GENOMICS

10,201-206

(1991)

Microdeletions within 22qll Associated with Sporadic and Familial DiGeorge Syndrome PETER 1. SCAMBLER,* *Department

ALISOUN H. CAREY,* RICHARD K. H. Wm,t SHERRY ROACH,**’ MAGNUS NORDENSKJOLD,$ AND ROBERT WILLIAMSON*

of Biochemistry and Molecular Genetics, St. Mary’s Hospital Medical School, London W2 IPG, United Kingdom; t Department of Paediatric Cardiology, Institute of Child Health, London WC1 N 1EH, United Kingdom; and *Department of Clinical Genetics, Karolinska Institute, S-104 01, Stockholm, Sweden ReceivedAugust31,

1990;

revised

willing to risk the birth of a severely affected child. o 1991 Press, Inc.

INTRODUCTION The DiGeorge syndrome (DGS; 188400; McKusick, 1990) is characterized by absent/hypoplastic thymus, absent/hypoplastic parathyroids, and heart defects, causing deficient cell-mediated immunity, hypocalcemic fits, and cardiovascular insufficiency; mental retardation and craniofacial dysmorphism are often additional features (Conley et al., 1979; Muller et aZ., 1988). The conotruncus, thymus, and parathyroids have a common embryonic origin, at about 4 weeks of gestation, in the third and fourth pharyngeal pouches. Abnormalities in the development of these structures coupled with the first branchial arch anomalies suggest a defect in neural crest interactions (Oster et al,

1 Current

address:

Green

College,

Oxford,

Decemberl8,

1990

1983; Bockman and Kirby, 1984; Lodewyk et al., 1986). DiGeorge syndrome usually occurs sporadically, although it can be inherited as an autosomal dominant (Rohn et al., 1984; Greenberg et al., 1984; Keppen et al., 1988). Cytogenetic analyses of several DGS patients have revealed chromosomal abnormalities in both sporadic and familial cases; among those patients with karyotypic abnormality, unbalanced translocations leading to monosomy for pter-22qll are particularly common (de la Chapelle et al., 1981; Greenberg et al., 1984, 1988). It is therefore thought that there is a major locus for a gene or group of genes involved in DGS at pter-22qll. Ring chromosome 22 has never been reported in association with DGS. Patients with DGS who have an interstitial deletion within 22qll have also been described. These observations map the disease locus to 22qll (Greenberg et al., 1988; Fibison et al., 1990). It is not known whether DGS can be caused by deletion of a single gene, or whether it is a contiguous gene syndrome in which differing extents of segmental aneuploidy account for phenotypic variability between cases. The Wilms tumor/aniridia (WAGR) complex is the best-characterized example of this group of diseases (Emanuel, 1988). There is evidence that the number of genes involved in DGS is small: there are no reports of a consistent subgrouping of signs as there is for the WAGR complex, and DGS has been described in association with an apparently balanced translocation disrupting 22qll (Augusseau et al., 1986). As a first step in the identification of the gene at the DGS locus (DGS), we have mapped 27 DNA markers to the region of pter-22qll deleted in patients with cytologically visible deletions and DGS (Carey et al., 1990). We describe here the molecular analysis of a group of patients with DGS and normal karyotypes and demonstrate that the dysmorphology can be as-

DiGeorge syndrome (DGS) is a developmental field defect of the third and fourth pharyngeal pouches. It is associated with deletion of 22qll in 11% of cases. Molecular genetic analysis with probes from 22qll-pter reveals that a subset of markers is hemizygous in DGS patients with normal karyotypes. There is no apparent difference in the phenotype or the severity of the disorder between patients with the smallest detectable submicroscopicdeletion and those with the largest cytogenetically visible abnormality. A microdeletion was found in a mildly affected child and in the severely affected child of a mildly affected father. Dysmorphology, especially cardiac outflow tract anomalies, resulting from 22qll deletion may be more common than currently realized since chromosomes are unlikely to he checked if the complete spectrum of DGS is not present. Antenatal diagnosis, through detection of hemizygosity at 22ql1, will be a possibility for mildly affected parents unAcademic

JAN P. DUMANSKI,~

UK. 201

All

Copyright 0 1991 rights of reproduction

OSSS-7543/91$3.00 by Academic Press, Inc. in any form reserved.

