The Role of BAALC Gene in the Transformation of Myeloid Progenitor Cells to Acute Myeloid Leukemia

Abstract

One of the unanswered questions in hematology is the question concerning disorders in the regulation of gene expression in different subtypes of acute myeloid leukemia (AML), leading to changes in the functional activity of certain genes and acting as a component of a series of events in the leukemogenesis. One example of such a gene is BAALC gene (brain and acute leukemia and cytoplasmic), localized in chromosome 8, which plays a role in the regulation of myeloid progenitors’ differentiation. This role is associated with several other oncogenes, such as HoxA9, ERK, and RUNX1. Gene interactions determine normal proliferation and differentiation of cells, and any disturbances could lead to leukemic development. What is the role of BAALC in normal/impaired balance? What are the connections of BAALC with the mutations established in AML: FLT3, NPM1, etc.? What are the correlations of its overexpression with clinical and laboratory findings in AML patients? What are the changes in the expression of BAALC, after successful therapy of AML and after therapy failure? Can we use it as a predictive marker in AML patients? This chapter summarizes available data about functions of BAALC gene, the frequency of overexpression, and its importance as a predictive marker in the development of AML.

1. Introduction

Detailed sequencing of the genome among 200 AML patients, done in the program “the cancer genome atlas (TCGA),” revealed AML as a heterogenic disease with a total of 2585 gene mutations. The authors report 2315 somatic single-nucleotide variants and 270 small insertions and deletions in the genome of AML patients, with an average of 13 mutations per sample [1]. TCGA identifies “at least one potential driver mutation per sample” and confirms the complexity of gene interactions in AML pathogenesis. Some of these mutations could be observed also in myelodysplastic syndromes (MDS) or in myeloproliferative disorders (MPN), even in healthy individuals, with age-dependent clonal hematopoiesis, increasing the risk for the development of AML [2]. This heterogeneity shows why AML remains one of the serious problems in front of hematology in the twenty-first century, despite the huge advances in the last years. The unclear etiopathogenesis, and the blurred boundaries between the different groups of myeloid diseases, connected with AML, still keep open many questions that continue to challenge researchers. Against the background of large-scale studies on a variety of gene mutations in recent years, some issues are still less clear, such as the abnormalities in the regulation of gene expression in various subtypes of AML, leading to functional impairment of some genes and respectively overexpression or lack of expression, resulting in overproduction/absence of normal proteins of key importance for cell biology. One of these questions was raised with the discovery of BAALC (brain and acute leukemia, cytoplasmic) gene in 2001 by Prof. S. Tanner’s team at Ohio State University. Increased BAALC expression was initially identified in a study of patients with AML and trisomy 8 and was later found in other types of AML as well as in acute lymphoblastic leukemia (ALL) [3, 4].

2. Genetics

BAALC gene is located on the long arm of chromosome 8 at position 22.3 (8q22.3) [5]. It contains 8 exons and extends to 89,613 bases of genomic DNA. The BAALC gene has eight different transcriptional forms, resulting in six different protein isoforms [6, 7]. The transcriptional proteins encoded by the corresponding exons consist of 54 to 180 amino acids [5, 8]. The BAALC gene is expressed in mammals but is not found in lower-level organisms, such as insects and fungi [2]. The high levels of expression in mammals, as well as the lack of an orthologous gene in lower vertebrates, suggest a specific role of the BAALC gene in the mammalian CNS [2]. The BAALC gene and the protein encoded by it have high expression in neural tissues, such as the brain and spinal cord, and significantly lower expression in other neuroectodermal tissues, such as the adrenal glands [2, 8]. In neuroectodermal tissue, the highest expression of BAALC is found in the frontal cortex of the brain and particularly in the hippocampus and neocortex [3]. Its expression has been found also in bone marrow in CD34-positive white blood cell progenitors [2]. Interestingly, the transcriptional isoforms of BAALC in neurons are localized to postsynaptic lipid rafts. The protein encoded by BAALC gene is found also in muscle cells (myocardial and skeletal) and localizes attached to the inner side of the cell membrane but polarized only to one end of the cell [8]. The BAALC-encoded protein is not expressed in peripheral blood leukocytes, lymph nodes, or non-neural tissue [2, 5]. The expression of BAALC by CD34-positive progenitors and neuroectodermal cells suggests that BAALC performs certain functions in these tissues [2, 6, 9].

