Case 4: Hyper-IgM Syndrome
Warren Strober

Case Report

"My baby has a high fever and a bad cough."

Clinical History and Initial Evaluation

Lee, a six-month-old infant, is brought to the pediatric emergency room (ER) by his mother at about 8 AM on a Saturday. She tells you that Lee was well until two days ago, when he developed a cough and seemed lethargic. At that time, she called her family physician. He advised her that it was probably a cold, to keep him well hydrated, and to administer one baby acetaminophen every 6 h if Lee develops a fever. Lee slept poorly that night and the next day was noted to have a low-grade fever of about 38°C. Since he seemed to be getting worse, the family physician was called again, but he continued to think Lee merely had a cold and should be treated symptomatically. Over the course of the day Lee's cough became more frequent, and by the evening he was breathing more rapidly than usual (tachypneic). These symptoms became more intense during the night and prompted the mother's visit to the ER with the baby.

Your physical examination of Lee discloses a child of appropriate length and weight (normal growth) who has a normal overall appearance, that is, no dysmorphic facial features or bone abnormalities. He has an elevated temperature of 38.5°C and obvious respiratory symptoms (coughing and rapid breathing) but no meningeal signs such as neck rigidity or inconsolable irritability. Auscultation of his chest reveals coarse breath sounds but no bronchospasm or evidence of pulmonary consolidation. Examination of the pharynx reveals no evidence of pharyngitis or tonsillitis (inflammation of the phyarynx and tonsils); in fact, the tonsils seem reduced in size.

You surmise that the patient has a pulmonary infection (probably a bacterial pneumonia) and have him admitted to the pediatric ICU (PICU). You also order a chest X ray, blood cultures, and a complete blood count (CBC).

Course in Hospital and Initial Diagnosis

The chest X-ray result is available shortly after admission to the PICU and reveals the presence of bilateral perihilar infiltration. A complete blood count shows an increased white blood cell (WBC) count (15,500 cells/µL, reference range < 12,000 cells/µL) comprised of 72% neutrophils, 7% monocytes, and 20% lymphocytes. Although this differential is compatible with the presence of an infection, you note that the total white count is lower than you expected for a bacterial pneumonia. In any case, you start the patient on IV cefotaxine, a broad-spectrum third-generation cephalosporine antibiotic to treat what you consider is the likely cause of his pneumonia, a bacterial pathogen. You also order a lung computer-aided tomography (CAT) scan to obtain an objective baseline for the extent of the infection.

Over the next several days you return frequently to the PICU to check on Lee's condition. Over this period Lee's respiratory status remains stable but does not improve despite the antibiotic therapy. He continues to be febrile and to be tachypneic. Pulse oxymeter readings show periodic episodes of oxygen desaturation (< 92%) that require treatment with an oxygen mask. The blood cultures have proved negative, so you know neither the identity of the organism nor its sensitivity to antibiotics. Thus, you cannot be sure his infection is being adequately treated and, in fact, his lack of more rapid clinical improvement suggests otherwise. On the morning of the fourth day after admission, Lee's status seems to have deteriorated; his breathing is more labored and oxygen desaturation episodes are more prolonged and frequent. Indeed, a repeat CAT scan shows more extensive lung involvement. After consultation with a pulmonologist, a bronchoscopy is performed to obtain bronchial lavage fluid and brushings for microscopic examination and culture. This is done without incident. The next morning the laboratory report is returned; to your surprise, Lee has an infection with Pneumocystis jiroveci (formerly called Pneumocystis carinii) a protozoal organism that ordinarily causes infection only in immunocompromised hosts. You immediately change the antibiotic to IV trimethoprim-sulfamethoxazole (Bactrim), a specific and effective therapy for this organism.

The results of this treatment are dramatic. Within 24 h Lee shows noticeable improvement. Over the next several days he defervesces and his pulmonary status returns to normal. After four days of IV Bactrim therapy you switch to oral therapy and two days later Lee is discharged to home with the proviso that he continue on oral Bactrim therapy for another two weeks.

• What is unusual about a pneumonia due to Pneumocystis jiroveci (carinii)?

  Show Answer

  Correct answer:
  As discussed in Case 1 on X-linked agammaglobulinemia (XLA), neonates are protected against infections, particularly bacterial infections, by maternal antibodies that have crossed the placenta (IgG) or been obtained through the mother's milk (IgM and IgA) until the infant's own immune system is more mature and the infant can mount his or her own antibody responses. Protection by placentally transmitted antibodies has almost completely dissipated by six months after birth; thus the occurrence of a pneumonia at this age, even a bacterial pneumonia, is unusual but not rare. It is the particular organism causing Lee's pneumonia, Pneumocystis jiroveci, which is unusual and which raises a red flag that an underlying immune problem may exist. Pneumocystis belongs to a category of organisms known as "opportunistic pathogens" because they do not cause disease in healthy individuals, but do so in those with a compromised immune system that allow them the opportunity to act as pathogens. Although infants with their incompletely mature immune systems may be partially immunodeficient relative to the general population, the occurrence of an infection due to an opportunistic pathogen in a young child should still alert one to the possibility that an underlying immunologic defect is present. In particular, it suggests the presence of an immunodeficiency disease characterized by an abnormal T-cell-mediated immunologic response, since T-cell immunity provides the main host defense against opportunistic pathogens. DiGeorge syndrome and the various forms of severe combined immunodeficiency disease (SCID) are among the immunodeficiency diseases characterized by defective T-cell-mediated immunity (see introduction to this unit). More commonly, however, an acquired immunodeficiency such as that associated with human immunodeficiency virus (HIV) infection accounts for the presence of an infection with an opportunistic organism.

