2.1 Overexpression of DAP3 Induces Cell Death and Mitochondrial Damage

Upregulation of DAP3 levels has been reported to induce cell death [22, 23]. However, the underlying mechanism remains poorly understood and controversial results have been observed in previous studies [18–23, 29–32]. To investigate the roles of DAP3 in cell death, we first overexpressed C-terminal BFP-tagged DAP3 in HeLa cells. After 24 h, DAP3–BFP overexpression induced cell death in approximately 30.7 ± 5.9% of the transfected cells. In contrast, the percentage of cell death observed in cells transfected with the control BFP vector was low at approximately 8.9 ± 0.8% (Figure 1A,B).

We and others have previously revealed that DAP3 is a mitoribosomal subunit involved in the regulation of mitochondrial dynamics [13, 29, 34]. Since mitochondrial fragmentation is a typical feature of apoptosis, we investigated the impact of DAP3 overexpression on mitochondrial morphology and dynamics. Mitochondria were visualized by staining with MitoTracker dye. As shown in Figure 1C,D, DAP3 disrupted the balance of mitochondrial dynamics, resulting in extensive mitochondrial fragmentation in approximately 37.0 ± 9.2% of the cells, in contrast to only about 5.0 ± 1.4 3% in the control group, which is consistent with the previous report [29]. Quantitative analyses of the mitochondrial average size revealed a decrease in mitochondrial mean area from 1.75 ± 0.01 to 0.59 ± 0.01 µm² (Figure 1E), and a reduction in mean perimeter from 9.47 ± 0.44 to 3.56 ± 0.07 µm (Figure 1F). On the other hand, the mean circularity increased from 0.55 ± 0.01 to 0.73 ± 0.01 (Figure 1G), confirming a shift in mitochondrial morphology from tubular-like to dot-like. To further validate the mitochondrial fragmentation phenotype, we conducted the fluorescence recovery after photobleaching (FRAP) assay to measure mitochondrial interconnectivity. As shown in Figure 1H,I, the mitochondria in control cells (GFP-overexpressed) exhibited rapid fluorescence recovery after photo-bleaching within the initial 30 s, whereas this recovery process was markedly impeded upon DAP3–GFP overexpression, resulting in sustained darkness in the bleached region. These results further support the notion that DAP3 serves as a potent inducer of mitochondrial fragmentation.

Notably, weaker MitoTracker signals, indicating a loss of ΔΨm, were observed in some of the DAP3-overexpressing cells (Figure 1C). To confirm this observation, we performed TMRM staining for quantitative analysis of the ΔΨm. Overexpression of DAP3 resulted in a significant decline in mean TMRM intensity by approximately 38.1% in the fragmented mitochondria (Figure 2A,B), implying ΔΨm collapse and mitochondrial damage. In addition to membrane potential, we evaluated the mitochondrial ATP levels using a fluorescence resonance energy transfer (FRET)-based indicator, ATeam (Figure 2C) [36, 37]. ATeam is designed by linking a CFP tag and a YFP tag with the ε subunit of F₀F₁–ATP synthase. Upon ATP binding, the ε subunit brings the two fluorescent proteins close to each other, thus enhancing YFP emission via FRET. In line with the reduced ΔΨm, DAP3 overexpression also impaired mitochondrial ATP production (Figure 2D,E).

Given that DAP3 is a subunit of mitoribosome, we assessed whether its overexpression perturbs mitochondrial protein synthesis. As shown in Figure S1A, the levels of mitochondrial gene-encoded proteins, ND5 and COXII, remained unimpaired, suggesting DAP3 regulates cell death and mitochondrial homeostasis through a mechanism independent of its role in protein translation. To ascertain whether DAP3-induced mitochondrial fragmentation is a common phenomenon in other cell types, we overexpressed DAP3–GFP in HEK 293T, U2OS, and ARPE-19 cells. Consistently, mitochondrial fragmentation, accompanied by reduced mitochondrial size, was observed in all these cell lines (Figure S1B,C).

