If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Pediatrics, University of Pittsburgh, Pittsburgh, Pennsylvania 15261Department of Cell Biology, University of Pittsburgh, Pittsburgh, Pennsylvania 15261
To whom correspondence should be addressed: Renal-Electrolyte Division, Dept. of Medicine, University of Pittsburgh School of Medicine, A919 Scaife Hall, 3550 Terrace St., Pittsburgh, PA 15261. Tel.: 412-647-3121; Fax: 412-648-9166.
Department of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania 15261Department of Cell Biology, University of Pittsburgh, Pittsburgh, Pennsylvania 15261Pharmacology and Chemical Biology, University of Pittsburgh, Pittsburgh, Pennsylvania 15261
* This work was supported by
National Institutes of Health
Grants
DK103834
(to S. S.), DK051391 (to T. R. K.), DK096990 (to G. A. S.), and DK079307 (to T. R. K.). The authors declare no conflict interests. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. 1 Present address: Dept. of Pediatrics, Washington University, St. Louis, MO 63110.
Mechanotransduction in Caenorhabditis elegans touch receptor neurons is mediated by an ion channel formed by MEC-4, MEC-10, and accessory proteins. To define the role of these subunits in the channel's response to mechanical force, we expressed degenerin channels comprising MEC-4 and MEC-10 in Xenopus oocytes and examined their response to laminar shear stress (LSS). Shear stress evoked a rapid increase in whole cell currents in oocytes expressing degenerin channels as well as channels with a MEC-4 degenerin mutation (MEC-4d), suggesting that C. elegans degenerin channels are sensitive to LSS. MEC-10 is required for a robust LSS response as the response was largely blunted in oocytes expressing homomeric MEC-4 or MEC-4d channels. We examined a series of MEC-10/MEC-4 chimeras to identify specific domains (amino terminus, first transmembrane domain, and extracellular domain) and sites (residues 130–132 and 134–137) within MEC-10 that are required for a robust response to shear stress. In addition, the LSS response was largely abolished by MEC-10 mutations encoded by a touch-insensitive mec-10 allele, providing a correlation between the channel's responses to two different mechanical forces. Our findings suggest that MEC-10 has an important role in the channel's response to mechanical forces.
). It has been proposed that C. elegans mechanosensitive channels are quiescent at rest. Upon mechanical stimulation, channels open transiently to allow cation influx (
). However, mechanisms by which these channels alter their gating in response to a mechanical force are not well understood.
To gain insights regarding the structural features of C. elegans degenerins that are important for mechanotransduction, we used a heterologous expression system (Xenopus oocytes) to identify subunits, domains, and specific sites that are required for degenerin channels to respond to a mechanical stimulus, laminar shear stress (LSS). We found that C. elegans channels comprising MEC-4 (or MEC-4d) and MEC-10 responded to LSS with an increase in channel activity. This response required MEC-10 and was substantially diminished in channels bearing touch-insensitive mutations of MEC-10, suggesting that the LSS response in oocytes provides a surrogate system for studying worm mechanosensation in vitro. We also identified key domains and sites within MEC-10 required for a robust LSS response. These findings suggest that MEC-10 plays an essential role in modulating the channel's response to mechanical stimuli.
Experimental Procedures
Domain Swap and Site-directed Mutagenesis
MEC-4 in the pGEM-HE vector and MEC-10 in the pSGEM vector were used as templates to generate five MEC-10/4 chimeras shown in Fig. 1: 1) MEC-10-4NTM1 in which the N terminus and the first transmembrane domain (TM1) of MEC-10 (residues Met1–Tyr146) were replaced with the corresponding MEC-4 region (residues Met1–Tyr133), 2) MEC-10-4N in which the N terminus of MEC-10 (residues Met1–Ala122) was replaced with that of MEC-4 (residues Met1–Ala109), 3) MEC-10-4TM1 in which the TM1 of MEC-10 (residues Ala123–Tyr146) was replaced with MEC-4 TM1 (residues Val110–Tyr133), 4) MEC-10-4ECD in which the whole extracellular domain (ECD; residues Gln147–Tyr666) of MEC-10 was replaced with MEC-4 ECD (residues Asn134–Tyr706), and 5) MEC-10-4TM2C in which the second transmembrane domain (TM2) and C terminus of MEC-10 (Gly667–Tyr724) were replaced with the corresponding residues (Gly707–Phe768) of MEC-4. Other MEC-10 variants were generated with the QuikChange II XL site-directed mutagenesis kit (Agilent Technologies, Santa Clara, CA). Standard DNA sequencing was performed to confirm desired mutations, deletions, or domain swaps between MEC-4 and MEC-10.
