Main

CGD is a rare genetic disease, which is classified into four genetic types(1). Patients with CGD suffer from recurrent episodes of catalase-positive bacterial or fungal infections due to a lack of superoxide production in the phagocytes. Impairment of superoxide production in the phagocytes results from a defect involving one of the essential components of the NADPH oxidase system. X-CGD is the most common type, whose responsible gene has been identified and termed CYBB, according to the gp91-phox, a subunit of cytochrome b558(2). Genetic analysis of patients with X-CGD revealed that the mutation responsible for the disease varies greatly(3). A report based on a database of X-CGD-causing mutations showed 192 unique mutations in 261 families(4). However, no precise molecular genetic study for the proportion of sporadic cases in X-CGD has been reported. In the present study, we performed a genetic analysis of six newly identified X-CGD patients together with the eight previously reported X-CGD patients(5–9). The mutations newly detected were all novel. Twelve mothers (two of them were from the same family) and four maternal grandmothers of the patients from 13 families were available. Molecular genetic analysis of the family members using the mutation information revealed that only one of the 12 X-CGD patients was a sporadic case, whereas two of the five mothers were sporadic carriers. A low proportion of sporadic patients and a high proportion of sporadic carriers suggest that mutation for the disease originated mainly from male gametes.

METHODS

Patients. Fourteen patients with X-CGD, eight of whom had already been reported(5–9), were included in this study. Two of them (CGD 2 and 3) were maternal cousins. Diagnosis of CGD was made from the clinical feature of recurrent pyogenic episodes and a functional assay of the patient's phagocytes, chemiluminescence production study, and/or a nitro blue tetrazolium reduction test. For determination of genetic classification, we considered the patient's sex, familial history (actually there were no positive familial histories except for CGD 2/3), and the results of Western blot studies for four components of NADPH oxidase. The final diagnosis of patients as having X-CGD was made by the detection of a mutation in the patient's gp91-phox gene. Two patients (CGD 5 and 22) were diagnosed as having cytochrome b-positive X-CGD.

Mutation analysis. High molecular weight DNA and total RNA were purified from peripheral blood lymphocytes, granulocytes, or established EBV-lymphoblastoid cell lines using standard phenol/chloroform methods and RNA isolation solvent (RNAzol™; Cinna/Biotecx, Houston TX), respectively. Single strand cDNA was synthesized from 2 µg of RNA using a First Strand cDNA Synthesis kit according to the manufacturer's instructions (Pharmacia LKB Biotechnology, Uppsala, Sweden). The cDNA of the gp91-phox was amplified with PCR in three overlapping fragments using three pairs of primers that were originally designed by Dinauer et al.(10). PCR was performed in 20 µL volumes using a buffer consisting of 50 mM KCl, 10 mM Tris-HCl, pH 8.0, 200 mM dNTPs, 100 pM of a pair of primers, and 2.5 U of Amplitaq polymerase (Perkin-Elmer, Foster City, CA). The conditions used for PCR were denaturation at 94°C for 30 s, annealing at 58°C for 30 s, and polymerization at 72°C for 30 s. Twenty-five to 30 cycles of amplification were performed using the GeneAmp® PCR System 2400 (Perkin-Elmer). The amplified fragments were purified and then directly sequenced by using the ABI PRISM Dye Terminator Cycle Sequencing Ready Reaction Kit (Perkin-Elmer) and the automated ABI 373A DNA sequencer. In some cases, the sequencing study was carried out using the Sequenase version II kit (U. S. Biochemical Corp., Cleveland, OH) with35 S-labeled dCTP (Amersham, Corp., Arlington, IL). Primers used for sequencing were the same as those used for PCR at a concentration of 3.2 pM. Each mutation was confirmed at least twice using a different set of the PCR products. Sequencing was performed in both directions. Furthermore, each mutation was confirmed in a fragment amplified by genomic PCR. We amplified the fragment including an exon containing the mutation. Primers used for genomic PCR were located in the flanking region of each exon of the gp91-phox gene, which were recently reported by Hui et al.(11). The amplified fragment including the mutation was directly sequenced in the same way.

