in which the
`
`PETITIONERS
`
`EXHIBIT NO. 1020 Page 3 of 12
`
`

`

`112
`
`S.-C. Lin
`
`hydrophilic (sugar) moiety is influenced by the carbon
`source. Fructose lipids are produced when this micro-
`organism is grown on fructose as the carbon source,
`whereas glucose and sucrose lipids predominate in
`sucrose-grown culture^.^**^^
`The fatty acid components of biosurfactants are syn-
`thesized by the rather well characterized pathways of
`lipid m e t a b o l i ~ m . ~ ~ The hydrophilic moieties, on the
`other hand, exhibit a greater degree of structural com-
`plexity which is the outcome of a wide variety of bio-
`synthetic mechanisms. Recent studies have begun to
`shed light on the formation of the amino acyl part of
`lipopeptide antibiotics, many of which display inter-
`esting surface active characteristic^.^^*^^ Lipopeptides
`are synthesized non-ribosomally by
`large multi-
`functional enzyme complexes exemplified by gramicidin
`S synthetase. The first step in the formation of the deca-
`peptide antibiotic gramicidin S is the activation of
`amino acids via adenylation by ATP. The activated
`intermediates are attached to specific sites on the grami-
`cidin S synthetase complex by thioether linkages. The
`amino acid intermediates are arranged on the enzyme in
`a linear fashion corresponding to the sequence with
`which they will be incorporated into the growing
`peptide. Assembly of the peptide involves a pantetheine
`cofactor having a reactive-SH group. The cofactor
`functions as an internal swinging arm to mediate the
`transport of the growing peptide between the sites of
`attachment of the activated amino acids. This mode of
`synthesis is called the thiotemplate mechanism. The
`antibiotic tyrocidine is formed by a similar process
`except that the tyrocidine synthetase complex consists
`of three rather than two enzymes.
`The synthesis of the surface active lipopeptide sur-
`factin has been investigated in detail. Genetic evidence
`has indicated that two putative components of the
`surfactin-synthesizing enzyme from the Bacillus subtilis
`complex share homology with tyrocidine synthetase 1
`and gramicidin S synthetase l.43 Recent biochemical
`studies demonstrated that surfactin synthesis occurs via
`a thiotemplate-based process. In-vitro, the synthesis of
`surfactin by a cell free system requires ATP, Mg2+ pre-
`cursors and sucrose, the latter presumably because of
`the need to stabilize the enzyme complex. Even though
`the peptide contains D-Leu, only the L-isomer of the
`amino acid can serve as a precursor. The fatty acid
`component is incorporated only as an acetyl-CoA deriv-
`a t i ~ e . ~ ’ . ~ ~
`
`scientific research and commercial development. Meta-
`bolic engineering, i.e. the application of genetic engin-
`eering to improve the synthesis of non-ribosomal
`
`p r o d u ~ t s , 4 ~ - ~ ~ has been exploited effectively in the anti-
`biotics area. So far the only example of metabolic engin-
`eering for biosurfactant production is the expression of
`the lactose utilization genes in Pseudomonas aeruginosa
`to allow growth and rhamnolipid production on lactose
`or cheese whey.”
