Identification of a novel PML exon 6 splice variant in atypical PML::RARα transcripts in acute promyelocytic leukemia

Abstract

The PML::RARα fusion gene resulting from the t(15;17) chromosomal translocation serves as the pathognomonic molecular marker of acute promyelocytic leukemia (APL), which has been shown to directly repress transcription of retinoic acid (RA)-responsive genes, ultimately inducing granulocytic differentiation arrest. In this study, we report an APL case harboring an atypical PML::RARα fusion transcript characterized by a novel splice site variant (GCCaggccc) within PML exon 6, resulting in an 80 base pairs deletion of the distal exonic sequence with concomitant insertion of 23 exogenous nucleotides (agagccttcttctctctgggacaag). To our knowledge, this isoform differs from all previously described PML::RARα fusion transcripts. This case emphasizes the importance of molecular characterization in APL diagnosis and minimal residual disease (MRD) monitoring, though further studies are required to establish its clinical correlation.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00277-025-06613-6.

Introduction

Acute Promyelocytic Leukemia (APL), a clinically distinct molecular subtype of Acute Myeloid Leukemia (AML), accounts for approximately 10–15% of all AML cases [1]. This hematologic emergency frequently accompanied by potentially severe coagulation dysfunction and thrombosis, with hemorrhagic complications being the primary cause of early mortality in patients. Prompt diagnosis and treatment can lead to a survival rate exceeding 95% for low-risk patients (leukocyte WBC count ≤ 10 × 10^9/L) [2, 3]. The majority of APL cases are characterized by the classical t(15;17) chromosome translocation, which generates the PML::RARα fusion gene. The fusion protein results in inhibited cell differentiation and insufficient apoptosis [4], forming the main molecular mechanism underlying the development of APL.

The typical PML::RARα fusion gene features the breakpoint cluster region (BCR) of the RARα gene situate within intron 2. Depending on the PML breakpoint location (intron 6, exon 6, or intron 3), the fusion gene is classified into three subtypes: bcr1, bcr2, and bcr3, also known as the long form (L-form), variant form (V-form), and short form (S-form), respectively. The bcr1 and bcr3 subtypes are generated by the splicing of RARα exon 3 with PML exon 6 or exon 3, respectively. The bcr2 subtype originates from a more complex splicing mechanism that creates a novel exon-exon junction between a cryptic donor splice site (GAAgtgagg) in PML exon 6 and the acceptor splice site of RARα exon 3 [5, 6]. Notably, over 30 atypical PML::RARα subtypes have been identified to date, predominantly associate with the V-form subtype [7, 8]. These atypical V-form subtypes share similar genetic architectures, characterized by deletions in the 3’ region of PML exon 6 (spanning from 8 to 146 nucleotides) accompanied by insertion of foreign DNA fragment, primarily originating from RARα intron 2 (3 to 127 nucleotides) [7]. Furthermore, in t(15;17)-negative variant APL patients, rearrangements involving RARβ or RARγ [9], as well as fusions between RARα and genes such as NPM1, NUMA1, STAT5B, PRKAR1A, FIP1L1, BCOR, and NABP1, have been reported [10]. The PML::RARα fusion protein not only contributes to leukemogenesis but also mediates the differentiation response to all-trans retinoic acid (ATRA) [11]. The distinct structural domains of PML retained in different PML::RARα subtypes may lead to varying clinical outcomes, influencing functionality and treatment responsiveness [12, 13], thereby holding prognostic significance. Although the biological implications of many of these variants remained debated, analyzing their sequences was crucial for understanding pathogenic molecular mechanisms and monitoring minimal residual disease (MRD).

In this study, we characterized the bone Marrow morphology, flow cytometry, cytogenetic and clinical features of a 48-year-old female patient with APL. The presence of the PML::RARα fusion gene was detected through reverse transcription polymerase chain reaction (RT-PCR), presenting three fusion gene transcripts. Sequence analysis revealed a new breakpoint within exon 6 of PML (Splice Site: ctctagCCA) with a distal 80 base pairs deletion and insertion of an exogenous DNA fragment (agagccttcttctctctgggacaag). Additionally, the RARα breakpoint site was Mapped to intron 2 (Splice Site: ctctagCCA), establishing a mechanistically distinct PML::RARα fusion configuration that diverges from all previously reported isoforms.