FIGURE 1Schematic representation of MEC-4, MEC-10, and five chimeric constructs in which specific domains of MEC-10 were replaced with the corresponding MEC-4 domains. Key sites where MEC-4 domains were introduced into the MEC-10 construct are labeled with arrowheads and residue numbers.
cRNAs of wild type and mutant subunits of degenerin channels and of renal outer medullary K+ (ROMK) channels were synthesized with a T7 mMESSAGE mMACHINE transcription kit (Ambion, Grand Island, NY) and purified with an RNeasy MinElute Cleanup kit (Qiagen, Valencia, CA). All procedures above were conducted following established protocols. Defolliculated stage V-VI oocytes of Xenopus laevis were injected with 1 ng of MEC-6 cRNA and 5 ng of cRNAs of MEC-2, MEC-10 (or MEC10-4 chimera), and either MEC-4 or MEC-4 A713T (MEC-4d). Following microinjection, oocytes were initially maintained at 18 °C in modified Barth's saline (88 mm NaCl, 1 mm KCl, 2.4 mm NaHCO3, 15 mm HEPES, 0.3 mm Ca(NO3)2, 0.41 mm CaCl2, and 0.82 mm MgSO4, pH adjusted to 7.4) supplemented with 10 μg/ml sodium penicillin, 10 μg/ml streptomycin sulfate, and 100 μg/ml gentamicin sulfate. On the following day, surviving oocytes were transferred into fresh modified Barth's saline solution supplemented with 50 μm benzamil and incubated for an additional 2–7 days at 18 °C for optimal expression. Excess benzamil was removed by rinsing oocytes in benzamil-free modified Barth's saline for at least 15 min before initiating current recordings.
LSS Response
Oocytes were placed in a recording chamber (20-mm diameter and 6 mm deep) and perfused with NaCl-110 solution (110 mm NaCl, 2 mm KCl, 1.54 mm CaCl2,and 10 mm HEPES, pH adjusted to 7.4 with NaOH) at a rate of 3.5 ml/min. Oocytes were clamped at −60 mV while whole cell Na+ currents were continuously recorded. LSS was applied by perfusion through a vertical pipette (1.8-mm internal diameter) submerged near the top of the oocyte. The vertical perfusion rate was adjusted to 1.5 ml/min, corresponding to 0.12 dyne/cm2 shear stress (
). At the end of each experiment, benzamil (20 μm for channels with MEC-10 and 100 μm for channels with lower benzamil sensitivity) was added to the bath perfusion to determine the leak current. To assess the magnitude of the LSS response, three measurements were taken: (i) the current measured just prior to the initiation of LSS (Ibasal), (ii) the current measured 40 s following the initiation of LSS (ILSS), (iii) and the current measured after the application of benzamil. Whole cell currents were corrected for the benzamil-insensitive component when determining ILSS and Ibasal. Measurements of ILSS/Ibasal for wild type and mutant from the same batches of oocytes are reported. Time constants for channel activation (τ) were determined by fitting the first 60 s of current increases following the initiation of vertical perfusion with an exponential equation as described previously (
). To measure the effect of shear stress on ROMK channels, oocytes were injected with 2 ng of rat ROMK1 cRNA. Oocytes were bathed with a solution containing 112 mm KCl, 1.54 mm CaCl2, and 10 mm HEPES, pH adjusted to 7.4 with KOH. The LSS response was examined by monitoring the increase in whole cell K+ current in the presence of LSS. At the end of each recording, 5 mm BaCl2 was added to the bath solution to determine the ROMK-independent leak current.
Benzamil Dose-Response Relationship
Oocytes were clamped at −60 mV while whole cell currents were continuously recorded. For the first 60 s, oocytes were perfused with NaCl-110 to measure basal currents (I0) in the absence of benzamil. A benzamil stock solution (0.1 m in DMSO) was diluted to a series of concentrations (10−9, 10−8, 10−7, 10−6, 10−5, 10−4, and 10−3m) with NaCl-110 and delivered to oocytes in sequence. Oocytes were perfused with each concentration of benzamil for 30 s to obtain stabilized currents (I). Normalized whole cell currents (I/I0 values) were plotted as a function of benzamil concentrations (m), and dose-response curves were generated by fitting data to the following sigmoidal equation.
(Eq. 1)
where X is the concentration of benzamil (m). The IC50 is defined as the concentration of benzamil that inhibits 50% of the whole cell Na+ current (y). Comparisons between the dose-response curves were analyzed with the extra sum-of-squares F test.
Single Channel Recording
Vitelline membranes were manually removed from oocytes following incubation in a hypertonic solution (NaCl-110 supplemented with 200 mm sucrose). Oocytes were transferred to a recording chamber filled with a low [Cl−] bath solution (100 mm sodium gluconate, 2 mm NaCl, 2 mm CaCl2, 5 mm Na2EGTA, and 10 mm HEPES, pH adjusted to 7.4 with NaOH) and maintained for at least 5 min before patching. Pipettes with a tip resistance of 4–10 megaohms were filled with the same bath solution. Single channel experiments were performed in the cell-attached mode with the holding potential (negative value of the pipette potential) clamped at −60 mV using a PC-One patch clamp amplifier (Dagan, Minneapolis, MN) and a DigiData 1322A interface connected to a personal computer. Recordings were acquired at 5 kHz, filtered at 1 kHz with a built-in Bessel filter. In selected experiments, the recording pipette was backfilled with benzamil (50 μm) to confirm degenerin channels were expressed in oocytes injected with a cRNA mixture of MECs. Channel open probability (Po) was estimated with the single channel search function of ClampFit 10.2. Briefly, open state probability (NPo) was calculated using the following equation.
