The endoplasmic reticulum, not the pH gradient, drives calcium refilling of lysosomes

1) You need at least one convincing experiment showing that inhibiting the V-ATPase increases pH (needs to be measured) but doesn't affect store [Ca²⁺]. In the Christensen studies, inhibiting V-ATPase with Baf (similar to the treatment used in this paper) increased luminal pH to ~ 7.0 (where the Ca²⁺ imaging dye should work well) and drastically decreased [Ca²⁺]lys from ~ mM to μM. Although you added new OG-BAPTA imaging, it needs calibration. You should show the data measuring the pH and [Ca²⁺].

We performed the experiments as suggested by the reviewers. First, we monitored lysosomal pH under different conditions using Oregon green 488-dextran, a commonly- used luminal pH indicator [1, 2] (see Methods and new Figure 3—figure supplement 3I). Using the calibration that we made for OG-488-dextran (see Figure 3—figure supplement 3I), we showed that lysosomal pH (pH_L), which is ~4.8 under the resting condition, increased dramatically within minutes to ~ pH 7.0 upon treatment of Baf-A1 (5 μM) or NH4Cl (10 mM) in HEK293 cells stably expressing ML1 (see Figure 3—figure supplement 3J, K). Second, based on the calibration for the luminal Ca²⁺ dye OG-BAPTA-dextran (see Figure 3—figure supplement 3C), we found that the intra-lysosomal [Ca²⁺] ([Ca²⁺]Ly), which was ~ 360 μM under the resting condition, dropped quickly to ~ 100 μM upon ML-SA1 application (see Figure 3H, I). Note that at pH 4.5, OG-BAPTA- dextran’s Kd is ~300 μM (Figure 3—figure supplement 3C), allowing measurement of Ca²⁺ ranging from 30 (0.1 Kd) to 3,000 (10 Kd) μM [3]. At the resting pH_L of 4.8 in the HEK cells, OG-BAPTA- dextran’s working range is 130-1,300 μM (Kd ~ 130 μM). Overall, all these values are in rough agreement with previous measurements [1, 3, 4].

Upon Baf-A1 and NH4Cl treatment to increase lysosomal pH, the fluorescence intensity of OG-BAPTA-dextran increased mildly, following the pH_L changes (Figure 3—figure supplement 3J, K). If the effect of lysosomal pH is exclusively on the Kd of OG-BAPTA-dextran, these results would suggest that [Ca²⁺]Ly dropped significantly, as concluded by Christensen et al. [5]. However, in addition to triggering lysosomal Ca²⁺ release, as proposed by Christensen et al. [5], lysosomal pH elevation is known to affect [Ca²⁺]Ly or its measurement via several additional mechanisms. Therefore, the interpretation offered by Christensen et al. might be too simplistic.

First, a change in lysosomal pH may cause a significant change in luminal Ca²⁺ buffering. In most studies dealing with cytosolic Ca²⁺, the total versus free [Ca²⁺] in the intracellular organelles (stores) are often not separately-considered since these two parameters are interrelated. Whereas the total [Ca²⁺]Ly is reported to be in the low mM range (5-10 mM), free [Ca²⁺]Ly is generally agreed to be in the high μM range (100-500 μM; 360 μM in the current study) [3]. Therefore, the lysosome lumen must contain a substantial amount of Ca²⁺ buffers [3]. Ca²⁺ buffers in the acidic compartments and ER are known to bind Ca²⁺ much better at neutral pH (H⁺ and Ca²⁺ compete with each other for Ca²⁺ buffers) [4]. Hence increasing pH_L from 4.8 to 7.0 may effectively reduce free [Ca²⁺]Ly without necessarily triggering lysosomal Ca²⁺ release and affecting total [Ca²⁺]Ly.

Second, lysosomal pH may act on luminal Ca²⁺ dyes by affecting their chromophore fluorescence and Ca²⁺-binding affinity (Kd) [3]. Because Kd is dropped more than 1,000 times when pH_L is increased from 4.8 to 7.0 (Figure 3—figure supplement 3C), perfect calibration is currently impossible. Note that in the Christensen studies, which were conducted in the activated macrophages, the basal pH_Lis ~ pH 3.8 [5]. When pH_Lwas increased from 3.8 to 5.0, the free [Ca²⁺]Ly was already dropped to ~ 10 μM (see Figure 3 of ref. [5]). However, in most studies including the current study, lysosomal pH is between 4.5 to 5.0 [1]. Hence, the mechanism proposed by Christensen et al., might not apply to most cell types whose basal pH_Lis close to pH 5.0. In addition, because of high levels of luminal Ca²⁺ buffers, the amount of releasable Ca²⁺ is high. Hence, a depletion of free [Ca²⁺]Ly to low μM is expected to produce a large cytosolic Ca²⁺ increase, which did not occur (see below). In the current study, upon ML-SA1 application to deplete the stores, free [Ca²⁺]Ly dropped from 360 to 100 μM (Figure 3I). Hence, within physiological ranges of lysosomal pH (4.8 -7.0) and Ca²⁺, the fluorescence signals of OG-BAPTA- dextran dyes might be nearly saturated, and this would prevent accurate determinations of [Ca²⁺]Ly changes when pH_Lis increased from 4.8 to 7.0 upon Baf-A1 or NH4Cl application. Note that at pH 7.0, OG-BAPTA-dextran’s Kd is ~0.1 μM (Figure—figure supplement 3C), which is not suitable for measuring Ca²⁺ over 10 μM.

