IgG4 deficiency
Eur. J. Irnmunol. 1992. 22: 227-233
Roberto Gallina., Andrea Bottaro., Cleide Boccazzio, Gerda DeLange., Paola Danesen, Gina Mazzolaav, Antonio Amorosoa, Mario DeMarchie and Angelo Oscar Carbonara. Dipartimento di Genetica, Biologia e Chimica Medica e Centro CNR Immunogenetica ed Oncologia sperimentale., Torino, Central Laboratory of the Netherlands Red Cross., Blood Transfusion Service, Amsterdam, Servizio Immunologia e TrapiantP, Ospedale Maggiore S. Giovanni Battista, Torino and Dipartimento di Biologia Molecularee, Universita di Siena, Siena
227
The genetics of IgG4 deficiency: role of the immunoglobulin heavy chain constant region and HLA loci* IgG4 deficiency is very common (1/400 in the Italian population) and provides a good model for analyzing the genetic factors involved in Ig subclass deficiencies. We have previously reported an association between some immunoglobulin heavy chain constant region (IGHC) polymorphisms and the IgG4 deficiency.The associated polymorphisms spanned the region between the G P and the G4 genes. A larger sample composed of 50 healthy blood donors with IgG4 deficiency (< 0.001 g/l IgG4), not carrying homozygous gene deletions, together with 82 first-degree relatives is now examined. The results confirmed the association of the deficiency with IGHC polymorphisms, and detected a new association with the HLA-D locus with a strong additive effect between the two systems. However, despite these associations and a highly significant risk for IgG4 deficiency within families, close linkage with either IGHC or HLA loci was not apparent by the affected sib pair method. These findings suggest that several concomitant, possibly cooperating, genetic factors may be involved in IgG4 deficiency.
1 Introduction IgG4 is the least abundant of the four human IgG subclasses, with adult values of 0.2-1.0 g/l [l]. Levels of IgG4 below 0.001 g/l are not unusual in the normal population, with a frequency of 1/400 among Italians [2]. Lack of IgG4 is therefore the most common type of Ig isotype deficiency. While generally asymptomatic, it has also been incriminated as a cause of recurrent infections [3, 41, though the unclear functional role of IgG4 makes this correlation uncertain. Its main role seems t o be mediation of responses against protein antigens [5] and this subclass predominates in prolonged antigenic challenges [6, 71. The IgG4 response function parallels that of IgE and their co-regulation has indeed been observed [8].
to homozygotes, we suggested that a significant fraction of cases was due to a minor structural defect in the IGHC region. This association with IGHC RFLP has now been confirmed and extended in a larger sample of IgG4-deficient subjects, who were also investigated for the association with HLA class I1 DRB, DQB and DQA RFLI? Moreover, 82 firstdegree relatives were analyzed to determine the familiar recurrence of IgG4 deficiency and the segregation of IGHC and HLA markers.
2 Materials and methods 2.1 Subject ascertainment and IgG4 evaluation
Previous work has shown that a small fraction of IgG4 deficiencies stem from deletions in the Ig H chain C region (IGHC) region [2] that are usually accompanied by other isotype deficiencies. Most cases do not present DNA rearrangements and for them no clear cause has been yet readily demonstrated. A significant association with IGHC RFLP in the GP-G4 genes region, in particular with an RFLP (IGHPG4) mapping 1.2 kb 5’ of the G4 gene switch (S) region was reported in one of our previous studies [9]. Since the increased relative risk (RR) values were restricted
Fifty unrelated Italian probands with IgG4 deficiency not due to gene deletions were examined (32 maledl8 females; age 45311.2). They were detected during a two-step serological screening [first by ELISA and then by hemagglutination inhibition assay, (HIA)] of 22 000 healthy blood donors (frequency = 0.0023) [2]. By this protocol, negative sera are estimated to have < 0.001 g/l of IgG4. Quantitative levels of Ig isotypes of the probands are shown in Table 1. Eighty-two first-degree relatives (14 parents, 60 sibs and 8 Table 1. Ig levels (g/l) in the IgG4-deficient subjects
[I 97141
Deficient subjects
Normala) population
1.43 f 0.82 10.80 f 2.09 2.16 f 0.96 33 UI/mlb)
1.25 f 0.5 12.50 f 3 2.10 f 0.5 32 UUml
* This work was supported by CNR Target Project “Biotechnology and Bioinstrumentation”, by MURST 60%, Torino and Siena, and Ricerca finalizzata Regione Toscana. Recipient of a fellowship from the Regione Piemonte. Correspondence: Angelo Carbonara, Dipartimento di Genetica, Biologia e Chimica medica, Via Santena 19, 1-10126 Torino, Italy Abbreviation: IGHC: Ig H chain C region
0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1992
IgM IgG IgA IgE
a) As indicated in [lo]. b) Geometric mean.
