Limits...
Chromosomal instability in near-diploid colorectal cancer: a link between numbers and structure.

Muleris M, Chalastanis A, Meyer N, Lae M, Dutrillaux B, Sastre-Garau X, Hamelin R, Fléjou JF, Duval A - PLoS ONE (2008)

Bottom Line: To address these questions, we karyotyped a consecutive series of 96 near-diploid colorectal cancers (CRCs) and distinguished chromosomal changes generated by either MSG or SR in tumor cells.Using a maximum likelihood statistical method, chromosomes affected by MSG or SR and likely to represent changes that are selected for during tumor progression were found to be different and mostly mutually exclusive.Taken together, these data provide new insights into the respective involvement of MSG and SR in near-diploid colorectal cancers, showing how these processes target distinct portions of the genome and result in specific patterns of chromosomal changes according to MSI status.

View Article: PubMed Central - PubMed

Affiliation: French National Institute for Health and Medical Research (INSERM), UMRS_762, Centre de Recherche (CDR) Saint-Antoine, Paris, France. martine.muleris@cephb.fr

ABSTRACT
Chromosomal instability (CIN) plays a crucial role in tumor development and occurs mainly as the consequence of either missegregation of normal chromosomes (MSG) or structural rearrangement (SR). However, little is known about the respective chromosomal targets of MSG and SR and the way these processes combined within tumors to generate CIN. To address these questions, we karyotyped a consecutive series of 96 near-diploid colorectal cancers (CRCs) and distinguished chromosomal changes generated by either MSG or SR in tumor cells. Eighty-three tumors (86%) presented with chromosomal abnormalities that contained both MSGs and SRs to varying degrees whereas all 13 others (14%) showed normal karyotype. Using a maximum likelihood statistical method, chromosomes affected by MSG or SR and likely to represent changes that are selected for during tumor progression were found to be different and mostly mutually exclusive. MSGs and SRs were not randomly associated within tumors, delineating two major pathways of chromosome alterations that consisted of either chromosome gains by MSG or chromosomal losses by both MSG and SR. CRCs showing microsatellite instability (MSI) presented with either normal karyotype or chromosome gains whereas MSS (microsatellite stable) CRCs exhibited a combination of the two pathways. Taken together, these data provide new insights into the respective involvement of MSG and SR in near-diploid colorectal cancers, showing how these processes target distinct portions of the genome and result in specific patterns of chromosomal changes according to MSI status.

Show MeSH

Related in: MedlinePlus

The two major chromosomal pathways in near-diploid CRCs.Schematic representation of the two-by-two association study presented in Supplementary Table S4. Positive (continuous lines) and negative (dotted lines) associations are represented by thick and thin lines indicating significant associations at p = 0.01 and 0.05 level, respectively. The size of each box is proportional to the frequency of the alteration, whole-chromosome or chromosome arm gains (red background), and chromosome losses or deletions (blue background) being distinguished. The preferential associations between alterations involving both arms of the same chromosome, such as -8p and +8q, or -17p and +17q, need to be interpreted with caution since they originate mostly from a single isochromosome formation event. For instance, gain of 17q seems to be a side effect of 17p deletion via isochromosome 17q formation (10/11 tumors with 17q gain also exhibited 17p loss whereas 17p loss alone was observed in an additional 30 tumors) whereas both 8p loss and 8q gain might provide tumor cells with a selective advantage since although associated in 20 tumors, these alterations were also observed independently in 8 and 7 tumors, respectively (see also Supplementary Figure S1). Some negative associations (between +13 and +13q, +8 and +8q, +8 and -8p) could also be due to the fact they involve the same chromosome.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2238794&req=5

pone-0001632-g003: The two major chromosomal pathways in near-diploid CRCs.Schematic representation of the two-by-two association study presented in Supplementary Table S4. Positive (continuous lines) and negative (dotted lines) associations are represented by thick and thin lines indicating significant associations at p = 0.01 and 0.05 level, respectively. The size of each box is proportional to the frequency of the alteration, whole-chromosome or chromosome arm gains (red background), and chromosome losses or deletions (blue background) being distinguished. The preferential associations between alterations involving both arms of the same chromosome, such as -8p and +8q, or -17p and +17q, need to be interpreted with caution since they originate mostly from a single isochromosome formation event. For instance, gain of 17q seems to be a side effect of 17p deletion via isochromosome 17q formation (10/11 tumors with 17q gain also exhibited 17p loss whereas 17p loss alone was observed in an additional 30 tumors) whereas both 8p loss and 8q gain might provide tumor cells with a selective advantage since although associated in 20 tumors, these alterations were also observed independently in 8 and 7 tumors, respectively (see also Supplementary Figure S1). Some negative associations (between +13 and +13q, +8 and +8q, +8 and -8p) could also be due to the fact they involve the same chromosome.

