Team:Edinburgh/RFC

From 2011.igem.org

BBF RFC xx: "BioSandwich" - a homology based assembly method using a library 
of standard parts

Chris French, Allan Crossman
September, 2011

1. Purpose

This Request for Comments (RFC) describes a strategy for using homology-based 
assembly methods to assemble parts from a library in any order.

2. Relation to other BBF RFCs

BBF RFC xx does not update or replace any earlier BBF RFC.

3. Copyright Notice

Copyright (C) The BioBricks Foundation (2011). All Rights Reserved.

4. Background

One of the main aspects of synthetic biology is the assembly of standard 
modular parts from a library to produce larger constructs. There are two main 
types of assembly strategy: hierarchical and homology-based.

The RFC10 format, which is standard for iGEM projects, is based on 
heirarchical assembly methods such as Standard Assembly and Three Antibiotic 
Assembly. The main advantage of RFC10 is flexibility: any combination of parts 
can be assembled in any order. The main disadvantage is speed.

Many iGEM teams now use homology-based assembly strategies, such as Gibson 
Assembly. These methods use PCR to create parts that have homologous ends. 
They are then assembled using the method of Gibson et al (2009). This method 
is relatively fast. However, the order of the parts in the final construct is 
entirely determined by the primers chosen for the initial PCR reactions. If a 
different order is desired, parts have to be remade with new PCR primers.

BioSandwich is a hybrid that combines many of the benefits of hierarchical 
assembly and homology-based assembly. In particular, it is useful for creating 
different combinations of the same parts in different orders.

5. Outline

BioSandwich parts are reusable because they come in a standard format (much 
like ordinary BioBricks) with restriction sites flanking the part. The 
restriction sites are BglII (agatct) in the prefix, and SpeI (actagt) in the 
suffix. However, these restriction sites are not used directly for assembly; 
instead they are used to attach short (~35 bp) oligonucleotides (hereafter 
"spacers"). These spacers serve two purposes:

	* They create homology between the end of one part and the start of 
another; this allows homology-based assembly.
	* They can incorporate short meaningful sequences such as ribosome binding 
sites, linkers for fusion proteins, etc. 

Any lab using BioSandwich will want to keep a small library of different 
spacers for different purposes. Once (carefully chosen) spacers have been 
attached to each part, homology based assembly can be carried out in a single 
reaction. Precisely which spacers have been attached to which parts will 
determine the order of the parts in the final assembly.

6. Formats

6.1. Normal Parts

Parts must be free of internal BglII restriction sites (agatct) and SpeI sites 
(actagt). Each part must be made with a BglII site at the start, and a SpeI 
site at the end. If compliance with RFC10 is desired, the full format becomes: 

	gaattcgcggccgcttctagagatct NNN NNN NNN NNt actagtagcggccgctgcag

After cutting with BglII and SpeI, we have the following (shown in frame, 
which is relevant for fusion proteins):

	5'    ga tct NNN NNN NNN NNt a         3'
	3'         a nnn nnn nnn nna tga tc    5'

When a part is used for fusion proteins, the "t" base at the start of the 
RFC10 suffix becomes the final base of the final codon in the part. Design of 
the part should take this into account. This is often done by making this 
final codon GGT, coding for an innocuous glycine residue. (If a fusion protein 
is not being made, no such consideration is needed. If RFC10-compliance is not 
required, this "t" base can be replaced with anything.)

6.2. Normal Spacers

Spacers are designed as a set of oligonucleotides that can be attached to any 
part via the part's restriction sites. Each type of spacer will be attached 
upstream of one part and downstream of another. It is necessary to create 
three different single-stranded oligonucleotides (spacer oligos or "spoligos") 
for each spacer so as to allow annealing and ligation at the restriction 
sites. They should be designed so that the spacer's non-ligating ends are blunt.

The format for the upstream spacers is as follows (both strands shown):

	5'    ct agc NNN NNN NNN g         3'               (forward spoligo)
	3'    ga tcg nnn nnn nnn cct ag    5'               (reverse spoligo ONE)

The format for the downstream spacers is:

	5'    ct agc NNN NNN NNN g         3'               (forward spoligo)
	3'         g nnn nnn nnn c         5'               (reverse spoligo TWO)

Note that the forward spoligo is used for both upstream and downstream 
attachment.

The content of different spacers should be varied to avoid homology. It is 
recommended that non-coding spacers have an in-frame stop codon, in case they 
are to be used after a coding part which lacks one.

6.3 Vector Part

A vector is made as a PCR product with a BglII site (agatct) at the 5' end and 
an XbaI site (tctaga) at the 3' end. Since XbaI produces sticky-ends 
compatible with SpeI, the vector is compatible with standard spacers.

If one wishes the final products (after insertion into the vector) to be 
RFC10-compliant, the format for a vector is as follows:

	agatct [nnn] tactagtagcggccgctgcag [ori, cmlR, etc] gaattcgcggccgcttctaga

Note that a few additional bases must also be present at the 5' and 3' ends to 
allow the restriction enzymes space to work. When spacers are to be annealed, 
the vector is cut with BglII and XbaI (not SpeI).

6.4. Start Spacer

For compliance with RFC10, the spacer that connects the vector to the first 
part must be in the following format:

Spoligos attaching upstream of first part:

	5'    ctagag NNNNNNNN g         3'               (forward spoligo)
	3'    gatctc nnnnnnnn cctag     5'               (reverse spoligo ONE)

Spoligos attaching downstream of vector:

	5'    ctagag NNNNNNNN g         3'               (forward spoligo)
	3'        tc nnnnnnnn c         5'               (reverse spoligo TWO)

This format ensures that the final product contains a standard RFC10 prefix 
and suffix.

7. Assembly

7.1. Digestion of Parts and Vector

Each part must be digested separately.

    * Parts: digest with BglII and SpeI in NEB Buffer 2
    * Vector: digest with BglII and XbaI in NEB Buffer 2 or 3 

Each tube gets:

	36 uL     Water
	 5 uL     DNA
	 5 uL     Buffer
	 2 uL     Enzyme 1
	 2 uL     Enzyme 2

The digestions are then left for 2 hours at 37 C. Afterwards they are purified 
with 5 uL glass beads, and eluted to 10 uL EB (protocol).

7.2. Ligation of Parts and Vector to Spacers

The parts and vector must now be ligated to the correct spacers, as follows.

Each tube gets: 

	10 uL     Water
	 5 uL     DNA
	 1 uL     Spacer pair #1
	 1 uL     Spacer pair #2
	 2 uL     Ligase buffer
	 1 uL     T4 DNA ligase

Tubes should then be left for 9 hours at 16 C. They must then be purified.

7.3. Homology-Based Assembly

The parts can now be assembled using a number of different homology-based 
assembly methods, including Gibson Assembly, Overlap Extension PCR, Circular 
Polymerase Extension Cloning, and others.

8. Authors' contact information

Chris French: c.french@ed.ac.uk
Allan Crossman: igem@nothinginbiology.com

References

Gibson DG, Young L, Chuang RY, Venter JC, Hutchison CA, Smith HO (2009) 
Enzymatic assembly of DNA molecules up to several hundred kilobases. Nature 
Methods 6:343-345 (doi: 10.1038/nmeth.1318).