Skip to main content
BioThera Solutions

EV marker panels explained: CD9, CD63, CD81, TSG101, Alix, and markers for plant, bacterial, fungal and algal vesicles

MISEV2023 asks for a marker panel, not a single marker. What CD9, CD63, CD81, TSG101, and Alix actually tell you about a preparation, why no single one is sufficient, and why the mammalian panel does not carry over to vesicles from plants, bacteria, fungi or algae.

FS

Frédéric St-Denis-Bissonnette, PhD

Founder & CEO, BioThera Solutions

MISEV2023 is most often misquoted on the marker question. The common shorthand, "EVs have CD9, CD63, and CD81 on them," is not what the guideline says. The guideline says that demonstrating a preparation is EV-enriched requires a panel of markers covering several categories, and that the panel choice depends on what is being claimed about the preparation. The single-marker positive result, especially on a Western blot, is not sufficient evidence on its own.

Understanding why is the difference between a defensible characterization record and a marketing-grade one.

What the canonical markers actually are

MISEV2023 organises EV markers into three working categories.

Category 1: Transmembrane EV-associated proteins. The tetraspanins CD9, CD63, and CD81 are the canonical examples. They are enriched in EVs across many biogenesis routes and serve as positive markers that the preparation contains material of EV origin. Different EV subpopulations express the three tetraspanins to different extents, which is also informative.

Category 2: Cytosolic proteins recoverable in EVs. TSG101 and Alix are the canonical examples. Both are associated with the ESCRT machinery involved in multivesicular body biogenesis. Their presence in a preparation supports an exosome (MVB-derived) origin and helps distinguish exosomes from microvesicles, which bud from the plasma membrane and are typically lower in ESCRT-associated proteins.

Category 3: Markers of non-EV contaminants. This is the category most often skipped in commercial characterization. MISEV2023 expects evidence of the absence of contamination, typically high-density lipoproteins (apolipoproteins like APOA1, APOB), serum albumin, and cellular debris markers, depending on the source biomass. A preparation that is enriched for CD9 but also enriched for apolipoproteins is co-purifying lipoproteins along with the vesicles. That is information the buyer needs.

Why no single marker is sufficient

Three reasons, all of them load-bearing.

First, no marker is exclusively expressed by EVs. CD9, CD63, and CD81 are enriched in EVs relative to whole-cell lysates, but they are also present on cellular membranes and in cellular subcompartments. A Western band for CD9 in a preparation tells you the preparation contains CD9-bearing material. It does not, on its own, tell you that material is exosomes versus, for example, plasma-membrane fragments.

Second, no marker is uniformly expressed across all EV subpopulations. Different EV-producing cells release vesicles with different marker profiles. An exosome population enriched for CD63 might be relatively depleted of CD9. A preparation negative for one tetraspanin is not necessarily not-EVs; it might be EVs of a subtype that does not strongly express that marker. This is why MISEV2023 expects a panel.

Third, contamination is invisible to single-marker positive testing. A preparation can be CD9-positive, TSG101-positive, look great on a Western, and still co-purify a substantial fraction of non-vesicular contaminating particles that drive the bulk of the measured biological activity. Without an absence-of-contaminants screen, the positive markers tell only half the story.

Beyond mammalian EVs: plants, bacteria, fungi and algae

CD9, CD63, CD81, TSG101 and Alix were defined in human and other mammalian cells. Every other kingdom that releases vesicles does so with its own membranes and its own machinery, and the mammalian panel does not carry over. MISEV2023 notes that its marker examples come from mammalian systems and that other organisms need markers suited to their own biology. For a buyer, the practical point is simple: a non-mammalian product that reports CD63 or CD81 positivity, detected with antibodies raised against the human proteins, deserves a follow-up question about what that antibody is actually binding.

