Identification of a molecular resistor that controls UCP1-independent Ca2+ cycling thermogenesis in adipose tissue

Abstract

Adipose tissue thermogenesis contributes to energy balance via mitochondrial uncoupling protein 1 (UCP1) and UCP1-independent pathways. Among UCP1-independent thermogenic mechanisms, one involves Ca²⁺ cycling via SERCA2b in adipose tissue; however, the underlying molecular basis remains elusive. Here, we report that an endoplasmic reticulum (ER) membrane-anchored peptide, C4orf3 (also known as another regulin [ALN]), uncouples SERCA2b Ca²⁺ transport from its ATP hydrolysis, rendering the SERCA2b-C4orf3 complex exothermic. Loss of C4orf3/ALN improved the energetic efficiency of SERCA2b-dependent Ca²⁺ transport without affecting SERCA2 expression, thereby reducing adipose tissue thermogenesis and increasing the adiposity of mice. Notably, genetic depletion of C4orf3 resulted in compensatory activation of UCP1-dependent thermogenesis following cold challenge. We demonstrated that genetic loss of both C4orf3 and Ucp1 additively impaired cold tolerance in vivo. Together, this study identifies C4orf3 as the molecular resistor to SERCA2b-mediated Ca²⁺ import that plays a key role in UCP1-independent thermogenesis and energy balance.

Keywords: Ca(2+) cycling; UCP1-independent; energy balance; obesity; thermogenesis.

Figure 1.. Direct recording of microsomal thermogenesis in adipose tissue

(A) Schematic of organelle thermogenesis assays using ITC. Isolated mitochondria (20 μg) were subjected to ITC in the presence of glutamate, pyruvate, malate, and succinate (complexes I and II) or palmitoyl-carnitine and malate (β-oxidation). To detect microsomal thermogenesis, isolated microsomes (20 μg) were subjected to ITC in the presence of ATP at 1 mM and free Ca²⁺ at 10 μM. Thapsigargin was added to determine SERCA-dependent signals. (B) Heat rate (μJ s⁻¹) in isolated mitochondria from the iBAT of wild-type (WT) and UCP1 KO male mice following the addition of palmitoyl-carnitine and malate. Heat rate was normalized by mitochondrial protein contents (μJ min⁻¹ μg⁻¹). n = 4. (C) Contributions of mitochondrial vs. microsomal thermogenesis in the iBAT and inguinal WAT. Mitochondrial thermogenesis was measured in the presence of glutamate, pyruvate, malate, and succinate (complexes I + II) or in the presence of palmitoyl-carnitine and malate (β-oxidation). Thapsigargin-independent heat rate in microsomes was excluded as background. n = 4. (D) Heat rate in isolated mitochondria from IngWAT of WT and UCP1 KO male mice. Mitochondrial thermogenesis (complexes I + II or β-oxidation) was measured and normalized by mitochondrial protein contents (μJ min⁻¹ μg⁻¹). n = 5. (E) Heat rate in isolated microsomes from the inguinal WAT of WT and UCP1 KO male mice in the presence of ATP and free Ca²⁺ (pCa 6.0). The heat rate was normalized by microsome protein contents (μJ min⁻¹ μg⁻¹). n = 4. (F) Immunoblotting of SERCA2 and ER-localized protein calreticulin as a loading control in the isolated microsomes from WT and UCP1 KO male mice. n = 3. Statistic (B and D–F): unpaired t test. Bars represent the mean and error shown as SEM.

Figure 2.. C4orf3 promotes Ca²⁺ cycling thermogenesis by reducing the energetic efficiency of SERCA2-dependent Ca²⁺ transport

