Cellular velocity, electrical persistence and sensing in developed and vegetative cells during electrotaxis
Isabella Guido, Douglas Diehl, Nora Aleida Olszok, Eberhard Bodenschatz
CCellular velocity, electrical persistence and sensing in developed and vegetative cellsduring electrotaxis
Isabella Guido, ∗ Douglas Diehl, Nora Aleida Olszok, and Eberhard Bodenschatz
1, 2, 3 Max Planck Institute for Dynamics and Self-Organization (MPIDS), 37077 G¨ottingen, Germany Institute for Dynamics of Complex Systems, Georg-August-University G¨ottingen, 37073 G¨ottingen, Germany Laboratory of Atomic and Solid-State Physics, Cornell University, Ithaca, NY 14853, United States
Cells have the ability to detect electric fields and respond to them with directed migratory movement.Investigations identified genes and proteins that play important roles in defining the migrationefficiency. Nevertheless, the sensing and transduction mechanisms underlying directed cell migrationare still under discussion. We use
Dictyostelium discoideum cells as model system for studyingeukaryotic cell migration in DC electric fields. We have defined the temporal electric persistenceto characterize the memory that cells have in a varying electric field. In addition to imposing adirectional bias, we observed that the electric field influences the cellular kinematics by acceleratingthe movement of cells along their paths. Moreover, the study of vegetative and briefly starved cellsprovided insight into the electrical sensing of cells. We found evidence that conditioned mediumof starved cells was able to trigger the electrical sensing of vegetative cells that would otherwisenot orient themselves in the electric field. This observation may be explained by the presence ofthe conditioned medium factor (CMF), a protein secreted by the cells, when they begin to starve.The results of this study give new insights into understanding the mechanism that triggers theelectrical sensing and transduces the external stimulus into directed cell migration. Finally, theobserved increased mobility of cells over time in an electric field could offer a novel perspectivetowards wound healing assays.
Electrotaxis, also known as galvanotaxis, is the directedmigration of biological cells in a DC electric field. Sinceit was first described over a century ago [1, 2], the eletro-tactic behavior of various cell types, including cancercells, neurons, fibroblast, keratinocytes, leukocytes, en-dothelial and corneal epithelial has been reported [3–12].Electrotaxis is thought to be involved in a wide rangeof physiological processes, such as embryogenesis, neu-ronal guidance, wound healing, and metastasis [13–18].Recently, also
Dictyostelium discoideum ( Dd ), the socialamoeba well-known as a model for studying cell motilityand chemotaxis [19] has proven to be a suitable model forinvestigating electrotaxis [20–22]. However, the mecha-nism triggering the local activation of the signal trans-duction cascade that leads to actin polymerization andmembrane protrusion and more generally the mechanismunderlying the directed cellular movement of Dd cells inthe electric field still awaits clarifications.A study on the involvement of cAMP receptors usingcAR1 − - cAR3 − cells showed that Dd cells, which areunable to sense cAMP, are electrotactically as efficientas wild type cells [21]. The same study showed thatthe cAMP binding transduction unit constituted by the Gα Gγβ complex does not play any rolein the transduction of the extracellular electric signalinto directional movement. Indeed, like cAR1 − - cAR3 − , Gα − and Gβ − mutants also exhibit sustained electro- taxis albeit with a reduced migration speed [21]. Anothermolecular study identified the genes required for the di-rectional switching of electrotactic migration. By genet-ically modulating both guanylyl cyclases (GCases) andthe cyclic guanosine monophosphate (cGMP)-bindingprotein C (GbpC) in combination with the inhibition ofthe phosphatidylinositide 3-kinases (PI3Ks) the cells re-versed their directed migration from the cathode to theanode [23]. Gao et al. [24] uncovered genes involved inthe electrotactic response by identifying 28 strains withdefective electrotaxis and 10 strains with a slightly higherdirectional response. They showed PiaA to be an essen-tial mediator of electrotaxis. This gene encodes a criti-cal component of TORC2, a kinase protein complex thattransduces changes in motility by activating the kinasePKB. Furthermore, they identified several genes thatencode other components of the TORC2-PKB pathway(gefA, rasC, rip3, lst8, and pkbR1) in playing importantroles in the signalling pathway for electrotaxis. Whilegenes and proteins that mediate electrical sensing anddefine the migration direction have been investigated in Dd cells, the cellular kinematic effects caused by electricfield, as well as the initial trigger mechanism of electricalsensing still present many riddles.In this study we characterize the effect of electric fieldson cells in terms of migration velocity and directional-ity as a function of time. In addition, we introduce the a r X i v : . [ q - b i o . CB ] A p r concept of electrical persistence to investigate how cellsinvert their trajectory when the electric field is reversed.We also studied the response of vegetative Dd cells andobserved that the presence of conditioned medium helpsthem to sense the electric field and orient themselves to-wards the cathode. We focus our attention on the con-ditioned medium factor (CMF), a protein that Dd cellssecrete when they begin to starve, as a possible triggerof the cellular electrical sensing. CMF has been shownto coordinate aggregation by regulating several aspectsof cAMP signal transduction such as the activation ofCa influx, adenylyl cyclase, GCases, and gene expres-sion [25, 26]. Besides influencing cAMP signalling, CMFalso participates in regulating cell shape. Moreover, cellmigration relies on pseudopod formation, and CMF ap-pears to allow cells to create pseudopodia more frequentlythan cells without CMF in their surroundings [27]. CMFis therefore a reliable candidate for such a triggering task. MATERIALS AND METHODS
Cell preparation.
