A plasma membrane divides the field of view. On one side is a dancing portion of eukaryotic cell, which lives, splits, floats and dies in the comparatively void fluid on the other. Magnified through the electron microscope, however, it’s unclear which half is enclosed and which is excluded.

As if to address that, the membrane begins to warp, flexing into the living cell like some ghostly little finger is pushing to pop it. A closer view would show that tiny particles from the surrounding fluid are pressing in, but the cell is actually in control of its own deformation. The membrane is invaginating the external material: turning inside-out to enwrap it – folding to form an interior sac that, expanding inwards, pulls together the two pivots from which it seems to open.

When those two ‘pseudopodia’ finally meet, an island of extracellular material buds from the membrane into the cell. Thus do the constituent parts consume, communicate, quarantine and destroy. Within the cell, an endosome will fuse with this membrane-wrapped vesicle to begin the process of breaking down or using its contents – protein, enzyme, bacterium or otherwise.

Then things just have to be balanced out. In order to maintain an equal flow of molecules in and out of the cell, some ‘late’ endosomes, themselves invaginated with numerous vesicular compartments, will reattach to the cell membrane and release their vesicles by an inverse process. The free-floating, plasma-bound sacs so ejected are called exosomes.

It’s hard to dispute that they have been underappreciated. Upon their discovery, the membranous outer bodies prompted more scientific shrugs than they did celebrations. As poetic as early appellation ‘platelet dust’ might sound, in practice exosomes were dismissed as the trash bags of their famous, discipline-defining parents, the cells. It was years before they were dignified with a name of their own. When they did eventually become exosomes, so did a host of completely different extracellular microvesicles (EVs). They were all still trash, too.

In good news for the nanoparticles’ self-esteem, not to mention humanity’s clinical capabilities, as the understanding of endocytosis and exocytosis has developed over the past 10–20 years, that hypothesis has been thoroughly disproved. As distinguished from other EVs, exosomes are now thought to provide an important means for cells to communicate with and influence each other, transmitting macromolecules like DNA and RNA over long distances. Numerous researchers have even proposed using them as more tolerable drug vectors than synthetic materials.

As fruitful as that might be, the most important advance for medical device developers was probably the discovery that diseased cells are already releasing exosomes for their own ends. Mirroring the biological features of their parent cells, these vesicles are an important biomarker for the early diagnosis of conditions like cancer, which is particularly aggressive in spreading them.

As a result, the otherwise outmoded conception of exosomes as waste disposal has taken on an instructive new valence. Long before a cancer presents for imaging equipment or tissue biopsies, by which time treatment options may already be drastically reduced, its exosomes are polluting the bloodstream. As Professor Yong Zeng of Kansas University – who recently published a proof-of-concept paper in Nature for a new lab-on-a-chip to detect them – sees it, exosomes still bear some of the hallmarks of trash. “By analysing all those things people think they have no use for, you can identify the pattern of how one person lives, and what they live on,” he says.

Thankfully, rather than facilitating identity theft, the data Zeng is interested in collecting forms a foundation for personalising and precision-engineering medicine. Liquid biopsy technologies, like his promise to catch and specifically identify cancers in the earliest stages without requiring cumbersome equipment or invasive procedures, make it more treatable than ever, wherever it might present.

of cancer deaths are caused by metastases.

“Each vesicle might not have all the information from the parent cells,” Zeng continues, “but with the opportunity and the technology to interrogate different ones at different times and in different situations, you can develop an overall understanding of the tumour by piecing them together.”

Generation waste

People create waste. One proposed name for the current geological epoch – the Anthropocene – even recognises the fact, but we’re burlesquing much more fundamental processes. Long before society started pumping out greenhouse gases and chemical sludge, tumour-derived exosomes (TDEs) were shaping internal environments. This isn’t a by-product of their primary function, but the function itself. By modulating the gene expression of healthy recipient cells, TDEs suppress immune responses, initiate tumour growth and promote therapy resistance. In the vicinity of a tumour, exosomes bring about angiogenesis; but others can travel through the bloodstream, invading tissue elsewhere in the body to form pre-metastatic niches that support the migration of circulating tumour cells, which can also be targeted by liquid biopsies.

