Genotyping Microarray for the Detection of More Than 200 CFTR Mutations in Ethnically Diverse Populations

Abstract

Cystic fibrosis (CF), which is due to mutations in the cystic fibrosis transmembrane conductance regulator gene, is a common life-shortening disease. Although CF occurs with the highest incidence in Caucasians, it also occurs in other ethnicities with variable frequency. Recent national guidelines suggest that all couples contemplating pregnancy should be informed of molecular screening for CF carrier status for purposes of genetic counseling. Commercially available CF carrier screening panels offer a limited panel of mutations, however, making them insufficiently sensitive for certain groups within an ethnically diverse population. This discrepancy is even more pronounced when such carrier screening panels are used for diagnostic purposes. By means of arrayed primer extension technology, we have designed a genotyping microarray with 204 probe sites for CF transmembrane conductance regulator gene mutation detection. The arrayed primer extension array, based on a platform technology for disease detection with multiple applications, is a robust, cost-effective, and easily modifiable assay suitable for CF carrier screening and disease detection.

Cystic fibrosis (CF) is a severe, common autosomal recessive disease due to mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene (OMIM number *602421; http://www.ncbi.nlm.nih.gov/Omim/). Asymptomatic carrier parents, who have no physiological or biochemical outcome that enables routine identification, typically have one CFTR mutation; whereas diseased progeny carry at least two mutations, one on each CFTR gene allele. CF has a high incidence in people of Northern European descent, occurring in approximately 1 in 2500 live births. Because of this, the American College of Medical Genetics and the American College of Obstetricians and Gynecologists (ACMG and ACOG) recently called for CF mutation carrier screening for expecting couples and for those contemplating pregnancy in high-risk groups (Northern European and Ashkenazi Jewish populations) as well as for proper counseling about diagnostic availability and limitations of screening in other groups with lower disease risk.1 By use of the inclusion criterion of mutations having a threshold of 0.1% frequency in the general U.S. population, a routine screening panel of 25 mutations was initially selected from the more than 1300 currently described CFTR sequence variants (http://www.genet.sickkids.on.ca/cftr/). This recommended panel was recently modified and currently includes 23 mutations.2 The selected panel chiefly reflects the mutations prevalent in the high-risk populations of Northern European and Askenazi Jewish descent. Mutations that are prevalent in other races and ethnicities are mostly not included in the routine CF 25 mutation panel because they do not reach inclusion criteria for the general population threshold.

Although development of commercial versions of the recommended screening panel (usually augmented by one to six mutations) has led to on-site CFTR mutation carrier screening, there are serious limitations of the currently available tests for widespread use, chiefly the detection rate, which depends on ethnicity. The risk of CF mutation carrier status varies by ethnicity, with the highest risk in people of Northern European ancestry at 1 in 25 and in people of Ashkenazi Jewish descent at 1 in 29.1,2 Other populations have discernible risk, however, including Hispanic Americans (1 in 46), African Americans (1 in 65), and Asian Americans (∼1 in 90). The latter three estimates may be lower than the actual risk in these ethnicities because of ascertainment bias due to the misperception that CF is a “Caucasian disease.” In addition, some ethnic groups (eg, Asians) have not, as yet, been studied at the molecular level as thoroughly as the Caucasian populations. Thus, whereas the current on-site tests can achieve a detection rate of 90% in Northern European Caucasians with CFTR carrier status, the ability to detect CF mutation carrier status in other populations falls off considerably, with the estimated detection rates of 69% in African Americans and 57% in Hispanic CF carriers.1

Currently, if the on-site carrier screening with the recommended panel of CFTR mutations is negative and if there is a family history or anxiety is high, DNA samples may be sent to reference laboratories for more comprehensive analysis that involves significantly higher costs. For non-Caucasians, carrier screening by a larger commercial panel is often preferentially chosen by counselors over the ACMG/ACOG recommended panel because of detection limitations. For diagnostic purposes in suspected CF patients, a carrier screening panel is sometimes used; but in those instances, second tier testing is frequently necessary. This second tier of more comprehensive testing generally involves either larger mutation panels or scanning methodologies, such as differential gradient gel electrophoresis, denaturing high-performance liquid chromatography, temporal temperature gradient gel electrophoresis, or single-strand conformation analysis, followed by direct DNA sequencing to characterize the mutations identified by scanning techniques.3 The gold standard of direct DNA gene sequencing is currently too costly for routine diagnostic use.

In light of the ethnic diversity of the U.S. population and the frequency of ethnically mixed marriages (http://www.census.gov/), we believe that a significantly expanded CF mutation panel, suitable for on-site use, would be beneficial as an option for routine CF mutation carrier screening. This may especially be desirable in non-Caucasians and individuals of mixed ethnicity. We describe a molecular diagnostic assay, using a new use of the arrayed primer extension (APEX) technology4 able to detect 204 different CFTR mutations that include a large number of mutations detected in non-Caucasians as well as those that are frequent in the Caucasian population and in the U.S. population as a whole. The assay is based on single-primer nucleotide extension, first described by Shumaker et al5 in 1996 and subsequently converted to an array format.4 It is a methodology that enables accurate mutation detection in a microarray format. Our CF APEX microarray described herein is suitable for carrier screening as well as molecular diagnosis of affected individuals.

Materials and Methods

Mutation Selection

The 204 mutations on the APEX microarray were selected from the CF Genetic Analysis Consortium (1994) (http//www.genet.sickkids.on.ca) and the literature,6,7,8,9 representing the most frequently screened mutations in Caucasians and those identified as recurring in specific Caucasian and non-Caucasian populations. The full set of mutations is listed in Table 1. The sequence numbering in this manuscript is according to the CFTR GenBank reference sequence NM_000492 (http://www.ncbi.nlm.nih.gov/GenBank/).9

Table 1.