All cell lines were derived fromthe axenically growing strain
Dictyostelium discoideum
AX2. Wild type, LimE-GFP, ACA − , and Amib − werecultivated in HL5 medium (Formedium) at 22 ° C onpolystyrene Petri dishes or shaken in suspension at 150rpm. For preparation of experiments, cells were starvedin shaking phosphate buffer (PB, 2 g KH PO and 0.36g Na HPO · O per 1 L, pH 6) for different durationsaccording to the corresponding experiment. For assayswith developed cells, they were starved for 1h, 5h, 6h, 8hat a density of ∼ × cells/mL. The shaking culturewas pulsed with 50 nM cAMP (Sigma) every 6 min overthe course of the starvation time when the experiment re-quired it. After the corresponding starvation time withor without cAMP pulses, the cells were harvested andwashed in PB. An aliquot of the cell suspension was in-jected into the chamber for the electrotactic assay, andthe cells were allowed to spread on the glass substrate for15 min at 22 ° C. During this time a PB flow of 30 µ l/hwas applied to the cells in order to wash away the cAMPproduced by the cells. It was switched to 50 µ l/h duringthe experiments. For experiments with vegetative AX2cells the cells cultivated in HL5 medium were detachedfrom the Petri Dish bottom, washed twice with PB andput directly into the experimental chamber without anyadditional starvation time. For the experiment with con-ditioned medium AX2 cells were shaken for 1 h withoutthe addition of cAMP. They were centrifuged, the bufferwas harvested and centrifuged twice again to eliminatepossible cells. The vegetative cells after having been de-tached from the Petri Dish were washed and afterwardsresuspended in the conditioned buffer. Microfluidic device.
Figure 1-A illustrates the designof the custom-made microfluidic device used during the experiments. Standard soft lithography was used to pro-duce microfluidic channels 1.5 mm wide, 100 µ m high,and 30 mm long. A master mold was fabricated transfer-ring via photholitography a pattern to a layer of photore-sist (SU-8, Micro Resist Technology), spin coated on Siwafer. To obtain the microfluidic device, polydimethyl-siloxane (PDMS, 10:1 mixture with curing agent, Sylgard184, Dow Corning Europe SA) was poured onto the waferand cured for 45 min at 75 ° C. A PDMS block containingthe channels was cut out, inlets and outlets for the PBwashing flow were punched through the PDMS by usinga syringe tip and two holes with a diameter of 6 mm werepunched at the two ends of the channels in order to in-sert the agar bridges. A glass coverslip (24x60 mm, . Electric connection.
The channel was connected tothe power supply through 2% agar salt bridges 13 cmlong. They were prepared in custom-made glass tubeswith an internal diameter of 3 mm (Glasgeraetebau Ochs,Germany) filled with PB supplemented with 2% (w/v)agar. One side of the agar bridge was inserted intothe 6 mm diameter hole at the end of the channel, theother side was placed into buffer reservoirs filled with PB.Ag/AgCl electrodes were immersed into the reservoirsin order to close the circuit channel-power supply. TheAg/AgCl electrodes (12 cm) were prepared by immers-ing two silver wires (99,9%, 1mm in diameter, WindausLaborthechnik, Germany) in household bleach for onehour. A programmable switch device (Siemens) was setto reverse the polarity of the electric field every 30 minwithin a few milliseconds. The resistance of the chan-nel was 840 K Ω and the flowing current 25 µ A. Uponthe application of direct current voltage of 10 V/cm, wemeasured that an electric field of 7 V/cm was applied tothe cells. This voltage drop was due to the agar bridgesthat we used to connect the channel with the power sup-ply generator in order to avoid the harmful effects of theelectric field on the cells, i.e. ions generated by electrol-ysis, changes in pH value, air bubble formation.
Microscopy and cell mobility analysis.
The cellsin the microfluidic channel were observed with an in-verted microscope (Olympus IX-71) in bright field with aDeltaVision imaging system (GE Healthcare), while themigration of the cells were recorded with a CCD camera(CoolSnap HQ2, Photometrics). Data acquisition started20 sec after the application of electric field. Cell imageswere acquired every 20 s for 2 hours. In the experimentswith cells developed for 5-8 h cell centroids were deter-mined manually, and the trajectories of the cell centroidswere traced using MTrackJ, an ImageJ plugin. The tra-jectory velocity was calculated by dividing the total pathlength of cell migration by the time interval. The cellularvelocity was presented as the mean value of the velocitiesof the cells recorded in each 30-minute time interval. Di-
FIG. 1.