`DNA transfer systems including shuttle vectors and
`transducing
`phages
`are
`available
`for many
`biosurfactant-producing microorganisms such as Bacil-
`lus and Rhodococcus sp. To successfully implement
`genetic studies for enhanced biosurfactant production,
`suitable plate assays for the screening of mutants is
`indispensable. To this end, direct colony thin-layer
`chromatography and blood agar plate assay, to be dis-
`cussed in the next section, have been developed and suc-
`cessfully used to isolate mutants with the desired
`proper tie^.^'.^' Similar approaches were
`taken
`to
`isolate null mutants of B. subtilis for surfactin pro-
`
`d ~ c t i o n . ~ ~ A Tn5 mutagenized population of P. aerugin-
`osa was screened for defective growth on hydrocarbon
`minimal media and two variants were characterized in
`detail. One strain was found to be defective for bio-
`surfactant production whereas the second exhibited a
`two-fold higher production when grown in minimal
`media with glucose as the carbon source.27
`Recent studies on the genetics of B. subtilis develop-
`ment and surfactin production have shed light on the
`complexities involved in the molecular-level regulation
`of biosurfactant
`which has been
`previously reviewed6 and therefore will not be covered
`in this review. Nevertheless, it is noteworthy that the
`srfA genes, required for surfactin production, have been
`placed under the control of an inducible promoter so
`that the production of surfactin is only dependent on
`the addition of the inducer isopropyl-/3-galactoside
`(IPTG) in the growth medium.43 Other recombinant
`microorganisms with enhanced biosurfactant pro-
`
`duction have also been c o n ~ t r u c t e d . ~ ~ . ~ ~ The construc-
`tion of a recombinant strain, Bacillus subtilis MI113
`with a plasmid-containing gene related to surfactin pro-
`duction from a wild-type surfatin producer, B. subtilis
`RB 14, has been reported. Under optimal conditions, the
`amount of surfactin production was eight times as high
`as that of the wild-type strain.57
`
`3.2 Genetics
`
`4 BIOSURFACTANT ASSAYS
`
`Although the genetic analysis of biosurfactant pro-
`duction is currently at an early stage, the use of recom-
`binant DNA
`techniques for
`the manipulation of
`biosurfactant production is slowly gaining ground and
`could become instrumental in future efforts for both
`
`The development of an effective biosurfactant assay is
`critical to the success in optimizing biosurfactant pro-
`duction by medium optimization and/or fermentation
`technology and in
`the selection of biosurfactant-
`producing microbes and/or their mutants. However, the
`
`PETITIONERS
`
`EXHIBIT NO. 1020 Page 4 of 12
`
`

`

`Biosurfactants
`
`113
`
`lack of common reactive groups or chromophores in
`most biosurfactant molecules has impeded the develop-
`ment of universal chemical or spectral assays for bio-
`surfactants. Recent developments
`in biosurfactant
`assays for medium optimization and strain selection will
`be discussed in this section.
`
`4.1 Biosurfactant assays for optimization studies
`
`The effective biosurfactant assays for optimization
`studies should have the capability of handling large
`amount of samples with relatively good specificity and
`sensitivity. So far, the most widely used methods for the
`detection of biosurfactants have been interfacial/surface
`tension measurements and thin-layer chromatography
`(TLC). However, these methods are inappropriate for
`quantitative studies for their lack of sensitivity and are
`also time-consuming. For example, interfacial/surface
`tensions of cell-free culture against organic phases are
`generally strongly affected by factors such as pH and
`ionic strength, which excludes their utility as a quanti-
`tative assay to investigate the effects of these factors on
`biosurfactant production. Therefore, interfacial/surface
`tension measurements are only good for the preliminary
`screening of biosurfactant-producing microbes.58 On
`the other hand, although TLC can provide reasonably
`good resolution and semi-quantitative information pro-
`vided with appropriate solvent systems, it is not appro-
`priate for analyzing large amount of samples obtained
`in optimization studies, because time-consuming pre-
`purification procedures such as precipitation and
`extraction are necessary for sample preparation. The
`determination of wetting activity by measuring the
`droplet diameter and contact angle have similar prob-
`lems to interfacial/surface tension measurement^.^^*^^
`Therefore, to effectively implement an optimization
`program, a new assay for biosurfactants must be devel-
`oped.
`High performance liquid chromatography (HPLC)
`has been widely used for the detection, quantification
`and purification of biomolecules. However, its applica-
`tion in biosurfactant analysis was not achieved until the
`development of a reverse phase HPLC method with a
`C18 column for the analysis of a biosurfactant produced
`by Bacillus lichenifomis JF-2.61 The assay has the
`advantages of high specificity and sensitivity, and the
`capability of handling large amounts of samples. Fur-
`thermore, the amount of sample required for accurate
`analysis is small, without the need for tedious pre-
`purification. This assay was successfully used
`to
`monitor the unique biosurfactant production profile, to
`be discussed later, which eventually led to the develop-
`ment of a procedure for the continuous production of
`Bacillus licheniformis biosurfactant.62
`The specific interactions between biomolecules, such
`as enzymes and substrates or substrate analogs and
`
`antigens and antibodies, have been extensively utilized
`for the development of novel purification and diagnostic
`techniques, such as affinity chromatography and
`enyzme-linked
`immunosorbent assay (ELISA). An
`ELISA procedure for lipopeptide biosurfactant by
`Bacillus lichengoformis JF-2 was reported recently.22 Such
`a biosurfactant ELISA is extremely specific and sensi-
`tive (at least as low as 0.01 mg dmP3) and capable of
`handling large numbers of samples simultaneously.