Immunologic Workup and Diagnosis

Initial Approach

Recognizing that Lee may have an immunologic disorder, you refer him to an immunologist for further workup. The immunologist starts the workup with a second look at the facts already in hand.

History

First, the history of normal development and lack of disease during Lee's first six months of life speak against SCID, an immunodeficiency that generally declares itself early and is associated with a failure to thrive. This tentative conclusion is corroborated by the CBC, which did not show the presence of a lymphopenia (reduced lymphocyte count). It should be noted, however, that Lee was observed to have a reduced tonsillar mass, possibly indicating some loss of ability to populate peripheral lymphoid compartments.
Re-review of the chest X-ray and CAT scans confirms the presence of a thymic shadow, with a size appropriate for age. Combined with the lack of evidence of hypoparathyroidism, dysmorphic facial features, and cardiac anomalies, the presence of a thymic shadow on radiologic examination eliminates consideration of thymic aplasia (DiGeorge syndrome).
Another major consideration in the differential diagnosis is the presence of a neonatal HIV infection (HIV transmitted to Lee from his mother during birth). This is unlikely because an HIV diagnostic test performed on Lee's mother during childbirth (as required by your state's law) was negative. Nevertheless, to be absolutely certain, Lee is tested for the presence of HIV using a Western blot-based technique to detect HIV protein, in case he lacks the ability to mount an antibody response to this organism. The results of this test are likewise negative, and HIV infection is definitively ruled out.
The immunologist also elicits a family history from the mother. He learns that Lee has a maternal uncle (his mother's brother) who died mysteriously at an early age due to pneumonia. This suggests that Lee may have an X-linked disease.
The conclusions drawn from the above review and family history are very valuable but do not identify a specific immunodeficiency state. A more extensive workup is therefore initiated. Although the Pneumocystis infection is drawing attention to the T cells or cell-mediated arm of the immune response, the status of his humoral immunity is also investigated.

Testing and Interpretations

As described in Cases 1 on XLA and the introduction to this unit, an orderly approach to the investigation of the capability of the immune system must be pursued, keeping in mind the stages of development of mature immune cells and the interactions that must take place among them to mount an effective immune response.

Lymphocyte Subset Analysis

Quantitation of peripheral lymphoid subsets by an initial flow cytometry study reveal normal percentages and absolute numbers of CD4⁺ T cells, CD8⁺ T cells, B cells, and natural killer (NK) cells. Additional studies are performed in a specialty laboratory that does a more extensive panel of CD markers. Analysis of the CD4⁺ cell subsets reveals that the percentage of CD4⁺ cells in the CD4⁺/CD45RA⁺ (naive) CD4 T-cell subset is high and, correspondingly, the percentage in the CD4⁺/CD45RO⁺ (mature) CD4 T-cell subset is low even if one considers the fact that the ratio of CD45RA⁺ cells to CD45RO+ cells is high in young children as compared to adults. In addition, while Lee has a normal number of B cells, virtually none bear surface IgG or IgA.

These findings suggest a defect in T- and B-cell maturation at a point governed by antigen exposure.

Immunoglobulin Levels

• How do the results obtained thus far differ from what would be found in studies of patients with XLA, SCID, and common variable immunodeficiency disease (CVID)?

  Show Answer

  Correct answer:
  In XLA, a primary B-cell defect leads to absence of circulating B cells and lack of production of all Ig classes by circulating cell populations. In SCID syndromes, a primary defect affecting T cells is usually present that reduces the capacity of T cells to elaborate the helper cytokines necessary for B-cell maturation and Ig production; in addition, this primary defect may also have a direct effect on B cells. The result is that the number of T cells in the circulation is usually reduced as are B cells and serum immunoglobulin levels. Finally, in CVID, B-cell function is also usually impaired by a primary B-cell defect, and production of all immunoglobulin classes, including IgM, is correspondingly decreased, although not to the level seen in either XLA or SCID. In CVID the levels of both T and B cells in the circulation are normal or only mildly decreased. Lee's immunoglobulin profile differs from that of all of these diseases in that his IgM level is normal or even elevated, whereas the other immunoglobulins are decreased. Lee has normal numbers and distribution of lymphocyte subsets, unlike patients with XLA, who exhibit selective absence of B cells, and most patients with SCID, who exhibit absence of T cells associated, in many cases, with absence of B cells as well. Finally, while some SCID patients do have normal numbers of circulating lymphocytes such as those with a T+B+ phenotype, Lee's capacity to produce IgM makes it unlikely that he fits into this category of immunodeficiency disease.

Proliferation Studies

Culture of Lee's cells with nonspecific mitogens, which induce proliferation of T cells (phytohemmaglutinin and concanavalin A) and B cells (pokeweed mitogen) show normal proliferative ability (see Case 1).

In Vitro Immunoglobulin Production

The ability of each cell type to function in the steps required to produce immunoglobulin is next tested.

The B Cells on Their Own

Lee's peripheral blood (circulating) B cells are isolated and cultured in vitro with a CD40L trimer (plus IL-4 and IL-10) to produce antibody in conditions circumventing T-cell dependence. CD40L (CD154) is a molecule appearing on activated T cells that interacts with CD40 on B cells and induces B-cell maturation and class switching; thus, soluble CD40L trimer mimics or replaces the stimulatory activity of T cells. The culture fluid is then assayed for the presence of secreted immunoglobulins.