Experimental set up and F-actin polymerisation . A. Experimental set up. The electric field in the channelis generated through an indirect contact with the electrodes. The buffer reservoirs provide the electric connection betweenelectrodes and agar bridges (for more technical details refer to Materials and Methods section). B. F-actin polymerisationthrough LimE-GFP. Cells with different morphology use different strategies to reverse their trajectory when the polarity ofthe electric field changes. The cell on the left rounds, extends a pseudopod in the new direction and forms a new leading edgewhereas the cell on the right reverses its trajectory by a doing U-turn. The dashed lines show the changing position of the twocells. Scale bar: 25 µ m. C. Histograms of fluorescence intensity within the cell representing the localization of F-actin. Theleft and right graphs correspond to the cells on the left and the right of the picture, respectively. The change in position ofF-actin from the front to the back in the cell on the left when the electric field is reversed is clearly visible, while in the U-turncell the F-actin remains at the front. rectionality of a cell with respect to the electric field, rep-resentative of the efficiency of the cell to migrate towardthe cathode, was defined as cos θ , where θ is the angle be-tween the vector connecting the starting and ending pointof the cell trajectory and the field line of the electric field.It was calculated every minute. For the visualization ofthe F-actin localization, the cell contours were automat-ically detected using the method described in [28]. Thecenter of mass of cells CM and the intensity-weightedcenter of mass CM W were computed considering the in-tensity of the pixels representing the actin localization.Every 20 seconds we considered that cells were movingtoward the direction of the electric field when the vectorCM-CM W was pointing towards the cathode. When thatvector was pointing in the opposite direction, we consid-ered the cells moving anti-parallel to the electric field (seeFig. S1). The bins in Fig 4-B represent the number ofcells moving parallel or antiparallel to the electric field ateach time point. Also in the experiments with vegetativeand briefly starved cells, cell contours were automaticallydetected using the method described in [28]. The cell cen-troid was then computed and with a time interval of 20sec between subsequent frames its position was trackedusing a custom-made MATLAB program. RESULTS AND DISCUSSION
In this study we present results on the movement of Dd cells in DC electric fields. We focus our analysis on theeffects of electric fields on the cell kinematics and on thecellular tendency to maintain the direction of motion im-posed by the electric field. We show how these effectsare reflected in the actin polarization process. Lastly wepresent the electrotaxis of vegetative and briefly starvedcells and suggest a possible mechanism regulating theelectric sensing independently from the cAMP-induceddevelopment. Electrotaxis of developed starved cells
We investigated the response to the electric field of AX2starving wild-type cells developed for 5 hours undercAMP pulsing prepared with the same development pro-cedure used for the study of chemotaxis (see Material andMethod section). The cells were seeded into a microflu-idic channel where the electric field was applied for twohours. The geometry of the channel guaranteed a uni-form electric field with parallel field lines along the chan-nel (Fig. 1-A). Under the influence of an electric field of7 V/cm cells polarized and exhibited migration towardsthe cathode. We chose not to apply voltages as high as re-ported by other studies (up to 20 V/cm) [20, 21, 24, 29] inorder to reduce the Joule heating and the associated tem-perature increase towards non-physiological conditions(see Material and Methods). We observed a high cel-lular death rate for voltages above 13 V/cm; as a result,we restricted our investigation to 7 V/cm. After exposingthe cells to 7 V/cm for two hours we observed no changein cellular viability and behaviour. This we tested by re-moving the electric field and the flow in the microfluidicdevice after the experimental time and verified that sub-sequently the cells aggregated and followed the naturallife cycle.In order to remove the cooperative effects of cAMP sig-nalling of the wild-type Dd cells, a flow of phosphatebuffer (PB) with a flow speed of 100 µ m/s [30] par-allel or antiparallel to the electric field lines was ap-plied. The successful removal of any extracellular sig-nalling molecule was confirmed by the absence of cellaggregates during the experiments, which would occurfor Dd cells developing normally. The flow speed wasadjusted such that the mechanical shear had no effect onthe directed motility of the cells. In our case we calcu-lated the shear stress to be 6 mPa, a value well below thecritical shear stress for mechanotactical response (0.8 Pa)[31, 32]. We also verified the effect of electro-osmotic flowon the cell motility. We analysed the movement of beadswith a diameter of approx. 