Complete List of Mutations Detectable with the CF APEX Assay

	CFTR location	Amino acid change	Nucleotide change
1	E 1	Frameshift	175delC
2	E 2,3	Frameshift	del E2, E3
3	E 2	W19C	189 G>T
4	E 2	Q39X	247 C>T
5	IVS 2	Possible splicing defect	296 + 12 T>C
6	E 3	Frameshift	359insT
7	E 3	Frameshift	394delTT
8	E 3	W57X (TAG)	302G>A
9	E 3	W57X (TGA)	303G>A
10	E 3	E60X	310G>T
11	E 3	P67L	332C>T
12	E 3	R74Q	353G>A
13	E 3	R75X	355C>T
14	E 3	G85E	386G>A
15	E 3	G91R	403G>A
16	IVS 3	Splicing defect	405 + 1G>A
17	IVS 3	Possible splicing defect	405 + 3A>C
18	IVS 3	Splicing defect	406 − 1G>A
19	E 4	E92X	406G>T
20	E 4	E92K	406G>A
21	E 4	Q98R	425A>G
22	E 4	Q98P	425A>C
23	E 4	Frameshift	444delA
24	E 4	Frameshift	457TAT>G
25	E 4	R117C	481C>T
26	E 4	R117H	482G>A
27	E 4	R117P	482G>C
28	E 4	R117L	482G>T
29	E 4	Y122X	498T>A
30	E 4	Frameshift	574delA
31	E 4	I148T	575T>C
32	E 4	Splicing defect	621G>A
33	IVS 4	Splicing defect	621 + 1G>T
34	IVS 4	Splicing defect	621 + 3A>G
35	E 5	Frameshift	624delT
36	E 5	Frameshift	663delT
37	E 5	G178R	664G>A
38	E 5	Q179K	667C>A
39	IVS 5	Splicing defect	711 + 1G>T
40	IVS 5	Splicing defect	711 + 1G>A
41	IVS 5	Splicing defect	712 − 1G>T
42	E 6a	H199Y	727C>T
43	E 6a	P205S	745C>T
44	E 6a	L206W	749T>G
45	E 6a	Q220X	790C>T
46	E 6b	Frameshift	935delA
47	E 6b	Frameshift	936delTA
48	E 6b	N287Y	991A>T
49	IVS 6b	Splicing defect	1002 − 3T>G
50	E 7	ΔF311	3-bp del between nucleotides 1059 and 1069
51	E 7	Frameshift	1078delT
52	E 7	Frameshift	1119delA
53	E 7	G330X	1120G>T
54	E 7	R334W	1132C>T
55	E 7	I336K	1139T>A
56	E 7	T338I	1145C>T
57	E 7	Frameshift	1154insTC
58	E 7	Frameshift	1161delC
59	E 7	L346P	1169T>C
60	E 7	R347H	1172G>A
61	E 7	R347P	1172G>C
62	E 7	R347L	1172G>T
63	E 7	R352Q	1187G>A
64	E 7	Q359K/T360K	1207C>A and 1211C>A
65	E 7	S364P	1222T>C
66	E 8	Frameshift	1259insA
67	E 8	W401X (TAG)	1334G>A
68	E 8	W401X (TGA)	1335G>A
69	IVS 8	Splicing changes	1342 − 6 poly(T) variants 5T/7T/9T
70	IVS 8	Splicing defect	1342 − 2A>C
71	E 9	A455E	1496C>A
72	E 9	Frameshift	1504delG
73	E 10	G480C	1570G>T
74	E 10	Q493X	1609C>T
75	E 10	Frameshift	1609delCA
76	E 10	ΔI507	3-bp del between nucleotides 1648 and 1653
77	E 10	ΔF508	3-bp del between nucleotides 1652 and 1655
78	E 10	Frameshift	1677delTA
79	E 10	V520F	1690G>T
80	E 10	C524X	1704C>A
81	IVS 10	Possible splicing defect	1717 − 8G>A
82	IVS 10	Splicing defect	1717 − 1G>A
83	E 11	G542X	1756G>T
84	E 11	G551D	1784G>A
85	E 11	Frameshift	1784delG
86	E 11	S549R (A>C)	1777A>C
87	E 11	S549I	1778G>T
88	E 11	S549N	1778G>A
89	E 11	S549R (T>G)	1779T>G
90	E 11	Q552X	1786C>T
91	E 11	R553X	1789C>T
92	E 11	R553G	1789C>G
93	E 11	R553Q	1790G>A
94	E 11	L558S	1805T>C
95	E 11	A559T	1807G>A
96	E 11	R560T	1811G>C
97	E 11	R560K	1811G>A
98	IVS 11	Splicing defect	1811 + 1.6 kb A>G
99	IVS 11	Splicing defect	1812 − 1G>A
100	E 12	Y563D	1819T>G
101	E 12	Y563N	1819T>A
102	E 12	Frameshift	1833delT
103	E 12	D572N	1846G>A
104	E 12	P574H	1853C>A
105	E 12	T582R	1877C>G
106	E 12	E585X	1885G>T
107	IVS 12	Splicing defect	1898 + 5G>T
108	IVS 12	Splicing defect	1898 + 1G>A
109	IVS 12	Splicing defect	1898 + 1G>C
110	IVS 12	Splicing defect	1898 + 1G>T
111	E 13	Frameshift	1924del7
112	E 13	del of 28 amino acids	1949del84
113	E 13	I618T	1985T>C
114	E 13	Frameshift	2183AA>G
115	E 13	Frameshift	2043delG
116	E 13	Frameshift	2055del9>A
117	E 13	D648V	2075T>A
118	E 13	Frameshift	2105–2117 del13insAGAA
119	E 13	Frameshift	2108delA
120	E 13	R668C	2134C>T
121	E 13	Frameshift	2143delT
122	E 13	Frameshift	2176insC
123	E 13	Frameshift	2184delA
124	E 13	Frameshift	2184insA
125	E 13	Q685X	2185C>T
126	E 13	R709X	2257C>T
127	E 13	K710X	2260A>T
128	E 13	Frameshift	2307insA
129	E 13	V754M	2392G>A
130	E 13	R764X	2422C>T
131	E 14a	W846X	2670G>A
132	E 14a	Frameshift	2734delGinsAT
133	E 14b	Frameshift	2766del8
134	IVS 14b	Splicing defect	2789 + 5G>A
135	IVS 14b	Splicing defect	2790 − 2A>G
136	E 15	Q890X	2800C>T
137	E 15	Frameshift	2869insG
138	E 15	S945L	2966C>T
139	E 15	Frameshift	2991del32
140	E 16	Splicing defect	3120G>A
141	IVS 16	Splicing defect	3120 + 1G>A
142	IVS 16	Splicing defect	3121 − 2A>G
143	IVS 16	Splicing defect	3121 − 2A>T
144	E 17a	Frameshift	3132delTG
145	E 17a	I1005R	3146T>G
146	E 17a	Frameshift	3171delC
147	E 17a	Frameshift	3171insC
148	E 17a	del V1022 and I1023	3199del6
149	E 17a	Splicing defect	3271delGG
150	IVS 17a	Possible splicing defect	3272 − 26A>G
151	E 17b	G1061R	3313G>C
152	E 17b	R1066C	3328C>T
153	E 17b	R1066S	3328C>A
154	E 17b	R1066H	3329G>A
155	E 17b	R1066L	3329G>T
156	E 17b	G1069R	3337G>A
157	E 17b	R1070Q	3341G>A
158	E 17b	R1070P	3341G>C
159	E 17b	L1077P	3362T>C
160	E 17b	W1089X	3398G>A
161	E 17b	Y1092X (TAA)	3408C>A
162	E 17b	Y1092X (TAG)	3408C>G
163	E 17b	L1093P	3410T>C
164	E 17b	W1098R	3424T>C
165	E 17b	Q1100P	3431A>C
166	E 17b	M1101K	3434T>A
167	E 17b	M1101R	3434T>G
168	IVS 17b	3500 − 2A>T	3500 − 2A>T
169	IVS 17b	Splicing defect	3500 − 2A>G
170	E 18	D1152H	3586G>C
171	E 19	R1158X	3604C>T
172	E 19	R1162X	3616C>T
173	E 19	Frameshift	3659delC
174	E 19	S1196X	3719C>G
175	E 19	S1196T	3719T>C
176	E 19	Frameshift and K1200E	3732delA and 3730A>G
177	E 19	Frameshift	3791delC
178	E 19	Frameshift	3821delT
179	E 19	S1235R	3837T>G
180	E 19	Q1238X	3844C>T
181	IVS 19	Possible splicing defect	3849 + 4A>G
182	IVS 19	Splicing defect	3849 + 10 kb C>T
183	IVS 19	Splicing defect	3850 − 1G>A
184	E 20	G1244E	3863G>A
185	E 20	G1244V	3863G>T
186	E 20	Frameshift	3876delA
187	E 20	G1249E	3878G>A
188	E 20	S1251N	3884G>A
189	E 20	T1252P	3886A>C
190	E 20	S1255X	3896C>A and 3739A>G in E19
191	E 20	S1255L	3896C>T
192	E 20	Frameshift	3905insT
193	E 20	D1270N	3940G>A
194	E 20	W1282R	3976T>C
195	E 20	W1282X	3978G>A
196	E 20	W1282C	3978G>T
197	E 20	R1283M	3980G>T
198	E 20	R1283K	3980G>A
199	IVS 20	Splicing defect	4005 + 1G>A
200	E 21	Frameshift	4010del4
201	E 21	Frameshift	4016insT
202	E 22	Inframe del E21	del E21
203	E 21	N1303K	4041C>G
204	E 24	Frameshift	4382delA

Oligonucleotide Microchips

Oligonucleotide primers were designed according to the wild-type CFTR gene sequence for both the sense and antisense directions. The 25-bp oligonucleotides with 6-carbon amino linkers at their 5′ end were obtained from MWG (Munich, Germany). Most scanning oligonucleotides were designed to scan 1 bp in the wild-type sequence, except in the case of deletions and insertions that have the same nucleotide in the 1-bp direction. In this case, we designed the oligo to extend further into the deletion or insertion to enable discrimination of the nucleotide change. For example: 5′-AGCCTGGCACCATTAAAGAAAATATCAT-3′ ΔF508 S; 5′-TTTCCTGGATTAT-GCCTGGCACCATTAAAGAAAATATCATCTTTGGTGTT-TCCTATGATGAATATAGATACAGA-3′; ΔF508 AS 3′-AA-CCACAAAGGATACTACTTATATC-5′ in which A repre-sents a deliberate mismatch to avoid strong secondary structure and CTT represents a deletion of three nucleotides. In case of the normal allele, we expect signals for the sense oligo in the cytosine (C) channel and for the antisense oligo in the adenine (A) channel (C/A). In case of the mutant allele, we will detect signals for the sense oligo in the thymine (T) channel. The signal corresponding to the antisense oligo will also appear in the T channel (T/T).