Third, prolonged lysosomal pH manipulations may also indirectly affect lysosomal Ca²⁺ homeostasis, for instance, via membrane fusion and fission between compartments containing different amounts of Ca²⁺, H⁺, and their buffers.

Fourth,although elevating lysosomal pH might trigger lysosomal Ca²⁺ release, the accompanied increase in cytoplasmic Ca²⁺ was rather small (20-40 nM) (see refs. [4, 5]). This minor increase is more consistent with the interpretation that other pH-dependent processes contribute to the changes of OG-BAPTA fluorescence upon Baf-A1 and NH4Cl application. In addition, the instantaneous changes (following pH increase and decrease) of Ca²⁺ probe fluorescence (see Figure 5 in ref. [5] and Figure 3 in ref. [4]) are inconsistent with the slow rates of Ca²⁺ leak (τ = 6 min; see Figure 3—figure supplement 1E) and re-uptake (τ = 2 min; see Figure 1—figure supplement 2F & Figure 3—figure supplement 2C). Furthermore, given the pH-sensitivity of the Fura-2 dyes, even the observed minor increase (20-40 nM) in cytosolic Ca²⁺, based on instantaneous increases of Fura-2 ratios, could be mostly due to the effects of lysosomal proton release on the Fura-2 dyes (see Figure 1—figure supplement 3).

Because of the reasons listed above, in the current study, we used the luminal Ca²⁺ dyes for free [Ca²⁺]Ly measurement only in the experiments during which the pH remains unchanged.

The bar graphs (Figure 3H and Figure 3—figure supplement 3F) do not provide information about the kinetics of the Ca2+ change and how they may (or may not) correlate with the Ca2+ changes reported by the GCamP ML1 indicator shown in the majority of the data. Example traces of the lysosomal Ca2+ indicator in response to lysosomal and ER Ca2+ store depletion should be shown.

The example traces were now added into new Figure 3I and Figure 3—figure supplement 3H. Note that due to the photo-bleaching and small signal, we had to minimize the problem by reducing data acquisition. With the calibration curve, we measured lysosomal Ca²⁺ changes under different experimental manipulations. Whereas ML-SA1 application significantly depleted lysosomal Ca²⁺ contents (see new Figure 3I), TG treatment only slightly reduced lysosomal Ca²⁺ after 5- 10 min (Figure 3—figure supplement 3H). Although a quick drop in [Ca²⁺]Ly may cause a transient increase in cytosolic [Ca²⁺], it may also liberate certain amount of Ca²⁺ from the Ca²⁺-bound buffers, resulting in a delayed reduction in free [Ca²⁺]Ly. Therefore, as suspected by the reviewer, the kinetics of the GCaMP3-ML1 and OG-BAPTA-dextran were not perfectly correlated.

This is an excellent suggestion – thank you! See response #1 regarding [Ca²⁺]Ly measurement.

In GCaMP3-ML1-positive vacuoles that were isolated from HEK-GCaMP3-ML1 cells and released into the dish, we moved patch pipettes containing either a “high Ca²⁺” (10 mM) internal solution or a “low pH” solution (140 mM KCl, 1 mM EGTA, 20 mM MES, 10 mM Glucose, pH adjusted to pH 2.0) close to the vacuoles to produce a “puffing” effect. We showed that GCaMP3- ML1 responded to both Ca²⁺ and pH (see Figure 1—figure supplement 3E), but not to the “puffing” of the standard internal solution. These data suggest that our GCaMP3-ML1 probe, in addition to being a good indicator for measuring Ca²⁺ released from lysosomes, may also respond to lysosomal proton release, for instance, upon GPN and Baf-A1 application. Our results suggest that caution is necessary in designing the experiments to study the interaction between lysosomal Ca²⁺ and lysosomal pH. We propose that BAPTA-AM control experiments should be routinely performed in any lysosomal Ca²⁺ studies.

Please provide more detail in the Discussion why you think the previously published measurements were erroneous.

We added a new paragraph in the Discussion (second paragraph), and included a discussion in the Results.

