For citation purposes: Liehr T. Copy number variations?is there a biological difference between submicroscopic and microscopically visible ones? OA Genetics 2013 Apr 01;1(1):2.

Critical review

Cytogenetics & Genomic Imaging

Copy number variations–is there a biological difference between submicroscopic and microscopically visible ones?

T Liehr*

Authors affiliations

Jena University Hospital, Friedrich Schiller University, Institute of Human Genetics, Kollegiengasse, Jena, Germany

*Corresponding author Email:



Copy number variations are nowadays most often understood as submicroscopic gains or losses of chromosomal material, either connected with a disease or just one of the many possible genetic variants in man. However, besides such submicroscopic copy number variations,already decades ago chromosome analysis revealed the existence of cytogenetic visible copy number variations; first discovered were the chromosomal heteromorphisms, more recently euchromatic variants and unbalanced chromosome abnormalities without phenotypic consequences. Recently, for submicroscopic copy number variations a so-called two-hit model suggested. This suggests that the combination of per se harmless size variants may be led to clinical aberrations if they are present together in a patient.

A literature based comparison of cytogenetic visible copy number variations and MG-CNVs revealed that, biologically, besides size, these two groups of copy number variations do not really differ. Both may be present as heterochromatic (gene-poor to gene-free) and euchromatic copy number variation-variants. Identical regions have even been reported as cytogenetic visible copy number variation and a submicroscopic copy number variationsindependently. This study reviews the differences between submicroscopic and microscopically visible CNVs.


Yet cytogenetic visible copy number variations, especially chromosomal heteromorphisms are by far not considered enough in evaluation of routine cytogenetic analysis as well as in array-based comparative genomic hybridisation analysis.


Since 1956 and the correct assessment of human chromosome numbers[1], cytogenetics plays a crucial role in pre- and postnatal, as well as tumour cytogenetic diagnostics and research. Currently, the GTG-banding approach (G-bands by Trypsin using Giemsa)[2] is still the gold-standard for all cytogenetic techniques and the preferred technique of molecular cytogenetics is fluorescence in situ hybridization (FISH)[3]. One additional major proceeding in molecular (cyto)genetics is nowadays the array-based variant of comparative genomic hybridisation (aCGH)[4]. All these mentioned approaches provide information on the human genome at different levels of resolutions5 (Figure 1). Public and patients expect from such analysis clear results in that sense that the tested person has (a) a normal or (b) an abnormal genetic condition. As in many cases nature is much more complex than layman—or even specialists—may expect. Besides the sheer impossibility per se to define what is ‘normal’, there is the problem of the so-called ‘copy number variant’ (CNV) regions in the human genome; CNVs can be larger or smaller in size in each clinically healthy individual tested, and any ‘norm’ can only be an artificial definition of a standard size[6].

Current approaches for analysis of the human genome at different resolutions.(A) Banding cytogenetics has an approximate resolution of 10 Mb; arrowheads highlight heteromorphisms (CG-CNVs) on the short arms of the acrocentric chromosomes 13 and 22.(B) Molecular cytogenetics provides resolutions between ~10Mb down to ~1Mb; in interphase resolution may be higher than in metaphase due to differences in DNApacking. 1. A locus-specific probe for the Kallmansyndrome (KAL1) specific region in Xp22.3 (red) and a centromeric probe (CEP X, green) are applied on a metaphase and an interphase nucleus of a healthy female. 2. The FISH-banding approach multicolour banding (MCB)[3] was used to characterise the breakpoints (arrowheads) of an inversion in one chromosome 5 to 5q15 and 5q31. (C) Here the most important segment of an aCGH result using the platform and software of Agilent is depicted. A clinical case with different clinical abnormalities was studied. The resolution of aCGH depends on the amount of the probes spotted on the so-called ‘DNAchip’; even though the results are given down to basepair level by the software, the resolution in diagnostic cases is considered to be between ~10 and ~200 kb. Here detected imbalances are shown in three different resolutions. The lowest resolution (‘all chromosomes’) gives an overview on all imbalances along the entire genome: one can see ~10 small and one large imbalance. While the smaller ones are considered as MG-CNVs, the large imbalance on chromosome 13 of ~37,577 Mb in size is the disease-causing one. Two higher magnifications of the affected regions are shown (‘chromosome 13’ and ‘affected region’). The final result would read as arr[GRCh37/hg19] 13q14.11q31.1(44,819,113x2,44,844,282-82,421,008x1,82,474,913x2) and the MG-CNV are not detailed in the report. Figure kindly provided by Dr. Kristin Mrasek, Human Genetics, Jena, Germany. (D) Next generation sequencing (NGS) is at present used for targeted analysis, for example, to detect single basepair mutations in a tumour suppressor gene. NGS provides the highest possible resolution for human genome. Here a result for a heterozygous, disease causing mutation with in the BRCA1 gene was found and is highlighted by arrowheads. NB: NGS is not suited for the analysis of higherorder repeat units, the latter being subject to CNVs. Figure kindly provided by M.Sc. Martin Voigt, Human Genetics, Jena, Germany.