+
0014-2980/92/0101-0227$3.SO .2S/O
228
Eur. J. Immunol. 1992. 22: 227-233
R. Gallina, A. Bottaro, C. Boccazzi et al.
All probands and their relatives were typed for Gm markers by HIA, as previously described 1121.
calculated using Woolf's formula [181; the significance was evaluated with Yates' corrected x2.The probability level of allele frequencies was corrected (pc) for the number of comparisons at each locus. Genotype RR were estimated by comparing the figures observed in patients with HardyWeinberg expectations in controls [191. The probability level was evaluated from the statistic (In RR)2N (V = variance), which has an asymptotic x2 distribution with one degree of freedom. The sib-pair test for detecting linkage was carried out as proposed by Penrose [20].
2.3 Southern blot analysis
3 Results
All deficient probands and 69 of their relatives (12 parents, 55 sibs and 2 children) were examined by Southern blotting for IGHC and HLA-D RFLP. Phenolkhloroform extraction of genomic DNA, endonuclease digestion (Bam HI, Hind 111, Sac1 and Taq I restriction enzymes; BoehringerMannheim, Mannheim, FRG), agarose gel electrophoresis, Southern blotting onto nitrocellulose (Schleicher and Schuell, Dassel, FRG) or nylon membranes (Hybond N, Amersham Int., Amersham, GB), hybridization with nick translated or primer-elongated DNA probes and autoradiography to XAR5 Kodak films followed standard protocols.
3.1 IGHC
Several IGHC markers were tested: eight belonging to the D, G and A genes, described in [13]; moreover, two highly polymorphic probes, IGHSA and p3.4, were employed. IGHSA (clone pHS.2-2.0) was a kind gift from N. Migone, Dipartimento di Genetica, Biologia e Chimica medica, University of Torino. It detects, SA1, SA2 and SM polymorphisms [14, 151. p3.4, a gift from R. White-Howard Hughes Medical Institute (Salt Lake City, UT) detects a variable number of tandem repeats region 5' of the IGHJ4 segment [16]. Family segregation was assessed with all the polymorphic probes described.
3.1.2 Structural IGHC defects
children) of 26 probands were examined by ELISA and/or HIA for the lack of IgG4. The controls were 95 random Italian subjects for IGHC markers and 109 for HLA class I1 RFLP [9, 111.
2.2 Gm typing
The HLA class I1 RFLP were analyzed with DRB, DQA and DQB cDNA probes, as described in [17]. Haplotypes and the corresponding HLA-D specificity were named according to nomenclature proposed by Bidwell and Bignon [17].
2.4 Statistical analysis Proband phenotype and deduced allele frequencies were compared with controls in 2 X 2 tables; RR values were
3.1.1 RFLP association The 9.3-kb PG4 allele (IGHPG4"Hl) showed a frequency of 0.57 vs. 0.41 in controls (pc = 0.0334) (Table 2). This variation is solely due to the increase in PG4*H1 homozygous subjects (RR = 3.01, p < 0.05) (Table 3). Significant associations were also found for the GP*Al, PG2*H1 and G2*A1 homozygotes, heterozygotes showing again no increase.
