10.3 Microarrays & Array Comparative Genomic Hybridization (aCGH)
Key Takeaways
Microarray platforms immobilize hundreds of thousands to millions of distinct, sequence-defined oligonucleotide or BAC probe features at known spatial coordinates on a solid glass or silicon surface for parallel nucleic acid interrogation.
Array Comparative Genomic Hybridization (aCGH) co-hybridizes differentially labeled patient (Cy5, red) and sex-matched normal reference (Cy3, green) genomic DNA to quantify copy number variations (CNVs), where normalized log2(Cy5/Cy3) ratios distinguish diploid normal (0.0), single-copy duplication (+0.58), and single-copy deletion (-1.0).
Single Nucleotide Polymorphism (SNP) microarrays measure both total hybridization intensity (Log R Ratio, LRR) and allelic fraction (B-Allele Frequency, BAF), uniquely identifying copy-neutral loss of heterozygosity (CN-LOH), uniparental disomy (UPD), and runs of homozygosity (ROH).
Chromosomal microarray (CMA) is recommended as the primary first-tier clinical diagnostic test for individuals with unexplained developmental delay, intellectual disability, autism spectrum disorder, and multiple congenital anomalies.
Microarrays cannot detect balanced chromosomal rearrangements (such as reciprocal translocations, balanced inversions, and Robertsonian translocations), low-level mosaicism (<10%–20%), or single-nucleotide point mutations, requiring complementary karyotyping, FISH, or NGS for comprehensive diagnostic assessment.
10.3 Microarrays & Array Comparative Genomic Hybridization (aCGH)
Quick Summary: Microarray technology immobilizes hundreds of thousands to millions of microscopic, sequence-defined nucleic acid probe features onto solid glass or silicon substrates, enabling massively parallel genomic interrogation. Array Comparative Genomic Hybridization (aCGH) employs two-color competitive hybridization (patient DNA labeled with Cy5 vs. reference DNA labeled with Cy3) to detect submicroscopic copy number variants (CNVs) across the entire genome based on normalized ratios. Single Nucleotide Polymorphism (SNP) microarrays provide dual orthogonal metrics: Log R Ratio (LRR) for total copy number and B-Allele Frequency (BAF) for allelic distribution. This dual capability allows SNP arrays to uniquely identify copy-neutral loss of heterozygosity (CN-LOH), uniparental disomy (UPD), and consanguinity, which are completely invisible to conventional aCGH. Chromosomal microarray (CMA) serves as the primary first-tier clinical diagnostic standard for developmental delay and congenital anomalies.
1. Microarray Architecture & Physical Substrates
A microarray is a miniaturized, ordered arrangement of immobilized single-stranded DNA oligonucleotides or cloned genomic fragments attached to a planar solid support (typically a silanized microscope glass slide or silica bead chip).
+---------------------------------------------------------------------------------------------------------+
| MICROARRAY PLATFORM ARCHITECTURES |
+---------------------+-----------------------------------+-----------------------------------------------+
| Microarray Platform | Substrate & Probe Synthesis | Analytical Capabilities & Resolution |
+---------------------+-----------------------------------+-----------------------------------------------+
| **Oligonucleotide** | High-density 60-mer synthetic | High resolution (10–25 kb); sequence-specific;|
| **aCGH (Agilent)** | oligonucleotides synthesized *in* | covers coding exons and intergenic backbones; |
| | *situ* via inkjet phosphoramidite | detects microdeletions and microduplications. |
+---------------------+-----------------------------------+-----------------------------------------------+
| **BAC / PAC** | Bacterial Artificial Chromosomes | Historical legacy platform; low resolution |
| **Arrays (Legacy)** | (100–200 kb genomic inserts) | (1–5 Mb); replaced by high-density synthetic |
| | spotted mechanically onto glass | oligonucleotide and SNP arrays. |
+---------------------+-----------------------------------+-----------------------------------------------+
| **Illumina Infinium**| 50-mer probes attached to 3 µm | Dual CNV + SNP genotyping (500k to >2M SNPs); |
| **BeadArray** | silica beads residing in etched | single-base extension incorporates labeled |
| | microwells on a silicon chip | dideoxynucleotide (ddNTP) hapten. |
+---------------------+-----------------------------------+-----------------------------------------------+
| **Affymetrix / Cyto**| High-density 25-mer oligos | Dual CNV + SNP interrogation (>2.6M markers); |
| **Scan (Thermo)** | photolithographically synthesized | whole-genome coverage of structural gains, |
| | directly on a quartz glass wafer | losses, LOH, and mosaicism down to 5–10%. |
+---------------------+-----------------------------------+-----------------------------------------------+
2. Array Comparative Genomic Hybridization (aCGH) Mechanics
In traditional metaphase CGH (developed by Kallioniemi in 1992), labeled patient and reference DNA were hybridized to normal metaphase chromosome spreads on a slide, limiting resolution to . Array CGH replaced metaphase chromosomes with thousands of sequence-defined genomic probes, improving diagnostic resolution down to .