The microarray slides used for spotting the oligonu-cleotides have a dimension of 24 × 60 mm and are coated with 3-aminopropyl-trimethoxysilane plus 1,4-phenylenedi-isothiocyanate (Asper Biotech, Ltd., Tartu, Estonia). Primers were diluted to 50 μmol/L in 100 mmol/L carbonate buffer, pH 9.0, and spotted onto the activated surface with BioRad VersArray (BioRad Laboratories, Hercules, CA). The slides were blocked with 1% ammonia solution and stored at 4°C until needed. Washing steps with 95°C distilled water (TKA, Toshvin Analytical, Germany) and 100 mmol/L NaOH were performed before the APEX reactions to reduce the background fluorescence and to avoid rehybridization of unbound oligonucleotides to the APEX slide.

Each selected CFTR sequence variant is identified by two unique 25-mer oligonucleotides, one for the sense and one for the antisense strand, although for some mutations, three oligonucleotides are used. On the other hand, fewer than expected oligonucleotides are used in some cases. For example, when different nucleotide substitutions (with a maximum of three) occur at the same nucleotide position, the accurate sequence can be determined using the same pair of detection oligos (sense and antisense). The APEX primers for mutations 3121-2A>G and 3121-2A>T, as an example, are the same, where A represents the location of the mutation under interrogation: ACCAACATGTTTTCTTTGATCTTAC 3121-2A→G,T S; 5′-ACCAACATGTTTTCTTTGATCTTAC A GTTGTTATTAATTGTGATTGGAGCTATAG-3′; CAACAA-TAATTAACACTAACCTCGA 3121-2A→G,T AS. These variables result in a total of 379 oligonucleotides annealed to the microarray slide.

Genomic and Synthetic Template Samples

Where possible, native genomic DNA was collected. Thus, 51 patient or cell line samples with known mutations were evaluated on the chip (Table 2). The presence of mutations in commercially available samples (Coriell Cell repositories; http://locus.umdnj.edu/ccr/) was verified in the Molecular Pathology laboratory at Stanford Hospitals and Clinics. Additional samples were obtained from the Molecular Diagnostics Centre of United Laboratories at Tartu University Clinics. Some DNA samples were a generous gift from Dr. Milan Macek, Jr., at Charles University. This set of samples was anonymized. When it was not readily possible to obtain native genomic DNA samples containing the screened mutations, synthetic approximately 50-bp templates were designed according to the mutated CFTR sequence for both the sense and antisense directions and optimized for melting temperature (MWG). In this case, poly(T) tracts were designed at the 5′ end to minimize the possibility of the self-extensions and/or self-annealing of the synthetic templates.

Table 2.

Genomic DNA Samples Used for Mutation Evaluation on the APEX Array

Mutations validated with native DNA
CFTRdel 2,3 (21 kb)
394delTT
G85E
R75X
574delA
Y122X
R117C
R117H
621 + 1G>T
621 + 3A>G
711 + 1G>T
I336K
R334W
R347P
IVS8-5T
IVS8-7T
IVS8-9T
A455E
ΔF508
ΔI507
1677delTA
1717 − 1G>A
G542X
G551D
R553X
R560T
S549N
1898 + 1G>A
1898 + 1G>C
2183AA>G
2043delG
R668C
2143delT
2184delA
2184insA
2789 + 5G>A
S945L
3120 + 1G>A
I1005R
3272 − 26A>G
R1066C
G1069R
Y1092X (C>A)
3500 − 2A>T
R1158X
R1162X
3659delC
S1235R
3849 + 10 kb C>T
W1282X

Template Preparation

The CFTR gene was amplified from genomic DNA in 30 amplicons. The PCR reaction mixture (50 μl) was optimized with the following: 10× TaqDNA polymerase buffer; 2.5 mmol/L MgCl₂ (Naxo, Estonia); 0.25 mmol/L dNTP (MBI Fermentas, Vilnius, Lithuania) (20% fraction of dTTP was substituted with dUTP); and 10 pmol of primer stock, genomic DNA (approximately 80 ng), SMART-Taq Hot DNA polymerase (3 U) (Naxo, Estonia), and sterile deionized water. After amplification (MJ Research DNA Thermal Cycler; MJ Research, Inc., Waltham, MA), the amplification products were concentrated and purified using Jetquick spin columns (Genomed GmbH, Lohne, Germany). In a one-step reaction, the functional inactivation of the traces of unincorporated dNTPs was achieved by addition of shrimp alkaline phosphatase (Amersham Pharmacia Biotech, Inc., Milwaukee, WI), and fragmentation of the PCR product was achieved by the addition of thermolabile uracil N-glycosylase (Epicenter Technologies, Madison, WI) followed by heat treatment.4

APEX Reactions

The APEX mixture consisted of 32 μl of fragmented product, 5 U of Thermo Sequenase DNA polymerase (Amersham Pharmacia Biotech), 4 μl of Thermo Sequenase reaction buffer (260 mmol/L Tris-HCl, pH 9.5, and 65 mmol/L MgCl₂) (Amersham Pharmacia Biotech) and 1 μmol/L final concentration of each fluorescently labeled ddNTPs: Cy5-ddUTP, Cy3-ddCTP, Texas Red-ddATP, Fluorescein-ddGTP, (PerkinElmer Life Sciences, Wellesley, MA). The DNA was first denatured at 95°C for 10 minutes. The enzyme and the dyes were immediately added to the DNA mixture, and the whole mixture was applied to prewarmed slides (58°C). The reaction was allowed to proceed for 10 minutes at 58°C, followed by washing once with 0.3% Alconox (Alconox, Inc.) and twice for 90 seconds at 95°C with distilled water (TKA, Germany). A droplet of antibleaching reagent (AntiFade SlowFade; Molecular Probes Europe BV, Leiden, The Netherlands) was applied to the slides before imaging. This process is depicted schematically in Figure 1.

Figure 1.

Schematic depiction of the APEX reaction.

During assay development, a few of the oligonucleotides designed from the wild-type CFTR sequence failed to perform the APEX reaction. The chief reason for APEX primer failure is the formation of self-annealing secondary structures that fail to hybridize or facilitate self-priming and extension. To obviate this problem, we designed new versions of such primers by incorporating a mismatch or a modified nucleotide at the 5′ or internal part of the primer. Such changes can reduce primer self-complementarity without compromising hybridization and primer extension. In the case of secondary structures at the 3′ end, which is essential for template annealing and extension, some alternative versions with internal base substitutions can be attempted, but not all work. After our final design of the APEX primers, 184 sequence variants (out of 204 variants at 181 sites) were detected in both of the sense and antisense directions, 7 from only the sense strand (the antisense strand does not work reliably), and 13 from only the antisense strand (the sense strand does not work reliably).

Analysis

The APEX array had four points of data analysis: both the forward and reverse primers are spotted in duplicate. This vastly reduced the possibility of false-positive signal interpretation due to dust particles or other nonspecific reactions and allowed distinction between homo- and heterozygosity. The array images were captured by means of detector Genorama QuattroImager 003 (Asper Biotech) at 20-μm resolution. The device combines a total internal reflection fluorescence-based excitation mechanism with a charge-coupled device camera.4 Sequence variants were identified using Genorama 3.0 genotyping software.

