Team:UNIZAR/Engineering

Overview

Nanobodies are proteins that selectively bind to an antigen. Due to their smaller size, they present advantages in diagnosis and therapeutics compared to antibodies. Their specificity relies on their complementarity-determining regions or CDRs (CDR1, CDR2, and CDR3), unique nanobody regions involved in antigen binding. Inspired by nature, we aim to recombine in vitro the different CDR genetic sequences of already characterized nanobodies at random. Our goal is to create a library of artificially developed nanobodies that bind to a specific antigen. These newly formed nanobody gene fragments are expressed in E. coli, generating a library for the screening of nanobodies with higher binding affinity than the starting ones. This way, our system improves the traditional method for nanobody production, being faster, cheaper, and without using animals.
Engineering cycle

Figure 1. Science, business, and law interrelation in technology transfer

RESEARCH. How does nanobody formation happen in camelids lymphocytes?

The natural process of nanobody formation consists of recombination of three different fragments called V, D and J (1). There are 17 different V regions, 8 Ds and 7 Js in Alpaca's (Vicugna pacos) genome that can form these nanobodies (2). The recombination of these regions creates a new nanobody with three CDRs (Complementary Determining Regions) that allow it to recognize a certain epitope of an antigen.
Usually, synthetic nanobodies libraries are done by randomizing CDRs from a stable nanobody called scaffold (3). From our point of view, this strategy has a problem, the 3D nanobody structure doesn´t change as it depends on FR regions. We want to tackle this limitation.

Figure 1. Nanobody structure. In blue, green and red are represented CDRs 1, 2 and 3 respectively.

IMAGINE. How can we design a nanobodies library by mimiquing the natural process that occurs in vivo?

Our idea at first was to mimic what happens in nature and artificially recombine these fragments by Gibson Assembly or similar methods. We wanted to have all the possibles V, D and J regions with homologous fragments to perform this recombination.

DESIGN. What homologous regions are going to be used?

We soon found out that it would not work as CDR3 is located in between fragments D and J. CDR3 sequences have a really high variability as terminal deoxynucleotidyl transferase insert mismatches and random nucleotides at intersection regions between D and J regions (4). Without homology between different D or J regions, it is impossible to bind them in vitro.

Figure 2. In vivo recombination and generation of nanobodies. V (pink), D (blue) and J (green) fragments recombine to generate new nanobodies. CDRs are represented in grey. As it can be seen, CDR3 is located in the intersection between D and J fragments, whose sequence variability makes impossible in vitro recombination of V, D and J fragments as there is no constant region.

Therefore, we decided to tackle the design with another strategy. We discovered that nanobody sequences have some really well-preserved regions, with scarce variations even between different species. They are known as FR and are responsible for nanobody's 3D structure (1). Two of them are just between the CDRs regions, which makes them perfect to use as overhangs for Gibson assembly.
We decided to select nanobodies from different origins against GFP as a "proof of concept" because it is easy to quantify antigen as it is a naturally fluorescent protein (measurements at 510nm after excitation at 488nm), perfect for evaluating at the end of the process the efficiency and enrichment on our newly formed nanobodies.
With our new strategy in which different species' nanobodies are mixed, not only the CDRs are randomized but also the FR regions allowing us to obtain a higher variety of nanobodies as 3D structure directly affects CDRs structure and exposition. In this way, we will solve synthetic libraries limitation described above.

BUILD. What nanobodies and which regions are we going to use for our design?

We divided each nanobody into three fragments separating them by the FR conserved regions, keeping these regions at the ends. As each fragment contains one CDR, from now on we will refer to each fragment as CDR1, CDR2 and CDR3, respectively (although CDRs in fact are just a region and not the entire fragment). All our nanobody sequences were chosen from the online Single Domain Antibody Database (5), except one, that was taken directly from Potsdam Bioware 2021 iGEM team part registry: BBa_K929104 (6).
We first chose different antiGFP nanobodies from 4 different origins: arabian camel (Camelus dromedarius), llama (Llama glama), alpaca (Vicugna pacos) and synthetic. The sequences' codons were adapted for expression in E. coli in order to improve the efficiency of the system.
The final nanobody regions we have designed can be found in Parts.
We compared all their sequences in order to find the perfect complementary regions (FRs). In this way any CDR1 can bind to any CDR2 through Gibson assembly as they share the same sequence and any CDR2 can bind as well to any CDR3 as they share another sequence.