Five heterozygotes for gene deletions were found among the 50 IgGCdeficient subjects. Proband NEG is hemizygous for the G4 gene only, carrying a deletion already reported in the literature (211. BER is carrying a deletion encompassing the genes from A1 to E similar to that described in other subjects [21-241. ROG carries a new type of deletion encompassing only the G2 and G4 genes (unpublished). In family MAR (see pedigree in Fig. l ) , Southern analysis demonstrated that the proband's father (subject 1-1) lacks GP-G2-GCassociated fragments (Fig. 2); analysis with the A and E probes (not shown) made it clear that the father is a compound heterozygote for one deletion spanning from GP to A2 and another from A1 to E (both already described [21-261); two normal sons (11-1,2) are heterozygous for the first deletion, the proband (11-3) for the second. In family TES (see pedigree in Fig. l ) , Southern analysis showed that all nine sibs are heterozygous for a deletion extending from A1 to E genes (Fig. 3). Here, too, one parent was most probably a deletion homozygote.
Table 2. Allele frequency of IGHC RFLP in normal and IgG4-deficient subjects Locus
IGHGP IGHPG2
Allele
(kb)
*A1 *A2 *H1 *H2 *H3
(9.7) (8.5) (3.2) (3.35) (3.4) (12.5) (22) (9.3) (9.5)
IGHG2
*A1
IGHPG4
*A2 *H1 *H2
Blood donors n Freq
52 45
44 0 52 50
46 55 41
.54 .46 .46
.OO .54 .52 .48 .57 .43
Controls n Freq
x2"'
124 66 52 1 137 126
8.68 8.68 8.94 N.S.b) 8.37 4.87 4.87 5.73 5.73
64 78 112
.35 .65 .27 .01 .72 .66 .34 .41 .59
PC
0.0064
0.0084 0.0546 0.0334
a) Yates' corrected, from 2 X 2 contingency tables of allele numbers. b) N.S. = not significant (p > 0.05).
Eur. J. Immunol. 1992. 22: 227-233
IgG4 deficiency
229
Table 3. IGHC phenotypes and IgG4 deficiency Locus
GP
9.7 9.7i8.5 8.5 3.2 3.2/Xe) x1x 22 22112.5 12.5 9.3 9.319.5 9.5
PG2 G2 PG4
bc
bd
Controls
n
%
observed n %
19 16 15 15 16 19 16 16 18 22 13 15
(38) (32) (30) (30) (32) (38) (32) (32) (36) (44) (26) (30)
10 46 39 8 36 51 11 42 42 17 44 34
HW expected n %
(10.5) (48.4) (41.1) (8.4) (37.9) (53.7) (11.6) (44.2) (44.2) (17.9) (46.3) (35.8)
11.64 43.22 40.14 6.93 37.45 50.63 10.98 42.64 41.38 15.97 45.96 33.07
(12.2) (45.5) (42.2) (7.3) (39.4) (53.3) (11.6) (44.9) (43.6) (16.8) (48.4) (34.8)
4.34 0.99 1.OO 5.81 1.14 1.00 3.36 0.87 1.00 3.01 0.63 1.00
6.18**) N.S.d) 0.00 8.82*** N.S. 0.00 4.23* N.S. 0.00 3.68* N.S. 0.00
~191. IJ) (In RR)*/Variance [19]. c) Significance: * = p < 0.05; ** =p 0.00s). e) X = All other alleles
3.3 Combined HLADGHC analysis
ad
yw yw
a c a d bc bc
bd ad a c b d a c
xw x z
xw yz y 2 yw xw
yw
XZ
The distribution of the joint HLA and IGHC genotypes is cross-tabulated in a 3 x 3 contingency table (Table 6). For F
TES
MAR
* ;T*5b $ ;J '
I
0
G
I
T
N l
C
v
l
) 1
~ ~ n t = i
probe
-22
4-
'2
1
a) RR for each genotype vs. X/X; as control genotype frequencies Hardy-Weinberg expectations were used
#*A*
I
XZ
Deficient probands
Phenotype
$$%I;;:$
p/
GI
b c bc ad
xw y z
ac
ad
ac
bc
yw yz
xw
xw
xw
Ig' K E SOM
8; 5
PES
CA F
I
G3-
GP-
I
II
l&;x,h k E
% B,d,
family MAR
Figure 2. Southern blot analysis of the IGHC genes in the MAR family. Note the lack of the G2, GP an G4 bands in subject 1-1. Band intensity and segregation analysis confirm that the three children are heterozygous for the same deletion. G 1 and G3 bands are normal.
