Tuberculosis 95 (2015) 26e30

Contents lists available at ScienceDirect

Tuberculosis journal homepage: http://intl.elsevierhealth.com/journals/tube

REVIEW

Deciphering the role of exosomes in tuberculosis Nicole A. Kruh-Garcia a, Lisa M. Wolfe a, b, Karen M. Dobos a, * a b

Department of Microbiology, Immunology & Pathology, Colorado State University, Fort Collins, CO 80523, USA Proteomics and Metabolomics Facility, Colorado State University, Fort Collins, CO 80523, USA

a r t i c l e i n f o

s u m m a r y

Article history: Received 15 August 2014 Accepted 27 October 2014

Exosomes were originally described as small vesicles released from reticulocytes during the maturation process. These 40e200 nm microvesicles were hypothesized to be a mechanism for the removal of membrane proteins in lieu of intracellular degradation by Harding et al. (1984) and Johnstone et al. (1987) [1,2]. Exosomes can be distinguished from other extracellular vesicles (ectosomes, apoptotic blebs) based on their size and the protein indicators intercalated in their membrane (also, linking their derivation from the endocytic pathway) by Simpson (2012) [3]. The exact role which exosomes play in cell-to-cell communication and immune modulation is a topic of intense study. However, the focus of most reports has been directed towards discovering aberrations in exosomal protein and RNA content linked to disease onset and progression, and also primarily related to cancer. Nonetheless, exosomes are now documented to be released from a wide variety of cell types by Mathivanan et al. (2012), Simpson et al. (2012) and Kalra et al. (2012) [4e6] and have been isolated from all bodily fluids; thus, exosomes are an excellent source of biomarkers. Here we describe the discoveries related to the role exosomes play in tuberculosis disease, as well as translational work in vaccine development and how circulation of these dynamic vesicles can be harnessed for diagnostic purposes. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Biomarkers Mass spectrometry Exosomes Proteomics

1. Introduction As early as 1994, Russell et al. described the ability of Mycobacterium (Mycobacterium avium/Mycobacterium tuberculosis) to traffic bacterial components, specifically lipoarabinomannan (LAM), in distinct intracellular vacuoles beyond the bacteriacontaining compartments [7]. Mycobacterial LAM and phosphatidylinositol mannoside (PIM) were found to accumulate in LAMP-1 positive vacuoles in late endosomal compartments and multivescular bodies (MVB), which later fused with the cellular membrane to release LAM/PIM-containing microvesicles to the extracellular environment [8]. The released vesicles were shown to have the ability to transfer the bacterial components to neighboring uninfected macrophages. Follow-up studies by Schorey et al. utilized flow cytometry to reveal a positive association of the released vesicles with the endosomal markers: LAMP1, LAMP2 and MHCII, as well as the exosomal-marker, tetraspanin CD81 [9], implicating a mechanism in which the internalized bacteria could communicate with other host cells.

* Corresponding author. Tel.: þ1 970 491 1891; fax: þ1 970 491 1815. E-mail address: [email protected] (K.M. Dobos). http://dx.doi.org/10.1016/j.tube.2014.10.010 1472-9792/© 2014 Elsevier Ltd. All rights reserved.

Extracellular exosomes charged with mycobacterial constituents have the ability to stimulate naïve macrophages in a proinflammatory manner, by inducing the production TNF-a, RANTES and iNOS [9,10]. Similarly, these exosomes can activate both CD4þ and CD8þ T cells, consistent with formation of a strong acquired immune response and indicative of an alternative route of antigen presentation to these cells in lieu of MHC presentation by macrophages and dendritic cells [11]. There appear to be multiple mechanisms in which exosomes can interact with other cells. Notably, exosomes shed from mycobacterium-infected antigen presenting cells (APCs) display MHC-II and have the ability to present processed antigen, as shown using a T cell line clonally restricted to M. tuberculosis Antigen 85B [12]. Later, it was shown that exosomes also contain whole mycobacterial proteins, including Antigen 85 complex proteins, HspX, DnaK, and a number of other mycobacterial proteins [13,14]. Whether a single exosome can contain both processed and unprocessed antigen is unknown, although it is likely that the form of antigen will change the cell and cellular response targeted by the exosome. It has been shown that the host heat shock protein 70 (Hsp70) is increased specifically in exosomes from M. tuberculosis-infected cells and is suggested to contribute to the pro-inflammatory response [15]. One can hypothesize that this response is exacerbated by the presence of bacterial heat shock proteins, including the Hsp70 homolog from

