Two novel variants of MOZ-CBP fusion transcripts in spontaneously remitted infant leukemia with t(1;16;8)(p13;p13;p11), a new variant of t(8;16)(p11;p13)

Translocation t(8;16)(p11;p13) involving the MOZ/MYST3 gene (8p11) and the CBP/CREBBP gene (16p13) is associated with the FAB M4/M5 subtype of acute myeloid leukemia (AML) and a poor prognosis. Five types of MOZ-CBP and three types of CBP-MOZ fusion transcripts have been identified in adult and adolescent patients by reverse transcriptase-polymerase chain reaction (RT-PCR).1,2 To date, 6 newborn infants with monocytic malignancies (five AML-M4/M5 and one myelosarcoma) associated with t(8;16)(p11;p13) have been reported.3–5 In these patients, only one AML patient had a variant of this translocation, t(8;16)(q11;p13).4 Surprisingly, 2 AML cases, including the case with t(8;16)(q11;p13) and the case of myelosarcoma, underwent spontaneous remission, indicating a more favorable outcome than older patients. Although the involvement of the MOZ gene and the CBP gene was indicated by fluorescence in situ hybridization (FISH) analysis in one patient,5 the presence of fusion transcripts was not described in any of the 6 cases. Therefore, we analyzed a case of infant AML with a variant of t(8;16)(p11;p13) by RT-PCR to determine whether MOZ-CBP and/or CBP-MOZ fusion transcripts were also involved in the development of infant leukemia.

A newborn girl presented with multiple skin nodules and hepatosplenomegaly after birth and was referred to our department on day 3 of life. Peripheral blood analysis showed a white blood cell count of 54.18×10/L with 6% blasts and 58% immature monocytic cells, a hemoglobin level of 14.1 g/dL, and a platelet count of 54×10/L. Bone marrow was hyperplastic with 44% blasts, and a diagnosis of AML (FAB M5b) was made. Surface marker analysis showed that the leukemic cells expressed CD4, CD13, CD14, CD33, and HLA-DR antigen. Cytogenetic study of bone marrow cells revealed an abnormal karyotype of 46,XX,t(1;16;8)(p13;p13;p11) in 17 out of 20 metaphases analyzed (Figure 1A). FISH analysis using probe sets for the centromeric and telomeric regions of the MOZ gene (BAC clones RP11-451C05 and RP11-142G13 labeled with Spectrum Green) and the CBP gene (RP11-507M07 and RP11-6C20 labeled with Spectrum Orange) demonstrated a fusion signal (Figure 1B). The skin nodules spontaneously regressed, and the patient’s clinical condition remained good, so we observed the patient without chemotherapy. She spontaneously underwent complete remission before three months of age and remains in remission 17 months after diagnosis.

Figure 1.Cytogenetic analysis of the patient’s bone marrow cells. (A ) The G-banded karyotype of the leukemic cells showing 46, XX, t(1;16;8)(p13;p13;p11). Three arrows indicate the rearranged chromosomes. (B) Dual-color FISH analysis using probe sets for the centromeric and telomeric regions of MOZ (labeled with Spectrum Green) and CBP (labeled with Spectrum Orange) demonstrated a fusion signal (a yellow signal indicated by an arrow head). Thin arrows and thick arrows indicate the signals for MOZ and CBP respectively. Each probe set was composed of two BAC clones: RP11-451C05 (centromeric) and RP11-142G13 (telomeric) for MOZ, and RP11-507M07 (centromeric) and RP11-6C20 (telomeric) for CBP.

Total RNA was extracted from bone marrow cells and cDNA was prepared. Since RT-PCR using previously published primer sets, MOZ3558F and CBP1201R, MOZ3536F and CBP404R,1 and MYST3 3319F and CBP1201R2 were negative for MOZ-CBP fusion, we designed a new primer set in exon 13 of the MOZ gene (MOZ2596F: 5’-cgaatgctggacttccgtagtg-3’) and a reverse primer in exon 9 of the CBP gene (CBP1904R: 5’-cca-gagttggagccatcgttc-3’). Two DNA fragments, a major fragment of 1,526 bp and a minor of 1,283 bp, were successfully amplified using this primer set (Figure 2A, left). RT-PCR using a primer combination of MYST3 3319F and CBP1904R also amplified two fragments of 815 bp and 572 bp (Figure 2A, right). Sequencing analysis revealed two novel in-frame fusion transcripts, a major fusion transcript between MOZ exon 16 and CBP exon 7 and a minor between MOZ exon 16 and CBP exon 8, indicating alternative splicing at the CBP gene (Figure 2B). The fusion transcripts were still positive at the age of one month but negative at two months. To detect reciprocal CBP-MOZ fusion transcripts, nested RT-PCR with the primer sets of CBP96F and MOZ3953R (first step) and CBP174F and MOZ3844R (second step) were performed as previously described.1 However, no amplification products were obtained.

