MHC-Independent Allorecognition of Invertebrates—A Link between Invertebrate Histocompatibility and Vertebrate Adaptive Immunity?

The molecular basis of the adaptive immunity of vertebrates, preserved from sharks to humans, is peptide presentation by the highly polymorphic major histcompatibiliy complex (MHC). Histocompatibility, i.e., the ability to distinguish the organism’s own cells from another organism’s cells, is found throughout the metazoan species. MHC, the key molecules for allorecognition in all vertebrates from sharks to humans, have not been found in the animal kingdom in species other than vertebrates, even when the closest relatives of vertebrates, the jawless fish, were studied (1).

It is against this background that the exciting findings of the Stanford group in the Monterey Marine Station are a genuine breakthrough. They define an ancient form of MHC-independent allorecognition (2). The authors studied a species of chordates, the ascidian Botryllus schlosseri, a small marine organism. The lifecycle of this species includes a free-living sexual stage and a sedentary asexual stage. In the asexual stage colonies of asexually derived, genetically identical individuals, so-called zooids, are embedded in a gelatinous mass, are united by a common vascular network, and grow by budding. When different colonies approach each other, the foremost protruding tissue, so-called ampullae, come into contact. It had been known that such encounters may lead to two opposite outcomes: Either the ampullae fuse and eventually circulating stem cells are exchanged through the interconnecting vascular systems, or a rejection reaction occurs (not unlike the rejection reaction in vertebrates), i.e., the ampullae are destroyed by necrosis, the vascular systems fail to fuse, and there is evidence of local inflammation.

Undoubted analogies exist with the allorecognition in jawed vertebrates where success or failure of grafts is determined by the molecules encoded by the MHC-I and MHC-II genes (3). In the nonvertebrate species B. schlosseri, the novel finding is that here allorecognition is under the control of a gene completely unrelated to MHC. The decision whether fusion or rejection occurs is governed by a single, highly polymorphic locus, the fusion/histocompatibility (FuHC) gene (4): When two colonies share one or both FuHC alleles they will fuse; if they don’t share any common allele, rejection occurs. The analogy to MHC goes even further: FuHC is highly polymorphic. Most populations of B. schlosseri contain tens to hundreds of alleles (4).

The aim of the study by De Tomaso et al. was the molecular characterization of the FuHC genes. To this end the authors bred B. schlosseri in captivity and developed inbred lines homozygous for different FuHC genes. Using such FuHC-defined cultures, the FuHC locus was mapped using techniques such as amplified fragment length polymorphism, bulk segregant analysis, and genomic walking. The FuHC gene was found to code for a transmembrane protein with an extracellular Ig domain. The open reading frame predicted a highly polymorphic protein of 1007 amino acids. Two Ig domains, which differed with respect to their genomic structure, were homologous to Ig superfamily members (IGSF4D) of vertebrate species. The authors also predicted a secreted form arising as a result of alternative splicing and a series of further variants of unknown function.

The functional importance of this highly polymorphic gene was then evaluated by testing its relation to the above described functional histocompatibility reaction as assessed by the fusion/rejection assay. The polymorphisms of the FuHC gene segregated absolutely with this histocompatibility reaction. Fusion was consistently seen in animals that shared an identical allele and, conversely, rejection was never seen when an allele was shared.

What are the implications of the finding of this first metazoan histocompatibility ligand identified outside of the vertebrate lineage? In a seminal paper, Burnet had argued that the ability to discriminate between multiple different ligands is shared by immunity and by histocompatibility (5). Based on this consideration Burnet put forward the hypothesis that the origins of the vertebrate adaptive immune system were rooted in the histocompatibility systems of more primitive metazoic organisms. Is the FuHC molecule the missing link?

Certainly there are structural similarities between the vertebrate MHC molecule and the FuHC type I transmembrane protein, which contains multiple extracellular Ig domains. Nevertheless, it is not a strict homologue. It is known that the adaptive immune system arose in ancestors of the jawed vertebrates approximately 500 million years ago. While homologues of immunoglobulins, T cell receptors, major histocompatibility complexes (MHC), and recombination activating genes (RAG) have been found in all jawed vertebrates, none of these genes or close homologues have been identified in jawless vertebrates or invertebrates (6). It has been postulated, however, that the molecules involved in adaptive immunity go back to “fossil” genes of invertebrates. It has been suspected that primitive species without the machinery for recombination (RAG) contain a “primordial receptor” (6)—is FuHC a candidate?

With respect to how the FuHC molecule works, the authors came up with an interesting suggestion. In B. schlosseri, individuals sharing only a single allele are compatible. One possibility would be that two identical molecules “recognize” each other by homotypic interaction—but an alternative appears more attractive. Apart from recognizing foreign antigens in the context of self MHC, a more primitive complementary defense system has recently been described in vertebrates, i.e., missing-self-recognition (7), which is operative in natural killer cells. When cells don’t express “self” molecules they are attacked and eliminated. This mechanism may be useful to prevent aberrant tumor cells that fail to express MHC products, thus avoiding detection and destruction. De Tomaso et al. speculate that this natural killer cell system in vertebrates, which has a forerunner in invertebrates, e.g., in the ascidian Ciona intestinalis (8), and at least the mode of operation has also been adopted by FuHC.

The most challenging possibility is that FuHC represents an undiscovered Ig-based system of allorecognition that still exists in vertebrates, analogous to the missing self recognition mode in natural killer cells, which interact with ligands other than members of the MHC family (non-MHC ligands) (9).

It is fascinating how much can be learned by going back to the roots and studying primitive model organisms that share key mechanisms with the higher species, beautifully illustrating the “unity of nature” that had been so well anticipated by the poet and philosopher Johann Wolfgang von Goethe (1749–1832).