In 2004 two keypublications[7,8] reported that no two clinically healthy individuals are alike and yet it is known[9] that each persons persondistinguishes from another by up to 0.5 megabasepair (Mb) of submicroscopic CNVs; the latter are abbreviated from now on as MG-CNVs, for ‘CNVs detectable by molecular genetics’. The description of MG-CNVs was celebrated in the years after 2004 as some kind of big surprise[10]. However, it was an already early finding of cytogenetics that on a chromosomal level there are lots of cytogenetically visible copy number variations (CG-CNVs), and that there are no individuals which are really the same on a chromosomal level[6,11,12]. The aim of this study was to review if there is a biological difference between submicroscopic and microscopically visible CNVs.


The author has referenced some of his own studies in this review.The protocols of these studies have been approved by the relevant ethics committees related to the institution in which they were performed.

What are CG-CNVs and what are MG-CNVs ones?

Any kind of CNV may be found in otherwise normal as well as abnormal karyotypes/genomes. In general, CNVs are considered as having no (major) clinical impact on the phenotype of its carrier and thus, may be present in clinical healthy and diseased persons.

CG-CNVs are here defined as gain or loss of genetic material, which leads to a different appearance of a chromosome visualised in banding cytogenetics; MG-CNVs are such, which are detectable in aCGH together with a normal karyotype and phenotype. For more comprehensive characterisation of a CG-CNV and/or an MG-CNV molecular cytogenetics may be applied[6].

In the literature, also clinically relevant changes in copy numbers, especially regions involved in microdeletion/microduplication syndromes[13] are sometimes reported as MG-CNVs. In this review, the major focus is on clinically primarily inert MG-CNVs. It has to be admitted, that some potentially harmless MG-CNVs have already been linked in population studies to an enhanced or diminish susceptibility for special disorders or (age-related) diseases. However, mechanisms, therefore, are still not understood yet[14].

A closer look on CG-CNVs

CG-CNVs comprise enlargement or reduction in the size of chromosomal subbands and/or appearance of additional band(s). Without knowledge about regions prone to develop CG-CNVs, and sometimes also if a cytogeneticist analyses GTG-banded chromosomes, such variants may easily be mixed up with clinically relevant gain or loss of chromosomal material[6].

However, specific chromosomal regions in the human karyotype are prone to variations more than others, like the centromeric satellite DNA of any of the 24 human chromosomes, the band Yq11.2 or the short arms of all acrocentrics. All these regions (Figure 2) are considered as heterochromatin, which are extremely gene-poor. It is important to notice that cytogenetically visible heterochromatic blocks may vary in size (in their original chromosomal location), but also may be inserted at any place in the human genome[15]; even mitochondrial DNA may be inserted and amplified in human chromosomes (Dr. M. Groß, Heidelberg, personal communication).

Heterochromatic regions (CG-CNVs) and euchromatic variants (EVs) of the human genome are highlighted in this schematic depiction of a haploid set of human chromosomes.

Overall, the majority of differences between individuals expressed as heterochromatic CG-CNVs are found in 1q12, 9q12, 13pter-q11, 14pter-q11.1, 15pter-q11.1, 16q11.2, 19p12-q12, 21pter-q11.1, 22pter-q11.1 and in male in Yq12; the cytogenetic approaches from the mid-1970s mid 1970s identified an average of four to five chromosomal heteromorphisms per person[6].

