Introduction

MicroRNAs (miRNAs) are endogenous non-coding single-stranded RNAs that negatively regulate eukaryotic gene expression through binding to complementary sequences in the 3′ untranslated regions of target mRNAs. It is currently believed that miRNAs regulate over 30 % of all human genes [1]. Thus, miRNAs play important regulatory roles in different pathways. It has been shown that the tissue levels of specific miRNAs correlate well with the pathological development of several different cancers, including hematological malignancies [2–7].

The search for non-invasive biomarkers for the diagnosis and management of cancer is a crucial part of cancer research. Studies have shown that miRNAs are present in plasma and serum in a stable form and that the expression levels of particular miRNAs differ between cancer patients and healthy controls [8–10]. Circulating miRNAs have been developed as a novel class of blood-based biomarker to diagnose and monitor tumors.

Several studies have investigated the role of miRNAs in multiple myeloma (MM). For example, the miR-17-92 cluster plays a role in MM tumorigenicity by targeting the apoptosis facilitator BCL2L11 in myeloma cells. Moreover, patients with high levels of miR-17, miR-20, and miR-92 had shorter progression-free survival times compared with those with low levels of these miRNAs [11]. miRNA-29b induces apoptosis of MM cells by downregulating Mcl1 [12]. The miR-106b–25 cluster, miR-181a/b, and miR-19a/b were upregulated in plasma cells of MM compared with those of healthy controls. In addition, the miR-106b–25 cluster, miR-32, miR-181a/b, and miR-19a/b can downregulate the interleukin 6 inhibitor SOCS1 and the p53 pathway component p300-CBP-associated factor; suppression of these miRNAs resulted in reduced myeloma tumor growth in nude mice [13].

To date, there have been no detailed studies of circulating miRNA expression in MM. Here, we conducted TaqMan quantitative reverse transcription PCR (qRT-PCR) analysis for global miRNA profiling in the plasma of MM patients and healthy donors. This approach identified six miRNAs (miR-148a, miR-181a, miR-20a, miR-221, miR-625, and miR-99b) that were significantly upregulated in MM patients compared with in healthy controls. Using quantitative analysis of these miRNAs in a large series of MM patients, we found that the levels of five of them (miR-148a, miR-181a, miR-20a, miR-221, and miR-99b) were increased in the plasma of patients. We correlated the expression of these five miRNAs with the patients’ clinicopathological data and survival and found that several were potential prognostic biomarkers for MM.

Methods

Study cohort and sample collection

All patients were diagnosed according to the NCCN clinical practice guidelines for MM [14]. Written informed consent was obtained from all participants after explanation of the nature of the study, and the study was approved by the research ethics board of our hospital.

We tested plasma samples from 40 patients diagnosed with MM, taken before chemotherapy: 12 were used for the microarray tests and 28 formed the validation set. The patient details are shown in Tables 1 and 2. From healthy controls, we used eight plasma samples for the microarray tests and 12 in the validation set.

Table 1 Summary of clinical details of MM patients used for microarray analysis of plasma samples (n = 12)
Full size table
Table 2 Summary of clinical details of MM patients used for q-PCR analysis of plasma samples (n = 28)
Full size table

Venous blood was collected in EDTA tubes (BD Biosciences, Franklin Lakes, NJ). The blood was centrifuged at 800g for 10 min, and the plasma was removed into RNase/DNase-free Eppendorf tubes, followed by a 10-min high-speed centrifugation at 16,000g to completely remove the cell debris. The plasma was transferred to a fresh tube and stored at −80 °C. All blood samples were processed within 2 h after they were obtained.

RNA isolation from human plasma

We isolated total RNA, which includes miRNAs, using the mirVana Paris RNA Isolation kit (Ambion, Austin, TX), following the manufacturer’s protocol for liquid samples. The RNA was stored at −80 °C until use.

miRNA profiling

miRNA expression in 12 MM patients and 8 healthy controls was profiled by accurate quantitation of 667 human miRNAs using TaqMan Low-Density Arrays (Part Number: 4400238; Applied Biosystems, Foster City, CA).

Total RNA from the plasma was subjected to RT (reverse transcription) using a TaqMan miRNA Reverse Transcription Kit and the TaqMan miRNA Multiplex RT Assays, Human Pool Set (Applied Biosystems). The diluted product of RT reaction was combined with TaqMan Universal Master Mix (Applied Biosystems) and then loaded onto the TaqMan Low-Density Arrays. Real-time PCR was performed on an ABI PRISM 7900HT instrument (Applied Biosystems), using the manufacturer’s recommended cycling conditions. Ct data were obtained using SDS v2.2 software (Applied Biosystems) with the default threshold settings.

