Author Correction: Isolation of large dense-core vesicles from bovine adrenal medulla for functional studies

Large dense-core vesicles (LDCVs) contain a variety of neurotransmitters, proteins, and hormones such as biogenic amines and peptides, together with microRNAs (miRNAs). Isolation of LDCVs is essential for functional studies including vesicle fusion, vesicle acidification, monoamine transport, and the miRNAs stored in LDCVs. Although several methods were reported for purifying LDCVs, the final fractions are significantly contaminated by other organelles, compromising biochemical characterization. Here we isolated LDCVs (chromaffin granules) with high yield and purity from bovine adrenal medulla. The fractionation protocol combines differential and continuous sucrose gradient centrifugation, allowing for reducing major contaminants such as mitochondria. Purified LDCVs show robust acidification by the endogenous V-ATPase and undergo SNARE-mediated fusion with artificial membranes. Interestingly, LDCVs contain specific miRNAs such as miR-375 and miR-375 is stabilized by protein complex against RNase A. This protocol can be useful in research on the biological functions of LDCVs.


Results
LDcV isolation. LDCVs were isolated from bovine adrenal medulla by combining differential and continuous sucrose gradient centrifugation; schematic overview and photos of individual steps during the purification are illustrated in Figs. 1 and 2. The quality of purified LDCVs can be checked using immunoblotting 18 (Fig. 3). We first tested the purity of LDCVs isolated using the conventional protocols. The conventional methods applied a sucrose step gradient of high density, which exploits the fact that LDCVs have the highest known buoyant density of all organelles 16,17 . Smith and Winkler used 0.3 M/1.6 M sucrose gradient to purify LDCVs 14 and we observed that the continuous sucrose gradient from 0.3 M to 1.6 M gives rise to the serious contamination of mitochondria (Fig. 3a). VAMP-2 and succinate dehydrogenase complex subunit A (SDHA) as protein markers represent the presence of LDCV and mitochondria, respectively. LDCVs migrate into the pellet fraction due to their high buoyant density, whereas other organelles largely remain on top of the sucrose layer. We only collected the pellet fraction for LDCVs after continuous sucrose gradient centrifugation. However, the buoyant density of mitochondria overlaps slightly with that of LDCVs. Mitochondria contaminants (SDHA) are present throughout the gradient and are still significant in the pellet fraction containing LDCVs in the conventional method that use 0.3 M/1.6 M sucrose gradient (Fig. 3a).
We next tried discontinuous sucrose gradient centrifugation with 0.3 M/1.8 M sucrose to isolate LDCVs (Fig. 3b). Mitochondria (SDHA) is still present in the pellet together with LDCVs; LDCVs are from the pellet fraction, proposing that 0.3 M/1.8 M sucrose gradient method is limited for isolating LDCVs with high purity. Then we examined mitochondrial contamination from LDCV pellets using the different sucrose gradient (Fig. 3c)  . Separation of LDCVs from mitochondria by increasing the maximal density. (a) Continuous sucrose gradient from 0.3 M to 1.6 M was used, which resembles conventional protocols. Profile of VAMP-2 (a marker for LDCV membranes) and succinate dehydrogenase complex subunit A (SDHA, a marker for mitochondria) across gradient fractions. Aliquots of equal volume from each fraction were subjected to SDS-PAGE and immunoblotting, as described elsewhere 18 . Mitochondria (SDHA) contaminants were present throughout the gradient and are still significant in the pellet fraction containing the LDCVs. (b) The LDCV fraction was recovered from the pellet after a discontinuous sucrose gradient in which a 1.   www.nature.com/scientificreports www.nature.com/scientificreports/ markers for LDCVs, are highly enriched in the continuous sucrose gradient with 0.3 M/2.0 M compared to the discontinuous sucrose gradient (Fig. 3c), suggesting that continuous sucrose gradient is much better for the high purity than discontinuous sucrose gradient. Taken together, we improved the purity of LDCV by i) increasing the maximal sucrose concentration to 2.0 M and ii) applying a continuous sucrose density gradient. This protocol is optimized to improve the purity of LDCVs by removing mitochondrial contamination.