(Eq. 2)
where i indicates the number of channels open, ti is the amount of time channels spend open, and T is the total recording time. Po was then determined by dividing NPo by N, the maximal number of open channels observed in a recording. To assess channel conductance, we recorded single channel current amplitudes at holding potentials between −100 and −20 mV in 20-mV intervals.
Gentle Touch Assay
Worms were assayed for their response to gentle body touch as described previously (
). Each animal was touched by an eyelash 10 times, alternating between touches to the anterior and posterior part of the animal. The animals were scored by the number of positive responses to the 10 delivered touches. All touch assays were performed in young adult worms, and results from three independent experiments are presented.
Data and Statistical Analyses
Data were expressed as the mean ± S.D. in the main text and table. Box-and-whisker diagrams were used to show the distribution of data: median (middle line), mean (filled square), 25th to 75th percentile (box), and 10th to 90th percentile (whisker). Experiments were repeated with a minimum of two batches of oocytes obtained from different frogs. When appropriate, statistical comparisons were obtained from unpaired Student's t test or one-way ANOVA followed by a Bonferroni test. A p value of less than 0.05 was considered statistically different.
Results
The C. elegans Degenerin Channel Is Activated by Shear Stress
To examine whether C. elegans degenerin channels respond to LSS, we used a perfusion system in which vertical fluid flow is delivered via a pipette placed in close proximity to the Xenopus oocyte surface to apply shear stress (
). MEC-4 and MEC-10 were co-expressed with an accessory protein (MEC-2) and a chaperone (MEC-6) as amiloride-sensitive Na+ currents in oocytes expressing MEC-4 alone or MEC-4/MEC-10 were significantly enhanced by co-expressing MEC-2 and MEC-6 (
). We observed that LSS of 0.12 dyne/cm2 elicited a 3.1 ± 1.5-fold increase in whole cell Na+ currents in oocytes expressing MEC-4/10/2/6 (n = 18) (Fig. 2). LSS-mediated channel activation was not observed in non-injected oocytes. In addition, LSS did not evoke an increase in whole cell Na+ current when oocytes expressing MEC-4/10/2/6 were treated with benzamil prior to the onset of the LSS stimulation (Fig. 2). Furthermore, LSS did not evoke an increase in whole cell K+ current in oocytes expressing ROMK (Fig. 2). Together, our observations suggest that MEC-4/MEC-10, the ion channel that senses gentle body touch in worms, is activated by LSS in Xenopus oocytes. However, prolonged incubation following cRNA injection (∼6–8 days) was required to detect MEC-4/MEC-10 activity even when MEC-2 and MEC-6 cRNAs were co-expressed, making it difficult to study wild type channels in oocytes. When we replaced MEC-4 with a gain-of-function mutation, MEC-4 A713T (MEC-4d; Ref.
), baseline channel activity prior to the LSS stimulation was dramatically enhanced (1.1 ± 0.8 (MEC-4; n = 18) versus 9.0 ± 4.6 μA (MEC-4d; n = 15)) and readily detected 3–4 days after injection. We also found that MEC-4d/MEC-10 channels responded to LSS with a significant (2.6 ± 1.0-fold, n = 15) increase in whole cell Na+ currents (Fig. 2c). Moreover, the magnitude of the LSS response of MEC-4d/MEC-10 channel was not different from that of MEC-4/MEC-10 channel (Fig. 2c). Therefore, MEC-4d was used in subsequent experiments.
FIGURE 2C. elegans degenerin channels are activated by LSS.a, representative traces are shown for recordings obtained from Xenopus oocytes injected with cRNA mixtures encoding MEC-4/MEC-10/MEC-2/MEC-6 (4/10/2/6), MEC-4 A713T (MEC-4d)/MEC-10/MEC-2/MEC-6 (4d/10/2/6) or rat ROMK1. The baseline channel activity prior to the LSS stimulation was dramatically enhanced (1.1 ± 0.8 μA (MEC-4; n = 18) versus 9.0 ± 4.6 μA (MEC-4d; n = 15) measured 7–8 days after cRNA injection). Note that the y axis scales differ. Oocytes were exposed to flow-mediated LSS of 0.12 dyne/cm2. At the end of the experiment, 20 μm benzamil (or 5 mm BaCl2) was added to the bath to block degenerin channel- (or ROMK)-mediated currents. No benzamil-sensitive currents or LSS response was observed in non-injected oocytes. b, channel activity was blocked by 20 μm benzamil prior to the initiation of the vertical perfusion with benzamil present. No channel activation by LSS was observed in oocytes expressing 4/10/2/6 or 4d/10/2/6. c, the magnitude of channel activation by LSS was determined as a ratio between the peak response of the whole cell current following the initiation of LSS (ILSS) and the basal whole cell current prior to initiation of LSS (Ibasal) (***, p < 0.001, determined with one-way ANOVA followed by a Bonferroni test). The number of oocytes (n) assayed for each group is indicated in c. Whiskers indicate the 10th and 90th percentiles. rROMK1, rat ROMK1.
