Introduction

The human endogenous retroviruses (HERVs) are composed of many families of endogenous retroviruses that were integrated into the germ DNA of the primate line over the last ∼30 million years [1]. Currently, HERVs are classified into 50–200 families depending on the primer binding site (PBS) sequences complementary to the 3′ end of a cellular tRNA used for reverse transcription [2, 3]. However, transcription of HERV elements is usually suppressed by epigenetic factors such as DNA methylation and heterochromatin silencing by histone adjustments; inserted LTRs act as alternative promoters to stimulate expression of adjacent genes causing the activation of oncogenes or inactivation of tumor suppressor genes [4, 5]. The re-activation of HERV transcription as a consequent to various factors (such as radiations, chemicals, other exogenous viral infections) could act as a promoter for the development of several malignancies through the modification in cell gene expression [6,7,8]. There are only few studies on the expression of endogenous retroviruses in pediatric leukemias [9,10,11,12]; moreover, data regarding the participation of HERV-K and W in the development of leukemia are contradictory [12,13,14]. One study compared the mRNA expression of HERVs in tumors to adjacent normal tissues and found high levels of HERV-K expression in testis tumor tissues, HERV-R (ERV3–1) in liver and lung tumor tissues; HERV-H in liver, lung, and testis tumor tissues; and HERV-P in colon and liver tumor tissues [13]. Specifically during hemato-oncological processes, several studies have reported overexpression of HERV genes and also the presence of retroviral particles in primary leukemia cells [15].

These considerations led us to study pol gene expression of HERV-P and HERV-R in bone marrow (BM) samples from a cohort of consecutive pediatric patients with leukemia. We expect that the results obtained will be able to provide some insight into the role of endogenous retroviruses in the pathogenesis of leukemia.

Material and methods

Patients and samples

BM cells were collected from 30 leukemia pediatric patients as reported in the Table 1 and from 20 pediatric healthy subjects with similar median age.

Table 1 Clinical data of patients entering in the study
Full size table

BM mononuclear cells were separated, total RNA extracted and reverse-transcribed as previously reported and [12]. All analyzed samples were BM cells which were collected from BM aspirates in the patients at the moment of the diagnosis and from BM collection in the pediatric related donors.

The study was conducted in accordance with Declaration of Helsinki for experiments involving humans, and informed consent was obtained in each case from the parents or their legal guardians.

Relative quantification by real-time PCR

Relative quantification of mRNA expression of pol genes was achieved by means of Taqman amplification and normalization to glyceraldehyde-3-phosphate dehydrogenase (GAPDH), which was chosen as a reference gene, using the ABI PRISM 7500 real-time system (Life Technologies, TX, USA). Approximately 100 ng cDNA were amplified in a 20 μl total volume reaction containing 2× GoTaq® Probe qPCR Master Mix (Promega, Madison, WI, USA), and 900 nmol of specific HERV-P primers (HERV-PF-5′-CAGGGTCCTCTCAAACCCTGT-3′) (HERV-PR-5′-TGGGTCAGCAAGGTATAGCGA-3′) and probe (HERV-P-6FAM-TGCGGCCAAAGTCCCAATCCATCCAGT-TAMRA) or HERV-R primers (HERV-RF-5′-GCACGAGTCAGCGGTGAAGA-3′) (HERV-RR-5′-GGGCTCAGGCAATTTCTGGT-3′) and probe (HERV-R-6FAM-CGGACCCGCTCAACCCAACAGTTCCG-TAMRA) or GAPDH specific primers 12. The amplifications were in a 96-well plate at 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Each experiment was repeated in triplicate.

Statistical analysis

The differences between HERV-P and -R expression were, respectively, evaluated comparing the patients with the donors by unpaired two-tailed Student’s t test with Welch’s correction using the Prism software (GraphPad Software, La Jolla, CA). In all analyses, p < 0.05 was taken to indicate a statistical significance between the analyzed groups.

Results

The study group enrolled 30 pediatric leukemia patients (15 acute myeloid leukemia (AML), 10 acute lymphoblastic leukemia (ALL), 4 chronic myeloid leukemia (CML), and 1 juvenile myelomonocytic leukemia (JMML)) diagnosed and treated in our center. Table 1 summarizes the patient’s characteristics. The median age of the patients and healthy cohort were 10.3 years, respectively, with a range of 0.8–18.6 and 3.9–16.4 years.