THE aCGH TWO-COLOR HYBRIDIZATION WORKFLOW
[ Patient Genomic DNA ] (500 ng) [ Reference Genomic DNA ] (500 ng)
| |
v [Random Primed Labeling (Klenow)] v [Random Primed Labeling (Klenow)]
[ Labeled with Cy5 (Red, 650 nm) ] [ Labeled with Cy3 (Green, 550 nm) ]
\ /
\ /
v v
[ Combine Equal Masses + Add Human Cot-1 DNA to Block Repetitive Sequences ]
|
v
[ Co-Hybridize onto Oligonucleotide Glass Array Slide (40–48 hrs at 65°C) ]
|
v
[ Dual-Laser Microarray Scanner (635 nm Laser for Cy5 / 532 nm Laser for Cy3) ]
|
v
[ Calculate Normalized Fluorescence Ratio: log2(Cy5 Intensity / Cy3 Intensity) ]
Mathematical Derivation of Ratios
The relative copy number at any specific genomic locus is calculated from the normalized fluorescence emission of the patient channel (Cy5, Red) divided by the reference channel (Cy3, Green):
+---------------------------------------------------------------------------------------------------------+
| THEORETICAL aCGH LOG2 RATIO INTERPRETATION |
+---------------------+-------------------+-----------------------+---------------------------------------+
| Genomic Copy State | Ratio (Pt : Ref) | Theoretical Log2 Value| Mathematical Derivation |
+---------------------+-------------------+-----------------------+---------------------------------------+
| **Diploid Normal** | 2 : 2 (1.0) | **0.00** | $\log_2(2/2) = \log_2(1.0) = 0$ |
+---------------------+-------------------+-----------------------+---------------------------------------+
| **Single Deletion** | 1 : 2 (0.5) | **-1.00** | $\log_2(1/2) = \log_2(0.5) = -1.00$ |
| (Heterozygous Loss) | | (Observed: -0.6 to -1)| (Shifted toward 0 by normal stromal) |
+---------------------+-------------------+-----------------------+---------------------------------------+
| **Homozygous Del** | 0 : 2 (0.0) | **$-\infty$** | $\log_2(0/2) = -\infty$ |
| (Complete Loss) | | (Observed: < -2.0) | (Residual background fluorescence) |
+---------------------+-------------------+-----------------------+---------------------------------------+
| **Single Duplication| 3 : 2 (1.5) | **+0.585** | $\log_2(3/2) = \log_2(1.5) = +0.585$ |
| (Heterozygous Gain) | | (Observed: +0.3 to +0.5| |
+---------------------+-------------------+-----------------------+---------------------------------------+
| **Double Duplication| 4 : 2 (2.0) | **+1.00** | $\log_2(4/2) = \log_2(2.0) = +1.00$ |
| (Tetrasomy / Gain)**| | (Observed: +0.8 to +1)| |
+---------------------+-------------------+-----------------------+---------------------------------------+
| **High Amplification| > 5 : 2 (> 2.5) | **> +1.32** | $\log_2(>5/2) = \log_2(>2.5) > +1.32$ |
| (e.g., Oncogene amp)| | | (e.g., MYCN, ERBB2 amplification) |
+---------------------+-------------------+-----------------------+---------------------------------------+
aCGH CHROMOSOME PLOT DISPLAY
log2 Ratio
+1.0 | [ DUPLICATION (+0.58) ]
| * * * * * * * * *
0.0 | - - * - * - * - * - * - - - - - - - - - - - - - - - - - - * - * - * - - Diploid Baseline
| * * * * * *
-1.0 | [ DELETION (-1.0) ]
+--------------------------------------------------------------------> Chromosome Position
pter qter
3. SNP Microarrays: Dual LRR & BAF Deconvolution
While aCGH measures total copy number via two-color competitive hybridization, Single Nucleotide Polymorphism (SNP) arrays measure two independent, orthogonal parameters at millions of individual single-nucleotide coordinates.
