Transgenic and Diet-Enhanced Silk Production for Reinforced Biomaterials: A Metamaterial Perspective.

Silk fibers, which are protein-based biopolymers produced by spiders and silkworms, are fascinating biomaterials that have been extensively studied for numerous biomedical applications. Silk fibers often have remarkable physical and biological properties that typical synthetic materials do not exhibit. These attributes have prompted a wide variety of silk research, including genetic engineering, biotechnological synthesis, and bioinspired fiber spinning, to produce silk proteins on a large scale and to further enhance their properties. In this review, we describe the basic properties of spider silk and silkworm silk and the important production methods for silk proteins. We discuss recent advances in reinforced silk using silkworm transgenesis and functional additive diets with a focus on biomedical applications. We also explain that reinforced silk has an analogy with metamaterials such that user-designed atypical responses can be engineered beyond what naturally occurring materials offer. These insights into reinforced silk can guide better engineering of superior synthetic biomaterials and lead to discoveries of unexplored biological and medical applications of silk. Expected final online publication date for the Annual Review of Biomedical Engineering, Volume 22 is June 4, 2020. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

Aside from silkworms and spiders, silk proteins have been produced in several different ways ( Table 1). Several alternative host systems (e.g., microorganisms, plants, and dairy animals) have been successfully exploited to efficiently produce spider silk proteins (24,27,29,33,36,38,44,54,80). Recently, silklike polypeptides have been produced by solid-phase and chemoenzymatic synthesis methods (25,26,63,83). Several engineering technologies using regenerated or 80 Leem •   Advances in silk-based material production using chemical modification and chemoenzymatic synthesis methods 83 synthetic silk proteins have been used to spin and pull silk fibers in a similar or superior manner that maintains the mechanical properties of natural spider silk fibers ( Table 2) (31,48,57,64,69,79). Given this wealth of information, we focus primarily on silkworm-produced silk with properties reinforced using silkworm transgenesis and functional additive diets. Transgenic silkworms have been exploited to produce recombinant proteins of interest in a scalable manner (96)(97)(98)(99)(100). Even with considerable research on synthetic silk protein production, silkworm transgenesis using silkworms as hosts is still considered an economical and practical production platform. Obviously, the genetic engineering of silkworms can offer additional functionalities for unique biophysical and biochemical properties. Several functional additives have also been incorporated into silk via direct feeding or injection, taking advantage of silkworms' open circulatory system (101)(102)(103)(104)(105)  In this review, we cover transgenic and diet-enhanced silk production with a focus on reinforced silk, given that many review articles about host cell-based silk synthesis, chemoenzymatic silk polymerization, and bioinspired silk spinning methods (24-27, 29, 31, 36, 38, 44, 48, 54, 57, 63, 64, 69, 79, 80, 83) are already available (Tables 1 and 2). To the best of our knowledge, there is no systematic review article on silkworm-based production of reinforced silk, although recent three reviews cover some aspects of silkworm transgenesis and nanomaterial feeding (72,82,106). We also cover research showing that reinforced silk is analogous to conventional metamaterials. Metamaterial research involves constructing materials to have user-designed atypical physical properties that often do not exist in nature (107)(108)(109)(110)(111)(112)(113).
First, we summarize the unusual physical properties of silk and important silk production methods. Second, we explain the basic methods of silkworm transgenesis and direct feeding of biologically friendly nanomaterials as scalable silk production platforms. Third, we discuss recent biomedical uses of fluorescent silk, mechanically reinforced silk, reactive oxygen species (ROS)generating silk, and artificial peptide-expressing silk. Finally, we present an outlook based on the current state of progress in silk research. We envision that an enhanced understanding of silkworm-based silk with superior physical and biological properties will allow us to explore new directions for the development of synthetic fibers and basic biomaterial research.

