Hidden Allotypes and Rheumatoid Arthritis. A Revolutionary Scenario Concerning the Etiology of R.A.

by Rune Grubb, Kerstin Eberhardt, Anders Grubb and Pierre Åman

Whenever Rheumatoid Arthritis (R.A.) is suspected the presence of Rheumatoid Factor (R.F.) is investigated. The presence of R.F. is one of the diagnostic criteria for R.A. listed by the American Rheumatism Association. Classical R.F., according to Waaler-Rose, consists of antibodies to rabbit immunoglobulin (Ig). R.F. is now defined as an antibody to the Fc portion of Ig of any species. A symposium dedicated to the 50th anniversary of the detection of R.F. was published in 1988 (1) but there is still a stumbling block as to why R.A. is associated with antibodies to rabbit Fc. Since then, it has also been found that R.F. as defined above occurs in several other diseases and in a few percent of normal elderly people. Consequently R.F. is not considered in vogue when it comes to understanding the nature of R.A.

It has however been shown that there is a multitude of different antibodies to Ig Fc and more than 50 species could easily be listed. Confusion is therefore created if R.F. is regarded as an entity.

By originally serendipitous observations (2, 3) it has been established that antibodies precisely detecting Mendelian Genetic markers of human Ig in the Gm system are common in R.A. and are rare in other diseases.

For some reason R.A. patients are thus prone to develop an immune response to human Mendelian Ig markers.

We will argue

By extensive studies (see 4, 5) the prerequisites have been established for a gene-technology based analysis of the possible role of Gm markers and anti-Gm´s in R.A. initiation.

Basic data on Gm allotypes

Figure 1.
H.Matsumoto and R.Grubb in front of wall charts of Gm distribution in Asia and the Americas.

Two apparent paradoxes

1. Numerous studies (6-11) have shown that anti-Gm´s of R.A. patients are often specific for other persons´ Mendelian Ig allotypes. This is surprising in a disease which is regarded as autoimmune.
2. Hidden or non-nominal allotypes are a paradox from the point of view of Mendelian genetics (see below).

Hidden allotypes

The Mendelian inheritance of Ig allotypes is firmly established by studies of tens of thousands of matings and their progeny. Absence of particular allotypes and their genes is therefore expected e.g. in homozygotes and in inbred strains. In apparent conflict with Mendelian law, "wrong", or hidden or non-nominal allotypes have been described in man, mice, rats, rabbits, mink and in the chicken. Hidden allotypes usually appear transiently and give rise to the corresponding anti-allotype antibody.

For example, the gene for G3m (g) should according to classical Mendelian law be absent in G3m (b+ g- ) persons homozygous for the G3m^b gene. Reciprocally, the G3m^b gene should be absent in G3m (b- g+) persons. PCR techniques appropriate for detecting fragments specific for one of these alleles in a surplus of fragments of the other allele were developed in our model experiments.

The data supporting the existence of hidden allotypes have hitherto been obtained by serological methods or by amino acid analyses of isolated serum proteins. Recently (12) we have defined human hidden allotypes at the genomic level. We have particularly studied the G3m(b) and G3m(g) allotypes because their gene sequences and restriction endonuclease cleavage sites are well-known (13). Partial sequence of the IgG3 gene containing the Gm b and g polymorphism is given in Fig 2.

Figure 2.  The  top image shows a physical map of the Ig heavy chain locus. Below is a partial sequence of the IgG3 gene carrying the polymorphic seqence encoding the G3m(g) and G3m(b) allotypes. The nucleotide differences between the Gm(b) and Gm(g) alleles are shown with the Gm(b) allele below on dark grey backround.Grey shadowwed sequences indicate the sequences used for PCR primers. Arrowheads show cleavage sites for the restriction enzyme RsaI that was used for analysis of the PCR products.

The result of nested PCR of G3m^b homozygotes showed that gene sequences characteristic for the hidden allotype G3m(g) was found in 14 of 14 R.A. patients and in 4 of 4 normal adults (12). The number of copies of these foreign genes is so small that Mendelian inheritance can be excluded.

Two possible explanations of the genetic enigma of hidden allotypes

I. Indications of a possible viral origin of hidden allotypes

Already in 1974 Bosma & Bosma (14) who demonstrated a "wrong allotype" in some Balb/c mice, suggested that the information could be carried by a virus. The knowledge of transduction of somatic genes by viruses has greatly expanded since then. We drew attention (15, 16) to the Herpes group of viruses in this context because

1. These viruses have been characterized as potential transducing viruses (17).
2. Several herpes viruses permanently reside in B cells.
3. They have sequence homologies to many human proteins. Ig HC and Ig switch sequences have been found in EBV and in herpes simplex as examplified below.

Table 1. Examples of the hijack of human genes by the Herpes group of viruses.

Using PCR Edinger 1997 (19) detected EB virus sequences in the synovia of 10 out of 11 R.A. patients.

