Reduced thymic output, increased spontaneous apoptosis and oligoclonal B cells in polyethylene glycol-adenosine deaminase-treated patients

Patient characterization

Samples were obtained from five patients with ADA-SCID (mean age 7.9±3.6 years; range 5.5–9 years) as part of the clinical protocol approved by the Institutional Review Board and after informed consent as well as from age-matched healthy controls admitted to our institution for traumas. The patients, three boys (patients #1, #2, and #4) and two girls (patients #3 and #5) were diagnosed at a median age of 4 months (range 2–5 months). At the time of this study, they had been under treatment with PEG-ADA for 6.7±0.9 years (range 5–8 years). In four out five patients, both ADA alleles contained mutations that severely reduced the expression of ADA activity: three unrelated patients (#1, #4, and #5), coming from different gypsy communities, were homozygous for the same mutation (R211H). A fourth patient (#2) showed compound heterozygosity for two severe mutations (H15D and deletion 1091–1092). At the time of the diagnosis, all children showed low total, B, and T cell counts, an impaired phytohemagglutinin (PHA) response, and levels of ADA activity and total deoxyadenosine nucleotide (dAXP) concentration in red blood cells compatible with the diagnosis of ADA-SCID.

After informed consent PEG-ADA treatment was started at a mean age of 1.1±0.4 years at a dosage of 60 U/Kg twice a week. The therapy resulted in a marked increase in plasma ADA activity to 132.6±24.1 μmol/h/mL (normal range 0.05–0.5 μmol/h/mL) and a fall in red cell dAXP concentration to almost undetectable levels. When an improvement in immunological function was achieved, PEG-ADA was reduced to a maintenance dose of 30 U/Kg twice a week. On this dose, we never observed pre-injection plasma ADA activity to fall below 10–12 µmol/hr/mL, and red cell dAXP remained very low. Antibodies to PEG-ADA were detected by ELISA in one patient, but his plasma ADA activity was stable.

Total lymphocyte, CD3⁺, and CD19⁺ cell counts increased within 5–14 months after initiation of PEG-ADA treatment, but they remained low in comparison to values in healthy controls for the entire period of follow-up (Fig. 1A–C). After an initial increase, highly fluctuating values of the in vitro proliferative response to PHA were observed (Fig. 1F). Four out of five patients showed a normal antibody response to tetanus vaccination. Patient #4 failed to respond to vaccination and is being treated with intravenous immunoglobulins. Patients have been under prophylactic antibiotic treatment since the diagnosis of ADA-SCID; none of them received stem cell transplantation, and only patient #2 received gene therapy (one cycle 6 years before this study). No major infections or other medical problems were observed in the patients since initiation of PEG-ADA therapy. The immunological features of the patients at the time of the latest evaluation are reported in Table 1.

Phenotypic analysis

All patients showed a decrease in the total lymphocyte count associated with low numbers of circulating CD19⁺, CD3⁺, and CD4⁺ lymphocytes (Table 1 and Fig. 1). The CD4/CD8 ratio was fairly homogeneous, resulting in a median of 1.1 (range 0.9–1.5), a value comparable with that reported by Comans-Bitter et al. 16 in a group of 35 age-matched healthy children (median 1.2; 5^th to 95^th percentile 0.9–2.6). The percentage of NK cells was consistently elevated (mean 33.5±16.5 vs. 9±4 in controls; p<0.05); the different subsets of NK cells expressing natural cytotoxicity receptors (NCR), killing inhibitory receptors (KIR), and the inhibitory form of CD94 receptor were present, although in variable percentages. HLA class I-specific KIR were heterogeneously expressed on NK cells from patients as well as from healthy donors. Interestingly, a significant percentage of NK cells from ADA-SCID patients also stained with AZ140 mAb recognizing the NKp44 protein (Table 1).

Analysis of thymic output and T cell repertoire

Since PEG-ADA-treated patients showed a similar proportion of a naive (CD45RA, 548/mL±777/mL) and memory (CD45R0, 491/mL±350/mL) CD3⁺ lymphocytes, we measured the level of TCR excision circles (TREC) in order to investigate whether T cell development in the thymus was occurring normally. Fig. 2A shows that TREC obtained at two different time points after PEG-ADA initiation were consistently and significantly low (p=0.01) in comparison to values obtained from 30 healthy age-matched controls (mean age 7.6±10.4 years). TREC and CD3⁺ lymphocytes were both very low, but there was no relationship between their values.

We analyzed the TCR diversity of cells bearing those TCRβV subfamilies that are known to be preferentially represented in healthy children 17, 18. The migration pattern of TCRβV-specific amplified products was similar in T cells obtained from the five patients and from a representative age-matched control, with most of TCRβV PCR products migrating in polyacrylamide gels as smears, indicating a dominant polyclonal T cell population in all children (Fig. 2B). However, homo and heteroduplex bands were observed in some TCRβV products prepared from the patients, implying the presence of minor monoclonal or oligoclonal lymphocyte expansions in the context of a background of heterogeneous T cells.

