© 1999 American Heart Association, Inc.
Original Contribution |
Analysis of the Presence and Physiological Relevance of Subconducting States of Connexin43-Derived Gap Junction Channels in Cultured Human Corporal Vascular Smooth Muscle Cells
From the Departments of Urology and Physiology & Biophysics (G.J.C.), Albert Einstein College of Medicine, Bronx, NY; Department of Physiology & Biophysics (P.R.B.), State University of New York at Stony Brook, Stony Brook, NY.
Correspondence to George J. Christ, PhD, Laboratory of Molecular & Integrative Urology, Room 744, Forchheimer Bldg, Albert Einstein College of Medicine, 1300 Morris Park Ave, Bronx, NY 10461. E-mail christ{at}aecom.yu.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Abstract—Subconductance states are a commonly observed feature of gap junction channels. Their overt frequency and consistent appearance in both single and multichannel records have led to speculation that they might be of physiological importance in terms of altering the rate of small solute transfer from cell to cell. Among the connexin gene family, connexin43 (Cx43) is the most ubiquitous connexin that has been shown to generate subconductive states. Therefore, it was the explicit aim of this investigation to more fully evaluate the potential contribution of human Cx43-derived subconducting states to intercellular communication in cultured human corporal vascular smooth muscle cells. To determine the weight of subconductive states in our records, we analyzed amplitude histograms of multichannel and single-channel data during the application of transjunctional voltages larger than expected for physiological conditions but still smaller than transjunctional voltages known to induce lower conductive states (Vo>Vj). The data clearly indicated that the subconducting states occupy only a small fraction of the total channel open time. This was reflected by the fact that the average open probability for the subconductive state(s) determined from the 9 records analyzed was
2%.
Closer inspection of the data revealed that the frequency of
subconductive states was actually higher than the frequency of the main
state conductance. In summary, recording conditions sufficient
for evaluation of the intrinsic gating characteristics of human
Cx43-derived gap junction channels have been used. Under these
conditions, our data clearly indicate that despite their greater
frequency, the duration of subconductance events is so short relative
to the main state duration as to render them physiologically
insignificant.
Key Words: connexin • vascular smooth muscle cell • gap junction
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Gap junction channels have a variety of properties that make them atypical ion channels. Chief among these are their long open time and correspondingly high open probability, as well as poor selectivity characteristics.1 Connexin43 (Cx43), the dominant gap junction protein in vascular smooth muscle, is no exception. Homotypic Cx43 channels show symmetric voltage dependence and are somewhat selective for cations over anions.2 Most connexins, including Cx43, have also been shown to allow the passage of small molecular weight probes (300 to 1200 Da1 3 ). Additionally, there is strong evidence that gap junction channels allow the diffusion of intracellular second messenger molecules such as Ca2+, cAMP, cGMP, and inositol 1,4,5-trisphosphate.4 5 These characteristics make the connexins important intercellular second messenger/ion pathways in many tissues. For example, it has already been demonstrated in vascular smooth muscle cells that Ca2+ can readily diffuse from cell to cell via channels composed of Cx43.5
In addition to their aforementioned permeability characteristics, the gating behavior of gap junction channels is also of interest. Cx43 has been well studied in this regard because it is one of the most ubiquitous connexins. Under steady-state conditions, estimates of mean open time (MOT) and mean closed time (MCT) have been made. A quantitative analysis of human Cx43 (hCx43) has yielded an MOT range of 0.43 to 5.25 s and an MCT range of 0.51 to 0.95 s.6 7 8 The study of hCx43 allowed determination of the open probability (Po), which was found to range from 0.5 to 0.95.6 This indicates that hCx43 remains open most of the time. Extrapolation of MOT and MCT to predicted values at Vj=0 mV yields a Po equal to 0.81.