`However, this approach may not be applied to other
`types of biosurfactants, because not all biosurfactants
`are immunogenic even upon conjugation with carrier
`proteins such as bovine serum albumin (BSA) or
`keyhole limpet hemocyanin (KLH).
`
`4.2 Biosurfactant assays for strain selection
`
`An efficient screening strategy is the key to the success
`in isolating the desired microbes or their variants,
`because a large number of strains need to be character-
`ized. The ex-situ biosurfactant assays described above
`are not appropriate for this purpose. Fortunately, some
`in-situ techniques utilizing the physiological and chemi-
`cal properties of biosurfactants have been developed.
`A modified version of TLC with bacterial colonies for
`screening biosurfactant-producing variants has been
`reported.51 Instead of spending days for TLC sample
`preparation, this technique involves the direct applica-
`tion of bacterial mass on a TLC plate and pre-
`development of the plate. The plate containing the
`bacterial extracts was subsequently developed following
`the removal of adhering bacterial mass and drying. This
`assay was employed successfully to identify and isolate
`the bacteria variants defective in biosurfactant pro-
`duction. However, this direct colony thin-layer chroma-
`tography may not be applied to microbes producing
`low levels of biosurfactant, because of its low sensitivity.
`Some biosurfactants possess antibiotic activity or
`hemolytic activity probably because of their amphi-
`philic properties. These activities have been utilized for
`the development of in-situ biosurfactant assays. For
`example, surfactin, a lipopeptide biosurfactant produced
`by B. subtilis, can rupture erythrocytes, although it is
`not a hemolytic enzyme per se.34 This hemolytic activity
`has enabled the development of a blood agar assay for
`~ u r f a c t i n . ~ ~ The sizes or diameters of colorless hemo-
`lytic zones around the colonies correspond well to the
`ability of the microbes in biosurfactant production. This
`assay was successfully employed to select for microbes
`capable of producing biosurfactants in media without
`hydrocarbons and later to isolate Bacillus subtilis
`mutants with enhanced biosurfactant prod~ctivity.~’
`This approach can be expanded to develop in-situ
`assays for biosurfactants with antibiotic activity such as
`the
`lipopeptide biosurfactant by Bacillus
`licheni-
`for mi^.^^,^^ However, it is important to keep in mind
`
`PETITIONERS
`
`EXHIBIT NO. 1020 Page 5 of 12
`
`

`

`114
`
`S.-C. Lin
`
`that the complex nature of blood agar makes the blood
`agar assay unsuitable for investigating the effect of
`medium formulation on biosurfactant production and
`that the extent of hemolytic zone formation may be
`affected by divalent ions and other hemolysins produc-
`ed by the microbes under
`Another new semi-quantitative agar plate bio-
`surfactant assay specific for anionic biosurfactants has
`recently been developed.66 The assay was developed
`based on the property that the concentration of anionic
`surfactants in aqueous solutions can be determined by
`the formation of
`insoluble ion pairs with various
`cationic substances. The cationic chemicals selected in
`the assay were cetyltrimethylammonium bromide
`(CTAB) and methylene blue. Under optimal conditions,
`dark blue halos were observed around rhamnolipid-
`producing colonies and the diameters of the halos could
`be directly related to the concentration of rhamnolipid
`produced. The assay was shown to be applicable to
`other anionic glycolipids such as sophoroselipids and
`cellobioselipids. The disadvantages of blood agar assay
`for biosurfactant discussed above were not observed in
`this agar plate assay. However, further modification of
`this assay with other cationic substitutes, such as N -
`cetylpyridinium chloride or benzethonium chloride,
`may be necessary for other biosurfactant-producing
`microorganisms, because CTAB inhibits the growth of
`most bacteria.66
`
`5 PRODUCTION AND RECOVERY OF
`BIOSURFACTANTS
`
`It is very difficult to draw general guidelines for optimal
`biosurfactant production because biosurfactants are a
`diverse group of compounds produced by a variety of
`microbial species. Biosurfactant-producing microbes
`can be divided into three categories:13 those producing
`biosurfactants exclusively with alkanes as the carbon
`sources, such as Corynebacterium sp. and Arthrobacter
`sp.; those producing biosurfactants only with water-
`soluble substrates as the carbon sources, such as Bacil-
`lus sp.; and those producing biosurfactants with alkanes
`and water-soluble substrates as the carbon sources, such
`as Pseudomonas sp. It is important to understand that
`the structures and yields of biosurfactants depend on
`the carbon sources as well as the microbial species
`used.6 Furthermore, the production patterns of bio-
`surfactants by different species might be different.