Remarkable Mechanical Properties of Spider Silk Fibers
Although silk is often considered to refer only to cocoons spun by silkworms, spider silk fibers have inspired countless scientific studies aiming to better understand silk's unusual mechanical properties and eventually produce superior synthetic fibers (1,3,28,(114)(115)(116)(117)(118)(119). Spiders have evolved the ability to produce as many as six or seven different types of silk fibers that vary in tensile strength and elasticity. Both dragline and flagelliform fibers have outstanding mechanical properties that permit the absorption of more energy prior to breaking than nearly any other common material ( Table 3). In particular, spider dragline silk fibers exhibit a unique combination of low density, high tensile strength, and extreme extensibility, resulting in superior toughness (87,119,120). Despite decades of research, it is still an engineering challenge to simultaneously realize these three mechanical properties (i.e., density, tensile strength, and extensibility) in synthetic fibers. Note that recent advances in biomimetic spinning dopes have enabled some 82 Leem •   regenerated artificial silk fibers to outperform natural spider silk fibers in terms of toughness (89,121). In addition, variations among different types of spider fibers are notable, as they encompass a fivefold range of tensile strength and a nearly 50-fold range of elongation ( Table 3).
In general, the superior mechanical properties of spider silk have motivated extensive research on silk.

Silk Proteins and Structures
The remarkable properties and characteristics of animal-produced silk are attributable to its constituent proteins and nanostructures. First, the nanostructures of silk fibers produced by silkworms and by spiders are very similar, as both are of core-shell type (2). Silkworms secrete silk fibroin synthesized in the silk glands through spigots of the spinnerets (i.e., silk-spinning organs). Subsequently, this insoluble protein crystallizes into nanofibrils in contact with the air while the nematic silk proteins are pulled out under shear stress and dehydration conditions; it is then assembled into fibroin filaments, with two filaments glued together with sericin into a silkworm cocoon (89,115,118,122,123). As a result, multiple parallel nanofibrils along the fiber axis form inside a single silk fiber. The diameter of silk fiber threads is dependent on the type and species of silkworm or spider (124). Typically, the silk fiber diameter of Bombyx mori silkworms, consisting of two core filaments coated with sericin, is approximately 20 μm. Spider dragline silk fibers have a diameter of 3-5 μm. The lustrous or silvery aspect of natural silk fibers can be explained by interactions of this exquisite nanostructure with light (18). Second, silk fibroin produced by domestic silkworms (B. mori) consists of heavy-chain (molecular weight, ∼350 kDa) and light-chain (molecular weight, ∼25 kDa) proteins covalently linked by a disulfide bond at the C terminus of the two subunits (22,(124)(125)(126). The molecular weight of spidroin proteins produced by orb-weaving spiders (araneids) is in the range of 250-400 kDa (27,34,124,127). The protein backbone (i.e., amide group) in the primary structure of fibroin and spidroin is composed of highly repetitive amino acid sequences. Both silkworm silk and spider dragline silk have the highly conserved and repetitive nature of protein sequences (2,114). Their secondary structures of fibroin and spidroin include β-sheet, random coil, α-helix, and antiparallel β-sheet secondary structures (123,124). The high content of α-helices and random coils in silk fibers is associated with higher elongation at break and higher toughness modulus, as the αhelices and random coil conformation consist of easily movable chains (11,128). Because animalproduced silk has the complexity of amino acid sequences and the reactive functional side groups of amino acids, it remains challenging to produce synthetic proteins that can mimic natural silk (23,83,114,119,124,129). Overall, an improved understanding of the protein structures and functions can provide a foundation to custom-design unique synthetic spider silk with mechanical properties ideally suited for specific applications.

