<<

toxins

Review Basics of Technology

Line Ledsgaard 1, Mogens Kilstrup 1 ID , Aneesh Karatt-Vellatt 2, John McCafferty 2 and Andreas H. Laustsen 1,* ID

1 Department of Biotechnology and Biomedicine, Technical University of Denmark, Kongens Lyngby DK-2800, Denmark; [email protected] (L.L.); [email protected] (M.K.) 2 IONTAS Ltd., Cambridgeshire CB22 3EG, United Kingdom; [email protected] (A.K.-V.); [email protected] (J.M.) * Correspondence: [email protected]; Tel.: +45-2988-1134

 Received: 31 May 2018; Accepted: 8 June 2018; Published: 9 June 2018 

Abstract: Antibody discovery has become increasingly important in almost all areas of modern medicine. Different antibody discovery approaches exist, but one that has gained increasing interest in the field of toxinology and antivenom research is phage display technology. In this review, the lifecycle of the M13 phage and the basics of phage display technology are presented together with important factors influencing the success rates of phage display experiments. Moreover, the pros and cons of different antigen display methods and the use of naïve versus immunized phage display antibody libraries is discussed, and selected examples from the field of antivenom research are highlighted. This review thus provides in-depth knowledge on the principles and use of phage display technology with a special focus on discovery of that target toxins.

Keywords: antibody discovery; recombinant antivenom; phage display; M13 phage; toxinology

Key Contribution: Antibody phage display technology is thoroughly presented and discussed in relation to antivenom research and toxinology.

1. Introduction With the recent inclusion of snakebite envenoming on the World Health Organization’s list of Neglected Tropical Diseases [1], focus on both prevention and treatment of this infliction has increased. This creates renewed hope for snakebite victims worldwide and could potentially lead to a mobilization of scientific efforts toward the development of novel snakebite envenoming . Several different avenues aimed at bringing innovation into the field of snakebite antivenoms have been pursued, including medicinal chemistry approaches, novel immunization techniques, and the use of biotechnological strategies [2–5]. One promising approach seems to be the use of human IgG antibodies [6] and/or camelid antibody fragments [7,8], as these molecules can be used to develop recombinant antivenoms with high efficacy and safety due to their compatibility with the human immune system [2]. Moreover, these therapeutic could be manufactured cost-competitively using modern cultivation methods employed for large scale production [6,9]. To discover and develop antibodies, different techniques can be harnessed. One approach is phage display selection [10], which is a robust, easy-to-perform, and inexpensive method by which specific antigen binders are selected from large combinatorial libraries containing billions of antibody fragments. As antibody phage display is gaining increasing interest in the field of toxinology, the intention with this review is to provide both basic and more advanced knowledge on the underlying science behind the technology and the lifecycle of the M13 phage.

Toxins 2018, 10, 236; doi:10.3390/toxins10060236 www.mdpi.com/journal/toxins Toxins 2018, 10, 236 2 of 15

Toxins 2018, 10, x FOR PEER REVIEW 2 of 15 2. The M13 2. TheCentral M13 to Bacteriophage phage display technology is the biology of the bacteriophage used to display antibodies. DifferentCentral bacteriophage to phage systems display can technology be utilized is for the phage biology display, of the including bacteriophage the T4, used lambda, to displayas well as theantibodies. filamentous Different bacteriophage [11 systems]. These can different be utilized phage for systems phage eachdisplay, have including their benefits the T4, and drawbacks.lambda, as However, well as the primarily filamentous the M13M13 phagebacteriophage has been [11] utilized. These extensively different phage in recent systems times, each and have to the besttheir of ourbenefits knowledge, and drawbacks. this is the However, only phage primarily system the that M13 has phage been exploredhas been withinutilized toxinology extensively [ 2in,12 ]. Therefore,recent times, this reviewand to the will best focus of our on knowledge, antibody phage this is display the only techniques phage system utilizing that has this been specific explored phage system.within Before toxinology giving [2,12] an. in-depth Therefore, description this review of will the steps focus involved on antibody in phage phage display display experiments,techniques anutilizing introduction this specific to the wild-type phage system. M13 phageBefore isgiving provided. an in-depth description of the steps involved in phageThe display M13 phage experiments, [13] belongs an introduction to a group to the of filamentouswild-type M13 phages phage collectivelyis provided. referred to as Ff phagesThe [14]. M13 The phage [13] onlybelongs infect to Escherichia a group of coli filamentousstrains that phages express collect the Fively referred as the adsorption to as Ff ofphages the phage [14] to. The the Ff bacterium phages only requires infect Escherichia binding of coli a strains phage that coat express the to F the pilus tip as of the the adsorption F pilus [15 ]. Theof M13the phage phage to is the neither bacterium temperate requires nor binding lytic. Instead, of a phage the phagecoat protein establishes to the atip chronic of the infectionF pilus [15] in. its The M13 phage is neither temperate nor lytic. Instead, the phage establishes a chronic infection in its host, where it continuously releases new phages. The phage contains a of single-stranded host, where it continuously releases new phages. The phage contains a genome of single-stranded DNA (ssDNA) with a length of 6407 bp [16] that consists of nine encoding 11 different proteins. DNA (ssDNA) with a length of 6407 bp [16] that consists of nine genes encoding 11 different proteins. Five of these proteins are coat proteins, and the remaining six proteins are involved in replication Five of these proteins are coat proteins, and the remaining six proteins are involved in replication and assembly of the phage. The M13 phage has a length of 900 nm and a width of 6.5 nm [17]. and assembly of the phage. The M13 phage has a length of 900 nm and a width of 6.5 nm [17]. The Themost most abundant abundant of the of the coat coat proteins proteins is the is thecapsid protein protein G8P, G8P,which which forms forms an envelope an envelope around around the thechromosome consisting consisting of ofapproximately approximately 2700 2700 protein protein units units (Figure (Figure 1).1 ). The The remaining remaining four four coat coat proteins,proteins, G3P, G3P, G6P, G6P, G7P, G7P, and and G9P, G9P, are are each each present present inin approximatelyapproximately five five copies. copies. Information Information on on the the genesgenes and and proteins proteins of of the the M13 M13 phage phage is is listedlisted inin TableTable1 1..

G 8P G 3P

G 7P G 9P

+ ssD N A G 6P FigureFigure 1. 1.Schematic Schematic representation representation ofof thethe M13M13 bacteriophage, which which is is a afilamentous filamentous phage phage carrying carrying a single-strandeda single-stranded DNA DNA (ssDNA) (ssDNA)chromosome. chromosome. The The genome genome contains contains nine nine genes, genes, which which encode encode 11 11 proteins.proteins. Five Five of of these these proteins proteins are are coat coat proteins proteins (G3P, (G3P, G6P, G6P G7P,, G7P, G8P G8P, and and G9P), G9P), while while the the remaining remainingsix proteinssix proteins are used are used for replication for replication of the of genome,the genome, assembly assembly of theof the phage, phage, and and phage phage extrusion. extrusion.

Table 1. name, protein name, protein size, and the function of the genes carried by the M13 Table 1. Gene name, protein name, protein size, and the function of the genes carried by the M13 phage [16]. phage [16].

GeneGene Name Name ProteinProtein Name Name (Abbreviation) (Abbreviation) Size (kDa)Size (kDa) FunctionFunction GeneGene 1 protein 1 protein (G1P) (G1P) 39.639.6 AssemblyAssembly I I GeneGene 11 protein 11 protein (G11P) (G11P) 12.412.4 AssemblyAssembly Replication-associatedReplication-associated protein protein (G2P) (G2P) 46.246.2 ReplicationReplication II II GeneGene 10 protein 10 protein (G10P) (G10P) 12.712.7 ReplicationReplication Coat protein IIIIII AttachmentAttachment protein protein (G3P) (G3P) 44.744.7 Coat protein Adsorption and extrusion Adsorption and extrusion IV IV VirionVirion export exp proteinort protein (G4P) (G4P) 45.945.9 AssemblyAssembly and and extrusion extrusion V V DNA-bindingDNA-binding protein protein (G5P) (G5P) 9.79.7 ReplicationReplication Coat protein VI VI HeadHead virion virion protein protein (G6P) (G6P) 12.412.4 Coat protein Infection and budding Infection and budding VII Tail virion protein (G7P) 3.6 Coat protein Assembly and budding Coat protein VII Tail virion protein (G7P) 3.6 VIII Capsid protein (G8P) 7.6Assembly Coat protein and budding IXVIII Tail virionCapsid protein protein (G9P) (G8P) 3.77.6 Coat protein AssemblyCoat protein and budding

