Ca2+-responses of white adipocytes to Ca2+-free medium application.
When the complete medium is replaced with calcium-free medium supplemented with 0.5 mM EGTA (Ca2+-free), different Ca2+-releases are observed in mature white adipocytes (9 days in vitro, DIV). Directly after stimulation, Ca2+-oscillations occur in 45 ± 11% of cells, Ca2+- oscillations are observed in 23 ± 8% of cells after a lag period of 5.2 ± 3.5 minutes, and transient Ca2+-signals are recorded in 32 ± 11% of adipocytes (Fig. 1А). Ca2+-oscillations in adipocytes occur without changing the basal level [Ca2+]i (Fig. 1А, black and green curves), and after transient signals, a new basal level of [Ca2+]i is established (Fig. 1А, red curve). When the Ca-free medium is replaced with a medium containing 1.2 mM Ca2+, no changes in [Ca2+]i are found in adipocytes (Fig. 1А).
Figure 1B shows the representative calcium signals of white adipocytes in response to repeated addition of Ca2+ -free buffer. After the first short-term (~ 160 seconds) stimulation of adipocytes by Ca2+-free medium, the complete medium (containing 1.2 mM Ca2+) was washed, and consequently, Ca2+-oscillations were suppressed. To recover of calcium signaling system of white adipocytes repeated application of Ca2+-free medium was performed after 30 minutes pause in the registration of [Ca2+]i. Repeated application of Ca2+-free medium induces Ca2+-release in 87 ± 21% of adipocytes from the population of cells that responded to the first stimulation. Diminution in the number of cells in response to the repeated stimulation is most likely associated with irreversible desensitization of the calcium-transport systems. Moreover, each of adipocytes is still able to respond to Ca2+ -free buffer exclusively in the form of Ca2+-response that occurs during the first short-term stimulation, i.e., adipocytes, which responded to the first stimulation with impulse rise in [Ca2+]i, react with the same type of signal to the repeated stimulation (Fig. 1D, black curve), but never with oscillations. Such a feature of white adipocytes may be due to an individual set of receptors and expression of calcium-transport systems, which is a good property for inhibitory analysis and establishing the mechanisms underlying the sensitivity of white fat adipocytes to changes in the concentration of extracellular calcium.
Therefore, the replacement of the extracellular medium of cultured white adipocytes with calcium-free medium induces generation of two types of Ca2+-signals, transient Ca2+-responses and Ca2+-oscillations, which are suppressed when the concentration of extracellular Ca2 + is restored.
Activation of Cx-43 connexin hemichannels promotes generation of Ca2+-oscillations in response to low extracellular Ca2+-level but this has no effect on Ca2+-transients
It is known that connexin hemichannels can be activated by a decrease in the concentration of extracellular Ca2+ ([Ca2+]ex), and Cx-43 type is most abundantly expressed in white fat adipocytes [32]. Connexin hemichannels blockers, such as carbenoxolone (CBX, 100 µM) and octanol (1 mM), completely suppress Ca2 + oscillations in all white adipocytes directly after 30-min preincubation (Fig. 2А), and their signals to addition of Ca2+-free medium convert into a single Ca2 + transient. Incubation of cells with proadifen (100 µM) (Fig. 2D) elicited a similar effect. In the population of white adipocytes that respond to the first addition of Ca2+-free medium, in the form of transient signals, connexin hemichannels blockers had no effects on the cells’ responses after the second addition of Ca2+-free medium (Fig. 2В). In general, no decreases in the amplitude of Ca2 + transients were found (Fig. 2D). Along with connexin hemichannels, pannexin 1 hemichannels (Pannexin-1) are also expressed in white adipocytes and perform several important physiological functions [45]. Incubation of white adipocytes with a pannexin blocker, probenecid (PROB, 1 mM), had no impact on generation of Ca2+-oscillations (Fig. 2А) and the amplitude of Ca2+-transients (Fig. 2D). After addition of Ca2+-free medium, the highly selective peptide blocker Pannexin-1 (10Panx, 100 µM) did not influence the Ca2+-signals of white adipocytes either (Fig. 2D).