A practical aspect of investigations on BAALC overexpression is the question of what to use for the analysis—bone marrow (BM) or peripheral blood (PB)? This question was addressed in different studies and a strong correlation of BAALC expression levels in both specimens has been reported and confirmed [10, 11, 12].

3. Functions

The BAALC gene was originally found in the neuroectodermal tissue [2]. The mechanism and the function of the protein encoded by BAALC, are still not clearly defined, but the current understanding is that the changes in BAALC gene expression are connected with a stop in cell differentiation, caused by changes in shape, motility, or adhesion of myeloid precursors [9]. The role of the BAALC gene in the development of leukemia has been studied in immature leukemia cells by eliminating gene function by RNA in a human leukemia cell line [13]. The elimination of overexpression of the BAALC gene results in a decrease in uncontrolled cell growth and an increase in programmed cell death [9, 14]. Evidence suggests that the BAALC protein is an intracellular protein that plays roles in the cytoskeletal network, including regulation of the actin cytoskeleton, associated with a role in the postsynaptic “lipid raft” [2, 3]. The BAALC protein isoforms were found to interact with Ca2+/calmodulin-dependent protein kinase II (CaMKII). CaMKII is a protein kinase whose activation is dependent on Ca2+/calmodulin complex. It is involved in many signaling cascades and is required for Ca2+ homeostasis. CAMKII is thought to be an important mediator in brain tissue in memory and learning [14], as well as in the differentiation and activation of the CD8 T-cell population [15]. BAALC interacts with the CAMKII alpha subunit but not with the CAMKII beta subunit [3]. The interaction with the CAMKII alpha subunit takes place in the regulatory region of the CAMKII alpha protein. The BAALC 1–6-8 isoform targets the postsynaptic “lipid raft”, suggesting that it has functions involved in signal transduction, transmembrane traffic, and actin cytoskeleton regulation [3]. BAALC may play a role in the regulation of the CAMKII protein by interacting with the alpha subunit [3].

According to Baldus et al. [16], BAALC expression is restricted to progenitor cells, and downregulation of BAALC occurs with cell differentiation. Authors conclude that BAALC is represented in an early progenitor cell common to the myeloid, lymphoid, and erythroid pathways. Heuser et al. also suggested that the function of BAALC in the hematopoietic system is to block the differentiation of the myeloid lineage and requires a secondary mutation that favors the proliferation of this clone to induce leukemia [6]. The study by Heuser et al. found that despite the constant activation of the BAALC gene, hematopoietic stem cell proliferation did not start, it blocks myeloid differentiation and thus starts leukemogenesis, when combined with a self-renewing oncogene promoter (e.g., HoxA9) [6]. The data of Heuser et al. provide a molecular basis for the role of BAALC in regulating the proliferation and differentiation of AML cells and demonstrate the possibility of targeting BAALC in AML patients with high BAALC expression. Morita et al. found, that BAALC-induced leukemic cell cycle progression by sustained activation of extracellular signal-regulated kinase (ERK) interacting with the cellular skeletal protein MEK kinase-1 (MEKK1) inhibiting the interaction between ERK and MAP kinase phosphatase 3 (MKP)/DUSP6). BAALC induces chemoresistance in AML cells through enhanced regulation (upregulation) of ATP-binding cassette proteins by ERK—a dependent way that can be therapeutically affected by a target MEK inhibitor. The authors also demonstrated that BAALC blocks ERK-mediated monocyte differentiation of AML cells by blocking Krüppel-like factor 4 (KLF4) in the cytoplasm and inhibiting its function in the nucleus. Therefore, MEK inhibition will synergize with the induction of KLF4, and high efficacy is expected against AML cells with high BAALC expression [9]. By sequencing the BAALC genomic region, Eisfeld et al. identified six informative single nucleotide polymorphisms and they tested them for possible association with BAALC overexpression. They showed that BAALC overexpression occurs in the presence of a T-allele that creates a binding site for RUNX1 activating transcription factor in the BAALC promoter region. The association of high BAALC expression with the T-allele and its correlation with RUNX1 expression status have been demonstrated in patients with cytogenetically normal (CN) AML from different populations. BAALC is very likely to act as one of the components of a complex high-risk series of changes in leukemogenesis [17]. There is evidence of a regulatory relationship between BAALC expression by FLI1 and c-MYC genes encoding transcription factors leading to cell proliferation [18].