DIM
Figure I . Pedigree of multiple-case families. The proband is indicated with an arrow. a, b, c, d = IGHC haplotypes; x, y. w, z = HLA haplotypes. Full shadow = lack of IgG4; half shadow < 10 kg/ml. In family MAR and TES all the siblings are heterozygous for deletions involving the G4 gene.
3.2 HLA The 50 probands were analyzed for HLA class I1 RFLP. Significant associations with the D R B l l (0.38 vs. 0.202, p c = 0.0228), DQB3b (0.46 vs.0.257, p c = 0.004) and DQA3b (0.41 vs. 0.225, p c = 0.0077) alleles were found. These alleles together form the HLA-Dw5 (25) haplotype (Table 4). As for the IGHC locus the association is higher for homozygotes; here, however, the frequency of heterozygotes is also significantly increased (Table 5 ) .
F N c y ? y w ~ - ~
~
~
H
H
H
pSPG4 PG4 PGI
~
W
W
-
A
0.6
-0.3
-
-7.7
-4.66
PG3-
-
-3.6 -3.6 3.4
PD< PG2
Eur. J. Irnmunol. 1992. 22: 227-233
Roberto Gallina., Andrea Bottaro., Cleide Boccazzio, Gerda DeLange., Paola Danesen, Gina Mazzolaav, Antonio Amorosoa, Mario DeMarchie and Angelo Oscar Carbonara. Dipartimento di Genetica, Biologia e Chimica Medica e Centro CNR Immunogenetica ed Oncologia sperimentale., Torino, Central Laboratory of the Netherlands Red Cross., Blood Transfusion Service, Amsterdam, Servizio Immunologia e TrapiantP, Ospedale Maggiore S. Giovanni Battista, Torino and Dipartimento di Biologia Molecularee, Universita di Siena, Siena
227
The genetics of IgG4 deficiency: role of the immunoglobulin heavy chain constant region and HLA loci* IgG4 deficiency is very common (1/400 in the Italian population) and provides a good model for analyzing the genetic factors involved in Ig subclass deficiencies. We have previously reported an association between some immunoglobulin heavy chain constant region (IGHC) polymorphisms and the IgG4 deficiency.The associated polymorphisms spanned the region between the G P and the G4 genes. A larger sample composed of 50 healthy blood donors with IgG4 deficiency (< 0.001 g/l IgG4), not carrying homozygous gene deletions, together with 82 first-degree relatives is now examined. The results confirmed the association of the deficiency with IGHC polymorphisms, and detected a new association with the HLA-D locus with a strong additive effect between the two systems. However, despite these associations and a highly significant risk for IgG4 deficiency within families, close linkage with either IGHC or HLA loci was not apparent by the affected sib pair method. These findings suggest that several concomitant, possibly cooperating, genetic factors may be involved in IgG4 deficiency.