Silkworm-and Spider-Free Methods of Producing Silk Proteins
Silk proteins have been produced by transplantation of silk-making DNA into a variety of host systems, including microorganisms, plants, and even goats, and by enzymatic or chemical synthesis of polypeptides (Table 1). Specifically, the isolation of spider silk gene sequences enables production of recombinant spider silk proteins in several different host systems, including bacteria (130,131), yeast (132), baculovirus/insect systems (133,134), mammalian cells (135), transgenic plants (136), and transgenic animals (137,138). Cloning and sequencing comparisons of complementary DNA encoding spider silk proteins successfully resulted in the identification of major ampullate spidroin 1, major ampullate spidroin 2, and flagelliform silk proteins of Nephila clavipes (139)(140)(141). The protein products of these genes are polymers consisting of highly repetitive blocks of amino acid sequence motifs, and specific sequence motifs are strongly correlated with the mechanical properties of spider silk fibers (116,142). One of the most common hosts is bacteria (e.g., Escherichia coli), which can be scaled up to industrial production. Key challenges to this approach include the bacterium's limited expressible gene size and distinct codon usage as well as the removal of repetitive sequences (143). In addition, biosynthesis of silklike polypeptide materials using protein engineering and chemoenzymatic methods has been successfully demonstrated for industrial production on a large scale (25,26,63,83). Note that after extraction of silk proteins, fiber spinning and pulling are required because of the insoluble characteristics of silk proteins ( Table 2).

Bioinspired Spinning Methods for Silk Fibers
When silk proteins are produced by biological or chemical synthesis, spinning and pulling of reconstituted silk fibers are an important step in engineering the same level of mechanical and physical properties of natural spider dragline silk. Several bioinspired spinning methods for constructing artificial silk fibers have been successfully demonstrated by mimicking the natural spinning process, using solution-based silk proteins (31,48,57,64,69,79). Silk fibers can be spun in a variety of methods, including electrospinning (31,59,144), wet spinning (26,145), dry spinning (64,146,147), self-assembly (148,149), and microfluidics (57, 150, 151) ( Table 2). The resultant mechanical properties of the artificial silk fibers are dependent on spinning process conditions, including solvents, coagulants, draw ratios, and other parameters (e.g., pH, temperature, viscosity, voltage, polypeptide molecular weight, blending, metal ions) (64,69). Typically, artificial silk fibers have mechanical properties comparable to those of silkworm silk fibers (64,151). More importantly, recent advances in bioinspired spinning of silk fibers have enabled artificial silk to outperform natural spider silk fibers in terms of toughness (89, 121).

Silkworm Transgenesis
The obvious host for expressing silk is the silkworm (B. mori) itself, which can produce recombinant silk cocoons on a large scale. One of the advantages of silkworm transgenesis is that the engineering demands of artificial fiber spinning and pulling can be minimized as silkworms spin silk fibers. The first successful transformation of B. mori was accomplished using the B. mori actin 3 promoter (BmAc3) to drive expression of the transposase on a helper plasmid, pHA3PIG, together with a piggyBac vector encoding a BmAc3-controlled gene encoding enhanced green fluorescent protein (eGFP) (100). The successful transformation was clearly evidenced by wholebody fluorescence in the F1 generation. Subsequent studies explored the utility of alternate promoters for expression and detection of fluorescent protein markers, including the 3×P3 eye-specific promoter (152) and the silk fibroin light-chain promoter (153,154). Specifically, the  genetic transformation of B. mori enables rapid functional characterization of silk gland promoters for silk protein production (Figure 1) (96). Silk gland-specific promoters exploit these glands for recombinant protein production and secretion in either tight or loose association with silk fibers, permitting noninvasive collection and simplifying downstream purification of recombinant protein products. The utility of such promoters for expression and detection of protein markers, including the silk fibrohexamerin gene promoter (155), the silk fibroin heavy-chain gene promoter (125), and the sericin1 gene promoter (156,157), has been extensively studied. An experimental animal model has also been widely used for recombinant silk protein production (154). Another advantage is that transgenic silkworms can provide high-level expression and efficient secretion of several recombinant proteins (96)(97)(98)(99)(100). Silkworm transgenesis can be an economical means of producing recombinant human type III procollagen, human serum albumin, human acid fibroblast growth factor, and antibodies (153,158,159). Silkworms produce a large amount of silk protein (0.2-0.5 g per worm), and their speed of protein biosynthesis is 10 6 -fold greater than that of mammalian cultured cells (160). Cost-effective breeding and relatively easy production scaleup also make transgenic silkworms a viable production system for biopharmaceuticals and other proteins, such as spider silk-based fibers. For example, the cost of silkworm rearing is less than five cents per larva, and it takes approximately 60 days to generate transgenic silkworms (160).