Mentions: We next investigated for possible preferential associations amongst the most frequent chromosome imbalances generated by either MSG or SR, i.e. those that were likely to be selected for in CRCs according to the likelihood method. A systematic analysis of two-by-two associations between 29 chromosomal imbalances (406 possibilities) demonstrated 51 associations (45 positive and 6 negative) that were significant in our tumor series (p = 0.05) (Supplementary Table S4). We are aware that considering that 406 analyses were performed, it could be expected that 20 out of these 51 significant associations were observed by chance alone (false positive) at the p = 0.05 level. Using a p = 0.01 level, 16 significant associations were found (Figure 3) that is four times more than the number of false positive expected, validating thus the existence of preferential associations. A p = 0.05 level was retained for the analysis. Amongst the 45 positive associations, most (84%) involved exclusively losses or gains (27 and 11, respectively) compared to 7 that showed gains mixed with losses (16%). Twenty (44.5%) showed associations of MSG with SR events compared to 15 (33.3%) and 10 (22.2%) that involved exclusively MSGs or SRs, respectively. By combining all results from two-by-two associations of chromosomal alterations observed in CRCs, two main groups of chromosomal abnormalities were delineated (Figure 3). The first group includes only whole-chromosome gains resulting from MSG (+13, +20, +7, +X, +12 and +6) whereas the second group mainly consists of chromosomal losses through MSG and SR (-18, -17p, -Y, -1p3, -8p, +8q, -14, +13q, -15, -4, +17q, +8, +6, -10q2, -11q, -9p, -8, -4p, -11p, -9q and -6, in decreasing order of occurrence). One minor group not related to the former two groups consisted of a preferential association between -4q and -5q.


Chromosomal instability in near-diploid colorectal cancer: a link between numbers and structure.

Muleris M, Chalastanis A, Meyer N, Lae M, Dutrillaux B, Sastre-Garau X, Hamelin R, Fléjou JF, Duval A - PLoS ONE (2008)

The two major chromosomal pathways in near-diploid CRCs.Schematic representation of the two-by-two association study presented in Supplementary Table S4. Positive (continuous lines) and negative (dotted lines) associations are represented by thick and thin lines indicating significant associations at p = 0.01 and 0.05 level, respectively. The size of each box is proportional to the frequency of the alteration, whole-chromosome or chromosome arm gains (red background), and chromosome losses or deletions (blue background) being distinguished. The preferential associations between alterations involving both arms of the same chromosome, such as -8p and +8q, or -17p and +17q, need to be interpreted with caution since they originate mostly from a single isochromosome formation event. For instance, gain of 17q seems to be a side effect of 17p deletion via isochromosome 17q formation (10/11 tumors with 17q gain also exhibited 17p loss whereas 17p loss alone was observed in an additional 30 tumors) whereas both 8p loss and 8q gain might provide tumor cells with a selective advantage since although associated in 20 tumors, these alterations were also observed independently in 8 and 7 tumors, respectively (see also Supplementary Figure S1). Some negative associations (between +13 and +13q, +8 and +8q, +8 and -8p) could also be due to the fact they involve the same chromosome.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2238794&req=5