N.A. Kruh-Garcia et al. / Tuberculosis 95 (2015) 26e30

M. tuberculosis, DnaK (Rv0350), as well as HspX (Rv2031c) d which have been identified from exosomes isolated from M. tuberculosisinfected macrophages and serum recovered from small animals experimentally infected with M. tuberculosis [13,14]. Exosomes released from M. tuberculosis-infected macrophages also have an inhibitory effect on cellular immune responses associated with protective immunity, specifically hindering INF-gregulated pathways which activate naïve macrophages [16]. The survival of mycobacteria within the host is a delicate balance of immune activation such as requisite phagocytosis through specific receptors and avoidancedas with interruption of phagosomeelysosome fusion [17]; it is clear that exosomes may play a more significant role than originally believed. 2. Clinical potential e vaccines The initial implications regarding the utility of exosomes as cancer vaccines are over a decade old. The first clinical trials using exosomes are now underway; phase I studies highlight the feasibility of this form of immunotherapy and importantly, that they can be administered safely and are well-tolerated [18,19]. Exosomes contain both adjuvant and antigenic properties, and as previously mentioned, have the ability to stimulate a robust acquired immune response [11]. Likewise, exosomes released from mycobacteriainfected macrophages can generate memory T cells in mice subsequent to intranasal vaccination e indicating their potential utility as an M. tuberculosis vaccine [20]. In 2010 it was noted that exosomes harvested from macrophages cultured with M. tuberculosis culture filtrate proteins (CFP) elicit a parallel T cell response in vitro and ex vivo [13]. Exosomes from macrophages co-cultured with CFP (CFP-exosomes) are significantly easier to prepare since they do not require manipulation with a live virulent organism; hence this option is much more plausible for the downstream up-scaling, standardization, and commercialization required for a vaccine or immunotherapeutic reagent. Mice vaccinated with CFP-exosomes show a comparable decrease in bacterial burden to Bacillus Calmrin (BCG) vaccinated mice upon challenge with virulent etteeGue M. tuberculosis [20]. In addition, mice boosted with CFP-exosomes post-BCG vaccination showed a significant decrease in bacterial counts in the lung upon challenge [20]. CFP-exosomes contain several antigens which overlap with those already in various stages of clinical trial (specifically, antigen 85A and 85B). It is not clear at this point whether there is a specific component(s) in the CFPexosomes or rather if the complexity of several highly immunogenic mycobacterial proteins correlates with protection; further work is underway to address these issues. 3. Clinical potential e diagnostic biomarkers Until 2010, evidence to support the incorporation of mycobacterial proteins in exosomes released from infected macrophages was exclusively from western blots. To support and build upon these findings, a comprehensive proteomic query was performed on exosomes released from macrophages challenged with live M. tuberculosis and dead g-irradiated M. tuberculosis. The initial proteomic discovery experiments demonstrated that proteome of these exosomes (exo-proteome) is predominately comprised of host proteins with only hundreds). Due to the overwhelming ratio of host to bacterial peptides (99:1) observed in cell culture models, the feasibility of an untargeted MS approach on more complex samples would be infinitely more challenging and likely to yield significantly less bacterial data. Fortunately, our previous studies in model systems afforded the confident identification of a number of mycobacterial proteins, providing a discreet list of candidate biomarker proteins to follow in human studies [27]. Thus our subsequent mass spectrometry work flows using human samples employed a targeted mass spectrometry approach [27]. Targeted methods, such as multiple reaction monitoring (MRM) MS, requires a priori knowledge of the protein/peptides to be screened for in a sample and has extraordinary specificity and sensitivity, allowing the abundant host peptides to be ignored while detecting only the defined set of proteins indicated. MRM assays were designed to target the presence of the top mycobacterial candidates (33 M. tuberculosis proteins represented by 76 peptides [27]). These methods were tested on exosomes (purified using ExoQuick) isolated from a series of guinea pig fluids that were collected 30 days after a low aerosol dose infection [28]

Figure 1. Nanoparticle tracking analysis via a NanoSight system comparing exosomes purified via (A) sucrose gradient (methods defined in [9]) and (B) Exoquick (methods defined in [27]) from a split human serum sample. Three traces represent the measurement made in triplicate.