Toxins 2018, 10, x FOR PEER REVIEW 3 of 15

Toxins 2018, 10, 236 3 of 15 Coat protein IX Tail virion protein (G9P) 3.7 Assembly and budding The first stage of the M13 infection is the adsorption process, which takes place through binding The first stage of the M13 infection is the adsorption process, which takes place through binding of the N2 domain of the G3P coat protein to the tip of a F pilus on the surface of E. coli hosts of the N2 domain of the G3P coat protein to the tip of a F pilus on the surface of E. coli hosts (Figure (Figure2)[ 18–20]. Under normal conditions, “male” E. coli cells (F+, containing F-pili) appear to trawl 2) [18–20]. Under normal conditions, “male” E. coli cells (F+, containing F-pili) appear to trawl the − E. coli theenvironment environment in inan aneffort effort to catch to catch “female” “female” (F−) (FE. coli) partnerspartners by a series by a seriesof F-pilus of F-pilus assembly assembly and anddisassembly disassembly events events [21] [21. Upon]. Upon phage phage binding binding to the to the tip tip of of one one of of these these F Fpili, pili, disasse disassemblymbly will will automaticallyautomatically bring bring the the phage phage closer closer to to the the surface surface of of the the bacterium bacterium [ 18[18]].. Contact Contact between between thethe G3P-N2G3P- domainN2 domain and theand F the pilus F pilus mediates mediates the the allocation allocation of of the the G3P-N1 G3P-N1 domaindomain and and allows allows it itto tobind bind TolA, TolA, whichwhich then then functions functions as as a co-receptora co-receptor on on the the surface surface of of the the bacterium. bacterium. ThreeThree TolTol proteinsproteins areare present in E. coliin E.; TolA,coli; TolA, TolR, TolR, and TolQ,and TolQ, and alland are all essentialare essential for infectionfor infection by by mediating mediating depolymerization depolymerization of of the phagethe phage coat and coat translocation and translocation of the of ssDNA the ssDNA into into the bacteriumthe bacterium [20 ,[20,2222–25–].25].

FigureFigure 2. 2Infection. Infection of ofEscherichia coliby by the the M13 M13 phage. phage. The The phage phage G3P G3P binds binds to to the the tiptip ofof thethe FF piluspilus on E. colion .E. Normal coli. Normal disassembly disassembly of the of F pilusthe F transportspilus transpo therts phage the phage to the to surface the surface of the bacterium,of the bacterium, where it interactswhere withit interacts the TolA with receptor, the TolA mediating receptor, mediating uptake of uptake the phage of the genome. phage genome.

Once the chromosome of the phage has been injected into the bacterium, the host DNA synthesis machineryOnce the will chromosome synthesize of a theDNA phage (−) strand has been complementary injected into theto the bacterium, ssDNA (+) the strand host DNA of the synthesis M13 machineryphage. The will two synthesize strands form a DNA a supercoiled (−) strand double complementary-stranded DNA to the (dsDNA) ssDNA phage(+) strand chromosome, of the M13 phage.termed The the two replicative strands form form (RF) a supercoiled (Figure 3A) double-stranded[26]. Replication of DNA the M13 (dsDNA) chromosome phage always chromosome, uses termedthe RF the and replicative is carried out form by (RF) continuous (Figure rolling3A) [ 26 circle]. Replication replication. of Initiatio the M13n of chromosome replication starts always when uses theG2P RF has and accum is carriedulated out to bya concentration continuous rollingthat allows circle it replication.to nick the (+) Initiation strand in of the replication RF and bind starts whencovalently G2P has to the accumulated 5′ end (Figure to a3B). concentration The 3′ end that that is generated allows it tofrom nick the the nick (+) will strand then be in elongated the RF and bindby covalently DNA polymerase to the 5 using0 end the (Figure (−) strand3B). The as a3 template.0 end that Throughout is generated the from elongation the nick process, will then the be elongatedoriginal byG2P DNA-bound polymerase (+) strand using is physically the (−) displaced strand as aby template. the Rep helicase Throughout [27,28] the (Figure elongation 3C). When process, theone original round G2P-boundof replication (+) is completed, strand is physically the old (+) displacedstrand is cu byt off the at Repthe origin helicase by a [ 27new,28 G2P,] (Figure which3C). Whenremains one bound round to of the replication new 5′ end. is After completed, dissociation the oldfrom (+) the strand RF, the is old cut (+) off strand at the is re origin-circularized, by a new G2P,ready which to be remains converted bound into toan theRF newor to 5be0 end. packaged After into dissociation new M13 fromphages the (Figure RF, the 3D). old In (+) the strand early is re-circularized,stages of infec readytion, all to newly be converted formed ssDNA into an (+) RF or to be packaged will be into converted new M13 into phages RFs, but (Figure in later3D). Ininfection the early stages stages, where of infection, the concentration all newly formed of G5P ssDNA is sufficient (+) chromosomes for fast sequestering will be of converted (+) strands, into formation of RFs will be prevented. When the G5P binds the ssDNA, it dimerizes in a back-to-back RFs, but in later infection stages, where the concentration of G5P is sufficient for fast sequestering conformation leading to the conversion from the circular appearance of the ssDNA to a more rod- of (+) strands, formation of RFs will be prevented. When the G5P binds the ssDNA, it dimerizes in shaped appearance (Figure 3E) [29]. The entire phage chromosome is covered by the G5P except for a back-to-back conformation leading to the conversion from the circular appearance of the ssDNA to an exposed hairpin loop termed the packing signal [30]. The packing signal is required for packing a more rod-shaped appearance (Figure3E) [ 29]. The entire phage chromosome is covered by the G5P except for an exposed hairpin loop termed the packing signal [30]. The packing signal is required for packing of the phage genome. G10P plays an essential but unknown role for the stable accumulation Toxins 2018, 10, 236 4 of 15 Toxins 2018, 10, x FOR PEER REVIEW 4 of 15 of theof ssDNA phage genome. (+) strands. G10P The plays ssDNA an essential bound by but the unknown G5P protein role for is the the substrate stable accumulation for phage assembly of ssDNA (+)(Figure strands.3F). The ssDNA bound by the G5P protein is the substrate for phage assembly (Figure 3F).

+ ssD N A R F

A B R eplication C H ost G 2P proteins

F D 5’ E + ssD N A : G 5P G 2P

FigureFigure 3. 3.InfectionInfection cycle cycle of ofthe the M13 M13 phage. phage. (A (A) Upon) Upon infection, infection, the the single single stranded stranded (+) (+) chromosome chromosome of theof M13 the M13 phage phage is converted is converted into into the the double double-stranded-stranded replicative replicative form (RF)(RF) (B(B)) After After proper proper accumulation,accumulation, the the G2P G2P nicks nicks the the(+) strand (+) strand in the in RF the and RF binds and binds covalently covalently to the to 5′- theend 5 (0C-end) The (C genome) The is thengenome replicated is then replicatedfrom the from3′ end the of 3 0theend nick, of the using nick, the using (−) the strand (−) strand as a template. as a template. The Theoriginal original G2P- boundG2P-bound (+) strand (+) strandis physically is physically displaced displaced by the by Rep the Rephelicase helicase throughout throughout the theelongation elongation process process (D) The(D old) The (+) old strand (+) strand is re-circula is re-circularizedrized by the by bound the bound G2P G2Pafter after dissociation, dissociation, ready ready to be to beconverted converted into an into RF an or RF to or be to packaged be packaged into into new new M13 M13 phages phages (D (D) )Genome Genome replication continues continues until until the the concentrationconcentration of ofG5P G5P has has accumulated accumulated to to sufficient sufficient levels to sequestersequester thethe ssDNA ssDNA (E ()E When) When sufficient sufficient G5PG5P is accumulated, is accumulated, G5Ps G5Ps will will bind bind the the ssDNA ssDNA in a back-to-backback-to-back dimericdimeric conformation, conformation, causing causing the the moremore rod rod-shaped-shaped appearance appearance of of the the ssDNA ssDNA ( (FF)) A pore isis formedformed inin thethe membrane, membrane, and and the the phage phage genome is translocated through this pore, while the phage coat is assembled. genome is translocated through this pore, while the phage coat is assembled.