202

SCAMBLER

TABLE Clinical

Cell

line

Summary

Karyotype

GM03577 GM05401 GM03479 GM07215 GM05876 GM07939” Th GOS4” SM”

22qll22qllN N N N N N N

1

of Patient Material Dosage Analysis Heart

A A A A A A A A (mild) A (mild)

Parathyroid

Thymus

A NI A NI A A A A A (mild)

A A A A A A NI E A (mild)

Used

Facial A A A NI A A NI A E

in

Other + + + -

Note. The full karyotypes of GM03577 and 05401 are published in the NIGMS catalogue. “A” represents a clinically significant abnormality, e.g., major heart malformation, absence of thymus and parathyroids, or hypocalcemic fits and T-cell dysfunction. “N” represents normal. “E” indicates examined and normal. “NI” indicates that no information was available. +/indicates presence/absence of additional dysmorphology less specific for DGS. GM03479 also bad trachea-esophageal fistula, clinodactyly, arrhinencephaly, and hypoplastic isthmus of the thyroid gland. GM07939 has polydactyly and micropenis. Tb has EEG abnormalities. ’ See text for discussion of these patients.

sociated with submicroscopic deletions in patients with no apparent karyotypic abnormality. Analysis of the shortest region of overlap (SRO) between the cytogenetically detectable and undetectable deletions will help determine the precise location of the genes involved in DGS. MATERIALS

AND METHODS

Cell Lines Cell lines from patients with DGS with the prefix GM were obtained from the NIGMS Human Genetic Mutant Cell Repository (Camden, NJ); cytogenetic investigation of these patients has only been conducted at 450-band resolution. Clinical details and karyotypes are summarized in Table 1. GL5 was a gift from Dr. K. Huebner (Philadelphia, PA) and contains the translocation derivative (17qter-17p13 : : 22qll22qter) with no other chromosome 22 sequencespresent. This hybrid can be used to divide 22qll into proximal (GL5 negative) and distal (GL5 positive) regions.

DNA Preparations DNA was prepared from blood and/or cell lines and blood by standard methods (Kunkel et al., 1977).

Phage and Plasmid

Preparations

The DNA markers used in this study have been described previously (Carey et al., 1990). For the pur-

ET

AL.

poses of the current investigation we analyzed probes that were known to be duplicated in der22 syndrome patients, since we know DGS lies proximal to this breakpoint (Carey et al., 1990). [der22 syndrome patients have dysmorphology and trisomy for pter22qll (Lin et al., 1986).] Some of these markers are present in hybrid GL5, and therefore located more distally within 22qll; others are absent and are located more proximally on pter-22qll. One marker distal to the der22 syndrome breakpoint, KI-205 (D22S112), was analyzed as a chromosome 22 control. A probe from chromosome 7, XV-2c (D7S23), was used as a non-chromosome-22 control (Estivill et al., 1987). Established miniprep protocols were used for phage and plasmid DNA preparation (Sambrook et al., 1989). Inserts to be used as radiolabeled DNA probes were released by digestion with the appropriate restriction enzyme and purified on a 1% lowmelting-point agarosegel (Bethesda Research Laboratories).

Labeling Reactions Inserts were labeled using the random oligonucleotide priming procedure (Feinberg and Vogelstein, 1983) or using oligonucleotides specific for sequences flanking the vector insertion site (Carey et al., 1990).

