Abstract

Objectives

   Nuclear receptor interacting protein 1 (NRIP1) suppresses energy
   expenditure via repression of nuclear receptors, and its depletion
   markedly elevates uncoupled respiration in mouse and human adipocytes.
   We tested whether NRIP1 deficient adipocytes implanted into obese mice
   would enhance whole body metabolism. Since β-adrenergic signaling
   through cAMP strongly promotes adipocyte thermogenesis, we tested
   whether the effects of NRIP1 knock-out (NRIP1KO) require the cAMP
   pathway.

Methods

   NRIP1KO adipocytes were implanted in recipient high-fat diet (HFD) fed
   mice and metabolic cage studies conducted. The Nrip1 gene was disrupted
   by CRISPR in primary preadipocytes isolated from control vs adipose
   selective GsαKO (cAdGsαKO) mice prior to differentiation to adipocytes.
   Protein kinase A inhibitor was also used.

Results

   Implanting NRIP1KO adipocytes into HFD fed mice enhanced whole-body
   glucose tolerance by increasing insulin sensitivity, reducing
   adiposity, and enhancing energy expenditure in the recipients. NRIP1
   depletion in both control and GsαKO adipocytes was equally effective in
   upregulating uncoupling protein 1 (UCP1) and adipocyte beiging, while
   β-adrenergic signaling by CL 316,243 was abolished in GsαKO adipocytes.
   Combining NRIP1KO with CL 316,243 treatment synergistically increased
   Ucp1 gene expression and increased the adipocyte subpopulation
   responsive to beiging. Estrogen-related receptor α (ERRα) was
   dispensable for UCP1 upregulation by NRIPKO.

Conclusions

   The thermogenic effect of NRIP1 depletion in adipocytes causes systemic
   enhancement of energy expenditure when such adipocytes are implanted
   into obese mice. Furthermore, NRIP1KO acts independently but
   cooperatively with the cAMP pathway in mediating its effect on
   adipocyte beiging.

   Keywords: Diabetes, Obesity, Adipose tissue, Adrenergic signaling,
   CRISPR, Implant

Highlights

     * •
       NRIP1KO adipocyte implants enhance systemic energy expenditure in
       HFD fed mice.
     * •
       The cAMP signaling pathway is dispensable for the upregulation of
       UCP1 by NRIP1KO.
     * •
       ERRα is dispensable for the regulation of thermogenesis by NRIP1.

1. Introduction

   Cell therapies have been successfully used in multiple fields of
   medicine such as oncology [[55][1], [56][2], [57][3], [58][4]], yet are
   still far from clinical application in type 2 diabetes and obesity.
   Nonetheless, preclinical studies indicate the strong potential of
   adipocytes to be used as cell therapy vehicles, as adipocyte
   dysfunctions are linked with the pathogenesis of disrupted systemic
   glucose and lipid homeostasis [[59][5], [60][6], [61][7], [62][8],
   [63][9], [64][10]]. Furthermore, numerous studies have shown efficacy
   of implanting thermogenic brown adipose tissue into obese mice to
   improve glucose tolerance [[65][11], [66][12], [67][13], [68][14]].
   More recently, CRISPR-based gene editing methods have been used to
   enhance the thermogenic phenotype of white adipocytes, which in turn
   have proved effective in enhancing systemic metabolic health when
   implanted into HFD fed mice [[69]15,[70]16]. Such studies from our
   laboratory targeted the thermogenic repressor Nrip1 gene to upregulate
   UCP1 and other genes of the thermogenic program in mouse and human
   adipocytes prior to implantation [[71]15,[72]17,[73]18]. While in other
   cell types, NRIP1 is known to display both corepressor and coactivator
   functions, in adipocytes it has been reported to interact with and
   negatively regulate PPARα and PPARγ as well as retinoid receptors and
   estrogen-related receptors that are transcriptional regulators of fat
   accumulation and thermogenesis [[74][19], [75][20], [76][21], [77][22],
   [78][23], [79][24]]. NRIP1-depleted adipocytes display increases in
   insulin-induced glucose uptake, fatty acid oxidation, energy
   expenditure and oxygen consumption, particularly in response to
   catecholamines [[80]15,[81]17,[82]18,[83]20,[84]22,[85]24]. Adipocyte
   implants have been reported to form fat pads that are fully innervated
   and vascularized within about 9 weeks from implantation
   [[86]25,[87]26]. In our studies, all implants including those
   containing control (NTC) and NRIP1KO adipocytes were observed to
   successfully engraft and are retrieved at the end of the experiments.
   No signs of rejection, other inflammatory or immune reactions or
   necrotic tissue have been observed. These CRISPR-engineered human and
   mouse adipocytes continued to consistently display the beige-like
   phenotype when retrieved four months following their successful
   implantation and engraftment [[88]15].