Direct Feeding of Artificial Additives
Directly feeding silkworms or spiders artificial additives can be a practical way to produce reinforced silk fibers (102,103,(162)(163)(164)(165)(166). In this approach, silkworms can easily produce silk containing functional additive nanomaterials. This is possible in part because silkworms have an open circulatory system. All of their organs float in hemolymph, which is a combination of lymph and blood cells that surrounds all tissues (Figure 2a) (101,167). In other words, the unique anatomy of silkworms is highly useful for producing functional silk by direct feeding methods (oral exposure and intake) of nanomaterials (Figure 2b). Nanoscale additives can diffuse out of the alimentary canal into the hemolymph and then into the glands and other tissues. Specifically, if nanomaterials are injected orally, they are absorbed by the digestive tract, pass through the digestive tract membrane barrier, and are circulated into the hemolymph and blood cells. They are transported concomitantly to the floating tissue compartments in the hemolymph, such as the silk glands, fat body, and Malpighian tubule. Several different functional nanomaterials (e.g., dye molecules, graphene, CNTs, titanium dioxide, and quantum dots) have been successfully incorporated into silk via direct feeding or injection, resulting in enhanced mechanical, thermal, electrical, and optical properties (101-105).

Fluorescent Silk
The production of fluorescent silk genetically hybridized with fluorescent proteins is a common example of silkworm transgenesis (168,169). A transformation gene vector is constructed by genetically encoding a fluorescent protein [e.g., eGFP, DsRed (derived from Discosoma spp.), Kusabira Orange, mKate2, or enhanced yellow fluorescent protein] as a color gene into the silkworm genome via the gene-splicing piggyBac transposase method (168)(169)(170)(171). Upon optical excitation using a common light source (e.g., light-emitting diodes), strong fluorescent emission intensity from fluorescent silk can be detected that corresponds to the fluorescent proteins in silk (Figure 3a,b). For widespread use in fabrics and textiles, it is important to maintain the mechanical strength of the transgenic fluorescent silk. In some cases, the mechanical properties were slightly diminished compared with those of nontransgenic silk (e.g., in the commercial race C146 × J137) (168). This side effect can be attributed to the presence of fluorescent proteins fused with fibroin heavy-chain N-and C-terminal domains, which may disturb silk crystallization. Nevertheless, transgenic fluorescent silk cocoons can be reeled and woven into fabrics by an automatic reeling machine without loss of fluorescence (168).