Results

We selected 204 CFTR sequence variants for a comprehensive diagnostic panel to increase CF carrier and disease detection across all racial and ethnic groups (Table 1). Sequence variants selected were obtained from the CF Genetic Analysis Consortium (1994) that compiled information from the screening of 43,849 chromosomes and from studies of relatively small-size samples or those isolated to specific ethnic populations (http://www.genet.sickkids.on.ca/cgi-bin/WebObjects/MUTATION). The population frequencies of the included mutations vary considerably according to ethnicity and number of samples screened, but they represent the most common reported mutations across population groups to date. Mutations include several prominent in non-Caucasian backgrounds, including N1303K, 3849 + 10 kb, 2789 + 5G>A, 3876delA, 406-1G>A, R75X, 2055delG>A, and S549N, which are each prevalent in the Hispanic population (personal observation);7,10,11,12,13 3120 + 1G>A, which is prevalent in the African-American population; and 1898 + 5G>T, which is prevalent in the Chinese population (http://www.genet.sickkids.on.ca/cftr/). The mutations originate from 26 exons (exons 1 to 22, including exons 6a and 6b, 14a and 14b, and 17a and 17b, and exon 24) and from 15 introns (2, 3, 4, 5, 6b, 8, 10, 11, 12, 14b, 16, 17a, 17b, 19, and 20). They include single-nucleotide substitutions, technically the most easy to detect with the APEX reaction, as well as insertions, deletions, including the large 21-kb deletion, which removes exons 2 and 3, and even repeats, such as the 5T/7T/9T repeats important in congenital bilateral absence of the vas deferens (CBAVD). Sample DNA is amplified in 30 amplicons with 29 pairs of PCR primers (Table 3) encompassing the mutations, with PCR mixtures that include 20% substitution of dUTPs for dTTPs, allowing for later fragmentation with uracil N-glycosylase as described previously.4

Table 3.

PCR Primers Used for Amplification of the Coding Sequence and Splice Sites in the CFTR Gene

Primer name	Sequence 5′ → 3′	Reference no.
1 F	TCTTTGGCATTAGGAGCTTG
1 R	CAAACCCAACCCATACACAC
IVS 1 F	GTTGTAAATCCTGGACCTGCAGAAG
IVS 1 R	CCCTCCTCTGATTCCACAAGGTAT
2 F	TCCATATGCCAGAAAAGTTG
2 R	ACAAGCATGCACTACCATTC
3 F	CTTGGGTTAATCTCCTTGGAT
3 R	CACCTATTCACCAGATTTCG
IVS 3 F	CCTTAGCAATTTGTATGAGCCCAA
IVS 3 R	CCATCATAGGATACAATGAATGCTGG
4 F	CCACTGTTGCTATAACAAATCCCAA
4 R	TTCAGCATTTATCCCTTACTTGTACC
5 F	ATTTCTGCCTAGATGCTGGG	9
5 R	AACTCCGCCTTTCCAGTTGT	9
6a F	GCTCAGAACCACGAAGTGTT
6a R	CATCATCATTCTCCCTAGCC
6b F	GGCACATAGGAGGCATTTACCAAA
6b R	GTCACAAACATCAAATATGAGGTGGA
7 F	GACCATGCTCAGATCTTCCATTCC
7 R	CCACTCTCATCCATCATACTGTCCA
8 F	AGTGGGTAATTCAGGGTTGCTTTG
8 R	CACCTGGCCATTCCTCTACTTCTT
9 F	GGCCATGTGCTTTTCAAAC
9 R	CTCCAAAAATACCTTCCAGCAC
10 F	TGTGCCCTTCTCTGTGAACCTCTA
10 R	TGATCCATTCACAGTAGCTTACCCA
11 F	TGTGGTTAAAGCAATAGTGTGA
11 R	ACAGCAAATGCTTGCTAGAC
IVS 11 F	TGTAGCTCTTGCATACTGTCTTC
IVS 11 R	AGTGCTGCCACAACTGTATAA
12 F	CTTCAGCAGTTCTTGGAATGTTGTG
12 R	GGAAGTAATCTTGAATCCTGGCCC
13 F	ATGCTAAAATACGAGACATATTGC
13 R	ACATGCTACATATTGCATTCTACTC
14a F	CATTATTCTGGCTATAGAATGACA
14a R	TGTATACATCCCCAAACTATCTTA
14b F	GCATGGGAGGAATAGGTGAA
14b R	TGCTTGGGAGAAATGAAACA
15 F	GTGCAGCTCCTGCAGTTTCTAAAG
15 R	ACCAACAAAACCACAGGCCCTA
16 F	GCAGGATGAGTACCCACCTATTCCT
16 R	GAAAGATGGTCCTTTGTGCCTCTC
17a F	CACTGACACACTTTGTCCACTTTGC
17a R	GAATGTCCTGTACACCAACTGTGG
17b F	TATTCAAAGAATGGCACCAGTGTG
17b R	GACAATCTGTGTGCATCGGTTTT
18 F	GTGCCCTAGGAGAAGTGTGA
18 R	GACAGATACACAGTGACCCTCA
19 F	CCTGTTAGTTCATTGAAAAGCCCG
19 R	TTCTGCTAACACATTGCTTCAGGC
IVS 19 F	GATTTCTGGAGACCACAAGG
IVS 19 R	ACCTCCTCCCTGAGAATGTT
20 F	ATCTTCCACTGGTGACAGGA
20 R	CTGCCTATGAGAAAACTGCAC
21 F	AATGTTCACAAGGGACTCCA	9
21 R	CCTGTTGCTCCAGGTATGTT
24 F	TGCCTTCTGTCCCAGATCTCACTA
24 R	TCAGTGTCCTCAATTCCCCTTACC

The general applicability of a proposed diagnostic will be determined by its clinical relevance, ease of use, low cost, and number of samples that can be processed per run. In general, the entire APEX reaction process, from PCR amplification (2 hours), PCR product purification (20 minutes), DNA fragmentation (1 hour), APEX reaction (15 minutes), to visualization of results (6 minutes) and analysis of results (10–15 minutes) can be accomplished in 4 to 5 hours. The time for sample preparation will be minimized in the future with the development of automated sample preparation and analysis. Because all steps leading up to the visualization of results on the QuattroImager can be performed in parallel, the rate-limiting factor in sample analysis will be determined by the number of imagers per test site. At 6 to 7 minutes analysis time per array, the throughput in 8 hours would be approximately 60 to 70 patients per imager unit.

The results allow reliable and reproducible detection of the wild-type (WT) versus mutation sequence at each array position on the APEX CF microarray. The assay is suitable both for screening of CF carriers (one heterozygous mutation in entire array) and for diagnosis of patients (two mutations, either heterozygous at two array sites or homozygous at one array site). Representative results are presented in Figures 2and 3. Figure 2 demonstrates results from three patient samples, one each of normal, heterozygous, and homozygous, at the grid position for mutation 2183AA>G, located in exon 13 (R domain). This mutation has a frequency of 2% in a screened sample of 2006 European probands (http://www.genet.sickkids.on.ca/cgi-bin/WebObjects/MUTATION). Figure 3 shows the results from three patient samples at the grid position for the common mutation ΔF508, again one each for normal, heterozygous, and homozygous at that position. In each case, the results accurately detect the sequence of both alleles for each patient sample. The 5T/7T/9T poly(T) tract polymorphism in IVS 8, of which the 5T variant is particularly important in CBAVD, is the most technically difficult mutation to accurately identify. Differentiation of the poly(T) length requires three sets of APEX primers for accurate detection. Representative results for three patient samples (5T/7T, 7T/9T, and 9T/9T) are presented in Figure 4. The first set of primers can distinguish between 5T and 7T/9T. The second set can distinguish between 5T and 9T. The third set of primers can distinguish between 7T and 9T. The first two of the three primer pairs present a technical difficulty previously alluded to, ie, despite several attempts to redesign and optimize the allele-specific primer pairs, an APEX oligo does not necessarily function reliably in both directions. Specifically the sense strand for the first primer pair and the antisense strand for the second primer pair are unreliable, and results from these must be disregarded. The antisense strand for the first pair and the sense strand for the second pair provide reliable identification of the probed base sequence, however, in each case. By testing at all three allele-specific oligo pairs for the IVS8 5T/7T/9T site, the length of the T repeat can be accurately and reproducibly determined.

Figure 2.

APEX analysis at mutation site 2183AA>G. Each numbered row represents the analysis of an individual patient sample. The row presents two sets of four-channel fluorescent images representing the bases adenine (A), cytosine (C), guanine (G), and thymine (T), respectively for the sense strand (top) and antisense strand (bottom). The histograms to the right of the fluorescent images are of the fluorescent intensities of the four channels at the mutation analysis site. The letters to the right of the histogram represent the base(s) identified on each strand. Row 3 presents the results of heterozygous target DNA derived from a CF patient (WT/2183AA>G). In this case, the sense strand is extended by both the wild-type (WT) complementary target sequence base A and the base G complementary for the mutation, whereas the antisense strand is extended by the WT base T and the base C complementary for the mutation. Row 4 contains the results of normal DNA derived from a non-CF individual at the target sequence (WT/WT), with the expected WT base A in the sense channel and WT base T in the antisense channel. Row 5 contains the results of a homozygous target DNA derived from a CF patient (2183AA>G/2183AA>G), with the base G complementary for the mutation in the sense channel and the base C complementary for the mutation in the antisense channel.