Carrier determination. Twelve mothers (two of them were sisters), four maternal grandmothers, and five sisters of the X-CGD patients were available for carrier diagnosis. Previously, we reported mutation analysis for eight patients with X-CGD(6–9) from seven families (two of them were maternal cousins). We also determined the carrier status for five of the seven X-CGD family members based on the patients' mutation information(5–7,9). In this study, we again used the mutation information to determine the carrier status of the remaining two families and the other six X-CGD family members. Genomic DNA was used for all of the carrier determination experiments.

We approached the diagnosis by allele-specific PCR (Al-PCR) (families of CGD 9 and 25), by digestion of a PCR fragment with a restriction enzyme(D-PCR) in cases where a recognition site was created (family of CGD 22;Nco I) or eliminated (family of CGD 24; ApaI) by the mutation, and by direct sequencing of the PCR fragment (Seq-PCR)(families of CGD 12, 21, 22, and 25). Primers used for A1-PCR/D-PCR are listed in Table 1. Restriction enzymes used were obtained from Takara Shuzo Co., Ltd., Kyoto, Japan.

Table 1 Primers used for Al-PCR and D-PCR
Full size table

Primers used for Al-PCR were made as previously reported(12). We set two pairs of 20-bp primers for each family. Primers Mu, for mutation allele, and primers No, for normal allele, were used. Both the primers included a common reverse primer. The difference between Mu and No was only 1 base, located at the 3′ end of forward primers, which originated from the sequence of a mutant or a normal allele.

For D-PCR, a pair of primers both located at the flanking mutation sites were synthesized. Because the mutation sites of patients CGD 22 and 24 were very close to each other, the same pair of primers were available. PCR products were electrophoresed after digestion with the proper restriction enzyme. Primers used for Seq-PCR were the same as those used for genomic PCR, which were used to confirm each patient's mutation, and then the products were directly sequenced as mentioned above.

RESULTS

Mutations detected in six patients with X-CGD are shown in Figs. 1 and 2. The mutations detected were three missense mutations (patients CGD 22, 24, and 25), two splice mutations(patients CGD 6 and 21), and one insertion mutation of 2 bases (patient CGD 7). All the mutations were novel. Patient CGD 22 was positive for cytochrome b558 (data not shown); a precise study for the defective gp91-phox is ongoing. The other five CGD patients were cytochrome b558-negative.

Figure 1
figure 1

The results of mutation analysis of patients CGD 7, 22, 24, and 25. The cDNA of the gp91-phox was amplified by RT-PCR in three overlapping fragments using three pairs of primers (see"Mutation Analysis" under "Methods"). Then, the amplified fragments were directly sequenced. The change of nucleotides or amino acids are underlined. The structure of the gp91-phox is also shown in which locations of each mutation detected are indicated with arrows.

Full size image
Figure 2
figure 2

The results of mutation analysis of patients CGD 6 and 21. (A) Structure of cDNA of the gp91-phox. A translation initiation codon (ATG) and a termination codon (TAA) are indicated. The locations of three overlapping fragments amplified by RT-PCR(frag. 1-3) are shown as arrow-end bars. (B) The results of RT-PCR. Abbreviations: P, patient; M, mother; GM, maternal grandmother; C, normal control subject. Fragment 1 of patient CGD 6 and fragment 2 of patient CGD 21 were abnormally short. Both of the abnormal fragments were processed for further analysis (indicated by arrows). (C) The results of sequencing analysis of the abnormally short fragments detected in both the patients. Junctional sites of exons are indicated as arrowheads; exon 5 of patient CGD 6 and exon 8 of patient CGD 21 were revealed to be skipped. (D) The results of sequencing analysis around the skipped exons of both the patients. The fragments including the skipped exons with flanking introns were amplified by genomic PCR and then were directly sequenced. The intron-exon junction is indicated as a vertical bar. Nucleotide changes from the normal sequence in both of the patients are indicated by arrows.

