A symbiotic-like biologically-driven regenerating fabric

Living organisms constantly maintain their structural and biochemical integrity by the critical means of response, healing, and regeneration. Inanimate objects, on the other hand, are axiomatically considered incapable of responding to damage and healing it, leading to the profound negative environmental impact of their continuous manufacturing and trashing. Objects with such biological properties would be a significant step towards sustainable technology. In this work we present a feasible strategy for driving regeneration in fabric by means of integration with a bacterial biofilm to obtain a symbiotic-like hybrid - the fabric provides structural framework to the biofilm and supports its growth, whereas the biofilm responds to mechanical tear by synthesizing a silk protein engineered to self-assemble upon secretion from the cells. We propose the term crossbiosis to describe this and other hybrid systems combining organism and object. Our strategy could be implemented in other systems and drive sensing of integrity and response by regeneration in other materials as well.


BLAST-KOALA analysis
The aim of this analysis was to get all protein sequences of the genes that were differentially expressed and submit them to BLAST KOALA. This search would result in a list of KOs that can be used for iPATH visualisation. 1. The table was split into 2 tables: upregulated (Condition VS control) and downregulated genes. 2. To retrieve the sequences of the genes in each list, the tables were fed to the "Fish_GFF_info_for_fasta_v1.py" script. The script takes both the table and a GFF file (downloaded from ensemble at: ftp://ftp.ensemblgenomes.org/pub/bacteria/release-30/gff3/bacteria_0_collection/bacillus_subtilis_subsp_subtilis_str_168/Bacillus_subtili s_subsp_subtilis_str_168.GCA_000009045.1.30.gff3.gz) and compares them. Each ID in the table (column 1 -"gene_ID") appears in the gff file in the attributes section as "locus_tag". It is written into an output file called "$INPUT_FILE_BASENAME_fetch_fasta.tab". 3. The "fetch_fasta" file is then fed into the script "connection2ncbi_v2.py" script (which takes it as an argument using a "-i" switch). The script will connect to NCBI Entrez and "fish" the fasta sequence of each gene based on the GI number (of the genome), the start and end positions of the feature, and its orientation (provided in the output of step 2). Please note that NCBI requires an email address of the user: use the "-e" switch to enter your email or write it into the script to use in default cases. Outputs were named using the "-o" switch. the default is named: "./myOutput.fa". 4. The resulting output is a nucleotide fasta. feed it to "translate.py" (requires biopython) to translate it to amino acids sequence. The output will be named automatically ($INPUT_FILE_BASENAME_protSeq.fa). 5. The amino acids multi-fasta file is then submitted to BLAST KOALA (http://www.kegg.jp/blastkoala/). We used the "family_eukaryotes + genus_prokaryotes" database because if a sequence will not be found in the prokaryotes DB for any reason it might be found in the EUKs DB. 6. Out of the BLAST-KOALA results, the gene_name to KO list was downloaded (Down/UpGenes_ko_list_geneName.txt) and the KO numbers were extracted (files are named "UpGenes_ko_list.txt" or "DownGenes_ko_list.txt").

DAVID analysis
This analysis is examining the enrichment of functions in a list of upregulated and downregulated genes extracted from differential expression analysis performed by the bioinformatics unit of Weizmann Institute using DESeq2. 1. Gene names as they appear in the gff file were used to match a UniProt accessions (that can be identified by DAVID, in contrast to the names appearing in the files). This was done by going to http://www.uniprot.org/uploadlists/ and uploading the list of genes. Procedure is: select from "gene name" to "UniProtKB" and click "go". Than list of identifiers is downloaded (file was named "Gene_name_to_uniprot.tab"). 2. Several IDs could not be found in the UniProt database (2 in the upregulated genes and 8 in the down regulated genes) -a list of "orphans" gene IDs, including gene name, product, and weather it was included in the DAVID analysis or not is provided (B.subtilis_orphan_IDs-no_uniProt_Accession.xlsx). Finally, only one of the orphans was included in the DAVID analysis from each list (from the list of the upregulated and of the downregulated genes). NOTE: Out of 204 upregulated genes submitted to DAVID, 201 found a hit. out of 284 downregulated genes 270 found a hit. 3. UniProt accessions were uploaded to DAVID and the CO_BP_FAT, CO_BP_ALL, KEGG_pathways, COG_ontology, and interPro_protein_domains tables were downloaded for visualization. Only GO_BR_ALL_UpGenes.txt and GO_BR_ALL_downGenes.txt were used 4. The optional parameters of the tables were first modified: all statistical values (FDR, Benferoni etc.) were included in the tables and the counts threshold was reduced from 2 to 1. The tables were saved as tab delimited text files and the headers were modified to match Shengwei's bbplot modified R script (provided -bbplot_visualization.R). The graphic is provided in the PDF file "GO_BP_FAT_enrichment.pdf". The red circles represent downpregulated and the green upregulated GO terms. the size of the circle is proportional to the number of genes assigned to the respective GO in the list and the opacity reflects the level of certainty for enrichment (the more solid the color, the more certain is the enrichment) Figure S6. DAVID analysis. Circle size is proportional to number of genes assigned to the respective function, the opacity is related to the probability of enrichment of this function (the more solid the color, the more probable this function is enriched in the provided list). Red is enriched in the list of down-regulated genes, Green in the list of up-regulated. This figure is based on GO terms. It can be generated for KEGG, COG, interPro ect. GO yielded the highest amount of functions.