and included bronchoalveolar lavage (BAL) fluid, urine and serum. All three fluids were collected from each animal when possible. The three fluids varied widely in exosome yield and protein content: with sera containing the highest levels of protein (~13 mg/mL), moderate levels in BAL fluid (0.26 mg/mL), and the lowest levels in urine (0.078 mg/mL). One mg of each exosome sample was injected to collect spectra indicating the presence of each of the 76 mycobacterial peptides using the MRM methods. Several peptides were present in all three sample types from multiple guinea pigs, including the GroES peptide (Figure 2). Urine samples had the least matrix effect (interference from co-eluting host peptide); the low number of mycobacterial peptides identified combined with the low exosome yield makes this sample type less suitable for additional discovery and downstream diagnostic utility with human specimen. Of the three fluids, exosomes harvested from sera yielded the highest number of mycobacterial peptides identified (our unpublished observations) e substantiating further exploration in human sera samples. The variety of proteins contained in serum, ranging from mg/mL down to pg/mL, can pose a significant hurdle when the biomarker proteins of interest are at the lower end of detection; however, the purification of exosomes allows us to reproducibly detect bacterial peptides of low abundance from such samples [29]. Exosomes isolated from pooled mouse BAL fluid samples were analyzed by MRM-MS and the data for several mycobacterial proteins was compared to that previously reported by spectral counting (SpC; [14]). From this analysis we observed that the relative abundance of proteins as measured by SpC was a bit deceiving in that the final mouse BAL time points (days 56 and 112) consistently demonstrated lower relative abundances for several mycobacterial proteins when compared to the initial two time points (days 14 and 28). In contrast, the MRM-MS detected several of these same proteins at sustained and heightened levels throughout the infection (our unpublished observations). This is exemplified in Figure 3, in which the relative quantification of the Ag85a and AcpM proteins benefit from the sensitivity of MRM-MS. Based on the promise of studies using cell culture and small animal models, MRM assays were further developed and extended to analyze banked human specimens. Twenty of the mycobacterial proteins queried were found in varying permutations across the exosomes of TB patients. Antigens 85b, BfrB, GlcB and Mpt64, among others, were identified by multiple peptides [27]. The study included two subsets of TB positive patients, those with pulmonary (PTB) and extrapulmonary tuberculosis (EPTB). Interestingly, pulmonary and extrapulmonary disease could be distinguished by the peptide markers found in their exosomes. For example, there was evidence of one or more MPT64 peptide in over 70% of PTB samples. When compared to non-TB patients, three peptides from Mpt64: VYQNAGGTHPTTTYK, AFDWDQAYR and EAPYELNITSATYQSAIPPR were significant (0.05, 0.04 and 0.04, respectively). Peptides from 5 different proteins were significant with respect to their presence in EPTB versus non-TB exosomes. The two peptides present in majority of the pulmonary samples: DGQLTIK and SEFAYGSFVR are from HspX protein and from this set of patients (culture-positive, from Uganda) was clear the strongest marker of active disease (p ¼ 0.0006 and 0.0007, respectively). HspX was detected in 90% of EPTB patients, versus ~50% of PTB sera exosomes [27]. The primary goal of this initial study was to find evidence of the candidate mycobacterial peptide markers in human serum exosomes during active disease. As mentioned, both active disease classes (PTB and EPTB) were compared to non-TB patients. Upon further stratification, about half of the non-TB patients showed evidence of latent disease (LTBI). While truly negative patients contained on average 3 peptides were detected per LTBI exosome sample. While this data is preliminary,

N.A. Kruh-Garcia et al. / Tuberculosis 95 (2015) 26e30

29

Figure 2. Screen shot from the Skyline analysis program depicting the presence of the GroES peptide DVLAVVSK by the three transition ions: 616.4028 (blue), 432.2817 (red), and 503.3188 (purple), in exosomes isolated from BAL fluid (A, D), serum (B, E) and urine (C, F) of two guinea pigs (guinea pig #136, A-C and guinea pig #138, D-F) 30-days post-infection with M. tuberculosis. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Figure 3. Comparison of SpC (red line; right axis) and MRM-MS (blue bars; left axis) for the two myocbacterial proteins: A. Ag85a and B. AcpM. SpC, which are the total spectra observed for all peptides within a protein, show an overall decrease for both mycobacterial proteins monitored by day 56. The MRM data, represented by the sum of the peak areas of the three transition ions monitored for a single peptide, show a sustained presence of the two proteins through day 112. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