CDR1-CDR2 assembly sequence: TGGTTCCGTCAGGCGCCGGGTAAA
CDR1-CDR2 assembly sequence: TGGTTCCGTCAGGCGCCGGGTAAA


Figure 4. Nanobody sequence alignment with MAFFT in Benchling. Red dots determine regions with sequence differences between nanobodies. Two internal regions, labelled constant 1 and 2 (in blue) are completely conserved and thus useful for recombination (7).	Figure 5. Representation of recombination step. CDR1-CDR2 assembly sequence in blue and CDR2-CDR3 assembly sequence in pink. Each green fragment represents a nanobody fragment, with this technology fragments from different nanobodies recombine creating a whole new nanobdy.

TEST. Do these regions recombine well and form new nanobodies?

Firstly, we successfully amplified by PCR each CDR fragment that was synthesized by IDT. In those PCRs we used these primers:

CDR1 forward: ACCTCAGAATTCTATGGCAGACGTTC
CDR1 reverse: TTTACCCGGCGCCTGACG
CDR2 forward: TGGTTCCGTCAGGCGCC
CDR2 reverse: GTCACGAGAGATGGTGAAACGACC
CDR3 forward: GGTCGTTTCACCATCTCTCGTGAC
CDR3 reverse: GCCGATCGACTTTGCTAAGCTTC

With enough quantity of CDRs we managed to assemble our CDRs into recombinant nanobodies composed by CDR1, CDR2 and CDR3. On our first approach, we used the CloneEZ PCR cloning kit with CDRs from one single Alpaca nanobody (nanobody A12). We succeeded in making the recombination:

Figure 6. Electrophoresis of nanobodies PCR-amplified nanobodies (NBs). Negative controls are labelled as c-. A12 NB was obtained from Clone EZ assembly of three fragments: its CDR1, CDR2 and CDR3. Cloning was successful as its band after PCR amplification Cl-Ez corresponds to the A12 nanobody sequence, generated in vitro through the combination of its CDRs. The rest of bands correspond to complete chemically synthesized nanobodies, which act as positive controls for the PCR.

LEARN. What happened?

The cloning efficiency of the recombination product in a digested plasmid with restriction enzymes was very low, so trying another cloning strategy is necessary in order to get better results. We focused in finding a new approach that allowed us to continue with the working pipeline.

Figure 1. Science, business, and law in</div>terrelation in technology transfer

IMPROVE. How can we increase the efficiency of the recombination process?

Since the two-step assembly was not successful we took another approach. We tried direct Gibson Assembly cloning by introducing in the CDR1 and CDR3 a homologous sequence with the plasmid sequence in the insertion site.

Figure 7. Gibson Assembly design. The blue and red regions in the plasmid are also present in the new nanobody sequence as it can be introduced by PCR using long primers.

DESIGN. New primers for Gibson Assembly!

We designed two new primers that allowed us to introduce a sequence in the CDR1 and CDR3 for recombination with the plasmid. The new primer sequences are shown below (region complementary to the plasmid sequence appear in bold type).

CDR1 forward with homology sequence: GCCGCCGCGGCCGCGAATTCTATGGCAGACGTTCAGCTGC
CDR3 reverse with homology sequence: TGTCGAGCGGCCGCAAGCTTCTAGAAGAAACGGTAACCTG

Figure 8. introduction of homologous sequences in the 3' and 5' end of the recombined nanobodies by PCR.

TEST. Does this new approach provide better results?

After cloning by Gibson Assembly the recombinant nanobodies pool in a plasmid, in order to know if the insert was well cloned, we performed PCR with primers that hybridized outside the MCS. If a short amplified product (200 pb) was detected it means that the cloning didn't go really well, if a long product is detected (500 pb) the strategy worked. After analyzing the amplification products in an agarose gel we observed that the cloning problems persisted, unfortunately.