Supplementary note 7: Protein expression in insect and bacterial systems
Alkaline lysis of transformed E. coli, and subsequent acidification of the lysate (using sodium acetate) and dehydration (using ethanol) resulted in rapid assembly of disordered mats of fibers of various diameters ranging from ~1 to ~20 µm (Figures 9-10). Figure S7. Assembly of silk fibers from transformed E. coli, following alkaline lysis, acidification, and dehydration (scale bars: C, 1000 µm; D, 100 µm). Figure S8. Silk fiber structures of single segments and a combination product. Panels A-E show structure of segments spsegI-V, respectively, while panel F shows the combination of spsegII+spsegV (all scale bars, 1000 µm; panel A is the same one used as panel C in Figure S7). Cubic objects appearing in some of the panels are salt crystals by EDS analysis. In order to study the physical, mechanical, and chemical properties of these silk fibers, a higher-scale production was required. In addition, we decided to switch to an insect system to get fibers which are as similar to the source material as possible. Our chosen expression system was the Drosophila melanogaster cell line Schneider's 2 (S2), which is a macrophage-like line derived from late stage primary cultures of embryonic cells.
S2 cells were cultured and evaluated daily for appearance and viability, to ensure <5% dead cells ( Figure S9). An expression vector was constructed for segment spsegII. The expression vector included three modules: BiP, a Drosophila secretion signal; V5 epitope for protein detection; and 6His for purification. The cells were transformed using three methods -calcium phosphate, lipofectamine, and cellfectamine, with calcium phosphate resulting in successful expression of the protein ( Figure S9). Figure S9. S2 cell culture and transformation. A, forward scatter/side scatter density plot of S2 cells. B, viability staining of S2 cells using propidium iodide (PI). C, result of transformation using either lipofectamine (lipo) or calcium phosphate (calc), compared with untreated cells (none). Cells were stained with anti-V5 antibody.
Successfully-transfected S2 cells were induced at 24, 48, and 72 hours post transfection for a period of 24 hours, after which culture supernatant and cell pellets were harvested for protein purification on a nickel column and analysis by western blot. A significant eluted fraction was detected and collected, and western blot (using weaklyreducing PAGE and anti-V5 antibody) analysis showed protein dimers in both supernatant and cell pellets from calcium phosphate transfected cells ( Figure S10). To produce assembled silk, purified proteins were dialyzed, concentrated, and acidified to pH ~6.0 to ~5.5. Within 1-2 min the silk protein self-assembled into fibers of ~5mm long. Fibers were analyzed for their diameter and composition using the SEM ( Figure  S11). Movie S12. A Z-stack visualization of formed fiber in the fabric-biofilm hybrid.

Supplementary note 8: hypothetic crossbiotic systems
Synthetic biological products potentially address three challenges: 1. Smartness: biological behaviors (response, self-organization, reproduction etc.) into mechanical systems, to achieve a system that is sustainable and resilient on one hand, and flexible on the other. 2. Precision: biological organizations mostly occur in nature under simple and neutral conditions, compared with those in a modern studio, manufacturing facility or laboratory. Synthetic biology offers high precision ranging from the sub-nanometer scale to kilometers, without the need for sophisticated, expensive setups. 3. Cost: while modern machines, materials etc. often display a high level of performance (for example, a laptop), their smartness and precision are accomplished by highly complex processes, rendering them relatively costly ( Figure X). In contrast, synthetic biology could reduce the cost by several orders of magnitude, e.g. biofabricated computers would cost $0.5/kg instead of $500.
Biological organisms produce a remarkable array of materials with properties often far exceeding those of their synthetic counterparts including unique optical, electronic, and physicochemical properties. It is important to note, that although novel materials with information processing capabilities may be generated through physical and chemical means as well, they might be inferior in terms of sustainability and environmental compatibility than those achieved using biological, albeit synthetically assembled, parts and processes.

Preparation of MSgg -Biofilm inducing medium
For 500 ml of medium (x1 or for plates prepare x2, see notes below): STEP 1 -(everything except CaCl 2 and FeCl 3 ): Add to a 500ml tube the ingredients in the table below: Add CaCl 2 as follows (to ~500ml medium): Look for a faint opaque white cloud in the medium -a sign that everything is good...

STEP 4 -FeCl3:
Just before using the medium, add FeCl 3 as follows (to ~500ml medium): If you prepare plates -add more FeCl 3 -about x2.5 of the amout listed above.
STEP 5 -Preparing the plates (1 Liter total): Add 7.5gr agar to 250ml dH 2 O (agar concentration of 3%) and autoclave. After autoclave mix well with 250ml MSgg medium x2 at RT, and pour solution to the plates. It is better to mix and pour in 50ml tube when preparing small quantities of plates -the agar tend to congeal rapidly.