Family name : VU

Middle name : Khoa

First name : Bang

Sex : Male

Date and place of birth : March 28th 1958, Hanoi - VIETNAM

Nationality : Vietnamese

Civil status : Married, Two children.

Correspondence address : Ultra-structure laboratory, VUB, 65 Paardenstraat, B-1640, Sint-Genesius-Rode, Belgium Tel. : 32-2-359-02-09 Fax : 32-2-359-02-89 E-mail : khoabang@vub.ac.be Permanent address : 61 Hang Buom, Hanoi, Vietnam. University studies : Agriculture University N°1 - Hanoi, Vietnam.

D.V.M. (Doctor of Veterinary Medicine).

Year study : From Dec. 1975 - to Dec. 1980.

Job in Vietnam : Research scientist, Virology department, National Institute of Veterinary, Vietnam.

(from Apr. 1981 to Sept. 1991)

MSc in Molecular Biology (Oct. 1991-Sept. 1993) (Institute of Molecular Biology, VUB, Brussels).

PhD in Molecular Biology.(Oct. 1993- Dec. 1998) (Institute of Molecular Biology, VUB, Brussels)

DVM (graduated in December 1980). Agriculture University - Hanoi, Vietnam

M.Sc. in Molecular Biology (Oct. 1991 - Sept. 1993) Vrije Universiteit Brussel (VUB), Belgium. Title of MScThesis " Induction and characterization of Hog Cholera Virus specific antibodies "

Ph.D. in Molecular Biology (Oct. 1993 - Nov. 1998) Ultra-structure laboratory, Institute of Molecular Biology and Biotechnology, VUB, Belgium. Title of Ph.D Thesis "Llama antibodies and single-domain heavy-chain antibody engineering "

Laboratory techniques including cell culture and virology diseases in veterinary medicine.

Most of the techniques required in molecular biology, from the RNA, DNA manipulation such as mRNA extraction, cDNA synthesis, PCR, DNA purification, southern blotting, gene cloning, sequencing and transgenic in plant, mammalian cells, transcription and translation in vitro etc. to the protein manipulation such as expression, protein purification, gel filtration, ELISA, Western blotting and RIA.

Bio-computing: GCG and Phillip softwares

1994 - 1998 : Ph.D. student at the Ultra-structure laboratory, Institute of Molecular Biology and Biotechnology, Vrije Universiteit Brussel, Belgium.

1992 - 1993 : M.Sc student in Molecular Biology, VUB, Belgium.

1981 - 1991 : Research scientist- Virology department National Institute of Veterinary, Hanoi, Vietnam.

1975 - 1980 : University student, (D.V.M.).

1972 - 1974 : Secondary school.

1965 - 1971 : Elementary school.

It has been demonstrated that in camel serum, beside the conventional immunoglobulin IgG antibodies (H2L2 Abs), also exist other IgG isotypes, the so-called heavy-chain antibodies (H2 Abs) in which the light chains and the CH1 domain are absent. These H2 Abs are referred to as IgG2 (g2) and IgG3 (g3) based on their different behaviour upon binding to protein A and protein G (Hamer-Castermans et al., 1993). The variable domain sequences (VHHs) derived from the H2 Abs bear residues at positions 11, 37, 44, 45 and 47 that differ from those of the human VHs. These differences result in the increased solubility of the H2 Abs. Additionally, the long CDR3 frequently found in dromedary VHHs is believed to be an addtional source of diversity generation, a mechanism that seems to compensate for the lack of the light chain (Muyldermans et al, 1994).

