Aldehyde dehydrogenase 1 a1 regulates energy metabolism in adipocytes from different species

Abstract

Background: Survival and longevity of xenotransplants depend on immune function and ability to integrate energy metabolism between cells from different species. However, mechanisms for interspecies cross talk in energy metabolism are not well understood. White adipose tissue stores energy and is capable of mobilization and dissipation of energy as heat (thermogenesis) by adipocytes expressing uncoupling protein 1 (Ucp1). Both pathways are under the control of vitamin A metabolizing enzymes. Deficient retinoic acid production in aldehyde dehydrogenase 1 A1 (Aldh1a1) knockout adipocytes (KO) inhibits adipogenesis and increases thermogenesis. Here we test the role Aldh1a1 in regulation of lipid metabolism in xenocultures.

Methods: Murine wide-type (WT) and KO pre-adipocytes were encapsulated into a poly-L-lysine polymer that allows exchange of humoral factors <32kD via nanopores. Encapsulated murine adipocytes were co-incubated with primary differentiated canine adipocytes. Then, expression of adipogenic and thermogenic genes in differentiated canine adipocytes was detected by real-time polymerase chain reaction (PCR). The regulatory factors in WT and KO cells were identified by comparison of secretome using proteomics and in transcriptome by gene microarray.

Results: Co-culture of encapsulated mouse KO vs WT adipocytes increased expression of peroxisome proliferator-activated receptor gamma (Pparg), but reduced expression of its target genes fatty acid binding protein 4 (Fabp4), and adipose triglyceride lipase (Atgl) in canine adipocytes, suggesting inhibition of PPARγ activation. Co-culture with KO adipocytes also induced expression of Ucp1 in canine adipocytes compared to expression in WT adipocytes. Cumulatively, murine KO compared to WT adipocytes decreased lipid accumulation in canine adipocytes. Comparative proteomics revealed significantly higher levels of vitamin A carriers, retinol binding protein 4 (RBP4), and lipokalin 2 (LCN2) in KO vs WT adipocytes.

Conclusions: Our data demonstrate the functional exchange of regulatory factors between adipocytes from different species for regulation of energy balance. RBP4 and LCN2 appear to be involved in the transport of retinoids for regulation of lipid accumulation and thermogenesis in xenocultures. While the rarity of thermogenic adipocytes in humans and dogs precludes their use for autologous transplantation, our study demonstrates that xenotransplantation of engineered cells could be a potential solution for the reduction in obesity in dogs and a strategy for translation to patients.

Keywords: adipocytes; adipogenesis; beige; bright adipocytes; brown; energy balance; neutrophil gelatinase-associated lipocalin; obesity; transplantation; vitamin A; xenocultures.

Figure 1. Canine adipocytes express higher levels of differentiation markers PPARg and Fabp4, and reduced levels of Ucp1 than non-differentiated ASC

ASC were isolated from falciform adipose depot of a female Mastiff dog and cultured for 3 passages. The resulted primary non-differentiated pre-adipocytes (ND) were differentiated in adipogenic medium (DMI and DM II) with or without 1µM BRL for 14 days (D for Differentiated). RNA expression of Pparg (A), Fabp4 (B), Atgl (C), Ucp1 (D) and LOC479623 (E) were measured in non-differentiated and differentiated canine adipocytes using Q-PCR assay and TaqMan probes. The values were normalized by TBP and represent the mean ± SD (n=3), one-way ANOVA.

Figure 2. Co-culture of encapsulated WT vs KO murine preadipocytes alter gene expression in canine adipocytes

Schematic presentation of an experiment in which encapsulated murine preadipocytes were co-cultured and differentiated with adherent canine preadipocytes (A). The upper panel shows encapsulated murine preadipocytes floating in media under light microscopy (20x magnification). In schematics, differentiated canine adipocytes were depicted as adherent cells containing yellow lipid droplets. Numerous murine adipocytes (blue ellipses) are floating in the media within microcapsules (circles) Canine preadipocytes (Mastiff dog) were differentiated in adipogenic medium with encapsulated WT or Aldh1a1^−/− (KO) preadipocytes for 14 days (B-F). Expression of Pparg (B), Fabp4 (C), Atgl (D), Ucp1 (E) and LOC479623 (F) were measured by Q-PCR assay TaqMan. The values were normalized by TBP and represent the mean ± SD (n=3), one-way ANOVA.