Materials and methods

Clinical data

A 48-year-old female patient presented to our hospital with gingival bleeding and a sore throat, both for no apparent cause. Table 1 summarized the patient’s initial peripheral blood tests, including complete blood count, coagulation tests, and blood chemistry analysis.

Table 1.

Clinicopathologic features of the patient

	Patient	Normal range
Complete Blood Count
WBC(×109/L)	1.7	4–10
Blasts in peripheral blood (%)	53	< 0.01
Hemoglobin (g/L)	95	115–150
Platelets×109/L	48	125–350
Blood Coagulation Profile
PT (s)	15.7	9.8–12.1
APTT (s)	27	22.7–31.8
TT (s)	20.1	14.0–21.0
Fbg (g/L)	0.42	1.8–3.5
D-dimer (mg/L)	33.99	0.01–0.55
Blood Chemistry Analysis
TBil (µmol/L)	7.3	3.42–20.5
DBil (µmol/L)	5.0	0–6.84
LDH (U/L)	198	120–250

WBC: White Blood Cell Count; PT: Prothrombin Time; APTT: Activated Partial Thromboplastin Time; TT: Thrombin Time; Fbg: Fibrinogen; TBil: Total Bilirubin; DBil: Direct Bilirubin; LDH: Lactate Dehydrogenase.

Morphometric analysis of bone marrow cells

Bone marrow samples were spread on slides and stained with Wright-Giemsa (Baso Diagnostics, Zhuhai, China), followed by light microscopic analysis of 200 bone marrow cells (Olympus BX53, Olympus Corporation Tokyo, Japan) and cytochemical examination.

Bone marrow myeloperoxidase (MPO) cytochemical staining

Bone Marrow sample was smeared to be dry, subsequently, added 10 drops of compound benzidine to cover the whole blood film and stained for 2 min. Then, continued by adding 5 drops of freshly prepared dilute H₂O₂ (10 ml of distilled water + 3 drops of 30% H₂O₂), mixed well and stained for an additional 4 min, after which the sample was rinsed with distilled water several times. Finally, the sample underwent with Wright-Giemsa re-staining and rinsed with distilled water several times before being air-dried. For analysis, 100 abnormal promyelocytes were counted using the following staining intensity criteria: negative (-): cytoplasm devoid of positive granules (no blue-black granules observed), weak positive (±): fine cytoplasmic granules, sparse and typically scattered, moderate positive (+~++): cytoplasmic granules of intermediate coarseness, moderately abundant, either aggregated or dispersed, strong positive (+++~++++): coarse cytoplasmic granules, densely packed and predominantly aggregated.

Chromosome karyotype analysis

6 mL of bone marrow of the patient was extracted aseptically, anticoagulated with heparin, and the chromosomes were prepared by short-term culture method, and karyotype analysis was performed by R-banding technique. Chromosomes were described according to the International System of Nomenclature for Human Cytogenetics (ISCN 2005) [14].

Flow cytometry

Bone marrow cell immunophenotyping was performed using a five-color flow cytometer (FACSCanto II, Becton, Dickinson and Company, NJ, USA) with a monoclonal antibody (Becton, Dickinson and Company, NJ, USA) to detect each cluster of differentiation (CD) on the gated cells, and when more than 20% of the gated cells showed fluorescence over the control stained background which was defined as positive.

