Journal of Histochemistry and Cytochemistry Priciples for Free Access to Science
  Search:   
    >> Advanced Search

Guidelines | Subscriptions | About | exPRESS - Current - Archive | Business Information | Contact
DOI: 10.1369/jhc.4A6436.2005
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Vermeesch, J. R.
Right arrow Articles by Devriendt, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Vermeesch, J. R.
Right arrow Articles by Devriendt, K.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
Journal of Histochemistry and Cytochemistry
Volume 53 (3): 413-422, 2005
Copyright ©The Histochemical Society, Inc.

Molecular Karyotyping : Array CGH Quality Criteria for Constitutional Genetic Diagnosis

Joris R. Vermeesch, Cindy Melotte, Guy Froyen, Steven Van Vooren, Binita Dutta, Nicole Maas, Stefan Vermeulen, Björn Menten, Frank Speleman, Bart De Moor, Paul Van Hummelen, Peter Marynen, Jean-Pierre Fryns and Koen Devriendt

Center for Human Genetics, University Hospital Gasthuisberg, Leuven, Belgium (JRV,CM,NM,J-PF,KD); Center for Human Genetics, Flanders Interuniversity Institute for Biotechnology (VIB4), Department of Human Genetics (GF,PM), MicroArray Facility, Flanders Interuniversity Institute for Biotechnology (VIB) (BD,PVH), Leuven, Belgium; ESAT-SISTA K.U. Leuven (SVV,BDM), Leuven, Belgium; and Department of Medical Genetics, Ghent University, Ghent, Belgium (SV,BM,FS)

Correspondence to: J.R. Vermeesch, Center for Human Genetics, Herestraat 49, 3000 Leuven, Belgium. E-mail: Joris.Vermeesch{at}uz.kuleuven.ac.be


   Summary
Top
 Summary
Introduction
 Materials and Methods
 Results
 Discussion
 Literature Cited
 
Array CGH (comparative genomic hybridization) enables the identification of chromosomal copy number changes. The availability of clone sets covering the human genome opens the possibility for the widespread use of array CGH for both research and diagnostic purposes. In this manuscript we report on the parameters that were critical for successful implementation of the technology, assess quality criteria, and discuss the potential benefits and pitfalls of the technology for improved pre- and postnatal constitutional genetic diagnosis. We propose to name the genome-wide array CGH "molecular karyotyping," in analogy with conventional karyotyping that uses staining methods to visualize chromosomes. (J Histochem Cytochem 53:413–422, 2005)

Key Words: array CGH • molecular karyotyping • constitutional cytogenetics • prenatal diagnosis • postnatal diagnosis


   Introduction
Top
 Summary
 Introduction
Materials and Methods
 Results
 Discussion
 Literature Cited
 
PRE- AND POSTNATAL cytogenetic testing is aimed at the genome-wide analysis for the detection of chromosome aneuploidies and segmental aneusomies. Traditionally, karyotyping has performed this role. In addition to its clinical importance, the identification of a chromosomal aberration in specific patients has proven to be a successful way to identify the implicated genes and to gain insight in the pathogenesis of different genetic conditions.

Over the last decades, improved resolution of cytogenetic techniques has lead to a significant increase in the detection rate of chromosomal aberrations in patients with mental retardation (MR) and/or congenital anomalies. However, resolution of traditional cytogenetic techniques is limited to at best ~5 Megabases (Mb), and smaller chromosomal aberrations often remain hidden. The identification of submicroscopic subtelomeric alterations in 3–7% of idiopathic MR patients (Flint et al. 1995Go; Knight et al. 1999Go; Slavotinek et al. 1999Go; Riegel et al. 2001Go; Rosenberg et al. 2001Go), as well as the sporadic reports of submicroscopic interstitial chromosomal rearrangements, suggests that a substantial portion of idiopathic MR may be caused by smaller chromosomal rearrangements. These observations make it clear that higher resolution screening techniques for the detection of small deletions or duplications at any chromosomal position will drastically increase the elucidation of human genetic diseases.

Although fluorescence in situ hybridization (FISH) has dramatically increased the sensitivity of detection of genomic imbalances, this approach requires prior knowledge of the chromosomal region(s) of interest and therefore is not applicable for whole genome screening approaches required in a diagnostic setting.

More recently, so-called array or matrix CGH (comparative genomic hybridization) utilizes mapped DNA sequences in a microarray format as a platform for the detection of chromosomal deletions/duplications. In this technique, genomic DNA from the patient is labeled with one fluorescent dye while a normal reference sample is labeled with a different dye, and these samples are co-hybridized to the array containing the genomic DNA targets. Chromosomal imbalances across the genome can thus be quantified and positionally defined by analyzing the ratio of the fluorescence of the two dyes along the targets. The resolution of array CGH depends on the size of the genomic fragments as well as on their density. Proof of principle was established in 1997 (Solinas-Toldo et al. 1997Go; Pinkel et al. 1998Go). Since then, only few laboratories mastered the technology, mainly for the detection of chromosomal amplifications in cancers (Solinas-Toldo et al. 1997Go; Pollack et al. 1999Go; Albertson et al. 2000Go; Bruder et al. 2001Go). Chromosomal duplications or deletions were initially technically challenging to detect (Carter et al. 2002Go). Recently, several other groups have successfully increased the sensitivity to the point where single copy changes can reliably be detected, which has led to the first reports of genome-wide screens in both human (Pollack et al. 1999Go; Snijders et al. 2001Go, Vissers et al. 2003Go; Schaeffer et al. 2004Go; Shaw-Smith et al. 2004Go; Schoumans et al. 2004Go) and mouse (Hodgson et al. 2001Go). However, due to these technical challenges, the technique has not made the transition to clinical practice. In this manuscript we explore some of the technical aspects essential for optimal results in array CGH, and we define the quality criteria an array experiment should reach to be reliable in a clinical setting.


