De Novo Autogenic Engineered Living Functional Materials Authors: Hoda M Hammad, Seth Swarnadeep, Erin C Jackson, Nicolas Burns, Hongyu Wang, Harrison Priode, Robert B Moore, Sanket Deshmukh, Avinash Manjula‐Basavanna, Anna M Duraj‐Thatte Published: 2026 ABSTRACT ABSTRACTAutogenic engineered living materials (ELMs) enable the in situ production and engineering of native extracellular matrix (ECM). However, existing autogenic ELMs remain limited in scope and functionality. Here, we present a versatile platform for de novo autogenic functional ELMs, leveraging protein mining, computational modeling, and synthetic biology. By analyzing 33,564 CsgA‐like homologs, we identify candidates for de novo ECM protein nanofibers. Using AlphaFold2 and molecular dynamics simulations, we elucidate the structural stability of these β‐solenoid proteins. By reprogramming the Escherichia coli curli machinery, we achieve the biosynthesis of CsgA‐like ELMs from non‐model bacteria, featuring up to a 9‐fold increased molecular weight and expanded β‐sheet repeat units. Furthermore, we fabricate macroscopic biomaterials with enhanced mechanical properties (a 3‐fold increase in storage modulus), and their extracellular fiber networks attenuate UV‐C irradiation, extending the survival of embedded cells by 5‐fold. We further demonstrate programmable functionalities, including 3D printability and selective binding to nanoparticles and antibodies. This work establishes a powerful framework for discovering, designing, and harnessing natural biomolecular systems to advance next‐generation autogenic ELMs.Keywords: curli nanofibers, engineered living materials, functional amyloids, protein hydrogels, protein structure prediction Introduction Engineered living materials (ELMs) are a rapidly growing field wherein living cells are engineered to produce materials with life‐like functionalities for sustainability, biomedicine, biosensing, biomining, and bioremediation applications [1, 2, 3, 4, 5, 6, 7, 8]. Based on the polymeric matrix, ELMs can be broadly classified into two types, namely, exogenic and autogenic ELMs [9, 10]. In exogenic ELMs, living cells are embedded in a synthetic/natural polymeric matrix obtained from an external source, whereas in autogenic ELMs, living cells are utilized/engineered for the in situ production of a native/functional polymeric matrix [11]. Evidently, autogenic ELMs are advantageous over exogenic ELMs, as they enable on‐demand in situ biosynthesis, self‐assembly, self‐organization, self‐regeneration, self‐regulation, environmental adaptability, and functionalization of the polymeric matrix [12, 13, 14]. Although the autogenic ELMs are crucial to realize the full potential of this emerging technology, there are only a few reported autogenic ELMs, emphasizing the need for innovative design strategies to (1) harness the programmability and biomanufacturing capabilities of living cells and (2) discover the wide variety of functionalities prevalent in the natural systems [15, 16, 17, 18, 19]. In the existing autogenic ELMs, the native extracellular matrix (ECM) of bacterial biofilms is utilized either directly or after functionalization. For example, the native protein nanofiber‐based ECMs of Escherichia coli and Bacillus subtilis, which comprise CsgA and TasA, respectively, have been genetically modified by fusing them with desired protein domains to obtain functional autogenic ELMs [15, 20, 21, 22] Alternatively, the native cellulose nanofiber‐based ECM of Komagataeibacter rhaeticus has been employed to obtain autogenic ELMs, but, unlike the above example of a protein‐based ECM, the genetic functionalization of cellulose/polysaccharide to controllably modulate their properties is still elusive [6, 23]. On the other hand, genetic modification of cell surface‐displayed proteins of E. coli, B. subtilis, and Caulobacter crescentus has resulted in autogenic ELMs. However, these surface‐displayed proteins are confined to the cell periphery, thereby limiting incorporation of functional domains and preventing the formation of a free‐standing three‐dimensional (3D) extracellular matrix. By contrast, secreted β‐solenoid nanofibers form a true 3D matrix that can be genetically functionalized at multiple positions along the nanofibers, enabling broader tunability of physical, chemical, and biological properties [12, 24, 25]. In the last 12 years, the ECM protein nanofibers of E. coli self‐assembled from CsgA are the most studied autogenic ELMs [7, 26]. These protein nanofibers, commonly known as curli, are attractive due to their (1) resistance to heat, solvents, pH, detergents, and denaturants, (2) mechanical properties, and (3) ability to modulate functionalities by genetically grafting the desired protein domains [27, 28]. CsgA, with its cross‐β structure, is energetically among the most stable protein folds, while the β‐solenoid architecture facilitates the positioning of the C‐ and N‐terminal domains on its periphery. As a result, the peptide/protein domains genetically grafted to CsgA via a flexible linker are often found to fold independently to their functional form (Figure S1) [26]. Moreover, the head‐to‐tail stacking of CsgA building blocks leads to the formation of curli nanofibers, and the genetically grafted peptide/protein domain gets displayed on its periphery to form the functional curli nanofibers, which have been demonstrated for various applications in the last several years [29, 30]. Thus, we hypothesized that the β‐solenoid architecture of the CsgA protein building block is crucial for the fascinating properties and applications of curli nanofibers‐based autogenic ELMs. In this work, we present de novo autogenic functional ELMs, wherein E. coli is engineered to produce de novo ECM of protein nanofibers (Figure 1). Motivated by the intriguing characteristics described above, we conducted protein database mining to discover CsgA‐like β‐solenoid proteins that can serve as de novo building blocks of synthetic ECM. We utilized the artificial intelligence (AI) tool AlphaFold2 to predict the structure of these β‐solenoid proteins, and the stability of the folded structure was deciphered using all‐atom (AA) molecular dynamics (MD) simulations in the explicit water model [31]. As these β‐solenoid proteins were from non‐model organisms thriving in different parts of the globe, we have harnessed the curli secretion machinery of the model organism E. coli to produce de novo ECM. Further, we employ these ECMs to produce functional materials like hydrogels and films and show that their physicochemical properties can be tailored. In addition, we demonstrate that the biological functions of these de novo ECM‐based ELMs could be controllably modified by genetically grafting the desired protein domains to facilitate specific binding to nanoparticles or antibodies. Schematic of the SECRETE platform to produce