The other fact concerning BAALC functions is described by Maki et al. [19], who identified that BAALC is connected with drebrin, in a protein complex. Drebrin—an actin-binding cytoplasmic protein located in the cell membrane is encoded from DBN1 gene, located on chromosome 5. DBN1 is identified in neural cells [20] and is also expressed in HSCs, which suggests that DBN1 plays a role in cell skeleton dynamics in hematopoietic cells, as well as in neuronal cells [21]. By remodeling the actin filament, drebrin transforms the cell membrane spine during cell–cell interactions, improving cell adhesion and mobility [22]. BAALC physically interacts with DBN1 and participates in cellular adhesive functions, which may induce drug resistance. The localization of BAALC-DBN1 complex in the cytoplasm is to the inner cellular membrane, and inhibition of DBN1 gene activity results in decreased adhesive capacity and better sensitivity to chemotherapy agents of leukemic cells. Taken together, these findings prompted us to hypothesize that BAALC plays a role as a molecular transporter between cytoplasm and cellular membrane, recruiting DBN1 to the cell membrane to activate cellular adhesion to the bone marrow niche. Considering that BAALC is exclusively expressed in the stem cell compartment in the hematopoietic system, BAALC-DBN1 interaction may be important, especially in the immature fraction [19].

4. Frequency of BAALC overexpression in myeloid neoplasms

The reported frequency of BAALC overexpression in AML patients varies according to different reports and authors, and one of the reasons for this fact is that there are still no defined cutoff ranges on high/low BAALC expression.

In a study, Taner et al. tested samples from BM or PB taken from 130 diverse AML. They found that 28% of all AML patients are BAALC-positive [4].

Damiani et al. evaluated the effect of BAALC overexpression on the outcome of 175 adult AML patients with different cytogenetic risks. BAALC was overexpressed in 87/175 (50%) patients, without association with cytogenetic status [23].

Amirpour et al. assessed 47 cases with normal cytogenetic AML and found BAALC gene up-regulation in 47% of the cases and down-regulation in 53% of patients [24].

A similar prevalence of 50% was reported in the investigation of Weber et al. in CN-AML [10].

Significant differences in BAALC expression were observed by Zhou et al. among AML subtypes. They investigated samples from 121 de novo patients and found that 67(55.3%) of them have high BAALC expression above the cutoff value of 2.35. Among AML FAB subtypes (M0/M1/M2/M3), they found a significantly higher incidence of BAALC overexpression in M0/M1 (8/9, 89%) and M2 subtypes (33/48, 68%) than in M3 subtype (6/27, 22%) [25].

The prevalence of BAALC overexpression was substantially higher in patients with intermediate (42/70, 60%) and poor karyotypes (9/11, 82%) than it was in patients with a favorable karyotype (13/34, 38%). In contrast to the other patients, those with t(15;17) had the lowest frequency of BAALC overexpression, while those with t(8;21) had the highest.

In MDS patients, the observed incidence of BAALC overexpression was 50%, with the highest rates in the subtypes: refractory anemia with blast excess (RAEB-2) (30% of all high-expression patients) and refractory anemia (RA) (20% of all patients with high expression) [18].

There are no data on the frequency of overexpression of BAALC in MPN—chronic myeloid leukemia, myelofibrosis, polycythemia vera, and essential thrombocythemia.

Concerning overexpression of BAALC outside myeloid diseases—in the same study cited before [4], Tanner et al. found 65% BAALC-positive ALL patients. In an investigation among pediatric ALL the frequency of BAALC-positive ALL patients are reported 60% [26].