pone-0001632-g003: The two major chromosomal pathways in near-diploid CRCs.Schematic representation of the two-by-two association study presented in Supplementary Table S4. Positive (continuous lines) and negative (dotted lines) associations are represented by thick and thin lines indicating significant associations at p = 0.01 and 0.05 level, respectively. The size of each box is proportional to the frequency of the alteration, whole-chromosome or chromosome arm gains (red background), and chromosome losses or deletions (blue background) being distinguished. The preferential associations between alterations involving both arms of the same chromosome, such as -8p and +8q, or -17p and +17q, need to be interpreted with caution since they originate mostly from a single isochromosome formation event. For instance, gain of 17q seems to be a side effect of 17p deletion via isochromosome 17q formation (10/11 tumors with 17q gain also exhibited 17p loss whereas 17p loss alone was observed in an additional 30 tumors) whereas both 8p loss and 8q gain might provide tumor cells with a selective advantage since although associated in 20 tumors, these alterations were also observed independently in 8 and 7 tumors, respectively (see also Supplementary Figure S1). Some negative associations (between +13 and +13q, +8 and +8q, +8 and -8p) could also be due to the fact they involve the same chromosome.
Mentions: We next investigated for possible preferential associations amongst the most frequent chromosome imbalances generated by either MSG or SR, i.e. those that were likely to be selected for in CRCs according to the likelihood method. A systematic analysis of two-by-two associations between 29 chromosomal imbalances (406 possibilities) demonstrated 51 associations (45 positive and 6 negative) that were significant in our tumor series (p = 0.05) (Supplementary Table S4). We are aware that considering that 406 analyses were performed, it could be expected that 20 out of these 51 significant associations were observed by chance alone (false positive) at the p = 0.05 level. Using a p = 0.01 level, 16 significant associations were found (Figure 3) that is four times more than the number of false positive expected, validating thus the existence of preferential associations. A p = 0.05 level was retained for the analysis. Amongst the 45 positive associations, most (84%) involved exclusively losses or gains (27 and 11, respectively) compared to 7 that showed gains mixed with losses (16%). Twenty (44.5%) showed associations of MSG with SR events compared to 15 (33.3%) and 10 (22.2%) that involved exclusively MSGs or SRs, respectively. By combining all results from two-by-two associations of chromosomal alterations observed in CRCs, two main groups of chromosomal abnormalities were delineated (Figure 3). The first group includes only whole-chromosome gains resulting from MSG (+13, +20, +7, +X, +12 and +6) whereas the second group mainly consists of chromosomal losses through MSG and SR (-18, -17p, -Y, -1p3, -8p, +8q, -14, +13q, -15, -4, +17q, +8, +6, -10q2, -11q, -9p, -8, -4p, -11p, -9q and -6, in decreasing order of occurrence). One minor group not related to the former two groups consisted of a preferential association between -4q and -5q.

Bottom Line: To address these questions, we karyotyped a consecutive series of 96 near-diploid colorectal cancers (CRCs) and distinguished chromosomal changes generated by either MSG or SR in tumor cells.Using a maximum likelihood statistical method, chromosomes affected by MSG or SR and likely to represent changes that are selected for during tumor progression were found to be different and mostly mutually exclusive.Taken together, these data provide new insights into the respective involvement of MSG and SR in near-diploid colorectal cancers, showing how these processes target distinct portions of the genome and result in specific patterns of chromosomal changes according to MSI status.

View Article: PubMed Central - PubMed

Affiliation: French National Institute for Health and Medical Research (INSERM), UMRS_762, Centre de Recherche (CDR) Saint-Antoine, Paris, France. martine.muleris@cephb.fr

ABSTRACT
Chromosomal instability (CIN) plays a crucial role in tumor development and occurs mainly as the consequence of either missegregation of normal chromosomes (MSG) or structural rearrangement (SR). However, little is known about the respective chromosomal targets of MSG and SR and the way these processes combined within tumors to generate CIN. To address these questions, we karyotyped a consecutive series of 96 near-diploid colorectal cancers (CRCs) and distinguished chromosomal changes generated by either MSG or SR in tumor cells. Eighty-three tumors (86%) presented with chromosomal abnormalities that contained both MSGs and SRs to varying degrees whereas all 13 others (14%) showed normal karyotype. Using a maximum likelihood statistical method, chromosomes affected by MSG or SR and likely to represent changes that are selected for during tumor progression were found to be different and mostly mutually exclusive. MSGs and SRs were not randomly associated within tumors, delineating two major pathways of chromosome alterations that consisted of either chromosome gains by MSG or chromosomal losses by both MSG and SR. CRCs showing microsatellite instability (MSI) presented with either normal karyotype or chromosome gains whereas MSS (microsatellite stable) CRCs exhibited a combination of the two pathways. Taken together, these data provide new insights into the respective involvement of MSG and SR in near-diploid colorectal cancers, showing how these processes target distinct portions of the genome and result in specific patterns of chromosomal changes according to MSI status.

Show MeSH
Related in: MedlinePlus