1 Introduction IgG4 is the least abundant of the four human IgG subclasses, with adult values of 0.2-1.0 g/l [l]. Levels of IgG4 below 0.001 g/l are not unusual in the normal population, with a frequency of 1/400 among Italians [2]. Lack of IgG4 is therefore the most common type of Ig isotype deficiency. While generally asymptomatic, it has also been incriminated as a cause of recurrent infections [3, 41, though the unclear functional role of IgG4 makes this correlation uncertain. Its main role seems t o be mediation of responses against protein antigens [5] and this subclass predominates in prolonged antigenic challenges [6, 71. The IgG4 response function parallels that of IgE and their co-regulation has indeed been observed [8].
to homozygotes, we suggested that a significant fraction of cases was due to a minor structural defect in the IGHC region. This association with IGHC RFLP has now been confirmed and extended in a larger sample of IgG4-deficient subjects, who were also investigated for the association with HLA class I1 DRB, DQB and DQA RFLI? Moreover, 82 firstdegree relatives were analyzed to determine the familiar recurrence of IgG4 deficiency and the segregation of IGHC and HLA markers.
2 Materials and methods 2.1 Subject ascertainment and IgG4 evaluation
Previous work has shown that a small fraction of IgG4 deficiencies stem from deletions in the Ig H chain C region (IGHC) region [2] that are usually accompanied by other isotype deficiencies. Most cases do not present DNA rearrangements and for them no clear cause has been yet readily demonstrated. A significant association with IGHC RFLP in the GP-G4 genes region, in particular with an RFLP (IGHPG4) mapping 1.2 kb 5’ of the G4 gene switch (S) region was reported in one of our previous studies [9]. Since the increased relative risk (RR) values were restricted
Fifty unrelated Italian probands with IgG4 deficiency not due to gene deletions were examined (32 maledl8 females; age 45311.2). They were detected during a two-step serological screening [first by ELISA and then by hemagglutination inhibition assay, (HIA)] of 22 000 healthy blood donors (frequency = 0.0023) [2]. By this protocol, negative sera are estimated to have < 0.001 g/l of IgG4. Quantitative levels of Ig isotypes of the probands are shown in Table 1. Eighty-two first-degree relatives (14 parents, 60 sibs and 8 Table 1. Ig levels (g/l) in the IgG4-deficient subjects
[I 97141
Deficient subjects
Normala) population
1.43 f 0.82 10.80 f 2.09 2.16 f 0.96 33 UI/mlb)
1.25 f 0.5 12.50 f 3 2.10 f 0.5 32 UUml
* This work was supported by CNR Target Project “Biotechnology and Bioinstrumentation”, by MURST 60%, Torino and Siena, and Ricerca finalizzata Regione Toscana. Recipient of a fellowship from the Regione Piemonte. Correspondence: Angelo Carbonara, Dipartimento di Genetica, Biologia e Chimica medica, Via Santena 19, 1-10126 Torino, Italy Abbreviation: IGHC: Ig H chain C region
0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1992
IgM IgG IgA IgE
a) As indicated in [lo]. b) Geometric mean.
+
0014-2980/92/0101-0227$3.SO .2S/O
228
Eur. J. Immunol. 1992. 22: 227-233
R. Gallina, A. Bottaro, C. Boccazzi et al.
All probands and their relatives were typed for Gm markers by HIA, as previously described 1121.
calculated using Woolf's formula [181; the significance was evaluated with Yates' corrected x2.The probability level of allele frequencies was corrected (pc) for the number of comparisons at each locus. Genotype RR were estimated by comparing the figures observed in patients with HardyWeinberg expectations in controls [191. The probability level was evaluated from the statistic (In RR)2N (V = variance), which has an asymptotic x2 distribution with one degree of freedom. The sib-pair test for detecting linkage was carried out as proposed by Penrose [20].