A burst of ROS has been widely recognized as a trigger of mitochondrial damage and cell death [38]. To extend the observations on mitochondrial health, intracellular ROS levels were measured using mitochondrial matrix- and cytosolic-specific dyes, respectively. As expected, overexpression of DAP3 elevated ROS levels in both mitochondrial matrix and cytosol (Figure 2F,G). The average ROS levels increased by approximately 30% in the matrix, while the rise in the cytosol was comparatively lower by 10%, but still significant.

To further examine ROS levels in different subcellular localizations, three types of redox-sensitive GFP (roGFP) probes were stably expressed in HeLa cells: Matrix-roGFP (localizes to mitochondrial matrix) [39], IMS-roGFP (specifically localizes to the mitochondrial intermembrane space) [40], and Cyto-roGFP (localizes in the cytosol) [39]. roGFP contains two surface-exposed cysteines, enabling the reversible formation of an intramolecular disulfide bridge, thereby acting as a sensor of redox status. roGFP exhibits two excitation peaks at ∼ 400 and ∼490 nm, and one emission peak at ∼510 nm. Oxidation of roGFP leads to the increase of the 400 nm excitation peak, while reduction enhances the 490 nm excitation peak, providing a ratiometric method for the detection of ROS (Figure 2H) [41]. Overexpression of DAP3–mCherry in roGFP stable cell lines led to a significant increase in ROS levels across all the measured subcellular localizations (Figure 2I–K). Particularly, the mitochondrial matrix exhibited the highest increase in ROS (1.4-fold) (Figure 2I,L), followed by the IMS (1.2-fold) (Figure 2J,M) and cytosol (1.1-fold) (Figure 2K,N), indicating that the ROS may originate from the mitochondrial matrix and subsequently spread into the IMS and cytosol after mitochondrial membrane damage or via diffusion.

Overall, the above results demonstrate that DAP3 can lead to mitochondrial fragmentation, loss of ΔΨm, ATP decline, and oxidative stress, indicating that DAP3 potentially triggers cell death via the intrinsic pathway.

2.2 Overexpression of DAP3 Induces Cell Death via Intrinsic Pathway

It is well established that mitochondrial outer membrane permeabilization, regulated by Bax and Bak, is a key event of intrinsic cell death [42]. To investigate the involvement of DAP3 in mediating the intrinsic cell death, we took advantage of using Bax/Bak knockout (KO) MEF cell lines. Approximately 24 h after transfection with DAP3–BFP or vector–BFP, the cells were harvested and stained with propidium iodide (PI)/Annexin V for cell death measurement. As shown in Figures 3A,B, and S2A, the deletion of both Bak and Bax (double KO, DKO) reduced cell death from approximately 26.1 ± 3.3 to 13.1 ± 4.5%. Interestingly, while Bak KO effectively inhibited the cell death rate, Bax KO had protective effect, with approximately 31.5 ± 5.4% of cells undergoing cell death after DAP3 overexpression. These findings reveal that DAP3 induces cell death via the intrinsic pathway in a Bak-dependent manner.

To further clarify that the observed cell death is apoptosis rather than another form of cell death, the apoptotic inhibitor Z-VAD-FMK (Z-VAD) was added to the cell culture. Surprisingly, the addition of Z-VAD failed to inhibit the cell death in WT cells overexpressing DAP3, with a cell death rate of 22.2 ± 1.8% in Z-VAD-treated cells (Figure 3D,H) compared with 25.8 ± 6.9% in the DMSO-treated control (Figure 3C,G). This phenomenon is reminiscent of necroptosis, a form of programmed necrotic cell death that can be triggered when apoptosis is blocked. To confirm this, the necroptosis blocker necrostatin-1 (Nec-1) was added into the cell culture, either alone (Figure 3E,I) or in combination with Z-VAD (Figure 3F,J). The combination of Nec-1 and Z-VAD significantly blocked cell death after DAP3 overexpression, lowering the percentage of apoptotic cells to 8.66 ± 2.4% (Figure 3F,J). In contrast, Nec-1 alone did not prevent the cell death (21.7 ± 6.3%) (Figure 3E,I). Therefore, these results indicate that DAP3 primarily induces intrinsic apoptosis, while in the presence of an apoptotic inhibitor, the mode of cell death changes to necroptosis.