). Benzamil-sensitive Na+ currents were only occasionally detectable in the absence of MEC-4 or MEC-4d where MEC-10 is the sole pore-forming subunit (observed in only four oocytes from one of 13 batches of oocytes; see Table 1). Consistent with previous studies (
), the activity of MEC-4d channels was significantly enhanced by co-expression of MEC-2 or MEC-6 but reduced by co-expression of MEC-10 (Table 1). Therefore, we examined the contribution of MEC-10, MEC-2, or MEC-6 to the channel's LSS response. Omitting MEC-2 or MEC-6 in the channel complex had little effect on the magnitude of the LSS response (Fig. 3), indicating that both MEC-2 and MEC-6 are dispensable for this response. This is not surprising as MEC-2 and MEC-6 do not contribute to the channel's pore (
). In contrast, the LSS response was largely abolished in channels lacking MEC-10. We observed that LSS (0.12 dyne/cm2) only elicited a 1.2 ± 0.2-fold increase in whole cell current of oocytes expressing homomeric MEC-4d channels (no MEC-10; n = 15) compared with a 2.2 ± 0.5-fold increase in oocytes expressing heteromeric MEC-4d/10 channels (n = 14) (Fig. 3). In addition, the rate of channel activation was much faster in homomeric MEC-4d channels (τ = 4.0 ± 0.7 s) than that of channels with MEC-10 (τ = 8.3 ± 1.8 s), indicating that MEC-10 also slows the channel's response to LSS. These observations suggest that channels with MEC-4d as the sole pore-forming subunit exhibit only modest sensitivity to LSS, whereas both MEC-10 and MEC-4d are required for a robust channel response to LSS.
TABLE 1Subunit composition affects C. elegans degenerin channel activity
FIGURE 3MEC-10 is required for the full sensitivity of the LSS response.a, representative traces of the LSS response recorded in oocytes injected with different combinations of mec cRNAs as noted. Vertical perfusion (1.5 ml/min) was initiated at t = 30 s and maintained for 120 s to apply LSS. Benzamil was added to the bath at the end of the experiment. Note that the y axis scales differ. b, summary of the magnitude (ILSS/Ibasal; top) and time constants of channel activation by LSS (τ; bottom) for each group. Whiskers indicate the 10th and 90th percentiles. The number of oocytes (n) assayed for each group is indicated in b. Statistical comparisons were made between channels lacking one subunit versus the 4d/10/2/6 channel (***, p < 0.001, determined with one-way ANOVA followed by a Bonferroni test).
We observed a larger benzamil-sensitive Na+ current in oocytes injected with MEC-4d, MEC-2, and MEC-6 (5.7 ± 3.4 μA, n = 73) than those injected with MEC-4d, MEC-10, MEC-2, and MEC-6 (2.9 ± 2.5 μA, n = 78, p < 0.001 (Table 1)). Previous studies suggest that LSS activates ENaC by increasing channel Po (
). The blunted LSS response and a higher channel activity raised the possibility that homomeric MEC-4d channels reside at a high Po state, which would limit any increase in Po due to LSS. To test whether the lack of LSS activation seen with these channels is due to a high baseline Po, we performed the single channel recording with a cell-attached configuration (Fig. 4). Single channels in oocytes injected with cRNA mixtures of MEC-4d, MEC-2, and MEC-6 (4d/2/6) or MEC-4d, MEC-10, MEC-2, and MEC-6 (4d/10/2/6) were inhibited by benzamil (50 μm) backfilled in the recording pipette (Fig. 5), suggesting that these currents were mediated by degenerin channels. The average Po of homomeric MEC-4d channels (0.33 ± 0.20, n = 14) was not significantly different from that of heteromeric MEC-4d/MEC-10 channels (0.23 ± 0.10, n = 17) (Fig. 4). Additionally, the Po value of MEC-4d channels was far below 1, indicating that the loss of the LSS response in MEC-4d homomeric channels is not due to a high Po channel under basal conditions. The absence of MEC-10 did not affect the single channel conductance (Fig. 4, bottom right). These results suggest that MEC-10 is required for LSS-induced conformational changes that lead to channel activation.
FIGURE 4Single channel properties of 4d/2/6 and 4d/10/2/6 channels.Top, representative traces of single channel recordings performed in the cell-attached configuration with oocytes clamped at −60 mV (holding potential). C, closed state; O, O1, and O2, open states. Recordings were filtered at 100 Hz with a low pass Gaussian algorithm of ClampFit 10.2 for display. Subconductance states were noted as described previously (
). Bottom left, Po was estimated using the single channel analysis function of ClampFit 10.2. Only recordings containing three or fewer open levels were included in the calculation. The average recording length was 8.1 ± 3.3 min. The results are from 14 patches from oocytes expressing 4d/2/6 and 17 patches from oocytes expressing 4d/10/2/6 (NS, not significant). Oocytes were patched 3–5 days after cRNA injection. Bottom right, unitary I/V relationship. Single channel Na+ conductance estimated from linear regression analysis of currents measured at a holding potential between −20 and −100 mV was 26.9 ± 5.4 picosiemens for 4d/2/6 (n = 9) and 26.9 ± 5.4 picosiemens for 4d/10/2/6 (n = 8). Error bars represent S.D. in the bottom figures. The averaged Po or single channel conductance of 4d/2/6 is not different from that of 4d/10/2/6 as determined by unpaired Student's t test. S, seconds.