Analysis of T cell function

At the time of the present study, a low proliferative activity (see Table 1) was observed in four out of five cultures prepared from PEG-ADA-treated patient cells stimulated with anti-CD3 mAb and pulsed with ³[H]-thymidine (³[H]-TdR) (39 600±26 440 vs. 169 000± 68 774 cpm in healthy controls; p=0.09). This defect was partially corrected by addition of exogenous IL-2 at sub-mitogenic doses (56 700±54 900 vs. 168 300±56 900 cpm). Similarly, the proliferative response to PHA was significantly reduced in all patients (17 400±16 100 vs. 170 100±64 000 cpm; p<0.01). These data demonstrate an impairment of T cell proliferation.

In order to further investigate the basis for impaired T cell proliferation, we used the CFSE fluorescent dye method, which measures the number of cell divisions and allows the identification of dividing cells 19. We found that the percentage of CD3⁺ cells that proliferates in response to PHA is comparable to that of healthy controls (data not shown). In separate experiments, the dilution of CFSE dye in PHA-stimulated cells was measured every day for 4 days (Fig. 3). In accord with the time course of the response to PHA, which follows a Gaussian distribution and reaches the exponential phase around the 3^rd day of culture, at 24 h of culture CD3⁺ cells from neither ADA-SCID children nor controls showed any halving of CFSE, indicating the absence of proliferation. At 48 h lymphocytes from control 1 and from patient #1 started to divide, and 24 h later CD3⁺ cells from the two controls and from patient #1 proliferated in response to PHA. Due to the different proliferation rates, we further analyzed cells from control 1 and patients #1 and #3 after 96 h of culture, and those from control 2 and patient #2 were harvested at 110 h of culture. At these last points, the majority of cells were dividing, making four divisions. These data indicate that T cells from PEG-ADA-treated patients are able to proliferate after mitogen stimulation, although with different kinetics than in controls.

Cells obtained from PHA cultures were also analyzed by flow cytometry in order to evaluate the expression of the anti-apoptotic Bcl-2 protein. The highest levels of Bcl-2 expression were detected just before the burst of proliferation, at 48 h of culture in healthy controls and at 72 h in the patients (data not shown). At 96 h of culture, the relative expression of Bcl-2 attained the same level in patients and controls.

Since an imbalance between proliferation and cell death could be another potential mechanism accounting for the low T cell numbers found in our patients, the percentage of apoptotic T lymphocytes was directly evaluated ex vivo by incubating the cells with Annexin V and anti-CD3 mAb. All patients showed higher levels of Annexin V⁺ T lymphocytes as compared to controls (Fig. 4), and the mean frequency of apoptotic cells among freshly isolated cells was significantly higher in PEG-ADA-treated children that in controls (16.6±6.8 vs. 4.3±1.7, p<0.05).

Finally, we evaluated the ability of lymphocytes from PEG-ADA-treated patients to produce cytokines (IFN-γ and IL-4) by incubating the cells with PMA and ionomycin in the presence of monensin. In resting cultures the percentage of IFN-γ-producing cells was irrelevant (<1%; data not shown). In contrast, after 4–6 h of culture, almost all cells were CD69⁺, and variable percentage of T lymphocytes produced IFN-γ (Fig. 5); these values are similar to those obtained for cells from control children (a representative example is reported in the figure) stimulated under the same conditions. We detected IL-4-producing cells (about 2%) in all stimulated cultures (data not shown).

B cell analysis

While B cell numbers remained lower in patients in comparison to age-matched controls over the entire period of treatment with PEG-ADA (Fig. 1), immunoglobulin levels increased after initiation of therapy (IgG from 296±60 to 847±257 mg/dL; IgA from 8.5±4.2 to 51±18 mg/dL; IgM from 11±7 to 57±24 mg/mL), and isoagglutinins appeared in all patients concomitant with the improvement in T cell function (data not shown). However, patient #4 was unable to respond to tetanus vaccination (Table 1), while patient #2 showed low levels of IgG for a long period of time (Fig 1E).

In order to determine whether the abnormalities in the B cell compartments of these two patients could be associated with defects in immunoglobulin variable gene somatic hypermutation, we examined the frequency of mutations in IGHV3–23 transcripts among total circulating lymphocytes. All the sequences from the two patients contained mutations (Fig. 6), but it was impossible to calculate the percentage of mutations because of the high number of shared mutations that was not explainable by chance, allelic variation, or selection by antigens. Indeed, two groups of nine and three sequences prepared from patient #4 shared, respectively, 25 and 34 identical mutations, and two groups of eight sequences each prepared from patient #2 showed 12 and 3 identical nucleotide substitutions. Therefore, these sequences had to be clonally related, and their presence suggests clonal B cell expansions in both PEG-ADA-treated patients. In contrast, we found that in the other three PEG-ADA-treated patients, the mutation rate was similar (3.6%) to that of age-matched children (2.5%).