1 µ m in the flow induced byelectro-osmosis. The velocity of the beads was found tobe around 8.8 µ m/s, i.e., much smaller compared to thevelocity of the external applied flow. Therefore the ef-fect of the electro-osmotic flow on the cell migration canbe neglected. Altogether, this shows that the directedmigration elicited by the electric field in our experimentswas not influenced by external factors such as the appliedflow, electro-osmotic flow, or chemical gradients.We also found no evidence that extracellular Ca is nec-essary for electrotaxis. The experiments presented herewere conducted with Ca -free phosphate buffer. More-over, the flow in the microfluidic device washes away anycompound released by the cells. This evidence contra-dicts the results by Shanley et al. [33], where it wasfound that electrotaxis was not possible in the absenceof extracellular Ca .When the electric field polarity was reversed, the cellsturned around and migrated towards the new cathode(See Movie S1), confirming the observations reportedby [20, 22]. In response to inverting the field polaritythe cells reversed their trajectory by using two differ-ent strategies, depending on their initial morphology: Inmost cases polarised cells with a single pseudopod or anaccentuated pseudopod achieved reorientation by makinga U-turn while maintaining their morphological polarity.Cells with a less polarised initial morphology reoriented by forming extensions of the pseudopod in the directionof the new cathode, i.e. reversing the front and the rear.With LimE-GFP as an in-vivo marker for F-actin we vi-sualised the distribution of actin during this process. Fig1-B (see also Movie S2) shows an example of cells with thetwo different cellular morphologies: the U-turn cell showsstrong localisation of F-actin at the leading pseudopod,while it is less pronounced in the other cell. In responseto the polarity change of the field, the U-turn cell retainsits F-actin localisation, while in less polarized cells F-actin is redistributed from back to front, forming a newleading edge. This behaviour was also observed duringchemotaxis [34] and demonstrates that the electric fieldtriggers migration by activating the molecular signallingpathway that transduces an external stimulus into actinpolymerisation. Therefore, an intracellular electric polar-ization due to passive electrostatic effects of the electricfield can be excluded. Cellular velocity, directionality and temporalelectric persistence
When the cells were exposed to the electric field for twohours, with the field reversing every 30 minutes, we ob-served that the migratory velocity increased continuouslyduring the observation period. The increasing length ofthe cellular trajectory over time is clearly visible in Fig 2,and the corresponding cell velocity increases from 3.27 ± µ m/min in the first 30 minutes to 8.3 ± µ m/minafter 2 hours (Fig 3-C ). To investigate whether this ob- FIG. 2.
Cell tracking diagrams of fully developed AX2cells . The length of the migration paths increases signifi-cantly over time. In each experiment at least 45 cells wereanalysed. The diagrams refer to cells that were starved for 5hours. They originated from at least three different experi-ments. servation depends on the stage of development inducedby the cAMP pulsing procedure, we tested the responseof cells that were starved and pulsed for 5, 6 or 8 hours.The cell speed for the case of 6 and 8 hours is shownin Fig. 3-A,B. The acceleration of the cells caused bythe electric field was also observed in these cases andcalculated as a = ∆v / ∆t , with ∆v the difference in ve-locities between time intervals and ∆t the time inter-val of 30 minutes. Figure 3-D shows that the three cellpopulations reacted to the electric field with an increaseof the initial velocity, regardless of their developmentalstage. We speculate that the cAMP pathway involved in FIG. 3.
Cell acceleration . A. Migration velocity of Ax2cells starved for 6 hours. B. Migration velocity of Ax2 cellsstarved for 8 hours. C. Migration velocity of Ax2 cells starvedfor 5 hours. D. Cell acceleration over time for cells starvedand pulsed 5, 6 or 8 hours. All data are represented as mean ± s.e.m the activation of the aggregation adenylyl cyclase (ACA)and associated production of cAMP could play a signif-icant role. In fact, we observed that two mutant strainsunable to produce cAMP and aggregate, namely ACA − and Amib − , did exhibit electrotaxis, but did not showany increase in velocity over time (See S2). Neverthelessat this stage this assumption is a speculation and only adetailed molecular study similar to [24] can elucidate thetrue reason for the speed up.Interestingly, the directionality of the cells did not changesignificantly (Fig 4-A) reaching a plateau value Dir max ,which ranged from 0.54 in the first 30 min to 0.65 in thelast interval. The transition phase between two plateausprovides information about the reaction of cells to thereversal of the polarity of electric field. By fitting thiscells response to the exponential function
Dir max − ce tτ we calculated that it is characterized by a time constant τ − = 2.94 min in the first electric field reversal, fol-lowed by τ − = 5 min and τ − = 5.64 min in thesecond and third one, respectively (where the subscriptof τ refers to the corresponding time interval in minutes).It appears clear that the longer the cells are under the in-fluence of the electric field, the longer it takes for them toreverse their trajectory when the polarity of the electric FIG. 4.