the promise of serum markers indicative of separating active and latent disease is very tantalizing for a number of reasons. First, these data suggest that latent bacilli are active and responding to environmental stimuli in a measurable way. Second, a diagnostic assay that identifies LTBI persons within endemic populations would potentially impact treatment and monitoring of at risk groups and thereby interrupt disease transmission. The diagnostic, prognostic and therapeutic value of microvesicles with respect to tuberculosis has yet to be fully realized. Interestingly, several consistencies have been observed amongst all reports ranging from pathogenesis to diagnostics, published to date. Notably, that exosomes isolated from mycobacterial infections using experimental models through to human subjects, contain a number of M. tuberculosis products that are conserved regardless of the system studied ([8,9,13,27], for example), that these exosomes interact with the host immune system [10e12,16,20,25] and serve as sentinels of a productive intracellular mycobacterial infection [7,10,12,27]. Despite these consistencies, the precise role of exosomes in a natural infection remains unknown, and may be

controversial since exosome concentration is linearly correlated with lung burden in experimental models ([14]). Thus, many unanswered questions remain. Specifically, what cellular populations are the major contributors to the exosomes that contain mycobacterial products? The serum exosomes isolated in the aforementioned animal and human studies are derived from all over the body. While infected cells release exosomes decorated with bacterial proteins (as exemplified here), the localization of these bacterial proteins is unknown, thus prohibiting their use to separate the exosomes which are shed from infected cells from those shed as part of normal homeostatic processes. While several markers, including HSP70 and tetraspanins (CD63, CD9), appear to be conserved on all exosome surfaces [1e3], there are unique proteins reflective of the cell type which the exosome originates [4,5,30]; these may be used to isolate exosomes from specific cell lineages [31]. From a tuberculosis diagnostics perspective, relevant lineages include cells such as macrophages, dendritic cells, and lung epithelial cells. However, this type of information is still limited-even within the established ExoCarta database (in fact,

30

N.A. Kruh-Garcia et al. / Tuberculosis 95 (2015) 26e30

there are only two proteins listed: heat shock proteins 1 and 8). Regardless of the site of origin, preliminary data suggests that the host protein content changes during infection e leading to the hypothesis that host proteins could be also used as biomarkers of active disease. 4. Concluding thoughts To summarize, exosomes are an ideal source for discovery and assay design for a number of therapeutic strategies. They are easily purified, reduce the complexity of the biofluid source and enrich mycobacterial proteins (and other products, such as mycobacterial lipids [8,9], and in other infectious disease systems, RNAs [32e34]) of clinical interest. Exosomes are dynamic and can be harvested from a wide-variety of biofluids [2,3,6]. Their persistence throughout infection [7,10,12,14,27] and increase during inflammatory processes [15] may be exploited for their utilization as vaccine delivery vehicles. Similarly, the potential of exosome-based biomarkers for the diagnosis of active tuberculosis has only just begun. Additional studies are required to address the predictive value of these biomarkers in diagnosing tuberculosis and predicting disease outcome in LTBI individuals. Funding:

None.

Competing interests: Ethical approval:

None declared. Not required.