Result
Lee's B cells, stimulated under these culture conditions, produced normal amounts of IgM as well as normal amounts of IgG and IgA.

This provided solid evidence that his B-cell responses were intrinsically normal and that the serum Ig levels did not reflect an intrinsic B-cell abnormality.

T-B Cell Interaction

In concomitant studies, Lee's T cells are cultured with control B cells and his B cells are cultured with control T cells under conditions to induce polyclonal T-cell stimulation (surface-bound anti-CD3 and soluble anti-CD28). T-cell stimulation leads to the expression of CD40L and cytokine production, and the activated T cell should be capable of inducing B-cell stimulation and isotype switching to IgG and IgA production.

Result
Lee's T cells, stimulated by anti-CD3/anti-CD28, fail to induce normal B cells to produce IgG or IgA, whereas control T cells demonstrate this capacity; on the other hand, control T cells, stimulated by anti-CD3/anti-CD28, induce Lee's B cells to produce IgG and IgA.

These results suggest that Lee's T cells have reduced "helper T-cell activity", that is he has basic T-cell defects, whereas his B cells are normal. Thus his T cells are unable to interact with B cells to induce isotype class switching. This may be the result of a defect in CD40L expression and/or function or a defect in cytokine production.

Further Diagnostic Studies

Cell-Mediated Immunity Studies

Lee's clinical presentation (infection with an opportunistic pathogen) and immunologic workup thus far suggest an underlying T-cell abnormality in spite of the fact that Lee appears to have a B-cell deficiency as manifested by low IgG and IgA levels. The immunologist places particular importance on the T-B-cell interaction studies described above which, given that Lee's B cells are capable of producing immunoglobulin, point to a T-cell defect rather than a B-cell defect. One type of immunodeficiency that fits this pattern is hyper-IgM syndrome, an immunodeficiency that may be caused by failure to express CD40L (CD154) on the surface of activated T cells.
CD40L-CD40 also mediates interactions between T cells and antigen-presenting cells (APCs; monocytes or dendritic cells) to produce cytokines. The absence of effective interaction between T cells and macrophages would be expected to manifest in defective cell-mediated immune responses. To investigate this possibility, the immunologist conducts further T-cell studies that evaluate both in vivo and in vitro T-cell function and T-cell/APC interactions.

Delayed-Type Hypersensitivity Test

These "in vivo" studies consist of intradermal skin tests with tetanus toxoid and diphtheria antigens. Both are antigens to which Lee has already been exposed as part of his normal vaccination schedule. The skin test sites are "read" at 48 h for the presence of induration and erythema to determine if Lee can mount a delayed-type hypersensitivity (DTH) response.

Result

The test proves negative in that neither of these antigens elicits a significant inflammatory response.

Since the DTH reaction depends on T-cell stimulation of APCs via CD40L to produce cytokines (normally resulting in induration and erythrema), the absence of response in these studies is compatible with a CD40L defect. This test of in vivo T-cell immunity also predicts that Lee will unfortunately have further difficulties with opportunistic infections.

In Vitro Activation and Monocyte Cytokine Production: An Investigation of Cell Communication

Lee's peripheral blood mononuclear cells are cultured either with anti-CD3 (an antibody that acts as a T-cell mitogen as noted above) or Staphylococcus aureus Cowan strain 1 (SAC) plus interferon_-γ (IFN_-γ). The latter combination acts as a direct stimulant of APCs, which, in peripheral blood, consists mainly of monocytes. This time the culture fluid is assayed for the cellular production of secreted cytokines (see Unit V). T cells, when activated by the anti-CD3, normally express CD40L (CD154), which then interacts with CD40 on the monocyte membrane, inducing the monocytes to produce inflammatory cytokines such as interleukin-12 (IL-12) and tumor necrosis factor (TNF)_-α. On the other hand, SAC plus IFN_-γ stimulates monocytes directly and results in production of IL-12 and TNF_-α in the absence of T cells.

Result
Lee's peripheral monocytes produce greatly reduced amounts of IL-12 and TNF-α when they are stimulated by his T cells (cultured with anti-CD3). In contrast, his monocytes produce normal amounts of these cytokines when they are directly stimulated with SAC plus IFN_-γ.

Lee's - cells do not stimulate monocytes to produce cytokines, but his monocytes are intrinsically intact. This is likely the result of defective CD40L expression on the T-cells. However, a defect in IFN_-γ production by the T-cells could also be the culprit; therefore that possibility is tested next.

FN_-γ Production

Lee's mononuclear cells are cultured with anti-CD3, but this time CD40L stimulation of the monocytes is "artificially" provided by addition of CD40L trimer, the soluble form of CD40L that can stimulate monocytes via CD40 in the absence of T cells. The CD40L trimer was used for a similar purpose in B-cell cultures (see above). T cells in cultures with anti-CD3 and activated monocytes should produce a cytokine such as IFN_-γ as a result of stimulation by the anti-CD3 and the helper cytokines produced by the CD40L-stimulated monocytes.

Result
Whereas Lee's T cells produced greatly decreased amounts of IFN_-γ in the absence of CD40L trimer, they produced normal amounts in its presence. The control T cells, on the other hand, produced IFN_-γ in the presence and absence of CD40L trimer.