AssemblyAssembly and and budding budding of of the the M13 M13 phage phage is a five-step five-step process process that that includes: includes: preinitiation, preinitiation, initiation,initiation, elongation, elongation, pretermination, pretermination, and and termination. termination. During During preinitiation, preinitiation, an an assembly assembly complex complex is formed,is formed, consisting consisting of G1P, of G1P, G11P, G11P, and and G4P, G4P, all all interacting interacting through through their their periplasmic periplasmic domains. domains. A cylindricalA cylindrical structure, structure, consisting consisting of of 12- 12-1414 G4P G4P monomers, monomers, mediates mediates close close contact contact between between the the cytoplasmiccytoplasmic membrane membrane and and the the outer outer membrane. membrane. In addition In addition to a to large a large structure structure generated generated by G4P, by G4P, a multimeric complex containing five to six copies of both G1P and G10P is generated, potentially a multimeric complex containing five to six copies of both G1P and G10P is generated, potentially forming a second channel [22]. forming a second channel [22]. Initiation awaits formation of the assembly complex during preinitiation as well as accumulation Initiation awaits formation of the assembly complex during preinitiation as well as accumulation of G5P-bound M13 chromosomes, and subsequently positions G7P and G9P at the tip of the complex. of G5P-bound M13 chromosomes, and subsequently positions G7P and G9P at the tip of the complex. Here, G7P and G9P interact with the packing signal in the phage chromosome, which facilitates contact Here,with G7P the G1P, and resulting G9P interact in binding with of the thioredoxin packing signal from the in cell the [ phage31]. chromosome, which facilitates contactDuring with the the G1P, elongation resulting step, in binding the G5P of boundthioredoxin to the from phage the genome cell [31] is. replaced with G8P for translocationDuring the of elongation the DNA through step, the the G5P membrane-spanning bound to the phage channel genome (Figure is 3 replacedF). The translocation with G8P for translocationcontinues until of the the DNA phage through genome hasthe becomemembrane completely-spanning coated channel with (Figure G8P, at which3F). The point translocation G3P and continuesG6P will until collaborate the phage in the genome release has of thebecome phage completely from the bacterium. coated with If either G8P, G3Pat which or G6P point is missing, G3P and G6Padditional will collaborate phage in the release can be of loaded the phage through from the the pore bacterium. resulting If in either formation G3P ofor aG6P much is longermissing, additionalphage particle phage [ 32genomes,33]. can be loaded through the pore resulting in formation of a much longer phage particleDuring pretermination,[32,33]. membrane-embedded G3Ps complexed with G6Ps are incorporated at the terminalDuring end pretermination, of the phage particle. membrane Termination-embedded involves G3Ps thecomplexed release of with the phage,G6Ps are brought incorporated about by at thea terminal conformational end of changethe phage in the particle. G3P-G6P Termination complex [32 involves,34]. the release of the phage, brought about by a conformational change in the G3P-G6P complex [32,34].

3. Using the M13 Phage as a Tool in Antibody Discovery

Toxins 2018, 10, 236 5 of 15

3. Using the M13 Phage as a Tool in Antibody Discovery Toxins 2018, 10, x FOR PEER REVIEW 5 of 15 Phage display technology was demonstrated in 1985 by Smith, who successfully incorporated foreignPhage DNA display into the technology M13 phage was chromosome demonstrated such in 1985 that by foreign Smith, peptideswho successfully were fused incorporated to the G3P coatforeign protein DNA of into the the M13 M13 phage phage [35 chromosome]. Five years such later, that foreign antibody phage were display fused selection to the G3P was coat first describedprotein of by the McCafferty M13 phage et[35] al.,. Five who years were later, able antibody to fuse genesphage encodingdisplay selection an entire was antibody first described binding domainby McCafferty (in the form et al. of, who single-chain were able variable to fuse fragments, genes encoding scFvs) an to entire gene III antibody(Figure 4binding)[ 36]. Thisdomain approach (in facilitatedthe form of a sufficientsingle-chain level variable of antibody fragments, display scFvs) on to thegene outer III (Figure surface 4) [36] of the. This phage approach virion facilitated to permit selectiona sufficient of antigen-recognizing level of antibody display phages on the [36 ].outer The surface method of the thus phage exploits virion the to possibility permit selection of directly of antigen-recognizing phages [36]. The method thus exploits the possibility of directly linking a protein linking a protein () to its cognate gene () through a phage. Since 1990, different (phenotype) to its cognate gene (genotype) through a phage. Since 1990, different antibody formats antibody formats have been employed in the construction of antibody-displaying phage libraries, have been employed in the construction of antibody-displaying phage libraries, including heavy- including heavy-domain human antibody fragments (VHs), heavy-domain camelid and shark antibody domain human antibody fragments (VHs), heavy-domain camelid and shark antibody fragments fragments (VHHs), scFvs, diabodies (bivalent scFvs), and entire fragments antigen binding (Fab) (VHHs), scFvs, diabodies (bivalent scFvs), and entire fragments antigen binding (Fab) antibodies [37– antibodies40]. Generally, [37–40 these]. Generally, antibody fragments these antibody are fused fragments to the G3P are of fused the M13 to the phage G3P (Figure of the 4), M13 and phage by (Figurecloning4), large and by numbers of large genes numbers encoding of an genes antibody encoding fragment, an antibody large phage fragment, display large antibody phage displaylibraries antibody can be generatedlibraries can from be which generated many from diverse which antibodies many diverse can be selected.antibodies Ascan an beexample, selected. AsSchofield an example, et al.Schofield successfully et al.selected successfully more than selected 7200 recombinant more than 7200 antibodies recombinant to 292 antigens antibodies from to a 292 10 antigenshuman fromscFv library a human containing scFv library 1.1 × containing1010 clones [41] 1.1 ×. 10 clones [41].

III-scFv

G 3P

scFv

FigureFigure 4. 4.Engineering Engineering of of the the M13M13 phagephage for phage display display experiments. experiments. The The G3P G3P is isgenetically genetically fused fused toto a humana human single-chain single-chain variable variable fragmentfragment (scFv)(scFv) with a a trypsin trypsin cleavage cleavage sit sitee shown shown in inyellow yellow as part as part ofof a peptidea linker linker connecting connecting thethe twotwo proteins.proteins.

4. Phage and Libraries 4. Phage and Phagemid Libraries In the earliest examples, antibody genes were cloned directly into the filamentous phage genome, In the earliest examples, antibody genes were cloned directly into the filamentous phage genome, which carries all the genes needed for infection, replication, assembly, and budding while also whichcarrying carries the allgene the encoding genes needed the antibody for infection,-G3P fusion. replication, Since each assembly, phage andparticle budding normally while incorporates also carrying thethree gene to encoding five copies the of antibody-G3P G3P, the use of fusion. phage Since vectors each potentially phage particle results normally in a multivalent incorporates display three of to fiveantibody copies- ofG3P G3P, fusion the useproteins of phage (although vectors proteolytic potentially degradation results in acan multivalent result in removal display of of a antibody-G3Pproportion fusionof the proteins fused antibody). (although As proteolytic an alternative, degradation phagemid can vectors result based in removal on smaller of a proportion “minimal ” of the fused antibody).may be used. As an Phagemid alternative, vectors phagemid contain vectorsthree key based elements on smaller (i) an anti “minimalbiotic marker plasmids” for selection may be and used. Phagemidpropagation vectors of the contain three (ii) key the elements gene encoding (i) an antibiotic the antibody marker-G3P for fusion selection protein and, propagationand (iii) the of theregions plasmid of the (ii) M13 the genechromosome encoding (phage the antibody-G3P ) fusion required protein, andfor rolling (iii) the circle regions amplification of the M13 chromosomeand the production (phage originof the (+) of replication)DNA strand requiredthat is capable for rolling of being circle packaged amplification into a phage. and the In productionorder to ofproduce the (+) DNA functional strand phage that isparticles capable displaying of being packaged antibody- intoG3P afusion, phage. E. In coli order harboring to produce a phagemid functional phagevector particles is infected displaying with a antibody-G3P “helper phage”. fusion, The helperE. coli harboring phage contains a phagemid the complete vector M13 is infected genome with a “helperencoding phage”. all the phageThe helper proteins phage needed contains for capsid the complete production, M13 phagegenome assembly, encoding chromosome all the phage proteinsreplication, needed and for budding capsid production,(the concept phage of helper assembly, phage ischromosome described in replication, section 5 in and detail). budding Upon (the conceptinfection, of helperwild-type phage G3P is from described the helper in Section phage 5competes in detail). wi Uponth the phagemid infection, encoding wild-type the G3P G3P from-scFv the helperfusion phage protein competes for incorporation with the into phagemid the phage. encoding Ninety the percent G3P-scFv of the fusionresulting protein phage for population incorporation do intonot the display phage. any Ninety fusion percent protein ofand the the resulting vast majority phage of population the phage particles do not display that bear any the fusion protein and will only contain a single copy [42]. Phagemid vectors are more typically used in library construction than phage vectors because higher transformation efficiencies can be achieved facilitating the construction of larger libraries. In a direct comparison of libraries created using phage versus

Toxins 2018, 10, 236 6 of 15 the vast majority of the phage particles that bear the fusion protein will only contain a single copy [42]. Phagemid vectors are more typically used in library construction than phage vectors because higher transformation efficiencies can be achieved facilitating the construction of larger libraries. In a direct comparison of libraries created using phage versus phagemid vectors, it was shown that a greater diversity of antibody binders was generated from libraries built using the phage vectors rather than phagemid vectors. The phage system allows for the emergence of a wider range of antibody affinities compared to the phagemid system [43]. Although the phage system is often referred to as a polyvalent display and the phagemid system as monovalent, in practice degradation occurs in both cases, and the beneficial effects of a phage vector-based system may also be attributed to the presence of a higher proportion of non-bald phage particles. Methods have been developed to restore antibody display levels in phagemid libraries to the same levels achieved in phage vector-based systems. This is based on the use of helper phages, which do not encode G3P (the so called hyperphage system [44]), so that only the G3P-scFv encoded by the phagemid is available for presentation. Since G3P is needed for packaging and infection of the helper phage and the hyperphage genome lacks the gene encoding G3P, it is necessary to provide this in trans by using which express G3P from within the bacterial genome.