   In mammalian brown adipose tissue, a classical mechanism of thermogenic
   induction in response to activation of the sympathetic nervous system
   by cold exposure is mediated via β-adrenergic signaling. This pathway
   involves release of norepinephrine which stimulates G-protein coupled
   receptors, including the β1, β2 receptors and the adipocyte-specific β3
   receptor that act through the Gsα protein [[89]10,[90][27], [91][28],
   [92][29], [93][30], [94][31]]. Activation of this pathway leads to
   stimulation of adenylyl cyclase and cAMP synthesis, which activates
   protein kinase A (PKA) and phosphorylation of downstream targets that
   induce thermogenic programming. These targets include phosphorylation
   and activation of p38 protein kinase that acts to phosphorylate the
   transcription factor ATF2 [[95]32,[96]33]. In addition, PKA directed
   phosphorylation of mTORC1/RAPTOR has been shown to mediate a required
   signal or signals that upregulate UCP1 transcription [[97]34]. Other
   regulators that are responsive to β-adrenergic stimulation in
   adipocytes include PGC1a, and PPARα [[98]28,[99]31,[100]35,[101]36].
   Chronic activation of this β-adrenergic pathway in adipocytes leads to
   the upregulation of UCP1 as well as many of the other genes that are
   also responsive to NRIP1KO [[102]15,[103]17,[104]18,[105]20,[106]22].
   Evidence that adipocytes depleted of NRIP1 show increased
   responsiveness to the effects of β-adrenergic stimulation has also been
   reported [[107]15,[108]19]. These considerations raise the important
   mechanistic question whether the cAMP pathway downstream of
   β-adrenergic signaling is required for and contributes to the
   upregulation of UCP1 expression induced by NRIP1KO. A related question
   is what nuclear receptor or receptors are directly repressed by NRIP1
   in its role of downregulating thermogenesis? NRIP1 is known to act as a
   corepressor of many nuclear receptors, and it was shown that ERRα is
   required for its regulation of glucose transport in adipocytes
   [[109]18]. ERRα and ERRγ are known to also function in liver and brown
   fat to enhance whole body glucose metabolism [[110]37,[111]38].

   The aim of the present studies was to address these questions on the
   relationships between NRIP1 signaling and the cAMP and ERRα pathways in
   mediating the thermogenic response. In addition, we addressed the
   related question of whether the increased thermogenesis observed in
   NRIP1KO adipocytes translates into increased energy expenditure in mice
   after implantation. Our results revealed that 1.) Implantation of
   NRIP1KO adipocytes into HFD fed mice does upregulate systemic energy
   expenditure, and 2.) neither upstream (Gsα) nor downstream (PKA)
   elements of the cAMP signaling pathway are required for the actions of
   NRIP1KO to upregulate UCP1 and thermogenic genes in adipocytes.
   Similarly, NRIP1 regulates thermogenesis independently of ERRα.

2. Material and methods

2.1. Animals and diets

   All animal work was approved by the University of Massachusetts Medical
   School Institutional Animal Care Use Committee (IACUC protocol
   no.202200006 to Michael P. Czech). Mice were housed at 20–22 °C on a
   12-hour light/12-hour dark cycle with ad libitum access to food and
   water. C57BL/6J male mice were purchased from Jackson Laboratory for
   implant studies. C57BL/6J (Jackson Laboratory) male mice were bred for
   primary preadipocyte cultures. To generate adipocyte-specific Gsα
   (Gnas) knock-out mice that were then used for iWAT SVF (inguinal white
   adipose tissue stromal vascular fraction) isolation, homozygous
   Gsα−Flox/Flox animals were crossed to Adiponectin-Cre mice (B6; FVB-Tg
   (Adipoq-cre)1Evdr/J) as previously described [[112]28,[113]30].
   Briefly, 10-week-old male mice arrived and were allowed to acclimate
   for a week prior to any procedures. Baseline glucose tolerance tests
   were performed three days before implantation procedures. Mice were
   implanted under anesthesia with edited primary mouse adipocytes between
   10 and 11 weeks of age. Anesthesia was achieved with an anesthesia
   vaporizer chamber with a continuous flow 500 cc/minute of 0[2] with 3%
   (v/v) isoflurane for induction and 1.5% (v/v) for maintenance. After
   the cell injections, animals were allowed to recover. Mice were
   maintained on a chow diet for 6 weeks following implantation and then
   were placed on a 60 kcal% fat diet (Research Diets, D12492i) for the
   remainder of the study. Glucose tolerance tests were performed after
   16-hour overnight fasting with intraperitoneal injection of 1 g/kg D
   (+) glucose. Insulin tolerance tests (ITTs) were performed with 0.75
   IU/kg which did not yield an appropriate response and then with 1 IU/kg
   after a 6-hour daytime fasting. Additional ITTs with increased insulin
   dosage could not be performed because of the time-sensitivity of the
   follow-up of the cohorts. As previously described, body composition was
   noninvasively measured in mice using 1H-magnetic resonance spectroscopy
   (Echo-MRI; Echo Medical System). Energy balance was noninvasively
   assessed with indirect calorimetry for 72 h using metabolic cages (TSE
   Systems, Inc) [[114]39].

