Volume 53 (3): 401-408, 2005 Copyright ©The Histochemical Society, Inc. An Approach for Quantitative Assessment of Fluorescence In Situ Hybridization (FISH) Signals for Applied Human Molecular Cytogenetics
National Center of Mental Health, Russian Academy of Medical Sciences, Moscow, Russia (IYI,IVS,VVM,YBY), and Institute of Pediatrics and Children's Surgery, Russian Ministry of Health, Moscow, Russia (SGV) Correspondence to: Y.B. Yurov, National Center of Mental Health, Russian Academy of Medical Sciences, Zagorodnoe sh.2, 119152, Moscow, Russia. E-mail: y_yurov{at}yahoo.com; i_yurov{at}mail.ru
A number of applied molecular cytogenetic studies require the quantitative assessment of fluorescence in situ hybridization (FISH) signals (for example, interphase FISH analysis of aneuploidy by chromosome enumeration DNA probes; analysis of somatic pairing of homologous chromosomes in interphase nuclei; identification of chromosomal heteromorphism after FISH with satellite DNA probes for differentiation of parental origin of homologous chromosome, etc.). We have performed a pilot study to develop a simple technique for quantitative assessment of FISH signals by means of the digital capturing of microscopic images and the intensity measuring of hybridization signals using Scion Image software, commonly used for quantification of electrophoresis gels. We have tested this approach by quantitative analysis of FISH signals after application of chromosome-specific DNA probes for aneuploidy scoring in interphase nuclei in cells of different human tissues. This approach allowed us to exclude or confirm a low-level mosaic form of aneuploidy by quantification of FISH signals (for example, discrimination of pseudo-monosomy and artifact signals due to over-position of hybridization signals). Quantification of FISH signals was also used for analysis of somatic pairing of homologous chromosomes in nuclei of postmortem brain tissues after FISH with "classical" satellite DNA probes for chromosomes 1, 9, and 16. This approach has shown a relatively high efficiency for the quantitative registration of chromosomal heteromorphism due to variations of centromeric alphoid DNA in homologous parental chromosomes. We propose this approach to be efficient and to be considered as a useful tool in addition to visual FISH signal analysis for applied molecular cytogenetic studies. (J Histochem Cytochem 53:401408, 2005)
Key Words: quantitative FISH differentiation of FISH signals aneuploidy scoring low-level chromosomal mosaicism chromosome heteromorphism
FLUORESCENCE IN SITU HYBRIDIZATION (FISH) is considered to be the most commonly used molecular cytogenetic technique, having a wide variety of applications. Although it has been enhanced and modified since the early 1980s, there are a number of unsolved problems with the use of FISH for different diagnostic as well as research purposes. One of these is the interpretation of FISH results during chromosome identification using interphase FISH. In the study of aneuploidy by FISH in preimplantation and prenatal or postnatal diagnosis, the technique of quantitative signal differentiation appears to be required. This is of significant importance in FISH studies of low-level fetal and placental mosaicism and single blastomers or polar bodies because FISH artifacts can significantly reduce the resolution of the results. Over-position of interphase chromosome, somatic pairing of homolog chromosomes in nuclei, different replication patterns of chromosomes in dividing cells, or simply absence of signals of one or both homologous chromosomes due to low efficiency of hybridization are the most common reasons for the misinterpretation of FISH results. One possible solution to this problem is the development and application of methods for the quantitative analysis and differentiation of individual FISH signals from FISH artifacts.
Originally, quantitative analysis of fluorescent images was carried out for the improvement of routine cytogenetic tests (Pinkel et al. 1986 Here we present a relatively simple and rapid approach for the quantitative assessment of FISH signals based on the digital capturing of microscopic images and the intensity measuring of hybridization signals by Scion Image software originally developed for analyzing electrophoresis gels. We have carried out the tests for this technique studying aneuploidy and the low-level chromosomal mosaicism involving different human chromosomes in interphase nuclei of different tissues (chorionic villi, fetal skin, placenta, and neuronal cells of the adult brain). We have also investigated the efficiency of the approach in the differentiation of homologous chromosome heteromorphism by the quantitative analysis of alphoid DNA variation of chromosomes 13, 21, and X. In spite of many technical and theoretical limitations in the quantification of FISH signals, we were able to demonstrate the high reproducibility of the measurements after application of Scion Image software. We have concluded that this approach could be useful as an additional tool to visual microscopic analysis and to assist in correct FISH signal interpretation.