We used DataAssist v2.0 software (Applied Biosystems) to obtain a suitable gene for normalization, entering the Ct values for the 50 most highly expressed miRNAs in the whole sample set. The expression level of the small nuclear RNA U6 was the most stable; there was no significant difference in the U6 Ct value between control and MM samples. The ΔCt method was used for analysis [ΔCt = average Ct (miRNA of target) − average Ct (U6)].

Differentially expressed miRNAs were identified using the two-class (unpaired) format within the Significance Analysis of Microarrays (SAM) platform [15] with a 20 % false discovery rate (FDR) threshold.

qRT-PCR for independent series validation

Total RNA was reverse-transcribed to cDNA using TaqMan MicroRNA Reverse Transcription Kit with miRNA-specific primers (Applied Biosystems) as described previously [9]. Then, real-time PCR was performed using the TaqMan MicroRNA Assay kit (Applied Biosystems) according to the manufacturer’s instruction to detect miR-148a, miR-181a, miR-20a, miR-221, miR-625, and miR-99b levels in plasma. Data normalization was performed as described above.

Cytogenetic classification of myeloma samples

To investigate miRNA expression in the different genetic subtypes of MM, 28 cases were cytogenetically classified using microfluorescence in situ hybridization (FISH) [16]. Cytogenetic abnormalities involving 13q deletions and immunoglobulin heavy-chain gene (IGH) rearrangements were investigated. The FISH probes were produced by Vysis (Abbott Diagnostics, Berkshire, UK). The distribution of cytogenetic abnormalities in the 28 MM patients is summarized in Table 2.

Statistical analysis

Comparison of global plasma miRNA levels in patients with those in healthy donors was performed using the one-sided Wilcoxon test; the expression levels of the selected miRNAs were compared using the Mann–Whitney independent t test (GraphPad Prism v.5.0), using the median value as the high/low cut-off point.

Relapse-free survival (RFS) curves were plotted separately for patients with low versus high expression of each miRNA, using the Kaplan–Meier method. Curves were compared by univariate (log-rank) analysis (GraphPad Prism v.5.0). RFS times were calculated from the time of diagnosis to the date of relapse, death, or last contact. P values less than 0.05 were considered significant.

Results

Profiling of plasma samples with miRNA TaqMan arrays

Before running the arrays, plasma RNA samples were probed with TaqMan assays for U6. The samples for which the Ct value was above 35 cycles were discarded from the array assays. Twelve plasma RNA samples from MM patients and eight from control individuals remained.

Total miRNA expression, which was determined from the mean expression levels of 135 expressed miRNAs, was higher in samples from patients with MM than in samples from healthy donors (P = 0.022; one-sided Wilcoxon test; Fig. 1).

Fig. 1
figure 1

Global expression of miRNAs in MM samples (n = 12) was significantly higher (P = 0.022) than in control samples (n = 8). The box extends from the 25th to the 75th percentile; the line in the box indicates the 50th percentile; the whisker caps indicate the maximum and the minimum values

Full size image

Table 3 shows the six miRNAs whose expression levels were significantly different (FDR < 20 %, SAM) in disease and healthy conditions. All six were consistently expressed at higher levels in samples from patients newly diagnosed with MM than in samples from healthy donors.

Table 3 MiRNAs differentially expressed in MM patients with respect to healthy controls
Full size table

Validation of miRNA differential expression

To validate the array results, the expression levels of the six selected miRNAs (miR-148a, miR-181a, miR-20a, miR-221, miR-625, and miR-99b) were measured by qRT-PCR in plasma from 28 healthy donors and 12 patients with MM. The expression level of five of them (miR-148a, miR-181a, miR-20a, miR-221, and miR-99b) was significantly different between the two groups (P values: 0.020, 0.001, 0.030, 0.040, and 0.006, respectively; Mann–Whitney test; Fig. 2). No significant difference was observed for miR-625 between MM and control samples (P = 0.19; Mann–Whitney test; Fig. 2).

Fig. 2
figure 2

Box plots showing miRNA expression levels in the plasma of MM patients and controls. The expression levels of a miR-148a, b miR-181a, c miR-20a, d miR-221, e miR-625, and f miR-99b in MM (n = 28) and control (n = 12) samples were measured by qRT-PCR. The box extends from the 25th to the 75th percentile; the line in the box indicates the 50th percentile; the whisker caps indicate the maximum and the minimum values

Full size image

Association of miRNA expression with genetic subtype

It is increasingly evident that the genetic features of the tumor cells largely dictate the clinical heterogeneity of MM [17, 18]. To investigate miRNA expression in the different genetic subtypes of MM, 28 cases were cytogenetically classified using micro-FISH [16]. A high level of miR-99b expression was associated with a t(4; 14) (IGH;FGFR3) translocation (P = 0.029; Mann–Whitney test; Fig. 3a). Moreover, low expression of miR-221 was associated with del(13q) (P = 0.04; Mann–Whitney test; Fig. 3b).