FIGURE 5MEC-4d or MEC-4d/MEC-10 channels are inhibited by benzamil. Single channel recordings were performed in the cell-attached configuration with oocytes clamped at −60 mV (holding potential). Representative traces with expanded views are shown for the control group (a and e) and the inhibitory effect of benzamil where recording pipettes were backfilled with 50 μm benzamil (b and f). C, closed state. The average recording duration was 14 (a) or 8 min (e) for control groups and 19 (b) or 13 min (f) for benzamil backfill groups. Po (c and g) and NPo (d and h) values were estimated from the start (1) and end (2) of recordings as described under “Experimental Procedures.” The significance of the benzamil effect was determined by unpaired Student's t test (**, p < 0.01; ns, not significant).
). tm1552 and ok1104 worms not only exhibit a partial loss of the gentle touch response but also are compromised in their harsh touch response, suggesting that MEC-10 is a key component of the ion channel transducing mechanical signals in C. elegans (
). Both tm1552 and ok1104 carry large deletions in the mec-10 gene, including the coding region (Fig. 6a). However, it has not been established whether these null alleles truly encode loss-of-function mutations. We isolated total RNA from tm1552 worms as well as control worms (N2) and performed RT-PCR to analyze the transcription products of tm1552. As shown in Fig. 6a, RT-PCR yielded two major bands in tm1552 worms compared with the single larger band in N2 worms. Sequencing the two tm1552 transcripts revealed that the large transcript introduced a premature stop codon after Ile189. The small but more abundant transcript encoded a MEC-10 protein with an in-frame deletion of 150 residues (Ala162–Lys311) from a peripheral part of the extracellular region that is poorly conserved among members of the ENaC/degenerin family, referred to as the finger domain (
Constraint-based, homology model of the extracellular domain of the epithelial Na+ channel α subunit reveals a mechanism of channel activation by proteases.
). We generated both gene products, MEC-10 L190X and MEC-10 Δ162–311, and examined the LSS response of channels containing these MEC-10 mutants. We found that the LSS response was largely blunted with either mutant when compared with wild type MEC-10 (Fig. 6c). The tm1552 worms also exhibited a blunted response to light touch (Fig. 6b). These findings suggest that tm1552 is indeed a mec-10 null allele and that MEC-10 is required for both touch sensation in worms and channel activation by LSS in oocytes.
FIGURE 6Putative mec-10 null allele, tm1552, is impaired in its response to LSS and to gentle body touch.a, exon-intron map of mec-10. The tm1552 allele carries a deletion of exon 5, intron 5, and part of exon 6 of mec-10. Protein products were predicted from the two RT-PCR products of the tm1552 allele (shown at the right). N2 worms express wild type mec-10. The sites of deletion and new amino acid residues are labeled in red and green, respectively. b, the gentle touch response in wild type N2 worms (n = 60), mec-4-null (u253; n = 61), and mec-10-null (tm1552; n = 64) worms was examined by stroking the anterior and posterior part of the worm body 10 times alternately with an eyelash. Loss of either MEC-4 (in u253) or MEC-10 (in tm1552) expression caused a reduction in the sensitivity to gentle body touch (***, p < 0.001, determined by one-way ANOVA followed by a Bonferroni test). c, we generated two tm1552 mutants, MEC-10 L190X and MEC-10 Δ162–311. The LSS response was assessed in oocytes expressing MEC4d/MEC2/MEC6 and either wild type (WT) MEC-10, a MEC-10 mutant, or no MEC-10 (homomeric MEC-4d channel). LSS-mediated channel activation was abolished by both mutants (reduced ILSS/Ibasal values compared with wild type MEC-10; ***, p < 0.001, determined by one-way ANOVA followed by a Bonferroni test). Significant changes when compared with homomeric MEC-4d channels are noted with ††† (p < 0.001). The number of worms or oocytes (n) assayed for each group is indicated in b and c, respectively. Whiskers indicate the 10th and 90th percentiles. d, representative traces of the LSS response of oocytes expressing MEC4d/MEC2/MEC6 with either wild type MEC-10, MEC-10 L190X, MEC-10 Δ162–311, or no MEC-10 (homomeric MEC-4d channel). Note that the y axis scales differ.