All channels display, to varying degrees, subconducting states or sublevels.9 10 Gap junctions are also no exception in this regard. Cx43, for example, can display a main state of 90 to 100 pS with a sublevel of 30 pS in 150 mmol/L of salt.11 A subconducting state of 60 pS has also been reported.12 These different conducting levels (sublevels) have been correlated with phosphorylation levels for Cx43.13
The occurrence of subconductance states in gap junction channels raises
two immediate questions. First, are the selectivity characteristics of
the substates the same as or different from the main state? Second, do
the substates play a role in the voltage-dependent processes of gap
junctions? Recent studies11 showed that one subconductance
state for Cx43 was more restrictive than the main state by 50%, but
the data also indicated that, as documented for the main state, the
subconducting state was still permeable to both cations and anions.
Further, it has been argued that the substates of Cx43 are less voltage
dependent than the main state and thus comprise a large portion of the
steady-state conductance observed with large Vj
gradients.14 For hCx37 this is clearly the
case.15
A third question arises as well. Is the frequency or dwell time of the subconducting states sufficient to be of significance with regard to affecting junctional conductance or permeation of intercellular messengers when Vj is in the voltage-independent portion of the Gj/Vj relationship (ie, Vj steps<Vo)? This last case questions whether the substate conductances occur with sufficient frequency and duration to affect intercellular communication under in situ conditions, and this issue was addressed by the present study. To this end, we studied Cx43-derived gap junction channel activity in the well-characterized cultured human corporal smooth muscle cells.5 6 12
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Records analyzed in the present study arose from the database published in Brink et al.6 The data were generated by using dual whole-cell patch-clamp techniques as described elsewhere in detail.6 Briefly, for all records analyzed, the bathing solution was composed of (in mmol/L) CsCl 165, TEACl 30, CoCl2 1, NiCl2 1, MgCl2 1, CaCl2 2, HEPES 10, aminopyridine 1, TTX 0.01, and ZnCl2 0.5 (pH 7.0). The pipette solution was the same except for the following: 0.1 mmol/L CaCl2 was used, 0.6 mmol/L EGTA was added, and there was no ZnCl2; the pH was 6.8. In 2 experiments, CsCl was replaced by NaCl with no detectable difference in channel activity. That is, qualitatively similar substate events were present in both the cesium and sodium solutions. Multichannel records were analyzed using the procedures of Ramanan and Brink16 and Ramanan et al.17 Amplitude histograms were fit using fit parameters outlined in Ramanan and Brink16 and Veenstra et al.18 For experiments (n=3) designed to examine Cx43 substates under conditions previously documented to promote their formation,13 we used 2 mmol/L of the membrane permeant cAMP analogue 8-bromo-cAMP (Sigma). Records were obtained after 5 minutes and 30 minutes of exposure. All data were low-pass filtered at 5 kHz and stored on videotape via Neurocorder. To digitize the data, they were played back and low-pass filtered at 0.5 kHz and acquired through a 16-bit A/D converter into a Next computer. The sampling rate was 0.180 ms.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The occurrence of substates in gap junction channels is well documented, and the current tracings provided in Figures 1
and 2
highlight this fact. Figure 1 shows
representative records from both cells of a pair,
clearly indicating, by virtue of the presence of paired junction
current signals of opposite polarity, the junctional origin of the
substates. More specifically, the recordings in Figure 1
illustrate the equal but opposite rule for gap junction channels. These
records show that in addition to main state transitions, substate
transitions are observable in both records, removing any doubt that
the lower conductance transitions represent another population
of channels. Subsequently, the records from the cell held at zero
or near zero potential are shown. Under the ionic conditions used, the
equilibrium potential for all nonjunctional channels is zero, with the
exception of calcium; therefore, no contaminating nonjunctional channel
activity can occur. In Figure 2, the top and middle panels show
typical transjunctional current recordings, which
show channel closures and subconducting states for hCx43. The top panel
shows the subconductance state in more detail. The duration of the
substate is quite short relative to the closure of the main state and
is very short relative to the duration of the open state. Rarely are
subconducting states observed arising from the ground state to the
subconducting level for Cx43 gap junction channels. More typical is the
incomplete closure or lower conductive state shown in Figure 2. The transjunctional voltage (Vj)
in these records was 40 mV.
|
|
Figure 3 is an all-points histogram of a
larger portion of the records shown in Figure 2
. The
histogram constitutes a total of 300 seconds of data. As illustrated,
several distinct fitting parameters were used. Clearly, the
record is best interpreted as a single channel with no substate.