Quantitative

Hybridization

Digestion, blotting, hybridization, washing, and autoradiographic techniques for quantitative hybridization were as reported previously (Carey et al., 1990). Scanning densitometry was accomplished using an LKB Ultrascan enhanced laser densitometer. The area under each deflection curve was used to calculate signal strength ratios; control probes hybridized to the same filters were used to correct for variations in DNA loading. Nylon membranes (Hybond N, Amersham) were rehybridized several times, the previous signal having been removed according to the manufacturer’s instructions. Each digestion, blotting, hybridization, and quantification was, wherever possible, performed in triplicate before the presence or absence of deletion was determined. Three signal intensities of ~~60% of control levels were taken as evidence of deletion. The test DNA:control DNA signal intensity ratios were corrected for loading differences as we have described previously (Carey et al., 1990) [the ratio of the test probe:control probe signal intensities in the patients is divided by the ratio of the test probe:control probe intensities in an appropriate normal control DNA (see Table 2)]. RESULTS AND DISCUSSION Twelve markers that map proximal to the der22 syndrome breakpoint were hybridized to DNA from

DIGEORGE

SYNDROME

AND

203

MICRODELETIONS

control

197

506 Pvu II 506

Pvull FIG, 1. Probes 197,506, and XV-2c (control) hybridized to the same PuuII-digested DNA from Th, GM05401, and a range of normal controls. Th is hemizygous with both probes (W-197 and KI506) (compare signal strengths of 197,506, and control probe in Th and control lane H2). See Table 2 for densitometry results and calculations.

patients with DGS and normal karyotype. Ten showed no evidence of hemizygosity in the patients studied. However, KI-197 (D22Slll) and KI-506 (D22S139) detected submicroscopic deletions in a subset of DGS cases with normal chromosomes. Figure 1 compares the hybridization signal obtained with KI-197, KI-506, and XV-2c (control) in patient Th. Both chromosome 22 markers detect hemizygosity; densitometry results are given in Table 2. Figure 2 shows the intensities of the hybridization signals obtained with KI-506 and control probe XVTABLE Densitometric

Th

506 197 Control

0.68 0.21 0.50

GM05401 2.02 0.52 0.65

2c in another karyotypically normal patient, GM07939; KI-506 detects hemizygosity in GM07939, but not in GM05401. Figure 3 demonstrates that only the locus detected by KI-506 is hemizygous in GM05876, whereas both KI-506 and KI-197 show hemizygosity in lines GM07215 and GM07939. Neither is hemizygous in GM03479. KI-716 (D22S144) is hemizygous in GM05401, but dizygous in GM07939 (data not shown). These data would place KI-716 proximal to DGS. We found that the seven other 22qll markers tested that were not present in the GL5 hybrid and hemizygous in patients with visible monosomy for pter-22qll were also dizygous in DGS patients with normal karyotypes (data not shown). This would place the majority of probes hemizygous

2

Analysis

of Fig. 1 Corrected signal intensity (% control)

Cell line Probe

FIG. 2. 506 and XV-2c hybridized to the same PuuII-digested DNA from GM05401 and GM07939. A microdeletion can be seen in GM07939, a cell line derived from the severely affected child of a mildly affected father (compare control and 506 hybridizations in control lane H4 and GM07939). The corrected signal strength ratio of 506 for GM07939/control is 0.42.

Control 1.44 0.40 0.54

(H2)

Th 51 57 -

GM05401

197

116 108 -

Note. The values given in the left half of the table refer to the area under the absorbance curve as calculated by the densitometer. Corrected signal intensities are calculated as in (14) and are given on the right. Compare the values for 506 in Th and control lane H2; the absolute values are 0.68 and 1.44, respectively. However, H2 is overloaded by 8% with respect to Th as demonstrated by the control probe (absorbances of 0.54 and 0.50). Therefore the Th test probe result needs adjusting to compensate for this, an 8% increase (of 0.68) giving a value of 0.73. Thus the ratio of corrected signal strengths for 506 in TH and control lane H2 is 0.73/1.44 = 51%. Both probes detect loci hemizygous in Th, but dizygous in GM05401.

506 FIG. 3. 506 and 197 hybridized to the same filter containing EcoRI-digested DNA from a range of patients with normal karyotypes. GM05878 is derived from an unaffected father who has a balanced 10;2 translocation; he has an affected child with an unbalanced form resulting in 22qll-. The DNA is overloaded in this lane; no rearrangement is seen. GM05876 has a microdeletion detectable with 506 but not 197 (compare control and cell line signal intensities for both probes in GM05876; the ratio of signal strength for these probes is 0.49).