). If a specific region of MEC-10 is required for a robust channel response to LSS, replacing it with its corresponding MEC-4 region should blunt the channel's response to LSS. We generated a series of chimeras in which (i) the N terminus and TM1 of MEC-10 were replaced with the corresponding regions of MEC-4 (MEC-10-4NTM1), (ii) the ECD of MEC-10 was replaced with the MEC-4 ECD (MEC-10-4ECD), and (iii) TM2 and the C terminus of MEC-10 were replaced with the corresponding regions of MEC-4 (MEC-10-4TM2C) (Fig. 1). The LSS response was assessed for each individual chimeric construct (MEC-10-4NTM1, MEC-10-4ECD, or MEC-10-4TM2C) in oocytes co-expressing MEC-4d, MEC-2, and MEC-6. The magnitude of channel activation by LSS was largely blunted in channels containing MEC-10-4NTM1 (24 ± 17% increase in current, n = 11), similar to that observed in channels lacking MEC-10 (26 ± 19% increase in current, n = 10 (Fig. 7)). Channels with MEC-10-4ECD had a modestly blunted LSS response compared with channels with wild type MEC-10 (Fig. 7b). The rate of LSS-induced current increase was more rapid in oocytes expressing channels containing MEC-10-4NTM1 or MEC-10-4TM2C when compared with oocytes expressing channels containing wild type MEC-10 (Fig. 7b). We generated additional chimeras where either the N terminus or TM1 of MEC-10 was replaced with the corresponding region of MEC-4 to further define key regions within MEC-10 that are required for a robust LSS response. The LSS response was largely abolished in channels lacking the MEC-10 N terminus (MEC-10-4N) or TM1 (MEC-10-4TM1) (Fig. 7b). In addition, both chimeras increased the rate of channel activation (Fig. 7b). Our results suggest that both the N terminus and TM1 of MEC-10 are required for a robust channel response to LSS.
FIGURE 7The cytoplasmic N terminus and TM1 of MEC-10 are required for a robust LSS response.a, representative traces of the LSS response of oocytes expressing MEC4d/MEC2/MEC6 with either WT MEC-10, individual MEC-10 chimeras (10-4N, 10-4TM1, and 10-4NTM1), or no MEC-10 (homomeric MEC-4d channel). Note that the y axis scales differ. b, the magnitude (ILSS/I basal; top) and kinetics (τ; bottom) of the LSS response of each group was assessed as described above. Whiskers indicate the 10th and 90th percentiles. The number of oocytes (n) assayed for each group is indicated in b. Statistical analyses were performed with one-way ANOVA followed by a Bonferroni test. Significant changes when compared with wild type MEC-10 are noted as follows: ***, p < 0.001; **, p < 0.01; *, p < 0.05. Significant changes when compared with homomeric MEC-4d channels are noted as follows: †††, p < 0.001; ††, p < 0.01.
As three MEC-10 chimeras (MEC-10-4NTM1, MEC-10-4N, and MEC-10-4TM1) reduced the magnitude of channel activation by LSS to the levels observed in the absence of MEC-10 (Fig. 7), it was important to confirm that these MEC-10 chimeras were components of the channel complex. We examined the effect of an amiloride binding site mutant of MEC-10 (G680E) on the efficacy of channel inhibition by benzamil. Mutations at the equivalent site of ENaC subunits dramatically reduced the channel's sensitivity to amiloride and other pore blockers (
Identification of amino acid residues in the α, β, and γ subunits of the epithelial sodium channel (ENaC) involved in amiloride block and ion permeation.
). Channels containing MEC-10 G680E exhibited a modest but significant rightward shift in the benzamil dose-response curve when compared with channels with wild type MEC-10 (Fig. 8). A similar rightward shift in the benzamil dose-response curve was also observed with the chimeric MEC-10 subunits bearing the G680E mutation (Fig. 8), suggesting that these chimeric subunits were components of the channel complex.
FIGURE 8G680E mutations alter the channel's sensitivity to benzamil. Benzamil dose-response curves for channels comprising MEC4d/MEC2/MEC6 and either wild type MEC-10, individual chimeric MEC-10 (10-4N, 10-4TM1, and 10-4NTM1), or each MEC-10 construct with the G680E mutation are shown. Whole cell Na+ currents were measured while oocytes were sequentially exposed to increasing concentrations of benzamil (10−9, 10−8, 10−7, 10−6, 10−5, and 10−4m). Currents measured in the presence of each benzamil concentration (I) were normalized to the basal current (I0) to estimate relative currents (I/I0). The number of oocytes (n) assayed for each group is indicated. Each data point represents the mean from five to 12 measurements. Error bars represent S.D. Data were fit to the dose-response equation described under “Experimental Procedures.” The rightward shift of each dose-response curve with the G680E mutant was statistically significant (comparisons between the curves in each panel were performed with the extra sum-of-squares F test; p < 0.001).
To further define key sites in MEC-10 required for an LSS response, we focused on TM1 as it is well conserved between MEC-4 and MEC-10. The sequence alignment showed a major region of divergence between MEC-4 TM1 and MEC-10 TM1 within residues 130–137 in MEC-10 (Fig. 9a). Two mutants were generated in which MEC-10 residues at 130–132 or 134–137 were replaced by their corresponding MEC-4 residues. We observed that channels containing either of these MEC-10 mutants exhibited a dramatically reduced response to LSS (Fig. 9), suggesting that MEC-10 residues at these sites (130–132 and 134–137) are required for a robust response to LSS.