Figure 9. Agarose 0.8% electrophoresis gel. The 5th lane corresponds (without taking the marker into consideration) to amplification of the insert with the nanobodies inserted in it. A very weak line was detected at 500 pb while there is an intense one at 200 pb.

However we were able to randomly recombine the whole pool of sequences and obtain newly formed nanobodies, we were able to generate in vitro libraries! This is a huge step forward in the project results. In order to confirm the variability generated by this method, we wanted to sequence the new nanobody pool obtained in collaboration with ARIA team from Barcelona . Although we were not able to send them a sample enough enriched in nanobody containing plasmids

Figure 10. Electrophoresis of nanobody sequences in a 0,8% agarose gel. Second-line corresponds to the recombination product of all CDR sequences available.

PLASMID DESIGN AND PURIFICATION

RESEARCH. What type of expression vector are we using for our nanobody library?

We wanted to have a system that allowed us to test the newly formed nanobodies in an easy way. We discovered that there are bacterial proteins called intimins and autotransporters that are used to display small proteins in the surface of bacteria such as E. coli (8). These proteins are currently being used to display nanobodies and other small proteins (9,10), so we decided that it could be a good expression strategy for our project.

DESIGN. Designing the cloning strategy

For the plasmid design, we asked pSEVA repository (11) and contacted Dr. Esteban Martínez, a researcher from Zaragoza that is currently working in the CNB. He sent us 3 plasmids from the pSEVA repository: pSEVA228-I, pSEVA238-I, pSEVA228-AT. They differed on its nanobody expression strategy: intimin-transported (228I, 238I) and autotransporter (228AT) and on their copy number, being low for 228I and medium for 238I. Both plasmids confer E. coli resistance against kanamycin.

Figure 11. pSEVA228-I and pSEVA228-AT maps. Both of the systems allow displaying nanobodies on the surface of bacteria when cloned in the MCS (multiple cloning site). The main difference lies in the MCS position, in pSEVA228-AT it is in 5' end of AT but in pSEVA228-Intimin it is in the 3' end.

In order to achieve nanobody display, their sequence must be cloned in the MCS of the plasmids. We used the restriction site EcoRI in the 5' end and HindIII in the 3' end (Figure 7), these sites were introduced in the sequence by chemical synthesis.

Figure 12. Representation of MCS of plasmids (blue). The restriction sites used are highlighted in yellow. Figure made in SnapGene.

It is important to design the sequence of the nanobody in order to maintain the ORF in frame with the signal peptide (pelB) included in the plasmid and the fusion protein that will allow membrane display. The signal peptide guides the system to the bacterial periplasm, and finally, the intimin (or the autotransporter) is located in the external membrane of E. coli exposing the nanobody. Taking this into consideration, the sequences were designed as follows.

Figure 13. Sequence of a consensus nanobody. Restriction sites for EcoRI and HindIII in purple, in by their sides there must be a nucleotidic tail to ensure that the restriction enzyme is able to cut within its site. In green, there are represented nucleotides that have been introduced in order to maintain the ORF and ensure good expression. The starting codon is highlighted in red and the STOP codon has been removed as it is included in the plasmid in the E-tag sequence in pSEVA228-I and pSEVA238-I and in the autotransporter sequence in pSEVA228-AT.

The plasmids also contain an E-tag that allows us to detect protein expression levels by using an anti-E-tag antibody in western and ELISA assays.

Figure 14. Expression system based in intimins displaying the nanobody in the external membrane of E. coli.

BUILD. What cloning technique will be used for nanobodies display?

Plasmids pSEVA228-I, pSEVA238-I and pSEVA228-AT were originally expressed in E. coli XL1-Blue and purified via miniprep. Afterward, they were digested using EcoRI and HindIII restriction enzymes in order to open the MCS. The recombinant nanobody DNA was then cloned into the plasmid and transformed into chemically competent E. coli.