   Materials and Methods
Top
 Summary
 Introduction
 Materials and Methods
Results
 Discussion
 Literature Cited
 
Array CGH
Arrays were constructed using a 1 Mb Clone Set, which contains 3527 BAC and PAC clones, 3275 with known and fixed genomic position (Fiegler et al. 2003Go). DNA was isolated from 1 ml bacterial culture using Millipore genomic DNA purification kit following manufacturer's instructions (Millipore; Billerica, MA). BAC and PAC DNA obtained after purification was reconstituted in storage buffer supplied by the manufacturer. The BAC/PAC DNA was amplified by "degenerate oligonucleotide primed PCR" (DOP PCR) using a modification of the protocol described by Fiegler et al. (2003)Go. Preamplification was performed in a 50-µl reaction mixture containing PCR buffer [20 mM Tris HCl (pH 8.4), 50 mM KCl], 5 mM MgCl2, 200 µM dNTPs (Amersham Biosciences, Piscataway, NJ), 2 µM primer mix (5'-CCGACTCGAGNNNNNNCTAGAA-3'; 5'-CCGACTCGAGNNNNNNTAGGAG-3'; 5'-CCGACTCGAGNNNNNNTTCTAG-3'), 2.5 U Platinum Taq DNA polymerase (Invitrogen; Carlsbad, CA) and 2–10 ng template DNA. Reactions were performed using the GeneAmp PCR System 9700 thermocycler (PerkinElmer; Applied Biosystems, Foster City, CA) using the following conditions: 3 min denaturation at 94C followed by nine thermal cycles with denaturation for 1.5 min at 94C, annealing for 2.5 min at 30C, subsequently ramping with temperature increment of 0.1C/sec to 72C and a 3-min extension at 72C. Additionally, 30 cycles were performed consisting of 1 min at 94C, 1.5 min at 62C and 2 min at 72C, followed by a final elongation step at 72C for 8 min.

In a second round of amplification, an aminolinked primer was used. A 100-µl reaction was performed by combining 2 µl of the first reaction product with 1.5 µM primer (5'-NH2–GGAAACAGCCCGACTCGAG-3'), 200 µM dNTPs (Amersham Biosciences), PCR buffer, 5 mM MgCl2 and 2.5 U Platinum Taq DNA polymerase (Invitrogen). Thermal cycles were performed as follows: 10 min denaturation at 95C; 35 cycles of 1 min at 95C, 1.5 min at 60C and 7 min at 72C, and a final elongation at 72C for 10 min. Following amplification, the PCR products were purified by Qiaquick 96-well PCR purification kit following the manufacturers instruction (QIAGEN Inc.; Valencia, CA). The purified DOP PCR products were EtOH precipitated.

DOP PCR products were spotted on either CodeLink Bioarray System slides (Amersham Biosciences), type VII star silane-coated mirror slides (Amersham Biosciences) or UltraGAPS amino-silane-coated slides (Corning, Corning, NY). For spotting on amino-silane-coated slides, regular DOP PCR primers (without aminolinker) were used to amplify the BAC DNA (Van Buggenhout et al. 2004Go). Purified DOP PCR product was reconstituted in 20 µl of 80% DMSO solution containing nitrocellulose (0.37 µg/ml) at an average concentration of 250 ng/µl. The products were arrayed using a Molecular Dynamics Generation III printer (Amersham Biosciences). Before hybridization, the slides were humidified, dried, and target DNA was cross-linked in a UV-Stratalinker with 50 mJ UV light (Stratagene; Amsterdam, The Netherlands). For spotting on CodeLink Bioarray System slides, 5' aminolinked DOP PCR products were dissolved in 250 mM sodium phosphate buffer (pH 8.5) containing 0.001% sarkosyl. Products were spotted at a concentration of 200 ng/µl with a Molecular Dynamics Generation III printer. The clones were printed in duplicate. After printing, slides were pretreated by incubating them in a humidified chamber saturated with NaCl solution for 24 hr. The next day, slides were treated with 1% NH4OH solution on a shaker followed by 0.1% SDS for 5 min. The slides were rinsed in water, placed in 95C water bath for 2 min, and then at –20C for 1 min. The slides were rinsed again in H2O and finally dried by centrifugation. These slides were kept under dehydrated conditions until further use.

To test printing quality, a pretreated slide was hybridized with 70 pmol of Cy3-labeled oligo primer (5'-Cy3-GGAAACAGCCCGACTCGAG-3') for 1 hr. Following hybridization, the slide was washed with H2O and scanned to check printing quality.

Genomic DNA (gDNA) from an anonymous cell line with karyotype 46,XX or 46,XY was used as a reference. DNA was extracted from the blood of a trisomy-13 carrier following standard procedures. Test and reference gDNA were labeled by a random prime labeling system (Bioprime DNA Labeling System; Invitrogen) using Cy3- and Cy5-labeled dCTPs (Amersham Biosciences) as described (Van Buggenhout et al. 2004Go). The labeling efficiency was checked with the Nanodrop ND-1000 spectrophotometer (Nanodrop Technologies; Rockland, DE). DNA concentration was determined by measuring the intensities at 260 and 280 nm. Cy5 and Cy3 incorporation was measured at 650 and 550 nm, respectively. For each probe, the corresponding specific activity was calculated by the following formula: (total ng of probe x 103)/(molecular weight of nucleotide times the total pmol of dye incorporated). Except for some small modifications, probe preparation and preblocking of the slide were performed as described by Fiegler et al. (2003)Go. In short, equal amounts (200 pmol) of Cy5- and Cy3-labeled probe were combined with (if not otherwise indicated) 100 µg Cot-1 DNA followed by an ethanol precipitation. Resuspension of the pellet was done in hybridization buffer (50% formamide, 10% dextran sulfate, 0.1% Tween 20, 2x SSC, and 10 mM Tris HCl, pH 7.5) containing 400 µg yeast tRNA to hybridize a spotting area of 24 x 60 mm.