2.3 Southern blot analysis
3 Results
All deficient probands and 69 of their relatives (12 parents, 55 sibs and 2 children) were examined by Southern blotting for IGHC and HLA-D RFLP. Phenolkhloroform extraction of genomic DNA, endonuclease digestion (Bam HI, Hind 111, Sac1 and Taq I restriction enzymes; BoehringerMannheim, Mannheim, FRG), agarose gel electrophoresis, Southern blotting onto nitrocellulose (Schleicher and Schuell, Dassel, FRG) or nylon membranes (Hybond N, Amersham Int., Amersham, GB), hybridization with nick translated or primer-elongated DNA probes and autoradiography to XAR5 Kodak films followed standard protocols.
3.1 IGHC
Several IGHC markers were tested: eight belonging to the D, G and A genes, described in [13]; moreover, two highly polymorphic probes, IGHSA and p3.4, were employed. IGHSA (clone pHS.2-2.0) was a kind gift from N. Migone, Dipartimento di Genetica, Biologia e Chimica medica, University of Torino. It detects, SA1, SA2 and SM polymorphisms [14, 151. p3.4, a gift from R. White-Howard Hughes Medical Institute (Salt Lake City, UT) detects a variable number of tandem repeats region 5' of the IGHJ4 segment [16]. Family segregation was assessed with all the polymorphic probes described.
3.1.2 Structural IGHC defects
children) of 26 probands were examined by ELISA and/or HIA for the lack of IgG4. The controls were 95 random Italian subjects for IGHC markers and 109 for HLA class I1 RFLP [9, 111.
2.2 Gm typing
The HLA class I1 RFLP were analyzed with DRB, DQA and DQB cDNA probes, as described in [17]. Haplotypes and the corresponding HLA-D specificity were named according to nomenclature proposed by Bidwell and Bignon [17].
2.4 Statistical analysis Proband phenotype and deduced allele frequencies were compared with controls in 2 X 2 tables; RR values were
3.1.1 RFLP association The 9.3-kb PG4 allele (IGHPG4"Hl) showed a frequency of 0.57 vs. 0.41 in controls (pc = 0.0334) (Table 2). This variation is solely due to the increase in PG4*H1 homozygous subjects (RR = 3.01, p < 0.05) (Table 3). Significant associations were also found for the GP*Al, PG2*H1 and G2*A1 homozygotes, heterozygotes showing again no increase.
Five heterozygotes for gene deletions were found among the 50 IgGCdeficient subjects. Proband NEG is hemizygous for the G4 gene only, carrying a deletion already reported in the literature (211. BER is carrying a deletion encompassing the genes from A1 to E similar to that described in other subjects [21-241. ROG carries a new type of deletion encompassing only the G2 and G4 genes (unpublished). In family MAR (see pedigree in Fig. l ) , Southern analysis demonstrated that the proband's father (subject 1-1) lacks GP-G2-GCassociated fragments (Fig. 2); analysis with the A and E probes (not shown) made it clear that the father is a compound heterozygote for one deletion spanning from GP to A2 and another from A1 to E (both already described [21-261); two normal sons (11-1,2) are heterozygous for the first deletion, the proband (11-3) for the second. In family TES (see pedigree in Fig. l ) , Southern analysis showed that all nine sibs are heterozygous for a deletion extending from A1 to E genes (Fig. 3). Here, too, one parent was most probably a deletion homozygote.
Table 2. Allele frequency of IGHC RFLP in normal and IgG4-deficient subjects Locus
IGHGP IGHPG2
Allele
(kb)
*A1 *A2 *H1 *H2 *H3
(9.7) (8.5) (3.2) (3.35) (3.4) (12.5) (22) (9.3) (9.5)
IGHG2
*A1
IGHPG4
*A2 *H1 *H2
Blood donors n Freq
52 45
44 0 52 50
46 55 41
.54 .46 .46
.OO .54 .52 .48 .57 .43
Controls n Freq
x2"'
124 66 52 1 137 126
8.68 8.68 8.94 N.S.b) 8.37 4.87 4.87 5.73 5.73
64 78 112
.35 .65 .27 .01 .72 .66 .34 .41 .59
PC
0.0064
0.0084 0.0546 0.0334
a) Yates' corrected, from 2 X 2 contingency tables of allele numbers. b) N.S. = not significant (p > 0.05).
