V m sequences of the monoclonal APL

The first question to be raised about the antibodies LJ-1, AH-2, DA-3 and UK-4 is how relevant they are to the pathogenesis of APS. The patients A.C., L.C. and D.H. had SLE and either serum APL or LA, but only A C. had

1 7 0

developed a major diagnostic feature of APS (fetal loss) and none of the three gave an unequivocal history of thrombosis. Furthermore, these mAb did not show the cofactor dependence characteristic of APL found in APS. One of the mAb (UK-4) demonstrates lupus anticoagulant activity.

On the other hand, these are the first (and so far the only) human monoclonal IgG APL which show specificity for anionic PL and do not also bind DNA. They are thus of great interest in determining the sequence characteristics which might distinguish the binding sites of these two important autoantigens.

AH-2 and DA-3 are truly distinct antibodies and not simply subclones that arose in vitro, because they are of different isotypes. The isolation of two distinct, but clonally derived, monoclonal APL from the patient L.C. is important because it shows that this particular clone of IgG producing cells must be highly expanded in this individual. The fact that the Vh regions of

these mAb show accumulations of R mutations in the CDRs suggests but does not prove that this clonal expansion was driven by antigen.

An unusual feature of AH-2 and DA-3 is their use of the gene V5-51. The Vh5 family is small, with only two functional genes (Sanz et al 1989a). Vh5

genes are rearranged in the neoplastic clones of 20 -30% of patients with CLL (Shen et al 1987) but have rarely been shown to be expressed in functional antibodies. Functional Vh5 products have been reported in an IgM anti-insulin mAb (Sanz et al 1989b), an IgG anti-HIV mAb (Andris et al 1991), a monoclonal anti-DNA IgM (Manheimer-Lory et al 1991) and in IgE antibodies from patients with atopic dermatitis (van der Stoep et al 1993).

171

A high affinity anti-dsDNA Fab produced by the phage display method from lymphocyte cDNA of a patient with SLE also uses the V5-51 gene (Roben et al 1996). There were mutations in CDRs as well as a cluster of R mutations in FR3. Three R mutations in FR3 are also seen in AH-2 and DA-3 Vh,

though they are not the same ones noted by Roben et al. Perhaps FR3 may play an important role in the binding properties of antibodies produced by some Vh genes, as has been suggested by experiments in which FR3 mutations altered the affinity of mouse mAb for DNA (Katz et al 1994). The presence of many mutations in the FRs of AH-2 and DA-3 is otherwise difficult to explain, particularly as only one of these occurs at a mutational "hotspot". As described in section 1.4.3, these hotspots are preferentially targeted by the intrinsic hypermutation mechanism in B cells (Wagner and Neuberger 1996). The consensus sequence of these hotspots is A/G-G-C/T.

The Vh region of UK-4 is unusual in that mutations occur more commonly in the FRs than the CDRs. This could be due to chance alone (since FRs are longer than CDRs) but it remains surprising that so many mutations exist in the sequence in the absence of evidence of antigen driven selection. Only two of the 11 mutations in the FRs occur at mutational hotpots. Alternatively the nucleotide differences in FRs might not be mutations at all, but represent a highly unusual allele of V3-74 present in patient D.H. However, this is unlikely since this gene shows very little polymorphism. Only three V3-74 alleles have been discovered, which differ from each other at only two positions. The V3-74 gene is also used in the IgG anti-dsDNA mAb 35.21 (Winkler at al 1992) but in that case the pattern of R and S mutations is consistent with antigen driven selection.

172

The Vh regions of these IgG APL differ from those of IgG anti-dsDNA mAb in their lack of accumulation of basic residues in CDRs. Neither arginine residues nor the YYGS motif are seen in VhCDR3.

The inability to amplify LJ-1 Vh in these experiments can be explained in

one of two ways. Firstly, the amount of VhcDNA in the LJ-1 hybridoma cells might have been too low to give a detectable amount of PGR product after amplification. This is unlikely, because LJ-1 V;^ was amplified successfully from cDNA produced by the same cells, and these cells were secreting detectable quantities of whole antibody at the time of RNA extraction. A more likely explanation is that the primers used in PCR were unable to amplify LJ- 1 Vh, perhaps because it was encoded by a Vh6 or Vh7 gene. This could have been investigated further by synthesizing primers suitable for amplifying Vh6 or Vh7 gene sequences. This avenue was not pursued in order to concentrate on the development of the expression system.