3. What Is the Evidence for Polar Lipid Coatings of Nanobubbles in Xylem Sap?
The discovery that nanobubbles in xylem sap are coated by surfactant shells (Figure 2
) was not a surprise, because uncoated nanobubbles would either shrink and dissolve or expand in response to the constant pressure and temperature changes inside a plant. However, this raised the further question of the chemical nature of these shells. Initial analyses of xylem sap residue found evidence both for polar lipids and proteins 
. The former self-assemble in monolayers at water–gas interfaces, and the latter can be surface-active as well depending on their composition 
. Polar lipids extracted from xylem sap of thirteen flowering plant species and analyzed via mass spectrometry 
included both phospholipids and galactolipids, with a high proportion of phosphatidic acid (PA), a phospholipid with a negatively charged headgroup. Other charged phospholipids in xylem include negatively charged phosphatidylglycerol (PG), phosphatidylinositol (PI), and phosphatidylserine (PS), as well as zwitterionic phosphatidylcholine (PC) and phosphatidylethanolamine (PE). The amount of neutral galactolipids varied widely between species, making up between 7 and 75% of all polar lipids in xylem sap, and about 30% on average (Figure 3
Concentrations and chemical composition of polar lipids in xylem sap in twelve flowering plant species, including data from (a
) California 
, and (b
) Germany 
. DGDG, digalactosyldiacylglycerol; MGDG, monogalactosyldiacylglycerol; PA, phosphatidic acid; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PI, phosphatidylinositol; PS, phosphatidylserine. Figure modified from 
. See Figure 1
for the full species names.
Polar lipids in xylem of flowering plants do not appear to change in concentration or composition seasonally 
or over the lifetime of a plant. Comparison with the polar lipid composition of living xylem cells showed that the lipid composition is essentially the same in sap and living cells 
, suggesting that the lipids in dead conduits remain there after the rest of the living cell components have been enzymatically removed during the maturation of the conduit from a living cell into a dead hollow conduit.
It is not possible to separate polar lipids that coat nanobubbles in xylem sap from other polar lipids that coat other surfaces or form micelles in xylem; however, because polar lipids self-assemble at gas–water interfaces, it is clear that any polar lipids found in sap can be part of surfactant shells on gas nanobubbles. A significant correlation was found between the concentrations of nanoparticles and the concentrations of polar lipids in sap for three out of five species studied 
, further suggesting that surfactant shells on nanobubbles are made up largely of polar lipids. Proteins could contribute to these surfactant shells as well, including lipid transfer proteins that have surface-active properties and which are commonly found in xylem sap 
. More research is needed to determine which proteins can be found in surfactant shells of xylem nanobubbles.
4. How Do Nanobubbles Form in Xylem?
Nanobubbles in xylem sap can only exist under the widely fluctuating pressure and temperature conditions experienced by plants if they are covered by surfactant shells. How do they form, though? As functioning xylem cannot be imaged at the nanoscale with currently available technology, there are no direct observations of nanobubbles forming in plants. There are two possibilities: (1) nanobubbles could form inside lipid micelles or lipid bilayers by diffusion of gas from sap through the lipids into the center of the micelles, which are occupied by the hydrophobic acyl chains of polar lipids and as such are largely hydrophobic and free of water 
; or (2) gas could move into xylem sap via mass flow from other xylem compartments that contain a gas phase 
Option (1), diffusion of gas into the center of lipid micelles or bilayers, would only create a bubble if the resulting gas pressure counteracts the molecular forces that pull the lipids together, including hydrogen bonds between hydrophilic headgroups and relatively weak dispersion forces between the hydrophobic tails. Negative pressure in the surrounding sap could provide the necessary pull. However, molecular dynamics situations have shown that at the spatial and time scale applicable to plants, i.e., the volumes of conduits and the duration of negative pressure conditions, it would take on the order of −10 MPa of negative pressure in the sap to pull micelles or bilayers apart and create a void that could be occupied by gas 
. Very few plant species ever reach −10 MPa of pressure in their xylem, and it could be that the instability of lipid bilayers and micelles under those conditions is the reason why plants cannot operate at more negative pressures 
. When lipid bilayers are pulled apart under negative pressure the process is very unlikely to result in nanobubbles, because the energy released during this cavitation event would favor creation of a large cavitation void and thereby create a xylem embolism, i.e., fill the whole conduit with gas, rather than create a stable nanobubble. While it is unlikely that cavitation in lipid bilayers or micelles occurs within the normal pressure range in the xylem of most plants, which ranges between 0 and −5 MPa 
, it is important to realize that the probability of such events increases with the volume of xylem and with the duration and strength of negative pressure 
. Therefore, xylem embolisms could occasionally form through this process.