The slide was blocked with 50 µg Cot-1 DNA and 300 µg salmon sperm DNA dissolved in 60 µl hybridization buffer. Blocking solution and probe mixture were denaturated for 10 min in a water bath at 75C. Blocking solution was then placed on the slide covered with a coverslip (24 x 60 mm) and placed in a humid chamber. Meanwhile, the probe was placed at 37C for preannealing. After 1 hr of blocking, the coverslip was removed and probe was placed on the slide. After placing a coverslip (24 x 60 mm), the slide was placed in a humid chamber saturated with 20% formamide and 2x SSC. Hybridization was allowed to take place for two nights at 37C. While optimizing the protocol, it was noticed that the targets at the outer ends of the slide often showed reduced signal intensities when sealing the slide with rubber cement. Eliminating the sealing of the coverslips by hybridizing the slides in small humid chambers yielded equal intensity ratios over the entire (full) slide. Post-hybridization washes were performed as described by Fiegler et al. (2003)Go. In short, the coverslip was removed by placing the slide in PBS with 0.05% Tween 20, followed by a 10-min wash in a fresh solution of PBS/Tween 20 at room temperature, 30 min in 50% formamide/2x SSC at 42C, and 10 min in PBS/0.05% Tween 20 at room temperature. Slides were spin dried at 1000 rpm for 5 min.

Three types of slide were compared: amino-silane-coated slides on which the target DNA is cross-linked by UV light, CodeLink Bioarray System slides on which the DNA is covalently linked via the amino group of the DOP primer, and type VII star silane-coated mirror slides. When hybridized with 250 pmol probe each, the signal-to-noise ratios on amino-silane-coated and CodeLink Bioarray System slides were 6–10 and 20–30 for the Cy3 and Cy5, respectively. The mirror slides were hybridized with 200 pmol probe resulting in signal-to-noise ratios of, respectively, 5.3 and 2.7 for the Cy3 and Cy5. These values are derived from at least 10 hybridizations.

Image and Data Analysis
Arrays were scanned at 532 nm (Cy3) and 635 nm (Cy5) using the GenIII scanner (Amersham Biosciences) for the mirror slides or the Agilent G2565BA MicroArrayScanner System (Agilent Inc.; Palo Alto, CA) for the CodeLink and UltraGAP slides. Image analysis was done using ArrayVision (Imaging Research Inc.; St Catharines, Ontario, Canada). All further data analysis was performed with Excel (Microsoft Inc.; Diegem, Belgium). Spot intensities were corrected for local background, and only spots with signal intensities at least twofold above background signal intensities were included in the analysis. For each clone, a ratio of Cy5 to Cy3 fluorescent intensity was calculated. Normalization of the data was achieved by dividing the fluorescent intensity ratio of each spot by the mean of the ratios of the autosomes. Finally, the normalized ratio values of the duplicates were averaged and a log2 value was calculated. Datapoints for which the variation between the two intensity ratios was larger than 10% were excluded from the analysis.


   Results
Top
 Summary
 Introduction
 Materials and Methods
 Results
Discussion
 Literature Cited
 
Parameters Influencing Array CGH Quality
The two main technical challenges when performing array CGH are obtaining adequate signal-to-noise ratios and low standard deviations (SD) of the intensity ratios. In contrast to expression arrays, genomic array CGH deals with a vastly more complex probe mixture, and the copy number variations are much lower. In pre- and postnatal diagnostic samples, copy-number ratios vary mainly between 1/2 for deletions and 3/2 for duplications. To obtain optimal signal-to-noise ratios, controlling the fluorescent dye incorporation during labeling, assaying probe quantity, type of slides used, and adequate hybridization skills proved essential. Table 1 provides an overview of these parameters and lists the measures that were taken to optimize them.


View this table:
[in this window]
[in a new window]
  		Table 1

Parameters influencing signal-to-noise and SD of the intensity ratios

 
Because usually only a small subset of the genome is expected to be aberrant in pre- and postnatal constitutional abnormalities, the intensity ratios of patient vs control DNA will be equal for most spots on the array. This offers a unique possibility to measure the quality of an array experiment by determining the SD of all intensity ratios. The log2 values of these intensity ratios behave in a linear fashion because the relation of the intensities of both dyes inversely influences the value of the intensity ratio at a spot. The normalized log2 of the intensity ratio values will typically fluctuate around 0. The lower this fluctuation, the better the array CGH experiment. To obtain the lowest SD of the intensity ratios, adequate signal-to-noise ratios are essential as well as an optimized Cot-1/probe ratio (Figure 1). During these optimization experiments we noted that not only insufficient Cot-1 (Figure 1C) but also an excess of Cot-1 (Figure 1A) leads to an increase in the SD. The SD of all pooled clones can be further reduced by averaging several independent hybridization experiments (Figure 2).



View larger version (26K):
[in this window]
[in a new window]
  		Figure 1

(A–C) Chromosome 9 array CGH ratio profiles using DNA from a patient with a duplication 9q22.33-> 31.1 using different ratios DNA/Cot-1 (indicated at each panel). For each panel, the x-axis represents the clones ordered from the 9p telomere to the 9q telomere according to their clone position in the February 2004 Ensembl freeze. The y-axis marks log2 transformed intensity ratios at each locus.

