Monitoring multiple myeloma by idiotype-specific peptide binders of tumor-derived exosomes

Multiple myeloma (MM) is a clonal B-cell malignancy accounting for more than 10% of hematologic cancers, and is characterized by the aberrant expansion of bone marrow plasma cells releasing a high level of monoclonal immunoglobulin (mIg) in the blood, so called paraprotein [1]. MM remains largely incurable due to the rapid development of aggressive, drug-resistant phenotypes [2]. Monitoring MM progression is a crucial step for determining the stage of disease and choosing the most appropriate therapy. In this context, there is an urgent need to develop novel diagnostic approaches allowing the non-invasive early detection of tumor growth and the efficient monitoring of tumor progression [3].

First identified in the mid-80s [4] and initially classified as unfunctional “garbage bags” containing unwanted cellular constituents, exosomes represent a promising tool for novel diagnostic options in the diagnosis of malignant diseases [5]. Indeed, recent evidence demonstrated the utility of microvesicles in detecting relapse weeks before existing clinical tests, highlighting the sensitivity and capacity for microvesicles in monitoring disease progression and minimal residual disease in myeloma patients [6]. Exosomes are vesicles of 30–130 nm in diameter released by different cell types and detectable in all biological fluids [7] and supernatants of cultured cells [8]. Exosomes contain a wide range of RNA and proteins, playing an important role in cell-to-cell communication [9]. In particular, exosomes are involved in the regulation of the immune response, antigen presentation [10], tumor survival [11], cell migration [12], tumor invasion [13], cell differentiation and angiogenesis [14]. Reflecting the genomic and proteomic profile of their parental cells, circulating serum exosomes are potential biomarkers in predicting cancer burden with relevant impact for personalized therapy [15]. Although several methods have been developed for exosome purification, none of them clearly distinguish between normal and tumor-derived exosomes (TDEs), or avoid contamination by shed membrane vescicles [16]. Even if the mechanism of expression remains not completely defined, it is worthwhile that MM-released exosomes constitutively express on their surface the immunoglobulin of B-cell receptor (Ig-BCR) derived from the parental tumor B-cell, and thus they can be reliable tumor markers [17, 18].

In the last few years, we successfully validated the screening of random peptide libraries (RPLs) as a method to identify peptides binders of soluble immunoglobulins (Igs) [19] transmembrane receptors [20, 21] and biomaterials [22]. In particular, we identified peptide binders of the Ig-BCR idiotypic determinants (hereafter named “Id-peptides”) that are expressed on the surface of the A20 murine B-cell lymphoma, which revealed to be sensitive tools for in vivo tumor detection and tumor-specific delivery of radionuclides, fluorophores, siRNAs and nanoparticles [23].

In this study, we addressed the question whether MM-released exosomes detected by Id-peptides could allow a more efficient monitoring of tumor growth compared to the standard paraprotein assay. To this end, we measured the tumor growth and serum MM-released exosomes in vivo in the 5T33MM murine model [24]. 5T33MM-engrafted mice develop a highly aggressive MM form, presenting biological and genetic characteristics similar to the human disease, and thus it represents one of the most reliable MM preclinical model [25].

The experimental plan undertaken is shown in Fig. 1. Briefly, we screened a C7C M13 phage-displayed RPL [NEB] using as bait the Igs isolated from cultured 5T33MM cells. The random peptide insert was flanked by a pair of cysteines resulting in cyclized peptide exposed on M13 coat protein. After three rounds of affinity selection, single phage clones were purified and tested by enzyme-linked immunosorbent assay (ELISA) for binding to the 5T33MM Igs or control Igs, as previously described [26]. The insert peptide sequence from selected phage, the percentage of clonal identity, and K_D values of identified synthetic Id-peptides are shown in Table 1. Based on the highest affinity binding to 5T33MM Igs, the insert amino acid sequence of phage clone 5 (CIGNSNTLC) was used for large-scale synthesis of the p5 peptide in order to evaluate the binding properties outside of the phage context. To this end, the biotinylated-p5 peptide was incubated with 5T33MM-Igs coated plates at different concentrations, and revealed with streptavidin-conjugated alkaline phosphatase. The p5 peptide showed a concentration-dependent specific reactivity to the cognate antibody, and did not react against an A20 secreted IgG or polyclonal mouse Igs (Fig. 2a).