2.2. Primary mouse preadipocyte isolation, culture and differentiation to
primary adipocytes

   3 to 4 week-old C57BL/6J male mice or Gnas Flox–Flox or Gnas Flox-
   Adiponectin Cre mice were euthanized and inguinal fat tissue was
   harvested (including lymph node) and placed in HBSS buffer (Gibco
   #14025) plus 3% (w/v) bovine serum albumin (BSA) (American
   Bioanalytical) for SVF isolation and the primary preadipocytes were
   grown, transfected and differentiated as described previously
   [[115]15,[116]18]. Briefly, the cells were cultured to sub-confluence
   in complete media containing DMEM/F12 media (Gibco #11330), 1% (v/v)
   Penicillin/streptomycin, 10% (v/v) Fetal bovine serum (FBS) (Atlanta
   Biologicals #S11550), 100 μg/mL Normocin (Invivogen #Ant-nr-1). For
   cAMP activation, the mature adipocytes were treated with 1 nM (unless
   otherwise specified) of CL 316,243 disodium salt compound (Tocris
   Biotechne, Catalog No. 1499) resuspended in ddH[2]O following serial
   dilutions. For 1 mM cAMP treatment, dibutyryl-cAMP sodium salt was
   purchased (Tocris Biotechne, Catalog No. 1141) and was resuspended in
   ddH[2]O to a concentration of 100 mM. The same volume of ddH[2]O was
   added in the cell cultures as vehicle.

2.3. Transfection of primary preadipocytes with ribonucleoprotein (RNP)

   For RNP transfection, we used the Neon Transfection System 10 μL or
   100 μL Kits (ThermoFisher, #MPK10096) depending on the expected
   targeted cell population and we prepared a mix consisting unless
   otherwise specified of sgRNA 4 μM (Synthego or IDT DNA) purified
   SpyCas9 protein 3 μM (PNA Bio, #CP02) or 3xNLS-SpCas9 [[117]40]
   (purified by the Scot Wolfe laboratory) in Buffer R provided in the
   Neon Transfection System Kit. The cells were resuspended in
   Resuspension Buffer R to a final density as suggested by the
   manufacturer per electroporation. For delivering the RNP complex into
   primary pre-adipocytes the electroporation parameters used were voltage
   1350 V, pulse width 30 ms; number of pulses 1 according to our previous
   optimization experiments [[118]15]. The electroporated cells were
   placed in complete media immediately following transfection, expanded,
   grown to confluence, and differentiated into mature adipocytes for
   downstream applications. Nrip1 gene was targeted with sgRNAs 5′
   ACAGGCTGTTGCCAGCATGG 3′ and 5′ ATAAGGTTTGGAGTCACGTC 3′ for the
   implantation and in vitro experiments, respectively. Esrra gene was
   targeted with sgRNA 5′ GATCACCAAGCGGAGACGCA 3'.

2.4. Implantation of primary mouse adipocytes

   For the implantation of edited adipocytes on day 6 post
   differentiation, we followed the protocol and procedures previously
   described [[119]15]. Briefly, the procedure included washing with 1X
   PBS and dissociation of the cells from the plate using a
   trypsin-collagenase solution. The detached cells are pelleted at 300
   rcf for 10 min at room temperature, washed twice with 1x PBS, then kept
   on ice for a brief time until implantation. Each pellet deriving from
   1 × 150 mm fully confluent plate was mixed with matrigel (Corning®
   Matrigel® Growth Factor Reduced Basement Membrane Matrix, Phenol
   Red-free, LDEV-free # 356231) up to a total volume of 500 μL and the
   cell and matrigel suspension was subcutaneously injected into the
   anesthetized recipient C57BL/6J mouse using a 20G needle. Each mouse
   received 2 × 150 mm plates of fully differentiated murine adipocytes in
   two subscapular, bilateral injections. [120]Figure 1B, and C include
   but are not limited to glucose tolerance test all time points and body
   mass data in week 0, 9, 12 and 15 from NTC and NRIP1KO implant
   recipients that have been previously reported [[121]15], in addition to
   new data from non – implanted mice and body masses from all groups in
   week 3 and 6. Due to the inclusion of the third NT group, we now
   employed 2-way ANOVA for the statistical analysis of these data.

