Massive endocytosis driven by lipidic forces originating in the outer plasmalemmal monolayer: a new approach to membrane recycling and lipid domains

Abstract

The roles that lipids play in endocytosis are the subject of debate. Using electrical and imaging methods, we describe massive endocytosis (MEND) in baby hamster kidney (BHK) and HEK293 cells when the outer plasma membrane monolayer is perturbed by the nonionic detergents, Triton X-100 (TX100) and NP-40. Some alkane detergents, the amphipathic drugs, edelfosine and tamoxifen, and the phospholipase inhibitor, U73122, are also effective. Uptake of the membrane tracer, FM 4-64, into vesicles and loss of reversible FM 4-64 binding confirm that 40-75% of the cell surface is internalized. Ongoing MEND stops in 2-4 s when amphipaths are removed, and amphipaths are without effect from the cytoplasmic side. Thus, expansion of the outer monolayer is critical. As found for Ca-activated MEND, vesicles formed are <100 nm in diameter, membrane ruffles are lost, and β-cyclodextrin treatments are inhibitory. However, amphipath-activated MEND does not require Ca transients, adenosine triphosphate (ATP) hydrolysis, G protein cycling, dynamins, or actin cytoskeleton remodeling. With elevated cytoplasmic ATP (>5 mM), MEND can reverse completely and be repeated multiple times in BHK and HEK293 cells, but not cardiac myocytes. Reversal is blocked by N-ethylmaleimide and a nitric oxide donor, nitroprusside. Constitutively expressed Na/Ca exchangers internalize roughly in proportion to surface membrane, whereas Na/K pump activities decrease over-proportionally. Sodium dodecyl sulfate and dodecylglucoside do not cause MEND during their application, but MEND occurs rapidly when they are removed. As monitored capacitively, the binding of these detergents decreases with MEND, whereas TX100 binding does not decrease. In summary, nonionic detergents can fractionate the plasma membrane in vivo, and vesicles formed connect immediately to physiological membrane-trafficking mechanisms. We suggest that lateral and transbilayer inhomogeneities of the plasma membrane provide potential energies that, when unbridled by triggers, can drive endocytosis by lipidic forces.

Figure 1.

Activation of MEND in a BHK cell by extracellular application of 120 µM TX100 for 10 s. Membrane area, i.e., C_m, decreases at a maximal rate of 13% per second, and the endocytic response stops abruptly upon the removal of detergent. During MEND, there is a barely detectable transient increase of cell conductance (G_m). Other cell electrical parameters do not change.

Figure 2.

Salient characteristics of TX100 and NP-40–induced MEND. In A and B, the pipette solution contained 2 mM ATP and 0.2 mM GTP. C is without nucleotides, D is with 2 mM AMP-PNP and no other nucleotides, and E is with 2 mM ATP and 0.2 mM GTP. (A) Average MEND responses of BHK cells upon the application of four different concentrations of TX100 (n = 5–7). (B) Concentration dependence of the maximal rate of change of C_m by applying TX100 in A. The response rate increases with roughly the fourth power of the concentration, given by solid line. (C) MEND in a small BHK cell at 22°C. Variance of RMS noise of C_m increases immediately upon the application of NP-40 and decreases during the decline of C_m. The inset shows the residual signal after subtracting an exponential function from the declining C_m signal. (D) MEND induced by 120 µM NP-40. Upon the application of NP-40, MEND is preceded by a rapid rise of cell C_m, and MEND stops within 2–3 s upon removing detergent. After a second MEND response, detergent causes small increases of C_m that reverse almost as quickly as solutions can be changed. (E) Rapid pipette perfusion of 150 µM NP-40 into a BHK cell by pipette perfusion. As in four additional experiments, NP-40 has no effect from the cytoplasmic side, nor does intracellular detergent have any evident influence on the MEND response to extracellular detergent.

Figure 3.

Characterization of TX100-induced MEND using simultaneous optical and electrical recordings in BHK cells. The pipette solutions contained 2 mM ATP and 0.2 mM GTP. (A) 5 µM FM 4–64 was applied and removed twice, and then 120 µM TX100 was applied for 30 s, inducing a 66% MEND response. FM dye was applied and removed twice more. Membrane binding of FM dye decreases 73% after MEND. (B) 5 µM FM 4–64 was applied once and washed. During the second application, 120 µM TX100 was applied for 30 s, inducing a 65% MEND response. After MEND, FM dye was applied and removed three more times. FM fluorescence remaining in the cell does not wash off and amounts to 70% of pre-MEND fluorescence, corresponding roughly to the amount of membrane internalized as per capacitive measurements.