Figure 3.

APEX analysis at mutation site ΔF508. Three patient samples, one each with ΔF508/ΔF508, WT/ΔF508, and WT/WT, are presented. The results are presented as described in Figure 1. Row 3 (top) contains the results of homozygous target DNA derived from a CF patient (ΔF508/ΔF508). In this case, both the sense and the antisense strands are extended by the base T complementary in both strands for the mutation. Row 10 (center) contains the results of heterozygous target DNA from a CF patient (WT/ΔF508). In this case, the sense strand is extended by both the WT base C and base T complementary for the mutation, whereas the antisense strand is extended by the WT base A and the base T complementary for the mutation. Row 11 contains the results from a non-CF individual (WT/WT), in which the sense strand is extended by the WT base C, and the antisense strand is extended by the WT base A.

Figure 4.

APEX analysis at mutation site IVS8-5T/7T/9T. Representative results for three patient samples are shown at mutation site IVS8-5T/7T/9T. This challenging mutation site requires three pairs of allele-specific primers for accurate identification. The first set of primers (A) consists of a sense strand that does not work reliably despite several iterations and thus should be discounted and an antisense strand predicted to give base C for 5T (–/C) and base A for either 7T or 9T (–/A). The second set of primers (B) consists of a sense strand that elongates with a C only for 9T and an antisense strand that extends only with a C only for 5T. Thus the expected results for this set of primers is 5T (–/C), 7T (–/–), and 9T (C/–). The third set of primers consists of a sense strand oligo that extends with A for 7T and T for 9T, whereas the antisense strand extends with C for 7T and A for 9T. Thus the expected set of results for the third set of primers is 5T (–/–), 7T (A/C), and 9T (T/A). Adding the three sets of results together, patient sample 30 can be identified as compound heterozygous 5T/7T, patient sample 31 as compound heterozygous 7T/9T, and patient sample 32 as homozygous 9T/9T.

The CF APEX microarray was validated in a pilot study by means of 51 patient samples (Table 2) with different CFTR mutations. Mutation sites in which relevant patient samples could not readily be obtained were tested by means of 136 synthetic primers (Table 4), designed as approximately 50-mer oligonucleotides based on the wild-type sequence but incorporating the mutation to be identified. Four sites were tested with patient samples and synthetic template DNA with highly comparable results. With this pilot validation series, sensitivity (TP/TP+FN) was 100%. No false negatives were observed. Specificity (TN/TN + FP) was 100% and no false positives were present. The APEX reactions are entirely reproducible under our testing conditions. Testing at each site occurred from 3 to 20 times with identical results, as long as the following requirements have been met: 1) good quality of the DNA sample; 2) clean PCR products, identified as a single band of expected size on agarose gels; 3) complete fragmentation of the uracil N-glycosylase-treated PCR product; and 4) optimization and validation of the arrayed oligos. The reported results constitute a pilot demonstration of the effectiveness of the APEX approach in CF mutation detection. Complete validation of the CF APEX chip, including true sensitivity and specificity, will require systematic evaluation of patient samples with mutations at each of the array sites, and studies performed in a blinded manner.

Table 4.

Primers Generated to Create Synthetic Templates That Serve As Positive Mutation Controls