Silk fibroin transgenically hybridized with fluorescent proteins can be processed and regenerated into various forms with nano-and microstructures for applications in optics, electronics, 86 Leem  optoelectronics, and medicine, as it is transparent, mechanically stable, edible, biocompatible, and implantable in the human body (2,7,69). For example, regenerated fluorescent silk fibroin was used for surgical inspection by enabling localization of gastrointestinal fistula lesions (169). Regenerated solutions and particles of fluorescent silk fibroin were also applied for bioimaging (169). Through the use of an eGFP fluorescent silk fibroin solution, an inspection of esophageal perforations was performed, permitting intraoperative surgical field imaging in an animal model. Note that the conventional fibroin processing methods, which include boiling, are inappropriate for the process of regenerating fluorescent silk because fluorescent proteins are highly susceptible to denaturation from high temperature and pH values (5,172,173). Silk fibroin fused with fluorescent proteins must be processed at low temperatures (≤60°C) in order to avoid heat-induced denaturation of the fluorescent proteins. Another way to produce fluorescent silk is to utilize direct feeding methods that rely on in vivo uptake of fluorescent dye molecules into the silk glands (Figure 3c,d) (104,167). Typically, fluorescent colored silk can be achieved using a modified diet of mulberry leaves containing dye molecules. Several different dyes have been successfully used, including rhodamine dyes (i.e., sulforhodamine 101, RhB, Rh101, Rh110, and Rh116) (104) and azo dyes (i.e., Brilliant Yellow, Congo Red, Acid Orange G, Acid Orange II, Mordant Black 17, Direct Acid Fast Red, and Sudan III) (167), all of which are common in the textile industry. Again, it is important to ensure the mechanical strength of dye-fed fluorescent silk. The typical average tensile strength and strain of colored silk range from 406 to 454 MPa and from 23.7% to 26.5%, respectively, which are comparable to those of typical white silk (approximately 455 MPa and 27.1%, respectively) (11,104). The combination of reliable mechanical properties and high biocompatibility allowed the construction of scaffolds made of intrinsically RhB dye-fed silk, demonstrating the feasibility of visualizing human colon fibroblast (CCD-112CoN) and lung cancer cell (A549) lines (104). This result indicates that the addition of fluorescent molecular dyes did not affect the crystallinity of silk fibers, relatively maintaining the mechanical properties of natural silk.
For fluorescent dye molecules to be successfully integrated with silk, hydrophilicity and hydrophobicity must be balanced, which can be estimated by the partition coefficient (i.e., measure of hydrophobicity) of the dye molecules (104,167). The dye's partition coefficient determines its preferential association with either sericin or fibroin in silkworm glands and, ultimately, with silk fibers. For example, dyes with partition coefficients of 0-1.5 (e.g., Rh110, Rh116, Acid Orange II, Mordant Black 17, and Direct Acid Fast Red) are well expressed in silk fibers. Hydrophilic dyes with negative partition coefficients (e.g., fluorescein sodium, sulforhodamine 101, Brilliant Yellow, Congo Red, and Acid Orange G), as well as hydrophobic dyes with extremely high partition coefficients (e.g., Sudan III), are not well expressed. Acridine Orange, with a partition coefficient of 1.8, is also not colored in silk fibers. Because Acridine Orange's nonamphiphilic structure does not allow tuning of hydrophobicity upon the formation of dimers (174), molecules of this dye cannot penetrate hydrophilic sericin and thus are retained in silk gland cells. Although Rh101 and RhB are hydrophobic dyes with high partition coefficients greater than 2, they accumulate in high concentrations in silk fibers. This finding might be attributed to the formation of dimers that can result in more efficient transfer of dye molecules into silk glands (175)(176)(177). In other words, hydrophobic dyes with good water solubility are promising candidates for producing fluorescent silk.
Another factor that affects the absorption of dye molecules into silk is molecular weight. Higher molecular weights limit the absorption of dyes into the silkworm's biochemical pathways. Dyes accumulate in the peritoneal membrane of the alimentary canal. Thus, molecular weight lower than 400 g/mol is essential for effective transport of dyes into the biochemical pathways of the silkworm body and in the production of naturally dyed silk fibers (167). In addition, the solubility 88 Leem •  of dyes in hemolymph and the differential permeation through the linings of various types of tissue are likely to control the transfer of dye molecules to the silk glands.