These data thus offer excellent functional evidence that Lee's T cells are functionally deficient either because they do not express CD40L or because they express an inactive CD40L, rather being unable to produce IFN_-γ.

Definitive Diagnosis and Classification

Detection of the Molecule

At this point the immunologist closes in on the diagnosis of X-linked hyper-IgM syndrome (XHIGM; HIGM-I) by performing flow cytometry studies using fluorochrome-labeled anti-CD40L. These studies indeed demonstrate that Lee's cells do not express CD40L. The immunologist also sends a sample of Lee's DNA to a sequencing facility for sequencing of the CD40L gene. The results show that Lee's CD40L gene has a mutation in the extracellular domain of the gene that results in a truncation of this domain and thus low expression of CD40L on the T-cell surface. The diagnosis of XHIGM is now secure. Given the family history of a maternal uncle with an illness that could have been caused by HIGM, Lee is assumed to have inherited this mutated X-linked gene rather than to have a new mutation. This was subsequently verified by sequencing of Lee's mother's DNA, which showed that the mother carries the mutated gene on one X chromosome.

• What is the explanation for the initial observations that Lee's tonsils are underdeveloped and that his peripheral helper T cells are skewed toward the naive (CD45RA) subset rather than the memory (CD45RO) subset?

  Show Answer

  Correct answer:
  Germinal center formation is dependent on interactions between T-cells and dendritic cells and T-cells and B cells. Both interactions occur in a CD40L-CD40-dependent fashion. The underdevelopment of germinal centers and predominance of only small primary follicles in the tonsils and even in the lymph nodes result in the reduced size of these structures. The change in the T-cell subset distribution is similarly the result of the inability of the T-cells to express CD40L. Without CD40L, they are unable to stimulate the dendritic cells to express B7, which in turn is required to induce the T-cell to mature. Thus Lee's peripheral T-cells are skewed to the naive T-cell subset (CD45RA⁺), with corresponding decrease of the memory T-cell compartment (CD45RO⁺).

Lee is placed on monthly intravenous immunoglobulin (IVIG) infusions of 20 mg/kg to bring his "trough" (low-point) IgG level above 500 mg/dL, the target IgG level for immunodeficiency patients who cannot produce normal amounts of IgG. In addition, he is placed on prophylactic Bactrim therapy to prevent recurrence of Pneumocystis infection.

Lee's Subsequent Medical Course

At the age of three years, Lee developed a neutropenia (2000 neutrophils/μL versus the reference level of 5000 neutrophils/μL), a known concomitant of XHIGM syndrome. At age six years, he was placed on granulocyte colony-stimulating factor (G-CSF) therapy and partial amelioration of the neutropenia was achieved. This neutropenia was associated with recurrent apthous mouth ulcers, which tended to improve when neutrophil levels increased. Extensive attempts to culture an organism from these ulcers were unsuccessful.
At age eight, Lee suddenly developed mild right-sided weakness. Unfortunately, this symptom was not investigated until three months later, when he had a grand mal seizure. At that point a magnetic resonance imaging (MRI) examination of the brain revealed multiple punched-out lesions with enhancement. This picture is typical of toxoplasmosis, a known complication of XHIGM. Lee was therefore put on long-term pyrimethamine-sulfadiazine therapy that had an immediate effect as evidenced by the decrease in the enhancement in the MRI and had the long-term effect of preventing progression of the toxoplasmosis.

Clinical Features of Hyper-IgM Syndrome

As will be detailed in a later section, the clinical syndrome of hyper-IgM has several possible pathophysiologic bases. Here we will discuss the clinical features of XHIGM syndrome. Although the name "hyper-IgM" places the emphasis on the inability to isotype switch to IgG and IgA immunoglobulin production, the infectious complications of XHIGM remind one that this is a deficiency in an interaction (CD40-CD40L) important in both humoral and cell-mediated immune responses. Thus there is overlap in its clinical features with those of some forms of SCID and T-cell immunodeficiency in that these patients have infections not normally found in conditions such as CVID (see below), a disorder with usually intact cell-mediated immunity.

Pulmonary Problems
Since patients with all forms of HIGM syndrome have increased or normal amounts of IgM levels but decreased IgG and IgA levels, it is not surprising that they are susceptible to infection with extracellular pathogens, similar to patients with CVID. Indeed, upper and lower respiratory infections by extracellular pathogens are the most common problems of hyper-IgM patients. In the absence of IVIG replacement therapy, bronchiectasis is likely to result as a consequence of recurrences. However, respiratory infection in X-linked hyper-IgM syndrome is also not infrequently due to opportunistic organisms (unlike CVID), and in over 40% of patients in whom pneumonia marked the onset of disease in the first year of life, Pneumocystis jiroveci was the cause of infection. Other opportunistic organisms causing infection in XHIGM patients include cytomegalovirus (CMV), cryptococcus, and atypical mycobacterial infection. These are the consequence of the inadequate interaction between T cells and macrophages via CD40-CD40L.

Gastrointestinal Problems
XHIGM patients frequently have diarrhea, and in 20-30% of cases, this can have a chronic course, associated with malabsorption and failure to thrive (again similar to what occurs in CVID). At times, the diarrhea results from opportunistic infection, such as with cryptosporidia or giardia, but in other instances it appears to be noninfectious in nature. In these latter cases, the patients may have intestinal nodular lymphoid hyperplasia and villous atrophy, suggesting the presence of an autoimmune process involving the GI tract.