Option (2), nanobubble formation from a preexisting gas phase such as an embolized conduit, is far more likely than option (1). Where an embolized (i.e., gas-filled) xylem conduit borders on another one, the gas and liquid phases are separated by mesoporous pit membranes. The discussion here addresses flowering plants, not conifers, which have a very different pit membrane morphology 
. In flowering plants, these are fibrous membranes ranging in thickness from less than 200 to more than 900 nm 
, largely consisting of aggregated cellulose microfibrils that are between 20 and 30 nm thick 
. Pores through such fibrous membranes have numerous pore constrictions or pore throats, with thicker membranes having a longer pore pathway and more pore constrictions 
. The largest constriction size tends to be about 20 nm in diameter 
. If nanobubbles form in pit membranes between gas-filled and sap-filled conduits, than they would have to form by gas overcoming the surface tension and moving through these constrictions (Figure 4
Diagram of nanobubbles forming in a pit membrane of a bordered pit between adjacent conduits. Lipid-coated nanobubbles are formed by a snap-off event due to the presence of surface-active lipids and pore constrictions between cellulose microfibrils. In addition to surfactant-coated nanobubbles, this image shows a wide range of nanoparticle sizes, which can be bilayer to multilayer micelles and vesicles associated with pit membranes. Reprinted with permission from Ref. 
According to the Young–Laplace equation, the pressure difference forcing a bubble through a 20 nm pore, assuming a contact angle of zero 
and a pore shape correction factor of 0.5 to account for the fact that pores in fibrous membranes are not cylindrical 
, would be 7.2 MPa, assuming surface tension of pure water of 72 mJ m−2
. Pit membranes are coated with polar lipids 
, which reduce the surface tension to about a third of pure water if the polar lipids are in equilibrium 
. At a surface tension of 24 mJ m−2
, a meniscus could pass through a 20 nm pore with a shape correction factor of 0.5 under a pressure difference between the gas and liquid phase of 2.4 MPa 
. If pores are slightly larger than 20 nm or surface tension is slightly lower, then the pressure difference required for gas movement through a pore constriction would be even lower. Such pressure differences are well within the range of pressures experienced in the xylem of many plant species 
Movement of a meniscus through a pore constriction would not result in a continuous stream of gas from the gas into the liquid phase, instead resulting in snap-off of a nanobubble inside the membrane if the radius of the constriction were less than half the radius of the pore behind it 
, which would be the case in fibrous membranes. In geometrically complex pore spaces, the invasion of a non-wetting fluid, such as gas invading a pit membrane wetted with water, occurs not as a simple wetting front, but in rapid snap-off events and so-called Haines jumps 
. Second, the entry of gas into the liquid phase increases the local liquid pressure, which causes bubble snap-off at the pore constriction. Due to the low compressibility of liquid water, the pressure release caused by bubble entry into water under negative pressure is substantial; a bubble that compresses water volume in the confined space within a pit border (around 0.5 to 10 μm3 
) by only 0.1% releases about 2 MPa of negative pressure 
. Third, in liquid that is under negative pressure, minimization of the surface area created by formation of a single small bubble is always thermodynamically favored over rupture of hydrogen bonds between water molecules at the gas–water interface, which requires far more tensile energy density 
. Therefore, movement of gas through nano-sized pore constrictions under negative pressure initially would produce nanobubbles, not a continuous stream of gas or a cavitation void. Thus, the multiphase interactions between gas, xylem sap, mesoporous pit membranes, and surfactants lead to surfactant-coated nanobubbles.
The pores in pit membranes play another important role for the persistence of nanobubbles in xylem conduits, as they are too small to allow passage of lipid-coated nanoparticles or most lipid micelles (see Figure 1
). Therefore, nanobubbles do not move with the sap from one conduit to another or into living cells, and do not accumulate in leaves, instead remaining locked inside individual xylem conduits 
The scenario described here for nanobubble creation in pore constrictions depends on the presence of polar lipids inside the pore, lowering the surface tension. Without polar lipids, it would take a far higher pressure difference to force a meniscus through. If polar lipids are present in the pores, then they would invariably coat the nanobubbles because of the strong surface activity of polar lipids. Considering the normal pressures in xylem, the sizes of pore constrictions in pit membranes, and the presence of polar lipids in pit membranes, it appears inevitable that lipid-coated nanobubbles will form in pit membranes that separate gas-filled and sap-filled xylem conduits. The question then arises as to how such bubbles can possibly be stable under the physical conditions that exist in xylem.