Lymphocyte phenotyping

Evaluation of lymphocyte surface membrane expression was performed on whole blood using the following mAb conjugated with FITC, PE or peridinin-chlorophyll protein (PerCP). FITC-conjugated CD3, TCRαβ, CD16, and CD19, PE-conjugated CD4, CD8, CD45RA, CD45R0, and CD69, and PerCP-conjugated CD3 mAb were purchased from Becton Dickinson (Biosciences, San Jose, CA). The NCR mAb AZZ20 (IgG1 anti-NKp30), AZ 140 (IgG1 anti-NKp44), BAB281 (IgG1 anti-NKp46), and FSTR172 (IgG2 activatory receptor anti-NKp50.3/CD158c) as well as the KIR mAb AZ 158 specific for HLA-A and -B (IgG2 anti-p70, -p140), Z27 specific for HLA-B (IgG1 anti-p70), AZZ 115 recognizing HLA-C alleles (IgG1 anti-p58.1 or 50.1/CD158a), Z270 specific for HLA-E (IgG1 anti-NKG2A/CD94 inhibitory receptor), and GL 183 (IgG2 anti-p58.2 or 50.2/CD158b) were kindly donated by Professor A. Moretta (Dipartimento di Medicina Sperimentale, University of Genoa, Italy). FITC- or PE-conjugated isotype-specific goat anti-mouse IgG F(ab)₂ fragment was used as the second reagent.

Analysis of thymic output and T cell diversity

Thymus function was measured by evaluating the TREC production by means of real-time PCR 41, and the TREC values were corrected according to the percentage of CD3⁺ cells in the sample and then expressed as the numbers of TREC/µg CD3⁺ cell DNA. TCR diversity was analyzed by heteroduplex analysis 17. TCRβV products, obtained after PCR amplification of PBMC with TCRβV-specific primers, were heated to 95°C for 5 min and then cooled to 50°C for 1 h. The annealed samples, kept on ice until used, were run for 5–6 h at 200 V on a 12% non-denaturing gel (PAGE; 29:1 acrylamide/bisacrylamide). Gels were stained for 30 min, at room temperature and in the dark, in a solution containing ethidium bromide and then photographed under UV light.

Lymphocyte cultures and proliferation analysis

PBMC, cultured for 3 days in the presence of 1.2 μg/mL PHA, 200 ng/mL anti-CD3 mAb (Becton Dickinson), or 200 ng/mL anti-CD3 mAb plus 20 U/mL recombinant human IL-2 (Biosource International, Camarillo, CA), were pulsed with 1 μCi ³[H]-TdR and harvested 18 h later. Alternatively, PBMC were incubated at 37°C in the dark for 20 min with 200 nM CFSE (Molecular Probes, Eugene, OR) 17 in dimethyl sulfoxide (Sigma), washed twice in cold RPMI/20% FCS, and then cultured with or without PHA. Finally, cells were washed, stained with anti-CD3 mAb, and analyzed by flow cytometry on the 4^th day of culture or, in a different set of experiments, every day for 4 consecutive days.

Annexin V, Bcl-2, and intracellular cytokine staining

PBMC were stained with PerCP anti-CD3 mAb, washed, and incubated with FITC anti-Annexin V (Bender Medical Systems, Vienna, Austria) for 10 min in the dark at room temperature. The percentage of CD3⁺/Annexin V⁺ cells was analyzed by flow cytometry. For the measure of the relative level of Bcl-2 expression, cells were fixed with 4% paraformaldehyde (Sigma) for 5 min on ice, washed, permeabilized in PBS containing 0.2% saponin (Sigma) and 0.2% BSA, placed on ice for 15 min, washed, and then incubated with FITC anti-Bcl-2 mAb (Ancell, Bayport, MN). For intracellular cytokine staining, PBMC (5 × 10⁵) were incubated with monensin (10 μg/mL; Sigma) in the presence of PMA (5 ng/mL; Sigma) and ionomycin (500 ng/mL; Sigma) or left unstimulated for 4–6 h in a 5% CO₂ humidified incubator. Cells were then labeled with FITC anti-TCRαβ mAb for 30 min at 4°C, washed in PBS, fixed with 4% paraformaldehyde, and permeabilized with 0.2% saponin. Finally, cells were incubated with PE-conjugated anti-IFN-γ, anti-IL-4, anti-CD69, or control isotype-specific mAb (Becton Dickinson) in the dark at 4°C for 30 min.

IgG somatic hypermutation

Somatic hypermutation analysis was performed on CD19⁺ cells separated using Dynabeads M-450 Pan-B (CD19). Briefly, total mRNA was extracted with MicroPoly (A) Pure kit (Ambion, Austin, TX), reversed transcribed using random hexamers provided with the RNA PCR Core Kit (Applied Biosystems, Foster City, CA) and the cDNA amplified with specific primers for the variable region IGHV3–23 leader and for the IgG constant region as already described 42. PCR products were cloned using the TOPO-TA cloning kit (InvitroGen, San Diego, CA) and, for each patient, 15–20 positive colonies were sequenced with the BigDye terminator cycle sequencing kit (Applied Biosystems) and analyzed with the ABI Prism 310 Genetic Analyzer and the Sequence Navigator Software (Applied Biosystems).