The unitary conductance used in this fit was 105 pS for the main state,
and on the basis of the transitions shown in Figure 2, a
subconducting state of 65 pS was presumed. Five distinct cases are
shown in Figure 3. In Figure 3A, a single 105-pS channel
is assumed with no weighting for a substate. Po
for the main state is 85%. In Figure 3B, a single 105-pS
channel with a 65-pS substate is assumed where Po
for the main state is 0.85 and the substate is 0.03. A small decline in
the peak of the closed state fit is the result. In Figure 3C, the Po for the main state was dropped to 0.80,
and a subconductance of 65 pS was given a Po of
0.06. The closed peak fit is only slightly decreased, but a significant
increase in the height between the open and closed states is
observable. In Figure 3D, Po for the main
state was further decreased to 0.725 to allow a substate weighting of
0.125 or 12.5%. The closed peak fit remains close to the fit shown in
Figure 3A, but a notable peak or increase in area exists between
the open and closed peaks, which represents the subconductance
weight. The dashed line fit of Figure 3A appears to fit the data
the best. In this case, no subconductive state was needed.
|
In summary, despite the documented presence of the subconductance
states, which generally ranged from 30 to 65 pS, the weighted
amplitude histogram shows little evidence for their contribution to the
overall open time of the channel. The reason for the apparent
discrepancy between the recordings (Figures 1 and 2) documenting that the substate exists, and the fits in the
amplitude histogram, which cannot account for the presence of the
substate, is related to the fact that the despite their frequency,
their duration renders them insignificant in the weighting of the
amplitude histogram.
In light of these considerations, we attempted fitting the data under
yet another condition. In Figure 3E, the dashed line
represents the predicted amplitude histogram if 2 independent
channels, one of 105 pS and one of 65 pS, are assumed to be in the
patch. As illustrated, the data are inconsistent with the 65-pS
channel representing an independent channel, or the
histogram would have the form of that shown by the dashed line in
Figure 3E.15
Figure 4 shows another example of the
data in which the amplitude histogram clearly shows subconductive
states. In this case, the data were taken from a 50-s record. To
fit this data set, 2 different Po values had to
be used for each of the 2 channels in the patch. This is not an
uncommon feature for multichannel gap junction
records.17 18 To fit the amplitude histogram, a
subconductive state weighting of 0.06 was required for both channels
(dashed line). In these records, the unitary conductance of the
main state was 115 pS and that of the subconductive state was 45 pS.
The entire recording is shown in Figure 5. The middle panel shows the whole
recording and the top and bottom panels show
representative main state openings and closures, as
well as transitions to the subconductive state.
|
|
All the histogramic data are summarized in the
Table (n=9 experiments). In 5 of these
experiments, no weighting for subconductive states was necessary. In 4
of the data sets, weighting for subconductive states was necessary to
fit the data. The average weighting for those 4 data sets was 0.042 or
4.2%. Taking all the records into account yielded an average of
0.02 or 2% weighting for the subconductive state(s). To further
examine the potential contribution of substates to intercellular
communication, we evaluated the effects of phosphorylating treatments
known to promote the formation of the subconducting
states13 on our junctional recordings. Thus, the
effects of 2 mmol/L 8-bromo-cAMP were tested in 2 additional
experiments (see Materials and Methods). There was no detectable change
in the weight of the subconductive states in the amplitude histogram
(data not shown).
|
Previous studies of human corporal smooth muscle cells have noted the
occurrence of various conductive states or substates and used
transition or event histograms to illustrate the various conductive
states observed but without consideration of dwell time. Figure 6 shows an event histogram taken from No.