Overexpression of BAALC has also been found in solid tumors: glioblastoma, melanoma, and gastrointestinal stromal tumors (in children), suggesting its role in oncogenesis [17].

5. Clinical significance and importance of BАALC overexpression

5.1 Correlation of BАALC overexpression with some clinical, laboratory characteristics, and AML mutational status

In most of the papers in the literature, there are no associations of high BAALC gene expression with patients’ and clinical characteristics, such as sex, age, white blood cell (WBC) count, PB blasts, BM blasts, or hemoglobin levels, at the time of AML diagnosis [14, 23]. In some of them, the authors found a trend that high BAALC expressers are younger age than the low expressers [10]. In an analysis from Shafik et al. a statistical significance between BAALC and age, CD34 expression, and close to significant results with some of FAB sub-groups was found [27].

Zhou et al. by applying multivariate analysis, which took into account factors, such as gender, age, WBC, HB, PLT, karyotype classifications, 10 gene mutations (mutant/wild type), and BAALC, identified high BAALC expression as an independent adverse prognostic factor in AML patients. Multivariate analysis using the same factors, except karyotype classification, among CN-AML patients also supported the conclusion that BAALC overexpression is a poor prognostic factor [25].

The results of a meta-analysis conducted in selected studies show a tendency for increased BAALC gene expression in poorly differentiated AML (M0/M1/M2) subtypes, while in well-differentiated subtypes the overexpression is rare [14]. There is no evidence of an association between overexpression of the BAALC gene and the type of chromosomal abnormalities found. However, an increased incidence of aberrant overexpression has been observed in patients with a normal karyotype.

Association between BAALC gene mutations connected with AML has been reported.

FLT3 (ITD/TKD) are some of the most explored mutations in AML. There are sufficient data about the prognostic significance (included in “WHO classification of tumors of hematopoietic and lymphoid tissues”), and in addition, FLT3 mutations already act as a therapeutic target. Therefore, BAALC overexpression and the FLT3 mutation is an objects of many studies. Some of them [14, 23, 27, 28], failed to find a link between BAALC and FLT3-ITD, but opposing this conclusion, other authors provide results concluding that there is a strong correlation between high BAALC expression and the FLT3-ITD mutation [10, 29, 30].

When comparing 307 adult patients with low and high BAALC expression, Baldus et al. found that high-expression patients had a higher incidence of primary resistant leukemia. High BAALC expression is associated also with a higher cumulative relapse rate and shorter overall survival (OS). According to their conclusion, higher BAALC expression is an independent predictive factor for disease resistance. They only confirmed that high BAALC expression and the presence of the high FLT3 mutation/wild-type ratio is a predictive factor for a high incidence of primary resistant leukemia, as well as for reduced OS [29].

FLT3-ITD and NPM1 mutations were analyzed in several studies and the data were summarized in a meta-analysis [14]. The results obtained do not show a correlation between the high expression of the BAALC gene and the FLT3-ITD status, but the authors reported an association with the NPM1 mutation. Patients with high levels of BAALC expression at diagnosis were correlated with the mutational status of NPM1. This comparison revealed a good correlation between the % BAALC/ABL1 levels with the NPM1 mutation [10].

As mentioned above, there is evidence that high BAALC gene expression blocks myeloid cell differentiation and when combined with a HOXA9 gene mutation that is associated with cell self-renewal, may induce leukemia [6]. Regarding IDH1 and IDH2 gene mutations, BAALC overexpression has been reported to be associated with wild-type IDH1 and IDH2, in patients with cytogenetically normal AML [31], but other authors [13], found that there is no direct association between IDH1 and IDH2 gene mutations and overexpression of BAALC.

Different associations of altered BAALC expression with specific molecular aberrations have been shown. For instance, high BAALC expression has been demonstrated to correlate with the mutation status of FLT3-ITD, CEBPA, and MLL-PTD as well as with NPM1wt. [28, 29, 30]. In a cohort of 326 CN-AML patients, the correlation of high BAALC expression with FLT3-ITD, CEBPAmut, MLL-PTD, and NPM1wt could be confirmed. Authors provided results that show a strong correlation of high BAALC expression with WT1mut, biCEBPA, IDH2R172mut, and especially RUNX1mut (31 of 33 RUNX1-positive AML patients have also high BAALC expression), but they did not find correlations to ASXL1mut, IDH1R132mut, IDH2R140mut, or TET2mut [10].