By flow cytometry, we analyzed the specific binding of FITC-conjugated p5 peptide to the 5T33MM, A20 and IM9 cells expressing the surface Ig-BCR. An irrelevant peptide (CGGNGPGLC) was included as a negative control (pCNT). The p5 peptide recognized 5T33MM target cells in a dose-dependent manner, while the pCNT peptide did not (Fig. 2b). The specific Id-peptide binding to 5T33MM was verified by testing other B-cells, including the IM9 and A20 cell lines and primary B cells from a multiple myeloma patient and healthy donor (Additional file 1: Figure S1). To visualize the binding of the p5 peptide to the surface Igs, we performed confocal microscopy of 5T33MM cells stained with the FITC-conjugated p5 peptide and the anti-mouse IgG detecting the surface Igs. The p5 peptide bound to the 5T33MM cell surface and also localized inside the cells (Fig. 2c). Since this analysis was conducted on focal planes of non-permeabilized cells, the intracellular localization of the p5 peptide was a likely consequence of the BCR-mediated internalization.

Exosomes were then purified from the supernatant of cultured 5T33MM, A20 and IM9 cells and from serum of MM patient and healthy donor. By scanning electron microscopy (SEM), the exosomes had a rounded shape (30–120 nm diameter) with some agglomerations occurring in the drying process during sample preparation (Fig. 3a). Confirmation of the size homogeneity of vesicles was verified using a Zetasizer Nano ZS90 (Fig. 3b). Purified exosomes were also analyzed by Western blotting using antibodies against typical exosomal markers, such as CD81, CD63, including anti-IgG antibody to verify the IgG expression on TDEs (Fig. 3c).

To validate the reliability of the p5 peptide in targeting the 5T33MM-released exosomes, we used an immuno-capture approach based on anti-CD63 decorated streptavidin magnetic nanoparticles (SMNPs) to trap exosomes [27]. The size of the exosomes-functionalized SMNPs (9.1 μm of diameter) supported the use of flow cytometric detection. Front/side scattering (FSC/SSC) plot of RedExo-labeled and un-labeled exosomes is shown in Fig. 4a. 5T33MM–derived exosomes were equally detected using anti-mouse IgG or FITC-conjugated p5 peptide, while undetected when stained with control peptide, or left unstained (Fig. 4b-e). As controls, the FITC-conjugated p5 peptide did not detect the exosomes released from the human IM9 multiple myeloma and murine A20 B-lymphoma cells (Fig. 4). Further, this analysis provided the evidence of similarity between murine and human multiple myeloma derived-exosomes in terms of IgG expression (Fig. 4).

For translating the analysis of MM-released exosomes to in vivo preclinical models, two groups of 10 mice were respectively intravenously inoculated with 5T33MM cells and with A20 cells. A third group of 10 un-injected mice were used as negative control. At day 0 and each 7 days, blood samples were collected by retro-orbital bleeding for measuring the serum MM-released exosomes and paraprotein levels. According to the Kaplan–Meier survival curve the fatal outcome of mice progressively occurred between 21 and 40 days post cells injection (Fig. 5a). By ELISA, the serum paraprotein was detected 21 days after tumor inoculation (Fig. 5b). Exosomes were purified from serum samples, conjugated with SMNPs and analyzed by flow cytometry using the FITC-conjugated p5 peptide. Among 5T33MM serum-derived exosomes, we identified a p5 peptide-positive exosome sub-population 7 days after tumor inoculation, which increased in a time-dependent manner until death (Fig. 5c). Serum-derived exosomes from A20 harboring mice were also purified and used as control (Fig. 5d). The percentage of p5-FITC positive TDEs population in all mice is reported in the Additional file 1: Table S1. These results indicated that analysis of serum MM-released exosomes allowed an earlier detection of MM growth compared to the conventional measure of paraprotein.

In conclusion, we describe the development of a method allowing the rapid and simple detection of MM-released exosomes using the Id-peptide binders of the Igs expressed by tumor B-cells. Being the IgBCR expressed on the surface of tumor B cells as well as tumor-derived exosomes, our methodology could be extended to the most common B cell neoplasia with potential application for monitoring the minimal residual disease. Furthermore, compared with the standard exosome purification and immuno-capturing systems, the advantage of our method relies on the use of fluorescent-labeled Id-peptides coupled to SMNPs, which represent a more powerful and sensitive tool for tumor detection. Based on specificity, flexibility, cost effectiveness and modularity, our capture system is ideal for novel non-invasive applications in personalized medicine and could be easily extended to other B-cell malignancies. Our experimental approach could open new avenues in devising lab-on-a-chip platforms and liquid biopsy options for an effective early detection of TDEs shed into the blood by tumor B-cells.