Full size image

Using mutation information detected in the patients with X-CGD, carrier diagnoses for the patients' female relatives were performed(Table 2). We found a sporadic patient and two sporadic carriers in three different X-CGD families (CGD 21, 22, and 25). The mother of CGD 21 seemed not to be a carrier as the result of RT-PCR (Fig. 2B), and this fact was confirmed by the Seq-PCR method (Fig. 3). The maternal grandmothers of CGD 22 and 25 were proved not to be carriers by D-PCR/Seq-PCR and A1-PCR/Seq-PCR, respectively (Figs. 3 and 4); however, the mothers of CGD 22 and 25 were determined as being carriers. Mothers of CGD 9(Fig. 4) and 12 (Fig. 3) were proved to be carriers. Mothers of CGD 6 and 7 were not available. The results of other X-CGD family members (CGD 1, 2/3, 4, 5, and 15) were previously reported(5–7,9).

Table 2 Summary of carrier status in family members of X-CGD patients
Full size table
Figure 3
figure 3

Carrier diagnosis of the patients' family members by direct sequencing of PCR products. Results showed the antisense strand of the PCR fragments for families of CGD 12, 21, and 25, and sense strand for the family of CGD 22. Arrowheads indicate the mutation site in each family. Mothers of CGD 12, 22, and 25 showed ambiguous nucleotides at this site.

Full size image
Figure 4
figure 4

Carrier diagnosis of the patients' family members by PCR-related methods. The results of allele-specific PCR(families of CGD 9 and 25), and digestion of PCR products with a restriction enzyme (family of CGD 22 with NcoI, family of CGD 24 with Apa I) are shown. Pat, patient; Mo, mother;M.GM, maternal grandmother; M.A, maternal aunt;Cont, control. Primers for normal and mutant alleles are indicated as No and Mu, respectively. Plus (+) and minus (-) indicate digestion with and without the restriction enzyme, respectively.

Full size image

DISCUSSION

Here we showed the results of genetic analysis of six newly identified patients with X-CGD in addition to the eight previously reported patients. Mutations detected in the gp91-phox gene were all novel. Six mutations in the gp91-phox were three missense mutations (patients CGD 22, 24, and 25), two splice mutations (patients CGD 6 and 21), and one insertion mutation of 2 bases (patients CGD 7).

Deduced amino acids substitution for patients CGD 22, 24, and 25 were Arg54 to Met, Ala55 to Asp, and Ser344 to Phe, respectively. It was interesting to note that patient CGD 22 was cytochrome b-positive. Most of the missense mutations in the gp91-phox gene resulted in a lack of cytochrome b558 from the phagocyte cell membrane. It has been reported that two cases of missense mutation located at a specific area of N-terminal domain, Arg54 to Ser(13) and Ala57 to Glu(6), were cytochrome b-positive. Mutation of our patients CGD 22 and 24 were also located in this area. It seemed that some missense mutations in this area, by some means, turned out to be cytochrome b-positive X-CGD, even though patient CGD 24 was cytochrome b-negative. In addition to the mechanism of dysfunction of these patients' abnormal gp91-phox (study of patient CGD 22 is ongoing), the mechanism of why most of the missense mutations in X-CGD result in a lack of cytochrome b, whereas some missense mutations in these area do not, should be defined.

We found splice mutations in patients CGD 6 and 21. An abnormally short fragment amplified by RT-PCR was observed in both patients, and a sequencing study revealed that exon 5 and exon 8 was missing in patients CGD 6 and 21, respectively. Finally, a sequencing study for exon/intron boundary showed 1 base change in an acceptor site of intron 4 (ag to aa) in patient CGD 6 and 1 base change in that of intron 7 (ag to gg) in patient CGD 21. Although X-CGD cases in which the same exons were skipped have been reported(3,14), they resulted from different base changes in other sites. The mother of CGD 21 did not show an abnormally small fragment by RT-PCR (Fig. 2B), suggesting she was not a carrier for the disease. However, the possibility of highly imbalanced expression of her normal gp91-phox gene was not ruled out.

The mutation detected in patient CGD 7 was insertion mutation of two bases. Slipped missparing might be the responsible mechanism for the mutation of TTT to TTTTT in exon 7.

We also determined the carrier status for the female family members of these X-CGD patients (Table 2). It was found that 11 mothers (two of them were from the same family) of 12 X-CGD patients were carriers and one mother was not. Two mothers, of CGD 6 and 7, were not available for the study. Four maternal grandmothers, of CGD 4, 21, 22, 25, were studied. It was shown that three of them (including a grandmother of CGD 21 whose mother was not a carrier) were not carriers. Although we could not confirm it in this study, a maternal grandmother of patients CGD 2/3 seemed to be a carrier for the disease, because she had two daughters who were proved to be carriers. Thus, only 1 out of the 12 mothers of X-CGD patients was not a carrier, and conversely, two of four carriers' mothers (maternal grandmothers) turned out not to be carriers.

