Water drives migration of a biological nomad

Naïve T cells lead a nomadic life, migrating between lymphoid organs in their search for foreign antigens, the first essential step of the adaptive immune response. From the bloodstream, T cells enter lymph nodes through specialised high endothelial venules (HEVs) following a signal from the chemokine CCL21 to the T cell CCR7 receptor. They use integrin-mediated adhesion to stick to the HEV endothelium, and then squeeze through into the lymph node parenchyma. Once inside, the hunt for antigen begins: the T cells migrate rapidly in response to CCL19 and CCL21 signalling, predominantly via an integrin-independent mechanism, and scan antigen-presenting cells for the presence of cognate antigen.

A T cell migrating in response to chemokine is polarised, tugged along by formation of a pseudopodium at the leading edge. The pseudopodium is stuffed with a highly dynamic network of actin filaments, which are stimulated to polymerise by PI3K activation of the RAC1, RAC2 and CDC42 GTPases. This actin polymerisation provides the motor that drives the plasma membrane forward, moving the cell.

Victor Tybulewicz’s lab came into the field of T cell migration somewhat by accident. After initiating a genetic screen in 2012 for kinases involved in T cell adhesion, they found that one of their hits, WNK1, was a negative regulator of integrin-mediated T-cell adhesion, but also positively regulated migration via a separate pathway involving the kinases OXSR1 and STK39 activating the ion co-transporter SLC12A2 [1]. Loss of WNK1 or inhibition of SLC12A2 in T cells caused them to home less efficiently to lymphoid organs and migrate more slowly once there. 

WNK1’s presence as a central player in T cell motility was intriguing: it’s a member of a family of serine/threonine protein kinases that regulate ion homeostasis in kidney epithelial cells, promoting net influx of Na+, K+ and Cl- ions from urine. If WNK1 had this same function in T cells, how might ion flux be able to affect migration? Finding the definitive answer to this question has caused a radical rethinking of how T cell motility, and potentially the motility of many other cell types, can be viewed [2].

To extend the original results from their 2016 paper, the Tybulewicz lab first used two standard assays (migration towards CCL21 under agarose on dishes coated with the integrin ligand ICAM1, or in a 3D collagen matrix) to mimic CCL21-driven migration of T cells in a lymphoid organ. If they stopped naïve CD4+ T cells from switching on SLC12A2 in response to WNK1 signalling, they saw that however this was effected—by loss or inhibition of WNK1, combined mutation of OXSR1 and STK39, or inhibition of SLC12A2 itself—the cells could not migrate efficiently. When activated T cells were put into the same assays, they too required the same WNK1 pathway for chemokine-induced migration.

Directly demonstrating the effects of ion movement on T cell migration was tricky. Victor enlisted the help of collaborators in the Department of Earth Science in Cambridge and the London Metallomics Facility at King’s, who used inductively-coupled plasma mass spectrometry (ICP-MS) to measure the change in Na+, K+ and Cl- ion concentration in cell lysates after CCL21 stimulation. Even with their expertise, the background noise of Na+ and Cl- ions in the extracellular medium was too much to allow accurate measurement, but it was clear that CCL21 treatment resulted in an influx of K+ ions into the T cells, and that this was blocked by treatment with the WNK1 inhibitor WNKi. The importance of Na+ and Cl- ions was demonstrated in a different way: if Na+ and Cl- ions were depleted from the extracellular medium, but isotonicity was maintained by substituting in neutral ‘stuffer molecules’, speed of migration was greatly decreased. Taken together, these data showed that CCL21 stimulation results in WNK1-dependent K+ ion entry into T cells, and that Na+ and Cl- ions are also required for migration.

Ion influx implies water uptake by osmosis, and hence cell swelling. The lab looked at cell size by dusting off an old-school Coulter-like counter to show that CCL21 signalling caused an increase in T cell volume, and that this was inhibited if WNK1 and SLC12A2 were absent or switched off. There was an extra wrinkle to this experiment: WNK1-deficient or inhibited cells were also about 15% smaller than controls, as were cells in which Oxsr1 and Stk39 were mutated, suggesting that the WNK1 pathway is required for regulatory volume increase, a mechanism by which cells recover after exposure to a hypertonic environment. SLC12A2 played no part in this homeostatic regulation. The volume change upon CCL21 stimulation was confirmed by imaging T cells migrating in response to CCL21 in the presence and absence of WNKi, and that this was due to water influx was demonstrated by NMR analysis showing uptake of deuterated water from spiked medium. 