5. Phagemid Libraries Are Amplified Using Helper Phages As mentioned earlier, an M13 helper phage is used for packaging of phagemid particles in E. coli. The helper phage carries all genes necessary for infection, replication, assembly, and budding and therefore provides the phagemid, which primarily carries the gene encoding the G3P-scFv fusion protein, with the proteins needed for amplification. One such helper phage is the M13KO7 phage [45]. This helper phage carries a heterologous, low copy origin of replication from the plasmid p15A inserted within the native M13 origin of replication. When the M13KO7 phage is present alone in a host bacterium (i.e., during preparation of helper phage), the replication is sufficient for producing high titers of M13KO7 phage. However, if a high copy number phagemid with an intact M13 origin is present, this will out-compete the helper plasmid during packaging, meaning that the majority of resulting phage particles will carry phagemid DNA [45]. The resultant phagemid particles will incorporate G3P encoded by the helper phage genome as well as G3P-scFv encoded within the packaged phagemid. In practice, there is preferential incorporation of the wild-type G3P from the helper plasmid, meaning that the majority of the will be ‘bald’ (not displaying an antibody fragment). The preferential presentation of G3P rather than the G3P-scFv fusion is probably due to a combination of differences in expression/ levels and the fact that a proportion of the G3P-scFv fusion is degraded (Figure5). An overview of which genetic elements are carried by the phagemid and the helper phage is provided in Table2.

Table 2. Genetic elements carried by phagemids and helper phages.

Genetic Element Phagemid Helper Phage Gene I X Gene II X Gene III X Gene IV X Gene V X Gene VI X Gene VII X Gene VIII X Gene IX X Gene III-scFv X Antibiotic resistance gene X X Origin of replication X Origin of replication (inefficient) X Toxins 2018, 10, 236 7 of 15 Toxins 2018, 10, x FOR PEER REVIEW 7 of 15 Toxins 2018, 10, x FOR PEER REVIEW 7 of 15 H elper phage H elper phage

G ene III G ene III

Phagem id Phagem id

G ene III-scFv G ene III-scFv

FigureFigure 5. Gene5. GeneIII IIIproduct product from from helper helper phage phage and and genegene III-scFvIII-scFv product from from phagemid. phagemid. The The helper helper Figure 5. Gene III product from helper phage and gene III-scFv product from phagemid. The helper phagephage carries carries the gene the gene encoding encoding a trypsin-sensitive a trypsin-sensitive ‘bald’ ‘bald’ G3P. The G3P. phagemid The phagemid carries thecarries gene the encoding gene phage carries the gene encoding a trypsin-sensitive ‘bald’ G3P. The phagemid carries the gene theen G3P-scFvcoding the fusion G3P-scFv protein. fusion As protein. M13 phagesAs M13 willphages be will assembled be assembled with threewith three to five to five copies copies of G3P,of encoding the G3P-scFv fusion protein. As M13 phages will be assembled with three to five copies of theG3P, number the number of assembled of assembled phages phages containing containing more more than than one G3P-scFvone G3P-scFv fusion fusion protein protein will will be be too too low G3P, the number of assembled phages containing more than one G3P-scFv fusion protein will be too to influencelow to influence the outcome the outcome of selections, of selections, because because more more ‘bald’ ‘ba G3Pld’ G3P than than G3P-scFv G3P-scFv fusion fusion proteins proteins will low to influence the outcome of selections, because more ‘bald’ G3P than G3P-scFv fusion proteins be expressed.will be expressed. Thereby, Thereby, the majority the majority of the of scFv-displaying the scFv-displaying phagemids phagemids will will only only be carrying be carrying one one scFv will be expressed. Thereby, the majority of the scFv-displaying phagemids will only be carrying one (monovalentscFv (monovalent display). display). scFv (monovalent display). Bald phages do not participate in positive antibody: antigen selection and only contribute to BaldBald phages phages dodo notnot participateparticipate in in positive positive antibody antibody:: anti antigengen selection selection and andonly onlycontribute contribute to background noise within the system. It is possible to reduce the background noise from this tobackground background noise noise within within the the system. system. It It is is possible possible to to reduce reduce the the background background noise noise from from this this population by rendering them non-infectious. This is achieved by engineering gene III within the populationpopulation by by rendering rendering them them non-infectious.non-infectious. This This is is achieved achieved by by engineering engineering gene gene IIIIII withinwithin the the helper phage to render the encoded G3P sensitive to trypsin [46]. Trypsin treatment before infection helperhelper phage phage to to render render the the encoded encoded G3PG3P sensitive to to trypsin trypsin [46 [46].]. Trypsin Trypsin treatment treatment before before infection infection will result in inactivation of bald phages, which rely on labile G3P encoded by the helper phage willwill result result in in inactivation inactivation of of baldbald phages,phages, which rely rely on on labile labile G3P G3P encoded encoded by by the the helper helper phage phage (Figure 6). In contrast, phages carrying a G3P-scFv fusion, encoded by the phagemid, will be resistant (Figure(Figure6). 6). In In contrast, contrast, phages phages carrying carrying a a G3P-scFv G3P-scFv fusion,fusion, encoded by by the the phagemid, phagemid, will will be be resistant resistant to G3P disruption and will retain infectivity after trypsin treatment (As previously shown in Figure toto G3P G3P disruption disruption and and will will retain retain infectivity infectivity after after trypsintrypsin treatmenttreatment (As previously shown shown in in Figure Figure 4, 4, the G3P-scFv fusion protein encoded by the phagemid contains a trypsin cleavage site, but this is the4, G3P-scFvthe G3P-sc fusionFv fusion protein protein encoded encoded by by the the phagemid phagemid containscontains a a trypsin trypsin cleavage cleavage site, site, but but this this is is situated between the scFv and the G3P, and thereby does not prevent infection (Figure 6)). situatedsituated between between the the scFv scFv and and the the G3P,G3P, andand therebythereby does not prevent prevent i infectionnfection (Figure (Figure 6)).6)).

scFv carrying phage Bald phage scFv carrying phage Bald phage

n

i

n

s

i

p

s

y

p

r

y

t

r

t

e

r

e

o

r

f

o

e

f

B

e

B

n

i

n

s

i

p

s

y

p

r

y

t

r

t

r

e

r

t

e

f

t

f

A

A Infective N on-infective Infective N on-infective

Figure 6. Trypsin treatment of G3P and G3P-scFv fusion proteins on the M13 phage. G3P encoded by Figure 6. Trypsin treatment of G3P and G3P-scFv fusion proteins on the M13 phage. G3P encoded by Figurethe helper 6. Trypsin phage treatment is trypsin of-sensitive G3P and and G3P-scFv can thusfusion be cleaved proteins by trypsin, on the rendering M13 phage. the G3Pphage encoded non- the helper phage is trypsin-sensitive and can thus be cleaved by trypsin, rendering the phage non- byinfective. the helper In phagecontrast, is the trypsin-sensitive G3P-scFv fusion and protein can encoded thus be by cleaved the phagemid by trypsin, contains rendering a myc-tag the in phage the infective. In contrast, the G3P-scFv fusion protein encoded by the phagemid contains a myc-tag in the non-infective.linker between In contrast, the G3P theand G3P-scFv the scFv, fusionwhich proteinwill be cleaved encoded during by the trypsin phagemid treatment. contains This a myc-tag leaves a in linker between the G3P and the scFv, which will be cleaved during trypsin treatment. This leaves a theG3P linker on betweenthe phage, the rendering G3P and it the infective. scFv, which will be cleaved during trypsin treatment. This leaves a G3PG3P onon the phage, rendering rendering it it infective. infective.