CsgA‐like β‐solenoid proteins for autogenic engineered living functional materials. CsgA‐like proteins discovered by mining the National Center for Biotechnology Information (NCBI) protein database were studied by using AlphaFold2 and molecular dynamics (MD) simulations to predict their structure. E. coli curli machinery was harnessed to produce de novo β‐solenoid proteins of non‐model bacteria to obtain materials with modular physicochemical properties, programmable biofunctionalities, and tunable mechanical properties. Parts of the schematics were created in BioRender. Created in BioRender. Duraj‐Thatte, A. (2026). Results Protein Mining and Structure Prediction CsgA is a 13 kDa protein of 131 amino acids with five cross‐β repeats, wherein the minimalistic curli repeat sequence is presented in the form X6QXGX2NX10 [32, 33]. To identify CsgA‐like proteins, we utilized the National Center for Biotechnology Information (NCBI) database to search for proteins having the X6QXGX2NX10 sequence, which resulted in 33 564 entries, and their lengths varied from 141 to 1390 amino acids. From these, we selected 50 protein sequences representing the full spectrum of identified lengths, and their protein structure was predicted using AlphaFold2, which revealed β‐solenoid architecture for all of them (Figure S2). Finally, for a detailed study, we selected five protein sequences of different lengths that belonged to diverse bacterial species, namely, Halomonas saliphila (Hs13‐CsgA), Alteromonas macleodii (Am18‐CsgA), Blastomonas sp. CACIA14H2 (Bc36‐CsgA), Erythrobacter longus (El43‐CsgA), and Ensifer sp. Root31 (Er46‐CsgA). These CsgA homologs, named according to their genus, species, and predicted number of cross‐β repeats, represent a range of ecological adaptations and functional properties, none of which had been experimentally validated for curli production to the best of our knowledge prior to this study (Figure S3 and Table S1). The 3D structure of CsgA and five other CsgA homologs predicted using AlphaFold2 revealed the signature sequence KR1‐Ω‐Ψ‐Ω‐Ψ‐Ω‐KR7 of cross‐β sheet structure, wherein Ω represents outward‐facing residues, while KR1, KR7, and Ψ represent inward‐facing residues (Figures 2 and S4). Each repeat unit in CsgA consists of two antiparallel β‐strands, β1 and β2, arranged in a cross‐β sheet structure, with each β‐strand made of seven residues from KR1 to KR7 that align across the entire folded monomer. These β‐strands within each repeat unit are connected by short loops. The first loop, connecting β1 to β2 within a repeat unit, is referred to as the intra‐repeat loop, while the loop connecting one repeat unit to the next is called the inter‐repeat loop. Both loops are typically composed of four residues, while the intra‐repeat loop is represented by the sequence X‐G‐X‐G, where X represents variable residues, and G signifies the conserved glycine, which allows for structural flexibility. In summary, the conserved β‐strand and loop sequence of each repeat unit follows the pattern: β1 strand (KR1‐Ω‐Ψ‐Ω‐Ψ‐Ω‐KR7), intra‐repeat loop (X‐G‐X‐G), β2 strand (KR1‐Ω‐Ψ‐Ω‐Ψ‐Ω‐KR7), inter‐repeat loop (X‐X‐X‐X), and so on. Different CsgA homologs vary in the number of repeat units: E. coli CsgA consists of five repeat units, while Hs13‐CsgA has 13, Am18‐CsgA has 18, Bc36‐CsgA has 36, El43‐CsgA has 43, and Er46‐CsgA has 46 repeat units (43 in the core β‐solenoid structure and three in the outlier region). Structures of β‐solenoid protein building blocks predicted by AlphaFold2. The predicted β‐solenoid protein models of (a) CsgA, (b) Hs13‐CsgA, (c) Am18‐CsgA, (d) Bc36‐CsgA, (e) El43‐CsgA, and (f) Er46‐CsgA. The longitudinal view of the β‐solenoid protein model shows the number of cross‐β repeats. Each β‐solenoid protein sequence is presented to showcase the conserved (N, Q, G indicate>50% of conservation probability) and non‐conserved residues (n, g indicate 25%–50% of conservation probability, and for <25%, it is indicated as X). On‐axis top view shows the β‐solenoid fold with inward‐facing residues (green), key residues (blue), non‐conserved residues (red), and outlier regions (orange). For clarity, only the top five cross‐β repeats are shown for the on‐axis top view. Within each β‐strand pair (β1 and β2) in a repeat unit, key residues (KR), located at the beginning (KR1) and the end (KR7) of each β‐strand, are reported to play a crucial role in stabilizing and ensuring proper folding of the monomer [32, 33]. In E. coli CsgA, KR1 of β1 is conserved as serine (S), while KR1 of β2 is asparagine (N), and KR7 of both β1 and β2 is glutamine (Q) [34]. However, these key residues KR1/KR7 in β1/β2 strands are conserved to various extents for the other five β‐solenoids, and the non‐conserved key residues, denoted as X in the consensus motif, are highlighted in red. The inward‐facing residues, Ψ, shown in green, are tightly packed between the β‐sheets, stabilizing the fibril core. These inward‐facing residues typically consist of hydrophobic amino acids (A, I, V, L, F), as well as serine (S) or threonine (T). Meanwhile, outward‐facing residues, Ω, shown in black, are usually hydrophilic and exposed to the surrounding environment, facilitating fibrillation. Additionally, our structural analysis revealed outlier regions (highlighted in orange and denoted as ∑), which extend beyond the canonical β‐strands and loop motifs, are observed in all five variants, except for E. coli CsgA. While the precise role of these outlier regions remains unclear, they may have functional implications or impact the overall architecture of the fibrils, warranting further investigation. Additional analysis sheds insights into the physicochemical characteristics of these six β‐solenoid proteins (Figure 3a–d; Table S2 and S3). The molecular weight of the proteins increased with a greater number of repeats, from 13.1 kDa for CsgA to 117.6 kDa for Er46‐CsgA, while their isoelectric points remained similar, ranging from 3.3 to 4.5. The predicted charge of CsgA at pH 7.4 was −6.3, and it increased by several folds for the five other β‐solenoid proteins, with the highest charge of −116.6 observed for El43‐CsgA. The hydrophobicity, as predicted by their GRAVY score (−0.7 to −0.1), indicated that all five β‐solenoid proteins are more hydrophobic than CsgA (Figure S5). Physicochemical characteristics of β‐solenoid proteins and their structural stability analysis by all‐atom molecular dynamics simulations. Comparison of the physicochemical characteristics of β‐solenoid proteins: (a) molecular weight, (b) isoelectric point, (c) net charge, and (d) hydrophobicity index. (e) Root mean square deviation (RMSD) of the β‐strands for β‐solenoid proteins throughout MD simulation. (f) Hydrogen bond autocorrelation function for β‐solenoid proteins. (g) Average normalized van der Waals, electrostatic, and total interaction energies (IE). IE1: between KR and