Directionality and F-actin localization uponelectric field reversal
A. Directionality of Dd cells in anelectric field. The polarity of the field is reversed every 30minute as indicated by the red arrows. The graph is obtainedfrom at least three different experiments. The data are dis-played as mean ± s.e.m . B. F-actin localization within thecell during cellular migration and its dynamics when the po-larity of electric field changes. At each point in time, thenumber of cells with F-actin localized towards the cathode(blue), towards the anode (green) and their difference (red)are displayed. The black dotted line indicated the time pointof polarity reversal. field is reversed. Thus, the adaptation to their environ-ment prevents the cells from reacting immediately to anychange in the electric field and rearranging the migratorymachinery. To characterize this behavior, we defined thetemporal electrical persistence or electrical memory, i.e.the time required by cells to reverse their trajectory inan inverted electric field. In this way, we could gain in-sights into the cellular sensing process and its transduc-tion into the molecular pathway. For this purpose weanalyzed the localization of the F-actin network by usingLimE-cells and we could observe how the adaptation be-havior was reflected by F-actin polymerization inside thecell. The cells were kept in the electric field for 90 min-utes. Afterwards, the polarity was reversed and the cellswere observed for 20 minutes (Fig. 4-B). We evaluatedthe occurrence of cells in which F-actin was localized to-wards the cathode and of cells in which it was localizedin the opposite direction. The difference between thesetwo populations represents the net population of cells po-larized in the direction defined by the electric field (seesection Material and Method for analysis details). Theseresults show that the cells need 4.3 min ±
20 sec to shiftthe F-actin in the new direction of the electric field afterthe polarity change.To better characterized the sensing mechanism, we anal-ysed the temporal electrical persistence of the cells whenthe electric field was removed and tested how long it tookfor the cells to“forget” the environmental stimulus andmove randomly. Cells migrating without any electricalstimulus showed no preferred direction (Fig. 5-A), andthe average directionality of their random movement was0 ± ±
20 sec (Fig. 5-B).Also in this case we could observe that the directed move-ment decreased with time when the electric stimulationwas switched off.
FIG. 5.
Electrical persistance without electric field
A.Directionality of Dd cells that were never exposed to electricfield and move randomly (blue) and cells where the electricfield was switched off after 90 min (red). The electric fieldwas removed at t=10 min. It is clear that the directionalitydecreases during the time after the field removal. The dataare represented as mean ± s.e.m . B. F-actin localization insidethe cell during the cellular migration and its dynamics whenthe electric field is switched off. Electrotaxis of initially vegetative cells
In this study we were also interested to understand themechanism involved in the initiation of the electricalsensing in cells. For this purpose we analysed the be-haviour of vegetative and briefly starved Dd cells underthe influence of electric fields. It allowed us to studythe response of cells before they entered the developmentphase and went though the program of gene-expressionchanges induced by cAMP pulses.Vegetative wild type AX2 cells cultivated in HL5 mediumwere washed and resuspended in PB, then seeded imme-diately into the microfluidic device. We analysed theirbehaviour under the influence of electric fields over a pe-riod of up to 7 hours, during which the polarity of theelectric field was reversed every 30 minutes. No directedmovement was observed; rather, the cells migrated ran-domly during the whole observation time. Figure 6-Ashows the behaviour of vegetative cells in the electric fieldfor 2 hours. After 7 hours deprived of nutrients in PB thecells did not show the typical characteristics of starvedcells in an electric field such as morphological changes andmigration towards the cathode. As a control experiment,we tested the effect of the electric field on cell physiologyand repeated the experiment by keeping the cells underthe flow for 5 hours without electric field and then apply-ing the electric field for 2 hours. Again, the cells showedno preferred direction of movement. In order to check theinvolvement of calcium (Ca ) in this behaviour we re-peated the experiments by substituting the Ca -free PBfor a buffer containing 1mM CaCl. No remarkable dif-ference was observed in the cells response to the electricfield. We concluded that under these experimental con-ditions the cells were not able to enter the developmentstage normally induced by starvation. Thus, entering thedevelopmental phase and activating genes related to thestarvation program are necessary steps for enabling cellsto sense and respond to the electric field. Electrotaxis of briefly starved cells
The influence of the initial starvation on the behaviourof cells in electric fields was investigated by experimentswith cells starved for 1 hour both with and without addi-tional cAMP pulses in shaking culture (see Materials andMethods). We observed that cells starved for 1 hour withand without additional cAMP pulses reacted similarly bymigrating or realigning their cell body towards the cath-ode when the electric field was applied for 30 minutes.They inverted their orientation towards the new cath-ode when the polarity of the electric field was reversed(Fig. 6-B-C, see also S3). Interestingly, after one hourthe directed movement decreased with time, independentof the change in the polarization of the electric field. Asin the case of vegetative cells, we repeated these exper-
FIG. 6.