PML::RARα fusion gene qualitative screening assay

RNA was isolated and purified from leukocytes of bone Marrow samples using the HemaFus Human 58 Fusion Gene Screening Kit (Healthybiotech, Wuhan, China), and then reverse-transcribed to obtain cDNA. All basic PML::RARα fusion types were detected through RT-PCR using specific 5’ primers targeting different PML loci, namely: ENF903 (TCTTCCTGCCCAACAGCAA) for bcr1, ENF906 (ACCTGGATGGGACCGCCTAG) for bcr2, which were both localized in exon 6 of PML; ENF905 (CCGGATGGACCGCCTAG) for bcr3, which was localized to exon 3 of PML. The universal 3’ primer RARα−1099R (ACATGCCCACTTCAAAGCAC) were localized to exon 4 of the RARα gene. E2A served as an internal control (Forward: CTACGACGGGGGGTCTCCAC, Reverse: AGGTTCCGCTCTCTCGCACTT), and the results were interpreted only if the internal control gene was positive. The PCR products were visualized on a 1.5% agarose gel stained with ethidium bromide.

Sequencing and analysis

PCR products were isolated and purified from agarose gels using TaKaRa MiniBEST Agarose Gel DNA Extraction Kit Ver. 4.0 (Takara, Shiga, Japan), and sequencing was performed on an ABI 3730XL Genetic Analyzer (Applied Biosystems, Carlsbad, USA), and all steps were carried out according to the manufacturer’s instructions. Sequence results were presented by the software Frinch TV and compared with the sequence information in the NCBI database.

Results

Morphologic features and karyotype analysis

Bone Marrow smear indicated an extremely active state of bone Marrow hyperplasia, with significant granulocytic proliferation. The abnormal promyelocytes accounted for 77.5%, and the proportions of other stages are reduced, with morphology generally normal (Fig. 1A). Peroxidase staining (POX) showed: −0%, + 1%, ++14%, +++58%, ++++27% (Fig. 1B). The results of chromosomal karyotype analysis were 46, XX, del(7)(q22),?add(8)(p23), t(15;17)(q24;q21), −21, +r[15]/46, XX[5] (Fig. 1C).

Fig. 1.

Morphologic features and karyotype analysis of the patient. A Bone Marrow smear showed the abnormal promyelocytes accounted for 77.5% (Wright-Giemsa stain. ×100). B POX staining indicated cells with large and denseones blue granules reached to 85% (Wright-Giemsa stain. ×100). C Karyogram showed 46,XY with t(15;17)(q24;q21)

Flow cytometry immunophenotyping analysis

We immunophenotyped this bone marrow sample by flow cytometry and set up a gate analysis on the CD45 versus side scatter (SSC), and a population of abnormal cells was visible in the distribution area of the myeloid lineage towards the primordial extension, which accounted for approximately 73.5% of the nucleated cells. This cell population expressed or strongly expressed CD33, CD9, CD58, CD123, MPO, and partially expressed CD34, CD117, CD64, CD38, CD13, with a small amount of berrant CD2 expression. Negative expression of HLA-DR, CD11b, CD15, CD16 was consistent with the immunophenotype of acute promyelocytic leukemia (Fig. 2). In particular, the expression of CD9 and CD2 molecules was associated with the PML::RARα fusion gene.

Fig. 2.

Flow cytometry showed the blasts express CD13, CD33, CD34, CD9, CD38, CD58, CD64, CD117, CD123, MPO and a small amount expression of CD2

Rare variant PML::RARα fusion gene

According to the common sites of BCR, we used three 5’ primers for different sites of PML gene to avoid the missed detection of PML::RARα fusion transcripts (Fig. 3A). The results showed that, in the case of passing internal quality control (690 bp of E2A gene was detected), only 5’ primer ENF905 amplification produced products of 765 bp, 621 bp and 419 bp, suggesting a positive PML::RARα fusion, and the remaining two sets of 5’ primers failed to amplify (Fig. 3B). However, all of these amplification products were of a different length than expected, and an atypical fusion may be present.

Fig. 3.