“A lot of this is still poorly understood,” says Zeng, “but we’re accumulating evidence that shows it is a very effective way for the cells to change their environment. Previously, people thought that circulating tumour cells might be the main route to metastases, but TDEs can actually remodel tissues and cells in remote areas.”

In contrast, tissue biopsies, whether they’re performed by surgery or fine-needle aspiration, can only be carried out once the tumour has been located and identified, which becomes possible specifically because of the action of TDEs.

“When tumours grow, they have to develop their vascular systems, they need nutrients and oxygen to expand and outgrow the normal cells,” says Zeng. “That’s why they dump those materials, including exosomes, into circulation. The blood carries them, and they carry the information we need to develop a more comprehensive overview of the disease and its pathological conditions.

“It’s not limited to the liver; it’s not limited to the stomach or the brain – it circulates, so I think it’s more likely to spell out something unusual that happens in the body.”

There’s just one slight problem. Cancer cells may spread their genetic material more aggressively than their healthy counterparts, but there are still approximately 37 trillion other cells constantly releasing nanoscale vesicles in every human, and exosomes are the smallest of them all, measuring 30–150 billionths of a metre. As numerous recent research papers understate, the task resembles finding “small needles in large haystacks”. Before now, there have been few, if any, liquid biopsies sensitive enough to detect TDEs in typical blood samples.

The biggest challenge comes from the differences between microfluidics and its macro-scale equivalent, even when they both involve blood. Circulation is a type of turbulent flow full of eddying swirls and fluctuations that continually mix and scramble fluid particles. The choppy velocity of the bloodstream will actually destroy tumour cells that don’t complete their migration into other organs quickly enough, but, by dashing them repeatedly against the surface of blood vessels, it also gives them plenty of opportunities to escape. By contrast, the particles in blood samples interact with microfluidic LOCs and liquid biopsy devices in slow, orderly and unmixed parallel lines. This laminar flow profile means that, unless the relevant biomarkers happen to end up in the skein of fluid directly in contact with the diagnostic device, they won’t be detected.

The Size of a blood sample from which the Kansas University liquid biopsy can detect tumourderived exosomes.
Kansas University

This has some stark consequences. “When it comes to drug development, some methods have to kill multiple animals in order to collect enough exosomes to run one or two experiments,” says Zeng. “To get the culture cells, they have to use litres of culture mediums, which costs a fortune. You cannot imagine having a child give you 10ml for every test.”

So, Zeng, co-lead Dr Andrew Godwin (deputy director of the Kansas University cancer centre) and their team needed to generate turbulence in their microfluidic channels. As their paper on ‘Ultrasensitive detection of circulating exosomes with a 3D-nanopatterned microfluidic chip’ makes clear, the obvious way to do so was with herringbone. For the past 20 years, the patterns so beloved of country gentleman and the compositors of brick pavements have also been the gold standard for disrupting laminar flow, forcing samples over zig-zagging obstacles to create microvortices. Nevertheless, significant hydrodynamic resistance is exerted by the sliver of liquid that remains in contact with the sensor surface. To overcome this, the Kansas researchers nanoengineered tiny holes in their herringbone to drain away the most stubbornly laminar strings of blood.

In the original press release that accompanied the team’s paper, Zeng compared these nanopores with an array of plugholes. “If you have a sink filled with water and many balls floating on the surface, how do you get all the balls in contact with the bottom of the sink where sensors could analyse them?” he asked. “The easiest way is to drain the water.”

As suspiciously simple as that sounds, it works. When tested with samples from 20 ovarian cancer patients (of particular interest because in 85% of cases it is detected at late stages, and often cannot be cured) and 10 age-matched controls, Zeng’s nano-herringbone liquid biopsy chip was able to detect TDEs from 2μLs of blood, or 1/25th of a drop.

Now, Zeng says, “We don’t have to sacrifice animals for these experiments. We can just draw, say, 15μLs of blood from the tail, and keep a single animal alive for the entire study, which will eliminate variation due to individual heterogeneity. Every animal is different, and every person is different. If you take the average you don’t see the effects sometimes.

“Not only can we reduce the cost for studies, we can actually be more ethical and improve the quality of our findings”.