Peripheral blood lymphocytes were obtained from patients with and without chromosome abnormality previously detected by routine cytogenetic studies. Metaphase spreads and interphase nuclei were prepared from blood lymphocytes obtained from these patients according to a standard protocol previously described in detail (Yurov et al. 1996 The chorionic villi samples were obtained from the material of spontaneous abortions (510 weeks of gestation). These tissues were processed for FISH as follows: chorionic villi samples were washed in physiological solution three times. To clean the specimens of the rest of maternal deciduas and blood, samples were washed several times in 70% ethanol. They were then rinsed for 30 sec with 60% acetic acid and placed in solution of 60% acetic acid for 1520 min at room temperature and periodically mixed by inversion. Dispersed single-cell suspensions were fixed in a methanol-acetic acid (3:1) fixative mixture two times for 30 and 50 min. The cells were dropped onto wet slides and air dried at room temperature. Three slides with two drops of cell suspensions each were prepared for each sample.
Fetal skin, placenta, and samples of postmortem adult brain tissues were processed as described previously (Yurov et al. 2001a Permission from the Ethics Committee of the National Center of Mental Health, Russian Academy of Medical Sciences was obtained. Written informed consent was obtained from the patients for whom molecular cytogenetic studies were carried out.
The set of DNA probes from the original collection developed at the Laboratory of Cytogenetics, National Center of Mental Health, Moscow, Russia, and including chromosome enumeration DNA probes specific to chromosomes 1, 7, 8, 9, 13, and 21; 14 and 22; 15, 16, 18, X, and Y was used (Yurov et al. 1996 For epifluorescence microscopy, a Leitz Orthoplan microscope (Leica Mikroskopie und Systeme; Wetzlar, Germany) equipped with a 100 W lamp was used with the following filter sets (Leica Mikroskopie und Systeme): A (No. 513,596) for DAPI fluorescence; I3 (No. 513,719) or GR (No. 513,821) for fluorescein isothiocyanate signals; GR (No. 513,821) for both fluorescein and cyanine signals; N2 (No. 513,609) for cyanine signals. All images were observed with the Plan-Neofluotar x40/1.30 or x63/1.300.60 oil immersion lenses. The relative intensity of FISH signals was obtained by digital capturing of microscopic image by the monochrome CCD camera (Cohu 4910 series; Cohu Inc., San Diego, CA), LG-3 grayscale scientific PCI frame grabber (Scion Corporation; National Institutes of Health, Frederick, MD), and subsequent measuring of the intensity of hybridization signals by Scion Image Beta 4.0.2 (Scion Corporation) acquired from www.scioncorp.com (accessed 12/07/2001). The quantification of FISH signals from each digital image was processed by the macros supplied by the manufacturer. Numerical values of the signal relative intensity were compared with each other in the case of interphase FISH study. For the homologous chromosome differentiation, the ratio of the signal relative intensity from the digital image was obtained and compared with the value from another digital image. The reproducibility of the intensity measuring was assessed by several quantitative analyses (510 times) of the same interphase nucleus or metaphase spread. All interphase nuclei suspected to have one signal were subjected to quantitative assessment of FISH signals. Ten to 20 metaphase spreads were analyzed for each sample.
Interphase FISH: Studies of Low-level Mosaicism and Somatic Chromosome Pairing To test the efficiency of the proposed approach, we studied chromosome complement in interphase nuclei of chorionic villi, fetal skin, placenta, and neurons of the adult brain. Examples of quantification of FISH signals using Scion Image software are depicted in Figure 1. Examples of quantitative assessment of FISH signals in metaphase spreads and interphase nuclei are shown in Figure 2. The data demonstrate the high reproducibility of the measurement results, allowing the possibility of the differentiation of homologous chromosomes X, 21, and 16 by means of the study of alphoid or classical satellite DNA size heteromorphism in metaphase spreads and interphase nuclei.
Interphase FISH analysis of 23 spontaneous abortion specimens (chorionic villi, fetal skin, and placenta) were performed using quantification of FISH signals. The results of the chromosome complement study using quantitative assessment of FISH signals in the samples with the mosaic form of monosomy are shown in Table 1. The technique applied has allowed us to confirm the mosaic form of monosomy of chromosomes 14/22, 15, 16, 18, X, and Y in 12 cases. In these studies, a confidence interval (threshold of percentage) of 95% for abnormal chromosome complement nuclei with scoring no less than 500 interphase nuclei was chosen. This means that no fewer than 25 nuclei (5%) from 500 analyzed nuclei should have one FISH signal indicating the presence of monosomy (mosaic case). In the present study of 12 spontaneous abortion specimens, an abnormal cell population varied from 6% to 93% (Table 1). The cells with the signal over-position in these specimens were present as well in low percentages (15%, mean value 3.2%). The approach has been successfully applied for discrimination of single signals with increased intensity ("pseudo-monosomy") due to over-position of heterochromatic regions (Figures 3A and 3B). Control samples (samples considered as normal, without regular cell population with abnormal chromosome complement) should have less than 5% of aneuploid cells involving target chromosomes in aberrant sample. In this study, monosomy was not confirmed for 11 control samples in fetal skin (seven specimens) and placenta (four specimens). In these samples, the average percentage of the FISH signal over-position ranged from 1.4% to 4.2%, with a mean value of 3.1%. The frequency of probable monosomic nuclei with the "normal" intensity of signals was less than 1%.