Plants

Plants have tetraspanins of their own, but they are not CD9, CD63 or CD81. In Arabidopsis, the tetraspanins TET8 and TET9 mark one population of extracellular vesicles and the syntaxin PEN1 marks another, which is itself evidence that plants release more than one vesicle type. These markers come from model-plant research, and antibodies for them are not available for most crop or medicinal species. Proteins that recur across plant EV proteomic surveys, such as heat shock protein 70, plasma membrane aquaporins and H+-ATPases, help confirm identity but are not specific to vesicles.

Many commercial plant preparations are made by homogenising or juicing tissue rather than collecting what the plant secretes. That material is better described as plant-derived nanovesicles (PDNVs), a mix of secreted vesicles and vesicles formed from membrane fragments. For it, the most useful contamination checks target the chloroplast, protein-pigment complexes and cell-wall material that travel with the vesicles through some isolation workflows.

Bacteria

Bacteria release extracellular vesicles, and the field studies them as such; the term "bacterial exosomes" also appears in the literature and on products. Bacteria have no endosomal system, though, so in the strict MISEV2023 sense these vesicles are not exosomes, and the ESCRT and tetraspanin markers do not apply. Gram-negative bacteria release outer membrane vesicles (OMVs) carrying outer membrane proteins such as OmpA, along with lipopolysaccharide (LPS). Gram-positive bacteria release cytoplasmic membrane vesicles that can carry lipoteichoic acid, and cell lysis produces further vesicle types that contain cytoplasmic content. Flagella, pili and protein aggregates co-purify readily and are the usual contaminants. Because LPS is an endotoxin, characterizing bacterial vesicles intended for skin or human use leads directly into safety testing.

Fungi

Fungi do use ESCRT machinery to make vesicles, yet in Candida albicans the yeast counterparts of TSG101 and Alix are not found as vesicle cargo, and tetraspanin homologues have not been identified in many fungi, including baker's yeast. Proteomic work in C. albicans proposed fungal-specific markers instead, notably the claudin-like Sur7 family proteins Sur7 and Evp1. Cell-wall fragments are the main contaminant to watch, since fungal vesicles cross a thick polysaccharide wall on their way out of the cell.

Algae

Microalgae are the newest source. Early studies on microalgal vesicles, sometimes called nanoalgosomes, reported positive immunoblots for conserved proteins such as Alix, enolase and HSP70. No algae-specific panel has been agreed, so proteomics and orthogonal physical characterization carry more of the weight. Pigment complexes and extracellular polysaccharides from the culture are the contaminants to rule out.

The common thread across all four: the panel should be justified by the biology of the source organism, it should include a contamination check matched to what co-purifies from that source, and the detection method should be disclosed. Borrowing the mammalian panel is a shortcut, not a characterization.

For our Aloe barbadensis material, identity is confirmed by mass spectrometry (proteomics), particle counts by NTA are paired with TEM imaging of vesicle morphology, and the particle-to-protein ratio is reported on every batch as a purity indicator. The combination, not any one of these elements, is what supports the vesicle-enrichment claim.

Where the marker panel belongs

MISEV2023 is a reporting guideline for research, not a product release specification. In manufacturing, the marker panel does its work when a process is characterized: it shows that the process yields vesicle-enriched material from that source. Each batch Certificate of Analysis then reports the release tests showing that the batch matches the characterized process. A defensible marker characterization record includes:

  1. At least one membrane-associated marker appropriate to the source organism, reported with a clear result on the stated detection method.
  2. A cytosolic marker where it is relevant to the claim, for example TSG101 or Alix when a mammalian exosome (MVB) origin is being claimed.
  3. An absence-of-contaminants screen targeting the contaminants most likely to co-purify given the source biomass and isolation method.
  4. Source-appropriate substitutions for plant, bacterial, fungal or algal preparations, using markers validated for that organism rather than mammalian tetraspanins alone.
  5. Method-of-detection disclosure (Western blot, ELISA, flow cytometry, mass spectrometry) with sufficient information that a third party can interpret the result.