 


View larger version (22K):
[in this window]
[in a new window]
  		Figure 2

Chromosome 9 array CGH ratio profiles using DNA from a patient with a duplication 9q22.33-> 31.1 using 200 pmol DNA/100 µg Cot-1. For each panel, the x-axis represents the clones ordered from the 9p telomere to the 9q telomere according to their clone position in the February 2004 Ensembl freeze. The y-axis marks the hybridization ratio plotted on a log2 scale. (A,B) represent the analysis of the same sample hybridization but with inversed dye labeling. Panel (C) represents the combined results of (A,B). The red lines indicates the lower thresholds for clone deletion or duplication (±4x SD) and the green line the upper threshold (log2(3/2) – 2x SD). The y-axis shows the log2 transformed intensity ratios at each locus; 97.84% of the spots were included in the analysis.

 
Quality Criteria and Threshold Values
When karyotyping, the resolution of an experiment is defined by the number of bands that can be discerned. When performing a genome-wide array CGH experiment, in theory all clones will provide a value, and the resolution of the experiment is defined by the number of genomic fragments on the array as well as by their size. However, in practice, a number of spots fail due to bad quality, e.g., bad spotting or hybridization efficiency. Therefore, it is essential to report the percentage of spots that were amenable to analysis. In our present study, typically more than 97.5% of the spots provided adequate intensity ratio values, i.e., signal-to-noise over two and the SD among the duplicates less than 10%.

When performing array CGH, the main challenge is to define a threshold value above which no false positives are retained without eliminating true positives (i.e., avoiding false negatives). Since the log2 transformed normalized intensity ratios fluctuate in a Gaussian fashion around 0, the SD can be used to define thresholds. A threshold level is described as mean plus or minus three or four times the SD (Schwaenen et al. 2004Go; Shaw-Smith et al. 2004Go). Using three times the SD as a cut-off level, 99.7% of the fragments will fall within the normal range. This would result in ~10 false-positive clones on an array with 3500 different loci. Using four times the SD as a cut-off level, 99.9936% of the fragments fall within the normal range, resulting in one false positive for every four analyses. Therefore, for an array containing 3500 loci, four times the SD is the threshold value of choice.

In addition, using four times the SD as a threshold value defines an important array quality value: the value of the SD of the log2 transformed normalized ratios. To be applicable in constitutional diagnosis, the value of four times the SD needs to be below the detection limit of an autosomal deletion or duplication. Because the ratio of a duplication (3/2) is closer to a normal ratio than the ratio of a deletion (1/2), four times the SD needs to be below the detection limit of a duplication. This threshold can be defined as the log2 transformed mean intensity ratio of the duplicated loci minus two times the SD or

4*SD ≤log2(3/2) – 2*SD or SD ≤0.096

Polymorphisms
While four times the SD seems to be an adequate cut-off level based on statistical grounds, this reasoning assumes a perfect Gaussian distribution of intensity ratios and neglects the biological and experimental causes of non-Gaussian behavior of a (small) subset of genomic fragments.

Polymorphisms can be identified on arrays as those clones for which the intensity ratios of independent experiments repeatedly fall outside the above-defined cut-off level. Identifying and reporting these polymorphic clones is thus an essential first step toward proper interpretation of array CGH data. When abnormal values are obtained in a number of separate experiments either using DNA from individuals without an obvious clinical phenotypic or in different experiments with DNA from individuals with very different clinical phenotypes, it can be assumed that these clones are polymorphic. Based on the DNA analysis of 30 individuals, aberrant signal intensity ratios were obtained in at least three different experiments in 18 loci (Table 2). However, their status as polymorphic loci awaits further experimental ascertainment.


View this table:
[in this window]
[in a new window]
  		Table 2

Clone list with suspected polymorphic loci*

 
Polymorphic loci will not only be detected by intensity ratios repeatedly surpassing the four times SD threshold value, but when the polymorphisms are embedded within a single clone on the array, they will be evidenced as smaller deviations from the average intensity ratio. Indirect evidence for such polymorphisms comes from experiments with a low SD. A low SD results in a low threshold level which, in turn, results contradictory in many more DNA fragments that are (false) positive for a deletion or duplication than in a less optimal experiment with a higher SD (data not shown).

Therefore, to eliminate further analysis of these "false positive" values, we suggest using not only the log2(1) ±4*SD as a threshold but, in addition, use the log2 of the mean of the duplicated loci minus two times the SD as a second higher threshold—97.72% of duplicated loci and ~100% of deleted loci will surpass this higher threshold. We propose that to trigger confirmatory experiments, the intensity ratio at a single locus has to surpass this higher threshold. However, in the case of two flanking clones the intensity ratios surpass the lower threshold, further experiments are warranted to confirm or disprove the suspected chromosomal anomaly.

To empirically test these theoretical figures, DNA from a normal cell line was hybridized vs DNA from a cell line trisomic for chromosome 13 (Figure 3A). The observed log2 transformed mean of the intensity ratios of duplicated loci was 0.53 rather than the theoretical 0.58. The SD of the log2 transformed intensity ratios at all normal spots was 0.08 while the SD of the duplicated loci was 0.1, rather than the theoretical SD of 0.08. From a total of 102 chromosome 13-derived loci, 88 were above the higher threshold, 10 between the higher and lower threshold, and 4 loci below the lower threshold. Hence, the intensity ratios of 13% rather than the expected 2.3% of the duplicated loci are below the higher threshold value. That more than expected values are below the threshold values is in part caused by the lower empiric mean value of the duplicated loci. This lower mean value could be caused by incomplete blocking or repetitive sequences and/or by the occurrence of some clones that contain low copy repeats which will render the theoretical value for a deletion lower than 3/2.