204

SCAMBLER

ET

Chr. 22 AL.

GLS

3577

5401

7215 7939 l-h.

5876

SM GOS 4

716

506

Hind

III

FIG. 4. 506 hybridized to PuuII-digested DNA from GOS4 and a range of controls demonstrates a microdeletion in a mildly affected patient who may desire a family in the future (compare control and 506 hybridizations in control lane H2 and GOS4). The corrected signal strength ratio of 506 for GOSl/control is 0.48.

in patients with visible deletions proximal to DGS, as defined by the microdeletions. Figure 4 demonstrates hemizygosity for the locus detected by KI-506 in the mildly affected patient GOS4. Table 3 summarizes the results for the two probes that detect microdeletions and for one marker proximal and one marker distal to DGS. A schematic map of the region and the various deletion breakpoints are shown in Fig. 5.

DGS Is Associated with Submicroscopic of 22qll

Deletions

We have shown that DGS can be associated with deletions within 22qll that are too small to be detectable by cytogenetic analysis. Five of the six patients with “normal” chromosomes had a microdeletion. This microdeletion can extend into regions of 22qll not deleted in patients with visible karyotypic deletion of pter-22ql1, since KI-197 is dizygous in

TABLE

3

Quantitative Hybridization Analysis of DGS Patient DNA Patient/cell Locus

03577

KI-716 KI-506 KI-197 KI-205

H H D D

05401 H D D D

GOS4

03479

D H NI D

D D D D

line

05876

07215

D H D D

D H H D

07939

Th

SM

GL5

D H H D

D H H D

NI H’ NI D

+ +

Note. H indicates a hemizygous locus. D indicates a dizygous locus. - indicates not present in GL5. + indicates present in GL5. NI indicates no information. a On the basis of a single result.

7 II

U

FIG. 5. Order of DNA markers around Filled in areas denote deletions. All cell lines, as described in the text. The SRO is defined

DGS (not patients, by ++.

to scale). and loci are

GM03577 and GM05401, and KI-506 is dizygous in GM05401, but both markers detect deletions in patients with normal chromosomes. Molecular genetic evidence for deletions in patients with normal karyotypes has been reported in other diseases such as Prader-Willi syndrome and MillerDieker syndrome (Gregory et aZ., 1990; van Tuinen et al., 1988). In each of these casesthe detection of submicroscopic deletions is of clinical importance, since the use of DNA probes in addition to cytogenetic analyses increases diagnostic accuracy in specific cases.

Microdeletions

Can Occur in Mildly

Affected Patients

Patient GOS4 has a normal karyotype and a relatively mild manifestation of DGS. The patient is now 9 years of age and presented shortly after birth with hypocalcemic fits. She has low-set ears and a minor degree of mandibular hypoplasia. She has no history of recurrent infection and she had normal T-cell number and erythrocyte rosettes. Two-dimensional echocardiography was normal, but a barium swallow revealed an aberrant drainage of the right subclavian artery. Apart from an expressive speech problem (secondary to subglottic stenosis), her mental and physical development was normal and she has made good progress at school. A diagnosis of DGS was considered, but rejected. Figure 4 demonstrates a microdeletion of 22qll encompassing KI-506, confirming the diagnosis of DGS and indicating that microdeletions may be associated with mild disease. There are two reported instances of severely affected children being born to a mildly affected parent with normal karyotype. In one case we have been able