4 of 9 experiments summarized in the Table; note that
Vj=-40 mV in all 4 cases and also that there is
no way to determine whether the subconductive states are symmetrically
voltage dependent as is the case for the main state of Cx43 gap
junction channels.6 19 Therefore, we analyzed
event histograms for the same voltage amplitude and polarity. Three
bins were used: 100 to 135 pS, 60 to 99 pS, and 20 to 59 pS. Note that
the most abundant conductance range in the event histogram is between
60 to 99 pS. Events or transitions are determined as any transition
from closed to open or vice versa.12 The data illustrate
that the occurrence of substates on the basis of transitions or number
of events is not useful in determining the weight of those events in
terms of open time and hence contribution to the movement of solutes
from one cell to another.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Recent studies have unequivocally documented that intercellular communication through Cx43-derived gap junction channels among smooth muscle cells plays a proximal role in the coordination of responses in physiologically diverse vascular tissues. Still at issue however is the precise manner, if any, in which the permeability/permselectivity of Cx43-derived gap junctions might contribute to the modulation of vascular smooth muscle responses. In this regard, one well-documented feature of hCx43 gap junctional channels is the consistent presence and high frequency of appearance of subconducting states of the main state. As the conductance of these states can be as little as 35% of the main state conductance, their presence has frequently been tacitly assumed to represent a priori evidence for a potentially important mechanism for regulating/altering junctional permeability/selectivity. However, the relative magnitude of their contribution to junctional conductance has never been rigorously evaluated in human vascular smooth muscle. It was the explicit aim of the present study, therefore, to evaluate the nature and extent of their contribution to junctional conductance in single and multichannel recordings of cultured human corporal vascular smooth muscle cells.
The data shown illustrate that the gating of the hCx43 subconductance
states is such that the number of (partial) transitions per unit of
time can be equal to or exceed the number of complete open-close
transitions. Nonetheless, the observed substates constitute a very
small fraction of the open time for the hCx43 gap junction channel when
transjunctional voltage is less than the Vo of
Cx43, 80 mV. This phenomenon is best illustrated by the amplitude
histograms, which are a simple and effective method for determination
of the total open time for the main open state, as well as any
detectable subconductance states. As illustrated, the data indicate a
maximum weight of the open peak(s) of 6% for the observed substates
(Figures 3 and 4; Table). However, more often than
not (Table), the subconductance states constitute an even
smaller percentage of the open portion of the amplitude histogram.
In light of these results, it seems relevant to ask the following question: On the basis of the current in vitro data, is there any role for gap junction channel subconductive states in the functioning of multicellular tissues that are reliant on intercellular communication for coordination of their respective responses? In short, the current data argue against such a possibility. Specifically, it is not a question of whether or not the biophysical characteristics of the substates themselves would confer any degree of permselectivity but rather an issue of whether or not the substates stay open for sufficiently long periods of time to alter the passage of enough solutes, relative to the main state, to be physiologically relevant. That is, if subconductance states are to have meaningful effects on cell-to-cell transfer of small solutes, then the open probability of the subconductance states would be predicted to occupy a larger fraction of the open dwell time than we estimate, on the basis of our current observations. As such, our observations indicate that the permselectivity of hCx43 subconductive states is secondary to the dwell time with regard to any possible effects on solute transit for Cx43.
In short, the main issue is not the selectivity but rather the relative
contribution of the substate to the total open time. For example, the
average mean open time for the main state of hCx43 has been estimated
to be 1.4 s.6 If the subconductance state or states
represent 2% of the average mean open time, then the average
mean open time or dwell time for the subconductive states is estimated
to be 28 ms. Clearly, such a substate frequency and duration are not
sufficient to alter the transit of solutes from cell to cell relative
to the expected contribution of the main state. To further evaluate the
potential contribution of subconducting states to intercellular
communication, we examined the effects of phosphorylating treatments
known to promote their formation. In 2 experiments, treatment of cells
with 8-bromo-cAMP had no detectable effect on the weight of the
subconductive states in the amplitude histogram. As such, it would seem
that under the experimental conditions used in the present study,
subconducting states play little, if any, role in modulating
intercellular communication between human vascular smooth muscle
cells.
