Engineered living materials (ELMs) integrate living cells within a biopolymer matrix to create materials with customized functions. While a bottom-up assembly of macroscopic ELMs with a de novo matrix offers superior control over material properties, current methods lack the ability to genetically encode a protein matrix leading to collective self-organization. Naturally occurring living biomaterials, like bones and wood, exemplify bottom-up growth from a small number of cells into macrostructures. ELMs aim to replicate this, using synthetic biology to incorporate tailored functions. These materials find applications as sensors, therapeutics, biomaterials, platforms, electronics, energy converters, and structural materials. The matrix is crucial for determining material properties, making the creation of ELMs with a synthetic biomolecular matrix that allows for autonomous growth into macroscopic structures a primary goal. However, this bottom-up, de novo approach presents significant challenges. Screening recombinant biopolymers at cell-growth-compatible concentrations is difficult, and assembling micrometer-sized cells into centimeter-scale materials requires self-organization across vast length scales. Existing methods for creating macroscopic ELMs typically employ top-down approaches like 3D printing or processing microscopic ELMs into larger structures. The few examples of autonomously produced macroscopic ELMs demonstrate limited genetic control over mechanical properties. This research proposes that strategies for developing synthetic biomolecular matrices for self-assembling bacteria into macroscopic ELMs can be informed by previous work on surface-engineered bacteria and surface-modified colloidal particles.

Prior research has engineered the surface of *Escherichia coli* to display interacting proteins, resulting in microscopic self-assembled aggregates. In contrast, studies on DNA-coated colloidal particles have shown that high-density surface modifications are crucial for self-assembly into both microscopic and macroscopic structures. This study hypothesizes that a matrix of self-interacting proteins displayed on bacteria at high density could lead to the formation of macroscopic solid materials. Previous work on the surface layer (S-layer) of *Asaccharobacter crenaesaccharolyticus* demonstrated high-density peptide display and biopolymer secretion. The S-layer's crystalline structure on the *C. crescentus* extracellular surface allows protein display at high density.

The researchers engineered the S-layer of *Caulobacter crescentus* (a bacterium lacking the adhesive holdfast to minimize native cell interactions) to display a high-density, surface-bound, self-assembling protein called the BUD protein. The BUD protein design incorporated four functional properties: (i) a surface-anchoring domain (first 250 residues of RsaA) for high-density display, (ii) an elastin-like polypeptide (ELP) biopolymer region (ELP0) to influence material properties, (iii) SpyTag and FLAG tags for functionalization, and (iv) a self-interaction and secretion domain (last 336 residues of RsaA). The engineered strain, termed BUD-ELM, was grown in liquid culture. The researchers used confocal microscopy, electron microscopy, and atomic force microscopy (AFM) to characterize the BUD protein's location (extracellular surface and secreted matrix) and the structure of the resulting BUD-ELMs (hierarchical structure with a secreted protein matrix). The roles of individual BUD protein domains in assembly were investigated by creating strains lacking the ELPα or the RsaA anchoring domain. To understand the assembly process, BUD-ELM cultures were imaged at various time points, revealing a multi-step process involving pellicle formation at the air-water interface, followed by pellicle collapse and material sinking. The influence of physical parameters, such as shaking speed and culture volume, on BUD-ELM assembly was studied, leading to a model relating modified volumetric power to material size. Rheological measurements characterized the viscoelastic properties of BUD-ELMs, demonstrating that genetic manipulation of the BUD protein significantly affected the storage and loss moduli. The self-regenerating, processability, and functional capabilities of BUD-ELMs were explored through experiments on desiccation, reshaping, and composite material formation. Heavy metal removal and enzymatic catalysis were also assessed. The researchers used various techniques such as immunoblotting, inductively coupled plasma mass spectrometry (ICP-MS), and colorimetric assays to analyze protein expression, heavy metal binding, and enzymatic activity. Specific procedures such as the construction of BUD-ELM strains through cloning and homologous recombination, growth conditions optimization, optical and confocal microscopy, AFM analysis, immunoblotting, rheological measurements, biosorption assays (Cd²⁺), functionalization with glucose dehydrogenase, and colorimetric activity assays are described in detail in the Methods section.

The BUD-ELM strain autonomously formed macroscopic (centimeter-scale) living materials composed of *C. crescentus* cells embedded in a de novo protein matrix. Confocal and electron microscopy revealed a hierarchical structure with a secreted protein matrix in addition to the surface-displayed protein. AFM imaging showed a brush-like structure on the BUD-ELM cell surface, different from wild-type cells. The BUD protein was found both surface-displayed and secreted, playing dual roles in cell aggregation and matrix formation. Genetic modifications of the BUD protein (deletion of ELPα or RsaA anchoring domain) allowed for tuning of the resulting material properties, leading to either cell-rich or matrix-rich ELMs. BUD-ELM assembly was shown to be a multi-step process involving pellicle formation at the air-water interface due to the hydrophobic nature of the protein and shaking-dependent collapse of the pellicle to form the final material. A modified volumetric power parameter was identified as a predictor of the final BUD-ELM size. Rheological measurements revealed that genetic alterations to the BUD protein could modulate the storage modulus and loss modulus of BUD-ELMs by up to 25-fold. BUD-ELMs showed self-regenerating properties, processability (extrusion, composite formation), and functionality (heavy metal removal and enzymatic catalysis). The BUD-ELMs were capable of removing over 90% of cadmium from a solution above the EPA limit, demonstrating potential for bioremediation. Functionalization with SpyCatcher-GDH enabled enzymatic catalysis. Experiments showed that BUD-ELMs with less cells than usual had an improved capacity for reshaping when wet.

This study successfully addressed the long-standing challenge of creating autonomously grown, macroscopic ELMs with tunable properties. The use of a genetically encoded, self-interacting protein matrix (BUD protein) enabled the formation of centimeter-scale materials, showcasing precise control over material characteristics. The discovery of the multi-step assembly process, involving pellicle formation at the air-water interface and subsequent collapse, provides valuable insights into the design principles for creating similar materials. The identification of a modified volumetric power parameter for predicting material size offers practical implications for scale-up and production optimization. The demonstration of self-regeneration, processability, and multifunctional capabilities highlights the potential of BUD-ELMs for various applications, including bioremediation and biocatalysis. This work establishes a robust platform for creating sophisticated living materials with tailored properties.

This research presents a novel approach to creating macroscopic engineered living materials (BUD-ELMs) through the genetic engineering of *Caulobacter crescentus*. The study successfully demonstrated the autonomous growth of centimeter-scale ELMs with tunable mechanical and functional properties. The findings provide significant advancements in the field of ELM design and fabrication and offer promising applications in various fields, including bioremediation and biocatalysis. Future research should focus on exploring different self-interacting protein domains, optimizing the assembly process, and expanding the range of functionalities achievable through genetic engineering.

The study primarily focused on the *Caulobacter crescentus* system. The generalizability of the findings to other bacterial species requires further investigation. The modified volumetric power parameter, while effective in predicting material size, is an empirical model and lacks a complete mechanistic understanding. The functional studies focused on heavy metal removal and enzymatic catalysis, and further investigation is needed to evaluate BUD-ELMs' performance in other applications. The long-term stability and viability of the cells within the BUD-ELM matrix require further study.