DIGEORGE

SYNDROME

to examine DNA from the severely affected child. GM07939 is a cell line from the second son of a mildly affected father (Rohn et aZ., 1984), who presumably has the same chromosome deletion as his son; Figs. 2 and 3 demonstrate the presence of a submicroscopic deletion in GM07939. The father is mildly dysmorphic (micrognathia, low-set ears), has persistent nonsymptomatic hypocalcemia, a T-cell count 59% of normal with normal mitogen stimulation, and no cardiovascular symptoms or signs. His other son, with DGS, died at 3 months following cardiac catheterization. We have also found preliminary evidence of 22qll microdeletion in the mildly affected father (SM) of a severely affected (deceased) proband (Keppen et aZ., 1988). His daughter had a type B interrupted aortic arch, aberrant origin of the right subclavian artery, patent ductus arteriosus, absent thymus, typical facial features, hypocalcemia, and Pseudomonas pneumonia. A high-resolution karyotype obtained at 9 days revealed no abnormalities; the patient died at 23 days. SM has a right-sided aortic arch (without clinical problems), a repaired cleft palate, a history of infantile hypocalcemia, recurrent oral candidiasis, and a normal karyotype. On testing, his T-cell number was 39% of normal (19% helper cells) (Keppen et al., 1988). Preliminary data (not shown) indicate that SM is hemizygous for KI-506, but this is an unconfirmed result because there was insufficient DNA for the duplicate experiments. DGS secondary to 22qll microdeletion can thus present with a wide range of severity; detection of submicroscopic deletions will allow antenatal diagnosis in cases like GOS4 where there may be the prospect of a severely affected child being born to a mildly affected parent. The birth defects found in DGS may also present in isolation or in patients without the complete constellation of DGS signs. For instance, 30% of cases of Fallot’s tetralogy have dysmorphic facies and 50% have severe or recurrent infections (Shimizu et al., 1984; Radford and Thong, 1989; Burn, 1989). On the basis of our observations of microdeletions in “forme fruste” DGS, we think it likely that examination of more patients at the molecular level will reveal that the incidence of 22qll deletion in cases with anomalies of the cardiac outflow tract is much higher than previously thought (e.g., if 15% of Fallot’s patients had deletions within 22ql1, then the incidence of monosomy 22qll would be l/10,000 live births). DiGeorge

Syndrome

as a Contiguous

Gene Defect

There is little support for the idea that the known clinical variability of DGS (Lammer and Opitz, 1986) is due to large differences in the extent of aneuploidy.

AND

205

MICRODELETIONS

Case GM05876, for which microdeletion was detected with one marker only, has the complete spectrum of abnormality associated with DGS with a severity similar to that of GM03577, which has the more extensive of the two visible deletions. We cannot rule out correlations between phenotype and extent of monosomy within microdeletions since the defects in GM07939 and GM7215 are more severe than those in GOS4 and SM, but if we assume that GM07939 and his father have the same microdeletion there is no need to invoke this phenomenon to explain our data. We do not conclude from these data that DGS is not a contiguous gene defect but would maintain that the region of 22qll within which these genes must lie has been substantially narrowed. No marker was found to be hemizygous in the severely affected patient GM03479. This may be because the deletion in this patient was smaller than that in the other cases with submicroscopic deletion or a point mutation. Alternatively, DGS in this case may be produced by a genetic abnormality at a different locus, e.g., lop (Greenberg et aZ., 1988). Flanking Markers for DGS

and a Shortest

Region of Overlap

On the basis of the data discussed above we conclude that both KI-506 and KI-197 map distal to DGS, with KI-506 the marker closer to DGS because it detects hemizygosity in GM05876 (see Fig. 5). Those markers that do not hybridize to GL5, which are hemizygous in GM05401 and GM03577 but dizygous in patients with submicroscopic deletions, will be proximal to DGS, e.g., KI-716. KI-716 is GL5 negative, hemizygous in both cases with visible deletions of pter-22ql1, and dizygous in all microdeletion patients tested. KI-716 is therefore proximal to DGS, making KI-716 and KI-506 flanking markers. At present, the distal border of the SRO for DGS is bounded by the distal border of the breakpoint of the deleted region in GM05401. The proximal border is the proximal breakpoint of the microdeletions (Fig. 5) which is as yet undefined; the breakpoints are shown at the same point in Fig. 5 because they cannot be distinguished, but it should be emphasized that there are no data supporting the hypothesis that they are coincident. We have no markers that map within this SRO, which could be more simply defined by a GM03479 microdeletion should one be detected. The variability in translocation breakpoints and deletion endpoints seen in DGS patients both with and without karyotypic abnormality emphasizes the high frequency with which 22qll rearrangements are found in various clinical abnormalities (McDermid et al., 1989). Coupled with a genetic map and pulse-field gel analysis, investigation of the variety of 22qll