Finally, Langer et al. analyzed BAALC expression in 172 primary CN-AML patients (<60 years). They found high BAALC expression associated with FLT3-ITD, wild-type NPM1, mutated CEBPA, MLL-PTD, absent FLT3-TKD, and high ERG expression. The multidrug resistance gene ABCB1 (MDR1) was shown to be one of the most substantially up-regulated genes, which is consistent with the resistant disease [32].

5.2 Predictive value of BАALC in AML patients

In contemporary hematology, a crucial part of the therapeutic decision-making process is the proper risk assessment, based on pretreatment genetic markers. The potential effect of BAALC in AML could be assessed in two directions: as a prognostic marker at diagnosis and as a minimal residual disease marker after treatment.

There are enough data, confirming that BAALC overexpression is associated with adverse prognosis in AML patients. In the last years, several studies have explored the relevance of using BAALC overexpression as a part of different score systems aimed at assessing the clinical efficacy of chemotherapy.

We know that approximately 55–60% of cases of AML have specific recurrent chromosomal aberrations, which can be identified by classical cytogenetic techniques. This information is the basis for classifying patients, at diagnosis, into prognostic categories: favorable, intermediate, and unfavorable risk. According to present recommendations and guidelines, patients with favorable risk AML are treated with chemotherapy only, while patients with adverse risk are referred to allogeneic stem cell transplantation if a suitable donor is available. The optimal therapeutic strategy in patients with intermediate-risk is still debatable. Approximately 40–45% of AML patients, who have no identified cytogenetic abnormality at diagnosis are classified in this group [33]. By finding overexpression of BAALC in this group and assessing the outcome, we are able to identify subgroups of patients at adverse risk among those with CN-AML. Xiao et al. reported that patients with AML who overexpress BAALC (BAALC-positive) have a median EFS (event-free survival) of 5 months, compared to 15 months in BAALC-negative patients. The prognosis of patients with AML and BAALC overexpression is unfavorable—the significant overexpression of BAALC and accumulation of gene products in the cell leads to drug resistance in these patients, which in turn results in a low overall response to therapy and hence to the reduced overall survival of these patients [14].

The importance of identifying prognostic molecular markers in CNAML patients to improve the therapeutic approach was confirmed by a study in which the authors investigated the prognostic effect of altered BAALC expression in CN -AML patients [34]. They analyzed bone marrow samples from 30 CN-AML patients for BAALC expression levels using RT-PCR. Patients were divided into BAALC low and high expression. High expression was present in 70% of patients and the level of expression was not correlated with patients’ clinical parameters. In a 2 year follow-up, patients with high BAALC expression had a lower incidence of clinical remissions and shorter overall survival (OS). Multivariate analysis confirmed the high expression of BAALC as an independent risk factor for OS. BAALC overexpression predicts an adverse clinical outcome and may be identified as an important risk factor in CN-OML patients [34].

According to Yahya et al. a similar conclusion about the prognostic relevance of BAALC gene expression in adult acute myeloid leukemia was formulated. High expression was found in 22 of 45 patients (48.9%), and there was no link between this expression and the most important treatment-outcome characteristics. High BAALC expressers have a higher mortality rate, a significantly shorter DFS, and a shorter overall survival. They also have a lower incidence of CR and a lower mortality rate. The conclusion is that high BAALC expression may be a significant prognostic indicator in individuals with a normal karyotype and an independent risk factor for both DFS and OS, according to the authors’ findings [35].

In a survey by Weber et al. patients with CN-AML and high expression of BAALC had significantly shorter EFS and OS than low expressers. High BAALC expression was also linked to a significantly lower survival when OS was analyzed and patients were censored on the day of transplantation, excluding the impact of allogeneic SCT. The authors were able to validate the usefulness of BAALC expression as a marker for identifying residual disease patients through their investigation. Parallel investigation of BAALC expression showed a substantial association between BAALC expression levels at diagnosis and after treatment. Additionally, the authors found that in the cohort of patients with relapsed AML, the mean BAALC expression levels at diagnosis and relapse were comparable, confirming BAALC expression as a potential MRD marker [10].