References [1] Harding C, Heuser J, Stahl P. Endocytosis and intracellular processing of transferrin and colloidal gold-transferrin in rat reticulocytes: demonstration of a pathway for receptor shedding. Eur J Cell Biol 1984;35(2):256e63. [2] Johnstone RM, Adam M, Hammond JR, Orr L, Turbide C. Vesicle formation during reticulocyte maturation. Association of plasma membrane activities with released vesicles (exosomes). J Biol Chem 1987;262(19):9412e20. [3] Simpson RJ, Mathivanan Suresh. Extracellular microvesicles: the need for internationally recognised nomenclature and stringent purification criteria. J Proteomics Bioinform 2012;5. [4] Mathivanan S, Fahner CJ, Reid GE, Simpson RJ. ExoCarta 2012: database of exosomal proteins, RNA and lipids. Nucleic Acids Res 2012;40(Database issue):D1241e4. [5] Simpson RJ, Kalra H, Mathivanan S. ExoCarta as a resource for exosomal research. J Extracell Vesicles 2012;1. [6] Kalra H, Simpson RJ, Ji H, Aikawa E, Altevogt P, Askenase P, Bond VC, Borras FE, Breakefield X, Budnik V, Buzas E, Camussi G, Clayton A, Cocucci E, FalconPerez JM, Gabrielsson S, Gho YS, Gupta D, Harsha HC, Hendrix A, Hill AF, Inal JM, Jenster G, Kramer-Albers EM, Lim SK, Llorente A, Lotvall J, Marcilla A, Mincheva-Nilsson L, Nazarenko I, Nieuwland R, Nolte-'t Hoen EN, Pandey A, Patel T, Piper MG, Pluchino S, Prasad TS, Rajendran L, Raposo G, Record M, Reid GE, Sanchez-Madrid F, Schiffelers RM, Siljander P, Stensballe A, Stoorvogel W, Taylor D, Thery C, Valadi H, van Balkom BW, Vazquez J, Vidal M, Wauben MH, Yanez-Mo M, Zoeller M, Mathivanan S. Vesiclepedia: a compendium for extracellular vesicles with continuous community annotation. PLoS Biol 2012;10(12):e1001450. [7] Xu S, Cooper A, Sturgill-Koszycki S, van Heyningen T, Chatterjee D, Orme I, Allen P, Russell DG. Intracellular trafficking in Mycobacterium tuberculosis and Mycobacterium avium-infected macrophages. J Immunol 1994;153(6): 2568e78. [8] Beatty WL, Rhoades ER, Ullrich HJ, Chatterjee D, Heuser JE, Russell DG. Trafficking and release of mycobacterial lipids from infected macrophages. Traffic 2000;1(3):235e47. [9] Bhatnagar S, Schorey JS. Exosomes released from infected macrophages contain Mycobacterium avium glycopeptidolipids and are proinflammatory. J Biol Chem 2007;282(35):25779e89. [10] Bhatnagar S, Shinagawa K, Castellino FJ, Schorey JS. Exosomes released from macrophages infected with intracellular pathogens stimulate a proinflammatory response in vitro and in vivo. Blood 2007;110(9):3234e44. [11] Giri PK, Schorey JS. Exosomes derived from M. Bovis BCG infected macrophages activate antigen-specific CD4þ and CD8þ T cells in vitro and in vivo. PLoS One 2008;3(6):e2461.