Figure 3. Encapsulated KO cells reduce lipid accumulation in canine adipocytes

Canine preadipocytes from a Mastiff dog were differentiated in adipogenic medium with encapsulated WT (upper panel in A, black bar in B) or KO preadipocytes (lower panel in A, patterned bars in B) for 14 days. Encapsulated WT and KO cells were removed. Neutral lipid accumulation in lipid droplet is shown in differentiated canine adipocytes using Oil Red O staining (A). Left panels showed enlarged (40x magnification) areas (square) in the right panels A. Lipid accumulation was quantified in isopropanol extract from these Oil red stained canine adipocytes by detecting absorbance of dissolved oil red staining at 510nm (B). Data are shown as mean ± SD (n=3), one-way ANOVA.

Figure 4. Co-culture of encapsulated murine WT and KO preadipocytes influence expression of genes in canine adipocytes isolated from a German Shepherd dog

Canine preadipocytes were differentiated in adipogenic medium with encapsulated WT or KO preadipocytes for 14 days (A-E). Expression of Pparg (A), Fabp4 (B), Atgl (C), Ucp1 (D) and LOC479623 (E) were measured by Q-PCR assay TaqMan. The values were normalized by TBP and represent the mean ± SD (n=3), one-way ANOVA.

Figure 5. Encapsulated KO cells reduce lipid accumulation in canine adipocytes from a German Shepherd dog

Canine preadipocytes (German shepherd) were differentiated in adipogenic medium with encapsulated WT (upper panel in A, black bar in B) or KO preadipocytes (lower panel in A, patterned bars in B) for 14 days. Encapsulated WT and KO cells were removed. Neutral lipid accumulation in lipid droplet is shown in differentiated canine adipocytes using Oil Red O staining (A). Left panels showed enlarged (40x magnification) areas (square) in the right panels A. Lipid accumulation was quantified in isopropanol extract from these Oil red O stained canine adipocytes by detecting absorbance of dissolved oil red staining at 510nm (B). Data are shown as mean ± SD (n=3), one-way ANOVA.

Figure 6. KO adipocytes secret higher levels of lipocalins RBP4 and LCN2 than WT adipocytes

DIGE was performed to identify different proteins (<50 kD) in the secretome from differentiated WT and from KO adipocytes (n=3 per condition). A representative 2D gel with protein spots that significantly changed by at least 1.5-fold (A). Principal component analysis (PCA) grouped of all proteins according to their similarity of standardized expression to determine how much variability between the proteins from WT and KO secretome (grey identification numbers, ID) (B). Red ID numbers showed secreted proteins that significantly changed by at least 1.5-fold between WT and KO secretome. Among standardized expression profiles, the high variability protein clusters were identified in WT (pink dots) and KO (blue dot) secretomes. There were 2 major clusters of proteins exhibiting KO-induced downregulation and 3 major clusters of proteins exhibiting KO-induced upregulation. RBP4 protein abundance in secretomes from WT (black circle) and KO mouse adipocytes (white square) (C). Different Rbp4 expression in WT and KO preadipocytes measured using Affymetrix GeneChip Mouse Gene ST 2.0 arrays (GEO file: ‘QS wild type and Aldh1a1 KO preadipocytes 2015’, n=3) (D). RBP4 protein abundance in secretomes from WT (black circle) and KO mouse adipocytes (white square) (E). Different Lcn2 expression in WT and KO preadipocytes measured using Affymetrix GeneChip Mouse Gene ST 2.0 arrays (F).