+---------------------------------------------------------------------------------------------------------+
| ORTHOGONAL SNP ARRAY METRICS |
+---------------------+-----------------------------------+-----------------------------------------------+
| Metric Name | Mathematical Definition | Diagnostic Information Provided |
+---------------------+-----------------------------------+-----------------------------------------------+
| **Log R Ratio** | $\text{LRR} = \log_2(R_{\text{obs}}| **Copy Number / Total Dosage:** |
| **(LRR)** | / R_{\text{exp}})$ where $R$ is | Reflects total hybridization signal intensity |
| | total probe fluorescence $(A + B)$| ($\text{LRR} \approx 0$ normal; $>0$ gain; $<0$ loss) |
+---------------------+-----------------------------------+-----------------------------------------------+
| **B-Allele** | $\text{BAF} = \frac{I_B}{I_A + I_B}$| **Allelic Imbalance & Genotype:** |
| **Frequency (BAF)** | Normalized proportion of signal | Reflects the ratio of B-allele to total A+B; |
| | contributed by the 'B' allele | reveals discrete genotype tracks ($0, 0.5, 1$) |
+---------------------+-----------------------------------+-----------------------------------------------+
Deconvoluting Genotypes on LRR and BAF Plots
+---------------------------------------------------------------------------------------------------------+
| SNP ARRAY GENOTYPIC PROFILES ACROSS LRR & BAF |
+---------------------+-------------------+-----------------------+---------------------------------------+
| Biological State | Log R Ratio (LRR) | BAF Track Pattern | Observed Genotypes |
+---------------------+-------------------+-----------------------+---------------------------------------+
| **Diploid Normal** | $\text{LRR} \approx 0$| **3 Tracks:** | $AA = 0.0$ |
| (Disomy, 2 copies) | | $0.0$, $0.5$, $1.0$ | $AB = 0.5$ |
| | | | $BB = 1.0$ |
+---------------------+-------------------+-----------------------+---------------------------------------+
| **Hemizygous** | $\text{LRR} < 0$ | **2 Tracks:** | $A = 0.0$ |
| **Deletion (Loss)** | ($\approx -0.55$) | $0.0$, $1.0$ | $B = 1.0$ |
| (1 copy remaining) | | (Loss of $0.5$ track) | (Heterozygous AB state destroyed) |
+---------------------+-------------------+-----------------------+---------------------------------------+
| **Homozygous** | $\text{LRR} \ll 0$| **No discrete tracks**| No DNA remaining |
| **Deletion (Null)** | ($<-2.0$) | (Random noise) | (Complete signal loss) |
+---------------------+-------------------+-----------------------+---------------------------------------+
| **Duplication** | $\text{LRR} > 0$ | **4 Tracks:** | $AAA = 0.0$ |
| **(Trisomy, Gain)** | ($\approx +0.35$) | $0.0, 0.33, 0.67, 1.0$| $AAB = 0.33$ |
| (3 copies) | | (Split middle track) | $ABB = 0.67$, $BBB = 1.0$ |
+---------------------+-------------------+-----------------------+---------------------------------------+
| **Copy-Neutral LOH**| **LRR = 0.0** | **2 Tracks:** | $AA = 0.0$ |
| **(CN-LOH / aUPD)** | **(Normal Copy!)**| $0.0$, $1.0$ | $BB = 1.0$ |
| (2 identical copies)| | (Absence of $AB=0.5$) | (Megabase loss of heterozygosity!) |
+---------------------+-------------------+-----------------------+---------------------------------------+
SNP ARRAY COPY-NEUTRAL LOH (CN-LOH) SIGNATURE
Log R Ratio (LRR)
+1.0 |
0.0 | - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Normal Baseline!