Dynamic light scattering analysis. Dynamic light scattering analysis was carried out to determine the size distribution of isolated LDCVs from bovine adrenal medulla, which give rise to 153.7 ± 42.2 nm (Fig. 4a). The diameter of purified LDCVs can also be measured using electron microscopy that showed a heterogeneous size distribution with an average diameter of 167.7 nm 18 , correlating with dynamic light scattering analysis (Fig. 4a). The average diameter of purified LDCVs with ~153 nm is smaller than that of total LDCVs in chromaffin cell; the average diameter is 356 nm 19 . It is because this method with 0.3 M/2.0 M continuous sucrose gradient purifies mainly mature LDCVs (smaller in size) whereas immature LDCVs are separated away 18 . Mature LDCVs become smaller and condensed through removal of water during the maturation process 20,21 , proposing that isolated LDCVs have high purity and mature LDCVs are mainly purified 18 .
Vesicle fusion assay using purified LDcVs. LDCVs frozen for several months are structurally and functionally intact as judged by functional assay. We measured the activity with respect to SNARE-mediated membrane fusion. The reconstitution of LDCV fusion was previously established using purified LDCVs 18,22 . Purified LDCVs fused with plasma membrane-mimicking liposomes that contain the stabilized soluble N-ethylmaleimide-sensitive factor attachment protein receptor (Q-SNARE) complex with high efficiency (Fig. 4b).
Vesicle acidification by V-ATPase. Another functional assay for intact LDCVs is to test the activity with respect to luminal acidification. A key role of LDCVs is the accumulation of monoamines such as adrenaline, noradrenaline, dopamine and serotonin prior to their release through SNARE dependent exocytosis. Monoamine loading into LDCVs relies on a proton electrochemical gradient (ΔμH + ) generated by the V-ATPase in the LDCV membrane. In presence of ATP the V-ATPase pumps protons into the vesicle lumen which results in the formation of a proton gradient (ΔpH) and an inside positive membrane potential (ΔΨ). In absence of negative counter ions ΔpH is comparably small, as the V-ATPase is slowed down by the accumulation of positive charges in the vesicle lumen. However, in presence of membrane permeable anions such as Cl − , anion co-influx neutralizes the positive charges, which stimulates proton pumping and thus facilitates lumenal acidification. Indeed, we observed an initial ATP dependent acidification that was strongly enhanced upon addition of Cl − (Fig. 4c), demonstrating that the purified LDCVs exhibit a pronounced proton pumping activity. Furthermore, blocking of the V-ATPase with the inhibitor Bafilomycin reversed acidification, suggesting V-ATPase-specific acidification.

Stability of miR-375 stored in LDCVs.
We have previously reported that LDCVs contain a variety of non-coding RNAs including miR-375, which is a dominant miRNA in LDCVs 11 . miR-375 is protected in the presence of RNase A as reported previously 11 . Here we tested how miR-375 can be stabilized after exocytosis. The presence of miRNAs in LDCVs are shown using qRT-PCR (Fig. 5) and the details of experiment set-up is described in Table 1. A synthesized Caenorhabditis elegans miRNA (cel-miR-39) as the spike-in control was added to normalize the RNA extraction efficiency. RNase A is not able to pass across the membranes so that it selectively degrades vesicle-free miRNA, but not vesicle-incorporated miRNA. RNase A degraded cel-miR-39, the spike-in control, but miR-375 remained intact, suggesting that miR-375 is incorporated in LDCVs 11 .
Next, we tested if miR-375 can be degraded by RNase A after Triton X-100 (TX-100) detergent treatment. TX-100 is a non-ionic detergent and completely disrupts the vesicle membrane. cel-miR-39 was mixed with LDCVs prior to TX-100 and then RNase A was applied after TX-100 treatment (Fig. 5a). We observed that miR-375 remains resistant to RNase A even in the presence of TX-100, whereas cel-miR-39 becomes degraded by RNase A (Fig. 5a). Then, we hypothesized that proteins might stabilize miR-375 and miR-375 was degraded by RNase A when proteinase K (PK) was included in TX-100-treated LDCVs (Fig. 5b). To confirm that proteins can protect miR-375, RNase A is applied after RNA extraction from protein complex using miRNeasy Mini Kit. Indeed, miR-375 is completely degraded by RNase A after proteins are removed; this result suggests that miR-375 is stabilized by proteins (Fig. 5c). Altogether, this protocol provides methods to test the presence of miRNAs in LDCVs and the stability of miR-375.