What is the role of BAALC in patients with AML and abnormal karyotype? In a study of 175 adult patients with AML with different cytogenetic risks, Damiani et al. assessed the effect of BAALC gene overexpression on disease outcome [23]. The patients in the study were distributed as follows: 13 with a favorable karyotype, 117 with an intermediate, and 45 with an unfavorable. The BAALC gene was overexpressed in 87/175 (50%) of patients without any association with the cytogenetic status. Complete remission (CR) was achieved in 111 patients (63%) and was maintained till 5th year in 52 ± 7%. BAALC overexpression had a negative effect on the achievement of CR but did not affect the likelihood of disease recurrence. The median survival was 22 months with overall survival (OS) of 35%. BAALC was identified as one of the factors with a negative impact on the OS. The authors observed a worse outcome in patients with high BAALC expression in all cytogenetic risk categories: 5-year OS was 100% versus 71% in patients with a favorable prognosis, 55% versus 40% in patients with an intermediate karyotype, and 34% versus 23% in the group with an unfavorable prognosis. In conclusion, overexpression of the BAALC gene identifies patients with an unfavorable prognosis in all cytogenetic groups, and the presence in patients with favorable or unfavorable karyotypes significantly worsens survival [23].

The potential effect of BAALC in cytogenetic abnormal AML is explored also by Santamaria et al. [36]. They evaluated BAALC expression in AML patients with abnormal karyotypes and found a connection between high BAALC overexpression and lower rates of CR and OS.

In meta-analysis involving over 7525 cases of AML from 25 studies, Yuen et al. concluded that at diagnosis, complete remission attainment and survival outcome were negatively impacted by BAALC overexpression. Therefore, stratifying AML patients based on BAALC gene expression levels at diagnosis could be beneficial for treatment decisions [37].

A survival analysis of 108 patients with follow-up data was conducted to examine the prognostic significance of BAALC expression in AML. Both in the entire AML cohort and in CN-AML category, patients with high levels of BAALC had significantly lower rates of complete remission (CR) than patients with low levels of BAALC. In the entire cohort of AML patients, patients with high BAALC expression had significantly shorter overall survival (OS) than those with low BAALC expression. Additionally, in the CN-AML subgroup, patients with high BAALC expression had also a significantly shorter OS compared to those with low BAALC expression [25].

Weber et al. showed data that BAALC expression is an independent predictor of shorter EFS, OS, and that it is a useful target for detecting residual disease. Authors showed that a reduction in BAALC expression after chemotherapy from the initially detected levels resulted in better EFS [10].

In conclusion, based on the described data, BAALC could be regarded as a relevant prognostic marker in AML patients and could be used for treatment response evaluation after treatment in patients with high BAALC expression at diagnosis. Future prospective analysis should confirm the predictive value of BAALC and if AML patients with high BAALC expression at diagnosis may have therapy benefits from BAALC detection over the course of their disease.

5.3 Predictive value of BAALC in MDS patients

Data on the role of BAALC in patients with MDS are scarce, and most of them come from studies investigating multiple genes (including BAALC) associated with a poor prognosis in myeloid neoplasms. In a study of 140 MDS patients, Thol et al. combined the expression data of four genes: MN1, ERG, BAALC, and EVI1 (MEBE) in an additive score that was validated with an independent cohort of 110 MDS patients. The high MEBE score, defined as high expression of at least two of the four genes, predicted significantly shorter overall survival and time to progression to AML. In the validation cohort of 110 MDS patients, the high MEBE score predicted shorter OS and time to progression to AML. Thus, it can be concluded that a high MEBE score is an unfavorable prognostic marker in MDS and is associated with an increased risk of progression to AML. Almost half of the MDS patients with high BAALC expression (48.6%) transformed into AML within the study, compared with only 11.4% of the BAALC low expression group [18]. The cited data outline the possible involvement of the BAALC gene in the pathogenesis of MDS and clearly indicate the need for further research in this direction.