(A) FPKM of indicated SERCA-binding peptides in isolated beige adipocytes of mice. n = 3. (B) Relative C4orf3/ALN protein expression levels in microsomes from IngWAT of UCP1 KO mice at the corresponding temperature. Calreticulin was used as a loading control. n = 3. Statistic: one-way ANOVA with Tukey’s post hoc HSD test. (C) Protein interaction between SERCA2b and endogenous C4orf3/ALN protein in beige adipocytes. Immunoprecipitants of a SERCA2 complex (FLAG-tagged) were immunoblotted using a polyclonal antibody for C4orf3. Inputs were included in the immunoblotting. (D) Endogenous C4orf3 protein expression in IngWAT of C4orf3^CRISPRi mice and control male mice. β-actin was used as a loading control. n = 3. (E) Intracellular Ca²⁺ flux assay in IngWAT-derived adipocytes from control and C4orf3^CRISPRi mice. Primary adipocytes were differentiated on collagen-coated glass-bottom dishes in which intercellular Ca²⁺ levels were determined by using the Fluo-8 dye. Control, n = 50; C4orf3^CRISPRi, n = 50. Statistic: two-way ANOVA with Šídák’s multiple comparisons test. (F) Ca²⁺ uptake in isolated microsomes from the inguinal WAT of male C4orf3^CRISPRi mice and control mice at pCa²⁺ 6.0. n = 5. (G) SERCA ATP hydrolysis assay in isolated microsomes from (F). (H) The bioenergetic efficiency of SERCA2 was calculated by Ca²⁺ uptake per ATP hydrolysis from (E) and (F). (I) Heat rate in isolated microsomes from IngWAT of control and C4orf3^CRISPRi male mice in the presence of ATP and Ca²⁺ (pCa 6.0). Thapsigargin was added to calculate SERCA-dependent thermogenesis. n = 5. (J) Relative mRNA levels of C4orf3 in adult dCas9-KRAB male mice 2 weeks after direct injection into IngWAT of AAV expressing scrambled control or gRNA targeting C4orf3. n = 4. (K) Heat rate in isolated microsomes from the inguinal WAT of dCas9-KRAB male mice 2 weeks after AAV injection into the inguinal WAT. n = 4. Statistic (D and F–K): unpaired t test. Bars represent the mean and error shown as SEM.

Figure 3.. A structural mechanism by which C4ORF3 alters the Ca²⁺ transport efficiency of SERCA2

(A) Side (left) and top (right) views of the SERCA2b-C4ORF3 complex structure predicted by AlphaFold3. The A-, N-, P-, and TM domains of Ca²⁺-unbound SERCA2b are shown in orange, pink, yellow, and green, respectively. C4ORF3/ALN is shown in blue. (B) Protein interaction interface of SERCA2b to C4orf3 as determined by the Turbo-ID proximity-labeling proteomics. Turbo-tag was fused to the N terminus (the cytoplasmic side) of C4orf3. Orange, TM domain; yellow, actuator domain; pink, nucleotide-binding domain; cyan, P-domain. (C) Cross-linking assays of SERCA2b with C4orf3/ALN^W48C in isolated microsomes at indicated SERCA2 catalytic states. The cross-linker BMH was added to induce the SERCA2-C4ORF3 complex formation. The assays were performed in three independent biological samples. Bottom: catalytic cycle of SERCA to transport Ca²⁺ into the ER lumen. Indicated compounds were used to induce the specific catalytic states of SERCA. (D) The amino acid sequence alignment of C4ORF3. The lower panel shows the schematics of a C4ORF3 mutant containing only the transmembrane (TM) domain and three mutants (mutants A, B, and C) lacking the evolutionarily conserved domains (shown in red boxes). (E) Ca²⁺ uptake in isolated microsomes from cells co-expressing SERCA2b together with empty vector control and indicated C4ORF3 constructs at pCa²⁺ 6.0. n = 3. (F) SERCA ATP hydrolysis assay in microsomes from (E). (G) ITC-based thermogenesis assays from microsomes in (E). Statistic (E–G): one-way ANOVA with Tukey’s post hoc HSD test. Bars represent the mean and error shown as SEM.

