Team:Hong Kong-CUHK/methods

The construction of the MFO

Figure 1: MFO

Our first construct is called the MFO, which stands for the magnetosome forming operon.

From research papers, we found that there are different operons that are used to produce and regulate the magnetosome formation. And among them, it is the mamAB operon that is the crucial operon that is needed for the production of magnetosome. Other operons like mamGFDC, mamXY, mms6 et cetera are more frequently used for regulating the size and shape and the biominerlization that facilitate the formation of magnetosome[1].

Therefore, in our project this year, we would only insert the mamAB operon from MSG to our bacteria Azotobacter Vinelandii, hoping to test on using minimal number of genes to produce functional magnetosome. This aims at assisting the magnetosome formation progress making it easier to form for future research developments.

Figure 2: mamAB operon

As it is shown in the diagram, the mamAB operon is a rather large operon consisting of 16.4 KB. Therefore it means the normal cloning strategy, by putting the gene of interest into a vector and transforming it into the targeted bacteria will have a lower successive rate.

Therefore, a new method was required to help the transfer of such a large operon into the Azotobacter vinelandii. Our new method involves the making of a template vector first, follow by homologous recombination and then random integration (Figure 3).

Figure 3: Cloning strategy for MamAB operon

Furthermore, as the mamAB operon is too long for normal insertion, we have decided to firstly split it into two. Once all recombination and integrations are done, the split mamAB operon should be recombined again inside the azotobacter genome.

The start of the mamAB operon is the mamH gene with the ending gene being mamU. The mamAB operon will be cut at around half way of the operon. And the cutting point is between the mamN and mamO genes. In consequence, the first half of mamAB will be starting with mamH, ending with mamN; and the second half will be starting from mamO, ending with mamU.

Hence, in order to put our gene of interest (mamAB operon) into a vector, we need to produce a template first and put it into E .coli (the strand of E. coli we are using is BL21). The template which we used is a vector which includes J13002 (consists of the constitutive promoter R0040 and the RBS); the flanking sequences; and the double terminator. All in between the multiple restriction sites.

Flanking sequences are actually the first 250 bp and also last 250 bp of the whole linkage of genes which we want to “flip into” our template. Therefore in order to flip in the first half of the mamAB operon, in other words, the one starting with mamH and ends with mamN, we need to have a total of two flanking sequences. This is done by making the first 250bp of mamH as the first flanking sequence and also the last 250 bp of mamN as the second flanking sequence in the template. As BL21 (the strand of E. coli) has a special enzyme RecA, therefore it is able to perform in vivo homologous recombination[2].

With these homologous parts as our template, we can then flip in both parts of the mamAB operon into the vector which should be much easier than ligating such a long mamAB operon into the vector by usual method. As vectors by usual method usually can’t uphold more than 10kb gene of interest.

Note that homologous recombination need a high concentration of the two parts of mamAB in order to flip in successfully.

Then, after homologous recombination, the complete mamAB operon should now be inside the plasmid.

Furthermore, by cutting one restriction site, the plasmid containing the whole mamAB operon with its promoter, RBS and double terminator and an antibiotic resistant gene will be linearized for random genome integration afterwards.

The linear fragment will then be feed into the Azotobacter vinelandii by the method called random integration. In other words, the linear DNA will just go inside the Azotobacter and be randomly integrated into its genomic DNA which is a “self-property” of the Azotobacter vinelandii.

(Gel photo to be added)

The insertion kit:

Figure 3: The structure of the trans-membrane mamC protein

This construct is actually a simple one, consisting of a mamC gene, a gene coding for a trans-membrane protein [3] (Figure 3) on the magnetosome membrane, in a vector. However, unlike usual recombinant methods in which we put our insert between the multiple restriction sites, we are putting our mamC gene in front of it. Through this method, it enables is to attach any protein we desired on the magnetosome membrane just by fusing it with the mamC gene by inserting it between the multiple restriction sites. (For your interest, this is done by removing the stop codon of the mamC gene and the start codon of the desired protein, for example an antibody, making it a mamC fused protein).

As we are now putting multiple restriction sites behind mamC, therefore we can insert any desire genes afterwards. Thus, we name it our insertion kit.

Hence, as you know, the mamC gene is put in front in the multiple restriction site. Therefore, normal method of inserting the gene of interest by restriction digestion and ligation is not possible.

Instead, the whole plasmid has to be synthesized by ourselves using PCR method.

The PCR of this plasmid involves the use of 8 primers which can produce four fragments: B0015 – the double terminator; backbone with kanamycin; J13002 (which consists of the constitutive promoter R0040 and the RBS); and mamC.

Through the over-lapping PCR of these four fragments, one linear fragment will be produced. And we have designed PstI sites at both the start and the end of the fragment.

Consequently, by cutting the PstI site and ligate it, a circular plasmid will be produced.

Figure 4: Pstl ligation


Then finally, the antibody is put into the plasmid by restriction digestion and ligation.



1. LOHßE, Anna, et al. Genetic dissection of the mamAB and mms6 operons reveals a gene set essential for magnetosome biogenesis in Magnetospirillum gryphiswaldense. Journal of bacteriology, 2014, 196.14: 2658-2669.

2. CHEN, Zhucheng; YANG, Haijuan; PAVLETICH, Nikola P. Mechanism of homologous recombination from the RecA–ssDNA/dsDNA structures. Nature, 2008, 453.7194: 489-494.

3. XU, Jun, et al. Surface expression of protein A on magnetosomes and capture of pathogenic bacteria by magnetosome/antibody complexes. Frontiers in microbiology, 2014, 5.