We analyzed the chromosome complement in interphase nuclei of neurons in seven postmortem adult brain samples. DNA probes for chromosomes 1, 9, 13, 16, 18, 21, 22, X, and Y were applied. The regular forms of aneuploidy (including true monosomy) were not observed in any of the samples studied. However, a significant fraction of neuronal cell interphase nuclei had one signal per nucleus after hybridization with classical satellite DNA probes for heterochromatic regions of chromosomes 1, 9, and 16. The frequent appearance of one signal per nucleus in the brain tissues could be explained by the phenomenon of somatic pairing of homologous chromosomes shown previously for chromosomes 1 and 17 (Arnoldus et al. 1989
Differentiation of Homologous Chromosome by FISH with Chromosome-specific DNA Probes The differentiation of homologous chromosome X in females by FISH was found to be applicable to the molecular cytogenetic approach to the study of X-chromosome inactivation patterns. We studied 40 families (propositus and her mothers) with offspring affected by mental retardation not related to chromosome abnormality. An alphoid DNA probe specific for chromosome X was used. The example of quantitative assessment of variable chromosome X FISH signals used for differentiation of homologous chromosomes X is shown in Figure 1D and Figure 2A. The data show the high reproducibility of the measurement results, allowing the possibility of the differentiation of homologous chromosomes X by alphoid DNA heteromorphism. In 19 cases (47.5%), the approach of quantitative FISH signal assessment was found to be sufficiently efficient for the differentiation of homologous chromosomes X (data not shown).
Although there have been an increasing number of attempts at quantitative analysis of FISH signals, and an increased availability of a number of software programs for the registration, processing, and quantification FISH results, to date, there are no standardized or commonly recommended protocols for the quantitative analysis of FISH signals. As mentioned previously, in some routine diagnostic as well as research molecular cytogenetic studies, there is a special need for accurate FISH signal interpretation. The determination of FISH signal relative intensity or the relative intensity ratio could significantly improve the accuracy of data obtained. Many commercially available programs for FISH studies include capabilities for scoring or quantification of FISH signals. Details concerning protocols and efficiency of these FISH programs can be found in the descriptions of the software. We selected the software for image processing developed by the Scion Corporation for a number of reasons. First, this software is freely available and therefore can be used by anyone in FISH studies for non-commercial purposes. This means that results obtained can easily be reproduced by subsequent independent studies. In addition, Scion Image has been successfully applied for the quantification of electrophoresis gels in many laboratories (including our experience); therefore, positive results with this software application in FISH studies can be expected. Despite the fact that Scion Image application for FISH analysis has not been previously published, we decided to present our original results in this field. Our experience indicates that the quantification of FISH signals by Scion Image is a relatively simple and rapid approach for the quantitative assessment of FISH signals. The comparative analyses of the FISH signal relative intensity value in interphase nuclei of different types of tissues followed by discrimination of the over-posed (artifact) signals have been found to improve significantly the molecular cytogenetic scoring of chromosomal abnormalities. The reproducibility of the approach was assessed by five to ten analyses of the same interphase nucleus or metaphase spread. The variation of the signal relative intensity has been in the range of 20%, and is therefore adequate for quantitative comparison of FISH signals. The reproducibility of measurements between different cells of the same preparation is also high enough to be useful for the interpretation of signal scoring results. It should be emphasized that in all the samples studied, over-position of signals was observed. As a result, the presence of a small but sufficient population of nuclei with one hybridization signal (instead of the two signals expected in a normal diploid cell) produces the problem of interpretation of the results. Additionally, the phenomenon of somatic pairing of homologous chromosomes observed with a higher frequency in brain tissues causes a misinterpretation of the aneuploidy FISH scoring. Therefore, the lack of an approach for signal discrimination could lead to misdiagnosis in the cases of low-level mosaicism. In the present paper, we show that digital quantification (rarely used in FISH analysis of aneuploidy) aids in avoiding the scoring of "pseudo-monosomy," the result of over-position (or somatic chromosome pairing), leading to more accurate study results. The application of the approach has allowed us to exclude or confirm monosomy in the samples studied. Our results show that a confidence interval in interphase FISH studies of aneuploidy could probably be significantly less than 95%. Therefore, on the basis of the data we have obtained, we conclude that this approach is efficient enough for studies of low-level chromosomal mosaicism in different tissues.