Eur. J. Immunol. 1992. 22: 227-233
IgG4 deficiency
229
Table 3. IGHC phenotypes and IgG4 deficiency Locus
GP
9.7 9.7i8.5 8.5 3.2 3.2/Xe) x1x 22 22112.5 12.5 9.3 9.319.5 9.5
PG2 G2 PG4
bc
bd
Controls
n
%
observed n %
19 16 15 15 16 19 16 16 18 22 13 15
(38) (32) (30) (30) (32) (38) (32) (32) (36) (44) (26) (30)
10 46 39 8 36 51 11 42 42 17 44 34
HW expected n %
(10.5) (48.4) (41.1) (8.4) (37.9) (53.7) (11.6) (44.2) (44.2) (17.9) (46.3) (35.8)
11.64 43.22 40.14 6.93 37.45 50.63 10.98 42.64 41.38 15.97 45.96 33.07
(12.2) (45.5) (42.2) (7.3) (39.4) (53.3) (11.6) (44.9) (43.6) (16.8) (48.4) (34.8)
4.34 0.99 1.OO 5.81 1.14 1.00 3.36 0.87 1.00 3.01 0.63 1.00
6.18**) N.S.d) 0.00 8.82*** N.S. 0.00 4.23* N.S. 0.00 3.68* N.S. 0.00
~191. IJ) (In RR)*/Variance [19]. c) Significance: * = p < 0.05; ** =p 0.00s). e) X = All other alleles
3.3 Combined HLADGHC analysis
ad
yw yw
a c a d bc bc
bd ad a c b d a c
xw x z
xw yz y 2 yw xw
yw
XZ
The distribution of the joint HLA and IGHC genotypes is cross-tabulated in a 3 x 3 contingency table (Table 6). For F
TES
MAR
* ;T*5b $ ;J '
I
0
G
I
T
N l
C
v
l
) 1
~ ~ n t = i
probe
-22
4-
'2
1
a) RR for each genotype vs. X/X; as control genotype frequencies Hardy-Weinberg expectations were used
#*A*
I
XZ
Deficient probands
Phenotype
$$%I;;:$
p/
GI
b c bc ad
xw y z
ac
ad
ac
bc
yw yz
xw
xw
xw
Ig' K E SOM
8; 5
PES
CA F
I
G3-
GP-
I
II
l&;x,h k E
% B,d,
family MAR
Figure 2. Southern blot analysis of the IGHC genes in the MAR family. Note the lack of the G2, GP an G4 bands in subject 1-1. Band intensity and segregation analysis confirm that the three children are heterozygous for the same deletion. G 1 and G3 bands are normal.
DIM
Figure I . Pedigree of multiple-case families. The proband is indicated with an arrow. a, b, c, d = IGHC haplotypes; x, y. w, z = HLA haplotypes. Full shadow = lack of IgG4; half shadow < 10 kg/ml. In family MAR and TES all the siblings are heterozygous for deletions involving the G4 gene.
3.2 HLA The 50 probands were analyzed for HLA class I1 RFLP. Significant associations with the D R B l l (0.38 vs. 0.202, p c = 0.0228), DQB3b (0.46 vs.0.257, p c = 0.004) and DQA3b (0.41 vs. 0.225, p c = 0.0077) alleles were found. These alleles together form the HLA-Dw5 (25) haplotype (Table 4). As for the IGHC locus the association is higher for homozygotes; here, however, the frequency of heterozygotes is also significantly increased (Table 5 ) .
F N c y ? y w ~ - ~
~
~
H
H
H
pSPG4 PG4 PGI
~
W
W
-
A
0.6
-0.3
-
-7.7
-4.66
PG3-
-
-3.6 -3.6 3.4
PD< PG2