5.4 Predictive value of BAALC in HSCT in AML patients

Concerning the role of BAALC in transplantation, we have two main questions: is it possible to use the level of BAALC expression as a suitable marker for MRD/response to therapy, and what is the prognostic value of BAALC expression on the treatment outcome in transplanted patients.

In an attempt to assess the impact of allogeneic transplantation on treatment outcome, Damm et al. proposed an integrative prognostic risk score (IPRS) for CN-AML patients based on clinical, hematological, and molecular markers, including the expression levels of BAALC. Depending on the assessment, authors divided patients with AML into three risk groups: low, intermediate, and high risk. The role of allogeneic stem cell transplantation (alloSCT) for patients in each of the three risk groups was investigated by intent-to-treat analysis, due to the fact that all patients with an available donor were referred to alloSCT. The results showed that the high-risk group of patients have longer overall survival (OS) and relapse free survival (RFS) when alloSCT is performed by a family-related donor than in patients with a nonfamily donor. The intermediate risk group had lower OS and RFS when the alloSCT was performed from a family-related donor compared to patients who were from a nonfamily donor. The effect in the low-risk group is not clear, most likely due to the low frequency of relapses. The data showed that the IPRS may be an additional tool for disease outcome prediction and treatment choice, reflecting the biological heterogeneity in CN-AML patients [38].

Assessing the prognostic value of BAALC expression, Weber et al. found that excluding the effect of allogeneic HSCT, patients with high BAALC expression have considerably shorter OS [10].

Jentzsch et al. tried to assess how high BAALC copy numbers, in peripheral blood, prior to allogeneic transplantation predict early relapse in acute myeloid leukemia patients. They performed absolute quantification of BAALC copy numbers in peripheral blood prior (median 7 days) to HSCT in complete remission or CR with incomplete peripheral recovery in 82 acute myeloid leukemia patients. Patients with high BAALC/ABL1 copy levels before HSCT were more likely to relapse within 100 days of HSCT [39].

In аnother study, Zhang et al. investigated 71 de novo AML patients who were treated with allo-HSCT and classified them as low or high expressers based on the median BAALC expression levels at diagnosis. They did not find significant differences in event-free survival or overall survival between the two groups [40].

5.5 The role of BAALC gene in developing chemoresistance

Interactions between HSCs and the bone marrow niche components, such as osteoblasts, sinusoidal endothelial cells, and mesenchymal stromal cells, are very important because they create a special bone marrow microenvironment, in which processes of self-renewal and differentiation of HSCs are strictly regulated [41]. Cell adhesion to the niche protects HSCs, leukemic cells from toxic exposure, and they can develop resistance to cytotoxic agents [42, 43, 44]. Based on the fact that BAALC plays an important role in the activation of cellular adhesion to the bone marrow niche some authors rise a hypothesis that high BAALC expression in leukemic cells induces drug resistance by adhesion to the niche components [19]. Previously mentioned data suggests that BAALC binds to drebrin, and this interaction develops chemoresistance against cytotoxic drugs, such as cytarabine [19].

Several other papers are in alignment with the hypothesis of connections between BAALC overexpression and a chemo-resistant disease.

Santamaria et al. analyzed the BAALC gene in 127 AML patients categorized as intermediate risk (98 cytogenetically normal and 29 with cytogenetic alterations). They reported that high expressers are more resistant to induction chemotherapy in comparison with low BAALC expressers. High expressers also show a lower complete remission rate after reinduction therapy, lower 3-year overall survival, and relapse-free survival. The authors also found similarity in the results when subgroups were analyzed separately according to their cytogenetic status [36].

In a study of gene expression profiles (GEPs) in 312 probe sets, Langer et al. found an association between high BAALC expressers with overexpression of one of the genes involved in multidrug resistance (MDR). High BAALC expression levels had distinctive features, according to the study. In individuals with high BAALC expression levels, the multidrug resistance gene ABCB1 (MDR1) was shown to be one of the most substantially up-regulated genes, which is consistent with the resistant disease that these patients are known to have. The ATP-dependent drug efflux pump MDR1 protein, also known as P-glycoprotein, is in charge of reducing drug accumulation in cells that are resistant to several treatments. It also frequently plays a role in the emergence of anticancer drug resistance [32].

