William C. Hwang
Burnham Institute for Medical Research, La Jolla, California
A structural biologist has great expectations for llamas' small antibodies.
Llamas aren't just unusual and exotic looking: their antibodies are also a reason for much excitement. Made entirely of heavy chains, they are about half the size of those found in humans and many other vertebrates, which are normally composed of both heavy and light chains. When it comes to therapeutic applications, these larger antibodies are hard to store and deliver. But llama and other camelid antibodies demonstrate superior heat-stability and solubility, without compromising affinity or specificity, making them an attractive alternative.
Robin Weiss of University College London and his colleagues isolated three llama antibodies, known as 'neutralizing' antibodies, that can broadly prevent multiple HIV subtypes from infecting cells (A. Forsman et al. J. Virol. 82, 12069–12081; 2008). They began by creating an antibody library from two llamas immunized with the HIV gp120 antigen. To select for neutralizing antibodies, antibodies were raised against one HIV subtype but cross-screened against multiple subtypes. The researchers also included a competitive elution step to select antibodies that can compete with binding by CD4, the primary HIV receptor on human T cells. It remains to be seen how these neutralizing antibodies fare in animal studies and where they bind in atomic detail.
Intriguingly, there have been reports of several potent, broadly neutralizing human antibodies (for example, F10 and CR6261 against influenza's haemagglutinin) in which only heavy chains are involved in antigen binding — reminiscent of the situation of llama antibodies. These studies corroborate that the heavy chain alone can mediate broad neutralizing activity, and invite speculation that this may be a special strategy engaged by the human immune system to reach cryptic binding sites. Llama antibodies may be even better suited for those hard-to-reach targets.
Comments
Dr. Hwang’s post is incorrect on several items.
Not all Llama or camelid antibodies are heavy chain only. This is mentioned in the Forsman paper, so it is disappointing Dr. Hwang did not even pay attention to the introduction of the paper. Llama’s have normal IgM, may or may not have IgD (literature is unclear), have at least one completely normal 2H2L IgG1 subclass, and have “VHH” only IgG2 and IgG3 (see S. Muyldermans et al (2009) Vet Immunol Immunopathol 128: 178-183, PubMed 19026455). The VHH only subclasses do not bind kappa or lambda light chains because they lack the CH1 domain. Heat stability of llama VHH antibodies may be enhanced by a novel intrachain disulfide bond between CDR1 and CDR3. Apparently not all VHH antibodies have this intrachain disulfide bond and it is unclear to me how much of the improved heat stability is due to this disulfide bond, vs not having CH1 vs not having CH1 and light chains.
Llama and camel VHH antibodies approximately 90 kDa, which according to my math is not “about half the size” of the 150 kDa of human and other 2H2L IgG’s. The lack of binding of the light chains accounts for a loss of 50 kDa, and the loss of the pair of CH1 domains accounts for a loss of about 10 kDa.
Are llama and camel VHH antibodies really that exotic? X. Zou et al (2007 J Exp Med 204: 3271-3283) reported that light chain deficient mice (kappa and lambda gene double knockout) make VHH antibodies that lack CH1. They also provide RT-PCR based evidence that normal mice make VHH type transcripts lacking CH1. This suggests that somewhere in the thousands of mouse IgG proteins gels that have been published there ought to be some with ~90 kDa mouse VHH bands that were missed. Zou et al also cite human H chain disease as also being VHH.
Posted by: George McNamara | May 31, 2009 07:32 PM