Interestingly, cellular Cx43 gene knock-down using Gja1 siRNA not only to completely inhibits Ca2+-free medium-induced Ca2+-oscillations (Fig. 2С), but also provokes a statistically significant decrease in the amplitude of Ca2+-transients (Fig. 2D). Herewith, cellular Cx43 knock-down does not change the density of a cell culture, cell morphology (not shown) and their response to physiological stimuli (Fig. 2С – NE); the response of white adipocytes is a high-amplitude increase in [Ca2+]i after application of 1 µM of noradrenaline, which is typical for normal cell cultures of mature white adipocytes [46].
It is known that connexin hemichannels in an opened state are permeable to carboxyfluorescein [41]. In the presence of carboxyfluorescein in the incubation media, application of Ca2+-free to white adipocytes facilitated intracellular accumulation of the dye (Fig. 3А, В), indicating increased open probability of membrane channels. Ca2+-free stimulation-induced dye accumulation was prevented by CBX (Fig. 3А, В) or peptide blockers of Cx43 – Gap-26 (рис. 3В) and by Cx43 gene knock-down using Gja1 siRNA (Fig. 3А, В). Preincubation of white adipocytes with the blocker of Panexin-1–10Panx (100 µM) did not affect intracellular accumulation of the dye (Fig. 3B), which confirms that pannexins are not involved in the response of white adipocytes to a decrease in the concentration of external calcium.
Thus, a decline in the concentration of extracellular Ca2 + induces generation of Ca2+-oscillations in white adipocytes due to the activation and opening of connexin, mainly Cx43 but not pannexin hemichannels. However, the blockers of connexin and pannexin hemichannels did not affect the generation and amplitude of Ca2+-transients, which probably occur due to the activation of other signaling pathways.
Decrease in extracellular Ca2 + ions induces vesicular ATP secretion by white adipocytes.
A fairly large range of active molecules, including ATP, can be secreted through hemichannels [16]. The cells were loaded with fluorescent probe quinacrine to visualize ATP-containing vesicles and to research the dynamics of their secretion in white adipocytes during a decrease of [Ca2+]ex and opening of Cx-43. Figure 4A shows a single white adipocyte loaded with quinacrine prior to stimulation with Ca2+-free medium, and the image obtained by TIRF-microscopy shows a great number of ATP-containing vesicles. After a 5-minute exposure to Ca2+-free medium, almost complete secretion of ATP-containing vesicles takes place (Fig. 3А, Ca2+-free). Time analysis of the dynamics of the secretion shows that the secretion of most ATP-containing vesicles occurs within the first 20–30 seconds after addition of Ca2+-free medium (Fig. 3В). In other non-excitable cells, for example, astrocytes, it was shown that ATP secretion is a Ca2+-dependent process [47] and incubation with Tetanus toxin (TeNT, 50 ng/mL), an inhibitor of Ca2+-dependent vesicular fusion, leads to the development of a lag phase followed by secretion of ATP-containing vesicles after addition of Ca2+-free medium (Fig. 3В, +TeNT). Furthermore, the number of secreted vesicles usually decreases (Fig. 3Е). This experiment shows that Ca2+-ions are needed for secretion of ATP-containing vesicles, and taking into account that after addition of Ca2+ -free medium, Ca2+-ions are absent in the external environment, then, probably, adipocytes use accumulated Ca2+. In fact, incubation of white adipocytes with Ca2+-chelator, BAPTA-AM, for an hour changes the form of Ca2+-free medium-induced Ca2+-oscillations. As a result, either the interval between oscillations becomes longer (Fig. 3С, black curve), or the amplitude of Ca2+-oscillations become smaller and their frequencies increase (Fig. 3С, red curve). Also, at the level of secretion of ATP-containing vesicles, incubation with BAPTA-AM, commonly, significantly reduces the number of secreted vesicles (Fig. 3Е). A decrease in the number of secreted vesicles in Ca2+-free medium was also observed upon incubated with Bafilomycin A1 (BafA) a vacuolar ATPase inhibitor (Fig. 3Е), or cellular knock-down of Cx43, when ATP secretion is completely suppressed in white adipocytes (Fig. 3D, E).