Besides heterochromatic CG-CNVs, there are also euchromatic ones, which are more recent findings. They were reported as so-called euchromatic variants (i.e. large scale CG-CNVs) (Figure 2) and unbalanced chromosome abnormalities without phenotypic consequences[6,16]. These CG-CNVs can be due to gain or loss of copy numbers at the original location; moreover, such regions can be translocated to another chromosomal region due to rearrangements. It is suggested that any region not containing dose-dependant genes may be amplified or lost in the genome without clinical consequences, as also seen in euchromatic small supernumerary marker chromosomes[6,17].

A closer look on MG-CNVs

A MG-CNV is a submicroscopic amplification or loss of DNAstretches. One of the first identified MG-CNV is a 28 kilobasepair(kb) deletion in 15q11.2[18]. MG-CNVs range from 1 kb to several Mb in size and are distributed along the entire human genome. As detected by aCGH, which for technical reasons only can identify euchromatic regions of the human genome, no submicroscopic heterochromatic CNVs were reported yet. However, there are hints from exceptional studies that such also exist, but as aCGH is (almost) blind for these regions yet, they cannot be analysed systematically at present. MG-CNVs which consist of low repetitive elements are summarised in Figure 3[19]. Nonetheless, clinically relevance was found already for some of them[19,20].

MG-CNVs which consist of low repetitive elements are highlighted according to Warburton et al.[19].

MG-CNVs are determined as variable copy numbers when compared to a reference genome and may include deletions and duplications of genomic loci[21]. They may encompass as much as 12% of the human genome[22]. Most of them are considered as benign MG-CNVs and are usually inherited from a parent[23]. When determined as de novo, genomic imbalances are considered more likely pathological[24]. It is also known that each human being carries about thousand MG-CNVs ranging from only a few hundred basepairs to over 1 Mb[25]. The major determinant for the clinical impact of a CNV seems to be if dosage sensitive genes are present in the corresponding DNA stretch[17,26]. However, ‘it was recently observed that more than one (submicroscopic) CNV (larger than 500 kb) can contribute to severe developmental delay and often is responsible for phenotypic variability associated with genomic disorders’; this phenomenon is called ‘two-hit’ model[9].

CG-CNVs and MG-CNVs—a comparison

CG-CNVs and MG-CNVs have in common 8 of 10 of the features listed in Table 1. The theories how they may be formed are identical—unequal crossing-over events or breakage-fusion-bridge model are considered. Both may be either inherited or de novo formed. Both are size variants of DNA stretches and may be visualised by molecular cytogenetics. Both may be euchromatic, but also heterochromatic, as outlined before. They only are different in two points: (i) MG-CNVs may be associated with disease, CG-CNVs may not, and (ii) CG-CNVs are microscopically visible, MG-CNVs are submicroscopic. Both points are matters of definition: for (i) CG-CNVs considered only as such if they are not associated with clinical signs and symptoms; for (ii) MG-CNVs are submicroscopic just due to limitations of optical microscopy[6].

Table 1

Comparison of CG-CNVs and MG-CNVs

Thus, overall there is no biological reason to distinguish MG-CNVs and CG-CNVs,i.e., besides size there is no real difference between them. Nothing could underline this fact better than identical CNV regions reported ‘independently’ by cytogenetic approaches on the one side and by molecular techniques on the other side; examples are CNVs in 8q21.2[27] (Figure 4) or 15q11.2[18].

FISH result in a case with a CG-CNV in the region 8q21.2; chromosome 8 with arrowhead shows an enlargement of the corresponding sub-band in GTG-banding and the BAC-probe RP11-96G1 gives an ~3 fold stronger signal on chromosome 8 with arrowhead than on the homologous one. In database of genomic variants, the corresponding region at position 86,898,422-86,955,420 (NCBI36/hg18) has a collection of 60 normal cases with gain and 150 cases with loss of copy numbers according to aCGH. Obviously this region can be expressed as MG-CNV or CG-CNV. A centromeric probe for chromosomes 8 (cep 8) in red was also applied in this FISH-experiment as a positive control.