KR, IE2: KR and all residues, and IE3: non‐KR and all residues, over the period of the production run. The energies were normalized per residue. Decoding β‐Solenoid Protein Structure Stability All six β‐solenoid proteins showed pLDDT (predicted local distance difference test) confidence scores between 71 and 89, which indicated a good backbone prediction (Figure S6). To investigate the conformational stability of individual proteins, we performed all‐atom (AA) molecular dynamics (MD) simulations starting from β‐solenoid structures predicted by AlphaFold2 (Video S1). These simulations were carried out in an explicit TIP3P water model for 500 ns at 300 K and 1 atm using the NPT ensemble. To assess the stability of β‐solenoid conformations, the simulation trajectories were analyzed for different structural—root mean square deviations (RMSD), root mean square fluctuations (RMSF), Ramachandran analysis, dynamical—hydrogen bond analysis, and energetic characteristics of β‐solenoid. Overall, this analysis provided molecular‐level insights into the stability and dynamic behavior of the β‐solenoid under physiological‐like conditions. For all six proteins, the root mean square deviation (RMSD) of all heavy atoms in the backbone and the side chains of amino acids in β‐sheets increases with increasing residues (initial equilibration of less than ∼5 ns), and it remains stable during the remainder of the simulations run (Figure 3e). RMSD of the entire protein showed slightly more perturbations and higher RMSD values, which could be due to fluctuations in the N‐terminal and outlier regions, which is consistent with the pLDDT scores (Figures S7 and S8). The residue‐wise root mean squared fluctuations (RMSF) were consistent with the RMSD analysis, showing that the N‐terminal and outlier residues had more fluctuations than the β‐sheet region residues (Figure S9). The Ramachandran plot showed a dense cluster of points with Φ ≈ −120° and 𝜓 ≈ 120° throughout 150 ns MD simulations, suggesting that the overall backbone dihedral angles are relatively preserved in the β‐sheet region, with minor presence of α‐sheets in all proteins (right α‐sheet: Φ = −60° and 𝜓 = −45°; left α‐sheet: Φ = 70° and 𝜓 = 35°; Table S4 and Figure S10) [35]. The hydrogen bond characteristics between all protein residue pairs forming hydrogen bonds were evaluated using a geometric criterion described in the Method section [36]. The hydrogen bond analysis indicates that the presence of bifurcated hydrogen bonds, in addition to hydrogen bonds in β‐sheets, may further contribute to the structural stability of the proteins [37]. A slower decay of the hydrogen bond autocorrelation function was observed for the five β‐solenoid proteins compared to CsgA, which indicates that the hydrogen bonds formed between amino acid pairs in the five β‐solenoid proteins are more stable compared to those in the CsgA protein (Figure 3f). A detailed structural analysis revealed that the five β‐solenoid proteins contain a higher number of charged residues (e.g., D, E, K, R) compared to CsgA (Figure S11). Additionally, the local environment of each residue (within 12 Å) is predominantly composed of glycine (G) and polar residues (e.g., N, Q, S, T) rather than non‐polar residues (e.g., I, L, M, F, W). The categorization of residues by polarity highlights the role of polar, non‐polar, and charged interactions in shaping the local environment, which is essential for understanding the protein's aggregation propensity and functional properties. The local environment of individual residues in the CsgA β‐solenoid protein and the five variants was analyzed. For each residue, the surrounding residues present within a 12 Å radius were identified, and their average distance was calculated. These data were aggregated by amino acid type, computing averages for surrounding residue counts over the last 50 ns of MD trajectory data. The results are presented as a heatmap in Figure S12, where circles represent the frequency of surrounding residue types and are categorized by residue polarity (polar, non‐polar, and charged). To investigate further the role of charged residues and their local environments, we calculated averaged non‐bonded interaction energies (IE) between KR and KR (IE1), between KR and all residues (IE2), and between non‐KR and all residues (IE3) (Figure 3g). Across all six proteins, the electrostatic interactions in IE1, IE2, and IE3 were generally more favorable (more negative) compared to van der Waals interactions but exhibited varying patterns. The electrostatic interactions in IE1 were consistent across most proteins, except for Bc36‐CsgA, which showed less favorable interactions (less negative). In contrast, electrostatic interactions in IE2 and IE3 displayed greater variability, with CsgA and Er46‐CsgA showing more favorable energies than the other proteins. Notably, IE2 electrostatic energies were more favorable than IE3. These results highlight that both charge placement and the local environment of these residues play an important role in determining electrostatic energies. Meanwhile, van der Waals energy contributions in IE1, IE2, and IE3 were distinct but consistent across all six proteins, likely due to their similar geometries [38]. Overall, MD simulations revealed that all six novel β‐solenoid proteins are stabilized by hydrogen bond formation among the β‐sheets and favorable electrostatic interactions between the key residues (KR1 and KR7) and all surrounding residues, as captured by interaction energies IE1 (KR‐KR), IE2 (KR‐all residues), and IE3 (non‐KR‐all residues). This computational validation is critical, as structural integrity of the monomer is a fundamental prerequisite for successful extracellular self‐assembly into functional nanofibers. Notably, the slower decay of hydrogen bond autocorrelation observed for the five de novo homologs compared to native CsgA indicates that their inter‐residue hydrogen bonds are more stable, representing a key distinction from that observed for native CsgA. Biomanufacturing of Modular De Novo Autogenic Engineered Living Materials The above‐described five de novo CsgA homologs from non‐model bacteria pose several challenges, such as culturing bacteria, scalable production, and genetic modifications. To circumvent these challenges, we have harnessed the curli secretion machinery of E. coli to develop a platform termed SECRETE, which stands for Secretion of Extra Cellular REcombinant β‐solenoid proteins for Tailorable Engineered living materials. Herein, we employed E. coli strain PQN4, wherein the curli operon is deleted, which makes it suitable for curli overproduction. These β‐solenoid protein monomers were designed by integrating Sec (N‐terminal signal sequence) and N22 (N‐terminal curli‐specific targeting sequence) sequences of E. coli to facilitate extracellular secretion and self‐assembly. Congo red dye assay