-1.0 |
+--------------------------------------------------------------------->
B-Allele Frequency (BAF)
1.0 | * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * (BB Genotype)
|
0.5 | * * * * * * * * * * [ NO AB TRACK! ] * * * * * * * (Heterozygous AB)
|
0.0 | * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * (AA Genotype)
+--------------------+-----------------------------------+------------->
Normal Diploid Flank REGION OF CN-LOH / aUPD Normal Flank
Why aCGH Cannot Detect Copy-Neutral LOH (CN-LOH)
Copy-Neutral Loss of Heterozygosity (CN-LOH)—also known as acquired uniparental disomy (aUPD)—occurs when a cell loses one parental allele and duplicates the remaining homologous allele via mitotic recombination or nondisjunction rescue.
- Total copy number remains exactly 2.
- In aCGH, because total patient DNA hybridizes at a ratio with normal diploid reference DNA, the calculated is exactly . aCGH is completely blind to CN-LOH.
- In SNP arrays, while the LRR is , the BAF track instantly reveals the complete loss of all heterozygous SNPs ( frequency) over contiguous megabase segments, exposing the abnormal isodisomy.
- Clinical Significance: Identifies imprinting disorders (e.g., Uniparental Disomy in Prader-Willi / Angelman syndromes on chromosome 15, Beckwith-Wiedemann syndrome on 11p15.5) and oncogenic homozygosity of tumor suppressor mutations (e.g., TP53, FLT3, JAK2 V617F in AML/MDS).
4. Clinical Indications & Professional ACMG Guidelines
The American College of Medical Genetics and Genomics (ACMG) and the American Academy of Pediatrics (AAP) endorse Chromosomal Microarray (CMA) as the first-tier diagnostic test for:
- Unexplained Developmental Delay (DD) and Intellectual Disability (ID)
- Autism Spectrum Disorders (ASD)
- Multiple Congenital Anomalies (MCA) not fitting a well-defined monogenic syndrome
+---------------------------------------------------------------------------------------------------------+
| MICROARRAY VS. G-BAND KARYOTYPING |
+---------------------+-----------------------------------+-----------------------------------------------+
| Parameter | G-Banded Karyotyping | Chromosomal Microarray (CMA) |
+---------------------+-----------------------------------+-----------------------------------------------+
| **Resolution** | **5 to 10 Mb** (low resolution) | **10 to 50 kb** (>100-fold higher resolution!)|
+---------------------+-----------------------------------+-----------------------------------------------+
| **Cell Culture** | Mandatory; requires living cells | None; uses extracted DNA from uncultured blood|
| | arrested in metaphase | saliva, skin, or fresh/frozen tissue |
+---------------------+-----------------------------------+-----------------------------------------------+
| **Diagnostic Yield**| 3% to 5% in DD/ID/ASD cohorts | **15% to 20%** in DD/ID/ASD cohorts |
+---------------------+-----------------------------------+-----------------------------------------------+
| **Submicroscopic** | Invisible (below optical limit) | Easily detects DiGeorge (22q11.2), Williams |
| **Microdeletions** | | (7q11.23), Prader-Willi/Angelman (15q11-q13) |
+---------------------+-----------------------------------+-----------------------------------------------+
| **Balanced Trans-** | **Detected easily** (visualizes | **Completely blind**; no copy number change! |
| **locations** | chromosome structural movements) | |
+---------------------+-----------------------------------+-----------------------------------------------+
5. Inherent Limitations of Microarray Platforms
Understanding what microarrays cannot detect is a primary focus of clinical competency and the ASCP MB examination.