As a result, activation of Cx43 hemichannels occurs in response to a decrease in [Ca2+]ex, thereby leading to vesicular secretion of ATP-containing vesicles, a process dependent on the intracellular concentration of Ca2+-ions.
Activation of Cx-43 in response to a decrease in [Ca2+]excontributes to phosphoinositide signaling pathway and activation of G-proteins.
Application of ATP (10 µМ) to the cell culture of white adipocytes contributes to generation of predominantly Ca2+-oscillations that occur without any change in the basal level of [Ca2+]i in 22 ± 16% of adipocytes (Fig. 5А, designation 2), or with an increase in the basal level of [Ca2+]i in 47 ± 11% of adipocytes (Fig. 5А, designation 3). Ca2+-free medium-induced Ca2+-oscillations are rapidly suppressed with application of apyrase (apyrase, 35 U/ml, Fig. 5В), an ATP hydrolyzing enzyme. At the same time, in one population of white adipocytes, the basal [Ca2+]i level returned to that observed at rest, while in another population, no similar effect is found.
To generate Ca2+-signals, cells can use both the input of Ca2+-ions from outside and mobilization from intracellular calcium stores. Depletion of calcium from the endoplasmic reticulum (EPR) of the cells incubated with thapsigargin (TG, 10 µM), the SERCA inhibitor, leads to disappearance of both Ca2+-oscillations and Ca2+-transients of white adipocytes following a decrease in [Ca2+]ex (Fig. 5С). U73122 (10 µM), a phospholipase C (PLC) inhibitor, completely inhibits Ca2+-release from white adipocytes after repeated addition of Ca2+-free buffer (Fig. 5D). Similarly, inhibition of the IP3R by Xestospongin C (XeC, 1 µM, Fig. 5Е) prevents generation of Ca2+-oscillations and transients in all adipocytes in response to the first addition of Ca2+-free buffer. However, the ryanodine receptor as an inhibitor of ryanodine receptors ryanodine (Rya, 10 µM, Fig. 5F) has no effect on generation of Ca2+-signals from adipocytes.
MRS-2179 (30 µM, Fig. 5G), the P2Y1 receptor antagonist, suppressed completely Ca2+-oscillations of white adipocytes, but the signals in response to addition of Ca2 +-free medium were in the form of single fast Ca2+-impulses. Furthermore, suramin (5 µM, Fig. 5Н), an uncoupler of G-proteins [48], inhibits completely Ca2+-release from adipocytes suggesting that G-proteins participate in activation of Ca2+-transients in response to a decrease in [Ca2+]ex; this issue requires further study.
The Ca2+-oscillations (red curves) and Ca2+-transients (black curves) of single cells are presented. Between the first and second applications of Ca2+-free, there was a 30 minutes pause in the Ca2+-dynamics registration. During recording pause, an inhibitor was added.
Thus, the signaling pathway involved in generation of Ca2+-oscillations by white adipocytes in response to a decrease in extracellular Ca2 + includes mobilization of Ca2+-ions from the thapsigargin -sensitive Ca2+-pool of endoplasmic reticulum with participation of phospholipase C and IP3R activation. In this case, paracrine activation of P2Y1 takes place in response to the opening of Cx43 hemichannels and secretion of ATP by adipocytes which responded to addition of Ca2+-free medium that leads to local signal propagation throughout the cells of white adipose tissue.
A decrease in [Ca2+]exstimulates lipolysis and correlates with generation of Ca2+-oscillations in white adipocytes.