was utilized to experimentally and qualitatively assess the cross‐β structure of Hs13‐CsgA, Am18‐CsgA, Bc36‐CsgA, El43‐CsgA, and Er46‐CsgA (Figure 4a). All five CsgA homologs showed good Congo red binding like that of CsgA; the relative differences in the Congo red binding could be attributed to slightly different (1) affinities due to their distinct amino acid compositions and structure and (2) levels of production of CsgA homologs. Wide‐angle X‐ray scattering (WAXS) analysis revealed characteristic d‐spacing (interplanar spacing) values of 0.98 and 0.46 nm, corresponding to inter‐β‐sheet and inter‐β‐strand distances, respectively, consistent with that of CsgA, indicating the cross‐β architecture of CsgA homologs (Figure 4b,c) [39]. Further, ultrastructural characterization of bacterial cultures by field‐emission scanning electron microscopy (FESEM) showed the protein nanofibers self‐assembled in the extracellular milieu (Figure 4d). Moreover, all CsgA homolog nanofibers exhibited characteristic birefringence under crossed polarizers, consistent with that of CsgA. (Figure 4e). Subsequently, to demonstrate the biomanufacturing of macroscopic modular ELMs, we produced hydrogels of all six variants by using the filtration protocol [18], and the FESEM images showed the dense network of nanofibers in the hydrogels (Figure 5a). In addition, the hydrogels were cast on templates and dried under ambient conditions to fabricate thin films, analogous to previously reported E. coli CsgA‐based “aquaplastics”. Such films could serve as biodegradable bioplastics, robust coatings, and sustainable materials (Figure 5b) [4]. SECRETE platform to produce β‐solenoid proteins of non‐model organisms by harnessing the curli secretion machinery of E. coli. (a) Congo red assay to determine the β‐solenoid proteins produced by engineered E. coli. E. coli cells with PQN4 and a sham plasmid (no curli operon) were used as a control. Biological replicates n = 4. Data represented as mean ± standard deviation. **** p ≤ 0.0001, one‐way ANOVA followed by Dunnett's test. (b) Wide‐angle X‐ray scattering (WAXS) analysis revealed cross‐β characteristics of β‐solenoid proteins. (c) d‐spacing (interplanar) values of 0.98 and 0.46 nm, corresponding to inter‐β‐sheet and intra‐β‐strand distances. (d) Field‐emission scanning electron microscopy (FESEM) images show the β‐solenoid protein nanofibers self‐assembled in the extracellular milieu. Scale bar 1 µm. (e) Congo red birefringence showed the cross‐β characteristics of β‐solenoid protein nanofibers. Scale bar 50 µm. Biomanufacturing of macroscopic autogenic engineered living materials from various β‐solenoid proteins. (a) Field‐emission scanning electron microscopy (FESEM) images of β‐solenoid protein hydrogels. Scale bar 1 µm. (b) Optical images of hydrogel and film manufactured using β‐solenoid protein nanofibers. Scale bar 0.5 cm. (c) Yield of β‐solenoid protein hydrogel obtained from engineered E. coli using a filtration protocol. Biological replicates n = 4. Data represented as mean ± standard deviation. **** p ≤ 0.0001, *** p ≤ 0.0009 one‐way ANOVA followed by Dunnett's test. (d) Storage modulus of β‐solenoid protein hydrogel. Biological replicates n = 3. Data represented as mean ± standard deviation. *** p ≤ 0.001, * p ≤ 0.014 one‐way ANOVA followed by Dunnett's test. (e) 3D printing of Am18‐CsgA protein hydrogel (3 layers). Scale bar 1 cm, (f) Protective effect of β‐solenoid protein nanofibers against UV‐C irradiation. E. coli cells embedded within β‐solenoid protein nanofibers exhibit enhanced survival following UV‐C exposure (10 min) compared to cells without nanofibers (negative control). Colony‐forming unit (CFU) data was collected before and after exposure. CFU values were log10(CFU+1)‐transformed before statistical analysis. Data represented as mean ± standard deviation. Significance was assessed by one‐way ANOVA with Welch's post hoc tests (* p < 0.05). Parts of the schematics were created in BioRender. Created in BioRender. Duraj‐thatte, A. (2026). The yield of the hydrogels, reported as the wet mass obtained per liter of bacterial culture, varied considerably across variants. The cell‐free hydrogels produced here, as per our previously reported protocol [18], are obtained by chemically removing bacterial cells, yielding a matrix comprising primarily β‐solenoid protein nanofibers. Am18‐CsgA (1938 ± 75 mg L−1) and Bs13‐CsgA (1075 ± 96 mg L−1) showed the highest yields, both exceeding that of E. coli CsgA (375 ± 42 mg L−1), making them more suitable for scalable biomaterial production (Figure 5c). As the five β‐solenoid proteins have more repeat units than CsgA and exhibit slightly different nanofiber network morphologies as observed by FESEM, we anticipated that their mechanical properties could be different. To test this, we investigated the storage modulus (G') of hydrogels, which revealed that all five β‐solenoids exhibited significantly higher G' values than CsgA. The storage modulus (G′) of hydrogels formed by the β‐solenoids was approximately twofold higher compared to those formed by E. coli CsgA. However, G' of Am18‐CsgA with 18 repeats was also the same as that of Er46‐CsgA with 46 repeats, which could be attributed to differences in the (1) physicochemical properties such as greater hydrophobicity and lower charge, and (2) aggregation of nanofibers and, in turn, the interactions with water that facilitate gelation (Figures 5d and S13). In addition, these curli hydrogels are suitable for extrusion bioprinting due to their shear‐thinning behavior, a key property that enables them to flow like a fluid under the influence of external stress, and upon removal of the external stress, it reverts to their initial viscosity. Building on this property, we demonstrated the 3D printability of β‐solenoid protein hydrogels by printing the Virginia Tech University logo using Am18‐CsgA, which was utilized due to its higher hydrogel yield (Figures 5e and S14; Video S2). The cross‐β architecture of amyloid proteins is known to confer exceptional resilience to desiccation, chemical stress, and radiation, owing to densely packed, hydrogen‐bonded β‐sheet networks. Prion‐derived amyloid fibrils exhibit extraordinary resistance to UV and even gamma irradiation, demonstrating that radiation tolerance can arise directly from the cross‐β fold rather than organism‐specific biology [40, 41]. Likewise, recent studies of the extremophile Geodermatophilus obscurus suggest that extracellular functional amyloids enhance survival under intense UV‐C exposure in Mars‐analog environments [42]. Based on these observations, we hypothesized that CsgA and the de novo CsgA homolog nanofibers identified in this study would similarly attenuate UV‐C and protect embedded cells. To test this, we evaluated all six β‐solenoid variants, including native E. coli CsgA. Five of the six variants provided significant