Haldane(15) proposed a formula for the proportion of sporadic cases in X-linked recessive disorders, m = (1 -f)u/(2 u + v) (m, proportion of sporadic cases; f, fertility;u, mutation rate in female gametes; and v, mutation rate in male gametes). Thus, the proportion of sporadic cases depends on fertility and the different mutation rates of both sexes. The prognosis of X-CGD has improved recently, and successful bone marrow transplantation is effective in some cases of the disease(16). Therefore, the m value in X-CGD will be decreased in the future with an increment of the f value. However, to make it simple, we counted f = 0 because patients with X-CGD still could rarely have children. To prevent the evaluation under ascertainment bias, we counted two patients in the same family as one. Accordingly, our results were m = 1/11. This value is significantly smaller than the predicted value of m = 1/3 (p < 0.08;F distribution). The sexual ratio of mutation rates (v/u) is then calculated as 9 if the value of m = 1/11 is correct. We admit that the number of families studied is small, which may, therefore, lead to inaccurate results. However, there are some lines to support the imbalance of sexual mutation rates. First, we showed that two out of four carriers' mothers (maternal grandmothers) were not carriers for the disease. Although the number of maternal grandmothers studied was small, we evaluated that the proportion of sporadic carriers for X-CGD (2/5; including two carriers in the same family) was much higher than that of sporadic patients (1/12; including two patients in the same family; Table 2). In sporadic carriers, the mutation might have originated mainly from the paternal gametes because a mutation which originated from male gametes resulted in a carrier, not a patient. In fact, two cases with X-CGD, whose responsible mutations were proved to have originated from grandparental gametes, were reported(17); however, no samples from patients' grandfathers were available in this study. Second, imbalanced sexual mutation rates have been reported in other X-linked recessive diseases such as Duchenne muscular dystrophy(18) and hemophilia A(19).

Without molecular genetic studies, the diagnosis of a patient as having X-CGD sometimes meets with difficulty. In general, X-CGD diagnosis depends on a carrier diagnosis of a patient's mother for the disease. The carrier diagnosis is made from the evaluation of her phagocyte function; an intermediate value between patients and normal individuals. Therefore, if we have a male cytochrome b-negative CGD patient whose mother shows normal values in the nitro blue tetrazolium reduction test, one may conclude the patient to be a sporadic case of X-CGD. However, there are two other possibilities. The patient may have CGD with a defective p22-phox. Cytochrome b558 consists of gp91-phox and p22-phox, and a defect of either would result in cytochrome b-negative CGD in most cases. In addition, carriers of CGD with a defective p22-phox (autosomal recessive disease) show normal phagocyte function. In other cases, the patient may have X-CGD, and his mother still be a carrier for the disease. Imbalance lyonization of the gp91-phox gene may lead to a misdiagnosis for the mother as a noncarrier, just like a carrier for X-CGD becomes symptomatic in some cases(20). From this point of view, we should not determine the mother of CGD 21 as a noncarrier just from the results of RT-PCR (Fig. 2B). We confirmed her as a noncarrier from the results of genomic PCR studies. Furthermore, X-CGD diagnosis may be confused in a patient with cytochrome b-positive X-CGD. Therefore, previous studies for X-CGD and carrier diagnosis based mainly on functional/biochemical results might contain a misdiagnosis. Because most of these misdiagnoses would make the proportion of sporadic X-CGD larger, reports of sporadic X-CGD cases as 3/19(21) or 4/9(22) might overestimate the proportion. No study for the proportion of sporadic cases for X-CGD based on molecular genetic study has been reported. Although the number of families we studied was not large, the diagnosis for X-CGD and carrier was definite because of genomic information. An increased number of X-CGD families for molecular genetic study would define the correct mutation rates in the two sexes. Finally, we have to take into account the change of fertility of X-CGD patients and also the occurrence of germinal mosaicism in some carriers(23).