Kuhnl et al. provided another confirmation that BAALC gene overexpression induces chemoresistance. They investigated the prognostic significance of BAALC in a different from AML area, that is, in B-precursor ALL, and concluded that high BAALC expression is associated with an immature chemo-resistant leukemic phenotype and inferior OS [45].

In the paper cited before [10], Weber et al. found that high BAALC expression is an independent predictor of shorter EFS and OS and chemotherapy-resistant disease.

In a study of BAALC-associated GEPs, Heesch et al. concluded that over-expression of BAALC could be associated with adverse outcomes and chemotherapy resistance in adult patients with cytogenetically normal AML [31].

6. Conclusions

Increased BAALC expression is identified and connected with the development of AML. Normally, the highest expression of BAALC is found in neural tissues, such as frontal cortex of the brain, in the hippocampus, neocortex, and bone marrow in CD34-positive white blood cell progenitors. BAALC gene is expressed in mammals but is not found in lower-level organisms.

The mechanism and the function of the protein encoded by BAALC are still not clearly defined, but the current understanding is that the changes in BAALC gene expression are connected with a stop in cell differentiation due to changes in shape, motility, or adhesion of myeloid precursors. Evidence suggests that the BAALC protein is an intracellular protein that participates in the cytoskeletal network, including the regulation of the actin cytoskeleton. There is a hypothesis that BAALC also plays a role as a molecular transporter between cytoplasm and cellular membrane, acting in the mechanism which activates cellular adhesion to the bone marrow niche.

The reported frequency of BAALC overexpression in AML patients varies (28%–55.3%) among different reports and authors because there is still no defined cutoff ranges on high/low BAALC expression. Some authors found variety in BAALC overexpression in different risk categories or in different AML subtypes according to FAB classification. To our knowledge, only one paper has investigated the frequency of BAALC overexpression in MDS–50%, so far. Currently, we have insufficient data about the involvement of the BAALC gene in the pathogenesis of MDS and this fact clearly indicates the need for further research in this direction.

In most of the scientific papers, there are no associations of high BAALC gene expression with patient and clinical characteristics, such as sex, age, white blood cell (WBC) count, PB blasts, BM blasts, or hemoglobin levels, at the time of diagnosis of AML. Concerning molecular anomalies—there is a discrepancy in the data connecting FLT3 (ITD/TKD) mutations with BAALC. Some of the authors failed to find a link between high BAALC expression and the FLT3-ITD mutation, but opposing this conclusion, others provide results concluding that there is a strong correlation between high BAALC expression and the FLT3-ITD mutation. Obviously, we need additional prospective studies to clarify this question.

Correlations between high BAALC expression with CEBPAmut/biCEBPA, MLL-PTD, NPM1wt, WT1mut, IDH2R172mut, high ERG expression, and a strong correlation with RUNX1mut and multidrug resistance gene ABCB1 (MDR1) have been demonstrated.

Based on the fact that BAALC plays an important role in the activation of cellular adhesion to the bone marrow niche, some authors rise a hypothesis that high BAALC expression in leukemic cells induces drug resistance.

Concerning the predictive value, BAALC could be a relevant prognostic marker in AML patients and could also be used for MRD assessment after treatment in patients with high BAALC expression at diagnosis. BAALC expression is an important risk factor in CN-AML and is associated with an inferior outcome and chemotherapy-resistant disease. In patients with AML with abnormal karyotype, BAALC overexpression identifies patients with an unfavorable prognosis across all cytogenetic groups, as survival is significantly worsened both in patients with favorable or unfavorable karyotype. The results about the role of BAALC in transplantation are again disputable. The available data cannot confirm it as a suitable marker for MRD or for the response to therapy.

Future prospective analyses to determine the potential role of BAALC expression assessment in the clinical assessment and decision-making process in AML and MDS patients are worthwhile.

Conflict of interest

The authors declare no conflict of interest.