206

SCAMBLER

rearrangements in DGS will aid in the precise and accurate mapping, and subsequent cloning, of DGS. The probes described here will be most useful in assessing the frequency of 22qll hemizygosity in a range of cardiac outflow tract abnormalities, such as tetralogy of Fallot, especially when they occur with other signs of neurocristopathy. It may be fruitful to search for evidence of hemizygosity in patients in whom there are single features of DGS but who do not show the spectrum of signs warranting the diagnosis of this anomaly. ACKNOWLEDGMENTS This work was supported by the Wellcome trust, the MRC! (UK), and the Swedish Cancer Society. We are indebted to Dr. B. White for p22/34 (D22S9), Dr. K. Huebner and Dr. A. G. van Kessel for providing cell hybrids, Dr. L. Keppen for blood from patient SM, and Dr. H. Punnett for clinical information concerning line GM03479. Note added in proof.Professor B. Emanuel has detected interstitial deletions in GM07939 and GM07212 using high-resolution cytogenetics. The microdeletion in SM has now been confirmed.

REFERENCES 1. AUGUSSEAU, S., JOUK, S., JALBERT, P., AND PRIJWR, M. (1986). DiGeorge syndrome and 22qll rearrangements. Hum. Genet.

74: 206.

2. BOCKMAN, D. E., AND KIRBY, M. L. (1984). Dependence of thymus development on derivatives of the neural crest. Science 223: 498-500. 3. BURN, J. (1989). The face and immune system in tetralogy of Fallot. Znt. J. Cardiol. 22: 237-239. 4. CAREY, A. H., ROACH, S., WILLIAMSON, R., DUMANSKI, J. P., NORDENSKOLD, M., COLLINS, V. P., ROULEAU, G., BLIN, N., JALBERT, P., AND SCAMBLER, P. J. (1990). Localization of 27 DNA markers to the region of human chromosome pter22qll deleted in patients with the DiGeorge syndrome and duplicated in the der22 syndrome. Gerwmics 7: 299-306. 5. CONLEY, M. E., BECKWITH, J. B., MANCER, J. K. F., AND TENCKHOFF, L. (1979). The spectrum of the DiGeorge syndrome. J. Pediatr. 94: 883-890. 6. DE LA CHAPELLE, A., HERVA, R., KOIVISTO, M., AND AULA, P. (1981). A deletion in chromosome 22 can cause DiGeorge syndrome. Hum. Genet. 67: 253-256. 7. EMANUEL, B. S. (1988). Molecular cytogenetics: Toward dissection of the contiguous gene syndromes. Am. J. Hum. Genet. 43: 575-578. 8. ESTIVILL, X., FARRALL, M., SCAMBLER, P. J., BELL, G. M., HAWLEY, K. M., LENCH, N. J., BATES, G. P., KRUYER, H. C., FREDERICK, P., STANIER, P., WATSON, E. K., WILLIAMSON, R., AND WAINWRIGHT, B. (1987). A candidate gene for the cystic fibrosis locus isolated by selection for methylation-free islands. Nature 326: 840-845. 9. FEINBERG, A. P., AND VOGELSTEIN, B. (1983). A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal. B&hem. 132: 6-13. 10. FIBISON, W. J., BUD-, M., MCDERMID, H., GREENBERG, F., AND EMANUEL, B. S. (1990). Molecular studies of DiGeorge syndrome. Am. J. Hum. Genet. 46: 888-895. 11. GREENBERG, F., CROWDER, W. E., PASCHALL, V., COLONLINARES, J-C., LUBIANSKI, B., AND LEDB~R, D. (1984).

ET AL.

12.

13.

14.

15.

16. 17.