Histograms of the propagation angle distribu-tion . Histograms of the propagation angle distribution forvegetative cells (A), briefly starved cells in a shaking culturewithout (B) and with (C) external cAMP pulses. Cell motionis biased towards the cathode with θ = 180 ◦ in the first 30min and θ = 0 ◦ or 360 ◦ when the polarisation of the electricfield has been reversed. The histograms represents the dis-tribution of the propagation angle θ ∈ [0 ◦ , ◦ ] of each timestep with respect to the electric field lines pointing towardsthe cathode. Every histogram resulted by the analysis of 100- 300 cells from three different experiments. iments using a buffer containing 1mM CaCl as washingout flow. Again, no difference in the attenuation effectafter 1 hour was observed.These results show that the induction of development bynutrient deprivation leads to an increased electrotacticcapacity, but not the presence of exogenous cAMP. Thisis not surprising since Yuen at al. have shown that theability of cells to transduce external cAMP occurs 2 hoursafter starvation [25]. Conditioned Medium restores the cellular sensingfor external electric stimuli
So far we have observed a difference in electrical sensingbetween vegetative and 1-hour-starved cells. We verified that the presence of external cAMP is not essential fortriggering the electrical sensing of cells. We therefore as-sume that the electrical sensing is triggered by a signalingmolecule that the cells immediately release at the onsetof starvation. In order to test this hypothesis, we studiedthe reaction of vegetative cells to electric fields when theywere resuspended in conditioned medium. We starvedwild-type cells in a shaking culture for 1 h, centrifugedthem, removed the conditioned buffer and used this con-ditioned medium to resuspend vegetative cells. Theseconditioned vegetative cells were immediately seeded intothe microfluidic channel and conditioned medium wasalso used as washing flow instead of PB during the en-tire observation period of 2 hours. Interestingly, underthese conditions the conditioned vegetative cells showedan electrotactic motion and reorient their cellular bodytowards the cathode. Fig. 7 clearly shows the reaction ofthe cells to the electric field. The involvement of condi-tioned medium in the electrical sensing and its triggeringeffect is clearly visible when comparing the cellular re-sponse of the initially vegetative cells (Fig. 6-A) andthe conditioned vegetative cells (Fig. 7). By comparingthe reactions to the electric field of the conditioned veg-etative cells (Fig. 7) and the briefly starved cells (Fig.6-B-C), we can see that conditioned medium rescues thesensing of cells for electric fields, but the cells react fora shorter time. Although there are dozens of autocrine
FIG. 7.
Histograms of the propagation angle distribu-tion for conditioned vegetative cells.
Every histogramresulted by the analysis of 300 - 350 cells from three differentexperiments agents in the conditioned medium of cells starved for onlyone hour, we speculate that the decisive factor triggeringthe cellular electrical sensing is the conditioned mediumfactor (CMF) [25], a protein that the cells release at theonset of starvation. It is secreted throughout develop-ment by Dd cells, but not by vegetative cells [35]. Var-ious studies suggest that CMF is essential for cell-cellcommunication and for the coordination of cell aggrega-tion. It acts during the earliest stages of starvation andis used by the cells to enter the developmental phase andinduce the expression of selected genes depending on thespatial density of cells [26, 36].While conditioned vegetative cells immediately showedelectrotaxis, vegetative cells were not able to sense theelectric field when seeded in a channel with a PB flow upto 7 hours. We attribute this behavior to the flow thatwashes away not only cAMP but also the CMF secretedby the cells. Under these conditions they cannot enter thedevelopmental phase. According to our hypothesis, CMFreceptors might be responsible for the signal transductionof the external electric field and thus for the activationof the signaling pathways that lead the cells to polariza-tion and directed migration. It is known that the geneencoding CMF receptors is expressed in vegetative cellsand CMF receptors accumulate on the cell membranewhen starvation sets in [37]. In addition, CMF bindsto CMF receptors and regulates PldB [27], a phospholi-pase D homologue known to regulate actin localization[38] and pseudopod formation [27], which are essentialfor cell migration. This is consistent with the fact thatin our microfluidic experiments vegetative cells did notshow any electrotaxis, while briefly starved cells and con-ditioned vegetative cells were electrotactic. The temporaldecrease in electrotactic directionality in all three casesmay be attributable to the lack of increase in CMF. CONCLUSION