In conclusion, if similar substate phenomena do indeed exist in the in situ environment, then substates most certainly would serve a similarly minor role in the modulation of intercellular communication in multicellular tissues. Clearly, this supposition awaits verification in vivo.
|
Acknowledgments |
|---|
This work was supported in part by National Institutes of Health United States Public Health Service grants DK46379 and DK42027 (to G.J.C.) and HL31299 and GM55263 (to P.R.B.).
Received July 22, 1998; accepted January 20, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-
Veenstra RD, Wang HZ, Beblo DA, Chilton MG, Harris
AL, Beyer EC, Brink PR. Selectivity of connexin-specific gap junctions
does not correlate with channel conductance. Circ Res. 1995;77:1156–1165.
[Abstract/Free Full Text] -
Wang HZ, Veenstra RD. Monovalent ion selectivity
sequences of the rat connexin43 gap junction channel. J Gen
Physiol. 1997;109:491–507.
[Abstract/Free Full Text] -
Simpson I, Rose B, Loewenstein WR. Size limit of
molecules permeating the junctional membrane channels.
Science. 1977;195:294–296.
[Abstract/Free Full Text] -
Tsien RS, Weingart R. Inotropic effect of cyclic AMP
in calf ventricular muscle studied by a cut end method.
J Physiol (Lond). 1976;260:117–141.
[Abstract/Free Full Text] -
Christ GJ, Moreno AP, Melman A, Spray DC. Gap
junction-mediated intercellular diffusion of Ca2+
in cultured human corporal smooth muscle cells. Am J
Physiol. 1992;263:C373–C383.
[Abstract/Free Full Text] -
Brink PR, Ramanan SV, Christ GJ. Human connexin 43 gap
junction channel gating: evidence for mode shifts and/or
heterogeneity. Am J Physiol. 1996;271:C321–C331.
[Abstract/Free Full Text] -
Burt JM, Spray DC. Inotropic agents modulate gap
junctional conductance between cardiac myocytes. Am J
Physiol. 1988;254:H1206–H1210.
[Abstract/Free Full Text] - Brink PR. Gap junction channel gating and permselectivity: their roles in coordinated tissue function. Clin Exp Pharmacol Physiol. 1996;23:1041–1046.[Medline] [Order article via Infotrieve]
-
Adhern G, Junankar PR, Dulhunty A. Subconductance
states in single channel activity of skeletal muscle ryanodine
receptors after removal of FKBP12. Biophys J. 1997;72:146–162.
[Abstract/Free Full Text] -
Patlak JB. Sodium channel subconductance levels
measured with a new mean-variance analysis. J Gen
Physiol. 1988;92:413–430.
[Abstract/Free Full Text] -
Valiunas V, Bukauskas FF, Weingart R. Conductances and
selective permeability of connexin43 gap junction channels examined in
neonatal rat heart cells. Circ Res. 1997;80:708–719.
[Abstract/Free Full Text] -
Moreno AP, Campos de Carvalho AC, Christ G, Melman A,
Spray DC. Gap junctions between human corpus cavernosum smooth muscle
cells: gating properties and unitary conductance. Am J
Physiol. 1993;264:C80–C92.
[Abstract/Free Full Text] - Saez JC, Nairn AC, Czernik AJ, Fishman GI, Spray DC, Hertzberg EL. Phosphorylation of connexin43 and the regulation of neonatal rat cardiac myocyte gap junctions. J Mol Cell Cardiol. 1997;29:2131–2145.[Medline] [Order article via Infotrieve]
-
Moreno AP, Rook MB, Fishman GI, Spray DC. Gap junction
channels: distinct, voltage-sensitive and -insensitive conductance
states. Biophys J. 1994;67:113–119.
[Abstract/Free Full Text] -
Veenstra RD, Wang HZ, Beyer EC, Ramanan SV, Brink PR.
Connexin37 forms high conductance gap junction channels with
subconductance state activity and selective dye and ionic
permeabilities. Biophys J. 1994;66:1915–1925.