It is known that Ca2+-oscillations can regulate many physiological processes both in excitable and electrically unexcitable tissues [49]. Replacement of the extracellular medium with a calcium-free medium results in generation of Ca2+-oscillations, on average, in 29 ± 14% of cells, which last for 52 ± 5 minutes and tend to decrease the amplitude of Ca2+-signals (Fig. 6А, red curve). At the same time, 47 ± 9% of white adipocytes show transient Ca2+-signals in the presence of Ca2+-free solution and no Ca2+-oscillations occur in this cell population for 60 minutes of recording, and increased basal [Ca2+]i level returned to the basal level observed at rest by the end of the cell response during the experiment (Fig. 6A, blue curve). The use of Ca2+-free solution and application of 100 µM of CBX (Fig. 6В, black curve) or Cx43 knockout (Cx-43-KD, Fig. 6B, green curve) to adipocytes generated Ca2+-transients exclusively in 54 ± 18% and 21 ± 7% of cells, respectively, the amplitudes of which were usually lower as compared to control (Fig. 6А, blue curve). Having recorded adipocyte Ca2+-dynamics after addition of Ca2+-free solution, the cell cultures were put into the CO2 incubator again for 24 hours. Total RNA was then isolated from one part of the cells and used for PCR analysis, while the other part of the cells was fixed and loaded with a probe that stained lipid inclusions (OilRed, Fig. 6D).
The expression of Gja1, a gene encoding Cx-43, does not significantly change after the use of Ca2+-free solution, in combination with the connexin blocker CBX (Fig. 6А), while the cell knockdown of Gja1 leads to almost complete suppression of expression of this gene. 24 hours after application to Ca2+-free solution in white adipocytes we observed a rise in the expression level of Lipe, Sirt1, Sirt3 and Atgl genes encoding hormone-sensitive lipase, sirtuins 1 and 3, triglyceride lipase, by a factor of 8.6, 3.1, 3.3 and 3.6, respectively (Fig. 6С). Simultaneously, there is a 54% suppression of the expression of the gene Igf2 encoding insulin-like growth factor-2 (Fig. 6С). Data on changes in gene expression coincide with those on diminution in the number of adipocytes loaded with lipid droplets (Fig. 6D, +Ca2+-free) as compared to the control (Fig. 6D, Control) (adipocytes on the 9 DIV, without Ca2+-free); this may indicate an activation of lipolysis. The addition of Ca2+-free solution for 60 min in the presence of CBX (100 µM) blocker leads to a significant increase in the expression level of 2 out of 5 genes studied, such as Igf2 (by 3,2 times) and Atgl (by 2,9 times) (Fig. 6С). The use of Ca2+-free solution in combination with CBX leads to smaller morphological changes in the white adipocytes (Fig. 6D, CBX + Ca2+-free) as compared to those observed when only Ca2+-free solution is used: the number of adipocytes loaded with lipid droplets begins to decline when compared to the control (Fig. 6D, Control), but increase when Ca2+-free solution is added (Fig. 6D, +Ca2+-free). A knockdown of the Gja1 gene suppresses not only Ca2+-oscillations in white adipocytes in response to application of Ca2+-free medium, but also contributes to a decrease in the expression of the Lipe (на 19%), Sirt1 (by 31%) genes and a 5.8-fold increase in Igf2 expression (Fig. 6С); this exactly correlates with the removal of the effect Ca2+-free medium has on intracellular accumulation of lipid droplets(Fig. 6D, Cx-43-KD + Ca2+-free); OilRed stained adipocyte culture is similar in morphology to the control (Fig. 6D, Control).
Thus, 24 h after Cx-43 activation and generation of Ca2+-oscillations in white adipocytes we observed a change in the expression of key genes involved in the regulation of lipolysis accompanied by a diminution in the number of adipocytes loaded with lipid droplets. It should be noted that inhibition or Cx-43 knockdown not only suppresses Ca2+-oscillations, but also tends to decrease the level of genes encoding key enzymes of lipolysis.