Discussion
This protocol obtains intact LDCVs from bovine adrenal medulla. These purified LDCVs can then be used to study their composition and content and to investigate their biochemical features such as proton pumping and SNARE-mediated membrane fusion as described in Fig. 4. More interestingly, purified LDCVs may be carriers in neurons and neuroendocrine cells that deliver miRNAs and other non-coding RNAs to extracellular fluid. miR-375 is stored in LDCVs and stabilized by protein complex (Fig. 5). (2020) 10:7540 | https://doi.org/10.1038/s41598-020-64486-3 www.nature.com/scientificreports www.nature.com/scientificreports/ The protocol of LDCV isolation begins with mild homogenization of bovine adrenal medulla to disrupt the plasma membrane and release free LDCVs into the supernatant. These extracts are then loaded on a continuous sucrose density gradient ranging from 0.3 M to 2.0 M and centrifuged, resulting in the separation from other organelles. Mitochondria (marker, SDHA), late endosomes/multivesicular bodies (LEs/MVBs) and lysosomes (marker, LAMP-1), early endosomes (marker, EEA-1), endoplasmic reticulum (marker, calnexin), proteasomes (marker, Rpt-4), and peroxisomes (marker, catalase) are mostly removed, whereas LDCVs are highly enriched 18 . This fractionation method yields primarily mature LDCVs with 100~200 nm in diameter 18 whereas immature LDCVs (identified here by the presence of VAMP-4) are also removed. Due to the vesicle maturation and condensation, mature LDCVs exhibit much higher buoyant density than immature LDCVs, and thus mature LDCVs are mainly enriched in the pellet 18 (Fig. 3c). Furthermore, our previous overlay assay shows ~95% purity of isolated LDCVs 11 , supporting that most mitochondria are removed and isolated LDCVs are highly pure.
LDCVs contain a variety of miRNAs, which are released by LDCV exocytosis 11 . A new paradigm suggests that secreted miRNAs constitute novel neuromodulators by regulating cell−to−cell communication that include gene silencing and cellular signaling. We propose 'ribomone (ribonucleotide + hormone)' , because miRNAs are stored inside the vesicles and released by active vesicle fusion in response to stimulation so that secreted miRNAs might regulate cell−to−cell communication including gene silencing and cellular signaling. Our data provide evidence that secreted miR-375 can be stabilized by protein complex against RNase A, implying that secreted miRNAs might function a physiological role for long-term with high stability.
In conclusion, this protocol improves the purity of LDCVs by combining differential and continuous sucrose gradient centrifugation. This technique can be useful in research on the biological functions of LDCVs, including vesicle fusion, vesicle acidification, monoamine transport, and the functions of miRNAs stored in LDCVs. The purification of LDCVs also contributes to study psychiatric and mental disorders, which might be caused by imbalance of serotonin, adrenaline, and dopamine, because LDCVs store most of monoamine transmitters such as serotonin, adrenaline, and dopamine, which are directly linked to psychiatric and mental disorders.
LDcV isolation. Fresh bovine adrenal glands were obtained from a local slaughterhouse. After trimming away the cortex and fat, the medullae were minced with a scissor in 300 mM sucrose buffer (300 mM sucrose, 20 mM HEPES, pH 7.4 adjusted with KOH) and then homogenized using a cooled a Glass-Teflon homogenizer at 1,000 rpm (H, homogenate). PMSF (200 μM) was added to prevent protein degradation. All subsequent steps www.nature.com/scientificreports www.nature.com/scientificreports/ were carried out at 0°-4 °C. After centrifugation at 1,000 g for 15 min at 4 °C, the pellet containing nuclei and cell debris (P1) was discarded. The supernatant (S1) was further centrifuged (12,000 g, 15 min, 4 °C), followed by the additional cycle of resuspension and centrifugation for washing step. The resulting pellet (P2, crude LDCV fraction) was resuspended in 300 mM sucrose buffer and loaded on top of a continuous sucrose gradient (from 300 mM to 2.0 M) to remove other contaminants including mitochondria. LDCVs were collected from the pellet after centrifugation at 110,000 g for 60 min in a Beckman SW 41 Ti rotor and resuspended with the buffer (120 mM K-glutamate, 20 mM K-acetate, 20 mM HEPES.KOH, pH 7.4). The fraction directly on top of the pellet was removed and the pellet was only resuspended in order to purify mature LDCVs. The purified LDCVs can be snap-frozen in liquid nitrogen and stored for several months at −80 °C. Make small aliquots of LDCV samples to reduce damage by freeze-thaw cycles. Size distribution of purified LDCVs was determined using dynamic light scattering (NanoPlus DLS, Particulate Systems).