Toxins 2018, 10, 236 8 of 15

Toxins 2018, 10, x FOR PEER REVIEW 8 of 15 6. Performing a Phage Display Selection Experiment 6. PhagePerforming display a Phage selection Display is a Selection high throughput Experiment method used to discover antibodies specific to differentPhage antigens. display The selection protocol is includesa high throughput rounds of method five steps used (Figure to discover7). The antibodies first step specific is addition to ofdifferent a phage antigens. library, where The protocol the library includes is added rounds to of a five well steps or vial(Figure in which7). The the first antigen step is addition is presented of (Figurea phage7A). library, Antigen where presentation the library canis added be achieved to a well via or vial direct in which coating the (adsorption) antigen is presented or via a (Figure capture system7A). Antigen (such as presentation streptavidin-biotin). can be achieved The second via direct step involves coating (adsorption) binding, where or via the a phages capture displaying system the(such highest as streptavidin affinity antibodies-biotin). bindThe second the epitopes step involves of the antigen binding, (Figure where 7 theB), phages after which displaying the vial the is washedhighest to aff removeinity antibodies non-binding bind phagesthe epitopes (Figure of the7C). antigen After washing,(Figure 7B), the after bound which phages the vial are is elutedwashed by enzymaticto remov digestione non-binding using phages trypsin (Figure or by 7 otherC). After means washing, of elution the bound (e.g., phages dilute acidare eluted or base) by enzymatic (Figure7D). Whendigestion enzymatic using digestiontrypsin or is by employed, other means this of step elution renders (e.g. bald, dilu phageste acid or non-infective base) (Figure [46 7D).] (Figure When6 ), enzymatic digestion is employed, this step renders bald phages non-infective [46] (Figure 6), thereby thereby improving the selection of antibody-displaying phages that are still infective (see previous improving the selection of antibody-displaying phages that are still infective (see previous section). section). In the fifth step, helper phages are added to allow for amplification of the eluted infective In the fifth step, helper phages are added to allow for amplification of the eluted infective phages in phages in E. coli (Figure7E). To accumulate phages displaying high affinity antibody fragments, these E. coli (Figure 7E). To accumulate phages displaying high affinity antibody fragments, these five steps five steps are usually repeated 1-3 times with the amplified phages from the preceding round of are usually repeated 1-3 times with the amplified phages from the preceding round of panning. panning.Titration Titration of phage of output phage outputnumbers numbers may also may be utilized also be utilizedto monitor to monitorprogress, progress, and polyclonal and polyclonal phage phageELISA ELISA used used to confirm to confirm selection selection of a population of a population of binders. of binders. SelectionsSelections can can also also be be performed performed inin solutionsolution withwith a subsequent pull pull-down-down step step to to isolate isolate high high affinityaffinity scFv-displaying scFv-displaying phages phages that that are are bound bound toto thethe antigen (which (which typically typically needs needs a atag tag that that can can be be pulledpulled down, down, such such as as a biotina biotin moiety) moiety) [ 47[47]].. ThisThis maymay also serve to to avoid avoid antigen antigen immobilization immobilization via via hydrophobichydrophobic regions, regions, which which may may diminish diminish thethe accessibility of of relevant relevant epitopes epitopes as asdiscussed discussed below. below. Conceptually,Conceptually, selections selections performed performed in in solution solution include the the exact exact same same steps steps with with a a few few practical practical changes.changes. Selections Selections in in solution solution may may be be beneficial beneficial for certain antigens, antigens, which which better better retain retain proper proper foldingfolding and and conformation conformation that that resembles resembles the the native native antigen antigen protein. protein. Possibly, Possibly, this isthis particularly is particularly useful foruseful antigens for antigens that natively that natively exist in exist solution in solution (e.g., blood (e.g., blood factors, factors, toxins, toxins, cytokines, cytokines, and and hormones). hormones).

Figure 7. The five steps in a phage display selection experiment. Addition of phage library (A) refers Figure 7. The five steps in a phage display selection experiment. Addition of phage library (A) refers to the addition of phages to an antigen-coated vial. The phages displaying the highest affinity scFvs to the addition of phages to an antigen-coated vial. The phages displaying the highest affinity scFvs willwill bind bind (B )(B the) the antigen, antigen, while while unspecific unspecific binders binders willwill be removed removed during during the the washing washing step step (C ()C.After).After washing,washing, the the antigen-specific antigen-specific phages phages cancan bebe elutedeluted (D) using trypsin trypsin digestion digestion (or (or other other means means of of elution)elution) Then, Then, phages phages are are amplified amplified ( E(E)) in inE. E. colicoli andand a new panning round round can can be be initiated initiated to tofurther further accumulateaccumulate phages phages displaying displaying high high affinity affinity antibody antibody fragments.fragments. Binding Binding can can be be evaluated evaluated using using both both ELISAELISA and and plate plate tests. tests.

ToxinsToxins2018 2018, 10, 10, 236, x FOR PEER REVIEW 99 of of 15 15

7. Antigen Quality and Presentation 7. Antigen Quality and Presentation In addition to having a phage display library of high quality, the most crucial element for the successIn addition of a phage to display having aselection phage displayexperiment library is antigen of high presentation. quality, the Different most crucial methods element of ant forigen the successpresentation of a phage exist display with different selection advantages experiment and is antigen disadvantages presentation. (Figure Different 8) [48,49] methods. The of simplest antigen presentationmethod of antigen exist with presentation different is advantages to coat the antigen and disadvantages directly to the (Figure plastic8 )[surface48,49 of]. a The well simplest or vial methodvia adsorption. of antigen The presentation major disadvantage is to coat of the direct antigen immobilization directly to theis that plastic the surfaceantigen ofmight a well undergo or vial viaconformational adsorption. The changes major disadvantage upon binding of todirect the immobilization well or vial is surface, that the potentially antigen might leading undergo to conformationalconformational changes distortion upon or bindingdenaturation to the of well the orantigen. vial surface, Selection potentially of antibodies leading against to conformational a distorted distortionor denatur ored denaturation antigen will result of the in antigen. accumulation Selection of phages of antibodies displaying against antibodies a distorted with orhigh denatured affinity antigento the distorted will result antigen in accumulation but low affinity of phages to the displayingnative antigen antibodies in its proper with conformation. high affinity to Additionally, the distorted antigenantigens but might low affinityhave preferences to the native regarding antigen how in its they proper bind conformation. to the surface, Additionally, which might antigensresult in mightareas haveof the preferences antigen not regarding being properly how they exposed bind to for the interaction. surface, which Therefore, might very result few in areas phages of thedisplaying antigen notantibodies being properly with high exposed affinity for to interaction. such an area Therefore, can be selected, very few which phages is displaying problematic antibodies if this area with is highimportant affinity to to target such anfor areaantigen can neutralization. be selected, which is problematic if this area is important to target for antigenAntigen neutralization. presentation by immobilization through a streptavidin-biotin capture system requires thatAntigen the antigen presentation is chemically by immobilization conjugated to a throughbiotin moiety a streptavidin-biotin via a linker. Biotinylation capture system of an antigen requires thatcan thebe antigena challenging is chemically task, espe conjugatedcially since to the a biotin biotin moiety-to-antigen via aratio linker. can Biotinylation be crucial but of difficult an antigen to cancontrol. be a challenging Over-biotinylation task, especially of the antigen since themay biotin-to-antigen heavily decrease ratio the available can be crucial surface but area difficult on the to control.antigen, Over-biotinylationthereby preventing ofphage the antigenbinding. may Also, heavily over-biotinylation decrease the may available alter the surface physico area-chemical on the antigen,properties thereby of the preventing antigen, possibly phage binding.leading to Also, undesirable over-biotinylation effects, such may as alterantigen the aggregation. physico-chemical The propertiesmain advantage of the antigen, of immobilization possibly leading through to undesirable a biotin linker effects, is that such the as antigen antigen will aggregation. be lifted from The main the advantageplastic surface of immobilization of the vial bottom through by first a biotin coating linker the vial is that with the streptavidin antigen will (or be neutravidin) lifted from the and plastic then surfaceadding of the the biotinylated vial bottom antigen by first to coating be captured. the vial This with allows streptavidin for an (orimproved neutravidin) display and of thenthe antigen adding theand biotinylated thus the selection antigen of to phages be captured. displaying This allowsrelevant for antibodies an improved against display more of antigen the antigen epitope ands thus[48]. the A selectiondisadvantage of phages of displayingimmobilization relevant through antibodies a biotin against linker more is antigen that addition epitopes [of48 ]. streptavidin A disadvantage (or ofneutravidin) immobilization introduces through an a additional biotin linker macromolecule is that addition that of the streptavidin phage can (orbind neutravidin) to, which may introduces lead to anaccumulation additional macromolecule of phages with that streptavidin/neutravidin the phage can bind to,- whichbinding may antibodies. lead to accumulation To avoid this of phages issue, withdeselection streptavidin/neutravidin-binding steps must be implemented antibodies.in the phage Todisplay avoid protocol this issue, when deselection working with steps antigens must be that are immobilized using such capture systems. implemented in the phage display protocol when working with antigens that are immobilized using Regarding optimal antigen display via capture systems, it can be worthwhile to consider which such capture systems. coupling chemistry should be employed. The benefit of having a site-specific coupling reagent Regarding optimal antigen display via capture systems, it can be worthwhile to consider which includes greater control; however, this may in some cases have a negative interference with a coupling chemistry should be employed. The benefit of having a site-specific coupling reagent includes potential binding site, which may lead to poor selection. As an example, site-specific coupling to the greater control; however, this may in some cases have a negative interference with a potential binding N-terminal is likely to be suboptimal if the N-terminal region is part of an important epitope-paratope site, which may lead to poor selection. As an example, site-specific coupling to the N-terminal is likely interaction. to be suboptimal if the N-terminal region is part of an important epitope-paratope interaction.

D irect coating Biotinylated toxin O ver-biotinylated toxin

N ative toxin D enatured toxin Streptavidin Biotin

FigureFigure 8. 8.Antigen Antigen immobilization immobilization strategies. strategies. Direct Direct coating coating can can result result in in denaturation denaturation of of the the antigen antigen or distortionor distortion of its of conformation, its conformation, whereas whereas immobilization immobilization through through a streptavidin-biotin a streptavidin- capturebiotin capture system, involvingsystem, involving biotinylation biotinylation of the antigen, of the can antigen, give rise can to over-biotinylationgive rise to over-biotinylation resulting in low resulting availability in low of bindingavailability sites. of binding sites.

