In the late 1970s the world witnessed the heart-wrenching tale of David Vetter, the “bubble boy.” Born with X-linked SCID (XSCID),2 a complete deficiency of T cells and NK cells, David lived his entire life inside a plastic bubble, deprived of all human contact. Sadly, David Vetter died in 1984, as a result of complications following an allogeneic bone marrow transplant designed to cure his disease. Nearly 10 years later, the genetic defect responsible for XSCID was identified by Warren Leonard and colleagues and was reported in a landmark paper published in Cell (1) that is reprinted in this issue of The Journal of Immunology. Published in April of 1993, this paper used genetic linkage analysis to demonstrate that the gene encoding the IL-2 receptor γ-chain (IL-2Rγ) mapped to the same chromosomal location as the gene responsible for XSCID. Further, Noguchi et al. showed that three unrelated patients with XSCID had three different nonsense mutations in the IL-2Rγ gene, resulting in premature translational termination of the predicted IL-2Rγ protein in these individuals (1). Together, these data provided compelling evidence supporting the conclusion that XSCID is caused by the absence of a functional IL-2R γ-chain protein.
However, despite this important breakthrough, the true underlying cause of XSCID remained a mystery. This congenital disease is characterized by a complete absence of circulating T cells and NK cells but normal numbers of B cells, although the B cells in these individuals are nonfunctional (2). Before the discovery of the underlying genetic basis of the disease, a diagnosis of XSCID also entailed a thymic biopsy, which would reveal a small underdeveloped organ with little evidence of developing T cells (3). Successful treatment of XSCID infants by bone marrow transplantation demonstrated that the observed structural defects in the thymus were secondary to the hematopoietic defects, as donor-derived T cells would develop normally and reconstitute the recipient’s immune system (2). Additional studies on the B cells from XSCID infants indicated that at least some of the impaired B cell function was due to the absence of T cells, as addition of normal T cells to in vitro assays with XSCID B cells could promote B cell activation and Ab secretion (2). Yet, there remained some lingering sense that all was not completely right with the XSCID B cells.
Some clarity regarding these issues came from an unexpected source. Studies of heterozygous carriers of the XSCID disease gene—in other words, the healthy mothers of the XSCID baby boys—led to remarkable insights on the cell-autonomous requirements for a functional XSCID gene. Because all somatic cells in human females undergo X chromosome inactivation, a process that normally affects one of the two X chromosomes in each cell randomly, it was possible to determine which cell types required a functional XSCID gene by examining the patterns of X chromosome inactivation in each hematopoietic cell subset of these mothers. At first glance the answers seemed obvious. T cells in the carrier females all showed inactivation of the X chromosome carrying the disease gene, indicating a requirement for this gene’s function in T cells (4), whereas B cells and myeloid cells showed random X chromosome inactivation. However, more careful study of the B cell subsets in these women led to the interesting finding that although all of the circulating naive IgM+ B cells had a random pattern of X chromosome inactivation, surface IgM-negative B cells (i.e., activated and/or memory B cells) all had the disease-linked X chromosome inactivated (5). Together, these observations indicated that T cells lacking a functional XSCID gene fail to develop, whereas B cells lacking this gene’s function will develop normally to the naive, mature B cell stage, but cannot undergo B cell activation and terminal differentiation.
Given this wealth of information, the identification of the XSCID gene as the IL-2Rγ-chain should have provided the final illumination regarding the underlying cause of this disease. Instead, it led to a new question: how is the IL-2R γ-chain involved in T cell development and B cell terminal differentiation?
In the years before 1993, the molecular characterization of the IL-2R had been well worked out. The structures of the three forms of the IL-2R, the high, medium, and low affinity receptors, had been shown to consist of different combinations of three receptor subunits, α, β, and γ, the genes for each of which had been cloned by that time (6). Both the high and medium affinity IL-2 receptors contained IL-2Rγ, clearly implicating defects in IL-2 signaling as the underlying cause of XSCID.
With that, the ‘XSCID paradox’ was born. Mapping of the XSCID disease gene to the IL-2R γ-chain locus was completely at odds with the prevailing wisdom at the time that IL-2 was not required for the development of T cells in the thymus nor for B cell activation and differentiation. Experiments performed in the early 1980s had indisputably demonstrated that the IL-2R was expressed on the earliest subsets of progenitor cells found in the thymus (7, 8) and to this day remains a characteristic hallmark of thymocytes committed to the T cell lineage. In addition, several early studies had demonstrated that IL-2R signaling was functional in developing T cells and might participate in T cell maturation (7, 9, 10). However, immunodeficiency patients with impaired IL-2 production had a milder form of disease than the XSCID patients, including the presence of circulating T cells (11, 12). Ultimately, the generation of IL-2-deficient mice in 1991 (13), one of the first knockout lines made to test the function of a gene involved in the immune system, provided ironclad evidence that IL-2 is not essential for T cell development—at least not in mice.
As an alternative to the hypothesis that humans and mice differ in their cytokine signaling requirements for T cell development, Noguchi et al. suggested that the IL-2R γ-chain may be a component of additional cytokine receptors beyond that for IL-2 (1). At the time, a precedent existed for shared cytokine receptor subunits, as gp130 was known to participate in multiple receptors in addition to IL-6 (14, 15), while the IL-3 receptor shared a β-chain with the receptors for IL-5 and GM-CSF (16, 17). So, the hunt was on for additional cytokine receptors that might also use the IL-2R γ-chain. A mere 8 mo later, the answer was in. Three back-to-back papers, published in Science in December of 1993 (18, 19, 20), reported that IL-2Rγ is shared with the receptors for IL-4 and IL-7, leading to a renaming of this protein as the ‘cytokine receptor common γ-chain, γc.’
Thus, the ‘bubble boy’ paradox was resolved. Although IL-2 is not required for T cell development in the thymus, IL-7 is (21, 22, 23). Further, the sharing of γc with the receptor for IL-4, along with IL-21, also accounted for the defects in B cell activation and differentiation (24). Also, just to eliminate any lingering doubt or concerns that perhaps mice and humans really are different in the signals regulating T cell development, a subset of SCID patients that completely lack T cells has since been shown to have deficiencies in expression of the IL-7R α-chain (25).
In addition to representing a tour de force of molecular immunology, the discovery of the genetic basis for XSCID presented a multitude of opportunities for diagnostic and therapeutic advances. With the knowledge that mutations in the IL-2Rγ gene lead to XSCID, which accounts for nearly 50% of all SCID cases (26), prenatal genetic screening and identification of carrier females became a possibility. Even more exciting was the potential for therapeutic intervention, as the single-gene, hematopoietic, cell-intrinsic nature of this defect provided a golden opportunity for the newly emerging field of gene therapy. In fact, in 1999 the first gene therapy clinical trial for XSCID was initiated with two patients, and the initial results of this study were reported in Science in 2000 (27). The good news was that the gene therapy was successful, leading to the treatment of additional cohorts of patients and representing the first example of a human disease cured by gene therapy; the bad news was that two of nine of the initial patients developed leukemia as a result of retroviral insertion into a proto-oncogene locus in their bone marrow stem cells (28). Although it is clear that the deleterious risks of gene therapy need to be eliminated, the overall success of the process for restoring IL-2R γ-chain expression in XSCID patients’ cells provides a strong impetus for the development of new gene therapy strategies. Of course, none of this would have occurred without the fundamental discovery by Warren Leonard and his colleagues that mutations in the IL-2R γ-chain cause the disease XSCID.
The author has no financial conflict of interest.
Abbreviation used in this paper: XSCID, X-linked SCID.