View larger version (27K):
[in this window]
[in a new window]
  		Figure 3

Array CGH ratio profiles using DNA from a female trisomy 13 cell line (A,C) and a mixture containing 20% DNA from the trisomy 13 cell line and 80% DNA from a normal female cell line (B,D) vs DNA from a normal male cell line. Panels A and B provide a genome-wide view. Clones are ordered from the short-arm telomere to the long-arm telomere and chromosomes are ordered from chromosomes 1–9, X, Y, and 10–22. The chromosomes that are not present in equal ratios (chromosomes X, Y, and 13) are indicated. Panels C and D show a partial karyotype enlarging the ratio profiles obtained for chromosomes 12, 13, and 14. Red lines indicate the lower thresholds for clone deletion or duplication (±4x SD) and the green line the upper threshold (log2(3/2) – 2x SD). The y-axis shows the log2 transformed intensity ratios at each locus; 97.84% of the spots were included in the analysis.

 
Mosaicisms
When using DNA from patients carrying mosaic segmental or chromosomal aneuploidies using array CGH, none of the datapoints derived from the aneuploid loci may reach the above-defined threshold levels. However, a series of intensity ratios will deviate from the mean value in a similar direction. Visual inspection of array CGH data can rapidly recognize the presence of non-random spreading of a series of flanking datapoints. DNA from a 100% trisomy-13 cell line DNA and a mixture containing only 20% of the trisomy-13 cell line DNA was hybridized vs DNA from a normal cell line. In Figure 3B, the presence of a trisomy 13 in 20% can be identified by viewing the intensity ratio values of all the chromosome 13 loci, which are all unidirectionally above the mean intensity ratio value. Molecular karyotyping can thus, in analogy with conventional karyotyping, readily be used to identify low-grade mosaicisms of at least 20%. However, smaller-sized mosaic segmental duplications or deletions that span only a few clones on an array will inevitably not be detected.


   Discussion
Top
 Summary
 Introduction
 Materials and Methods
 Results
 Discussion
Literature Cited
 
Because both genome-wide array CGH analysis and conventional karyotyping, based on staining chromosomes, aim to identify chromosomal aberrations by screening the genome, we propose to call this technology molecular karyotyping. Molecular karyotyping is likely to become part of the routine genetic diagnosis and to replace in part current karyotyping technologies. Advantages over conventional karyotyping include a higher resolution, direct mapping of aberrations to the genome sequence, amenability to automatization and quality control procedures and, probably, higher throughput and shorter reporting times. Essential for the successful introduction of the technology in current cytogenetic laboratories are (a) robust protocols and a good understanding of the critical technical factors, (b) quality criteria which define a successful array experiment, and (c) reporting guidelines to enable correct interpretation of the results obtained by different laboratories that includes a good knowledge of polymorphic loci in the human population.

Parameters Influencing Array CGH Quality
While array formats may differ in that they may use BACs (Solinas-Toldo et al. 1997Go; Pinkel et al. 1998Go), cDNAs (Pollack et al. 1999Go), oligos (Lucito et al. 2003Go) or PCR fragments (Mantripragada et al. 2004Go) as targets, the main steps involved in an array experiment are similar and include array production, probe preparation, hybridization and washing, and data analysis. When comparing different described protocols, a large variation in the details of the method can be discerned (Carter et al. 2002Go). Whereas all aspects of the protocol are important for optimal array CGH data, we identified the DNA/Cot-1 ratio as well as the efficiency of fluorescent dye incorporation as the main causes of inter-experimental variation. Batch-to-batch variability of commercial Cot-1 preparations has been previously pinpointed as the cause of variability among different experiments (Carter et al. 2002Go). The fluorophore incorporation efficiency (specific activity) has a direct effect on the sensitivity of an experiment and is thus an obvious key element for a successful array experiment.

Quality Criteria
In analogy with conventional karyotype, a molecular karyotype should have a well-defined quality. However, the few reports thus far that use array CGH for the detection of constitutional chromosomal imbalances lack uniformity in addressing the quality of an array and identifying appropriate threshold values (Snijders et al. 2001Go; Vissers et al. 2003Go; Schaeffer et al. 2004Go; Schoumans et al. 2004Go; Shaw-Smith et al. 2004Go). First, the number of clones on the array with successful hybridizations should be reported. Second, the SD of the log2 transformed intensity ratios should not exceed 0.096. Third, a minimum threshold can be defined as plus or minus four times the SD of all the log2 transformed intensity ratios of the normal clones. To define the SD of the clones with normal intensity ratios, a reiterative process to define the threshold can be used. First, the SD is calculated using the values of all clones and, subsequently, the values of the clones surpassing the threshold are eliminated and the SD is recalculated. Finally, a higher threshold value was defined as log2(3/2) – 2*SD. If the intensity ratio of only a single locus surpasses the lower threshold, it should also surpass the higher threshold to trigger confirmatory experiments. However, if two flanking loci surpass the lower threshold, this equally triggers confirmatory experiments.