`Therefore, process development and fermentations must
`be optimized on a case by case basis. As with other
`microbial fermentations the goal in the production of
`biosurfactants is to maximize the productivity (i.e.
`g dm-3 h-'), to increase the yield of biosurfactant from
`the carbon source and to achieve high final concentra-
`tions. In addition it is important to reduce the accumu-
`
`lation of other metabolic products that may interfere
`with the physical properties or the recovery of the
`surface active agent. Like the production and isolation
`of most biomolecules, the optimal recovery strategies
`are frequently dictated by the production patterns as
`well as the physicochemical characteristics of the bio-
`surfactants.
`
`5.1 Medium formulation
`
`The carbon source is known to be critical for the struc-
`tures and yields of biosurfactants. Depending on the
`physiology of the producing microbes, the production
`of biosurfactants can either be induced or inhibited by
`the addition of hydrocarbon^.^^*^^.^^ Furthermore, the
`addition of biosurfactant precursors such as sugars
`might result in the substitution of the hydrophilic moi-
`eties of glycolipids with the added sugar. As mentioned
`earlier, A. parafineus produces biosurfactants with dif-
`ferent hydrophilic moieties when grown on different
`carbon source^.^^,^^ Likewise, the structure of the fatty
`acid domains of biosurfactants can also be dictated by
`the chain length of the alkanes used for some microbes,
`such as the glycolipid by Rhodococcus e r y t h r ~ p o l i s . ~ ~
`Other nutrients that could affect the production of
`biosurfactants include nitrogen sources, phosphates
`source, metal ions and other additives. It has been
`shown that the production of biosurfactants by Pseudo-
`monas sp., Acinetobacter sp. and Torulopsis sp. can be
`regulated by the ratio of nitrogen to carbon source or
`
`concentration of yeast e x t r a ~ t . ~ ~ . ~ ' - ~ ~ The production
`of biosurfactant might also be affected by phosphate
`
`m e t a b ~ l i s m . ~ ~ For example, the production of lipopep-
`tide biosurfactant by Bacillus licheniformis JF-2 can be
`increased from 35 mg dm-3
`to 110 mg dm-3 by
`reducing
`the
`phosphate
`concentration
`from
`100 mmol dm-3 to 50 mmol dm-3.61 The addition of
`iron or manganese salts has also been shown to increase
`the yield of surfactin production by Bacillus subtilis.28
`The effect of multivalent ions on biosurfactant pro-
`duction might be correlated to nitrogen m e t a b o l i ~ m . ~ ~
`The yield of biosurfactant production can be either
`enhanced or inhibited by the addition of antibiotics,
`such as penicillin or chloramphenicol.7s~76 In some
`cases the addition of biosurfactant precursors can
`modify both the structure and yield of bios~rfactants.'~
`Finally, like any other fermentation, the growth tem-
`perature, medium pH and level of dissolved oxygen or
`agitation rate exert an important effect on biosurfactant
`production.'
`
`8 9 7 8 - 7 9
`
`5.2 Fermentation
`
`Most biosurfactants are released into the culture
`medium either at the stationary phase or throughout
`
`PETITIONERS
`
`EXHIB