Vesicle fusion assay. Vesicle fusion reactions were performed at 37 °C. 50 μg of purified LDCVs and 10 μl of plasma membrane-mimicking liposomes were mixed in 1 ml of buffer containing 120 mM K-glutamate, 20 mM K-acetate, 20 mM HEPES-KOH (pH 7.4). Plasma membrane-mimicking liposomes contained the stabilized Q-SNARE complex 23 . Fluorescence dequenching signal was measured with wavelengths of 460 nm for excitation and 538 nm for emission. Fluorescence values were normalized as the percentage value of the maximum donor fluorescence induced by 0.1% Triton X-100 detergent treatment at the end of experiments. Control represents basal fusion without any treatment.
immunoblotting. Glycerol-containing gels with 0.1% SDS were used to separate low molecular weight proteins with high resolution. Proteins were transferred to a nitrocellulose membrane and then blocked with 5% non-fat dry milk in solution (20 mM Tris-HCl, pH 7.4, 137 mM NaCl, and 0.1% tween-20). Proteins as indicated in manuscripts were detected using horseradish peroxidase-conjugated secondary antibodies.
Vesicle acidification assay. Acidification measurements were performed as described previously using acridine orange (AO, Molecular Probes) as a pH sensitive dye 25 . Changes in absorbance at 492 nm (ΔAU) were monitored in an Aminco dual-wavelength spectrophotometer using absorbance at 530 nm as reference, giving a read-out of lumenal pH changes 26 . Usually, 600-650 µl buffer (300 mM glycine, 10 mM MOPS, pH 7.3, and 2 mM MgSO4) were mixed in a 1 ml glass cuvette with purified LDCVs containing 10 µM AO and measured at 32 °C. The measurements were performed in 300 mM glycine, 10 mM MOPS, 2 mM MgSO4, pH 7.3 buffer.
Quantification of miR-375 using qRT-PCR. Several reactions to test the stability of miR-375 in the presence of RNase A, TX-100, and/or proteinase K was described in details in Table 1. 0.5 pmol of syn-cel-miR-39 was incubate with ~50 µg of LDCVs for 5 min at RT (~22 °C). 1% (vol/vol) Triton Tm X-100 (TX-100) was treated with LDCVs for 5 min at room temperature. Then, 10 µg/ml (final concentration) of RNase A was applied for 15 min in the presence or absence of TX-100. If RNase A is used, then 1 U/µl RNase inhibitor was added for 5 min to stop the RNase activity. In cases of proteinase K treatment 2 µl of 100 µg/ml of proteinase K was applied for 30 min. A protease inhibitor cocktail was added for 5 min to stop proteinase activity. Then 10 µg/ml RNase A was applied to degrade RNA.
Total RNA from LDCV samples with different treatments (Table 1) was isolated using miRNeasy Mini Kit according to the manufacturer's protocol. Isolated RNAs were transferred into 0.2 ml PCR Strip tubes, then cDNA syntheses was performed using miScript RT II kit according to the manufacturer's protocol. 5X miScript HiSpec buffer was used to prepare the reaction mix. miR-375 and cel-miR-39 were quantified using qRT-PCR according to miScript SYBR Green PCR Kit protocol. For miRNA quantification of miR-375 and cel-miR-39, a standard curve should be established using syn-bta-miR-375 and syn-cel-miR-39. For this task, serial dilutions (0.02 pM to 2 nM) of syn-bta-miR-375 and syn-cel-miR-39 were prepared. qRT-PCR analysis was performed using Bio-Rad, CFX Connect ™ Real-Time PCR Detection System. Crossing point-PCR-cycle (Cp) values were used to plot the standard curve and to analyze the concentraction of miR-375 stored in LDCVs and cel-miR-39, which is a spike-in control.

Data availability
The datasets generated during the current study are available from the corresponding author on reasonable requests.