8.8. Naïve Naïve versus versus Immunized Immunized PhagePhage DisplayDisplay AntibodyAntibody Libraries PhagePhage display display antibody antibody libraries libraries can be can derived be derived from either from non-immunized either non-immunized (naïve) or immunized (naïve) or donorsimmunized depending donors on depending whetherthe on donors,whether fromthe donors, which thefrom antibody which the genes antibody were isolatedgenes were and isolated used to

Toxins 2018, 10, 236 10 of 15

Toxins 2018, 10, x FOR PEER REVIEW 10 of 15 create the library, have been immunized with an antigen or not. The benefit of naïve libraries (often fromand theused IgM to repertoirecreate the fromlibrary, non-immunized have been immu donors)nized with is that an it antigen can be usedor not. for The discovery benefit of anaïve wide rangelibraries of antigens, (often from although the IgM the repertoire discovered from antibody non-immunized fragments may donors) generally is that have it can lower be used affinities for thandiscovery antibodies of a wide discovered range of from antigens, an immunized although sourcethe discovered [7,50]. In antibody a phage fragments display antibody may generally library fromhave an lower immunized affinities source than antibodies (from the discovered IgG repertoire), from an the immunized antibody fragments source [7,50] have. In increased a phage display affinity antibody library from an immunized source (from the IgG repertoire), the antibody fragments have towards the type of antigen used for immunization as the antibodies have gone through an in vivo increased affinity towards the type of antigen used for immunization as the antibodies have gone affinity maturation process. This can be an advantage in antibody discovery as it will more easily through an in vivo affinity maturation process. This can be an advantage in antibody discovery as it generate high affinity hits. Antibodies from the IgG pool of immunized sources will generally have will more easily generate high affinity hits. Antibodies from the IgG pool of immunized sources will less diversity than naïve IgM-based libraries, and they are therefore not as broadly applicable in terms generally have less diversity than naïve IgM-based libraries, and they are therefore not as broadly of their antigen scope. applicable in terms of their antigen scope. Different techniques can be employed for affinity maturation of antibody fragments Different techniques can be employed for in vitro affinity maturation of antibody fragments from naïve phage libraries. The basic principle of in vitro affinity maturation is to first diversify the from naïve phage libraries. The basic principle of in vitro affinity maturation is to first diversify the sequencessequences of of the the antibody antibody fragments fragments andand thenthen carrycarry out new rounds of of phage phage display display selections selections to to selectselect antibody antibody fragments fragments with with increasedincreased affinityaffinity toto thethe target antigen. Several Several different different methods methods to to achieveachieve diversification diversification exist. exist. One One suchsuch methodmethod isis chain shuffling shuffling [51] [51],, where where the the antibody antibody fragments fragments obtainedobtained after after selection selection either either have have their their heavy heavy or or their their light light chain chain replaced replaced with with the the full full repertoire repertoire of thisof chainthis chain from from a naïve a naïve library library (often (often the samethe same library library from from which which the initialthe initial antibody antibody fragment fragmen wast selected)was selected) (Figure (Figure9A). Thereby, 9A). Thereby, a new phagea new phage display display antibody antibody library library is created is created with a with constant a constant heavy orheavy light chain or light and chain the full and range the full of therange opposite of the chainopposite as partners chain as (i.e., partners one heavy(i.e., one chain heavy combined chain withcombined all available with all light available chains, ligh ort chains, one light or chainone light combined chain combined with all with available all available heavy heavy chains). chains). Phage displayPhage selectionsdisplay selections are then are conducted then conducted again to again identify to id theentify best the match best match between between heavy heavy and light and chains, light whichchains, often which results often in antibodyresults in fragments antibody fragments with increased with affinity increased to affinitytheir targets. to their Other targets. methods Other of diversificationmethods of diversification used for in vitro usedaffinity for in vitro maturation affinity maturation employ mutagenesis employ mutagenesis of the variable of the regions variable of theregions antibodies. of the Theseantibodies. mutations These canmutations either becan introduced either be introduced through random through mutagenesis random mutagenesis of the whole of variablethe whole region variable of the region antibodies of the [ 52antibodies] (Figure 9[52]B) or (Figure through 9B) different or through types different of site-directed types of site mutagenesis-directed techniquesmutagenesis that techniques specifically that target spec theifically different target complementarity-determining the different complementarity- regionsdetermining (CDRs) regions in the variable(CDRs) regions in the variable of the antibodies regions of (Figurethe antibodies9C) [53 ].(Figure 9C) [53].

Figure 9. Diversification methods used for in vitro affinity maturation of antibody fragments. Chain Figure 9. Diversification methods used for in vitro affinity maturation of antibody fragments. Chain shuffling (A) is where one heavy or light chain is paired with chains of the opposite type from a naïve

Toxins 2018, 10, 236 11 of 15

shuffling (A) is where one heavy or light chain is paired with chains of the opposite type from a naïve library. The principle of both light and heavy chain shuffling is presented in the figure. Random mutagenesis (B) is used to introduce mutations (red lines) in the entire variable regions of the antibody fragment. Site-directed mutagenesis (C) is used to specifically introduce mutations in one or more of the complementarity-determining regions (CDR) regions of the antibody fragment.

9. Selected Examples of the Use of Antibody Phage Display Selection within Toxinology Phage display selection has been employed in the field of toxinology to obtain antibody fragments with specificities to different toxins [3] and was first employed within snakebite antivenom research in 1995 by Meng et al., who used a murine scFv library to discover antibody fragments against crotoxin from Crotalus durissus terrificus [54]. The employed scFv library was affinity matured as it was generated from mice that had been immunized with crotoxin. The dissociation constant between one obtained scFv and crotoxin was determined to be 7.0 × 10−10 M, reflecting a very strong binding interaction between the antibody fragment and the target toxin. The antibody fragment was tested for its ability to inhibit phospholipase A2 activity in vitro without any effects even at 5:1 antibody-to-toxin molar ratios. However, the scFv successfully delayed death and in some cases even prevented death in mice injected with lethal doses of Mojave toxin (a crotoxin homolog) pre-incubated with the scFv [54]. Since 1995, phage display technology has been employed using different target toxins from different animal species (including snake, scorpion, spider, and bee toxins), using libraries containing murine, human, and camelid antibodies [2,12]. As an example, Kulkeaw et al. were the first to use a human antibody phage display library [55]. This library consisted of human svFvs and was used to discover antibody fragments against α-cobratoxin (a long chain α-neurotoxin) from Naja kaouthia. The scFvs selected from this library were tested for their ability to neutralize one LD100 of α-cobratoxin upon pre-incubation of each scFv and the toxin in vivo. This experiment demonstrated that the selected human scFvs mainly provided prolonged survival of the mice and failed to protect them from lethality. However, the best scFv was able to protect 33% (2/6) of the mice administered with scFv and toxin from death. In comparison, Richard et al. employed a llama VHH phage display library to discover antibody fragments against α-cobratoxin from N. kaouthia [7]. Here, the library was affinity matured as it was based on VHH genes from a llama immunized with N. kaouthia crude venom. The obtained antibody fragments had dissociation constants as low as 4 × 10−10 M and proved effective both in protecting and rescuing mice in a lethality study at antibody-to-toxin molar ratios as low as 0.75:1. Several examples of phage display selection being employed for discovery of antibody fragments with specificity to scorpion toxins can be found in the literature. One such example was reported by Rodríguez-Rodríguez et al., where a human scFv library was used to discover scFvs against Mexican scorpion venom toxins [56]. Here, scFvs with the ability to bind Cn2 toxin (a β-neurotoxin) from Centruroides noxius were discovered. However, the scFvs were not able to neutralize the Cn2 toxin in vivo. To improve the affinity of the antibody fragments, in vitro affinity maturation methods involving , site-directed mutagenesis, and random mutagenesis were employed. As a result, the researchers successfully improved an scFv to become effective in protecting mice from both the Cn2 toxin from C. noxius, the Cll1 toxin from Centruroides limpidus, the Css2 toxin from Centruroides suffusus, and whole venom from C. suffuses [56]. Another example is provided by Pucca et al., where the human scFv library, Griffin.1, was employed to discover scFvs against Tityus serrulatus toxins [57]. One obtained scFv was investigated for its neutralizing abilities both through in vitro biochemical assays and in a murine model, where it showed partially neutralizing effects. In a later study of the same scFv, it was additionally shown that the scFv could neutralize the effects of similar β-neurotoxins from other scorpion genera [58]. Finally, phage display selection has also been employed to discover antibody fragments against bee venom toxins [59]. In a study by Pessenda et al., the human scFv library, Griffin.1, was used to Toxins 2018, 10, 236 12 of 15

find scFvs against melittin and phospholipase A2 from the Africanized bee Apis mellifera. These scFvs were shown to inhibit the hemolytic activity of the venom in vitro and displayed the ability to reduce the edematogenic activity when the scFvs were pre-incubated with the venom and injected into the paws of mice. Additionally, prolongation of survival was observed when mice were subjected to lethal doses of venom pre-incubated with the scFvs. However, full protection against lethality was not observed [59]. In relation to toxin-targeting, it is important to note that although antibodies with specific toxin-binding specificities can often easily be discovered using phage display selection, such binding does not necessarily translate into toxin-neutralizing ability in vivo [60–62]. Different modes of toxin neutralization likely exist [63], and any phage display selection experiment must be followed by vigorous testing of functional activity of the isolated toxin-binding antibodies, as antibodies binding to non-neutralizing epitopes may have limited therapeutic value.