An additional application of the proposed approach is the differentiation of homologous chromosome parental origin. Homologous chromosomes in normal chromosomal complement usually have no morphological differences (with the exception of acrocentric chromosomes with variable short arms or heteromorphism of chromosomes 1, 9, and 16 after C-banding). However, variations of alpha satellite DNA present in the centromeric regions of all human chromosomes can be visualized by in situ hybridization with chromosome-specific alphoid DNA probes (Yurov et al. 1987
Initially, the molecular cytogenetic technique based on the identification of heteromorphism of homologous X chromosomes was found to sufficient for the study of X-chromosome inactivation patterns in Rett syndrome (Yurov et al. 2001b
Supported by Copernicus 2 grant no. ICA2-CT-2000-10012 and INTAS grant no. 03-55-4060.
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 1718, 2004. Received for publication May 29, 2004; accepted September 23, 2004
Arnoldus EP, Noordermeer IA, Peters AC, Raap AK, Van der Ploeg M (1991) Interphase cytogenetics reveals somatic pairing of chromosome 17 centromeres in normal human brain tissue, but no trisomy 7 or sex-chromosome loss. Cytogenet Cell Genet 56:214216[Medline] Arnoldus EP, Peters AC, Bots GT, Raap AK, van der Ploeg M (1989) Somatic pairing of chromosome 1 centromeres in interphase nuclei of human cerebellum. Hum Genet 83:231234[CrossRef][Medline] Iourov IY, Vorsanova SG, Villard L, Kolotii AD, Yurov YB (2003) The study of X chromosome inactivation in mental retardation: comparative analysis of molecular-cytogenetic and polymerase chain reaction-based techniques in Rett syndrome. Balkan J Med Genet 6:3337 Levsky JM, Singer RH (2003) Fluorescence in situ hybridization: past, present and future. J Cell Sci 116:28332838 Pinkel D, Straume T, Gray JW (1986) Cytogenetic analysis using quantitative, high-sensitivity, fluorescent hybridization. Proc Natl Acad Sci USA 83:29342938 Soloviev IV, Yurov YB, Vorsanova SG, Fayet F, Roizes G, Malet P (1995) Prenatal diagnosis of trisomy 21 using interphase fluorescence in situ hybridization of postreplicated cells with site-specific cosmid and cosmid contig probes. Prenat Diagn 15:237248[Medline] Soloviev IV, Yurov YB, Vorsanova SG, Malet P (1994) Microwave activation of fluorescence in situ hybridization: a novel method for rapid chromosome detection and analysis. Focus 16:115116 Soloviev IV, Yurov YB, Vorsanova SG, Marcais B, Rogaev EI, Kapanadze BI, Brodiansky VM, et al. (1998) Fluorescent in situ hybridization analysis of Yurov YB, Mitkevich SP, Alexandrov IA (1987) Application of cloned satellite DNA sequences to molecular-cytogenetic analysis of constitutive heterochromatin heteromorphisms in man. Hum Genet 76:157164[CrossRef][Medline] Yurov YB, Soloviev IV, Vorsanova SG, Marcais B, Roizes G, Lewis R (1996) High resolution fluorescence in situ hybridization using cyanine and fluorescein dyes: ultra-rapid chromosome detection by directly fluorescently labeled alphoid DNA probes. Hum Genet 97:390398[Medline] Yurov YB, Vorsanova SG, Kolotii AD, Iourov IY (2001b) Molecular-cytogenetic investigation of skewed chromosome X inactivation in Rett Syndrome. Brain Dev 23(suppl 1):214217[CrossRef] Yurov YB, Vorsanova SG, Soloviev IV, Demidova IA, Alexandrov IA, Sharonin VO, Beresheva AK (2002) Original collection of DNA probes for preimplantational, fetal prenatal and postnatal diagnosis of chromosomal analysis by FISH. In Macek M Sr, Bianchi D, Cuckle H, eds. Early Prenatal Diagnosis, Fetal Cells and DNA in Mother. Present State and Perspectives. Prague, The Karolinum Press, 275283 Yurov YB, Vostrikov VM, Vorsanova SG, Monakhov VV, Iourov IY (2001a) Multicolor fluorescent in situ hybridization on post mortem brain in schizophrenia as an approach for identification of low-level chromosomal aneuploidy in neuropsychiatric diseases. Brain Dev 23(suppl 1):186190[CrossRef]
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||