Liver Problems
Severe liver problems occur in almost 20% of XHIGM patients. These can take the form of sclerosing cholangitis (frequently due to cryptosporidiosis) as well as infection with hepatitis B and hepatitis C or even CMV. The hepatic infections can lead to cirrhosis and liver decompensation, which may necessitate a liver transplant. Historically, hepatitis in hyper-IgM syndrome has been associated with exposure to contaminated blood products, thus, IVIG used for replacement therapy must come from an unimpeachable source. Another complication of chronic hepatitis in XHIGM patients is hepatocellular carcinoma, which usually has a fatal outcome.

Hematologic Manifestations
The main hematologic problems in XHIGM patients are anemia and neutropenia. The former occurs in about one-third of patients and is usually due to iron deficiency; however, the anemia can occasionally be due to bone marrow failure (aplastic anemia) or peripheral lysis of the red cells (hemolytic anemia). Neutropenia, the more common hematologic abnormality, occurs in almost 70% of patients. It takes the form of a block in myeloid cell development, occurring at the promyelocytic-myelocytic stage. In some cases, the neutropenia is chronic and in other cases it is cyclic. It is frequently associated with gingivitis and perioral ulceration.

Neoplasia in XHIGM patients
Visceral neoplasms involving the gall bladder, liver, and pancreas have been reported, but this is not a frequent phenomenon. Lymphomas, often increased in other immunodeficiency states, are not associated with XHIGM, possibly because patients do not live long enough to develop this complication.

Miscellaneous Problems
A degenerative encephalopathy is not infrequently seen in XHIGM patients and is usually due to infection with any of several opportunistic pathogens, including Cryptococcus, toxoplasmosis, CMV, Echo virus, and Mycobacterium bovis. Because some of these infections are treatable, vigorous search for a causative agent should be done at the onset of CNS symptoms. Once degenerative changes occur, they can rarely be completely reversed.
Finally, secondary arthritis occurs in a small group of patients. This is usually associated with neutropenia and is presumably due to infection, but in most cases a causative agent cannot be identified.

Normal B-Cell Stimulation and Function

The maturation of B cells occurs in two phases. The first is an antigen-independent phase occurring in the fetal liver and bone marrow and the second is an antigen-dependent phase that occurs in lymphoid peripheral tissues such as the Peyer's patches, the spleen, and the lymph nodes. The steps necessary to advance the B cell from the progenitor stage to a naive mature B cell leaving the bone marrow were reviewed in Case 1 on XLA. In that first phase, the "germline" B-cell repertoire is established in individual B cells by rearrangement of both the heavy- and light-chain immunoglobulin genes. These rearrangements result in the formation of a VDJ segment from the variable segments, diversity segments, and joining segments (V, D, and J segments, respectively) as a consequence of a random recombination process. At the conclusion of these molecular events, the Ig heavy-chain gene consists of a "recombined" VDJ segment that determines antigen specificity. This segment sits upstream of a string of heavy-chain constant-region segments, the first of which are the Cμ and Cδ segments. The cell is now able to produce IgM and IgD antibodies.

In the second (antigen-dependent) stage of development, which occurs in the periphery, B cells undergo proliferation. At the same time, their Ig genes may undergo further changes, that of somatic hypermutation in the variable region (resulting in affinity maturation) and that of isotype class switch rearrangement. Somatic hypermutation is the result of point mutations in the VDJ segments, which lead to subtle changes in the affinity of the immunoglobulin for antigen. This process allows for the generation of cells producing high-affinity antibodies. Isotype class switch rearrangement results in a substitution of the constant-region portion of the heavy-chain molecule. Thus, the VDJ becomes associated with various downstream "C-region" heavy-chain genes such as γ, α, or ε C regions. In this way cells producing the various classes and subclasses of immunoglobulins are formed and combined with the hypermutated variable portion; these antibodies thus produced have a generally higher affinity for antigen than IgM.

In patients with hyper-IgM syndrome, the first phase of B-cell maturation, the antigen-independent phase in the bone marrow, is intact. It is in the second, antigen-dependent stage in the periphery that progression is halted (Fig. 4.1).

Figure 4.1 Sites of defective lymphopoietic development associated with primary immunodeficiency syndromes. Developing and mature lymphoid populations remaining in patients with hyper-IgM syndrome. Gray cells denote absent populations. Bar denotes block in interaction.

The Structural Basis of Class Switching

• What is the molecular basis of the process by which B cells switch from the production of IgM to IgG or IgA; in other words how do they undergo Ig class switch differentiation?

  Show Answer

  Correct answer:
  A switch region (S) sits in the intron 5' of each heavy-chain constant region with the exception of δ which has a 5' pseudo-switch region (Fig. 4.2). These switch regions are repetitive, conserved base sequences of 1-10 kilobases. When the cell receives the signal to switch (see below), two switch regions line up with the intervening DNA looping out. Recombination then takes place between these repetitive (and somewhat homologous) sequences and the DNA between them is cleaved. "Fingerprints" of this event are evident from the small loops of this cleaved DNA, which can be transiently detected before degradation. Switching may occur in an adjacent constant region or a constant region further downstream or sequential switching events may occur, with the cell progressing first to IgG and then to IgA. Isotype switching requires transcription and is initiated by promoters. Transcription is presumed to be required so that the DNA of the switch region can be opened up and accessed by the enzymes needed to accomplish the feat. Activation-induced cytidine deaminase (AID), an enzyme expressed exclusively in B cells and integral to the mechanism of somatic hypermutation, is also involved in isotype switching. It functions in both of these processes by deaminating cytidine directly on DNA. AID is a member of a class of molecules capable of "RNA editing"; it may be essential to class switching because it is necessary to edit the RNA of a key switch molecule. Obviously the breaking and ligation of DNA involve a host of coordinated enzyme activities and enzymes, including an endonuclease necessary for genetic DNA breaks and DNA repair enzymes necessary to anneal such breaks. Interestingly, as opposed to the V-region genes, in which rearrangements not infrequently cause disruption of the gene and are therefore nonproductive, all isotype switch recombinations are successful, resulting in a productive immunoglobulin gene. This preserves the ongoing and effective immune response.