A record that reports "CD9 positive" with no contamination screen, no second marker, and no detection method is not characterizing the preparation. It is asserting a single fact that, on its own, does not support the conclusions a marketing page draws from it.

What clinicians and partners can ask

Three questions on the marker panel, paralleling the ones on isolation and counting:

  1. Which markers were used to characterize the material or its process, and which MISEV2023 category does each one address?
  2. Is a contamination check included, and what does it target?
  3. For plant, bacterial, fungal or algal preparations, were the markers chosen for that organism, or were mammalian markers used as a stand-in?

The answers, taken together with the isolation method and the particle count, are what a defensible EV product specification consists of. Anything less is a partial picture, and MISEV2023 is explicit about not wanting to settle for that.

Want more EV science in your Google results?

Add BioThera Solutions as a preferred source on Google

Primary sources

  1. Welsh JA, Goberdhan DCI, O'Driscoll L, Buzas EI, et al. (MISEV Consortium, incl. St-Denis-Bissonnette F). Minimal information for studies of extracellular vesicles (MISEV2023): from basic to advanced approaches. Journal of Extracellular Vesicles, 2024. doi.org/10.1002/jev2.12404
  2. Kowal J, Arras G, Colombo M, et al.. Proteomic comparison defines novel markers to characterize heterogeneous populations of extracellular vesicle subtypes. PNAS, 2016. doi.org/10.1073/pnas.1521230113
  3. Jeppesen DK, Fenix AM, Franklin JL, et al.. Reassessment of exosome composition. Cell, 2019. doi.org/10.1016/j.cell.2019.02.029
  4. Pinedo M, de la Canal L, de Marcos Lousa C.. A call for Rigor and standardization in plant extracellular vesicle research. Journal of Extracellular Vesicles, 2021. doi.org/10.1002/jev2.12048
  5. Rutter BD, Innes RW.. Extracellular vesicles isolated from the leaf apoplast carry stress-response proteins. Plant Physiology, 2017. doi.org/10.1104/pp.16.01253
  6. Cai Q, Qiao L, Wang M, et al.. Plants send small RNAs in extracellular vesicles to fungal pathogen to silence virulence genes. Science, 2018. doi.org/10.1126/science.aar4142
  7. Toyofuku M, Nomura N, Eberl L.. Types and origins of bacterial membrane vesicles. Nature Reviews Microbiology, 2019. doi.org/10.1038/s41579-018-0112-2
  8. Dawson CS, Garcia-Ceron D, Rajapaksha H, et al.. Protein markers for Candida albicans EVs include claudin-like Sur7 family proteins. Journal of Extracellular Vesicles, 2020. doi.org/10.1080/20013078.2020.1750810
  9. Picciotto S, Barone ME, Fierli D, et al.. Isolation of extracellular vesicles from microalgae: towards the production of sustainable and natural nanocarriers of bioactive compounds. Biomaterials Science, 2021. doi.org/10.1039/d0bm01696a

Questions on this topic?

Clinicians, researchers, and partners are welcome to go deeper. We treat scientific dialogue as a working part of the platform, not a sales channel.

BioThera Solutions

A fully integrated extracellular vesicle company, from research and intellectual property to manufacturing, with a process designed to scale. Extracellular vesicles (EVs/exosomes) are among biology's most promising delivery systems, and we are building the infrastructure the field has been missing.

Ottawa, Ontario, Canada
477-150 Louis-Pasteur Private Ottawa, Ontario K1N 9B4 Canada
Add BioThera Solutions as a preferred source on Google

© 2026 BioThera Solutions Inc. All rights reserved.

For informational purposes only. Nothing on this website is medical advice. BioThera develops and manufactures characterized extracellular vesicles (exosomes) and, for cosmetic applications, licenses its technology exclusively to Vesera™, a skincare company and BioThera subsidiary. Vesera™ products are cosmetics and are not intended to diagnose, treat, cure, or prevent any disease. BioThera does not sell products to consumers, and its work on therapeutic applications is research only.