To our knowledge, these fusion transcripts will be the sixth and seventh types of MOZ-CBP, demonstrating the heterogeneity of the fusion transcripts. Although amplification of MOZ-CBP and CBP-MOZ fusion transcripts by RT-PCR has been reported to be difficult,1,2 some of the cases negative for RT-PCR may have fusion transcript variants that could not be detected by previously reported primers. Our primer combinations may help to detect MOZ-CBP fusion transcripts in more AML cases with t(8;16)(p11;p13). CBP-MOZ fusion transcripts have not always been detected in MOZ-CBP-positive patients, indicating that MOZ-CBP is responsible for the leukemogenesis. However, it remains unclear whether CBP-MOZ fusion transcripts contribute to the development of leukemia because it has not been excluded that CBP-MOZ transcripts failed to be detected for technical reasons.6 CBP-MOZ fusion transcripts were also negative in our patient, but false negatives caused by technical problems were excluded because der(16)t(8;16) encoding CBP-MOZ was not generated in this three-way translocation. Translocation t(8;18;16)(p11;q21;p13), another three-way translocation variant of t(8;16)(p11;p13) found in a 15-month-old boy with AML M5b, also has a structure that could not generate CBP-MOZ but MOZ-CBP,while R PCR was not performed in this case.7 Moreover, Murati et al. reported a 72-year-old man with AML M5a having a complex t(8;16)(p11;p13), in whom only MOZ-CBP but not CBP-MOZ was detected as expected because of the insertion of 8q material between 16p and 8p on the der(16).2 These results clearly show that the CBP-MOZ fusion transcripts are not essential for the development of AML with t(8;16)(p11;p13) and its variants.

Figure 2.Analysis of the MOZ-CBP fusion transcripts. (A) RT-PCR using a primer set of MOZ2596F and CBP1904R amplified a major fragment of 1,526 bp and a minor of 1,283 bp (left). Two fragments of 815 bp and 572 bp were also amplified with a primer combination of MYST3 3319F and CBP1904R (right). M, size marker; lane 1, present case; lane 2, negative AML case; lane 3, no template. (B) Sequencing analysis revealed that MOZ exon 16 was fused in-frame to CBP exon 7 in the major fusion transcript (top) and also in frame to CBP exon 8 in the minor (bottom). ( C ) Predicted novel MOZ-CBP fusion proteins lack C/H1 domain of CBP compared to previously reported ones, but retain most of the other functional domains, including the bromodomain and the HAT domain. The thick arrows with the bolded and underlined numbers 6 and 7 indicate the breakpoints of the novel fusions. The arrows indicate the breakpoints of the type I-V MOZ-CBP with the numbers 1–5. MYST: MYST domain; acidic: acidic domain; NR: nuclear receptor binding domain; C/H 1–3: cysteine/histidine-rich domains 1–3; KIX: KIX domain; Br: bromodomain; Q-rich: glutamine-rich domain; HAT: histone acetyltransferase domain. *Two breakpoints in MOZ exon 17 were reported in the type III fusion.

We find for the first time that MOZ-CBP fusion transcripts are expressed in a case of infant leukemia with a variant of t(8;16)(p11;p13), and the fusion transcripts were shown to be two novel variants. Although predicted MOZ-CBP fusion proteins lack a protein interaction domain, C/H1 (cysteine/histidine-rich domain 1), compared to previously reported ones, these fusion proteins retain most of the other functional domains of CBP, including the bromodomain and the HAT domain (Figure 2c) which were reported to be important for MOZ-CBP to inhibit the Runx1-mediated transcription and myeloid cell differentiation.8 This suggests that the novel MOZ-CBP proteins are also leukemogenic and contribute to leukemic development in this case. While the reason for the favorable outcome of infant leukemia with t(8;16)(p11;p13) is not clear, secondary mutations may be required to develop aggressive disease as observed in adult patients. The difference of MOZ-CBP structures between infant and adult patients might also be related to the different outcomes. To clarify these questions, more cases of both infant and adult AML with t(8;16)(p11;p13) need to be analyzed.

Footnotes

• Funding: this work was supported by Grants-in-Aid for Scientific Research and Grants-in-Aid for Scientific Research on Priority Areas from the Ministry of Education, Science, Sports and Culture. Keyw ords: MOZ-CBP, CBP-MOZ, t(8;16), infant leukemia.