Mechanically Enhanced Silk
While several transgenic microorganisms, plants, and animals have been successfully used to produce spider silk proteins, none of these hosts were less expensive than spiders to scale up or naturally equipped to spin silk fibers. One approach that overcomes these limitations is to use silkworms as surrogate hosts for spider silk production. Indeed, silkworms are preferred as a host system for producing recombinant spider silk fibers because they are naturally equipped to spin silk fibers. The relationship between spider silk protein structures and mechanical functions is relatively well understood (142), allowing us to design artificial fibers for specific biomedical applications. Specifically, silkworm transgenesis for spider silk production can be efficiently performed using piggyBac vectors (100,178,179) or CRISPR/Cas9 direct gene replacement (180,181). For example, recombinant protein production can be targeted to silk glands with tissue-specific promoters (153,155,(182)(183)(184). Natural and customized spider silk fibers have been produced by assembling DNA sequences encoding synthetic spider silk proteins with mixed motifs for expression in transgenic silkworms (127,181,185,186). The more recent introduction of CRISPR/Cas9 has led to direct replacement of the silkworm's endogenous fibroin genes with synthetic spider silk genes; the properties of the resulting silk fibers closely mimic those of native spider silk fibers (180). Direct feeding of silkworms or spiders with functional nanomaterials in order to produce reinforced silk has been demonstrated with carbon-containing materials (e.g., graphene and CNTs) (102,(162)(163)(164), titanium dioxide (103), metal nanoparticles (e.g., silver and copper) (165), and ion precursors (e.g., Ca 2+ and PO 4 3− ) (166). Table 4 summarizes the mechanical properties of silk fibers from silkworms or spiders fed various artificial diets containing nanomaterials. One of the underlying mechanisms is that such nanomaterials can act as knots in fibers, resulting in cross-linked networks with crystallites, lower crystallinity, and higher elongation at break (187). Typically, artificial diets were generally fed to silkworms and spiders by spraying aqueous solutions containing reinforcing nanomaterials onto normal diets. Control groups of silkworms or spiders were routinely fed normal diets (e.g., mulberry leaves). Importantly, simple feeding with such functional nanomaterials resulted in improved mechanical properties of the silk fibers, with stronger fracture strength and higher elongation at break, that did not strongly depend on silkworm or spider strain, type of mulberry leaf, or rearing environment.
Although a functional nanomaterial diet can be relatively well incorporated into silk fibers, there are several caveats. First, the key parameter for successful integration of a nanomaterial with the silk glands is the size of the nanomaterial, because diffusion of large-sized nanoparticles is significantly limited by the silkworm's open circulatory system, which typically requires sizes smaller than 30 nm (101,188). Second, the presence of nanocarbon materials (e.g., graphene and CNTs) in the silk matrix may hinder the transformation of α-helices and random coils into β-sheet structures, resulting in noncovalent interactions between nanocarbon materials and silk fibroin through physical adsorption (102). Third, excessive nanomaterials in silk can reduce the mechanical strength of silk fibers, because the nanomaterial-conjugated nanofibrils may aggregate and act as defects instead of link points (102,103). Fourth, silver nanoparticles are toxic and should be used with caution (189,190). They can affect silkworms' metabolic cycle, signal transduction, apoptosis, and ion transport, weakening metabolic function and increasing energy storage and utilization (191). Silver nanoparticles can also degrade the ability of silkworms to withstand oxidative stress, interfere with programmed cell death, and attenuate the expression of detoxification proteins.