An interesting observation is that although Z-VAD alone failed to prevent cell death in WT cells, it showed protective effects in Bax KO cells, reducing the cell death to 10.0 ± 1.9% (Figure 3D,H). Thus, the combination of Z-VAD treatment and Bax depletion had a similar effect as the combination of Z-VAD and Nec-1, suggesting that Bax may play a role in regulating necroptosis.

The Bcl-2 family proteins are known to regulate cell death by direct interaction with Bax and Bak. To examine whether DAP3 induces cell death via a similar mechanism, we performed a coimmunoprecipitation (co-IP) assay. However, no detectable interaction between DAP3 and Bax/Bak was observed, as shown in Figure S2B,C. To further investigate the DAP3's effects, we examined mitochondrial morphology in Bak KO cells after DAP3 overexpression. The results revealed that while Bak deletion was sufficient to block cell death, it did not prevent mitochondrial fragmentation (Figure S2D). Likewise, the ROS accumulation was not alleviated in Bak KO cells (Figure S2E,F). These data suggest that Bak may act as a downstream effector in DAP3-induced mitochondrial fragmentation and ROS burst.

2.3 DAP3 Induces Mitochondrial Impairment Independently of the Canonical Fusion–Fission Machinery

Mitochondrial dynamics, regulated by fusion and fission proteins such as Mfn1/2, OPA1, Drp1, Fis1, and Mff, plays a crucial role in controlling mitochondrial morphology. Enhanced mitochondrial fragmentation can arise from either increased fission or decreased fusion. To elucidate how DAP3 triggers mitochondrial fragmentation, we examined whether DAP3 overexpression affects the levels of the fusion–fission proteins. As shown in Figure 4A, the protein levels of these mitochondrial dynamics mediators remained unchanged following DAP3 overexpression. To further explore the mechanism, we performed a co-IP assay to investigate potential associations between DAP3 and these fusion–fission proteins. Interestingly, DAP3 was found to interact with Fis1 and Mfns (Figure 4B).

To determine whether the fission mediator Fis1 is responsible for the DAP3-induced mitochondrial fragmentation, we utilized a Fis1 KO cell line (Figure S3A). Overexpression of DAP3 in Fis1 KO cells still resulted in substantial mitochondrial fragmentation (46.23 ± 9.05%; Figure 4C,D), thus excluding the involvement of Fis1. Although Drp1 showed no interaction with DAP3 (Figure 4B), it is widely characterized as a key regulator of mitochondrial fission. We, therefore, investigated its potential involvement. Depletion of Drp1 using siRNA significantly inhibited mitochondrial fission, leading to a highly fused mitochondrial network (Figures S3A and 4E; GFP control). However, overexpression of DAP3–GFP still induced mitochondrial fission in approximately 22.9 ± 4.6% of the Drp1 knockdown (KD) cells (Figure 4F), albeit with a slight decrease compared with that in WT cells (37.0 ± 9.2%) (Figure 1D). These results suggest that Fis1 and Drp1 are largely dispensable for DAP3-induced mitochondrial fragmentation.

