Heterozygous deletion of the PU.1 locus in human AML

The lineage-specific transcription factor PU.1 is essential for myeloid development, and its disruption leads to a block of myeloid and B-cell development.^1-4  Consequently, PU.1 expression is tightly regulated.^5-9  Several studies have indicated that PU.1 function is suppressed in subsets of acute myeloid leukemia (AML) patients, such as in the presence of the leukemic fusion proteins PML-RARA or AML1-ETO.^10,11  In addition, genomic mutations of PU.1 in AML patients have been reported.^11-16 

A highly conserved distal upstream regulatory element (URE) is mediating transcriptional control of Pu.1 expression. Targeted disruption of this element suppresses Pu.1 expression to 80% and induces AML in mice.^17-20  Here, we hypothesized that mutations within the URE could cause dysregulation of PU.1 and thus contribute to leukemogenesis in human AML. We therefore analyzed the URE in AML patients by direct sequencing. We identified one AML patient with heterozygous loss of the entire URE sequences. Our data suggest that heterozygous deletion of the PU.1 locus can be associated with human AML.

DNA was extracted from blasts of 120 AML patients at diagnosis. Characteristics of the patients are depicted in supplemental Table 1 (available on the Blood website; see the Supplemental Materials link at the top of the online article). In addition, DNA was studied from peripheral blood of 141 healthy volunteers, with informed consent of patients and volunteers in accordance with the Declaration of Helsinki. The studies were approved by the cantonal ethics committee of Bern, Switzerland.

The primers for distal regulatory unit (DRU) were (607 bp): 5′-AGAGGAAACTGAGGCCAAGTG-3′ and 5′-TGGCAGTCCTCACTGAGGCCATTG-3; and for PRU (634 bp) 5′-CAATGGCCTCAGTGAGGACTGCCA-3′ and 5′-TCTTGGCGGAAGCTGTTAGGGAAG-3′. Sequencing reactions were performed on an Applied Biosystems 3730 Analyzer (Applied Biosystems).

The primers were as follows: 5′-TGTTACAGGCGTGCAAAATGGAAGG-3′, 5′-CTCGTGCGTTTGGCGTTGGTATAGA-3′; and 5′-FAM-CCTCGTCCCCCCTCCATCAGAAGACCTGG-TAMRA-3′. Expression levels were obtained using standard curves, and normalized to ABL expression.

For locus-specific fluorescence in situ hybridization (FISH), 2 probes were used: a polymerase chain reaction (PCR)-generated 1.4-kb fragment covering the entire URE sequences (red signal) and a probe directed against the centromer of chromosome 11 (green signal). Labeling, probe preparation, and hybridization of slides were performed as previously described.^21,22 

We analyzed the DRU and proximal regulatory unit (PRU) of the URE of the hematopoietic master transcription factor PU.1 for genomic mutations by direct sequencing (Figure 1). We screened DNA from malignant cells of 120 AML patients and peripheral blood of 141 healthy volunteers, and no mutations were detected. This suggests that genomic mutation in the URE of PU.1 is a rare event in human AML.

We observed one single nucleotide polymorphism (SNP) in the DRU (DRU1) and 3 SNPs in the PRU (PRU1-3) of the URE sequences as illustrated in Figure 1. DRU1 and PRU3 correspond to the 2 SNPs described by Steidl et al.²¹  We could not detect a higher incidence of the SNP DRU1 in AML patients with complex abnormalities (n = 24) as previously reported,²⁰  possibly because of the small size of this group.

In mice, targeted disruption of the entire URE sequences suppresses Pu.1 mRNA by 80% and induces AML.¹⁹  We therefore hypothesized that loss of heterozygosity (LOH) of the URE might be observed in AML patients. Accordingly, we focused on AML patients with noninformative SNPs in all 4 locations of the URE. We identified homozygosity for all SNPs in 37 of 120 patients (30.8%) and in 47 of 141 healthy volunteers (33.3%).