Results: “In contrast, treatment of cells with Baf-A1 or NH₄Cl markedly increased lysosomal pH from 4.8 to 7.0 (Figure 3—figure supplement 3J, K). Such large pH elevations may cause dramatic changes in both K_d of OG-BAPTA-dextran (see Figure 3—figure supplement 3C) and luminal Ca²⁺ buffering capability (Dickson et al., 2012; Morgan et al., 2015), preventing accurate determinations of [Ca^2+]LY under these pH manipulations”.

Discussion:“Lysosomal pH gradient is thought to be essential for the maintenance of high free [Ca²⁺]Ly (Calcraft et al., 2009; Christensen et al., 2002; Dickson et al., 2012; Lloyd-Evans et al., 2008; Shen et al., 2012). However, in addition to triggering lysosomal Ca²⁺ release, as proposed by Christensen et al. (Christensen et al., 2002), lysosomal pH elevation is also known to affect [Ca²⁺]Ly or its measurement via several other mechanisms. Whereas the total [Ca²⁺]Ly is reported to be in the low mM range (5-10 mM), free [Ca²⁺]Ly is generally agreed to be in the high μM range (100-500 μM) (Morgan et al., 2015). Therefore, lysosome lumen must contain substantial amount of Ca²⁺ buffers (Morgan et al., 2015). Ca²⁺ buffers in the acidic compartments and ER are known to bind Ca²⁺ much better at neutral pH (Dickson et al., 2012). Hence increasing pH_Lfrom 4.8 to 7.0 may effectively reduce free [Ca²⁺]Ly without necessarily triggering lysosomal Ca²⁺ release and affecting total [Ca²⁺]Ly. Consistent with such interpretation, a compelling study recently demonstrated that in secretory granules and the ER, increasing luminal pH changed the Ca²⁺ buffering capacity of both Ca²⁺ containing compartments intraluminally to reduce free [Ca²⁺], while causing a minimal (20 nM) increase in cytosolic Ca²⁺ (Dickson et al., 2012). Additionally, lysosomal pH may act on luminal Ca²⁺ dyes by affecting their chromophore fluorescence and Ca²⁺-binding affinity (Kd) (Morgan et al., 2015). Because Kd is dropped more than 1,000 times when pH_Lis increased from 4.8 to 7.0, perfect calibration is near impossible. Furthermore, prolonged lysosomal pH manipulations may also indirectly affect lysosomal Ca²⁺ homeostasis, for instance, via membrane fusion and fission between compartments containing different amounts of Ca²⁺, H⁺, and their buffers. Finally, although elevating lysosomal pH may trigger lysosomal Ca²⁺ release, the accompanied increase in cytoplasmic Ca²⁺ was rather small (20- 40 nM) (Christensen et al., 2002; Dickson et al., 2012). Moreover, the instantaneous changes (following pH increase and decrease) of Ca²⁺ probe fluorescence (Christensen et al., 2002; Dickson et al., 2012) are inconsistent with the slow rates of Ca²⁺ leak and re-uptake demonstrated in the current study”.

3) A triple IP3R knockout cell line was used to assess the proposed role of IP3Rs in refilling the lysosomal store. However, in contrast to the authors' theory, this does not have a lysosomal Ca2+ defect. The authors attribute this to an unknown compensatory mechanism. Please examine whether the Xestospongin abolishes ML1-SA1 responses in the triple IP3R KO cells and show the kinetics of the response. If the responses in the IP3R KO are not affected, and the kinetics now differ, the results would provide an additional layer of confidence for the proposed role of IP3Rs in lysosomal refilling.

This is another excellent suggestion. We performed these experiments as suggested, and showed that although lysosomal Ca²⁺ refilling still occurred in IP3R TKO cells, the refilling kinetics were much slower than WT cells (see Figure 3—figure supplement 2C), and more importantly, refilling were completely insensitive to IP3R inhibitions (see Figure 3D-F).

References:

1) Johnson, D.E., et al., The position of lysosomes within the cell determines their luminal pH. J Cell Biol, 2016. 212(6): p. 677-92.

2) DiCiccio, J.E. and B.E. Steinberg, Lysosomal pH and analysis of the counter ion pathways that support acidification. J Gen Physiol, 2011. 137(4): p. 385-90.

3) Morgan, A.J., L.C. Davis, and A. Galione, Imaging approaches to measuring lysosomal calcium. Methods Cell Biol, 2015. 126: p. 159-95.

4) Dickson, E.J., et al., Orai-STIM-mediated Ca2+ release from secretory granules revealed by a targeted Ca2+ and pH probe. Proceedings of the National Academy of Sciences of the United States of America, 2012. 109(51): p. E3539-48.

5) Christensen, K.A., J.T. Myers, and J.A. Swanson, pH-dependent regulation of lysosomal calcium in macrophages. Journal of cell science, 2002. 115(Pt 3): p. 599-607.