TEST. First purification of pSEVA plasmids

Firstly, plasmid purification was evaluated through quantitative and qualitative methods: nanodrop DNA concentration measurement and visualization at agarose gel electrophoresis. We found that our purification was inefficient as the plasmid concentration we obtained was lower than expected for low-copy plasmids (see results).
The lack of high plasmid concentrations influenced the rest of the experiments performed. Digestions, clonings, transformations and gel visualization weren´t optimal as very few plasmids was obtained and its quality differed in every extraction.

LEARN. What can we do now?

Doing good plasmid purification is a critical step in order to have good results in the next steps of the molecular design. In order to improve the low quantities of purified plasmid, we decided to try different purification protocols miniprep kits, but all of them were proven insufficient. As it seemed the miniprep kit was not a problem, we began to think that the issue could be the E. coli strain we were working with.
Moreover, we noticed that by purifying directly from a plate colony, we could obtain higher concentrations than from liquid culture. Consequently, we decided to prepare new glycerol stocks from the plates we obtained a greater amount of plasmid from.

DESIGN. Is it worth it to change the strain that keeps the plasmids?

Bearing in mind what we have learned from our previous experiences, we decided to transform our plasmid into a new E. coli expression strain: DH5α as it lacks RecA so no homologous recombination between plasmids can occur and endonucleases so plasmid degradation is difficult (12). This strain is usually used for plasmid keeping and purification.

BUILD. E. coli DH5α transformation and plasmid purification

This E. coli strain was transformed with the best previous plasmid purification by thermic shock with the three plasmids.

Figure 15. Representation of E. coli chemical transformation protocol and selection in plaque in presence of kanamycin.

TEST. Plasmid purification results

Unfortunately, plasmid concentration results remained below our expectations and complicated further work with the plasmids.

LEARN. Asking for help is always a great option!

In sight of the problems we were having with plasmid purification, we contacted ARIA iGEM team and Dr. Esteban Martínez in order to get their support in this step we were stuck. Thanks to the advice and protocols they provided us with, we were able to increase the plasmid concentration.
Now we are able to perform experiments with higher quality and efficiency as the plasmid is no longer an issue!

References

Muyldermans S. Nanobodies: natural single-domain antibodies. Annu Rev Biochem. 2013 Jun ;82:775–97.
IMGT Home page [Internet]. [cited 2021 Oct 10]. Available from: http://www.imgt.org/
Muyldermans S. A guide to: generation and design of nanobodies. FEBS J. 2021;288(7):2084–102.
Richard A. Goldsby, Thomas J. Kindt, Janis Kuby BAO. Kuby Immunology. 2002
Single Domain Antibody Database [Internet]. [cited 2021 Oct 10]. Available from: http://www.sdab-db.ca/
Part:BBa K929104 - parts.igem.org [Internet]. [cited 2021 Oct 10]. Available from: http://parts.igem.org/wiki/index.php/Part:BBa_K929104
Editor · Benchling [Internet]. [cited 2021 Oct 10]. Available from: https://benchling.com/editor
Dalbey RE, Kuhn A. Protein Traffic in Gram-negative bacteria - how exported and secreted proteins find their way. FEMS Microbiol Rev. 2012;36(6):1023–45.
Veiga E, De Lorenzo V, Fernández LA. Autotransporters as scaffolds for novel bacterial adhesins: Surface properties of Escherichia coli cells displaying Jun/Fos dimerization domains. J Bacteriol. 2003;185(18):5585–90.
Salema V, Fernández LÁ. Escherichia coli surface display for the selection of nanobodies. Microb Biotechnol. 2017;10(6):1468–84.
SEVA plasmids – Standard European Vector Architecture – : Your database and repository of standard, modular plasmid vectors for constructing complex bacterial phenotypes [Internet]. [cited 2021 Oct 11]. Available from: http://seva-plasmids.com/
NEB® 5-alpha Competent E. coli (Subcloning Efficiency) | NEB [Internet]. [cited 2021 Oct 19]. Available from: https://www.neb.com/products/c2988-neb-5-alpha-competent-e-coli-subcloning-efficiency#Product Information