Every human being provides besides a unique DNA-primary sequence also an individual combination of CG-CNV and MG-CNV. Monozygote twins are considered as exception from that rule, however, distinct MG-CNVs were already found there[28]. In genetic analyses, irrespective which of the presently available approaches are applied, the geneticist has always to be aware of the fact that one never obtains the whole picture of a genome (Figure 1). (Molecular) cytogenetics provides a genome wide view with low resolution; however, also heterochromatic CG-CNVs are visible there. Molecular genetic analysis provide a high-resolution of the analysed genome, but low- and massive-copy repeats are neither ascertainable by aCGH nor by next generation sequencing approaches[29] (Figure 1).


The future has to show if we start to consider CG-CNV and MG-CNV as the same biological phenomenon or keep the artificial borders of thinking here. If these obstacles are driven down the way may also be free to check the phenomenon ‘two-hit’ model for patients with multiple CG-CNVs or with CG-CNVs and MG-CNVs and syndrome or disease. Also, then it might be necessary to list all detected MG-CNVs in a CGH reports and all CG-CNVs, especially heterochromatin variants, in cytogenetic reports.

Authors Contribution

All authors contributed to the conception, design, and preparation of the manuscript, as well as read and approved the final manuscript.

Competing interests

None declared.

Conflict of interests

None declared.


All authors abide by the Association for Medical Ethics (AME) ethical rules of disclosure.