Reactive Oxygen Species-Generating Silk
Functional silk with disinfecting properties has received considerable attention for a variety of antimicrobial applications (76,80). Typical approaches to producing it rely on treatment (i.e., simple attachment or coating) of organic or inorganic nanomaterials (e.g., titanium dioxide, zinc oxide, CNTs, silver/gold nanoparticles, natural dyes) (192)(193)(194) to generate ROS upon light illumination. In this respect, silkworm transgenesis can provide a more effective hybridization alternative by genetically encoding phototoxic fluorescent proteins into silk proteins (161,195,196). Integration of phototoxic fluorescent proteins such as GFP, KillerRed, SuperNova, TurboGFP, and mKate2 is another way to generate and release ROS upon light excitation (197)(198)(199)(200). The phototoxic action of red fluorescent proteins (e.g., KillerRed, SuperNova, and mKate2) is known to originate from a cleftlike opening filled with water molecules, allowing for enhanced generation and release of ROS (201,202). The exact types of ROS vary among different red fluorescent protein variants, depending on the type of photoreaction and the concentration of local molecular oxygen (i.e., electron acceptor) (197). Superoxide (O 2 •− ) and other ROS, such as hydroxyl radical (−OH • ) and hydrogen peroxide (H 2 O 2 ), are generated via the type I photoreaction by electron transfer to and from the substrate in the excited triplet state (T 1 ). The most common electron acceptor is molecular oxygen [O 2 ( 3 g − )] (i.e., 3

O 2 ). The resultant O 2
•− can further interact with its surroundings to produce other reactive oxygenated products (e.g., −OH • and H 2 O 2 ). In contrast, singlet oxygen ( 1 O 2 ) [i.e., O 2 ( 1 g )] is generated by energy transfer from T 1 to 3 O 2 in a type II photoreaction (197). mKate2, a phototoxic far-red fluorescent protein, has been genetically hybridized with silk using the piggyBac method (Figure 4) (161). Specifically, in order to combine mKate2 and silk protein, the mKate2 gene was fused with N-terminal and C-terminal domains of the fibroin heavy-chain promoter (pFibH), giving rise to a p3×P3-eGFP-pFibH-mKate2 transformation vector. Its antimicrobial activity was demonstrated by the inactivation of E. coli (DH5α) bacteria and the photodegradation of dye molecules (i.e., methylene blue) (161). Under visible (green) light illumination, the mKate2 silk generated two types of ROS, O 2 •− and 1 O 2 , by type I and II photoreactions, respectively. The ROS generated by mKate2 silk significantly inactivated the bacteria and photodegraded the dye. This result supports the idea that phototoxic fluorescent protein-expressing silk can offer an alternative means of generating ROS that is comparable to visible light-driven plasmonic photocatalysis. Note also that ROS generation and photoelectric conversion are two sides of the same coin in terms of redox reactions and electrochemistry, as photoinduced electrons are closely related to photoelectric conversion. Thus, silk fused with phototoxic fluorescent proteins can have basic current-voltage characteristics that can be exploited for biological photosensors and energy-harvesting devices (203)(204)(205)(206).

Antimicrobial Peptide-and Noncanonical Peptide-Expressing Silk
Silkworm transgenesis has been expanded to incorporate antimicrobial peptides into silk proteins. Antimicrobial peptide-expressing silk is a viable way to impart a disinfection functionality to silk. A transgenic silkworm carrying a recombinant B. mori fibroin light chain fused to GFP was successfully hybridized with antimicrobial peptide cecropin B (CEC B) by use of a genetargeting technique (i.e., homologous recombination) (195). Strong antimicrobial activity of CEC B-expressing silk cocoons was demonstrated by counting the number of E. coli colonies. Different transgenic silkworms were able to express CEC B or Moricin (MOR) antimicrobial peptides as a result of piggyBac-mediated germline transformation (i.e., pFibH-CEC B1/3×P3DsRed and pFibH-MOR/3×P3DsRed constructs) (196). Transgenic silk fused with CEC B or MOR antimicrobial peptides also inhibited the growth of E. coli. Furthermore, a silk yarn maintained antibacterial activity against E. coli. Importantly, these results indicate that the silk fibers retained active CEC B and MOR antimicrobial molecules after the degumming process. The inactivation mechanisms of antimicrobial peptides are generally explained by two major actions: direct killing and immune modulation ( Figure 5). Antimicrobial peptides bind to the bacterial membrane through electrostatic interactions, either to disrupt the membrane or to enter the bacterium in order to inhibit intracellular functions. Some antimicrobial peptides also modulate host immunity by recruiting or activating immunocytes or by influencing Toll-like receptor recognition of microbial products and nucleic acids that are released upon tissue damage (207,208).  