Familial DiGeorge syndrome and associated partial monosomy of chromosome 22. Hum. Genet. 65: 317-319. GREENBERG, F., ELDER, F. F. B., HAFFNER, P., NORTHRUP, H., AND LEDBETTER, D. (1988). Cytogenetic findings in aprospective series of patients with DiGeorge anomaly. Hum. Genet. 43: 605-611. GREGORY, C. A., KIRKILIONIS, A. J., GREENBERG, C. R., CHUDLEY, A. E., AND HAMERTON, J. L. (1990). Detection of molecular rearrangements in Prader-Willi syndrome patients by using genomic probes recognizing four loci within the PWCR. Am. J. Med. Genet. 35: 536-545. KEPPEN, L. D., FASULES, J. W., BURKS, A. W., GOLLIN, S. M., SAWYER, J. R., AND MILLER, C. H. (1988). Confirmation of autosomal dominant transmission of the DiGeorge malformation complex. J. Pediatr. 113: 506-508. KUNKEL, L. M., SMITH, K. D., BOYER, S. H., BORGAONKAR, D. S., WACHTEL, S. S., MILLER, 0. J., BREG, W. R., JONES, H. E., AND RARY, J. M. (1977). Analysis of human Y-chromosome reiterated DNA in chromosome variants. Proc. Natl. Acad. Sci. USA 74: 1245-1249. LAMMER, E. J., AND OPITZ, J. M. (1986). The DiGeorge anomaly as a developmental field defect. Am. J. Med. Genet. (Suppl.) 2: 113-127. LIN, A. E., BERNAR, J., CHIN, A. J., SPARKES, R. S., EMANUEL, B. S., AND ZACHAI, E. H. (1986). Congenital heart disease in supernumerary der(22)t(11;22) syndrome. Clin. Genet. 29: 269-275.

18. LODEWYK, H., VAN MIEROP, S., AND KUTSCHE, L. M. (1986). Cardiovascular anomalies in DiGeorge syndrome and importance of neural crest as a possible pathogenic factor. Am. J. Cardiol. 58: 133-137. 19. MCDERMID, H. E., BUDARF, M. L., AND EMANUEL, B. S. (1989). Toward a long-range map of human chromosomal band 22qll. Genomics 5: l-8. 19a. MCKUSICK, V. A. (1990). “Mendelian Inheritance in Man,” 9th ed., Johns Hopkins Univ. Press, Baltimore. 20. MULLER, W., PETER, H. H., WILKEN, M., JUPPNER, H., KALLFELZ, H. C., KROHN, H. P., MILLER, K., AND REIGER, C. H. L. (1988). The DiGeorge syndrome. I. Clinical evaluation and course of partial and complete forms of the syndrome. Eur. J. Pediutr. 147: 496-502. 21. OSTER, G., KILBURN, K. H., AND SEAL, F. P. (1983). Chemically induced congenital thymic dysgenesis in the rat: A model of the DiGeorge syndrome. Clin. Zmmunol. Zmmurwpathol. 28: 128-134. 22. RADFORD, D. J., AND THONG, Y. H. (1989). Facial and immunological anomalies associated with tetralogy of Fallot. Znt. J. Cardiol.

22: 229-236.

23. ROHN, R. D., LEFFELL, M. S., LEADEM, P., JOHNSON, D., RuBIO, T., AND EMANUEL, B. (1984). Familial third-fourth pharyngeal pouch syndrome with apparent autosomal dominant transmission. J. Pediatr. 106: 47-51. 24. SAMBROOK, J., FRITSCH, E. F., AND MANIATIS, T. (1989). “Molecular Cloning: A Laboratory Manual,” Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 25. SHIMIZU, T., TAKAO, A., ANDo, M., AND HIRAYAMA, A. (1984). Conotruncal anomaly face syndrome: Its heterogeneity and association with thymus involution. In “Congenital Heart Disease, Causes and Processes” (J. J. Nora and A. Takao, Eds.), pp. 29-41, Futura, New York. 26. VAN TUINEN, P., DOBYNS, W. B., RICH, D. C., SUMMERS, K. M., ROBINSON, T. J., NAKAMIJRA, Y., ANIJ LEDBETTER, D. D. H. (1988). Molecular detection of microscopic and submicroscopic deletions associated with Miller-Dieker syndrome. Am. J. Med. Genet. 43: 587396.