+---------------------------------------------------------------------------------------------------------+
| WHAT MICROARRAYS CANNOT DETECT |
+-----------------------------+---------------------------------------------------------------------------+
| Structural Variant Type | Biochemical / Biophysical Reason for Diagnostic Invisibility |
+-----------------------------+---------------------------------------------------------------------------+
| **Balanced Reciprocal** | Involves mutual exchange of chromosomal segments without any gain or loss |
| **Translocations** | of genomic material ($\log_2\text{ Ratio} = 0.00$; $\text{LRR} = 0.00$). |
+-----------------------------+---------------------------------------------------------------------------+
| **Balanced Inversions** | DNA segment is inverted 180° in orientation; total copy number is normal. |
+-----------------------------+---------------------------------------------------------------------------+
| **Robertsonian Trans-** | Fusion of acrocentric p-arms (containing non-essential rDNA); balanced. |
| **locations (Balanced)** | |
+-----------------------------+---------------------------------------------------------------------------+
| **Gene Relocations / Tandem**| Arrays identify gain of a segment but cannot determine its spatial |
| **vs. Insertional Position**| location (e.g., tandem duplication on chr 1 vs inserted into chr 12). |
+-----------------------------+---------------------------------------------------------------------------+
| **Low-Level Mosaicism** | Minor clonal sub-populations (<10% to 20% mutant cells) are submerged in |
| | background diploid signal noise. |
+-----------------------------+---------------------------------------------------------------------------+
| **Single-Nucleotide SNVs** | Point mutations (e.g., *CFTR* F508del, *BRAF* V600E) and tiny indels |
| **and Small Indels (<10 kb)**| are not detected unless directly interrogated by specific single-base SNPs|
+-----------------------------+---------------------------------------------------------------------------+
| **Trinucleotide Expansions**| Repetitive expansions (*FMR1*, *HTT*) do not alter copy number probes. |
+-----------------------------+---------------------------------------------------------------------------+
6. Microarray Quality Control Metrics & Clinical Troubleshooting
To guarantee diagnostic validity under CAP/CLIA regulations, laboratories analyze mathematical quality metrics before releasing clinical array data.
+---------------------------------------------------------------------------------------------------------+
| ARRAY QUALITY CONTROL BENCHMARKS |
+---------------------+-----------------------------------+-----------------------------------------------+
| QC Metric | Clinical Diagnostic Benchmark | Significance & Failure Root Cause |
+---------------------+-----------------------------------+-----------------------------------------------+
| **Derivative Log** | **DLRSpread < 0.20 to 0.30** | Measures probe-to-probe noise across adjacent |
| **Ratio Spread** | (Agilent aCGH platforms) | chromosomal features. Elevated by degraded |
| **(DLRSpread)** | | DNA, poor labeling, or uneven hybridization. |
+---------------------+-----------------------------------+-----------------------------------------------+
| **SNP Call Rate** | **> 98.5% to 99.0%** | Percentage of SNP features assigned a high- |
| | (Illumina / Affymetrix arrays) | confidence discrete genotype. Reduced by poor |
| | | DNA quality, low input mass, or inhibitors. |
+---------------------+-----------------------------------+-----------------------------------------------+
| **Specific** | **> 8.0 to 10.0 pmol dye / µg** | Measures fluorophore incorporation efficiency |
| **Activity** | for Cy5 and Cy3 | during random priming with Klenow fragment. |
+---------------------+-----------------------------------+-----------------------------------------------+
| **Signal-to-Noise** | **SNR > 10.0** | Ratio of mean feature foreground intensity to |
| **Ratio (SNR)** | for all scanning channels | surrounding background noise. |
+---------------------+-----------------------------------+-----------------------------------------------+
Systematic Array Troubleshooting Guide