Next, we analyzed samples obtained at diagnosis and at remission in the patients with noninformative SNPs in the URE. Appropriate material at remission was not available in 14 of the 37 patients because of early death (8 patients), treatment failure and thus unavailability of remission material (5 patients), or loss during follow-up (1 patient). In 22 patients, identical results were observed at diagnosis and remission. However, 1 patient showed a heterozygous pattern in the SNPs DRU1, PRU1, and PRU3 at remission. This result was verified in a second remission sample obtained at a later time point (Figure 2). These data indicate LOH of the URE sequences in leukemic cells at diagnosis.

The cytogenetic analysis of 20 metaphases in this patient indicated a normal karyotype at diagnosis. No mutations were detected in the genes encoding CEBPA, FLT3, and NPM1. The patient was diagnosed at the age of 32 years with AML-M2, the blasts were positive for CD11b, CD13, CD33, CD34, and HLA-DR, and leukocytes at diagnosis were 2.3 g/L, with 3% blasts. Hemoglobin was 92 g/L, platelets 72 g/L, and lactate dehydrogenase 398 U/mL. The bone marrow indicated 80% blast infiltration. The patient was treated with 3 cycles of chemotherapy and enjoys ongoing complete remission.

Because the karyotype was normal, LOH detected in SNP of the URE suggests a submicroscopic deletion in band 11p11.2. However, the size of the deletion remained to be determined. We therefore investigated whether the PU.1 coding region located 17 kb downstream of the URE might also be affected. We sequenced 625 basepairs of the 5′ PU.1 promoter region (PROM) starting immediately upstream of the translation start site in all 120 AML patients and in the healthy volunteers. Again, no mutations were detected. Remarkably, the patient with LOH in the URE sequences also showed homozygosity in all 4 promoter SNPs at diagnosis, whereas heterozygosity was found for PROM1 and PROM3 at remission. We also tested 2 additional SNPs located in intron 3 (SNPi3) and intron 4 (SNPi4) both just flanking exon 4. Both SNPs were homozygous at diagnosis but heterozygous at remission. This indicates that the deletion spans at least from the URE to exon 4, thus covering 40 kb.

Additional SNP analysis within the telomeric region of chromosome 11p and the centromeric region of chromosome 11 indicated conserved heterozygosity of both regions at diagnosis (supplemental Figure 1). Furthermore, SNPs were analyzed located 20, 50, and 100 kb upstream and downstream of the PU.1 translation start site. Interestingly, heterozygosity was conserved at diagnosis in an SNP 20 kb upstream, narrowing down the upstream breakpoint of the deletion between 20 kb and 17 kb (location of the URE), whereas the downstream distant SNPs flanking the LOH of the PU.1 locus were not informative.

We ultimately confirmed the deletion by FISH analysis using a probe that corresponded to the URE region (Figure 2). The probe showed loss of one signal in leukemic cells at diagnosis, whereas 2 signals were detectable using a probe against the centromeric region of chromosome 11.

Finally, we assessed PU.1 mRNA expression by real-time PCR available in 107 AML patients (Figure 2). We observed that the AML-M2 patient with the heterozygous PU.1 deletion expressed significantly lower PU.1 mRNA compared with our collection of AML-M2 patients, and the expression was 19% of the mean expression observed in AML-M0 patients, which served as a reference (100%).

Our data indicate that heterozygous deletion of the PU.1 locus can be associated with AML; however, this appears to be a rare event. Submicroscopic deletions spanning several kilobases as identified in our patient will necessarily escape the detection by conventional sequencing strategies. Our findings may therefore contribute to further clarify divergent reports on the incidence of PU.1 mutations in AML patients.^11-15,23 

The authors thank Andreas Steege for help with the SNP analysis, Ludwig Wilkens for technical assistance with the FISH analyses, and Marianne Eyholzer for help with sample preparation.

This work was supported by the Swiss National Science Foundation (SF 310000-113761; B.U.M.) and the Swiss Cancer League (OCS 01731082005; B.U.M.).

Contribution: N.B. performed research; T.P. analyzed data and wrote the paper; and B.U.M. designed research, analyzed data, and wrote the paper.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Beatrice U. Mueller, Department of Internal Medicine, University Hospital, 3010 Bern, Switzerland; e-mail: beatrice.mueller@insel.ch.