[Abstract/Free Full Text] -
Ramanan SV, Brink PR. Multichannel recordings
from membranes which contain gap junctions, II: substates and
conductance shifts. Biophys J. 1993;65:1387–1395.
[Abstract/Free Full Text] - Ramanan SV, Manivannan K, Mathias RT, Brink PR. Heterogeneous channel gating in gap junctions. In: Hall JE, Zampighi G, Davis RM, eds. Gap Junctions. New York, NY: Elsevier Science; 1993. Progress in Cell Research Series; vol 13.
- Veenstra RD, Wang HZ, Beyer EC, Brink PR. Selective dye and ionic permeability of gap junction channels form by connexin45 channels. Circ Res. 1994;268:706–712.
- Brink PR, Cronin K, Banach K, Peterson E, Westphale EM, Seul EH, Ramanan SV, Beyer EC. Evidence for heteromeric gap junction channels formed from rat connexin43 and human connexin37. Am J Physiol. 1997;273:C1386–C1396.
This article has been cited by other articles:
 |
|
 |
|
  G. Kanaporis, G. Mese, L. Valiuniene, T. W. White, P. R. Brink, and V. Valiunas Gap Junction Channels Exhibit Connexin-specific Permeability to Cyclic Nucleotides J. Gen. Physiol., March 31, 2008; 131(4): 293 - 305. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-Z. Wang, P. R. Brink, and G. J. Christ Gap junction channel activity in short-term cultured human detrusor myocyte cell pairs: gating and unitary conductances Am J Physiol Cell Physiol, December 1, 2006; 291(6): C1366 - C1376. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Vogel, V. Valiunas, and R. Weingart Subconductance States of Cx30 Gap Junction Channels: Data from Transfected HeLa Cells versus Data from a Mathematical Model Biophys. J., September 15, 2006; 91(6): 2337 - 2348. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. J. Christ, N. S. Day, M. Day, W. Zhao, K. Persson, R. K. Pandita, and K.-E. Andersson Increased connexin43-mediated intercellular communication in a rat model of bladder overactivity in vivo Am J Physiol Regulatory Integrative Comp Physiol, May 1, 2003; 284(5): R1241 - R1248. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Lagaud, V. Karicheti, Harm. J. Knot, G. J. Christ, and I. Laher Inhibitors of gap junctions attenuate myogenic tone in cerebral arteries Am J Physiol Heart Circ Physiol, December 1, 2002; 283(6): H2177 - H2186. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Valiunas, E. C. Beyer, and P. R. Brink Cardiac Gap Junction Channels Show Quantitative Differences in Selectivity Circ. Res., July 26, 2002; 91(2): 104 - 111. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Qu and G. Dahl Function of the voltage gate of gap junction channels: Selective exclusion of molecules PNAS, January 22, 2002; 99(2): 697 - 702. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Li and J. M. Simard Connexin45 gap junction channels in rat cerebral vascular smooth muscle cells Am J Physiol Heart Circ Physiol, November 1, 2001; 281(5): H1890 - H1898. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Valiunas, J. Gemel, P. R. Brink, and E. C. Beyer Gap junction channels formed by coexpressed connexin40 and connexin43 Am J Physiol Heart Circ Physiol, October 1, 2001; 281(4): H1675 - H1689. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.-E. Andersson Pharmacology of Penile Erection Pharmacol. Rev., September 1, 2001; 53(3): 417 - 450. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-Z. Wang, N. Day, M. Valcic, K. Hsieh, S. Serels, P. R. Brink, and G. J. Christ Intercellular communication in cultured human vascular smooth muscle cells Am J Physiol Cell Physiol, July 1, 2001; 281(1): C75 - C88. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Kitakaze, T. Minamino, H. Funaya, K. Node, Y. Shinozaki, H. Mori, and M. Hori Vesnarinone Limits Infarct Size via Adenosine-Dependent Mechanisms in the Canine Heart Circulation, April 15, 1997; 95(8): 2108 - 2114. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||