protection against UV‐C irradiation, increasing the survival time of embedded E. coli by approximately fivefold compared to samples lacking a β‐solenoid matrix under direct germicidal exposure (Figure 5f). Notably, El43‐CsgA was the only variant that conferred no significant UV‐C protection. This result indicates for the first time that extracellular β‐solenoid fiber networks can act as a radiation‐mitigating ECM, with protective capacity varying across homologs. Having demonstrated the modulation of physicochemical properties of β‐solenoid protein building block‐based nanofibers and macroscopic materials, we next explored their potential for programming biological functionalities. To achieve this, we selected Hs13‐CsgA as the β‐solenoid protein building block, which was genetically fused with a functional peptide or protein to assess whether the engineered variants could be secreted and self‐assembled into biofunctional nanofibers. Herein, the C‐terminus of Hs13‐CsgA was genetically fused with an 8‐amino‐acid iron‐binding peptide (Hs13‐CsgA‐IronBP) or an 185‐amino‐acid Immunoglobulin G (IgG) antibody‐binding protein [43] (Hs13‐CsgA‐IgG‐BD) via a 36‐amino‐acid flexible linker. AlphaFold2 structure predictions confirmed that both Hs13‐CsgA‐IronBP and Hs13‐CsgA‐IgG‐BD retained their characteristic 13‐repeat β‐solenoid architecture with no structural disturbances caused by the fused peptide or protein domain (Figures 6 and S15). Congo red assay and FESEM images of Hs13‐CsgA‐IronBP and Hs13‐CsgA‐IgG‐BD confirmed their cross‐β structure and nanofibrillar assemblies, respectively (Figure S16). The iron‐binding assay showed that Hs13‐CsgA‐IronBP had a significantly higher weight of bound iron oxide nanoparticles than the wild‐type Hs13‐CsgA nanofibers. The elemental composition, studied using energy‐dispersive X‐ray analysis (EDAX) and transmission electron microscopy (TEM), also confirmed that Hs13‐CsgA‐IronBP has a significantly higher iron content (Figures 6a, S17, and S18). Similarly, Hs13‐CsgA‐IgG‐BD was found to bind red fluorescent IgG antibodies nearly twice as strongly as wild‐type Hs13‐CsgA nanofibers, as indicated by fluorescence intensity measurements and imaging (Figure 6b). Additionally, FTIR analysis of the amide I region revealed that both Hs13‐CsgA‐Iron‐PB and Hs13‐CsgA‐IgG‐BD fusions retained characteristic β‐sheet peaks at 1620–1630 cm− 1, confirming structural integrity and validating our computational design approach (Figure S19). These results highlight the utility of β‐solenoid nanofibers to tailor their functionalities for biomining, bioremediation, biosensing, and biomedical applications. Programming the biological functionalities of de novo β‐solenoid proteins. a. Genetic design of E. coli programmed to produce β‐solenoid protein nanofibers displaying an iron oxide‐binding peptide (Hs13‐CsgA‐IronBP) along with a FESEM image of the protein nanofibers self‐assembled in the extracellular milieu. Weight analysis and energy‐dispersive X‐ray (EDAX) analysis show the enhanced binding of iron oxide nanoparticles to engineered nanofibers, Hs13‐CsgA‐IronBP. Scale bar 1 µm. Biological replicates n = 4. Data represented as mean ± standard deviation. *** p ≤ 0.0001 two‐way ANOVA for EDAX analysis and *** p ≤ 0.0008 t‐test followed by Welch's test for weight analysis. (b) Genetic design of E. coli cells to produce β‐solenoid protein nanofibers displaying IgG antibody binding domain (Hs13‐CsgA‐IgG‐BD) along with FESEM image of the protein nanofibers self‐assembled in the extracellular milieu. Fluorescent IgG antibody binding assay shows higher binding for Hs13‐CsgA‐IgG‐BD than wild‐type Hs13‐CsgA nanofibers. Scale bar 1 µm. Biological replicates n = 3. Data represented as mean ± standard deviation. ** p ≤ 0.0024 t‐test followed by Welch's test. The structures of de novo β‐solenoid proteins, Hs13‐CsgA‐IronBP and Hs13‐CsgA‐IgG‐BD, were predicted using AlphaFold2. Discussion There are only a few examples of functional amyloids, and therefore, it was initially surprising for us to find that 33,564 proteins listed in the NCBI database are homologs of CsgA. Interestingly, recent work has reported that 22% (43279) of the genomes out of the 201210 bacterial genomes searched contained one or more predicted CsgA sequences with one or more curli repeat signatures [32, 44]. Thus, our protein mining results closely corroborate with the previously reported genome mining results and, on the other hand, emphasize the enormous scale to which CsgA‐like proteins are prevalent in the natural world [44]. Moreover, the curli system is not exclusively limited to E. coli, and ∼33% of identified CsgA homologs are secreted by bacteria phylogenetically spread across at least four major bacterial phyla, each with its own distinct curli system. This evolutionary diversity and abundance of functional proteins highlight the untapped potential that could be explored and exploited for advanced autogenic ELMs. The rapid advancement of AI‐driven structure prediction tools such as the open‐source AlphaFold2, combined with publicly available protein databases such as NCBI, has revolutionized our ability to explore the structural properties of diverse proteins and significantly accelerated the design, screening, and development of novel functional protein‐based materials. As cryo‐TEM (transmission electron microscopy) and single‐crystal X‐ray diffraction studies of CsgA and its homologs have proven extremely difficult, structure prediction tools like AlphaFold2 are beneficial in getting critical insights. MD simulations provide valuable insights, demonstrating the critical role of KRs in stabilizing β‐solenoid structure. However, the evolutionary diversity observed in de novo analogs indicates that even non‐conserved KR1 and KR7 residues lead to stable structures. Thus, it opens a broad landscape for tinkering and rationally designing de novo functional protein materials. Conclusion In summary, we have demonstrated that E. coli curli machinery could be meticulously utilized to produce de novo autogenic ELMs by employing β‐solenoid protein building blocks of non‐model bacteria. Despite having distinct amino acid compositions and molecular weights (up to nine times that of native CsgA), their biosynthesis, secretion, and extracellular self‐assembly into stable β‐solenoid protein nanofibers highlight the robustness and versatility of the SECRETE platform to biomanufacture de novo autogenic ELMs. Moreover, the variations in the yields of hydrogels (375 to 1938 mg L−1) and storage modulus (335 to 959 Pa) might be influenced by the physicochemical properties of β‐solenoid building blocks, such as the number of β‐sheet repeats, hydrophobicity, charge, and amino acid compositions, which need further investigation. Additionally, the variation in UV‐C protection observed across the β‐solenoid variants, including the absence of protection