Mfns mediate mitochondrial fusion by forming homodimers or heterodimers that bring two mitochondria together. Given the observed interaction between DAP3 and Mfns, it is conceivable that DAP3 might disrupt Mfns’ dimer formation, thereby promoting mitochondrial fragmentation. To explore this hypothesis, we carried out co-IP assays to examine whether DAP3 interferes with Mfns interaction. As shown in Figure 4G, overexpression of DAP3 did not reduce the quantity of Mfn2 pulled down by Mfn1. Besides, comparing the precipitated protein levels with the input protein levels revealed that the Mfn1–DAP3 interaction was notably weaker than that of Mfn1–Mfn2. To further validate that DAP3 does not disrupt the interaction between Mfns, we performed a co-IP assay of Mfns in the presence of a substantial quantity of purified recombinant DAP3 (Figure S3B,C). Notably, the addition of 200 µg of DAP3 protein to the cell lysate (containing 400 µg of total protein) did not impair the Mfn1–Mfn2 interaction. We also tried to rescue the mitochondrial fragmentation by overexpressing Mfn1. However, Mfn1 overexpression resulted in a clustered, grape-like mitochondrial network that was distinct from the original tubular structure (Figure S3D), consistent with previous reports [43-45]. Consequently, it remains challenging to ascertain whether Mfn1 overexpression effectively rescues DAP3-induced mitochondrial fragmentation.

Although DAP3 is a mitochondrial ribosomal protein, previous research has associated it with the extrinsic cell death pathway, which takes place in the cytosol [20, 21]. Additionally, very recent studies have identified novel functions of DAP3 in RNA editing and splicing, revealing that a portion of DAP3 localizes in the nucleus [7, 46]. This raises a pivotal question regarding the subcellular localization of DAP3 during its role in mitochondrial fragmentation. To address this, we constructed an N-terminal GFP-tagged DAP3 (GFP–DAP3). The existence of GFP obstructs the recognition of the N-terminal mitochondrial targeting sequence, resulting in the retention of GFP–DAP3 in the cytosol. Remarkably, cells expressing GFP–DAP3 did not exhibit mitochondrial fragmentation (Figure S3E), suggesting that DAP3's mitochondrial localization is indispensable for its function.

In summary, these data suggest that DAP3 overexpression induces mitochondrial fragmentation through a mechanism that operates independently of the canonical fusion–fission mechanisms.

2.4 Miro1 Mediates the DAP3-Induced Mitochondrial Impairment

Miro1, a mitochondrial outer membrane protein, regulates mitochondrial movement by forming a transporting complex with kinesin heavy chain isoform 5 (KIF5) and trafficking kinesin-binding protein 1 and 2 (TRAK1/2) [47]. Recent research has identified Miro1 as a mediator of mitochondrial fragmentation independent of the canonical fusion–fission machinery [48, 49]. Therefore, we sought to investigate the involvement of Miro1 in DAP3-induced mitochondrial fragmentation. To this end, we generated stable KD HeLa cell lines using lentiviral short hairpin RNA (shRNA) constructs targeting either Miro1 gene or LacZ as a control. We then overexpressed DAP3–GFP in these cell lines. As shown in Figures 5A,B and S4A, Miro1 KD significantly reduced the percentage of cells with fragmented mitochondria to 4.1 ± 3.8%, compared with 28.7 ± 9.7% in shCtrl cells. The FRAP assay showed similar recovery curves after DAP3–GFP and vector–GFP overexpression in shMiro1 cell lines, further confirming that Miro1 is responsible for the DAP3-induced mitochondrial fragmentation (Figure 5C–E).

Given the pivotal role of Miro1 in mitochondrial movement, we next investigated whether the fragmentated mitochondria in DAP3-overexpressing cells have altered mobility. To evaluate mitochondrial movement, we acquired two images at a 10-s interval, labeling them green and magenta, respectively. The two images were subsequently overlaid, with increased mitochondrial mobility indicated by a greater nonoverlapping mitochondrial area [50]. The percentage of nonoverlapping areas was measured for quantitative comparison. As shown in Figure 6A,B, fragmented mitochondria in DAP3–GFP-overexpressing cells displayed significantly more nonoverlapping areas (35.5 ± 0.5%) compared with controls (22.8 ± 0.9%), indicating increased mitochondrial mobility. These results support the notion that Miro1 takes effects to induce mitochondrial fragmentation in DAP3-overexpressing cells. In contrast, overexpression of fission protein (Fis1 or Mff) or deletion of fusion proteins (Mfn1/2) resulted in extensive mitochondrial fragmentation but did not increase mitochondrial mobility as DAP3 did (Figures 6A–D and S4B), consistent with our previous conclusion that DAP3 induces mitochondrial fragmentation independently of the canonical fusion–fission machinery.