Figure 4.2 Mechanism of class switching in Ig synthesis; S = switch region, upstream of each heavy-chain constant region. Switching requires T cell interaction.

• What signals the B cell to undergo isotype switching?

  Show Answer

  Correct answer:
  The T cells are the prime players providing the signals for isotype switching to the B cells. CD40 engagement on the surface of the B cell and signaling from T-cell-derived cytokines are required. The helper T cell, in response to its specific antigen and to costimulatory molecules presented by the B cell, up-regulates CD40L (CD154) on its surface and secretes cytokines (see the pathogenesis section below). Additional adhesion is achieved through LFA-1-ICAM-1, ICOS-B7, and CD30-CD30L interactions, the first molecule listed in the pair being on the T cells and the second on the B cells. This cognate interaction between the T and B cells ensures the specificity of the immune interaction by being antigen driven and by focusing the signals to the area of contact between the two cells. It is the specific cytokines received by the B cell which govern the heavy-chain isotype that is produced and these cytokines differ depending on the T helper subset responding. Thus, in mice, viruses and many bacteria will activate T_H1 cells to secrete IFN_-γ, stimulating isotype switching to IgG_2a and IgG_2b subclasses; helminths will stimulate TH2 cells to produce IL-4, the main cytokine resulting in IgE, as well as IgG1 production. T_H2 cells also secrete transforming growth factor (TGF) β (resulting in IgG_2b and IgA switching) and IL-5 (causing IgA secretion). Thus the T cell has a central role in controlling the character of the antibody response.

The "Purpose" of the Different Immunoglobulin Constant Regions

Whereas the variable region governs the antigen-binding properties of the immunoglobulin molecule, the constant region determines its biologic properties. This region is required for recruiting other cells via their Fc receptors, binding complement, and specifically transporting the immunoglobulin molecule via active transport mechanisms to other sites. Macrophages and neutrophils, the body's phagocytic cells, have Fc receptors for IgG₁ and IgG₃, whereas high-affinity FcRε are found on mast cells, basophils, and eosinophils, the cells containing mediators needed for response to parasitic infections. Complement, which may directly destroy pathogens or activate phagocytes, is bound most efficiently by IgM, IgG₃, and IgG₁. IgG₁ and IgG₃ cross the placenta and make their way into the fetal circulation. IgA is specifically and actively transported across the epithelial cell into mucous secretions, tears, and milk (Table 4.1).

Table 4.1 Most Important Features of Immunoglobulin Isotopes

^a. Natural killer.

Pathogenesis of Hyper-IgM Syndrome

As its name implies, HIGM syndrome is defined by the fact that patients have B cells that produce IgM but not IgG, IgA, or IgE. Thus, they have a defect in class switch recombination. As is evident from the discussion above, class switch recombination is a multistep process and thus the HIGM syndrome could result from molecular defects affecting any of these steps. As shown in Figure 4.3, five distinct defects leading to the HIGM syndrome have so far been identified. Three of these defects (designated HIGM-I, III, and IV) affect the function of B cells, T cells, and APCs; two of these defects, HIGM-II and V, affect the function of B cells only. The most common form is X-linked HIGM, described in our patient, Lee.

Figure 4.3 Defects in hyper-IgM syndromes. Some defects interfer with T-macrophage interaction as well as T-B interaction. Others interfere with mechanisms of class switching within the B cells.

Molecular Basis of XHIGM, HIGM-I Syndrome
An important clue to the molecular origin of the most prevalent form of hyper-IgM syndrome came from the discovery that class switch rearrangement is an antigen-driven process that occurs in germinal centers of lymphoid follicles. Thus, when it was discovered that interaction between CD40 on APCs and CD40L on T cells was necessary for germinal center formation, investigators considered the possibility that this interaction was important for class switch rearrangement and by extension was somehow involved in hyper-IgM syndrome. This suspicion was heightened by the fact that most patients with HIGM syndrome have lymphoid organs characterized by poorly developed germinal centers. It was ultimately shown that an X-linked form of hyper-IgM syndrome (XHIGM or HIGM-1), representing about 50% of all cases, is due to a mutation of the CD40L gene, resulting in nonexpression of CD40L or expression of a defective CD40L that cannot exert signaling function. Figure 4.4 shows a diagram of the CD40L molecule and the approximate locations in the molecule of mutations associated with XHIGM. Most of the mutations occur in the gene segment encoding the extracellular domain of the molecule and result in stop codons or base substitutions that cause loss of expression of CD40L on the cell surface or a CD40L molecule that cannot signal via CD40.

Figure 4.4 Structure of CD40L (CD154) gene. Arrows denote positions of mutations.