Rare variant PML::RARα fusion gene. A Typical sites of BCR in t(l5;17) APL cases are shown by dotted line. The position and orientation of PCR primers used in this study are shown with black triangle symbol. B Electrophoretic analysis of RT-PCR results, the first three lanes was APL patient sample, and the remaining lanes were other patients screened during the same period. Lane 903 + E2A: the amplification product of the 5’ primer ENF90F and the internal control gene E2A were spotted in one well. Lane M: 1000 bp DNA Ladder. C Sequencing analysis of the obtained fragments, the dotted line indicates the breakpoint, and the blue box contains 23 inserted bases of unknown origin. D Schematic representation of the exon assembly predicted from the RT-PCR amplification products. PML exons are indicated in green and RARα exons in yellow; asterisks denote truncated exons; the shaded area represents the inserted DNA fragment; splicing events are indicated by the broken zigzag line

We further sequenced these three PCR products and showed that the 419 bp product was formed by the juxtaposition of PML exon 4 and RARα exon 3; the 765 bp product was formed by the juxtaposition of truncated PML exon 6 (deletion of 80 base pairs at the distal region) and RARα exon 3, with an insertion of 23 nucleotides at the junction (agagccttcttctctctgggacaag); the 621 bp product was obtained by deleting PML exon 5 in the 765 bp product. Based on the results of sequencing analysis, we can determine that the breakpoint in the RARα gene was located in intron 2 (Splice Site: ctctagCCA), and the breakpoint in the PML gene was located in exon 6 (Splice Site: GCCtagccc). There was a difference in the assembly of PML exons 5 and 6 in the PML::RARα fusion transcripts, probably as a result of alternative splicing (Fig. 3C, D).

Treatment and outcome

The patient initially received combined induction therapy consisting of ATRA and Compound Huangdai Tablets (CHT). Upon molecular diagnosis confirming the presence of the PML::RARα fusion gene, the dual induction therapy was continued. Subsequently, the patient developed abdominal pain and diarrhea with bloody stools, suggesting drug-induced lower gastrointestinal hemorrhage potentially associated with the CHT. Consequently, the CHT were discontinued and replaced with a combined consolidation therapy of ATRA and arsenic trioxide (ATO), to which the patient responded well. Currently, the patient is undergoing the seventh course of consolidation therapy, during which the PML::RARα gene has turned negative, and has been evaluated to be in complete remission without recurrence.

Discussion

Accurate characterization of the type of fusion gene at the molecular level is important for the management of APL patients. Most of the uncommon fusion products resulting from rare atypical PML::RARα fusion genes occur in the V-form subtypes [7]. In this study, we identified for the first time a novel case of the type V PML::RARα isoform, which showed typical APL morphological, immunological and cytogenetic features, however, qualitative screening of the PML::RARα fusion gene identified three rare fusion transcripts. These fusion transcripts originated from the fusion of truncated PML exon 6 with RARα exon 3, accompanied by an additional insertion of a 23-nucleotide fragment at the truncated PML exon 6.

A defining molecular characteristic of PML::RARα fusion events is the detection of structural heterogeneity in PML::RARα transcripts. The in-frame/out-of-frame configuration of fusion junctions critically determines the protein-coding potential of the fusion transcript and has significant clinical implications [15]. Through sequence analysis we observed that the 765 bp and 621 bp transcripts preserved co-linear alignment of the canonical ORFs from PML and RARα. The maintenance of this co-linear ORF architecture enables translation of chimeric proteins retaining critical functional domains implicated in leukemogenesis [16, 17], mechanistically linking these isoforms to APL pathogenesis. In contrast, the 419 bp truncated transcript exhibited frameshift iscordance between PML and RARα ORFs, introducing a premature UAG termination codon at codon 12 of the RARα reading frame. This frameshift mutation disrupts the canonical open reading frame, strongly suggesting that this transcript is unlikely to produce a stable, full-length chimeric protein with the oncogenic properties characteristic of in-frame PML::RARα fusions [18, 19]. However, it is crucial to note that frameshifted fusion transcripts, while typically non-coding at the protein level, have been reported in other contexts to potentially function as regulatory non-coding RNAs (e.g., as competing endogenous RNAs [ceRNAs]) [20, 21]. Therefore, while the protein-coding function appears compromised, the potential biological role, of this specific 419 bp transcript as an RNA molecule requires further investigation. Variations in exon-exon junctions involving PML exons 5 and 6 across these transcripts suggest alternative splicing mechanisms, with the observed splicing patterns indicating potential genomic breakpoint clustering within PML exon 6. This hypothesis is corroborated by unsuccessful amplification of PML::RARα transcripts using 5’ primers (ENF906/ENF905) designed against exon 6 sequences. The expression levels of each PML::RARα transcript varied significantly among APL patients. To elucidate the prognostic value of these different transcripts in all APL patients, studies with rigorous characterization and sequencing of all atypical subtypes, including functional assessment beyond protein coding potential, are required [22].