In this study we present results on the electrotacticbehavior of Dd cells at different developmental stages.We showed that fully developed wild-type Dd cells re-spond to the electric field by increasing their migratoryvelocity over time. However, the mutant strains ACA − and Amib − migrate at a constant velocity over theobservation period. At this stage it is unclear whetherthe enzyme ACA, which is known to be essential for os-cillatory cAMP signals, is involved in the determinationof the electrotactic migratory velocity. We introducedthe temporal electrical persistence as the time delay ofcells to respond to the variation of the external electricfield. We quantified the time interval in which the cellsrearrange their migratory machinery either to reversetheir migratory trajectory or to “forget” the electricstimulus after switching off of the field. This providesinsight into the mechanism by which cells transducethe sensing of the electric stimulus into a kinematicresponse. By analysing vegetative cells that are notcapable of directed movement, we show that the electricsensing can be rescued by using conditioned medium.We believe that among the autocrine molecules that thecells release at the onset of starvation, CMF is the primecandidate for triggering the electrical sensing. Onlywhen cells have been exposed to CMF (self-produced orsupplied with conditioned medium) do they respond tothe electric field with directed migration. This assump-tion requires an analysis using the protein CMF insteadof the conditioned medium. Until now, the protein could not be made available to test our assumption. Weencourage researchers to follow this line of inquiry. Acknowledgments
E.B. and I.G. acknowledge sup-port from the MaxSynBio Consortium which is jointlyfunded by the Federal Ministry of Education and Re-search of Germany and the Max Planck Society. Weacknowledge Albert Bae, Narain Karedla, and ChristianWestendorf for fruitful discussion and suggestions; MarenM¨uller, Katharina Gunkel, Achim Hillebrandt, and Ric-cardo Guido for technical support. ∗ [email protected][1] Dineur E. Note sur la sensibilite des leucocytesa l’electricite. Bull Seances Soc Belge Microsc.1891;18:113–118.[2] Verwon M. Untersuchungen ¨uber die polare Erregungder lebendigen Substanz durch den konstanten Strom.IIIMitteliung Pflugers Arch Eur J Physiol. 1896;62:415–450.[3] Djamgoz MBA, Mycielska M, Madeja Z, Fraser SP, Ko-rohoda W. Directional movement of rat prostate cancercells in direct-current electric field. Journal of Cell Sci-ence. 2001;114(14):2697–2705.[4] Fraser SP, Diss JKJ, Chioni AM, Mycielska ME, PanH, Yamaci RF, et al. Voltage-Gated Sodium ChannelExpression and Potentiation of Human Breast CancerMetastasis. Clinical Cancer Research. 2005;11(15):5381–5389. doi:10.1158/1078-0432.CCR-05-0327.[5] Huang YJ, Hoffmann G, Wheeler B, Schiapparelli P,Quinones-Hinojosa A, Searson P. Cellular microen-vironment modulates the galvanotaxis of brain tu-mor initiating cells. Scientific Reports. 2016;6:21583.doi:10.1038/srep21583.[6] Cooper MS, Keller RE. Perpendicular orientation anddirectional migration of amphibian neural crest cells in dcelectrical fields. Proceedings of the National Academy ofSciences of the United States of America. 1984;81(1):160–164.[7] Sheridan DM, Isseroff RR, Nuccitelli R. Impo-sition of a Physiologic DC Electric Field Altersthe Migratory Response of Human Keratinocyteson Extracellular Matrix Molecules. Journal ofInvestigative Dermatology. 1996;106(4):642 – 646.doi:http://dx.doi.org/10.1111/1523-1747.ep12345456.[8] Allen G, Mogilner A, Theriot J. Electrophore-sis of Cellular Membrane Components Createsthe Directional Cue Guiding Keratocyte Galvan-otaxis. Current Biology. 2013;23(7):560 – 568.doi:http://dx.doi.org/10.1016/j.cub.2013.02.047.[9] Zhao M, McCaig CD, Agius-Fernandez A, Forrester JV,Araki-Sasaki K. Human corneal epithelial cells reori-ent and migrate cathodally in a small applied elec-tric field. Current Eye Research. 1997;16(10):973–984.doi:10.1076/ceyr.16.10.973.9014.[10] Li X, Kolega J. Effects of Direct Current Electric Fieldson Cell Migration and Actin Filament Distribution inBovine Vascular Endothelial Cells. Journal of VascularResearch. 2002;39(5):391–404. [11] Rapp B, de Boisfleury-Chevance A, Gruler H. Galvano-taxis of human granulocytes. European Biophysics Jour-nal. 1988;16(5):313–319. doi:10.1007/BF00254068.