As a Ca²⁺-binding protein, Miro1 regulates the mitochondrial transport in a Ca²⁺-dependent manner [51]. High levels of intracellular calcium have been demonstrated to suppress Miro1-mediated mitochondrial movements [52, 53]. To further substantiate that DAP3 promotes mitochondrial movements via Miro1, we triggered a surge in the intracellular Ca²⁺ levels by adding ionomycin, a calcium ionophore, to the cell culture. The addition of ionomycin immediately suppressed mitochondrial movement (Figure 6E,F). In line with this, KD of Miro1 also attenuated the increase in mitochondrial motility induced by DAP3 overexpression (Figure 6G,H). These results collectively confirm that Miro1 promotes mitochondrial movement in DAP3-overexpressing cells.

Previous studies have demonstrated that mitochondrial fragmentation can lead to increased mitochondrial ROS generation [54, 55]. Therefore, we sought to investigate whether Miro1-mediated mitochondrial fragmentation contributes to ROS generation in DAP3-overexpressing cells. To assess mitochondrial ROS levels, we overexpressed DAP3–GFP and vector–GFP in shCtrl and shMiro1 cells, followed by MitoSox dye staining. As shown in Figures 7A,B and S4C, KD of Miro1 attenuated the increase of mitochondrial ROS, revealing that Miro1 is necessary for DAP3 to promote ROS generation. Moreover, as mitochondria-derived ROS is one of the key factors for apoptosis [56, 57], we examined whether Miro1 KD could alleviate DAP3-triggered cell death. shCtrl or shMiro1 HeLa cells were transfected with DAP3–BFP or vector–BFP, followed by a cell death assay. As shown in Figure 7C,D, shMiro1 cells exhibited a lower apoptosis rate (16.1 ± 5.1%) compared with shCtrl cells (28.5 ± 5.3%) after DAP3 overexpression, confirming that Miro1-mediated mitochondrial fragmentation and the following ROS burst contribute to DAP3-induced cell death.

2.5 DAP3 Regulates Intracellular Ca²⁺ Levels

Miro1 acts as a Ca²⁺ sensor that regulates mitochondrial transport [51]. Perturbations of cytosolic Ca²⁺ levels ([Ca²⁺]_c) have been reported to cause Miro1-mediated mitochondrial fragmentation [48]. To examine whether DAP3 overexpression affects intracellular Ca²⁺ levels, the [Ca²⁺]_c and mitochondrial Ca²⁺ level ([Ca²⁺]_m) were evaluated using the ratiometric probe GEM–GECO1 [58], which specifically localizes to the cytosol and mitochondria matrix. Ca²⁺ binding to the GEM–GECO1 probe causes a shift in emission wavelength from ∼510 toward ∼450 nm (Figure 8A), thus allowing detection of changes in Ca²⁺ levels by comparing the emission ratio at the two wavelengths. As depicted in Figure 8B–D, DAP3–mCherry overexpression led to increased [Ca²⁺]_m and decreased [Ca²⁺]_c in cells with fragmented mitochondria, indicating that DAP3 can induce alterations in basal Ca²⁺ levels.

We next investigated whether DAP3 accelerates mitochondrial Ca²⁺ uptake, thereby leading to the observed changes in basal Ca²⁺ levels. Mitochondrial Ca²⁺ uptake was triggered by adding ionomycin in the presence of 500 nM CaCl₂. As shown in Figure 8E, DAP3–mCherry overexpression markedly enhanced mitochondrial Ca²⁺ uptake, suggesting that DAP3 regulates intracellular Ca²⁺ homeostasis by promoting mitochondrial Ca²⁺ uptake. Interestingly, a recovery of mitochondrial morphology was observed during the measurement (Figure S5A), supporting our earlier finding that DAP3 induces mitochondrial fragmentation via altering intracellular Ca²⁺ levels.