  • l. Tijo JH, Levan A. The chromosome number in man. Hereditas 1956 May;42(1-2):1-6.
  • 2. Claussen U, Michel S, Mühlig P, Westermann M, Grummt UW, Kromeyer-Hauschild K. Demystifying chromosome preparation and the implications for the concept of chromosome condensation during mitosis. Cytogenet Genome Res 2002 Feb;98(2-3):136-46.
  • 3. Liehr T, Claussen U. Current developments in human molecular cytogenetic techniques. Cur Mol Med 2002 May;2(3):269-84.
  • 4. Ren H, Francis W, Boys A, Chueh AC, Wong N, La P. BAC-based PCR fragment microarray: high-resolution detection of chromosomal deletion and duplication breakpoints. Hum Mutat 2005 May;25(5):476-82.
  • 5. Shinawi M, Cheung SW. The array CGH and its clinical applications. Drug Discov Today 2008 Sep;13(17-18):760-70.
  • 6. Liehr T . Benign & pathological chromosomal imbalances. 1st ed.. New York: Oxford Academic Press; 2014.
  • 7. Iafrate AJ, Feuk L, Rivera MN, Listewnik ML, Donahoe PK, Qi Y. Detection of large-scale variation in the human genome. Nat Genet 2004 Sep;36(9):949-51.
  • 8. Sebat J, Lakshmi B, Troge J, Alexander J, Young J, Lundin P. Large-scale copy number polymorphism in the human genome. Science 2004 Jul;305(5683):525-8.
  • 9. Girirajan S, Rosenfeld JA, Cooper GM, Antonacci F, Siswara P, Itsara A. A recurrent 16p12.1 microdeletion supports a two-hit model for severe developmental delay. Array 2010 Mar;42(3):203-9.
  • 10. Buckley PG, Mantripragada KK, Piotrowski A, Diaz de Ståhl T, Dumanski JP. Copy-number polymorphisms: mining the tip of an iceberg. Trends Genet 2005 Jun;21(6):315-7.
  • 11. Ferguson-Smith MA, Ferguson-Smith ME, Ellis PM, Dickson M. The sites and relative frequencies of secondary constrictions in human somatic chromosomes. Cytogenetics 1962 Feb;1325-43.
  • 12. Makino S, Muramoto JI, Tabata S. A survey of a familial transmission of an anomalous autosome in group 13-15. Chromosoma 1966 Jun;18(3):371-9.
  • 13. Weise A, Mrasek K, Klein E, Mulatinho M, Llerena JC Jr, Hardekopf D. Microdeletion and microduplication syndromes. J Histochem Cytochem 2012 May;60(5):346-58.
  • 14. Uyan O, Omür O, Ağim ZS, Ozoğuz A, Li H, Parman Y. Genome-wide copy number variation in sporadic amyotrophic lateral sclerosis in the Turkish population: deletion of EPHA3 is a possible protective factor. PLoS One 2013 Aug;8(8):e72381.
  • 15. Bucksch M, Ziegler M, Kosayakova N, Mulatinho MV, Llerena JC Jr, Morlot S. A new multicolor fluorescence in situ hybridization probe set directed against human heterochromatin: HCM-FISH. J Histochem Cytochem 2012 Jul;60(7):530-6.
  • 16. Barber JC, Maloney V, Hollox EJ, Stuke-Sontheimer A, du Bois G, Daumiller E. Duplications and copy number variants of 8p23.1 are cytogenetically indistinguishable but distinct at the molecular level. Eur J Hum Genet 2005 Oct;13(10):1131-6.
  • 17. Liehr T . Small supernumerary marker chromosomes (sSMC). A Guide for Human geneticists and clinicians; with contributions by UNIQUE (The Rare Chromosome Disorder Support Group).. 1st ed. Berlin, New York: Springer 2012.
  • 18. Buiting K, Dittrich B, Dworniczak B, Lerer I, Abeliovich D, Cottrell S. A 28-kb deletion spanning D15S63 (PW71) in five families: a rare neutral variant?. Am J Hum Genet 1999 Dec;65(6):1588-94.
  • 19. Warburton PE, Hasson D, Guillem F, Lescale C, Jin X, Abrusan G. Analysis of the largest tandemly repeated DNA families in the human genome. BMC Genomics 2008 Nov;9533.
  • 20. Lemmers RJ, Tawil R, Petek LM, Balog J, Block GJ, Santen GW. Digenic inheritance of an SMCHD1 mutation and an FSHD-permissive D4Z4 allele causes facioscapulohumeral muscular dystrophy type 2. Nat Genet 2012 Dec;44(12):1370-4.
  • 21. Feuk L, Carson AR, Scherer SW. Structural variation in the human genome. Nat Rev Genet 2006 Feb;7(2):85-97.
  • 22. Redon R, Ishikawa S, Fitch KR, Feuk L, Perry GH, Andrews TD. Global variation in copy number in the human genome. Nature 2006 Nov;444(7118):444-54.
  • 23. Lee C, Iafrate AJ, Brothman AR. Copy number variations and clinical cytogenetic diagnosis of constitutional disorders. Nat Genet 2007 Jul;39(7 Suppl):S48-54.
  • 24. Tyson C, Harvard C, Locker R, Friedman JM, Langlois S, Lewis ME. Submicroscopic deletions and duplications in individuals with intellectual disability detected by array-CGH. Am J Med Genet A 2005 Dec;139(3):173-85.
  • 25. Conrad DF, Pinto D, Redon R, Feuk L, Gokcumen O, Zhang Y. Origins and functional impact of copy number variation in the human genome. Nature 2010 Apr;464(7289):704-12.
  • 26. Canales CP, Walz K. Copy number variation and susceptibility to complex traits. EMBO Mol Med 2011 Jan;3(1):-4.
  • 27. Manvelyan M, Cremer FW, Lancé J, Kläs R, Kelbova C, Ramel C. New cytogenetically visible copy number variant in region 8q21.2. Mol Cytogenet 2011 Jan;4(1):1.
  • 28. Bruder CE, Piotrowski A, Gijsbers AA, Andersson R, Erickson S, Diaz de Ståhl T. Phenotypically concordant and discordant monozygotic twins display different DNA copy-number-variation profiles. Am J Hum Genet 2008 Mar;82(3):763-71.
  • 29. Sipos B, Massingham T, Stütz AM, Goldman N. An improved protocol for sequencing of repetitive genomic regions and structural variations using mutagenesis and next generation sequencing. PLoS One 2012 Aug;7(8):e43359.
Licensee to OAPL (UK) 2013. Creative Commons Attribution License (CC-BY)

Comparison of CG-CNVs and MG-CNVs

Formation by unequal crossing-over events + +
Formation by breakage–fusion–bridge model + +
May be inherited + +
May be de novo + +
Are size variants of DNA stretches + +
May be visualised by molecular cytogenetics + +
May be euchromatic + +
May be heterochromatic + +
May be associated with disease (-)* +
Are microscopically visible + -

*Due to definition of CG-CNVs.