Antimicrobial functionalization of silk with genetic hybridization of peptides. Two antimicrobial actions involving antimicrobial peptides are direct killing and immune modulation (208) More interestingly, silkworm transgenesis has recently been used to express noncanonical peptides and amino acids that do not exist in nature. Azidophenylalanine-expressing silk was obtained from a genetically engineered silkworm (B. mori) by constructing a piggyBac plasmid vector (209). Azido groups in living systems play an important role in bioorthogonal chemistry, in which a chemical reaction occurs without interfering with natural biochemical processes (210,211). In order to screen azido derivative azidophenylalanine-recognizing variants, a pool of B. mori phenylalanyl-transfer RNA synthetase (BmPheRS) variant genes was introduced into E. coli cells (209). Selected BmPheRS variants were then examined for possible adverse effects in B. mori cultured cells. Finally, only safe variants were expressed in the silk glands of transgenic B. mori larvae to produce azido-functionalized silk. The feasibility of bioorthogonal chemistry using azido-functionalized silk was successfully demonstrated in a click modification experiment (209), while maintaining the basic mechanical properties for use in other general applications.

OUTLOOK AND CONCLUSION
Silkworm engineering via transgenesis and diet-enhanced methods can produce silk with reinforced or superior material properties that often do not exist in naturally occurring materials. This idea has an analogy to conventional metamaterials. Research on metamaterials and metastructures involves the construction of materials to offer user-designed atypical responses that are often absent in nature. Optical metamaterials are often realized through control of their electric and magnetic properties, and are not limited to negative refractive index and invisibility cloaks (107,109,110,113). On one hand, mechanical metamaterials take advantage of artificially engineered structures to determine mechanical properties, relatively independently of material compositions (108,111,112,212,213). On the other hand, conventional metamaterials are not intended for biomedical applications, and their construction relies primarily on exotic materials and on complex nanofabrication and nanomanufacturing. Often, metamaterials not only are limited by the material toxicity and biocompatibility but also are not amenable to scalable, economical, and eco-friendly production. For example, typical nanomanufacturing processes for nanoproducts require a large amount of toxic raw materials and fossil fuels and are often difficult to apply in sustainable mass production (214,215). In this respect, the concept of reinforced or superior silk can be extended to a type of metamaterial. In other words, reinforced silk produced by transgenic silkworms and silkworms fed with special diets could be considered biomedical metamaterials that are designed for direct use in biological and medical applications.
Reinforced or superior silk (metamaterial-like silk) produced by transgenic silkworms and functional diet additives can serve as an ab initio foundation for new opportunities. Biofactory and bioreactor strategies will be very useful for producing such silk in a scalable and eco-friendly manner, given that there is always a need for continuous and scalable production of nanobiomaterials. Incorporating the superior biocompatibility and bioresorbability of natural silk, such silk can be directly used in biological and medical applications. For example, silk integrated with desirable and enhanced functionalities can be used in wound dressings with monitoring or sensing features; tissue engineering scaffolds with antibacterial, anticoagulant, or anti-inflammatory features; vaccine manufacturing; photodynamic therapy; and bioenergy harvesting. In particular, antimicrobial peptides or ROS-generating fluorescent silk will offer exploitable and scalable photocatalyst-like biomaterials, potentially ruling out hazards associated with foreign semiconductor nanomaterials. Because interactions between light and fluorescent proteins are often understood on the basis of quantum mechanics (216-219), fluorescent protein-expressing silk could provide an alternative model system for studying quantum biology and quantum biophotonics. Silkworm transgenesis and diet-enhanced silk production could help guide synthetic biology approaches to enable the www.annualreviews.org • Transgenic and Diet-Enhanced Silk Production 93 , . • · � -Review in Advance first posted on March 11, 2020. (Changes may still occur before final publication.)