Most atypical V-form PML::RARα transcripts contain variable segments derived from RARα intron 2, potentially resulting from chromosomal rearrangements during t(15;17) translocation or dysregulated splicing leading to intronic retention [15]. Consistent with this, our study identified a 16-nucleotide insertion (tcttctctgggacaag) within truncated PML exon 6, mapping to RARα intron 2 (chr17: 40334421–40334436, NC_000017.11). The distal region of PML exon 6 encodes a serine/proline-rich (SP-rich) domain critical for ATRA-induced degradation of PML::RARα, as it undergoes extensive phosphorylation and harbors a key caspase cleavage site [5]. Concurrently, the structural integrity of the PML B-box2 zinc finger domain is essential for mediating clinical resistance to ATO, truncation of exon 6 may impair the functionality of the B-box2 zinc finger domain by disrupting the assembly of PML nuclear bodies (PML-NBs) [23]. Therefore, the extent of PML exon 6 truncation, which can disrupt these functional domains, may strongly correlate with clinical response to ATRA/ATO combination therapy. However, the precise impact of exon 6 truncation remains complex. Bai-Wei Gu et al. demonstrated that an aberrant spacer sequence (not the truncation itself) was responsible for resistance to ATRA, deletion of this spacer restored ATRA sensitivity [24]. Conversely, James L. Slack et al. found no significant correlation between the degree of PML exon 6 truncation/insertion type and clinical outcome in their analysis of 18 V-form cases [25]. In our specific case, the patient achieved molecular remission with combined ATRA + ATO therapy. Given the conflicting reports and the functional importance of the domains affected by truncation, further research with larger cohorts is necessary to elucidate the impact of specific V-form PML exon 6 alterations on clinical diagnosis and treatment outcomes in atypical APL subtypes.

In summary, this case highlight the complexity of atypical PML::RARα fusion gene testing and the limitations of traditional RT-PCR methods. It suggests that complementary multi-omics techniques should be considered to avoid missed diagnosis. Therefore, accurate and comprehensive laboratory tests are crucial for the diagnosis of APL, treatment efficacy, assessment of prognosis and prediction of recurrence. However, due to the limitations of this study, such as the sample size constrained by the rarity of variants, insufficient functional validation and the inadequate long-term follow-up, future research should concentrate on (i) elucidating the pathogenic mechanisms of different transcripts, (ii) correlating structural variant-drug sensitivity patterns, and (iii) optimizing individualized therapeutic strategies through multicenter collaboration. Concurrently, we advocate for the development of a treatment response prediction model based on subtype fusion, to enhance the precision of APL diagnosis and treatment.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

L.Y. and Y.L. designed the experiments. H.C., L.Y. and M.C. executed the experiments. H.C., L.Y., and M.C. performed the data analysis. L.S. and Z.Y. collected the clinical information. H.C. and L.Y. wrote the manuscript with input from all of the other authors. All the authors contributed to the article and approved the submitted version.

Funding

This work was supported by Startup Fund for Scientific Research of Fujian Medical University (NO. 2022QH1281 to L.Y. and No. 2021QH1274 to M.C.) and the Fujian Provincial Health Technology Project of Fujian Provincial Health Commission (No. 2023QNA005 to M.C.).

Data availability

The raw data that support the findings of this study are not openly available due to privacy and ethical restrictions. However, the de-identified data can be accessed from the corresponding author upon reasonable request.

Declarations

Ethics approval

The study was approved by the Medical Ethics Committee of The Fuzhou University Affiliated Provincial Hospital. All analyses and therapies were performed after informing consent from the patient and her legal guardian.

Consent for publication

Written informed consent for publication was obtained from the patient or her legal guardian.

Competing interests

The authors declare no competing interests.