Current Pitfalls for Pre- and Postnatal Diagnosis: Polymorphic Loci
Variations in chromosome lengths and banding are well known to the cytogenetic community. Common chromosomal variants have been documented over the last 30 years and practitioners have good knowledge of the most common benign variants. Novel benign variants are still being uncovered. Polymorphisms may extend several megabases (and thus multiple array loci) (Hand et al. 2000Go; Martin et al. 2002Go; Starke et al. 2003Go). Not surprisingly, at the higher resolution level obtained by molecular karyotyping, similarly polymorphic loci are detected and, due to the higher resolution, the number of variants that are observed has equally increased. The first reports using whole genome array CGH in constitutional cytogenetic analyses have identified, respectively, 2/20 and 5/50 single clone anomalies in patients that were also present in their phenotypically normal parents (Vissers et al. 2003Go; Shaw-Smith et al. 2004Go). During the analysis of 30 patients we estimated a total of 18 loci (or 0.5% of all loci on the array) within our clone set to be polymorphic as defined by intensity ratios in different independent experiments that are concordant with a deletion or duplication of a locus. However, proof that these loci are truly polymorphic sites awaits confirmation. While currently a polymorphism is described as a benign genetic variant, it seems likely that the distinction between benign polymorphisms and disease-causing genomic alterations will become blurred. Segmental duplications without phenotypic alterations have been detected (unpublished data). Recently, using a high-density molecular karyotype, Sebat et al. (2004)Go detected many copy number alterations in the population, some of which have previously been linked or associated with various phenotypes such as predispositions for cancers or neurological diseases. In addition, it is likely that smaller intraclonal variations and polymorphisms also exist. Albertson and Pinkel (2003)Go reported on the variation of the intensity ratios of a single spot derived from DNA on chromosome 6 containing the apolipoprotein. They suggest that this variation is due to the known polymorphisms within this gene. Indirect evidence that more intraclonal polymorphisms exist comes from our experiments with very low SD. In such optimal experiments the low cut-off levels result in many more DNA fragments that are (false) positive for a deletion or duplication than in a less optimal experiment with a higher SD. However, such non-Gaussian behavior of certain spots could also have other causes such as suboptimal printing or hybridization efficiencies.

Polymorphic clone information is likely to become integrated in the genome annotations. However, in the absence of a large-scale concerted effort, the question can be raised how such polymorphic sites in the genome will be ascertained. It is likely that continued feedback from a series of dedicated laboratories may lead to a validated database of candidate or proven genomic polymorphisms.

Future Prospects
Although we defined molecular karyotyping as a genome-wide array CGH experiment, no resolution has been defined. The minimum resolution of a molecular karyotype should equal but preferentially surpass the resolution obtained by conventional karyotyping, The maximum resolution that can be obtained using BAC- or PAC-based clone arrays would be a genomic tiling path array of 32,000 targets, which has already been achieved (Ishkanian et al. 2004Go). Even higher resolutions can be obtained by using arrays with overlapping BAC/PAC clones (Ishkanian et al. 2004Go), smaller insert fragments (Bruder et al. 2001Go), PCR products (Mantripragada et al. 2004Go), or oligonucleotides (Lucito et al. 2003Go). However, molecular karyotyping at higher resolution will likely await clinical use. First, based on the statistical principle that 4x SD will define the cut-off at 99.9936%, two spots would still result in false positives. One possibility is to increase the cut-off to 5x SD or even 6x SD, which will require more stringent quality criteria for an array CGH experiment. In addition, considering an estimated degree of 0.5–1% polymorphic clones in the genome, the identification and validation of these polymorphic loci will be a daunting task. Possibly, novel improved approaches to identify polymorphic loci may resolve this issue. Finally, low copy repeats, often containing small copy number variations, make up 5–10% of the genome. Aberrant ratios at these loci will be more difficult to identify.

Molecular karyotyping is likely to replace, in part, current karyotyping technologies based on staining chromosomes both in pre- and postnatal diagnosis. In addition to the many advantages, molecular karyotyping also has some caveats as compared with conventional karyotyping, e.g., it fails to identify balanced translocations and ploidy variations. Because arrays rely on the occurrence of genomic copy number differences between patient and control samples, balanced translocations, which by definition do not have genomic losses or gains, cannot be detected. Also, ploidy variations are likely to escape detection by array CGH as the technique relies on normalization of the intensity ratios. Double-dye intensity derived from a triploid DNA sample would subsequently be normalized and thus not be detected. Because of these limitations, it seems likely that we will continue to enjoy the view of banded chromosomes in the foreseeable future, and that banded chromosomes will remain an invaluable tool in the genetic diagnostic laboratory.


   Acknowledgments
 
This work was made possible by grants G.0200.03 from the FWO (Fund for Scientific Research–Flanders) and OT/O2/40 from the University of Leuven.

We would like to thank the Mapping Core and Map Finishing groups of the Wellcome Trust Sanger Institute for initial clone supply and verification.


   Footnotes
 
Presented in part at the 14th Workshop on Fetal Cells and Fetal DNA: Recent Progress in Molecular Genetic and Cytogenetic Investigations for Early Prenatal and Postnatal Diagnosis, Friedrich-Schiller-University, Jena, Germany, April 17–18, 2004.

Received for publication June 9, 2004; accepted November 12, 2004


   Literature Cited
Top
 Summary
 Introduction
 Materials and Methods
 Results
 Discussion
 Literature Cited
 

Albertson DG, Ylstra B, Segraves R, Collins C, Dairkee SH, Kowbel D, Kuo WL, et al. (2000) Quantitative mapping of amplicon structure by array CGH identifies CYP24 as a candidate oncogene. Nat Genet 25:144–146[CrossRef][Medline]

Albertson DG, Pinkel D (2003) Genomic microarrays in human genetic disease and cancer. Hum Mol Genet 12(Spec No 2):R145–152[Abstract/Free Full Text]

Bruder CE, Hirvela C, Tapia-Paez I, Fransson I, Segraves R, Hamilton G, Zhang XX, et al. (2001) High resolution deletion analysis of constitutional DNA from neurofibromatosis type 2 (NF2) patients using microarray-CGH. Hum Mol Genet 10:271–282[Abstract/Free Full Text]