Primer name	Sense strand 5′ → 3′	Name	Antisense strand 5′ → 3′
175delC synt F	T(15)ATTTTTTTCAGGTGAGAAGGTGGCCA	175delC synt R	T(15)ATTTGGAGACAACGCTGGCCTTTTCC
W19C synt F	T(15)TACCAGACCAATTTTGAGGAAAGGAT	W19C synt R	T(15)ACAGCTAAAATAAAGAGAGGAGGAAC
Q39X synt F	T(15)TAAATCCCTTCTGTTGATTCTGCTGA	Q39X synt R	T(15)AGTATATGTCTGACAATTCCAGGCGC
296 + 12T>C synt F	T(15)CACATTGTTTAGTTGAAGAGAGAAAT	296 + 12T>C synt R	T(15)GCATGAACATACCTTTCCAATTTTTC
359insT synt F	T(15)TTTTTTTCTGGAGATTTATGTTCTAT	359insT synt R	T(15)AAAAAAACATCGCCGAAGGGCATTAA
E60X synt F	T(15)TAGCTGGCTTCAAAGAAAAATCCTAA	E60X synt R	T(15)ATCTATCCCATTCTCTGCAAAAGAAT
P67L synt F	T(15)TTAAACTCATTAATGCCCTTCGGCGA	P67L synt R	T(15)AGATTTTTCTTTGAAGCCAGCTCTCT
R74Q synt F	T(15)AGCGATGTTTTTTCTGGAGATTTATG	R74Q synt R	T(15)TGAAGGGCATTAATGAGTTTAGGATT
R75X synt F	T(15)TGATGTTTTTTCTGGAGATTTATGTT	R75X synt R	T(15)ACCGAAGGGCATTAATGAGTTTAGGA
W57X(TAG) synt F	T(15)AGGATAGAGAGCTGGCTTCAAAGAAA	W57X(TAG) synt R	T(15)TATTCTCTGCAAAAGAATAAAAAGTG
W57X(TGA) synt F	T(15)AGATAGAGAGCTGGCTTCAAAGAAAA	W57X(TGA) synt R	T(15)TCATTCTCTGCAAAAGAATAAAAAGT
G91R synt F	T(15)AGGGTAAGGATCTCATTTGTACATTC	G91R synt R	T(15)TTAAATATAAAAAGATTCCATAGAAC
405 + 1G>A synt F	T(15)ATAAGGATCTCATTTGTACATTCATT	405 + 1G>A synt R	T(15)TCCCTAAATATAAAAAGATTCCATAG
405 + 3A>C synt F	T(15)CAGGATCTCATTTGTACATTCATTAT	405 + 3A>C synt R	T(15)GACCCCTAAATATAAAAAGATTCCAT
406 − 1G>A synt F	T(15)AGAAGTCACCAAAGCAGTACAGCCTC	406 − 1G>A synt R	T(15)TTACAAAAGGGGAAAAACAGAGAAAT
E92X synt F	T(15)TAAGTCACCAAAGCAGTACAGCCTCT	E92X synt R	T(15)ACTACAAAAGGGGAAAAACAGAGAAA
E92K synt F	T(15)AAAGTCACCAAAGCAGTACAGCCTCT	E92K synt R	T(15)TCTACAAAAGGGGAAAAACAGAGAAA
444delA synt F	T(15)GATCATAGCTTCCTATGACCCGGATA	444delA synt R	T(15)ATCTTCCCAGTAAGAGAGGCTGTACT
574delA synt F	T(15)CTTGGAATGCAGATGAGAATAGCTAT	574delA synt R	T(15)AGTGATGAAGGCCAAAAATGGCTGGG
621G>A synt F	T(15)AGTAATACTTCCTTGCACAGGCCCCA	621G>A synt R	T(15)TTTCTTATAAATCAAACTAAACATAG
Q98P synt F	T(15)CGCCTCTCTTACTGGGAAGAATCATA	Q98P synt R	T(15)GGTACTGCTTTGGTGACTTCCTACAA
457TAT>G synt F	T(15)GGACCCGGATAACAAGGAGGAACGCT	457TAT>G synt R	T(15)CGGAAGCTATGATTCTTCCCAGTAAG
I148T synt F	T(15)CTGGAATGCAGATGAGAATAGCTATG	I148T synt R	T(15)GTGTGATGAAGGCCAAAAATGGCTGG
624delT synt F	T(15)CTTAAAGCTGTCAAGCCGTGTTCTAG	624delT synt R	T(15)TAAGTCTAAAAGAAAAATGGAAAGTT
663delT synt F	T(15)ATGGACAACTTGTTAGTCTCCTTTCC	663delT synt R	T(15)CATACTTATTTTATCTAGAACACGGC
G178R synt F	T(15)AGACAACTTGTTAGTCTCCTTTCCAA	G178R synt R	T(15)TAATACTTATTTTATCTAGAACACGG
Q179K synt F	T(15)AAACTTGTTAGTCTCCTTTCCAACAA	Q179K synt R	T(15)TTCCAATACTTATTTTATCTAGAACA
711 + 5G>A synt F	T(15)ATACCTATTGATTTAATCTTTTAGGC	711 + 5G>A synt R	T(15)TTATACTTCATCAAATTTGTTCAGGT
712 − 1G>T synt F	T(15)TGGACTTGCATTGGCACATTTCGTGT	712 − 1G>T synt R	T(15)TATGGAAAATAAAAGCACAGCAAAAAC
H199Y synt F	T(15)TATTTCGTGTGGATCGCTCCTTTGCA	H199Y synt R	T(15)TATGCCAATGCTAGTCCCTGGAAAATA
P205S synt F	T(15)TCTTTGCAAGTGGCACTCCTCATGGG	P205S synt R	T(15)TAAGCGATCCACACGAAATGTGCCAAT
L206W synt F	T(15)GGCAAGTGGCACTCCTCATGGGGCTA	L206W synt R	T(15)TCAAGGAGCGATCCACACGAAATGTGC
Q220X synt F	T(15)TAGGCGTCTGCTTTCTGTGGACTTGG	Q220X synt R	T(15)TATAACAACTCCCAGATTAGCCCCATG
936delTA synt F	T(15)AATCCAATCTGTTAAGGCATACTGCT	936delTA synt R	T(15)TGATTTTCAATCATTTCTGAGGTAATC
935delA synt F	T(15)GAAATATCCAATCTGTTAAGGCATAC	935delA synt R	T(15)TATTTCAATCATTTCTGAGGTAATCAC
N287Y synt F	T(15)TACTTAAGACAGTAAGTTGTTCCAAT	N287Y synt R	T(15)TATTCAATCATTTTTTCCATTGCTTCT
1002 − 3T>G synt F	T(15)GAGAACAGAACTGAAACTGACTCGGA	1002 − 3T>G synt R	T(15)TCTAAAAAACAATAACAATAAAATTCA
1154insTC syntwt F	T(15)ATCTCATTCTGCATTGTTCTGCGCAT	1154insTC syntwt R	T(15)TTGAGATGGTGGTGAATATTTTCCGGA
1154insTC syntmt F	T(15)TCTCTCATTCTGCATTGTTCTGCGCAT	1154insTC syntmt R	T(15)TAGAGATGGTGGTGAATATTTTCCGGA
DF311 mt syntV1 F	T(15)CCTTCTTCTCAGGGTTCTTTGTGGTG	dF311 mt syntV1 R	T(15)GAGAAGAAGGCTGAGCTATTGAAGTATC
G330X synt F	T(15)TGAATCATCCTCCGGAAAATATTCAC	G330X synt R	T(15)ATTTGATTAGTGCATAGGGAAGCACA
S364P synt F	T(15)CCTCTTGGAGCAATAAACAAAATACA	S364P synt R	T(15)GGTCATACCATGTTTGTACAGCCCAG
Q359K/T360K mt synt F	T(15)AAAAAATGGTATGACTCTCTTGGAGC	Q359K/T360K mt synt R	T(15)TTTTTTACAGCCCAGGGAAATTGCCG
1078delT synt F	T(15)CTTGTGGTGTTTTTATCTGTGCTTCC	1078delT synt R	T(15)CAAGAACCCTGAGAAGAAGAAGGCTG
1119delA synt F	T(15)CAAGGAATCATCCTCCGGAAAATATT	1119delA synt R	T(15)CTTGATTAGTGCATAGGGAAGCACAG
1161delC synt F	T(15)GATTGTTCTGCGCATGGCGGTCACTC	1161delC synt R	T(15)TCAGAATGAGATGGTGGTGAATATTT
T338I synt F	T(15)TCACCATCTCATTCTGCATTGTTCTG	T338I synt R	T(15)ATGAATATTTTCCGGAGGATGATTCC
R352Q synt F	T(15)AGCAATTTCCCTGGGCTGTACAAACA	R352Q synt R	T(15)TGAGTGACCGCCATGCGCAGAACAAT
L346P synt F	T(15)CGCGCATGGCGGTCACTCGGCAATTT	L346P synt R	T(15)GGAACAATGCAGAATGAGATGGTGGT
1259insA synt F	T(15)AAAAAGCAAGAATATAAGACATTGGA	1259insA synt R	T(15)TTTTTGTAAGAAATCCTATTTATAAA
W401X(TAG)mtsynt F	T(15)AGGAGGAGGTCAGAATTTTTAAAAAA	W401X(TAG)mtsynt R	T(15)TAGAAGGCTGTTACATTCTCCATCAC
W401X(TGA) synt F	T(15)AGAGGAGGTCAGAATTTTTAAAAAAT	W401X(TGA) synt R	T(15)TCAGAAGGCTGTTACATTCTCCATCA
1342 − 2A>C synt F	T(15)CGGGATTTGGGGAATTATTTGAGAAA	1342 − 2A>C synt R	T(15)GGTTAAAAAAACACACACACACACAC
1504delG synt F	T(15)TGATCCACTGTAGCAGGCAAGGTAGT	1504delG synt R	T(15)TCAGCAACCGCCAACAACTGTCCTCT
G480C synt F	T(15)TGTAAAATTAAGCACAGTGGAAGAAT	G480C synt R	T(15)ACTCTGAAGGCTCCAGTTCTCCCATA
C524X synt F	T(15)ACAACTAGAAGAGGTAAGAAACTATG	C524X synt R	T(15)TCATGCTTTGATGACGCTTCTGTATC
V520F synt F	T(15)TTCATCAAAGCAAGCCAACTAGAAGA	V520F synt R	T(15)AGCTTCTGTATCTATATTCATCATAG
1609delCA synt F	T(15)TGTTTTCCTGGATTATGCCTGGCACC	1609delCA synt R	T(15)CAGAACAGAATGAAATTCTTCCACTG
1717 − 8G>A synt F	T(15)AGTAATAGGACATCTCCAAGTTTGCA	1717 − 8G>A synt R	T(15)TAAAAATAGAAAATTAGAGAGTCACT
1784delG synt F	T(15)AGTCAACGAGCAAGAATTTCTTTAGC	1784delG synt R	T(15)ACTCCACTCAGTGTGATTCCACCTTC
A559T synt F	T(15)ACAAGGTGAATAACTAATTATTGGTC	A559T synt R	T(15)TTAAAGAAATTCTTGCTCGTTGACCT
Q552X synt F	T(15)TAACGAGCAAGAATTTCTTTAGCAAG	Q552X synt R	T(15)AACCTCCACTCAGTGTGATTCCACCT
S549R(A>C) synt F	T(15)CGTGGAGGTCAACGAGCAAGAATTTC	S549R(A>C) synt R	T(15)GCAGTGTGATTCTACCTTCTCCAAGA
S549R(T>G) synt F	T(15)GGGAGGTCAACGAGCAAGTATTTC	S549R(T>G) synt R	T(15)CCTCAGTGTGATTCCACCTTCTCCAA
L558S synt F	T(15)CAGCAAGGTGAATAACTAATTATTGG	L558S synt R	T(15)GAAGAAATTCTCGCTCGTTGACCTCC
1811 + 1.6 kb A>G synt F	T(15)GTAAGTAAGGTTACTATCAATCACAC	1811 + 1.6 kb A>G synt R	T(15)CATCTCAAGTACATAGGATTCTCTGT
1812 − 1G>A synt F	T(15)AAGCAGTATACAAAGATGCTGATTTG	1812 − 1G>A synt R	T(15)TTAAAAAGAAAATGGAAATTAAATTA
D572N synt F	T(15)AACTCTCCTTTTGGATACCTAGATGT	D572N synt R	T(15)TTAATAAATACAAATCAGCATCTTTG
P574H synt F	T(15)ATTTTGGATACCTAGATGTTTTAACA	P574H synt R	T(15)TGAGAGTCTAATAAATACAAATCAGC