[12] Lin F, Baldessari F, Gyenge CC, Sato T, Chambers RD,Santiago JG, et al. Lymphocyte Electrotaxis In Vitro andIn Vivo. The Journal of Immunology. 2008;181(4):2465–2471. doi:10.4049/jimmunol.181.4.2465.[13] L F Jaffe RN. Electrical Controls ofDevelopment. Annual Review of Bio-physics and Bioengineering. 1977;6(1):445–476.doi:10.1146/annurev.bb.06.060177.002305.[14] Erickson C, Nuccitelli R. Embryonic fibroblast motilityand orientation can be influenced by physiological electricfields. The Journal of Cell Biology. 1984;98(1):296–307.[15] Robinson KR. The responses of cells to electrical fields:a review. The Journal of Cell Biology. 1985;101(6):2023–2027.[16] McCaig CD, Rajnicek AM, Song B, Zhao M. Control-ling Cell Behavior Electrically: Current Views and Fu-ture Potential. Physiological Reviews. 2005;85(3):943–978. doi:10.1152/physrev.00020.2004.[17] McCaig CD, Song B, Rajnicek AM. Electrical di-mensions in cell science. Journal of Cell Science.2009;122(23):4267–4276. doi:10.1242/jcs.023564.[18] Mycielska ME, Djamgoz MBA. Cellular mechanismsof direct-current electric field effects: galvanotaxisand metastatic disease. Journal of Cell Science.2004;117(9):1631–1639. doi:10.1242/jcs.01125.[19] Van Haastert PJM, Devreotes PN. Chemotaxis: sig-nalling the way forward. Nat Rev Mol Cell Biol.2004;5(8):626–634.[20] Zhao M, Jin T, McCaig CD, Forrester JV, Devreotes PN.Genetic analysis of the role of G protein-coupled recep-tor signaling in electrotaxis. The Journal of cell biology.2002;157(6):921–927. doi:10.1083/jcb.200112070.[21] Zhao M, Song B, Pu J, Wada T, Reid B, Tai G,et al. Electrical signals control wound healingthrough phosphatidylinositol-3-OH kinase-gammaand PTEN. Nature. 2006;442(7101):457–460.doi:10.1038/nature04925.[22] Sato MJ, Ueda M, Takagi H, Watanabe TM,Yanagida T, Ueda M. Input-output rela-tionship in galvanotactic response of Dic-tyostelium cells. BioSystems. 2007;88(3):261–272.doi:10.1016/j.biosystems.2006.06.008.[23] Sato MJ, Kuwayama H, van Egmond WN, TakayamaALK, Takagi H, van Haastert PJM, et al. Switching di-rection in electric-signal-induced cell migration by cyclicguanosine monophosphate and phosphatidylinositol sig-naling. Proceedings of the National Academy of Sciencesof the United States of America. 2009;106(16):6667–72.doi:10.1073/pnas.0809974106.[24] Gao R, Zhao S, Jiang X, Sun Y, Zhao S, Gao J, et al.A large-scale screen reveals genes that mediate electro-taxis in Dictyostelium discoideum. Science Signaling.2015;8(378):ra50. doi:10.1126/scisignal.aab0562. [25] Yuen IS, Jain R, Bishop JD, Lindsey DF, DeeryWJ, Van Haastert PJ, et al. A density-sensingfactor regulates signal transduction in Dictyostelium.The Journal of Cell Biology. 1995;129(5):1251–1262.doi:10.1083/jcb.129.5.1251.[26] Clarke M, Gomer RH. PSF and CMF, autocrinefactors that regulate gene expression during growthand early development ofDictyostelium. Experientia.1995;51(12):1124–1134. doi:10.1007/BF01944730.[27] Gomer RH, Jang W, Brazill D. Cell density sens-ing and size determination. Development, growth &differentiation. 2011;53(4):482–494. doi:10.1111/j.1440-169X.2010.01248.x.[28] Amselem G, Theves M, Bae A, BodenschatzE, Beta C. A Stochastic Description of Dic-tyostelium Chemotaxis. PLOS ONE. 2012;7(5):1–11.doi:10.1371/journal.pone.0037213.[29] Gao R, Zhang X, Sun Y, Kamimura Y, MogilnerA, Devreotes PN, et al. Different Roles of Mem-brane Potentials in Electrotaxis and Chemotaxis of Dic-tyostelium Cells. Eukaryotic Cell. 2011;10(9):1251–1256.doi:10.1128/EC.05066-11.[30] Beta C, Frohlich T, Bodeker HU, Bodenschatz E. Chemo-taxis in microfluidic devices-a study of flow effects. LabChip. 2008;8:1087–1096. doi:10.1039/B801331D.[31] D´ecav´e E, Garrivier D, Br´echet Y, Fourcade B,Bruckert F. Shear flow-induced detachment kinet-ics of Dictyostelium discoideum cells from solid sub-strate. Biophysical journal. 2002;82(5):2383–2395.doi:10.1016/S0006-3495(02)75583-5.[32] D´ecave E, Rieu D, Dalous J, Fache S, Brechet Y,Fourcade B, et al. Shear flow-induced motilityof Dictyostelium discoideum cells on solid substrate.Journal of Cell Science. 2003;116(Pt 21):4331–4343.doi:10.1242/jcs.00726.[33] Shanley LJ, Walczysko P, Bain M, MacEwan DJ, ZhaoM. Influx of extracellular Ca2+ is necessary for elec-trotaxis in Dictyostelium. Journal of Cell Science.2006;119(22):4741–4748. doi:10.1242/jcs.03248.[34] Van Haastert PJM, Devreotes PN. Chemotaxis: sig-nalling the way forward. Nat Rev Mol Cell Biol.2004;5(8):626–634.[35] Gomer RH, Yuen IS, Firtel RA. A secreted 80 ××