Carter NP, Fiegler H, Piper J (2002) Comparative analysis of comparative genomic hybridization microarray technologies: report of a workshop sponsored by the Wellcome Trust. Cytometry 49:43–48[CrossRef][Medline]

Fiegler H, Carr P, Douglas EJ, Burford DC, Hunt S, Scott CE, Smith J, et al. (2003) DNA microarrays for comparative genomic hybridization based on DOP-PCR amplification of BAC and PAC clones. Genes Chromosomes Cancer 36:361–374[CrossRef][Medline]

Flint J, Wilkie AO, Buckle VJ, Winter RM, Holland AJ, McDermid HE (1995) The detection of subtelomeric chromosomal rearrangements in idiopathic mental retardation. Nat Genet 9:132–140[CrossRef][Medline]

Hand JL, Michels VV, Marinello MJ, Ketterling RP, Jalal SM (2000) Inherited interstitial deletion of chromosomes 5p and 16q without apparent phenotypic effect: further confirmation. Prenat Diagn 20:144–148[CrossRef][Medline]

Hodgson G, Hager JH, Volik S, Hariono S, Wernick M, Moore D, Albertson DG, et al. (2001) Genome scanning with array CGH delineates regional alterations in mouse islet carcinomas. Nat Genet 29:459–464[CrossRef][Medline]

Ishkanian AS, Malloff CA, Watson SK, DeLeeuw RJ, Chi B, Coe BP, Snijders A, et al. (2004) A tiling resolution DNA microarray with complete coverage of the human genome. Nat Genet 36:299–303[CrossRef][Medline]

Knight SJ, Regan R, Nicod A, Horsley SW, Kearney L, Homfray T, Winter RM, et al. (1999) Subtle chromosomal rearrangements in children with unexplained mental retardation. Lancet 354:1676–1681[CrossRef][Medline]

Lucito R, Healy J, Alexander J, Reiner A, Esposito D, Chi M, Rodgers L, et al. (2003) Representational oligonucleotide microarray analysis: a high-resolution method to detect genome copy number variation. Genome Res 13:2291–2305[Abstract/Free Full Text]

Mantripragada KK, Tapia-Paez I, Blennow E, Nilsson P, Wedell A, Dumanski JP (2004) DNA copy-number analysis of the 22q11 deletion-syndrome region using array-CGH with genomic and PCR-based targets. Int J Mol Med 13:273–279[Medline]

Martin CL, Waggoner DJ, Wong A, Uhrig S, Roseberry JA, Hedrick JF, Pack SD, et al. (2002) Molecular rulers" for calibrating phenotypic effects of telomere imbalance. J Med Genet 39:734–740[Abstract/Free Full Text]

Pinkel D, Segraves R, Sudar D, Clark S, Poole I, Kowbel D, Collins C, et al. (1998) High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays. Nat Genet 20:207–211[CrossRef][Medline]

Pollack JR, Perou CM, Alizadeh AA, Eisen MB, Pergamenschikov A, Williams CF, Jeffrey SS, et al. (1999) Genome-wide analysis of DNA copy-number changes using cDNA microarrays. Nat Genet 23:41–46[Medline]

Riegel M, Baumer A, Jamar M, Delbecque K, Herens C, Verloes A, Schinzel A (2001) Submicroscopic terminal deletions and duplications in retarded patients with unclassified malformation syndromes. Hum Genet 109:286–294[CrossRef][Medline]

Rosenberg MJ, Killoran C, Dziadzio L, Chang S, Stone DL, Meck J, Aughton D, et al. (2001) Scanning for telomeric deletions and duplications and uniparental disomy using genetic markers in 120 children with malformations. Hum Genet 109:311–318[CrossRef][Medline]

Schaeffer AJ, Chung J, Heretis K, Wong A, Ledbetter DH, Lese MC (2004) Comparative genomic hybridization-array analysis enhances the detection of aneuploidies and submicroscopic imbalances in spontaneous miscarriages. Am J Hum Genet 74:1168–1174[CrossRef][Medline]

Schoumans J, Anderlid BM, Blennow E, Teh BT, Nordenskjold M (2004) The performance of CGH array for the detection of cryptic constitutional chromosome imbalances. J Med Genet 41:198–202[Free Full Text]

Sebat J, Lakshmi B, Troge J, Alexander J, Young J, Lundin P, Maner S, et al. (2004) Large-scale copy number polymorphism in the human genome. Science 305:525–528[Abstract/Free Full Text]

Shaw-Smith C, Redon R, Rickman L, Rio M, Willatt L, Fiegler H, Firth H, et al. (2004) Microarray based comparative genomic hybridisation (array-CGH) detects submicroscopic chromosomal deletions and duplications in patients with learning disability/mental retardation and dysmorphic features. J Med Genet 41:241–248[Abstract/Free Full Text]

Slavotinek A, Rosenberg M, Knight S, Gaunt L, Fergusson W, Killoran C, Clayton-Smith J, et al. (1999) Screening for submicroscopic chromosome rearrangements in children with idiopathic mental retardation using microsatellite markers for the chromosome telomeres. J Med Genet 36:405–411[Abstract/Free Full Text]

Snijders AM, Nowak N, Segraves R, Blackwood S, Brown N, Conroy J, Hamilton G, et al. (2001) Assembly of microarrays for genome-wide measurement of DNA copy number. Nat Genet 29:263–264[CrossRef][Medline]

Solinas-Toldo S, Lampel S, Stilgenbauer S, Nickolenko J, Benner A, Dohner H, Cremer T, et al. (1997) Matrix-based comparative genomic hybridization: biochips to screen for genomic imbalances. Genes Chromosomes Cancer 20:399–407[CrossRef][Medline]