1833delT synt F	T(15)ATTGTATTTATTAGACTCTCCTTTTG	1833delT synt R	T(15)CAATCAGCATCTTTGTATACTGCTCT
Y563D synt F	T(15)GACAAAGATGCTGATTTGTATTTATT	Y563D synt R	T(15)CTACTGCTCTAAAAAGAAAATGGAAA
T582R synt F	T(15)GAGAAAAAGAAATATTTGAAAGGTAT	T582R synt R	T(15)CTTAAAACATCTAGGTATCCAAAAGG
E585X synt F	T(15)TAAATATTTGAAAGGTATGTTCTTTG	E585X synt R	T(15)ATTTTTCTGTTAAAACATCTAGGTAT
1898 + 5G>T synt F	T(15)TTTCTTTGAATACCTTACTTATATTG	1898 + 5G>T synt R	T(15)AATACCTTTCAAATATTTCTTTTTCT
1924del7 synt F	T(15)CAGGATTTTGGTCACTTCTAAAATGG	1924del7 synt R	T(15)CTGTTAGCCATCAGTTTACAGACACA
2055del9>A synt F	T(15)ACATGGGATGTGATTCTTTCGACCAA	2055del9>A synt R	T(15)TCTAAAGTCTGGCTGTAGATTTTGGA
D648V synt F	T(15)TTTCTTTCGACCAATTTAGTGCAGAA	D648V synt R	T(15)ACACATCCCATGAGTTTTGAGCTAAA
K710X synt F	T(15)TAATTTTCCATTGTGCAAAAGACTCC	K710X synt R	T(15)ATCGTATAGAGTTGATTGGATTGAGA
I618T synt F	T(15)CTTTGCATGAAGGTAGCAGCTATTTT	I618T synt R	T(15)GTTAATATTTTGTCAGCTTTCTTTAA
R764X synt F	T(15)TGAAGGAGGCAGTCTGTCCTGAACCT	R764X synt R	T(15)ATGCCTGAAGCGTGGGGCCAGTGCTG
Q685X synt F	T(15)TAATCTTTTAAACAGACTGGAGAGTT	Q685X synt R	T(15)ATTTTTTTGTTTCTGTCCAGGAGACA
R709X synt F	T(15)TGAAAATTTTCCATTGTGCAAAAGAC	R709X synt R	T(15)ATATAGAGTTGATTGGATTGAGAATA
V754M synt F	T(15)ATGATCAGCACTGGCCCCACGCTTCA	V754M synt R	T(15)TGCTGATGCGAGGCAGTATCGCCTCT
1949del84 synt F	T(15)AAAAATCTACAGCCAGACTTTATCTC	1949del84 synt R	T(15)TTTTTAGAAGTGACCAAAATCCTAGT
2108delA synt F	T(15)GAATTCAATCCTAACTGAGACCTTAC	2108delA synt R	T(15)ATTCTTCTTTCTGCACTAAATTGGTC
2176insC synt F	T(15)CCAAAAAAACAATCTTTTAAACAGACTGGAGAG	2176insC synt R	T(15)GGTTTCTGTCCAGGAGACAGGAGCAT
2184delA synt F	T(15)CAAAAAACAATCTTTTAAACAGACTGG	2184delA synt R	T(15)GTTTTTTGTTTCTGTCCAGGAGACAG
2105–2117 del13 synt F	T(15)AAACTGAGACCTTACACCGTTTCTCA	2105–2117 del13 synt R	T(15)TTTCTTTCTGCACTAAATTGGTCGAA
2307insA synt F	T(15)AAAGAGGATTCTGATGAGCCTTTAGA	2307insA synt R	T(15)TTTCGATGCCATTCATTTGTAAGGGA
W846X synt F	T(15)AAACACATACCTTCGATATATTACTGTCCAC	W846X synt R	T(15)TCATGTAGTCACTGCTGGTATGCTCT
2734G/AT synt F	T(15)TTAATTTTTCTGGCAGAGGTAAGAAT	2734G/AT synt R	T(15)TTAAGCACCAAATTAGCACAAAAATT
2766del8 synt F	T(15)GGTGGCTCCTTGGAAAGTGAGTATTC	2766del8 synt R	T(15)CACCAAAGAAGCAGCCACCTGGAATGG
2790 − 2A>G synt F	T(15)GGCACTCCTCTTCAAGACAAAGGGAA	2790 − 2A>G synt R	T(15)CGTAAAGCAAATAGGAAATCGTTAAT
2991del32 synt F	T(15)TTCAACACGTCGAAAGCAGGTACTTT	2991del32 synt R	T(15)AAACATTTTGTGGTGTAAAATTTTCG
Q890X synt F	T(15)TAAGACAAAGGGAATAGTACTCATAG	Q890X synt R	T(15)AAAGAGGAGTGCTGTAAAGCAAATAG
2869insG synt F	T(15)GATTATGTGTTTTACATTTACGTGGG	2869insG synt R	T(15)CACGAACTGGTGCTGGTGATAATCAC
3120G>A synt F	T(15)AGTATGTAAAAATAAGTACCGTTAAG	3120G>A synt R	T(15)TTGGATGAAGTCAAATATGGTAAGAG
3121 − 2A>T synt F	T(15)TGTTGTTATTAATTGTGATTGGAGCT	3121 − 2A>T synt R	T(15)AGTAAGATCAAAGAAAACATGTTGGT
3132delTG synt F	T(15)TTGATTGGAGCCATAGCAGTTGTCGC	3132delTG synt R	T(15)AATTAATAACAACTGTAAGATCAAAG
3271delGG synt F	T(15)ATATGACAGTGAATGTGCGATACTCA	3271delGG synt R	T(15)ATTCAGATTCCAGTTGTTTGAGTTGC
3171delC synt F	T(15)ACCTACATCTTTGTTGCAACAGTGCC	3171delC synt R	T(15)AGGTTGTAAAACTGCGACAACTGCTA
3171insC synt F	T(15)CCCCTACATCTTTGTTGCTACAGTGC	3171insC synt R	T(15)GGGGTTGTAAAACTGCGACAACTGCT
3199del6 synt F	T(15)GAGTGGCTTTTATTATGTTGAGAGCATAT	3199del6 synt R	T(15)CCACTGGCACTGTTGCAACAAAGATG
M1101K synt F	T(15)AGAGAATAGAAATGATTTTTGTCATC	M1101K synt R	T(15)TTTTGGAACCAGCGCAGTGTTGACAG
G1061R synt F	T(15)CGACTATGGACACTTCGTGCCTTCGG	G1061R synt R	T(15)GTTTTAAGCTTGTAACAAGATGAGTG
R1066L synt F	T(15)TTGCCTTCGGACGGCAGCCTTACTTT	R1066L synt R	T(15)AGAAGTGTCCATAGTCCTTTTAAGCT
R1070P synt F	T(15)CGCAGCCTTACTTTGAAACTCTGTTC	R1070P synt R	T(15)GGTCCGAAGGCACGAAGTGTCCATAG
L1077P synt F	T(15)CGTTCCACAAAGCTCTGAATTTACAT	L1077P synt R	T(15)GGAGTTTCAAAGTAAGGCTGCCGTCC
W1089X synt F	T(15)AGTTCTTGTACCTGTCAACACTGCGC	W1089X synt R	T(15)TAGTTGGCAGTATGTAAATTCAGAGC
L1093P synt F	T(15)CGTCAACACTGCGCTGGTTCCAAATG	L1093P synt R	T(15)GGGTACAAGAACCAGTTGGCAGTATG
W1098R synt F	T(15)CGGTTCCAAATGAGAATAGAAATGAT	W1098R synt R	T(15)GGCGCAGTGTTGACAGGTACAAGAAC
Q1100P synt F	T(15)CAATGAGAATAGAAATGATTTTTGTC	Q1100P synt R	T(15)GGGAACCAGCGCAGTGTTGACAGGTA
D1152H synt F	T(15)CATGTGGATAGCTTGGTAAGTCTTAT	D1152H synt R	T(15)GTATGCTGGAGTTTACAGCCCACTGC
R1158X synt F	T(15)TGATCTGTGAGCCGAGTCTTTAAGTT	R1158X synt R	T(15)ACATCTGAAATAAAAATAACAACATT
S1196X synt F	T(15)GACACGTGAAGAAAGATGACATCTGG	S1196X synt R	T(15)CAATTCTCAATAATCATAACTTTCGA
3732delA synt F	T(15)GGAGATGACATCTGGCCCTCAGGGGG	3732delA synt R	T(15)CTCCTTCACGTGTGAATTCTCAATAA
3791delC synt F	T(15)AAGAAGGTGGAAATGCCATATTAGAG	3791delC synt R	T(15)TTGTATTTTGCTGTGAGATCTTTGAC
3821delT synt F	T(15)ATTCCTTCTCAATAAGTCCTGGCCAG	3821delT synt R	T(15)GAATGTTCTCTAATATGGCATTTCCA
Q1238X synt F	T(15)TAGAGGGTGAGATTTGAACACTGCTT	Q1238X synt R	T(15)AGCCAGGACTTATTGAGAAGGAAATG
S1255X (ex19)synt F	T(15)GTCTGGCCCTCAGGGGGCCAAATGAC	S1255X (ex19) synt R	T(15)CGTCATCTTTCTTCACGTGTGAATTC
S1255X;L synt F	T(15)AAGCTTTTTTGAGACTACTGAACACT	S1255X;L synt R	T(15)TATAACAAAGTAATCTTCCCTGATCC
3849 + 4A>G synt F	T(15)GGATTTGAACACTGCTTGCTTTGTTA	3849 + 4A>G synt R	T(15)CCACCCTCTGGCCAGGACTTATTGAG
3850 − 1G>A synt F	T(15)AGTGGGCCTCTTGGGAAGAACTGGAT	3850 − 1G>A synt R	T(15)TTATAAGGTAAAAGTGATGGGATCAC
3905insT synt F	T(15)TTTTTTTGAGACTACTGAACACTGAA	3905insT synt R	T(15)AAAAAAAGCTGATAACAAAGTACTCT
3876delA synt F	T(15)CGGGAAGAGTACTTTGTTATCAGCTT	3876delA synt R	T(15)CGATCCAGTTCTTCCCAAGAGGCCCA
G1244V synt F	T(15)TAAGAACTGGATCAGGGAAGAGTACT	G1244V synt R	T(15)ACCAAGAGGCCCACCTATAAGGTAAA
G1249E synt F	T(15)AGAAGAGTACTTTGTTATCAGCTTTT	G1249E synt R	T(15)TCTGATCCAGTTCTTCCCAAGAGGCC
S1251N synt F	T(15)ATACTTTGTTATCAGCTTTTTTGAGACTACTG	S1251N synt R	T(15)TTCTTCCCTGATCCAGTTCTTCCCAA
S1252P synt F	T(15)CCTTTGTTATCAGCTTTTTTGAGACT	S1252P synt R	T(15)GACTCTTCCCTGATCCAGTTCTTCCC
D1270N synt F	T(15)AATGGTGTGTCTTGGGATTCAATAAC	D1270N synt R	T(15)TGATCTGGATTTCTCCTTCAGTGTTC
W1282R synt F	T(15)CGGAGGAAAGCCTTTGGAGTGATACC	W1282R synt R	T(15)GCTGTTGCAAAGTTATTGAATCCCAA
R1283K synt F	T(15)AGAAAGCCTTTGGAGTGATACCACAG	R1283K synt R	T(15)TTCCACTGTTGCAAAGTTATTGAATC
4005 + 1G>A synt F	T(15)ATGAGCAAAAGGACTTAGCCAGAAAA	4005 + 1G>A synt R	T(15)TCTGTGGTATCACTCCAAAGGCTTTC
4010del4 synt F	T(15)GTATTTTTTCTGGAACATTTAGAAAAAACTTGG	4010del4 synt R	T(15)AAAATACTTTCTATAGCAAAAAAGAAAAGAAGAA
4016insT synt F	T(15)TTTTTTTCTGGAACATTTAGAAAAAACTTGG	4016insT synt R	T(15)AAAAAAATAAATACTTTCTATAGCAAAAAAGAAAAGAAGA
CFTRdele21 synt F	T(15)TAGGTAAGGCTGCTAACTGAAATGAT	CFTRdele21 synt R	T(15)CCTATAGCAAAAAAGAAAAGAAGAAGAAAGTATG
4382delA synt F	T(15)GAGAGAACAAAGTGCGGCAGTACGAT	4382delA synt R	T(15)CTCTATGACCTATGGAAATGGCTGTT