The CD40L molecule is a member of the TNF_-α receptor family of molecules; its interaction with CD40 results in activation of important signaling pathways that affect the function of the CD40-expressing cells, B cells, and APCs, particularly dendritic cells. As noted above, CD40 signaling in the B cell is necessary for B-cell proliferation, germinal center formation, and class switch rearrangement. In dendritic cells, CD40 signaling is important in the production of key cytokines such as IL-12 and in the expression of costimulatory molecules such as CD80 and CD86. Thus, as shown in Figure 4.5, the consequences of a mutation of CD40L include not only the functional abnormalities in B cells that are the hallmark of the disease, but also T-cell abnormalities due to defective interactions with APCs. The latter is illustrated in Figure 4.5C, which shows the sequential activation of T cells leading to expression of CD40L and the signaling of dendritic cells to produce IL-12. Both the B-cell and APC "defects" are actually secondary to the absence of signal from the real culprit, the T cell that lacks functional CD40L. The main immunologic consequence of the inadequate APC-T cell interaction is decreased IL-12 production, leading in turn to decreased production of IFN_-γ by T cells. This explains why hyper-IgM patients with CD40L mutations (XHIGM patients) have increased susceptibility to infection with opportunistic organisms. With the advent of Ig replacement therapy, this defect in XHIGM patients is now the most significant clinical abnormality remaining and is one that explains their clinical course.

Figure 4.5 Sequence of events in T cell-APC interaction through CD40-CD40L following antigen stimulation.

Interestingly, XHIGM patients with mutations in CD40L have IgM⁺/IgD⁺ B cells (unswitched B cells) that express CD27 (a maturation marker) and that exhibit evidence of some somatic hypermutation. These cells may have been generated in non-germinal center areas (e.g., marginal zones) as a result of other interactions, such as those involving OX40/OX40L, another TNF receptor family member. They can probably react with thymic independent (polysaccharide) antigens and thus provide patients some protection against disease. Along these lines, it also has been noted that a substantial fraction of patients produce IgA but not IgG. This may arise from class switching in certain types of B cells that does not require CD40L.

Other Forms of Hyper-IgM Syndrome
In recent years several other forms of hyper-IgM syndrome have been described; they are not due to a CD40L mutation, but like XHIGM they present as a defect in class switch recombination. As discussed below, some of these syndromes are characterized by considerably different clinical patterns than that found in XHIGM.

Hyper-IgM Syndrome due to CD40 Mutations: HIGM-III

Several patients with hyper-IgM syndrome have been reported to have an autosomal recessive form of the disease in which various mutations in the CD40 gene lead to failure to express CD40 on the cell surface. Inasmuch as this abnormality leads to the same immunologic defects as seen in XHIGM (absence of CD40-CD40L interaction), it is not surprising that the clinical pattern found in these patients closely resembles that of XHIGM. This form of hyper-IgM syndrome, which has been found in females (and can potentially occur in males as well), is classified as HIGM-III.

Hyper-IgM Syndrome Due to NEMO Mutations: HIGM-IV

Another form of HIGM syndrome is X linked but has a different molecular defect than XHIGM (HIGM-I). This is known as XHIGM-ED or HIGM-IV. It occurs in association with hypohydrotic ectodermal dysplasia (ED) and has been shown to be due to mutation in nuclear factor _κB (NF_-κB) essential modulator (NEMO). NF_-κB is a master transcription factor controlling expression of numerous gene programs and functioning in different cell types. NEMO is a docking molecule for I_κκα and I_κκβ, two phosphorylases involved in the phosphorylation and degradation of I_κBa and thus in the activation of NF_-κB components; it is also involved in the translocation of these components to the nucleus. As shown in Figure 4.6, interaction of CD40L with CD40 leads to the activation of second messengers that interact with NEMO and initiate the activation and nuclear translocation of NF_-κB components. Thus, a mutation in NEMO that results in impairment of its function leads to a defect in CD40L signaling and another form of the HIGM syndrome. It should be noted, however, that intact NEMO function is also necessary for the activation of NF_-κB by numerous other cell stimulants, such as Toll-like receptors; this explains the fact that cells of patients with this form of HIGM display abnormal responses to these stimulants as well as CD40L.

Figure 4.6 CD40L signaling at the cell surface leads to activation of NF_-κB with subsequent translocation of its components to the nucleus. The inital step of NF_-κB activation involves the interaction of TRAF with NEMO.

The clinical features of this form of hyper-IgM syndrome are thus similar to those with XHIGM due to CD40L deficiency. However, given the ubiquitous nature of NF_-κB and its interacting molecules, it is not unexpected that these patients have additional abnormalities. XHIGM-ED is associated with ectodermal dysplasia because the development and function of ectodermal tissue depend on another TNF family receptor that also depends on NEMO for signaling (the DL or ED receptor). The ectodermal abnormalities include inability to sweat (hypohydrosis), teeth malformations (hypodontia), and hairlessness. Susceptibility to infection is confined to high-grade extracellular pathogens rather than including opportunistic pathogens as well but is aggravated by the fact that the patients have thick secretions that impede elimination of organisms from the airways.