Fig. 3
figure 3

Box plots showing miRNA expression levels in the plasma of genetic subtypes of MM patients. The expression levels of a miR-99b and b miR-221 were measured by qRT-PCR. The box extends from the 25th to the 75th percentile; the line in the box indicates the 50th percentile; the whisker caps indicate the maximum and the minimum values

Full size image

Association of miRNA expression with RFS

We then performed Kaplan–Meier analyses to assess the relationship between miRNA expression levels and clinical outcome, using the median value as a high/low cut-off. The median follow-up was 31.8 months (range 3–52). High expression levels of mir-20a and miR-148a in MM patient plasma were associated with shorter RFS (P = 0.01 and P = 0.02, respectively; Fig. 4).

Fig. 4
figure 4

Kaplan–Meier survival curves of relapse-free survival in MM patients based on high or low (median) miRNA expression levels. The curves were compared by univariate (log-rank) analysis. a miR-148a, b miR-20a

Full size image

Discussion

Aberrant miRNA profiles in plasma have been reported for several types of hematological malignancy [19–21]. Changes in the expression levels of specific circulating miRNAs in plasma offer the potential for tumor detection, classification, and prognosis determination [9, 10].

In our study, we used microarray technology to elucidate the complete miRNome (miRBase version 10.0) of plasma samples from 12 MM patients and compared the miRNA expression levels with those from eight healthy controls. The total miRNA expression level was higher in samples from MM patients than in those from controls, and six miRNAs (miR-148a, miR-181a, miR-20a, miR-221, miR-625, and, miR-99b) were specifically upregulated in MM. Then, we verified the increased levels of five of them (excluding miR-625) in more clinical samples. Moreover, our data showed that the levels of miR-20a and miR-148a were related to the prognosis of MM and that miR-221 and miR-99b were associated with the karyotype of the disease.

miR-181a has been previously shown to act as an onco-miRNA in MM cells, modulating the expression of proteins critical to myeloma pathogenesis [13]. However, although we found significantly higher levels of miR-181a in MM patients than in controls, we failed to find a relationship between its gene expression and the karyotype or RFS.

Recent studies have suggested that miR-20a (belonging to the miR-17-92 cluster) may function as an oncogene [22, 23]. This miRNA may be involved in lung, colon, and prostate tumors, chronic leukemia, and mixed lineage leukemia-rearranged acute leukemia [23–27]. A recent study reported that the miR-17-92 cluster activated by Myc could down-modulate proapoptotic proteins and subsequently inhibit apoptosis in an MM cell line [11]. In our study, the expression of miR-20a was higher in samples from MM patients than in control samples; this is consistent with the findings of other studies [13, 28, 29]. Moreover, we detected higher miR-20a levels in the group of MM patients with a shorter RFS time. Thus, we suggest a possible association between miR-20a and a poor prognosis of MM.

Several independent reports have shown that miR-148a is an anti-oncogene that suppresses tumor cell proliferation and metastasis in human gastric and ovarian cancer [30–32]. However, recent research reported that miR-148a promotes gastric cancer cell growth by targeting p27 [33]. Moreover, Murata et al. [34] found that miR-148a promoted LNCaP prostate cell growth by repressing CAND1 expression. In our study, miR-148a expression was upregulated in MM patient plasma than in control plasma. Furthermore, we showed that high expression of miR-148a was related to a shorter RFS time in our cohort of patients. The mechanisms by which miR-148a is associated with a worse clinical outcome should be explored further.

Low expression of miR-221 was associated with the presence of del(13q). Increasing evidence indicates a crucial role for chromosome 13 deletion as a prerequisite for the clonal expansion of tumors [35, 36]. Therefore, this finding suggests a tumor-suppressor role for miR-221, although it has been shown to function as an oncogene in a wide range of cancers [37–39]. A possible explanation for this apparent paradox is that miR-221 has been proposed to act as either an anti-oncogene or an oncogene depending on the cellular context.

The chromosome abnormality t(4; 14) is a suggested marker of poor prognosis in MM patients [40, 41]. Here, we found higher levels of miR-99b in patients with t(4; 14); this is consistent with the findings of other studies [29, 42]. Our study supports the oncogenic role of miR-99b that has been suggested in previous studies [43, 44].

To our knowledge, this is the first report of quantitative assessment of plasma miRNAs in MM patients to find that the expression levels of particular miRNAs are related to genetic subtype and survival time. If our findings are verified in other populations, these miRNAs could be useful non-invasive biomarkers in MM.