Water enters cells through aquaporin channels. Of the 11-member aquaporin family, CD4+ T cells predominantly express Aqp3, although Aqp9 and Aqp11 are present at lower levels. There is a selective AQP3 inhibitor, DFP00173 (hereafter AQP3i), and the lab showed that AQP3i significantly slowed CCL21-induced migration of T cells and also reduced CCL21-induced polarisation. Just like the SLC12A2 inhibitor, AQP3i had no effect on the basal volume of the T cells but eliminated the CCL21-induced volume increase. Therefore, AQP3 is required for CCL21-induced migration of T cells, likely as a result of water transport. This puts another link in the chain from WNK1: the chemokine CCL21 switches on the WNK1 pathway, which causes ion influx, which in turn causes regulated water influx. Water influx is required for polarisation and migration, a startling result.

Migration of T cells requires actin polymerisation at the leading edge, resulting in F-actin filaments pushing forwards against the plasma membrane. This process turns out to be what WNK1-driven ion and water influx is regulating, which Victor and his colleagues determined by a series of elegant experiments.

They began by checking that the pathway was operative at the right time and place, by visualising the subcellular localisation and activation status of the key protein players. GFP tagging of WNK1, OXSR1 and SLC12A2 in T cells migrating in response to CCL21 showed that all three proteins were accumulating at the leading edge and depleted at the trailing edge, and that their activation, as measured by phosphorylation, was also enriched at the leading edge, making it likely that ion and water influx were concentrated there too. Further, CCL21 stimulation could induce an increase in total F-actin, which was partially decreased by inhibition of either WNK1 and SLC12A2, or mutation of Oxsr1 and Stk39.

What might the role of this localised ion flux and water entry be? Forward protrusion of F-actin filaments is driven by addition of G-actin monomers to the filament tip, and this had previously been suggested to happen by a Brownian ratchet mechanism: thermally generated Brownian motion would cause random fluctuations of both the cell membrane and individual actin filaments, and if a gap opened up between the two that was greater than 2.7nm, a G-actin monomer could slip in and lock into place, thereby extending the filament. 

Victor and his colleagues proposed a modified version of this—a facilitated Brownian ratchet. Instead of it being a wholly random process, water influx would drive the ratchet by inducing localised swelling of the membrane, forcing a gap to open into which G-actin could move to extend the filament tip. To prove this, they turned to instant structured illumination microscopy, a super-resolution method which allowed them to image the plasma membrane and F-actin in CD4+ cells migrating in response to CCL21. As expected, there was a bigger gap between the plasma membrane and the F-actin at the leading edge compared to other parts of the migrating cells. Treatment of the T cells with inhibitors against either SLC12A2 or AQP3 caused a significant shrinkage of the gap, and crucially, when the gap shrunk, the cells were unable to move as fast, linking gap width directly with motility.

To reinforce these results, two other parameters of leading edge formation were assessed in the presence or absence of inhibitors of WNK1, SLC12A2 or AQP3. If the pathway to water influx was inhibited, actin retrograde flow, which can be used as a surrogate for the actin polymerisation rate at the leading edge, was reduced, and pseudopodial branched actin networks, which occur at lower densities at active leading edges, increased.

The conclusive experiment was to show that water entry itself was required for T cell migration. The researchers asked whether simply forcing water into the cells was sufficient to rescue migration in WNK1-inhibited cells. They did this by reducing the tonicity of the medium in the presence of a dose of WNKi that slowed the cells down but was low enough to allow them to polarise. Remarkably, just this forced osmosis increased migration speed in cells where WNK1 had been partially inhibited, and also restored retrograde actin flow and the dearth of branched actin at the leading edge. The effect was not a result of reduced NaCl concentration, as replacing the lost NaCl in the hypotonic medium with L-glucose, an inert osmolyte, could not effectively rescue migration.

This work is a classic example of how studying a well-established process, in this case T cell migration, can still uncover an entirely unexpected regulatory pathway. WNK1 is expressed in many cell types, which begs the question of whether its control of ion and water influx may be important for driving migration in other cell types. It is likely to be doing more too: inhibition of WNK1 has more effect than just inhibiting the other proteins studied, suggesting that it has other as yet undiscovered roles in T cell migration.

Addendum: In 2025, the Tybulewicz lab showed that the WNK1 pathway's role in T cells extends beyond migration and adhesion [3]. The same pathway operating at a later point in a CD4+ T cell's life—when its T cell receptor is engaged by MHC-antigen and switches on the T cell activation machinery—is required for entry of the activated T cell into the cell cycle, and suppression of the ATR-mediated G2/M cell cycle checkpoint. Without WNK1 signalling the T cells cannot activate, and there is no T-cell dependent antibody response. Again, given the broad expression of WNK1, WNK1-dependent water influx may be a common feature of mitogenic pathways in many cell types, both within the immune system and beyond.