Our decision to pursue colorectal carcinoma suppression arose from two key points obtained from preliminary research: 1) bone morphogenetic protein 2 (BMP-2) suppresses the growth of colon cancer cell growth in vivo, and 2) the phage display peptide RPMrel confers specific and preferential binding to non-differentiated colon cancer cells.

With these two pieces of knowledge we had respectively: 1) our carcinoma suppression drug, and 2) a tool for specifically targeting cancerous cells. Thus the objective of this module was to identify and then construct a suitable mechanism making use of the RPMrel peptide to make the delivery of BMP-2 targeted. 

DESIGN

Considering limitations.

Design of a solution starts with considering existing limitations. Since this is iGEM, the clearest limitation was that the solution must be a biological one and thus must involve a living component. Only a certain set of living organisms lie within our reasonable capacity to engineer them, and of these we decided on Bacillus subtilis (see Chassis page).

We then examined the treatment environment. Carcinomas of the colon protrude into the digestive tract and items within the tract can interact with them directly. We decided on the concept that our biological system would be ingested, then produce and release the drug in the vicinity of the tumor for direct action.

To do this, RPMrel had to be expressed on the cell surface in a functional form. We conducted research into several methods to do this on B. subtilis and concluded that cell wall expression of the peptide was ideal. Imperial College London’s 2010 team had performed that same task using the cell wall binding domain of the hydrolase lytC as their peptide anchor and a helical linker of their design. We decided to the employ that same system for surface expression of RPMrel.

Using phage display to compile peptide libraries that confer specific binding to certain antigens is a now common way to come up with useful peptides. RPMrel, a 9 amino acid disulfide-constrained peptide, was screened out of the New England Biolabs Ph.D.-C7C library for positive binding to poorly differentiated HT-29 cells, and negative binding to well differentiated HCT-116 cells. All peptides in the Ph.D.-C7C library have random sequences of 7 amino acids bounded by cystines at the N- and C- terminals. Further screening of the peptides was done by performing 6 successive incubation-wash-elution cycles against HT-29. See Kelly & Jones (2003). ‘RPM’ - for an arginine-proline-methionine amino acid sequence immediately before the C-terminal cystine - emerged as a consensus motif for late selection high-affinity peptides, thus giving the peptide’s name. RPMrel’s full amino acid sequence is n-CPIEDRPMC-c.

The binding properties of RPMrel were identified during Kelly & Jones’ study when it was fused to the surface-exposed p3 minor coat protein of the bacteriophage M13KE. This module will lead to its novel fusion to the cell wall binding domain of lytC, exposing it to the extracellular environment.

lytC, and its cell wall binding domain.

lytC, a cell surface hydrolase, is native to B. subtilis and binds non-covalently to its cell wall interacting with it electrostatically. This property was previously determined to make it superior for exposure of bound peptides. Furthermore, as compared to other surface expression methods we investigated (including peptide expression on an engineered S-layer), the lytC model is much better studied.

According to a study by Yamamoto et al (2003) lytC is localized uniformly on B. subtilis cells grown past log phase, making it more ideal than the more specifically localized lytE and lytF for expression of RPMrel. It was further found that - compared to the others - lytC was particularly resistant to degradation by the cell surface protease WprA and extracellular protease Epr, both B. subtilis products. In the same study, 3xFLAG (a standard peptide epitope designed by Sigma-Aldrich Corp.) was successfully fused to the protein via a short linker and was successfully exposed to specific antibodies.

The full sequence of lytC is encoded in 1488bp, but its cell wall binding domain was isolated by Imperial College London’s 2010 team as the region encoded by the first 954bp. This means the natural function of lytC - cell wall turnover and autolysis for cell growth and separation - is removed from the recombinant protein we will use.

A GFP reporter.

Preliminary work of generating a reporter expression construct was done first. This reporter construct in question was produced by PCR amplification of the sequence encoding for green fluorescence protein (GFP) and double terminator from BBa_E0840 with simultaneous attachment of the B. subtilis consensus RBS embedded within the forward primer.

The product of this reaction was thus [consensus RBS + GFP + double terminator] and was inserted into pSB1C3 for future use. When expressed at the same time as RPMrel fused to the lytC cell wall binding domain the recombinant bacteria will resolve in green under UV illumination.

Amplification of BBa_K316037.

Exploratory work began by taking the construct submitted in BBa_K316037 as the starting point. BBa_K316037, originally submitted by Imperial College London’s 2010 team, contains the following regions in order: [Pveg promoter, spoVG RBS, cell wall binding domain of lytC, helical linker, elastase cleavage site, auto-inducing peptide and his tag]. The regions we wanted include Pveg promoter through to helical linker only.

RPMrel - a 9 amino acid peptide - can be synthesized de novo by PCR when embedded within a primer. We chose to codon-optimize the amino acid sequence for expression in B. subtilis and embed the resultant 27 nucleotide sequence in the design of a reverse primer. This reverse primer was then used in a PCR reaction that amplified [Pveg + spoVG + lytC + linker] from BBa_K316037 and simultaneously attached the sequence of RPMrel to the end of the linker. 

Design of that reverse primer is detailed as follows.
5' - [8bp cap] [7bp SpeI restriction site] [6bp reverse-complementary double stop codon] [27bp reverse-complementary sequence of codon optimized RPMrel] [15bp reverse-complementary overlap with linker] - 3'

The exact sequence is as follows.
5’ - GTTTCTTCACTAGTATTATTAACACATCGGGCGATCTTCGATCGGACAGGCCGCGGCTTTCGC - 3’ (63bp)

Assembly of PCR products.

The product of this PCR reaction in linear form comprised [Pveg + spoVG + lytC + linker + RPMrel]. Standard assembly methods were then used to join it in front of the aforementioned [consensus RBS + GFP + double terminator] construct in pSB1C3. Following this step, the module’s code comprised [Pveg + spoVG + lytC + linker + RPMrel + consensus RBS + GFP + double terminator]. The module’s was submitted to the Registry in this form (see BBa_K733007).

Further cloning work was done to ligate the module’s code into pDG1661, an integration vector for B. subtilis which was its final destination for characterization. For more details on pDG1661, see this document produced by the Bacillus Genetic Stock Center (BGSC).

BBa_K733007 is to result in the production of two gene products: 1) the lytC + RPMrel fusion protein localized to the cell wall, and 2) green fluorescence protein. B. subtilis cells transformed with the construct are expected to exhibit both binding affinity for HT-29 cells and green fluorescence under UV illumination.

The module is to be tested by washing fixed and stained HT-29 colorectal cancer cells with B. subtilis subtilis 168 transformed with the construct contained in BBa_K733007 in mammalian cell media. Following successive washes with phosphate buffered serum (PBS), the fixed cells will be imaged by fluorescence microscopy at a magnification resolving the HT-29 cells.

The same process will be conducted using B. subtilis subtilis 168 cells transformed with only GFP as a negative control. By comparing the relative localization of fluorescence between the two images, the effect of the lytC + RPMrel fusion protein on the recombinant bacteria’s binding ability can be visualized.