FIGURE 9Key sites within MEC-10 TM1 affect the channel's LSS response.a, sequence alignment of the TM1 domains of MEC-4 and MEC-10. Identical residues are labeled as red in a yellow background; homologous residues are highlighted in green. Sites of two MEC-10 mutants (MEC-10-4(130-132) and MEC-10-4(134-137)) are noted. b, representative traces of the LSS response in oocytes expressing MEC4d/MEC2/MEC6 and either wild type MEC-10, an individual MEC-10 mutant, or no MEC-10 (homomeric MEC-4d channel). Benzamil was added to the bath at the end of the experiment. Note that the y axis scales differ. c, the magnitude (ILSS/Ibasal; left) and kinetics (τ; right) of the LSS response is summarized. Whiskers indicate the 10th and 90th percentiles. The number of oocytes (n) assayed for each group is indicated in c. Significant changes when compared with wild type MEC-10 are noted as follows: ***, p < 0.001; **, p < 0.01 (determined by one-way ANOVA followed by a Bonferroni test). Significant changes when compared with homomeric MEC-4d channels (no MEC-10) are noted as follows: †††, p < 0.001; †, p < 0.05 (determined by one-way ANOVA followed by a Bonferroni test).
In the above experiments (FIGURE 7, FIGURE 9), we applied domain swap and site-directed mutagenesis to screen key region and sites within MEC-10 that are required for a robust LSS response in the context of MEC-4d, a gain-of-function mutation within the pore-lining region of the worm degenerin channel (
). We performed additional studies to confirm our key findings in the context of wild type MEC-4. In agreement with studies using MEC-4d, we found that MEC-10 was required for a robust LSS response when we co-expressed wild type MEC-4, MEC-2, and MEC-6 (Fig. 10). Consistent with our findings with the MEC-4d channel, the LSS response of wild type MEC-4 channel was largely blunted by co-expression of one of four MEC-10 chimeras, 10-4TM1, 10-4(130–137), 10-4(130–132), and 10-4(134–137) (Fig. 10). In addition, replacing the TM1 or key sites within MEC-10 TM1 with the corresponding MEC-4 resides enhanced the rate of channel activation by LSS (Fig. 10). Mutating both TM1 sites with MEC-10 (residues 130–132 and 134–137) did not have an additive effect on the LSS response (Fig. 10).
FIGURE 10Key sites within MEC-10 contribute to the LSS response of channels containing MEC-4.a, representative traces of the LSS response in oocytes expressing MEC4, MEC2, MEC6, and either wild type MEC-10, individual MEC-10 chimeras, or no MEC-10 (homomeric MEC-4 channel). Benzamil was added to the bath at the end of the recording. Note that the y axis scales differ. The magnitude (ILSS/Ibasal; b) and kinetics (τ; c) of the LSS response are summarized. Whiskers indicate the 10th and 90th percentiles. The number of oocytes (n) assayed for each group is indicated in b and c. Significant changes when compared with channels containing wild type MEC-10 are noted as follows: ***, p < 0.001; **, p < 0.01 (determined by one-way ANOVA followed by a Bonferroni test). Significant changes when compared with homomeric MEC-4 channels (no MEC-10) are noted as follows: †††, p < 0.001 (determined by one-way ANOVA followed by a Bonferroni test).
). This fast adapting transient current is sensitive to body indentation as well as the velocity and frequency of stimulus but not the amount of the applied force (
). We found that LSS elicited a robust activation of MEC-4/MEC-10 channels (2–3-fold increases in whole cell Na+ currents) expressed in Xenopus oocytes (Fig. 2). LSS-mediated channel activation is specific to members of the ENaC/degenerin family as LSS had no effect on non-injected oocytes or on oocytes expressing ROMK (Fig. 2). The response of MEC-4/MEC-10 channels to LSS allowed us to identify a specific subunit (MEC-10) as well as sites within this subunit that are required for the response of C. elegans degenerin channels to LSS in Xenopus oocytes.
MEC-4 and MEC-10 share similar structures but perform distinct roles in the nematode gentle touch response. Rather than being functionally redundant, the mec-4 and mec-10 alleles complement each other (
). No mechanosensitive currents were detected in mec-4 null animals, suggesting that MEC-10 alone is not sufficient for assembling a functional mechanosensory channel (
). However, the co-expression of MEC-10 with MEC-4 (or MEC-4d) in oocytes resulted in readily detectable Na+ currents as well as a robust LSS response (FIGURE 2, FIGURE 3). In the absence of MEC-10, the response to LSS was largely blunted (Fig. 3), although the homomeric MEC-4d channels had greater currents than MEC-4d/MEC-10 channels prior to the initiation of LSS (Table 1). Furthermore, MEC-10 affected the kinetics of the LSS response as the rate of LSS-induced channel activation was slower when MEC-10 was co-expressed with MEC-4.
The markedly blunted LSS response in channels lacking MEC-10 is not due to a high Po state at baseline as the Po of homomeric MEC-4d channels was similar to that of MEC-4d/MEC-10 channels (Fig. 4). Consistent with the requirement of MEC-10 for a robust LSS response in oocytes, we found that MEC-10 touch-insensitive deletion mutants also exhibited a blunted LSS response (Fig. 6). Our results suggest that MEC-10 is required for degenerin channels to exhibit a robust LSS response.