10. Closing Remarks With the renewed international focus on snakebite envenoming, more focus on innovation and development within the field of antivenom will likely arise. Phage display technology represents a robust, inexpensive, and easy to perform discovery approach that is particularly suitable for large-scale antibody discovery projects within antivenom research. In this area, multiple antibodies against a vast range of different toxin targets need to be developed, and phage display selection offers the opportunity to do so in a parallelized manner. Although no other animal envenomings than snakebite are currently included in the World Health Organization’s list of Neglected Tropical Diseases, these analogous fields may likely benefit from the inclusion of snakebite envenoming on this list. The use of phage display technology within toxinology and antivenom research is still at its infancy. However, several promising examples have already been reported in the literature, supporting the notion that phage display selection is indeed a feasible antibody discovery approach. Additionally, the many developments and advances within antibody technologies and manufacturing, championed by other fields, such as oncology and autoimmune and infectious diseases, represent a golden opportunity that antivenom research may piggy-back on. It therefore seems relevant that antivenom researchers become acquainted with phage display technology as well as other antibody discovery approaches.

Acknowledgments: The authors thank Cecilie Knudsen (Technical University of Denmark) and Timothy P. Jenkins (University of Cambridge) for helping proofread the final manuscript. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Chippaux, J.-P. Snakebite envenomation turns again into a neglected tropical disease! J. Venom. Anim. Toxins Trop. Dis. 2017, 23.[CrossRef][PubMed] 2. Laustsen, A.H.; Engmark, M.; Milbo, C.; Johannesen, J.; Lomonte, B.; Gutiérrez, J.M.; Lohse, B. From Fangs to Pharmacology: The Future of Snakebite Envenoming . Curr. Pharm. Des. 2016, 22, 5270–5293. [CrossRef][PubMed] 3. Laustsen, A.H. Guiding recombinant antivenom development by omics technologies. New Biotechnol. 2017, in press. [CrossRef][PubMed] 4. Harrison, R.A.; Cook, D.A.; Renjifo, C.; Casewell, N.R.; Currier, R.B.; Wagstaff, S.C. Research strategies to improve snakebite treatment: Challenges and progress. J. Proteom. 2011, 74, 1768–1780. [CrossRef][PubMed] 5. Knudsen, C.; Laustsen, A.H. Recent Advances in Next Generation Snakebite Antivenoms. Trop. Med. Infect. Dis. 2018, 3, 42. [CrossRef] 6. Laustsen, A.H.; Johansen, K.H.; Engmark, M.; Andersen, M.R. Recombinant snakebite antivenoms: A cost-competitive solution to a neglected tropical disease? PLoS Negl. Trop. Dis. 2017, 11, e0005361. [CrossRef] [PubMed] Toxins 2018, 10, 236 13 of 15

7. Richard, G.; Meyers, A.J.; McLean, M.D.; Arbabi-Ghahroudi, M.; MacKenzie, R.; Hall, J.C. In Vivo

Neutralization of α-Cobratoxin with High-Affinity Llama Single-Domain Antibodies (VHHs) and a VHH-Fc Antibody. PLoS ONE 2013, 8, e69495. [CrossRef][PubMed] 8. Julve Parreño, J.M.; Huet, E.; Fernández-Del-Carmen, A.; Segura, A.; Venturi, M.; Gandía, A.; Pan, W.-S.; Albaladejo, I.; Forment, J.; Pla, D.; et al. A synthetic biology approach for consistent production of -made recombinant polyclonal antibodies against snake venom toxins. Plant Biotechnol. J. 2018, 16, 727–736. [CrossRef][PubMed] 9. Laustsen, A.H.; Johansen, K.H.; Engmark, M.; Andersen, M.R. Snakebites: Costing recombinant antivenoms. Nature 2016, 538, 41. [CrossRef][PubMed] 10. Roncolato, E.C.; Campos, L.B.; Pessenda, G.; Costa e Silva, L.; Furtado, G.P.; Barbosa, J.E. Phage display as a novel promising antivenom therapy: A review. Toxicon 2015, 93, 79–84. [CrossRef][PubMed] 11. Bazan, J.; Całkosi´nski, I.; Gamian, A. Phage display—A powerful technique for immunotherapy. Hum. Vaccines Immunother. 2012, 8, 1817–1828. [CrossRef][PubMed] 12. Laustsen, A.H.; Solà, M.; Jappe, E.C.; Oscoz, S.; Lauridsen, L.P.; Engmark, M. Biotechnological Trends in Spider and Scorpion Antivenom Development. Toxins 2016, 8, 226. [CrossRef][PubMed] 13. Hofschneider, P.H. Untersuchungen über „kleine “E. coli K 12 Bakteriophagen. Z. Für Naturforschung B 2014, 18, 203–210. [CrossRef] 14. Rasched, I.; Oberer, E. Ff coliphages: Structural and functional relationships. Microbiol. Rev. 1986, 50, 401–427. [PubMed] 15. O’Callaghan, R.; Bradley, R.; Paranchych, W. The effect of M13 phage infection upon the F pili of E. coli. Virology 1973, 54, 220–229. [CrossRef] 16. van Wezenbeek, P.M.; Hulsebos, T.J.; Schoenmakers, J.G. sequence of the filamentous bacteriophage M13 DNA genome: Comparison with phage fd. Gene. 1980, 11, 129–148. [CrossRef] 17. Berkowitz, S.A.; Day, L.A. Mass, length, composition and structure of the filamentous bacterial fd. J. Mol. Biol. 1976, 102, 531–547. [CrossRef] 18. Jacobson, A. Role of F Pili in the Penetration of Bacteriophage fl. J. Virol. 1972, 10, 835–843. [PubMed] 19. Riechmann, L.; Holliger, P. The C-terminal domain of TolA is the coreceptor for filamentous phage infection of E. coli. Cell 1997, 90, 351–360. [CrossRef] 20. Click, E.M.; Webster, R.E. Filamentous phage infection: Required interactions with the TolA protein. J. Bacteriol. 1997, 179, 6464–6471. [CrossRef][PubMed] 21. Clarke, M.; Maddera, L.; Harris, R.L.; Silverman, P.M. F-pili dynamics by live-cell imaging. Proc. Natl. Acad. Sci. USA. 2008, 105, 17978–17981. [CrossRef][PubMed] 22. Clackson, T.; Lowman, H. Phage Display: A Practical Approach; Oxford University Press: Oxford, UK, 2004. 23. Branston, S. An Investigation of the Properties of Bacteriophage M13 and the Implications for Its Large-Scale Bioprocessing. Ph.D. Thesis, University College London, London, UK, 2009. 24. Russel, M.; Whirlow, H.; Sun, T.P.; Webster, R.E. Low-frequency infection of F- bacteria by transducing particles of filamentous . J. Bacteriol. 1988, 170, 5312–5316. [CrossRef][PubMed] 25. Webster, R.E. The tol gene products and the import of macromolecules into Escherichia coli. Mol. Microbiol. 1991, 5, 1005–1011. [CrossRef][PubMed] 26. Suggs, S.V. Replication of bacteriophage M13. J. Mol. Biol. 1977, 110, 147–163. [CrossRef] 27. Weigel, C. Bacteriophage replication modules. FEMS Microbiol. Rev. 2006, 30, 321–381. [CrossRef][PubMed] 28. Asano, S.; Higashitani, A.; Horiuchi, K. Filamentous phage replication initiator protein gpII forms a covalent complex with the 50 end of the nick it introduced. Nucleic Acids Res. 1999, 27, 1882–1889. [CrossRef][PubMed] 29. Stassen, A.P. Single-stranded DNA binding protein encoded by the filamentous bacteriophage M13: Structural and functional characteristics. Mol. Biol. Rep. 1994, 20, 109–127. [CrossRef][PubMed] 30. Russel, M.; Model, P. Genetic analysis of the filamentous bacteriophage packaging signal and of the proteins that interact with it. J. Virol. 1989, 63, 3284–3295. [PubMed] 31. Russel, M.; Model, P. The role of thioredoxin in filamentous phage assembly. Construction, isolation, and characterization of mutant thioredoxins. J. Biol. Chem. 1986, 261, 14997–15005. [PubMed] 32. Rakonjac, J. Filamentous phage are released from the bacterial membrane by a two-step mechanism involving a short C-terminal fragment of pIII. J. Mol. Biol. 1999, 289, 1253–1265. [CrossRef][PubMed] 33. Rakonjac, J.; Model, P. Roles of pIII in filamentous phage assembly. J. Mol. Biol. 1998, 282, 25–41. [CrossRef] [PubMed] Toxins 2018, 10, 236 14 of 15