conferred by El43‐CsgA, suggests that fiber network organization and packing density, governed by the distinct physicochemical properties of each β‐solenoid sequence, are key determinants of radiation attenuation capacity. Understanding the precise sequence–structure–function relationships that govern UV‐C resistance across β‐solenoid variants is an important direction for future investigation. Further characterizing these relationships will be essential to leveraging β‐solenoids as a programmable platform for engineering protective, resilient biomaterials suitable for harsh terrestrial and extraterrestrial environments. Beyond radiation resistance, the functionalities demonstrated here highlight the broader potential of the SECRETE platform for advanced ELMs development. Genetically engineered novel β‐solenoid ELMs capable of binding iron oxide nanoparticles could enable magnetically responsive living materials for applications such as targeted drug delivery or magnetically actuated hydrogels. IgG antibody display on the nanofiber matrix opens avenues for living biosensors and diagnostic materials capable of capturing specific antigens in situ. Altogether, the fabrication of hydrogels, films, and well‐defined 3D architectures with programmable physical, chemical, and biological functionalities will unlock new opportunities, enabling the rational design of sophisticated de novo autogenic ELMs for biomedical and environmental applications. Experimental Section Cell Strains and Plasmids All experiments were conducted using PQN4, E. coli strain derived from LSR10 (MC4100, ΔcsgA, λ(DE3), CamR) with the deletion of curli operon (∆csgBACEFG) [45]. The genes encoding CsgA and its five homologs and the two engineered domains were synthesized by Twist BioScience and cloned into the pET21d vector using overlap extension and isothermal Gibson Assembly (New England Biolabs). The whole plasmid sequences were confirmed by Plasmidsaurus. These plasmids also included genes encoding proteins essential for curli biosynthesis, such as csgC, csgE, csgF, and csgG. Five different β‐solenoid proteins, namely, Hs13‐CsgA, Am18‐CsgA, Bc36‐CsgA, El43‐CsgA, and Er46‐CsgA were inserted in place of the E. coli wild‐type CsgA after the SEC (N‐terminal signal sequence) and the N22 (N‐terminal curli‐specific targeting sequence) to facilitate the secretion of CsgA variants into the extracellular space. Additionally, two engineered β‐solenoid protein variants were created, namely, Hs13‐CsgA‐IronBP and Hs13‐CsgA‐IgG‐BD by fusing iron‐binding peptide (IronBP) and antibody IgG binding domain (IgG‐BD) to the C‐terminus of Hs13‐CsgA via a 36‐amino‐acid flexible linker. All gene sequences, protein sequences, and primers used in this study are fully provided in Figure S20 and Table S5. Protein Sequence Mining and AlphaFold2 Modeling of CsgA and Its Homologs We utilized the protein structure database of the National Center for Biotechnology Information (NCBI) and employed the search query—Major [Organism] OR major [All Fields] AND curlin [All Fields]—to identify CsgA homologs. This search yielded 33,564 entries, annotated as having the curlin subunit CsgA, either through computational analysis or by identifying sequences using a regular expression based on the minimalistic curli repeat (X6QXGX2NX10) described by Chapman et al. [46]. The lengths of the resulting protein sequences varied from 141 to 1390 amino acids. From these 33 564 entries, we selected 50 sequences, and to predict their 3D structure, we employed AlphaFold2 (version 2.1.1) [31] on the Tinkercliffs HPC cluster at the Advanced Research Computing Center, Virginia Tech, utilizing NVIDIA A100‐80G architecture. The source code for AlphaFold2 is available at https://github.com/deepmind/alphafold. Structural images were generated using PyMOL (version 1.16) [47]. Batch protein structure predictions were performed, running for six recycles and generating five models per sequence. The models were ranked using the pLDDT score, with the highest‐ranked model being used for subsequent analyses. Out of these 50 sequences, five sequences were selected, having lengths across the whole spectrum of 141 to 1390 amino acids. Genes of these five sequences were synthesized to validate their production in the E. coli PQN4 strain experimentally. These variants were named based on the first letter of the species, followed by the number of cross‐β amyloid repeats and CsgA—to indicate that it is a homolog. For example, the CsgA homolog of Halomonas saliphila with 13 cross‐β amyloid repeats was named Hs13‐CsgA. Similarly, 18, 36, 43, and 46 cross‐β amyloid repeats of CsgA homologs of Alteromonas macleodii, Blastomonas sp. CACIA14H2, Erythrobacter longus, and Ensifer sp. Root31 were labelled as Am18‐CsgA, Bc36‐CsgA, El43‐CsgA and Er46‐CsgA, respectively. In addition, two engineered variants, Hs13‐CsgA‐IronBP and Hs13‐CsgA‐IgG‐BD, were also predicted in a similar fashion. Molecular Dynamics (MD) Simulation All‐atom MD simulations of the six β‐solenoid proteins—CsgA, Hs13‐CsgA, Am18‐CsgA, Bc36‐CsgA, El43‐CsgA, and Er46‐CsgA in explicit water were conducted using the NAMD‐2.14 software [48]. The initial structures were obtained from the AlphaFold2 server using the experimental FASTA sequences [31]. These β‐solenoid proteins were solvated using TIP3P water in a cubic box with a minimum of 15 Å padding in each direction. The solvated β‐solenoid proteins were neutralized by adding sodium ions to balance the net charges of these proteins (Figure S7). The atomistic MD simulations were performed in 2 fs timestep in the NPT ensemble with periodic boundary conditions applied in all directions. The Charmm‐36m (July 22) force field (FF) was used to model the bonded and non‐bonded interactions [49]. The standard 12‐6 Lennard‐Jones potential was applied with a 12 Å cutoff, and the Particle Mesh Ewald (PME) method was used to calculate the long‐range electrostatic interactions [50]. The Langevin thermostat and Parrinello‐Rahman barostat were implemented to maintain a temperature of 300 K and a pressure at 1.01325 bar, respectively. This simulation setup maintained system stability while accommodating fluctuations in density. Energy minimization of 50 000 steps was followed by 150 ns MD simulations, in which the last 50 ns were treated as a production run. The trajectory data were stored in 1 ps intervals during these MD simulations. All the analyses were carried out using the last 50 ns production run except for the RMSDs, which were calculated for the entire 150 ns trajectory to compare and estimate the structural changes from the initial AlphaFold2‐generated structures. Analysis Method of MD Simulation The RMSD and RMSF of the amyloids were calculated using the VMD software [51] package, and the MD Analysis package was used to calculate the hydrogen bond autocorrelations and Ramachandran plots. The hydrogen bond