Hyper-IgM Due to Activation-Induced Cytidine Deaminase (AID) Mutations: HIGM-II
This fourth form of hyper-IgM syndrome, HIGM-II, which is inherited in an autosomal recessive manner, is caused by a mutation in the gene encoding AID, a molecule essential for both somatic hypermutation and class switching in B cells. As in the case of XHIGM-ED, these patients have normal CD40L and CD40 expression. Because somatic hypermutation is severely impaired in the absence of AID, the B cells of patients with HIGM-AID elaborate only low-affinity IgM antibodies.
Patients with this form of hyper-IgM syndrome differ from patients with other hyper-IgM syndromes in that they have a defect exclusive to their B cells. The CD40-CD40L interactions between T cells and macrophages and T cells and dendritic cells are intact; therefore these patients are not subject to opportunistic infection or other problems related to cell-mediated immunity. In this form of the syndrome, patients do manifest lymphoid hyperplasia and various autoimmune manifestations.

Hyper-IgM Syndrome Due to Uracil-DNA Glycosylase Mutations: HIGM-V
Yet another form of hyper-IgM syndrome has been identified and this one (HIGM-V) is due to mutations in a gene known as uracil-DNA glucosylase (UNG) This enzyme is expressed in activated B cells and operates downstream of AID during the class switch recombination: AID deaminates cytosine to form uracil and then the uracil is removed by UNG. The clinical picture in this hyper-IgM syndrome is similar to that seen due to AID deficiency.

Treatment of Hyper-IgM Syndrome

Replacement of IgG by IVIG administration is the main thrust of therapy and, indeed, has resulted in a marked decrease in infection in these patients. In some (but certainly not all) cases, this leads to normalization of elevated IgM levels and even correction of the neutropenia. Adjunctive therapy consists of treatment of the neutropenia with G-CSF, and prophylactic Bactrim therapy for the prevention of Pneumocystis pneumonia in XHIGM patients. Despite these therapies, many patients with XHIGM continue to experience opportunistic infections. Because of this, survival of these patients beyond the second decade has been less than 20%.
On a more fundamental level, several patients have been treated successfully with bone marrow transplantation, usually from a human leukocyte antigen (HLA)-matched sibling. In view of the poor prognosis of XHIGM patients, this therapy should probably be considered more frequently, especially when a matched donor is available.

References

• DiSanto JP, Bonnefoy JY, Gauchat JF, Fischer A, de Sain Basile G (1993): CD40 ligand mutations in x-linked immunodeficiency with hyper-IgM. Nature 11: 41-43.
• Ferrari S, Gillani S, Insalaco A, Al-Ghonalum A, Soresina AR, Loubser M, Avanzini MA, Marconi M, Badolato R, Ugazio AG, Levy Y, Catalan N, Durandy A, Tbakhi A, Notarangelo LD, Plebani A (2001): Mutations of CD40 gene cause an autosomal recessive form of immunodeficiency with hyper IgM. Proc Natl Acad Sci USA 98: 12614-12619.
• Imai K, Slupphaug G, Lee WI, Revy P, Nonoyama S, Catalan N, Yel L, Forvelle M, Kavil B, Krokan ME, Ochs HD, Fischer A, Durandy A (2003): Human uracil-DNA glycosylase deficiency associated with profoundly impaired immunoglobulin class-switch recombination. Nature Immunol 4: 945-946.
• Jain A, Atkinson TP, Lipsky PE, Slater JE, Nelson DL, Strober W (1999): Defects of T-cell effector function and post-thymic maturation in X-linked hyper-IgM syndrome. J Clin Invest 103: 1151-1158.
• Jain A, Ma CA, Liu S, Brown M, Cohen J, Strober W (2000): Specific missense mutations in NEMO result in hyper-IgM syndrome with hypohydrotic ectodermal dysplasia. Nature Immunol 2: 223-228.
• Muramatsu M, Kinosita K, Fagarasan S, Yamada S, Shinkai Y, Honjo T (2000): Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell 102: 553-563.
• Prasad ML, Velickovic M, Weston SA, Benson EM (2005): Mutational screening of the CD40 ligand (CD40L) gene in patients with X linked hyper-IgM syndrome (XHMI) and determination of carrier status in female relatives. J Clin Pathol 58: 90-92.
• Quartier P, Bustamamente J, Sanai O, Plebani A, Debre M, Deville A, Litzman J, Levy J, Fermand JP, Lane P, Horneff G, Aksu G, Yakin I, Davies G, Texcan I, Ersoy F, Catalan N, Imai K, Fischer A, Durandy A (2004): Clinical immunologic and genetic analysis of 29 patients with autosomal recessive hyper-IgM syndrome due to activation-induced cytidine deaminase deficiency. Clin Immunol 110: 22-29.
• Stavnezer J (1996): Immunoglobulin class switching. Curr Opin Immunol 8: 199-205.
• Tomizawa D, Imai K, Ito S, Kajiwara M, Minegishi Y, Nagasawa M, Morio T, Nonoyama S, Mizutani S (2004): Allogeneic hematopoietic stem cell transplantation for seven children with X-linked hyper-IgM syndrome: A single center experience. Am J Hematol 76: 33-39.
• Winkelstein JA, Marino MC, Ochs H, Fuleihan R, Scholl Pr, Geha R, Stiehm ER, Conley ME (2003): The X-linked hyper-IgM syndrome: Clinical and immunologic features of 79 patients. Medicine 82: 373-384.
• Zhu Y, Nonoyama S, Morio T, Muramatsu M, Honjo T, Mizutani S (2003): Type two hyper-IgM syndrome caused by mutation in activation-induced cytidine deaminase. J Med Dent Sci 50: 41-46.