Figure 4.. C4orf3 is required for UCP1-independent thermogenesis in vivo

(A) Cold tolerance test of male C4orf3^CRISPRi mice and littermate controls. Mice kept at room temperature were exposed to cold (6°C) for indicated time points. n = 11. Statistic: two-way ANOVA with Šídák’s multiple comparisons test. (B) OCR in iBAT and IngWAT of male C4orf3^CRISPRi mice and control male mice following cold exposure. A subset of isolated tissues was stimulated with NE. n = 12 per group for iBAT; 10 per group for Ing WAT. Statistic: Mann-Whitney U test. (C) Relative mRNA expression of thermogenic genes in IngWAT of male mice following cold exposure. n = 5. (D) Mitochondrial respiration in IngWAT of male mice following cold exposure. JO₂ at indicated states was measured following substrate injection. n = 5. (E) Heat rate in isolated mitochondria (state 4, complexes I and II) from IngWAT of male mice following cold exposure. Data are normalized by mitochondrial protein contents. n = 4. (F) Cold tolerance test of male UCP1 KO mice and DKO mice (UCP1 KO × C4orf3^CRISPRi). The mice were chronically treated with the β3-adrenergic receptor agonist CL316,243 for 5 days to stimulate beige fat biogenesis in both groups. n = 8 per group. Statistic: two-way ANOVA with Šídák’s multiple comparisons test and individual unpaired t tests. (G) Ca²⁺ uptake in isolated microsomes from IngWAT of male DKO mice and control UCP1 KO mice at pCa²⁺ 6.0. n = 4. (H) SERCA ATP hydrolysis assay in isolated microsomes from (G). The values were normalized by microsomal protein contents. (I) The bioenergetic efficiency of SERCA2 was calculated by Ca²⁺ uptake per ATP hydrolysis from (G) and (H). (J) Heat rate in isolated microsomes from IngWAT of UCP1 KO and DKO male mice at pCa 6.0. Thapsigargin was added to calculate SERCA-dependent thermogenesis. n = 4 per group. Statistic (C–E and G–J): unpaired t test. Bars represent the mean and error shown as SEM.

Figure 5.. Compensatory responses to C4orf3 and UCP1 loss

(A) GO pathway enrichment analysis for 276 genes that were uniquely upregulated in IngWAT of DKO mice (UCP1 KO × C4orf3^CRISPRi) relative to other genotypes following cold exposure. n = 6 per group. (B) Relative mRNA levels of indicated genes in IngWAT of WT control, UCP1 KO, C4orf3^CRISPRi, and DKO mice (UCP1 KO × C4orf3^CRISPRi) following cold exposure at 6°C for 5 h. Data represented as Z score heatmap for each gene in each sample representing quantitated value. GO terms and pathways for each gene are listed on the right. n = 6. (C) The respiratory exchange ratio of mice at 30°C. n = 7 per group. (D) Quantification of respiratory exchange ratio in (C). Statistic: unpaired t test.

Figure 6.. Loss of C4orf3 resulted in increased adiposity and systemic insulin resistance

(A) Relative mRNA expression of indicated genes in the subcutaneous adipose tissue of people in Table 1. Lean: n = 6 female, n = 3 males; overweight: n = 5 female, n = 4 males; obese classes 1 and 2: n = 6 female, n = 3 males; obese class 3: n = 9 female, n = 1 male; obese with T2D: n = 5 female, n = 4 males. Statistic: one-way ANOVA with Tukey’s post hoc HSD test. (B) Fat mass and lean mass of male C4orf3^CRISPRi and littermate control mice at 30°C on a regular chow diet. n = 10. (C) Daily food intake of mice in (B). (D) Body weight of mice in (B) at 20 weeks old. (E) Indicated tissue weight of male mice on a regular diet at 30°C. n = 11. (F) H&E staining and adipocyte size of IngWAT (n = 917 cells for control, 623 cells for C4orf3^CRISPRi) and epididymal WAT (n = 699 cells for control, 364 for C4orf3^CRISPRi). Male mice at 30°C on a regular chow diet (n = 3 per group). Statistic: unpaired t test with Welch’s correction. (G) Insulin tolerance test of male C4orf3^CRISPRi mice and controls after 4 h fasting (0.5 U per kg⁻¹). n = 8. (H) Fasting insulin levels in male C4orf3^CRISPRi mice and control mice. n = 6. (I) Blood triglyceride (TG) levels in male C4orf3^CRISPRi mice and littermate control mice. n = 6. Statistic (C–E and G–I): unpaired t test. Bars represent the mean and error shown as SEM.