criteria used were a donor‐hydrogen distance cutoff of 1.2 Å, a donor–acceptor distance cutoff of 3.0 Å, and a donor‐hydrogen‐acceptor angle cutoff of 150° [36]. We developed an automated tool utilizing the NAMD executable to calculate the residue‐residue interaction energies for the large β‐solenoid protein variants (https://github.com/Deshmukh‐Group/ResidueResidue_Interaction_Energies_NAMD). These energies were calculated based on the non‐bonded interactions, including electrostatic and van der Waals contributions. Water and ions were removed from the system to calculate the energies. The last 50 ns trajectory is analyzed with a 0.1 ns stride. A percent cutoff of 60% and a 12 Å filtering distance were applied, meaning that only pairs of residues whose center‐of‐mass come closer than 12 Å in at least 60 percent of trajectory frames were included in the non‐bonded energy calculations. We verified our energy data with the gRINN software package [52]. Microbial Production of β‐solenoid Protein Nanofibers All plasmids (CsgA, Hs13‐CsgA, Am18‐CsgA, Bc36‐CsgA, El43‐CsgA, Er46‐CsgA, Hs13‐CsgA‐IronBP, and Hs13‐CsgA‐IgG‐BD) were transformed into the E. coli strain PQN4. The transformed cells were streaked onto lysogeny broth (LB) agar plates containing 100 µg mL−1 carbenicillin and 0.5% glucose (w v−1) for catabolite repression of T7RNAP and incubated overnight at 37 °C. A single colony from each plate was picked and separately cultured at 37 °C in 5 mL of LB medium with 100 µg mL−1 carbenicillin and 2% glucose (w v−1). These overnight cultures (PQN4_CsgA, PQN4_Hs13‐CsgA, PQN4_Am18‐CsgA, PQN4_Bc36‐CsgA, PQN4_El43‐CsgA, PQN4_Er46‐CsgA, PQN4_Hs13‐CsgA‐IronBP, and PQN4_Hs13‐CsgA‐IgG‐BD) were then transferred to fresh 500 mL LB media containing 100 µg mL−1 carbenicillin. The cultures were incubated in shaking incubators (225 rpm, 37 °C) for 48 h to achieve expression of β‐solenoid proteins and assembly into functional amyloid nanofibers. As a negative control, PQN4 was transformed with a sham (all curli genes, including csgA, were absent), and the pET21d plasmid was used. Quantitative Congo Red Dye Binding Assay One milliliter of bacterial culture (48 h, 500 mL) was centrifuged at 6000 rpm for 10 min. The resulting cell pellet was resuspended in a 0.025 mM solution of Congo red in phosphate‐buffered saline (PBS) and incubated for 10 min. The cells were then pelleted again at 14 000 rpm for 10 min, and the absorbance of the supernatant (200 µL) at 490 nm was measured using a microplate reader. This absorbance value was subtracted from 0.025 mM Congo red in PBS, which was subsequently normalized by the OD600 of the original bacterial culture to quantify the production of functional amyloid nanofibers. The assay was performed in four biological replicates. Congo Red Birefringence Functional amyloid nanofibers were stained with Congo red and evaluated for birefringence [53]. One milliliter of bacterial culture was centrifuged at 6000 rpm for 10 min. The resulting pellet was resuspended in 10 µL of 500 µM Congo red solution in 80% ethanol and incubated for 60 min at room temperature. XploRA PLUS Confocal Raman Microscope, equipped with an inverted camera and crossed polarizing light filters, was utilized to visualize and assess birefringence. Preparation of β‐solenoid Protein Hydrogels E. coli PQN4 strain with the desired plasmid was cultured in 500 mL of LB media in a shaking incubator at 37 °C, 225 rpm for 48 h to produce the β‐solenoid protein nanofibers. The resulting 500 mL culture was treated with urea to a final concentration of 0.8 M and kept at 4 °C for 1 h. The treated cultures were then concentrated by vacuum filtration through a 90‐mm diameter polycarbonate membrane with 10‐µm pores (EMD Millipore). The resulting concentrated biofilm (nanofibers) was washed thrice with 50 mL of sterile deionized (DI) water on the filter membrane. The biofilm was subsequently incubated with 100 mL of 8 M urea solution in water for 5 min, followed by vacuum filtration and washing with 200 mL of DI water to remove the bacterial debris lysate. The resulting biomass on the filter membrane was then treated with 50 ml of 5% (w v‐1) SDS (sodium dodecyl sulfate) solution (gelator/plasticizer) in water for 5 min, followed by vacuum filtration and an additional wash with 500 mL of DI water. The β‐solenoid protein hydrogel formed on the filter membrane was then collected and stored at 4°C. Fabrication of Thin Films From β‐solenoid Protein Nanofibers To fabricate 2D thin films, β‐solenoid protein nanofibers‐based hydrogels were cast onto a flat, flexible substrate such as plastic wrap or aluminum foil. These were left to dry overnight under ambient conditions to form thin films of lateral dimensions 1 cm * 1 cm. Once thoroughly dried, the thin film was gently peeled off the substrate and stored at room temperature. Field‐emission Scanning Electron Microscopy Sample Preparation and Imaging 100 µL of the E. coli cultures or 10 mg of hydrogel samples were filtered/spread onto polycarbonate membranes with a 0.22 µm pore size under vacuum and then placed in a fixative solution containing 1 mL of 4% glutaraldehyde and 1 mL of 4% paraformaldehyde buffer for 2 h at room temperature. After fixation, the membranes were gently rinsed with water and subjected to a series of 200‐proof ethanol washes with increasing concentrations (25%, 50%, 75%, 100%, and 100% v v‐1), each for 15 min. The samples were then processed using a critical point dryer. The dried membranes were mounted on Scanning Electron Microscopy sample holders with carbon adhesives and coated with a 10 to 20 nm layer of Pt/Pd. Images were captured using a JEOL IT500 SEM equipped with a field‐emission gun operating at 5–10 kV. Wide‐Angle X‐ray Scattering Wide‐angle X‐ray scattering (WAXS) experiments on the thin films were performed using a Xenocs Xeuss 3.0 SAXS/WAXS equipped with a GeniX 3D Cu HFVLF microfocus X‐ray source utilizing Cu Kα radiation (λ = 0.154 nm). The sample‐to‐detector distance was 42 mm, and the q‐range was calibrated using a Lanthanum hexaboride standard. Two‐dimensional scattering patterns were obtained using a Dectris EIGER 4 M detector with an exposure time of 30 min. Data reduction was performed using XSACT software provided by Xenocs. Rheology Studies of the Hydrogels The viscoelastic properties of the β‐solenoid protein hydrogels were assessed using a µ‐volume sample holder with the ElastoSens Bio (Rheolution, Montreal, Canada). 250 µg gel sample was cast into the sample holder. The storage modulus (G′, Pa) of each sample was measured to evaluate its viscoelastic characteristics. Each test was performed in technical duplicates for three biological replicates, resulting in a total of six readings per time point. G′ was recorded every 5 s over a 1‐min period. 