Bold, mutation allele of interest; bold and italicized, modified nucleotide. 

Discussion

The CF APEX array system presents a new technology for comprehensive screening for CFTR mutations, one that offers several advantages. It is an integrated system with DNA microarray chip, multiplex primer extension on the array, and fluorescence imaging and data analysis.4,14 Because of the unprecedented number and selection of mutations on the CFTR array, this CF test has a higher mutation detection capability than assays currently available on-site in diagnostic laboratories. It also more than doubles the number of offered mutations on currently available panels at reference laboratories. The inclusion of mutations that have a higher frequency in specific ethnic populations increases this mutation detection capability for patients affected with cystic fibrosis (who have two mutations), as well as unaffected mutation carriers (who have one mutation). The CF APEX test is specific, reproducible, and technically robust. Test costs (including labor, supplies, and capital equipment) will ultimately depend on the scale of production and deployment, but the costs are clearly within the range of current commercial molecular diagnostic tests. In addition, the APEX system in general is applicable to many other molecular diagnostic tests for genetic and infectious diseases. For example, it has recently been developed for detection of more than 400 mutations in the ABCR (ABCA4) gene associated with five distinct retinal phenotypes15 and for detection of β-thalassemia and G6PD mutations.16 Thus, it is a “platform” technology that in turn amortizes labor and capital costs.

One concern with any platform that allows screening for a large number of mutations, of which some are rare in the general population, is the approach to quality control. Ideally, a positive control is run for each mutation every time the assay is performed. However, even with the widely used commercial CF screening panels that enable testing for approximately 25 mutations, this approach is prohibitively expensive and impractical. Thus, most laboratories have resorted to the inclusion of only a fraction of the positive controls in each assay so that all controls are assessed on a rotating basis. With the CF APEX array, the inclusion of positive controls is even more challenging because of the large number of mutations on the array and the fact that the majority of these are not commercially available. To perform the assay in a clinical diagnostic setting, native and/or synthetic positive control samples could be pooled for an assessment of the entire mutation list in only a few runs. This, of course, will require optimization to achieve ideal conditions. False-positive samples are avoided through several checkpoints in the assay: 1) a negative sample void of DNA is included in each run; 2) every sample is automatically assessed in duplicate; and 3) every sample, when technically possible, is evaluated in both the forward and reverse direction for direct result confirmation. Because it may remain impractical to obtain genomic DNA for routine quality control of all mutations tested, we expect that the use of synthetic samples may remain necessary. Synthetics are dependent on the correctness of the original sequence reported in the literature and cannot reflect the performance of true test conditions, which include the possibility of genomic polymorphisms at or near the PCR primer sites and differences in PCR efficiency compared with genomic DNA. Thus, an initial full validation with genomic samples will be optimal.

Although the CF APEX array, including the 204 mutations presented herein, improves the mutation detection possibilities of first-line CFTR mutation screening, it is clear that many, if not most, mutations that contribute to CF in Hispanics, African Americans, and Asians remain unknown. It is essential that future research endeavors determine the prevalent mutations in each population, followed by inclusion of newly identified mutations on the CF APEX chip (or other diagnostic tests), so that appropriate genetic counseling and understanding of the pathogenesis and outcomes of CF can be achieved in all ethnic groups. Considering that more than 35 million Hispanic Americans reside in the United States and that this is the fastest growing segment of the population (http://www.census.gov/; http://www.dhs.ca.gov), as well as one with a relatively high carrier frequency, it is imperative that more comprehensive screening routinely be provided. In addition, delayed diagnosis in children with CF can result in more severe health problems and can adversely affect prognosis. Diagnostic delay is more frequently observed in ethnic minorities.17

In addition to CF carrier screening, demand for molecular CFTR gene testing has been growing for another reason as well. It is increasingly recognized that mutations in the CFTR gene play a role in a variety of diseases other than classic CF, including CBAVD,18 chronic pancreatitis,19 disseminated bronchiectasis,20 and allergic pulmonary aspergillosis.21 The CF APEX array can serve as the initial research tool to screen for mutations in these diseases, coupled with other more comprehensive but labor- and capital-intensive technologies if results are negative. As other important disease-contributive CFTR mutations are identified, these can be added to the CF APEX array for routine analysis.