| Visual / Analytical Anomaly | Root Cause | Corrective Action |
|---|---|---|
| Elevated DLRSpread () / High Wave Noise | Degraded input genomic DNA; salt/solvent contamination; bubbles during hybridization | Verify DNA integrity via Agilent Bioanalyzer / TapeStation (DIN required); re-purify DNA with silica columns; roll out bubbles during array chamber assembly. |
| Wavy Baseline Across Chromosomes (GC Wave) | Systematic labeling bias or hybridization artifact correlating with local genomic GC content | Apply bioinformatic GC-wave correction algorithms (LOESS / spline smoothing); ensure equal DNA input quantities for Cy5 and Cy3. |
| Severe Dye Bias / Unbalanced Channels | Unequal labeling efficiency between Cy5 and Cy3; Cy5 ozone degradation | Measure specific activity spectrophotometrically prior to pooling; protect Cy5 from ambient atmospheric ozone () or use ozone-free enclosures. |
| Localized Donut / Ring Artifacts on Scan | Air bubble trapped under array gasket during 65°C incubation; dust on glass | Ensure hybridization rotator oven operates at correct RPM; handle slides in certified cleanroom / HEPA-filtered laminar flow hoods. |
| Low SNP Call Rate () | Inadequate genomic DNA input mass (); incomplete single-base extension enzymology | Accurately quantify input DNA using fluorometry (Qubit dsDNA BR); verify reagent expiration dates on enzyme extension master mix. |
A clinical cytogenetics laboratory evaluates an infant with suspected Beckwith-Wiedemann syndrome using a genome-wide SNP microarray. Analysis of chromosome 11p15.5 demonstrates a normal Log R Ratio (LRR ~ 0.0) across the entire 15-megabase terminal region, but the B-Allele Frequency (BAF) plot reveals a complete absence of heterozygous markers (no 0.5 BAF track), displaying only homozygous 0.0 and 1.0 tracks. What genetic mechanism explains this finding, and why would standard two-color aCGH fail to detect it?
A balanced reciprocal translocation between chromosomes 11 and 22; aCGH fails because translocations alter only centromere positioning
A submicroscopic 50-kb microdeletion; aCGH fails because oligonucleotide probes cannot hybridize to telomeric regions
Copy-Neutral Loss of Heterozygosity (CN-LOH) / Paternal Uniparental Disomy; aCGH fails because total DNA copy number is unchanged (2 copies present), resulting in a normal log2 ratio of 0.0
A triploid mosaic duplication; aCGH fails because Cy5 and Cy3 fluorescent dyes cross-quench in the 11p15 region
A pediatric patient with severe developmental delay and dysmorphic facial features is evaluated using a high-density oligonucleotide chromosomal microarray. The microarray report is returned as completely normal with no copy number variations detected. However, subsequent classical G-banded karyotyping reveals a de novo balanced reciprocal translocation between chromosomes 4 and 8: t(4;8)(q21;q24). Why was this structural chromosomal abnormality undetectable by chromosomal microarray?
Microarray slide scanning lasers operate at 532 nm and 635 nm, which cannot excite chromosome 4 oligonucleotides
Balanced reciprocal translocations involve a mutual exchange of chromosomal segments without any net gain or loss of genomic sequence, resulting in a normal log2 copy number ratio of 0.0 across all array probes
Chromosomal microarrays can only evaluate ribonucleic acid (RNA) and cannot bind genomic DNA fragments
The DLRSpread quality metric was excessively high, which automatically filters out all translocations during bioinformatic analysis
A molecular technologist is analyzing two-color array Comparative Genomic Hybridization (aCGH) data for a patient suspected of having a contiguous gene syndrome. Across a 3-megabase region on chromosome 22q11.2, the software reports a consistent, normalized log2(Cy5/Cy3) ratio of +0.585. What does this quantitative value indicate regarding the patient's copy number status at this locus?
A complete homozygous deletion resulting in zero remaining copies of the 22q11.2 region
A normal diploid state with two intact copies of the 22q11.2 region
A single-copy hemizygous deletion resulting in only one functional copy of the 22q11.2 region
A single-copy duplication resulting in three total copies (heterozygous gain) of the 22q11.2 region
Sections you finish are checked off in the contents.