The existance of the H2L2 and H2 Abs in llama and camel is a specific feature, not encountered in other species. Therefore, we focused on the study of the llama antibody genes. The variable domain including hinge genes were amplified by PCR, ligated into pBluescript vector, and sequenced. Analysis of forty clones revealed the presence of at least five different IgG gene isotypes. Two isotypes exhibit common features with the conventional Abs: they contain a CH1 exon followed by a hinge of 19 and 12 a.a. These clones are allocated to the IgG1 isotypes (g1a, g1b). Three other isotypes are classified as the VHHs in which CH1 exon is deleted. These VHHs are immediately followed by hinges of different lengths (35, 29 and 12 a. a.). The hinges of 35 a. a. and 12 a. a. are identical to the camel g2 and g3 identified earlier (Hamers-Casterman et al., 1993). The VHH clones that contain a 29 a. a. were shown to have a partial CH2 sequence more similar to that of the g2, therefore they are designated as the g2b. The llama VHs and VHHs were shown to belong to the VH III family. Llama VH (g1a, g1b) retain key amino acids (Gly44, Leu45 and Trp47) necessary for making contact with the VL to make an antigen binding site, whereas the VHHs (H2 Abs) adopt crutial amino acid substitutions at positions 11, 37, 44, 45 and 47. The long CDR3-loop (13.9 a. a. on average) is considered as a very important feature in interaction with an antigen. The appearance of the Cysteines in the CDRs of the VHHs suggests that they are required for the stability/functionality of the molecules by making the disulfide bond(s) between loop 1 and either loop 2 or loop 3. Based on the variability of amino acids within the H1-loop of H2 Abs in llama, we propose that the H1-loop in heavy-chain Abs may significally contribute to antigen binding (Vu et al., 1997).

A set of specific primers were designed to amplify selectively the Ig constant (C) domain genes from the hinge to the polyadenylation tail. Six different genes coding for the hinges followed by their C domains were obtained. Two of them belong to H2L2 Abs, whereas the other four C domain genes derive from the H2 Abs. The llama C domains were compared between them and with those of the mammalian IgGs classes. The essential features of amino acids for domain fold, for the effector function, glycosylation sites and the C1q complement binding site were found to be well preserved. Based on our results, it is expected that the llama CH domains fold as do those of the conventional IgGs. The differences between isotypes might have major consequences for biological activities such as phagocytosis, trans-epithelial transport, lymphocyte and complement activation and similarity to the human IgGs.

Based on the known sequences of the llama H2 Abs genes, four intact specific H2 Ab isotypes (g2a, g2b, g2c and g3) were constructed for expression in E. coli. The antigen binding VHHs used is the cAb-Lys3 (anti-lysozyme) (Ghahroudi et al., 1995). The recombinant gene was placed behind the pelB secretion signal for periplasmic expression and the H2 Abs expression is controlled by the lacZ promotor. The H2 Abs were purified by absorption on protein A, protein G and the functionality of the recombinant Abs was proven by ELISA. Yield of the H2 antibodies expressed in E. coli was estimated to be about 5 mg/litre culture. H2 Abs were also engineered for expression in mammalian cells (COS-7). The specific VHH used is cAb-TT2 (anti-tenanus toxoid) instead of the cAb-Lys3. The expression was under the control of the cytomegalovirus (CMV) promoter in pcDNA3 vector. The secreted H2 Abs in the medium were found to recognise specific antigen in ELISA. The yield of Abs was estimated to be about 5 mg/litre medium. Interestingly, the binding characteristics of H2 Abs expressed in E. coli and COS-7 were found to be the same as those of the corresponding subclasses purified from serum. These results show that the recombinant H2 Abs fold properly.

We also addressed the posibility to alter the affinity/specificity of a VHH (cAb-Lys3) by mimicking the gene conversion mutation used in chickens, rabbits or occasionally observed as a VH drift in mouse Abs. The rationale of this strategy is that the presence of conserved framework amino acid sequences around the CDRs should allow the recombination. Our result indicated that the nucleotides at the FRs regions should contain sufficient homology (at least 6 nucleotides) to allow gene conversion between two VHHs cloned in tandem, so that a new recombinant VHH is generated. Using gene conversion approach, we demonstrated that if one of the CDR is changed, or when the disulfide bond orientation is altered, the specificity of the original VHH (in case of cAb-Lys3) is lost. These preliminary results show the reasibility to use this approach for generating new antibodies.

To test the feasibility to express VHH in plant, the cAb-TT2 was recloned in pBinAR-TP-metC vector, and transformed into Nicotiana tabacum via Agrobacterium tumefaciens mediated gene transfer. Proteins in crude extract leaves were measured according to the Bradford’s. The specific recognition of the corresponding antigen by a VHH expressed in plant was confirmed by ELISA. However, the yield of the functional VHH expressed in plants was low (about 0.1% of total leaf proteins).

In addition, a VHH was expressed in plants as a protein fused to the N-terminus of the chloroplast transit peptide. This implements that the transgenic gene product (cAb-TT2 protein) should be transported into the chloroplast lumen, after which the transit peptide should be cleaved off. However, no antibody activity could be found in the extracts from chloroplasts of transgenic plants.