3D Printing of β‐solenoid Protein Hydrogels (Microbial Ink) The bioinks were transferred into a 10‐mL Luer‐Lok syringe and centrifuged at 112 × g for 2 min to remove air bubbles. Bioprinting was conducted at room temperature using a Hyrel HYDRA 21 Bioprinter (Hyrel3D, Norcross, GA, USA) equipped with a 10‐mL syringe dispensing head (SDS‐10) for gel printing. 22‐ and 20G needles with a premade 0.25‐inch blunt end were utilized. The printing speed of 2 mm s−1 and the blunt needle diameters of 0.41 (22G) and 0.60 mm (20G) were selected after optimization. This process involved assessing various printing speeds and needle sizes with internal diameters ranging from 0.26 to 0.60 mm to achieve optimal pattern fidelity, reproducibility, and consistency across different polymer concentrations and viscosities of the amyloid hydrogel inks. Initial bioprinting trials were performed at feed rates ranging from 2 to 10 mm s−1. The 3D STL files were prepared using the slic3r engine integrated with the Hyrel HYDRA 21 Bioprinter software. Optical Images Optical images were obtained using a Canon EOS Rebel T7 DSLR Camera. Iron Oxide Binding Analysis by Genetically Engineered β‐solenoid Protein The plasmid pET21d‐Hs13‐CsgA‐IronBP and pET21d‐Hs13‐CsgA were separately transformed into PQN4 cells to produce engineered and non‐engineered functional amyloid nanofibers. Engineered nanofibers were produced as previously described in the section “Microbial production of β‐solenoid protein nanofibers”. 50 mL of culture with expressed nanofibers was spun down, and the supernatant was discarded. This cell pellet was added with 5 mg of iron (III) oxide (<50 nm, Sigma–Aldrich) in 50 mL of water. Cell pellets with nanofibers were incubated at ambient temperature with light agitation for 2 h to facilitate binding of iron oxide nanoparticles to nanofibers. Next, the entire mixture was deposited on a 47‐mm diameter polycarbonate membrane with 10‐µm pores (EMD Millipore) and washed with water several times to remove unbound iron oxide nanoparticles. The weight of iron oxide nanoparticles bound to Hs13‐CsgA‐IronBP was determined and compared to control samples with non‐engineered nanofibers of Hs13‐CsgA. Additionally, the engineered and non‐engineered nanofibers with bound iron oxide nanoparticles deposited on the filter membrane were tested by energy‐dispersive X‐ray analysis (EDAX) to obtain elemental composition. Antibody Binding Analysis of Genetically Engineered β‐solenoid Protein The plasmid pET21d‐Hs13‐CsgA‐IgG‐BD was transformed into PQN4 cells to express the engineered functional amyloids containing IgG‐binding protein fusion. The expressed engineered nanofibers were deposited on the filter membrane (dry weight of nanofibers 1.875 ± 0.62 mg) and incubated with a solution of antibody Goat anti‐Rabbit IgG with Alexa Fluor 594 (Thermo Fisher Scientific) for 2 h. Next, filter membranes with deposited, engineered (Hs13‐CsgA‐IgG‐BD) and non‐engineered nanofibers (control, Hs13‐CsgA) were washed with water several times to remove unbound antibodies. The bound antibodies were detected by fluorescence intensity of Alexa Fluor 594 using a ChemiDoc Imaging System (BIO‐RAD). Images were analyzed by utilizing ImageJ software. UVC Irradiation Assays E. coli PQN4 strain with the desired plasmid was cultured in 500 mL of LB media with 100 µg mL−1 carbenicillin in a shaking incubator at 37 °C, 225 rpm for 48 h to produce the β‐solenoid protein nanofibers. 100 mL of culture was collected and pelleted at 4000 x g for 10 min, then resuspended in 10 mL of 8 M Urea for 45 min. Samples were then pelleted again, and 10 mL of 8 M Guanidine HCl was added for 15 min to lyse all E. coli PQN4. Samples were pelleted and washed twice with PBS, and then resuspended in PBS to a final volume of up to 5 mL, corresponding to a 20x concentration. To determine the UV‐C protective capabilities, the E. coli BL21 strain carrying the pDawn plasmid (OD 1.0) was combined with prepared β‐solenoid protein nanofiber samples in a 1:1 (v/v) ratio. E. coli BL21 alone without β‐solenoid protein nanofiber resuspended in PBS was used as a control. The resulting mixture was then placed on a quartz glass slide and placed in the UVClave (UV wavelength: 254 nm, UV intensity: 500 µW/cm2 at the center of the chamber, Benchmark, USA). After the defined period of exposure (10 min), the samples were serially diluted and plated on LB agar containing kanamycin, and then incubated at 37°C overnight. Data were collected in biological triplicate. All statistical analyses were performed using GraphPad Prism 10. Raw values were log10‐transformed before analysis. Because some groups included very small values (≈1), we applied a log10(Y+1) transformation to stabilize variance and avoid undefined log(0). One‐way ANOVA was used to test for overall group differences, followed by Tukey's HSD post hoc test for pairwise comparisons. Results were considered statistically significant at p < 0.05. Statistics and Reproducibility All experiments presented in this article were repeated at least three times (n = 3) as biological replicates or distinct samples, and they are clearly stated in the figure captions and relevant method sections. No data transformation or normalization was applied, and no outliers were excluded. All data are presented as the mean and standard deviation (SD). Statistical significances were assessed using ordinary one‐way ANOVA or Welch t‐test (two‐sided). Where applicable, post‐hoc tests were performed as indicated in the figure captions. All statistical analyses and data plotting were performed using GraphPad Prism 10 software. Author Contributions A. D. T. and A. M. B. developed the initial concept. A. D. T. streamlined the central concept and methodology. A. D. T. and H. H. designed experiments. H. H. performed protein mining, AlphaFold2 structure prediction, and engineering of E. coli to produce β‐solenoid protein nanofibers, hydrogels, and their characterization. S. S. performed all molecular dynamics simulations and all analyses. E. J. performed WAXS analysis of β‐solenoid proteins. N. B. performed the UVC irradiation assay; H. P. performed FTIR spectroscopic analysis; W. H. performed transmission electron microscopy (TEM); A. M. B. performed the SEM analysis. R. M. supervised E. C.; S. D. supervised S. S.; A. D. T. supervised H. H, N. B., and H. P. The manuscript was drafted by H. H., A. M. B., and A. D. T. with input from all authors. All authors discussed the results and commented on the manuscript. Conflicts of Interest A.D.T., H.H., and A.M.B. are inventors on a patent application related to the SECRETE platform and de novo autogenic engineered living materials described in this manuscript. US Application No. 19293006 “Production of csga‐like functional amyloids for engineered living materials”. Supporting information Supporting File 1: adma74312‐sup‐0001‐SuppMat.pdf. Supporting File 2: adma74312‐sup‐0002‐VideoS1.mp4. Supporting File 3: adma74302‐sup‐0003‐VideoS2.mp4.