Back to Search Start Over

Array CGH analysis of nodal T-Cell lymphomas: Identification of genomic alterations specific to angioimmunoblastic and unspecified subtypes, and correlation with transcriptomic data

Authors :
Philippe Gaulard
Aurélien de Reyniès
Emilie Thomas
Louis Huang
Thierry Molina
Josette Brière
Laurence de Leval
Christian Gisselbrecht
David S. Rickman
Luc Xerri
Francoise Berger
Source :
Publons

Abstract

Genetic alterations underlying angioimmunoblastic and unspecified peripheral T-cell lymphomas (AITL and PTCL-u) are largely unknown. Seventeen AITL and 16 PTCL-u previously characterized by gene expression profiling, were analyzed by CGH on DNA microarrays comprising 4434 BAC clones with a resolution of about 600 KB. In the PTCL-u group, the mean number of chromosomal aberrations per case was 302 (range, 55 to 892). Gains (n=237, 41 to 587) were more frequent than losses (n=65, 8 to 305). AITL samples had, on average, a lesser number of genomic alterations than PTCL-u cases (n=243, range, 55 to 485), comprising more gains (n=201, 42 to 541) than losses (n=42, 9 to 262). Overall, the most frequent recurrent gains, present in 50% of all samples, were observed at 1p36.1; 1p36.3 ; 1q32 ; 2q37 ; 4p16 ; 5p15.3 ; 6q12 ; 7p22 ; 7p12 ; 7p11.2 ; 7q35-36 ; 8q24.3 ; 9q34 ; 11p15 ; 11q13 ; 16p13.3 ; 16q24 ; 17q12,q21,q25 ; 19p13.3 ; 19q13.2-q13.3 ; 20q11.2-q13.3 ; 22q11.1-q11.2 ; Xp11 ; Xp21-22 ; Xq27-28. The comparison of the genomic profiles of AITL and PTCL-u identified 73 genomic alterations (at clones or zones of clones) significantly associated with either group of tumors (Fisher test, p < 0,05). Six genomic gains mapping at 5p15 and 22q11 were associated with the AITL subtype. Thirty-four gains (mapping at 6p25, 7p1, 7q3, 8q24, 11p14, 14q32, 17q, 22q) and 33 losses (mapping at 6q, 10p and 13q), were overrepresented in PTCL-u. The coordinate analysis of the transcriptomic and CGH array data identified 10 regions with genomic imbalances containing genes differentially expressed in AITL versus PTCL-u. Seven regions amplified in PTCL-u contained genes overexpressed in PTCL-u, mostly related to metabolic pathways. Conversely, loss of genomic material at 13q12 correlated with decreased expression in PTCL-u of a few genes of the AITL signature. In AITL tumors, gain at 22q11 correlated with increased transcription of the LIF gene, previosuly characterized as part of the tumor cell signature in AITL. CD30+ PTCL-u samples had on average a higher number of genomic aberrations than CD30-negative cases (n=408 versus 238). Thirty-three genomic gains and 22 losses were exclusively seen in CD30+ tumors, and regions with chromosomal imbalances at 1q, 6q, 10p contained genes differentially expressed in CD30+ and CD30− tumors. For example, reduced transcription of FYN in CD30+ PTCL-u correlated with deletion of the corresponding chromosomal region. In conclusion, all 33 nodal PTCL analyzed harbor genomic imbalances (gains>losses), of which many are common to both AITL and PTCL-u subgroups; the pattern of genomic aberrations differs between the two subgroups, with certain aberrations being overrepresented in PTCL-u, and only a few specific for AITL; coordinate appraisal of transcriptomic and genomic data highlights correlations between genomic imbalances and gene expression signatures in subgroups of tumors.

Details

Database :
OpenAIRE
Journal :
Publons
Accession number :
edsair.doi.dedup.....42702173983c984892a956f9c8075d55