Starke H, Nietzel A, Weise A, Heller A, Mrasek K, Belitz B, Kelbova C, et al. (2003) Small supernumerary marker chromosomes (SMCs): genotype-phenotype correlation and classification. Hum Genet 114:51–67[CrossRef][Medline]

Schwaenen C, Nessling M, Wessendorf S, Salvi T, Wrobel G, Radlwimmer B, Kestler HA, et al. (2004) Automated array-based genomic profiling in chronic lymphocytic leukemia: development of a clinical tool and discovery of recurrent genomic alterations. Proc Natl Acad Sci USA 101:1039–1044[Abstract/Free Full Text]

Van Buggenhout G, Melotte C, Dutta B, Froyen G, Van Hummelen P, Marynen P, Matthijs G, et al. (2004) Mild Wolf-Hirschhorn syndrome: microarray CGH analysis of atypical 4p16.3 deletions enables refinement of the genotype-phenotype map. J Med Genet 41:691–698[Free Full Text]

Vissers LE, de Vries BB, Osoegawa K, Janssen IM, Feuth T, Choy CO, Straatman H, et al. (2003) Array-based comparative genomic hybridization for the genomewide detection of submicroscopic chromosomal abnormalities. Am J Hum Genet 73:1261–1270[CrossRef][Medline]


Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 Eur Heart JHome page
B. Thienpont, L. Mertens, T. de Ravel, B. Eyskens, D. Boshoff, N. Maas, J.-P. Fryns, M. Gewillig, J. R. Vermeesch, and K. Devriendt
Submicroscopic chromosomal imbalances detected by array-CGH are a frequent cause of congenital heart defects in selected patients
Eur. Heart J., November 2, 2007; 28(22): 2778 - 2784.
[Abstract] [Full Text] [PDF]

Home page
 J. Med. Genet.Home page
N. M C Maas, T. Van de Putte, C. Melotte, A. Francis, C. T R M Schrander-Stumpel, D. Sanlaville, D. Genevieve, S. Lyonnet, B. Dimitrov, K. Devriendt, et al.
The C20orf133 gene is disrupted in a patient with Kabuki syndrome
J. Med. Genet., September 1, 2007; 44(9): 562 - 569.
[Abstract] [Full Text] [PDF]

Home page
 J. Med. Genet.Home page
E. Chabchoub, L. Rodriguez, E. Galan, E. Mansilla, M. Luisa Martinez-Fernandez, M. Luisa Martinez-Frias, J.-P. Fryns, and J. Robert Vermeesch
Molecular characterisation of a mosaicism with a complex chromosome rearrangement: evidence for coincident chromosome healing by telomere capture and neo-telomere formation
J. Med. Genet., April 1, 2007; 44(4): 250 - 256.
[Abstract] [Full Text] [PDF]

Home page
 J. Med. Genet.Home page
M-L Jacquemont, D Sanlaville, R Redon, O Raoul, V Cormier-Daire, S Lyonnet, J Amiel, M Le Merrer, D Heron, M-C de Blois, et al.
Array-based comparative genomic hybridisation identifies high frequency of cryptic chromosomal rearrangements in patients with syndromic autism spectrum disorders
J. Med. Genet., November 1, 2006; 43(11): 843 - 849.
[Abstract] [Full Text] [PDF]

Home page
 J. Med. Genet.Home page
B Menten, N Maas, B Thienpont, K Buysse, J Vandesompele, C Melotte, T de Ravel, S Van Vooren, I Balikova, L Backx, et al.
Emerging patterns of cryptic chromosomal imbalance in patients with idiopathic mental retardation and multiple congenital anomalies: a new series of 140 patients and review of published reports
J. Med. Genet., August 1, 2006; 43(8): 625 - 633.
[Abstract] [Full Text] [PDF]

Home page
 J. Med. Genet.Home page
L Rickman, H Fiegler, C Shaw-Smith, R Nash, V Cirigliano, G Voglino, B L Ng, C Scott, J Whittaker, M Adinolfi, et al.
Prenatal detection of unbalanced chromosomal rearrangements by array CGH
J. Med. Genet., April 1, 2006; 43(4): 353 - 361.
[Abstract] [Full Text] [PDF]

Home page
 Hum Mol GenetHome page
O. Maertens, H. Prenen, M. Debiec-Rychter, A. Wozniak, R. Sciot, P. Pauwels, I. De Wever, J. R. Vermeesch, T. de Raedt, A. De Paepe, et al.
Molecular pathogenesis of multiple gastrointestinal stromal tumors in NF1 patients
Hum. Mol. Genet., March 15, 2006; 15(6): 1015 - 1023.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
C. Le Caignec, C. Spits, K. Sermon, M. De Rycke, B. Thienpont, S. Debrock, C. Staessen, Y. Moreau, J.-P. Fryns, A. Van Steirteghem, et al.
Single-cell chromosomal imbalances detection by array CGH.
Nucleic Acids Res., January 1, 2006; 34(9): e68 - e68.
[Abstract] [Full Text] [PDF]

Home page
 Hum Mol GenetHome page
H. Van Esch, K. Hollanders, L. Badisco, C. Melotte, P. Van Hummelen, J. R. Vermeesch, K. Devriendt, J.-P. Fryns, P. Marynen, and G. Froyen
Deletion of VCX-A due to NAHR plays a major role in the occurrence of mental retardation in patients with X-linked ichthyosis
Hum. Mol. Genet., July 1, 2005; 14(13): 1795 - 1803.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Vermeesch, J. R.
Right arrow Articles by Devriendt, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Vermeesch, J. R.
Right arrow Articles by Devriendt, K.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?


Guidelines | Subscriptions | About | exPRESS - Current - Archive | Business Information | Contact