34. Rakonjac, J.; Jovanovic, G.; Model, P. Filamentous phage infection-mediated gene expression: Construction and propagation of the gIII deletion mutant helper phage R408d3. Gene. 1997, 198, 99–103. [CrossRef] 35. Smith, G.P. Filamentous fusion phage: Novel expression vectors that display cloned antigens on the virion surface. Science. 1985, 228, 1315–1317. [CrossRef][PubMed] 36. McCafferty, J.; Griffiths, A.D.; Winter, G.; Chiswell, D.J. Phage antibodies: Filamentous phage displaying antibody variable domains. Nature. 1990, 348, 552–554. [CrossRef][PubMed] 37. Dooley, H. Selection and characterization of naturally occurring single-domain (IgNAR) antibody fragments from immunized sharks by phage display. Mol. Immunol. 2003, 40, 25–33. [CrossRef] 38. Sidhu, S.S. Engineering M13 for phage display. Biomol. Eng. 2001, 18, 57–63. [CrossRef] 39. Mandrup, O.A.; Friis, N.A.; Lykkemark, S.; Just, J.; Kristensen, P. A Novel Heavy Domain Antibody Library with Functionally Optimized Complementarity Determining Regions. PLoS ONE. 2013, 8, e76834. [CrossRef] [PubMed] 40. Abraham, R.; Buxbaum, S.; Link, J.; Smith, R.; Venti, C.; Darsley, M. Determination of binding constants of diabodies directed against prostate-specific antigen using electrochemiluminescence-based immunoassays. J. Mol. Recognit. 1996, 9, 456–461. [CrossRef] 41. Schofield, D.J.; Pope, A.R.; Clementel, V.; Buckell, J.; Chapple, S.D.; Clarke, K.F.; Conquer, J.S.; Crofts, A.M.; Crowther, S.R.; Dyson, M.R.; et al. Application of phage display to high throughput antibody generation and characterization. Genome Biol. 2007, 8, R254. [CrossRef][PubMed] 42. Hoogenboom, H.R.; de Bruïne, A.P.; Hufton, S.E.; Hoet, R.M.; Arends, J.-W.; Roovers, R.C. Antibody phage display technology and its applications. Immunotechnology. 1998, 4, 1–20. [CrossRef] 43. O’Connell, D.; Becerril, B.; Roy-Burman, A.; Daws, M.; Marks, J.D. Phage versus phagemid libraries for generation of human monoclonal antibodies. J. Mol. Biol. 2002, 321, 49–56. [CrossRef] 44. Rondot, S.; Koch, J.; Breitling, F.; Dübel, S. A helper phage to improve single-chain antibody presentation in phage display. Nat. Biotechnol. 2001, 19, 75–78. [CrossRef][PubMed] 45. Vieira, J.; Messing, J. Production of single-stranded plasmid DNA. In Methods in Enzymology; Recombinant DNA Part D; Academic Press: Cambridge, MA, USA, 1987; Volume 153, pp. 3–11. 46. Kristensen, P.; Winter, G. Proteolytic selection for protein folding using filamentous bacteriophages. Fold. Des. 1998, 3, 321–328. [CrossRef] 47. Lou, J. Affinity Maturation by Chain Shuffling and Site Directed Mutagenesis. In Antibody Engineering, 2nd ed.; Springer: Berlin/Heidelberg, Germany, 2010; Volume 1, pp. 377–396. 48. Laustsen, A.H.; Lauridsen, L.P.; Lomonte, B.; Andersen, M.R.; Lohse, B. Pitfalls to avoid when using phage display for snake toxins. Toxicon. 2017, 126, 79–89. [CrossRef][PubMed] 49. Laustsen, A.H. Recombinant Antivenoms, 1st ed.; University of Copenhagen: Copenhagen, Denmark, 2016; ISBN 978-87-93086-61-6. 50. Stewart, C.S.; MacKenzie, C.R.; Hall, J.C. Isolation, characterization and pentamerization of alpha-cobrotoxin specific single-domain antibodies from a naïve phage display library: Preliminary findings for antivenom development. Toxicon. 2007, 49, 699–709. [CrossRef][PubMed] 51. Marks, J.D. By-passing immunization. J. Mol. Biol. 1991, 222, 581–597. [CrossRef] 52. Hawkins, R.E.; Russell, S.J.; Winter, G. Selection of phage antibodies by binding affinity: Mimicking affinity maturation. J. Mol. Biol. 1992, 226, 889–896. [CrossRef] 53. Lowe, D.; Wilkinson, T.; Vaughan, T.J. Affinity Maturation Approaches for Antibody Lead Optimization. In Antibody ; Molecular Medicine and Medicinal Chemistry; Imperial College Press: London, UK, 2011; Volume 4, pp. 85–119, ISBN 978-1-84816-628-8. 54. Meng, J.; John, T.R.; Kaiser, I.I. Specificity and binding affinity of an anti-crotoxin combinatorial antibody selected from a phage-displayed library. Biochem. Pharmacol. 1995, 50, 1969–1977. [CrossRef][PubMed] 55. Kulkeaw, K.; Sakolvaree, Y.; Srimanote, P.; Tongtawe, P.; Maneewatch, S.; Sookrung, N.; Tungtrongchitr, A.; Tapchaisri, P.; Kurazono, H.; Chaicumpa, W. Human monoclonal ScFv neutralize lethal Thai cobra, Naja kaouthia, neurotoxin. J. Proteom. 2009, 72, 270–282. [CrossRef][PubMed] 56. Rodríguez-Rodríguez, E.R.; Olamendi-Portugal, T.; Serrano-Posada, H.; Arredondo-López, J.N.; Gómez-Ramírez, I.; Fernández-Taboada, G.; Possani, L.D.; Anguiano-Vega, G.A.; Riaño-Umbarila, L.; Becerril, B. Broadening the neutralizing capacity of a family of antibody fragments against different toxins from Mexican scorpions. Toxicon. 2016, 119, 52–63. [CrossRef][PubMed] Toxins 2018, 10, 236 15 of 15

57. Pucca, M.B.; Zoccal, K.F.; Roncolato, E.C.; Bertolini, T.B.; Campos, L.B.; Cologna, C.T.; Faccioli, L.H.; Arantes, E.C.; Barbosa, J.E. Serrumab: A human that counters the biochemical and immunological effects of Tityus serrulatus venom. J. Immunotoxicol. 2012, 9, 173–183. [CrossRef][PubMed] 58. Pucca, M.B.; Cerni, F.A.; Peigneur, S.; Arantes, E.C.; Tytgat, J.; Barbosa, J.E. Serrumab: A novel human single chain-fragment antibody with multiple scorpion toxin-neutralizing capacities. J. Immunotoxicol. 2014, 11, 133–140. [CrossRef][PubMed] 59. Pessenda, G.; Silva, L.C.; Campos, L.B.; Pacello, E.M.; Pucca, M.B.; Martinez, E.Z.; Barbosa, J.E. Human scFv antibodies (Afribumabs) against Africanized bee venom: Advances in melittin recognition. Toxicon. 2016, 112, 59–67. [CrossRef][PubMed] 60. Tung, C.-P.; Chen, I.-C.; Yu, C.-M.; Peng, H.-P.; Jian, J.-W.; Ma, S.-H.; Lee, Y.-C.; Jan, J.-T.; Yang, A.-S. Discovering neutralizing antibodies targeting the stem epitope of H1N1 influenza hemagglutinin with synthetic phage-displayed antibody libraries. Sci. Rep. 2015, 5, 15053. [CrossRef][PubMed] 61. Rosenfeld, R.; Alcalay, R.; Mechaly, A.; Lapidoth, G.; Epstein, E.; Kronman, C.; J Fleishman, S.; Mazor, O. Improved antibody-based ricin neutralization by affinity maturation is correlated with slower off-rate values. Protein Eng. Des. Sel. 2017, 30, 611–617. [CrossRef][PubMed] 62. Bakker, A.B.H.; Marissen, W.E.; Kramer, R.A.; Rice, A.B.; Weldon, W.C.; Niezgoda, M.; Hanlon, C.A.; Thijsse, S.; Backus, H.H.J.; de Kruif, J.; et al. Novel human monoclonal antibody combination effectively neutralizing natural rabies virus variants and individual in vitro escape mutants. J. Virol. 2005, 79, 9062–9068. [CrossRef][PubMed] 63. Laustsen, A.H.; María Gutiérrez, J.; Knudsen, C.; Johansen, K.H.; Bermúdez-Méndez, E.; Cerni, F.A.; Jürgensen, J.A.; Ledsgaard, L.; Martos-Esteban, A.; Øhlenschlæger, M.; et al. Pros and cons of different therapeutic antibody formats for recombinant antivenom development. Toxicon 2018, 146, 151–175. [CrossRef][PubMed]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).