Because retro-transcription efficiency was equivalent in all samples, as suggested by GAPDH expression (data not shown), we made the hypothesis that the strong variations of signal intensity between mononuclear cells from leukemia patients and normal BM cells might be related to differences in HERV-P and -R pol transcriptional levels.

Relative expression of HERV-P and HERV-R pol gene in leukemia samples vs healthy donor

As shown in the Fig. 1, 9 of 30 samples (30%) showed a higher expression of HERV-P in comparison to healthy donors but the difference was not statistically significance (p = 0.6014). Conversely, HERV-R expression resulted significantly higher in the leukemia patients than healthy subjects (p < 0.0001). In particular, all samples out of 30 (100%) showed higher HERV-R levels than the average of healthy subjects.

Fig. 1
figure 1

Expression of the pol gene of HERV-P (a) and -R (b) evaluated by quantitative PCR real-time in the leukemia patients and the healthy donors expressed as 2−ΔΔCt. HERV-R expression resulted higher in the leukemia patients than healthy subjects .The symbol *** indicate a statically significant difference with p < 0.001 calculated by Mann-Whitney U test for the expression of HERV-R in the leukemia patients in comparison with donors

Full size image

Relative expression of HERV-P and HERV-R pol gene analyzing their expression in the different types of leukemia

The comparison of HERV-P expression between the ALL patients and the healthy donors, patients with myeloid malignancies (ML), and healthy donors and ALL and ML patients did not show any significant differences (p = 0.7226, 0.3681 and p = 0.3735, respectively) as shown in the Fig. 2a–c.

Fig. 2
figure 2

Expression of the pol gene of HERV-P (a–c) and -R (d–f) expression evaluated by quantitative real-time PCR. a, d Comparison of pol env expression between ALL patients and healthy donor subjects. b, e Comparison between myeloid malignancy patients and healthy subjects. c, f Comparison between ALL and myeloid malignancies patients. The symbol *** indicates a statistically significant difference (p < 0.001) calculated by an unpaired t test with Welch’s correction

Full size image

HERV-R expression was higher in the ALL and in ML patients in comparison to the donors, and the differences were both statistically significant (p < 0.0001) as shown in the Fig. 2d, e. This significant difference on HERV-R was not observed when we compared ML patients with ALL patients (p = 0.2810; Fig. 2f).

Relative expression of HERV-K and HERV-W pol gene analyzing their expression in the different types of karyotypes

Then, we analyzed data obtained from the amplification of the pol gene of HERV-P and -R in specific subgroups of patients with an altered karyotype. We compared their expression in BCR/ABL, FLT3 mutated patients or showing other translocations (t(6,11),(10,11), (8,16)) with leukemia patients without mutations. We observed a trend of higher expression of HERV-P in the subgroups with an altered karyotype, but without significant statistical differences (p = 0.4169 for HERV-P and p = 0.7468 for HERV-R, respectively, for the analysis between no mutations and BCR/ABL+, FLT-3+, and other translocations subgroups).

Discussion

The physiological role of HERV is still unknown, even if some HERVs have been found to show organ-specific expression, proposing an association between HERV and the development and differentiation of human tissues [16]. In malignancy, a relationship between HERV family type and histological type has been proposed [13].

The aim of this study was to evaluate the overall endogenous retroviral transcription profile in BM samples of a representative cohort of high-risk leukemia pediatric patients. Our findings show that HERV-R expression resulted higher in leukemia patients (ALL and LM) vs healthy subjects with the same median of age. These findings might suggest that the HERV-R pol gene could play a role in the cell proliferation activity. This specific expression in leukemia patients also suggests that HERV-R pol gene could be considered as a target for gene therapy.

The number of papers in the literature exploring the role of HERVs in leukemia patients are reduced and show data to be confirmed. In particular, Januszkiewicz-Lewandowska and colleagues observed a high relative expression of the env gene in acute ML suggesting a participation of env sequences in the pathogenesis of leukemia [11]. Fischer and colleagues recently presented that only 2 out of 25 chronic lymphocytic leukemia patients exhibited elevated HERV-K gag expression (5-fold greater than healthy donors) [15]. Ahn and Kim investigated the HERV expression between the cancer tissues and adjacent normal tissues and observed high levels of HERV-R, -K, P, and -R observed in various cancer tissues compared to the adjacent normal tissues [13]. Otherwise, it must be taken into account that chronic antigen stimulation could also trigger HERV overexpression as a consequence of cytokine production and inflammation [17, 18].