Other members of the ENaC/degenerin family are mechanosensitive channels. For example, ENaC is activated by shear stress in native kidney tubules, endothelial cells, and heterogeneous expression systems (
). We have proposed a model where LSS-induced motions within the extracellular region of ENaC lead to movement of the transmembrane helices where the channel gate resides, transitioning the channel to a higher Po state (
). As there is abundant structural conservation among members of the ENaC/degenerin family, we hypothesize that mechanisms by which mechanical forces regulate channel activity are evolutionally conserved in this family. Our domain swap studies indicate that the large ECD of MEC-10 has a role in conferring sensitivity to LSS (Fig. 7). The finger domains within the ECD of MEC-4 and MEC-10 are noticeably longer than that of their mammalian orthologues, including blocks of cysteine-rich sequences at the start and end of the finger domains, and a 22-amino acid degenerin-unique sequence has been proposed to influence channel gating (
). It has also been proposed that the finger domains are responsible, in part, for the diversity of extracellular factors that regulate members in this family (
). It will be important to examine the roles of nematode-specific finger domain motifs in the channel's response to LSS in future studies.
Interestingly, the N terminus and TM1 of MEC-10 have important roles in evoking a robust response to LSS in heteromeric MEC-4d/MEC-10 channels (Fig. 7). Furthermore, within the MEC-10 TM1, we identified two sites (residues 130–132 and residues 134–137) that are required for MEC-4d/MEC-10 channels to exhibit a robust LSS response (FIGURE 9, FIGURE 10). The channel gate of the ENaC/degenerin family is thought to reside within the transmembrane domains (specifically in the vicinity of the degenerin site in TM2). Structural studies of a related acid-sensing ion channel suggest that rotational movements of the transmembrane domains occur in conjunction with the transition from a closed to an open state (
). Consistent with this model, we found that a transmembrane region (TM1) of MEC-10 and specific sites within MEC-10 TM1 have important roles in the response of MEC-4d/MEC-10 channels to LSS (Figs. 7, 9, and 10).
Previous studies from our group have suggested that TM2 and adjacent residues have important roles in the response of ENaC to LSS as both the magnitude and kinetics of the LSS response were altered by substitutions introduced within this region (
). Interestingly, the magnitude of the LSS response of the MEC-4d/MEC-10 channel was not altered by replacing the TM2 and C terminus of MEC-10 with the corresponding region of MEC-4 (MEC-10-4TM2C chimera) (Fig. 7). This result is not surprising as the pore-forming TM2 region is highly conserved between MEC-4 and MEC-10 (26 of 36 residues are identical). It is possible that mutations of specific residues within the TM2 region of MEC-10 will alter the response of the MEC-4/MEC-10 channel to LSS.
MEC-2 and MEC-6 are both required for expression of mechanoreceptor currents and the gentle touch response in C. elegans as these were abolished by null mutations of either gene (
). In contrast, omitting MEC-2 or MEC-6 did not affect the channel's LSS response in oocytes (Fig. 3). It is possible that endogenous Xenopus proteins compensate for the missing components of the channel complex. Alternatively, these accessory proteins are not required for the LSS-induced motions that lead to channel activation. MEC-2 belongs to the stomatin family. Mammalian homologues of MEC-2 have been implicated in transducing mechanical stimuli in vertebrates (
). Although MEC-6 and MEC-2 might alter channel activity by modifying the local lipid environment, our results suggest that it is not required for LSS-mediated channel activation. Similar to our observations, previous work suggested that the LSS-mediated ENaC activation was not influenced by deleting cholesterol (
). Furthermore, MEC-6 has been suggested to function as an endoplasmic reticulum-resident chaperone that facilitates degenerin channel folding, assembly, and maturation (
In summary, the C. elegans channel comprising MEC-4 and MEC-10 is activated by LSS in Xenopus oocytes. MEC-10 is required to elicit a robust response to LSS. Furthermore, key domains (N terminus and TM1) and TM1 residues within MEC-10 are required for this response. Given the abundant structural conservation among members of the ENaC/degenerin family, an understanding of C. elegans degenerin mechanosensation will facilitate studies focused on the regulation of ENaCs and other ENaC/degenerin family members by mechanical forces.
Author Contributions
S. S. and T. R. K. designed research, analyzed data, and prepared the manuscript. S. S. performed experiments and collected the data. C. J. L., M. T. M., and G. A. S. provided C. elegans strains and technical assistance. All authors discussed and approved the manuscript.
Acknowledgments
The original MEC-4, MEC-10, MEC-2, and MEC-6 plasmids and the bacterial strain SMC4 were kindly provided by Dr. Monica Driscoll and Dr. Laura Bianchi. We deeply appreciate their generosity.
References
Syntichaki P.
Tavernarakis N.
Genetic models of mechanotransduction: the nematode Caenorhabditis elegans.
Constraint-based, homology model of the extracellular domain of the epithelial Na+ channel α subunit reveals a mechanism of channel activation by proteases.
Identification of amino acid residues in the α, β, and γ subunits of the epithelial sodium channel (ENaC) involved in amiloride block and ion permeation.
36749-1&id=gr1.jpg){kind=link}
36749-1&id=gr2.jpg){kind=link}
36749-1&id=gr3.jpg){kind=link}
36749-1&id=gr4.jpg){kind=link}
36749-1&id=gr5.jpg){kind=link}
36749-1&id=gr6.jpg){kind=link}
36749-1&id=gr7.jpg){kind=link}
36749-1&id=gr8.jpg){kind=link}
36749-1&id=gr9.jpg){kind=link}
36749-1&id=gr10.jpg){kind=link}