To delve further into the mechanism by which DAP3 regulates intracellular Ca²⁺, we searched the BioGRID protein interaction database. Interestingly, a recent study on mitochondrial interaction networks revealed a potential interaction between DAP3 and the MCU, a subunit of the calcium channel in the inner mitochondrial membrane that controls the mitochondrial Ca²⁺ influx [59]. To verify the interaction, we coexpressed Flag-tagged DAP3 and V5-tagged MCU in HEK 293T cells, followed by co-IP with anti-Flag beads. As demonstrated in Figure 8F, DAP3 successfully pulled down MCU, suggesting that DAP3 physically associates with MCU. We then constructed a stable HeLa cell line with MCU KD using shRNA (Figure S5B). In this cell line, mitochondrial fragmentation was significantly reduced (17.6 ± 2.5%) compared with the control (36.9 ± 5.6%; Figure 8G,H). Furthermore, KD of MCU attenuated the DAP3-induced increase in mitochondrial ROS (Figure 8I). Together, these findings indicate that the MCU channel is essential for DAP3-induced changes in basal Ca²⁺ levels and mitochondrial stress.

4.1 Flow Cytometry Analysis

For the ROS assay, cells were harvested and washed once with PBS. Mitochondrial and cytosolic ROS were measured by incubating live cells in cell imaging buffer (140 mM NaCl, 2.5 mM KCl, 1.8 mM CaCl₂, 1.0 mM MgCl₂, 20 mM HEPES, pH 7.4) supplemented with 5 µM MitoSox or DHE, respectively, at 37°C for 20 min. After the incubation period, cells were washed twice and resuspended in imaging buffer for flow cytometry assay (Beckman; Cytoflex LX). For cell death assay, all cells, including floating and attached, were collected and washed once with PBS. Subsequently, they were incubated in binding buffer supplemented with Annexin V/propidium iodide (PI) (BioLegend; 640906). About 10,000 events were recorded for each experiment and the results were analyzed using CyExpert software. Cells showing negative staining were regarded as living cells, whereas Annexin V staining indicated apoptosis. Cells showing positive staining for both Annexin V and PI were classified as late apoptotic or necrotic cells.

4.2 Fluorescence and Confocal Microscopy Imaging

Fluorescence and confocal microscopy imaging were conducted following previously established protocols [83]. Cells were seeded onto a glass-bottom dish and transfected with 1.5 µg of plasmid. Live cells adhered to the dish were observed and imaged using the Olympus FV3000 confocal microscopy system or the Olympus IX83 microscopy system with a chamber to maintain a temperature of 37°C and 5% CO₂. ImageJ/Fiji software was used for quantitative analyses of the mitochondrial size, number, and fluorescence. Mitochondrial segmentation was performed using the Mitochondria Analyzer plugin in ImageJ [84], and morphological parameters were measured by particle analysis with the segmented mitochondria. The levels of ROS, mitochondrial potential, and calcium levels were calculated by quantifying the fluorescence values of mitochondrial or cytosolic probes per cell.

MitoTracker Red CMXRos staining was performed at a concentration of 100 nM for 20 min. TMTM staining was performed at a concentration of 50 nM for 20 min. MitoSox and DHE staining were performed at concentration of 5 µM for 20 min. Hoechst 33342 staining of nuclei was performed at a concentration of 1 µg/mL in combination with other dyes.