It remains also to ascertain whether the high HERV values we found in leukemia patients are related to the clinical outcome. The transcriptional activation of HERVs is a common feature in many human cancers. RNAs of HERV-K were found in testicular and ovarian germ cell tumors, as well as in gestational trophoblastic disease [19,20,21]. Indeed, in association to the DNA hypomethylation, an aberrant expression of syncytin-1 (HERV-W) has been found in biopsies from seminoma patients [22]. Remarkably, the expression of env of HERVs such as H, K, R, and P was found high in patients with breast cancer and resulted reduced during chemotherapy [23]. In prostate cancer, HERV-E and HERV-K expression was found in cancer tissue, cell lines, and peripheral mononuclear cells from patients [24]. HERV-K expression was found high in leukemia [9], in childhood acute lymphoblastic leukemia [10], and in pediatric high-risk leukemia patients [12]. In addition, HERV-W transcription levels were found elevated in lesions of the primary cutaneous in T cell lymphomas (CTCLs) [25, 26]. Recently, HERV-H transcripts were found be associated to colon cancer [27, 28]. Recently, HERV-K expression has been associated with soft tissue sarcoma [29] and hepatocellular carcinoma prognosis [30]. Moreover, HERV-K (HML-2) env expression was detected on seven pancreatic cancer cell lines and in 80% of pancreatic cancer biopsies, but not on normal pancreatic cell lines or on normal tissues [31].

To date, a growing body of findings supports the idea that HERVs may be a causative agent or at least a cofactor in carcinogenesis. Given to the potential ability to retrotransposon [32], it is the belief that HERV tumorigenicity could depend on retroviruses movement and consequently in de novo formation of proviruses in the host genome. Moreover, global DNA hypomethylation is a key event for the initiation and progression of cancer [33]. In agreement, the strong association of HERV expression with DNA hypomethylation [34, 35] and with genome instability reflects the transcriptome changes [36].

Oncogenic mechanisms mentioned above involve HERVs as regulators of cell gene expression, but HERVs could also exert a direct action through own proteins. The env protein of HERV-K could contribute to tumorigenesis inducing cell-cell fusion as suggested by studies in melanoma [37], while syncytin-1 seems to be involved in fusogenic events in endometrial carcinoma [38] and in breast cancer [39].

In pancreatic cancer has been also demonstrated that HERV-K (HML-2) Env protein plays a role in cell proliferation and tumorigenesis, as well metastasis formation, by regulating KRAS through hyperactivation of the RAS/MEK/ERK pathway [31].

Another mechanism by which HERVs could contribute to tumorigenesis regards the promotion of the immune escape, as in the case of transmembrane env protein of HERV-H that possesses immunosuppressive properties. [40]. Moreover, the interference of env of HERV-H improved immune responses in vitro and in vivo accompanied by tumor regression [41].

To the best of our knowledge, this is the first study of HERV-R pol mRNA expression in pediatric leukemia showing that this molecule is overexpressed in the BM cells of patients compared with healthy subjects. In this study, we found a relative overexpression of the endogenous retrovirus HERV-R in the BM cells from the preponderance of leukemia samples analyzed. After considering the different mechanisms potentially involving HERVs in carcinogenesis, we suggest that overexpression of HERVs might participate in leukemogenesis. Furthermore, precise and larger quantitative study is needed for a possible application of the HERVs pol gene to clinical use such as leukemia marker.

Although the functional associations have not been yet completely clarified, a mounting body of evidence allows to highlight the HERVs as potential biomarkers for some complex diseases and provide an adequate evidence that HERV reactivation, even if not the direct cause of the disease, acts mostly as triggering factor [42]. Moreover, HERV expression studies could be supplemented by tests such as Western blot or immune assays using anti-HERV-P or R antibody to support the above results. More studies are warranted to evaluate the potential impact of HERVs pol expression on leukemia patients and its role in the regulation of signaling pathways involved in lymphoblastic leukemogenesis.