Figure 4.

TX100-induced MEND does not depend on canonical endocytic proteins. (A) TX100-induced MEND after extensive disruption of cytoskeleton, phosphoinositides, and ATP-dependent processes (5 µM latrunculin with 2 µM Ca, no ATP or GTP, and 2 mM AMP-PNP). FM 4–64 was applied three times before MEND. During the fourth application, 120 µM TX100 was applied, causing a 77% MEND response. Thereafter, 85% of FM dye fluorescence was retained, and FM dye was reapplied twice. The reversible FM fluorescence amounts to 15% of pre-MEND fluorescence. (B) Further characteristics of MEND induced by 120 µM. As indicated by the open bar graphs, TX100-induced MEND was not inhibited when the cytoplasmic solution contained 0.2 mM guanosine 5′-[γ-thio]triphosphate (GTPγS), 5 µM latrunculin with 2 mM AMP-PNP, 50 µM of an unmyristolated dynamin inhibitory peptide (DynPep; Tocris Bioscience), or 0.2 mM of the organic dynamin inhibitor, dynasore.

Figure 5.

Analysis of NP-40 responses of HEK293 cells by scanning and transmission elecronmicroscopy (SEM and TEM). (A) Demonstration of NP-40–induced MEND in a cell population. Cells were incubated with FM 1-43FX during NP-40 application for 30 s, and then washed and fixed as described in Materials and methods without FM dye. A bright rim of fluorescence was generated in cells treated with detergent, thereby verifying the operation of MEND in this protocol. (B) Scanning electron micrographs of HEK293 cells before (left) and after (right) treatment with NP-40. Control cells show an extensive mesh of membrane ruffles, whereas NP-40 cells appear nearly smooth in comparison. Bar, 10 µm. (C) TEM image of an NP-40–treated HEK293 cell. Detergent-treated cells show large DAB-stained vacuoles (>20 observations). Bar, 5 µm. Stained vesicles are indicated with arrows. (D) DAB-stained vacuoles are significantly increased in HEK293 cells after treatment with NP-40 for 60 s (n ≥ 26; P < 0.001).

Figure 6.

Reversal of TX100-induced MEND in BHK cells. (A) TX100-induced MEND reverses completely when the cytoplasmic solution contains 8 mM ATP, 0.2 mM GTP, no lithium, and a low free cytoplasmic Ca concentration (1 mM EGTA with 0.2 mM total Ca; free Ca calculated to be 0.15 µM; >20 observations). In this recording, MEND is induced with 200 µM TX100, and it reverses with a progressively longer halftime of 5–16 min over five repetitions. (B) MEND does not recover in the absence of ATP and the presence of a nonhydrolyzable ATP analogue (2 mM; AMP-PNP). (C) MEND recovery is completely blocked by 0.5 mM NEM added to both extracellular and cytoplasmic solutions (five similar observations). (D) MEND recovery is strongly inhibited by cytoplasmic application of 16 µM of the NO donor, nitroprusside (four similar observations), whereas it is unaffected by 0.3 mM cGMP (five similar observations). (E) When MEND recovery has progressed partially for 3 min, a Ca transient can induce complete recovery of membrane area in seconds. In this experiment, a Ca transient was evoked before MEND to determine the immediately available membrane pool for exocytosis. Thereafter, MEND amounts to a loss of 66% of the cell surface, and after 3 min, a Ca transient causes exocytosis of a three times greater amount of membrane than was initially available.

Figure 7.

Functional and optical analysis of the influence of MEND on NCX1. (A) MEND responses and NCX1 currents in a BHK cell perfused with a high (20 mM) EGTA concentration to negate Ca transients and high ATP (8 mM) to promote MEND reversal. For the most part, MEND reverses within 2 min. We note that this experiment was performed with standard solutions, not modified standard solutions that promote recovery. NCX1 current is decreased after each MEND response and recovers partially during recovery of C_m. (B) Time-lapse confocal imaging of an HEK293 cell stably expressing NCX1–pHluorin fusion protein before and after exposing the cell to 150 µM NP-40 for 10 s. Fluorescence arising from NCX1 at the cell surface is determined as the fluorescence activated by jumping extracellular pH from 6.0 to 8.0. MEND induced by NP-40 causes a C_m drop of 48%, whereas the fluorescence intensity of cell surface NCX1 (i.e., fluorescence jump on changing solutions from pH 6.0 to 8.0) decreases by 37%. The presence of fluorescence at pH 6.0 after MEND indicates that NCX1 has entered an internal compartment that does not immediately acidify.

Figure 8.