Figure 1.

   [122]Figure 1
   [123]Open in a new tab

   Mouse NRIP1KO adipocyte implants, unlike control adipocyte implants,
   improve glucose tolerance, adiposity, insulin sensitivity and increase
   energy expenditure in recipient mice on fat-enriched diet. A. Schematic
   graph and timeline of the cohort study and follow up. B. Time-course
   areas under the curve of glucose tolerance tests with 1 g/kg of
   d-glucose following a 16-hour overnight fasting in mice without
   implants (NT), mice with implants of non-targeted control adipocytes
   (NTC), and in mice with implants of NRIP1KO adipocytes (NRIP1KO). 0
   week is the baseline pre-implantation measurement. The dashed line
   separates the measurements on chow (left) and on high-fat diet (right)
   C. Time-course total body mass follow up of recipient mice. D. Plasma
   insulin measurements following a 16-hour overnight fasting. One value
   in the NRIP1KO mice was below the detecting point of the assay so it
   was replaced with the minimum detectable value (0.09). E. Insulin
   tolerance test curves F. areas under the curve following a 6-hour
   daytime fasting with 1IU/kg of insulin administered i.p. in week 16 of
   study (10 weeks on HFD). G. Body composition studies with 1H-MRS at
   week 16 following implantation in NT and NTC mice (Control) and NRIP1KO
   implant recipient mice (NRIP1KO). H. Average food intake in a three-day
   metabolic cage study in NT and NTC mice (Control) and NRIP1KO implant
   recipient mice (NRIP1KO). I. Physical activity in a three-day metabolic
   cage study in NT and NTC mice (Control) and NRIP1KO implant recipient
   mice (NRIP1KO) full-day, and in the light and dark hours separately. J.
   O[2] consumption in a three-day metabolic cage study in NT and NTC mice
   (Control) and NRIP1KO implant recipient mice (NRIP1KO) full-day, and in
   the light and dark hours separately. K. Average daily energy
   expenditure in a three-day metabolic cage study in NT and NTC mice
   (Control) and NRIP1KO implant recipient mice (NRIP1KO) full-day, and in
   the light and dark hours separately. L. Energy expenditure time-point
   plot during a 72-hour metabolic cage study in NT and NTC mice (Control)
   and NRIP1KO implant recipient mice (NRIP1KO) full-day, and in the light
   and dark hours separately. In A-C,E,F, N^NT = 9, N^NTC = 9,
   N^NRIP1KO = 11. In D, N^NT = 9, N^NTC = 9, N^NRIP1KO = 9. In G-L,
   N^NT = 3, N^NTC = 4, N^NRIP1KO = 5Bars denote mean and error bars
   denote mean +S.E.M. In panels B,C,F statistical analysis was performed
   with 2-way ANOVA multiple comparisons Dunnett's test where each
   condition was compared to NTC as control. In panel D, and E unpaired
   2-tailed t-test was used between NT vs NTC and NTC vs NRIP1KO. In G-I,
   statistical analysis was performed with unpaired 2-tailed t-test was
   used between control and NRIP1KO in each condition and in J-K, general
   linear regression analysis was used via the CalR software with the
   total body mass used as adjusted covariant. ∗p < 0.05, ∗∗p < 0.01,
   ∗∗∗p < 0.001.

2.5. DNA harvest from cells and tissue

   At two distinct time-points, 72 h following transfection and after
   primary adipocyte differentiation (approx. 10–13 days post
   transfection), genomic DNA was isolated from the transfected cells
   using DNA QuickExtract™ Buffer (Lucigen) in adherence to the
   manufacturer's instructions.

2.6. Indel analysis by TIDE and ICE

   Genomic DNA was PCR amplified for downstream analysis using locus
   specific primers designed with MacVector 17.0 and purchased from IDT
   DNA and Genewiz, spanning the region 800 bp around the expected double
   strand break (DSB). For the PCR amplification, Kappa 2x Hot start HiFi
   mix was used and PCR products were purified using the Qiagen QIAquick
   PCR purification kit following the manufacturer's instructions and
   submitted to Genewiz for Sanger Sequencing. Sanger sequencing trace
   data were analyzed with TIDE and ICE webtools
   ([124]http://shinyapps.datacurators.nl/tide/,
   [125]https://ice.synthego.com/#/) that decipher the composition of
   indels created at the sites of DSBs [[126]41,[127]42]. All primers used
   for the assessment of on-target editing are listed in Supplementary
   Table in the [128]Appendix.