[12] Ramachandra L, Qu Y, Wang Y, Lewis CJ, Cobb BA, Takatsu K, Boom WH, Dubyak GR, Harding CV. Mycobacterium tuberculosis synergizes with ATP to induce release of microvesicles and exosomes containing major histocompatibility complex class II molecules capable of antigen presentation. Infect Immun 2010;78(12):5116e25. [13] Giri PK, Kruh NA, Dobos KM, Schorey JS. Proteomic analysis identifies highly antigenic proteins in exosomes from M. tuberculosis-infected and culture filtrate protein-treated macrophages. Proteomics 2010;10(17):3190e202. [14] Kruh-Garcia NA, S. J, Dobos KM. Exosomes: new tuberculosis biomarkers e prospects from the bench to the clinic. In: Cardona P-J, editor. Understanding tuberculosis e global experiences and innovative approaches to the diagnosis. InTech; 2012. [15] Anand PK, Anand E, Bleck CK, Anes E, Griffiths G. Exosomal Hsp70 induces a pro-inflammatory response to foreign particles including mycobacteria. PLoS One 2010;5(4):e10136. [16] Singh PP, LeMaire C, Tan JC, Zeng E, Schorey JS. Exosomes released from M. tuberculosis infected cells can suppress IFN-gamma mediated activation of naive macrophages. PLoS One 2011;6(4):e18564. [17] Pieters J. Mycobacterium tuberculosis and the macrophage: maintaining a balance. Cell Host Microbe 2008;3(6):399e407. [18] Morse MA, Garst J, Osada T, Khan S, Hobeika A, Clay TM, Valente N, Shreeniwas R, Sutton MA, Delcayre A, Hsu DH, Le Pecq JB, Lyerly HK. A phase I study of dexosome immunotherapy in patients with advanced non-small cell lung cancer. J Transl Med 2005;3(1):9. [19] Escudier B, Dorval T, Chaput N, Andre F, Caby MP, Novault S, Flament C, Leboulaire C, Borg C, Amigorena S, Boccaccio C, Bonnerot C, Dhellin O, Movassagh M, Piperno S, Robert C, Serra V, Valente N, Le Pecq JB, Spatz A, Lantz O, Tursz T, Angevin E, Zitvogel L. Vaccination of metastatic melanoma patients with autologous dendritic cell (DC) derived-exosomes: results of thefirst phase I clinical trial. J Transl Med 2005;3(1):10. [20] Cheng Y, Schorey JS. Exosomes carrying mycobacterial antigens can protect mice against Mycobacterium tuberculosis infection. Eur J Immunol 2013;43(12):3279e90. [21] Malen H, Berven FS, Fladmark KE, Wiker HG. Comprehensive analysis of exported proteins from Mycobacterium tuberculosis H37Rv. Proteomics 2007;7(10):1702e18. [22] Rosenkrands I, Weldingh K, Jacobsen S, Hansen CV, Florio W, Gianetri I, Andersen P. Mapping and identification of Mycobacterium tuberculosis proteins by two-dimensional gel electrophoresis, microsequencing and immunodetection. Electrophoresis 2000;21(5):935e48. [23] Prados-Rosales R, Baena A, Martinez LR, Luque-Garcia J, Kalscheuer R, Veeraraghavan U, Camara C, Nosanchuk JD, Besra GS, Chen B, Jimenez J, Glatman-Freedman A, Jacobs Jr WR, Porcelli SA, Casadevall A. Mycobacteria release active membrane vesicles that modulate immune responses in a TLR2dependent manner in mice. J Clin Invest 2011;121(4):1471e83. [24] Prados-Rosales R, Weinrick BC, Pique DG, Jacobs Jr WR, Casadevall A, Rodriguez GM. Role for Mycobacterium tuberculosis membrane vesicles in iron acquisition. J Bacteriol 2014;196(6):1250e6. [25] Singh PP, Smith VL, Karakousis PC, Schorey JS. Exosomes isolated from mycobacteria-infected mice or cultured macrophages can recruit and activate immune cells in vitro and in vivo. J Immunol 2012;189(2):777e85. [26] Thery C, Amigorena S, Raposo G, Clayton A. Isolation and characterization of exosomes from cell culture supernatants and biological fluids. Curr Protoc Cell Biol 2006;3:22e9 [Chapter 3]. [27] Kruh-Garcia NA, W.L., Chaisson LH, Worodria WO, Nahid SJP, Davis JL, Dobos KM. Detection of Mycobacterium tuberculosis peptides in the exosomes of patients with active and latent M. tuberculosis infection using MRMMS. PLoS ONE 2014;9:e103811. [28] Izzo A, Brandt L, Lasco T, Kipnis AP, Orme I. NIH pre-clinical screening program: overview and current status. Tuberculosis (Edinb) 2005;85(1e2):25e8. [29] Caradec J, Kharmate G, Hosseini-Beheshti E, Adomat H, Gleave M, Guns E. Reproducibility and efficiency of serum-derived exosome extraction methods. Clin Biochem 2014;47(13e14):1286e92. [30] Mathivanan S, Simpson RJ. ExoCarta: a compendium of exosomal proteins and RNA. Proteomics 2009;9(21):4997e5000. [31] Tauro BJ, Greening DW, Mathias RA, Ji H, Mathivanan S, Scott AM, Simpson RJ. Comparison of ultracentrifugation, density gradient separation, and immunoaffinity capture methods for isolating human colon cancer cell line LIM1863-derived exosomes. Methods 2012;56(2):293e304. [32] Wu SC, Yang JC, Rau CS, Chen YC, Lu TH, Lin MW, Tzeng SL, Wu YC, Wu CJ, Hsieh CH. Profiling circulating microRNA expression in experimental sepsis using cecal ligation and puncture. PLoS One 2013;8(10):e77936. [33] Canitano A, Venturi G, Borghi M, Ammendolia MG, Fais S. Exosomes released in vitro from Epstein-Barr virus (EBV)-infected cells contain EBV-encoded latent phase mRNAs. Cancer Lett 2013;337(2):193e9. [34] Narayanan A, Iordanskiy S, Das R, Van Duyne R, Santos S, Jaworski E, Guendel I, Sampey G, Dalby E, Iglesias-Ussel M, Popratiloff A, Hakami R, KehnHall K, Young M, Subra C, Gilbert C, Bailey C, Romerio F, Kashanchi F. Exosomes derived from HIV-1-infected cells contain trans-activation response element RNA. J Biol Chem 2013;288(27):20014e33.