Comparison of TX100-induced MEND and its influence on Na/K pump activity in a mouse myocyte, a BHK cell, and an HEK293 cell. All results use modified cytoplasmic solution (Lariccia et al., 2011) with 8 mM ATP and 0.2 mM GTP, and MEND is induced with 200 µM TX100. Pump currents are activated by applying and removing 6 mM KCl in exchange for 6 mM NaCl. Pump activation is evaluated both as pump current and as a capacitive signal that reflects a decrease of extracellular cation binding during pump activity. (A) Mouse myocyte. TX100 induces a 42% MEND response (>10 observations), whereas pump signals are decreased by 62%. Neither membrane area nor pump activity recovers over a 7-min observation period. (B) BHK cell. MEND amounts to 70% of membrane area, pump activity decreases by the same percentage, and both membrane area and pump activity recover over 8 min (four observations). (C) HEK293 cell. Na/K pump activity decreases over-proportionally to the loss of membrane area associated with MEND, and pump activity does not recover during the recovery of membrane area over 8 min (10 observations).

Figure 9.

MEND responses induced by the application and removal of SDS. Cytosolic solutions contained 2 mM ATP and 0.2 mM GTP. (A) Within a few seconds, 50 µM SDS caused a small increase of C_m, and then C_m continued to rise nearly linearly. MEND occurred within seconds upon SDS removal. (B) Rapid pipette perfusion of 150 µM SDS into a BHK cell with 2 mM ATP was without effect on C_m. Thereafter, membrane fusion occurred normally upon the activation of Ca influx by NCX1. Finally, the application and removal of 50 µM of extracellular SDS caused a 50% MEND response over 20 s after SDS was removed.

Figure 10.

Capacitive binding signals of detergents before and after MEND. Standard cytoplasmic solutions with 2 mM ATP and 0.2 mM GTP in the cytoplasmic solution. In each panel, the rising phases of capacitive binding signals, indicated by filled circles in the complete C_m records, are shown in the inset at higher resolution. (A) TX100 capacitive binding signals are unchanged by MEND, causing a 60% loss of plasma membrane. As indicated below the Cm record, a low concentration of TX100 (70 µM) was applied and removed multiple times, with each application giving rise to an ∼5-pF increase of C_m that reverses with a time constant of 2–3 s. MEND was induced by applying 200 µM TX100 for 25 s, and thereafter the capacitive binding signals for the low TX100 concentration were unchanged (five similar observations). (B) Capacitive binding signals for 30 µM SDS are markedly decreased by SDS-induced MEND. Because SDS does not cause MEND during its application, the capacitive binding signal can be evaluated with the same detergent application that induces MEND. The capacitive binding signals decrease by a fractional amount that is similar to the fractional loss of membrane. (C) Capacitive binding signals for 30 µM DDG are markedly decreased by DDG-induced MEND. (D) Capacitive binding signals for 500 µM deoxycholate are markedly decreased by TX100-induced MEND.

Figure 11.

Activation of MEND by diverse amphipaths in BHK cells. Cytoplasmic solutions contain 2 mM ATP and 0.2 mM GTP. (A) 5 µM U73122, but not the inactive “control” analogue, U73343, induces MEND within 30 s. (B) 20 µM edelfosine induces MEND. (C) 40 µM tamoxifen induces MEND. (D) 20 µM LPC induces MEND. Similar to detergent-induced MEND, LPC-induced MEND stops abruptly upon the removal of LPC.

Figure 12.

Endocytosis driven by lipidic forces: a hypothesis. (1) The outer monolayer consists of Lo and Ld domains of small size and equal prevalence. Lipids diffuse rapidly between domains, with affinity differences for Lo versus Ld domains being less than one log unit. “Lipid shells” around membrane proteins (Anderson and Jacobson, 2002) need not be synonymous with Lo and Ld domains. (2) Nonionic amphipaths expand Ld domains and promote cap formation with buckling of the plasmalemma, which is associated with domain coalescence and protein sorting. Ca transients may trigger the generation of endocytosis-promoting lipids and their movement into the outer monolayer (Hilgemann and Fine, 2011; Lariccia et al., 2011). (3) Membrane internalization. Coalescence of Lo domains within expanded Ld domains promotes negative curvature, vesiculation, and fission without adapters or dynamins, similar to the generation of ceramide domains via SMases (Lariccia et al., 2011). Nonionic amphipaths remain in Ld domains at the cell surface. (4) MEND generates vesicles that follow normal trafficking pathways to endosomes with recycling back to the plasmalemma via ATP-dependent processes that are inhibited by NEM and oxidative stress.