The basal level of Ca²⁺ was measured in the cell culture medium at 37°C. GEM–GECO1 was excited at 405 nm with emission filter settings as approximately 450 and 510 nm. For imaging of ionomycin-induced mitochondrial Ca²⁺ influx, cells were washed twice and incubated with 1 mL HEPES (20 mM)-buffered Hanks balanced salt solution (HHBSS, Mg²⁺-free) containing 500 nM CaCl₂. Time-lapse images were collected with 6.5 s intervals for 4 min. Approximately 45 s after the initial recording, 1 mL HHBSS containing 500 nM CaCl₂ and 5 µM ionomycin was added to the dish using a dropper. The changes in mitochondrial and cytosolic Ca²⁺ levels were monitored by the fluorescence of GEM–GECO1.

For the measurement of mitochondrial ATP levels, mitochondria-localized ATeam1.03-nD/nA probe was excited at 405 nm, and emissions around 475 and 527 nm were scanned.

4.3 FRAP-Based Mitochondrial Fusion Assay

A circular region of interest (ROI) was selected and photobleached by a single-pulse 488 nm laser. Time-lapse images were recorded with 0.94 s intervals. Ten cells were analyzed in each replicate. The fluorescence intensity of mito-RFP in the ROIs at each time point was recorded and normalized to the initial fluorescence.

4.4 co-IP and Western Blot

co-IP experiments were performed as previously described. Briefly, HEK 293T cells were transfected with the target plasmids in 10 cm dishes. After 24 h, the cells were washed and harvested in 1 mL mammalian lysis buffer (50 mM Tris–HCl, 10% glycerol, 1% Triton X-100, 100 mM NaCl, and 0.5 mM MgCl₂, pH 7.4) supplemented with protease inhibitors, including 10 µg/mL of aprotinin, 1 mM phenylmethylsulfonyl fluoride (PMSF), 1 µM pepstatin, and 10 µM leupeptin, followed by centrifugation under 21,130×g for 20 min at 4°C. The supernatant was incubated with 5 µL M2 anti-flag beads (Sigma) for 2 h at 4°C. After incubation, the beads were washed three times and suspended in 50 µL 2× SDS loading buffer for 10 min at 95°C to denature the precipitated proteins.

For the Western blot assay, protein samples were separated on 10–15% SDS-PAGE and transferred to PVDF membranes. The membranes were blocked with 5% (W/V) skimmed milk in TBST buffer (20 mM Tris, 150 mM NaCl, 0.1% Tween20, pH 7.4). Incubation with primary antibodies was performed at 4°C overnight, followed by incubation with secondary antibodies (Santa Cruz) at room temperature for 1–2 h. The membranes were imaged using a gel-doc Amersham Imager 600 system after brief incubation with ECL substrate.

4.5 Protein Expression and Purification Using E. coli System

To purify DAP3 for the co-IP assay, the pET42b (+) plasmid harboring the DAP3 sequence was transformed into BL21 E. coli. Bacterial cultures were shaken at 200 rpm and 37°C until the OD₆₀₀ value reached 0.6. DAP3 expression was initiated after the addition of 0.4 mM IPTG, followed by further cultivation at 180 rpm and 18°C overnight. Bacteria were then harvested by centrifugation (5000×g for 10 min), and resuspended in lysis buffer (0.4% Sarkosyl, 25 mM Tris–HCl, 500 mM NaCl, 10 mM β-mercaptoethanol, 1% Tween-20, pH 7.0) containing 1 mM PMSF, lysed by sonication, and centrifuged at 20,000×g for 1 h at 4°C. The supernatant containing DAP3 protein was purified by HisTrap HP column (Cytiva), HiTrap SP HP column (Cytiva), and Superdex 75 gel filtration column (Cytiva) according to the previous study [85]. The protein concentration was measured using the Bio-Rad Protein Assay Kit (#5000006).

4.6 Statistical Analysis

Statistical analysis was conducted using GraphPad Prism software. Student's t-test was used for comparing means between two groups, while one-way ANOVA was used for comparing means among three or more groups. For assays with two independent variables, two-way ANOVA was used. A p value < 0.05 was considered statistically significant