Gm genes are normally excised in B cell progression. The Ig switch process may facilitate the transfer of these particular gene segments.
The well-known order of the Ig heavy chain constant region class and subclass genes on chromosome 14 band q32 is schematically depicted below i fig. 3. The Ig switch takes place by looping out and excision of segments of the IgC genes. Switching proceeds sequentially in 5´-3´ order. This means that IgG3 sequences, containing G3m genes, are excised in all switches to IgG1, IgG2, IgG4, IgA1, IgA2 or IgE. In contrast, IgA2 sequences are never excised in the switch process. Excision of IgG3 determining sequences is thus most frequent and followed, in that order, by IgG1, IgA1, IgG2, IgG4 and IgE sequences. It can be noted that the anti-allotypes observed in R.A. are chiefly directed against G3m and G1m allotypes, the genes of which are most frequently excised. To our knowledge, anti-A2m has not been observed in R.A.

Figure 3.
Scheme of the organization of the human immunoglobulin heavy chain locus. (S=Switch regions)

Comment: Anti-Ig G3 and anti-Ig G1 are commonly observed in R.A. but anti Ig A2 has not yet been observed.

II. Microchimerism

Microchimerism, i.e. the continued presence of rare foreign cells, especially in a female from a fetus, is an established phenomenon also in man (20). Cells with the characteristics of the transferred cells have been observed for up to 27 years and do, of course, regularly carry foreign genes. Based on observations of microchimerism, Nelson in 1996 (21) raised the question whether some autoimmune diseases may, indeed, be auto-alloimmune or allo-autoimmune. Conceivably, microchimerism could explain the presence of genes for hidden allotypes, particularly in women. A classical chimera is depicted in Fig 4.

Figure 4.
Classical chimera.

Why are hidden allotypes of interest in R.A.?

Anti-immunoglobulins detecting human allotypes within the Gm system are at least as common as conventional R.F. in R.A. patients. These anti-allotypes are present early in the disease and frequently persist for life. In contrast to "R.F." anti-Gm antibodies are rarely found in other diseases, except in EB virus infections and then only transiently. The presence of anti-Gm correlates with progressive R.A. disease (22). As mentioned, studies of anti-Gm´s in R.A. have surprisingly revealed that these anti-Gm´s may detect MENDELIAN ALLOTYPES of OTHER INDIVIDUALS. Caucasian R.A. patients may for example have anti-Gm´s uniquely specific for Mongolian allotypes. Recent British and Scandinavian studies employing monoclonal reagents showed that anti-Gm´s specific for other individuals genetic markers are significantly more common in R.A. patients lacking that very allotype.

As shown, genes for hidden allotypes are commonly present in the genome. Normally these genes are silent. If they come or are influenced by viral or other promotors/ enhancers they express themselves.

Aleutian disease of mink as an animal model for "hidden allotype disease"

Hidden allotypes appear in this lethal disease caused by a parvovirus and characterized by adundant R.F.-production (Fig 5.)

Figure 5.

Two recent findings of particular interest.
By reverse transcriptase PCR followed by sequencing of the PCR fragments it has been shown that mRNA encoding hidden allotypes were expressed in the synovia of R.A. patients. An Ig intron was spliced out as in normal Ig expression (12).

Table 2. Gm g and b allotype specific Reverse Transcriptase PCR amplification from inflammatory joint tissues in b/b and g/g R.A. patients.

The synovial protein p. 205 of Hain et al. (23), isolated from R.A. patients specifically stimulates T- and B-cells from disease sufferers. As illustrated in Fig 6 below, p. 205 contains a stretch of 11 amino acid identical with the Ig G3 constant chain in the immediate proximity of allotype specific sequences.

Figure 6.
Synovial protein p.205 of Hain et al (23).

A scenario for the initiation of R.A.

As outlined above

These observations call for an interpretation. When it comes to immunoglobulins, the very special feature of hidden or non-nominal allotypes is observed in man and all species investigated. They are not derived from Mendelian inheritance and their corresponding genes are present in low copy number. These genes may derive from viral transduction or from microchimerism - two possible ways for spontaneous gene transfer. These non-Mendelian genes may carry information for polymorphic Ig traits which are foreign to the recipient.

If they are expressed after the tolerance induction period in early life, an immune response in the form of a specific anti-Gm is expected - just as is observed in R.A. This corresponds to a minigene-transplantation and the consequence of an exposure to foreign gene products. Ig - anti-Ig immune complexes will appear and these might be arthritogenic as observed in experimental and iatrogenic serum sickness.

After the initation of an immune response to epitopes which are highly similar to the host´s own, epitope spreading is a well-known consequence for example in juvenile diabetes and many other so-called autoimmune diseases (24). Epitope spreading means that additional epitopes become involved during the progression of an immune response and this may be the basis of the polyclonality of anti-Ig´s in R.A. and the plethora of Ig - anti-Ig complexes observed in this disease.

In summary

R.A. is initiated when enduring expression of hidden allotype genes occurs. These non-nominal genes may arise by viral transduction of Ig genes which are normally expelled in the B-cell progression or by microchimerism. Epitope spreading gives rise to conventional anti-Ig-antibodies and a plethora of Ig – anti-Ig complexes. Some of these are arthritogenic and T-cell stimulants.

Link to references