2.7. RNA isolation

   Transfected cells were harvested for RNA by removing media and washing
   once with 1xPBS and adding Trizol reagent to lyse the cells. The
   protocol for RNA isolation was performed according to manufacturer's
   instruction with the following modifications: 1 μL of Glycol blue
   (Invitrogen #AM9516) was added to the isopropanol to precipitate the
   RNA and was stored overnight at −20 °C. The isolated RNA was
   resuspended in RNase-free water, then treated with recombinant DNase I
   (DNA-free DNA removal kit, Ambion #1906) according to the
   manufacturer's instructions. RNA concentrations were determined by
   Nanodrop 2000.

2.8. RT-PCR

   For this procedure 1 μg of RNA was used in 20 μL reaction with Bio-Rad
   iScript cDNA synthesis kit according to manufacturer's instructions for
   cDNA synthesis. cDNA was diluted up to 100–120 μL with ddH[2]O and 5 μL
   of the diluted cDNA was used as template for RT-PCR with Bio-Rad Sybr
   Green Mix and gene-specific primers for a final concentration of 0.3 μM
   primers. Expression of genes was determined with the ΔΔCT method using
   36B4 as housekeeping gene. All primers used for RT-PCR are listed in
   Supplementary Table in the [129]Appendix.

2.9. Protein isolation

   Adipocytes grown in culture dishes were washed once with 1 x PBS at
   room temperature, followed by adding boiling 2% SDS (w/v) with 1 X HALT
   protease inhibitors and scraping to lyse the cells. The cell lysates
   were sonicated at 60% amplitude with a probe sonicator tip for 10 s at
   room temperature. Protein concentration of cell lysates was determined
   using a bicinchoninic acid kit (BCA Protein Assay Kit, Pierce #23227).
   Protein samples were prepared for western blotting at a final
   concentration of 1 mg/mL protein, 1x Laemmli loading buffer (BioRad)
   with 2.5% (v/v) β-Mercaptoethanol, followed by placing in a heat block
   at 100 °C for 10 min.

2.10. Western blotting

   Protein lysates were run on 4–15% Gradient Mini-Protean TGX stain-free
   pre-cast protein gels, followed by transferring the proteins to
   Nitrocellulose membrane. Unless otherwise stated, a total of 20 μg of
   protein lysate was loaded per well and nitrocellulose membranes were
   blocked using 5% (w/v) Non-fat milk in Tris buffered saline with 0.1%
   (v/v) Tween-20 (TBST) for 4 h at room temperature. Primary antibody
   incubations were carried out in 5% (w/v) BSA in TBST at the following
   antibody concentrations: UCP1-Abcam#10983; Anti-pHSL (Ser563) Cell
   Signaling Technology Cat# 4139; Anti-phospho PKA substrate (RRXS∗/T∗)
   Cell Signaling Technology Cat# 9624; ERRa-antibody Cell Signaling Cat#
   13826, 1:1000; Vinculin Cell Signaling Technology Cat# 18799, OXPHOS
   Abcam #110413, Tubulin Sigma Cat# T5168. Blots and primary antibodies
   were incubated overnight with a roller mixer at 4 °C. Membranes were
   washed with TBST prior to secondary antibody incubations.
   HRP-conjugated secondary antibodies were diluted with 5% BSA (w/v) in
   TBST at 1:10,000 for 45 min at room temperature with constant shaking.
   Membranes were washed in TBST, followed by incubating with Perkin Elmer
   Western Lightning Enhance ECL. The Bio-Rad Chemi-Doc XRS was used to
   image the chemiluminescence and quantifications were performed using
   the system software, or Image J v.1.51.

2.11. Immunofluorescence preparation and imaging

   For the immunofluorescence, a previously described protocol was used
   [[130]43]. Briefly, the cells were seeded on cover slips in the setting
   of the experiment. At the time-point as specified by the experimental
   design, the wells were washed with PBS 1X and fixed with 4% (v/v)
   paraformaldehyde in PBS for 10 min at room temperature. After three
   short washes with PBS 1X, 100% ice cold methanol was applied for 10 min
   at 4 °C for permeabilization. After 3 short washes with PBS 1X, normal
   goat blocking solution was used for blocking for 60 min. The primary
   antibody incubation (UCP1-Abcam#10983, 1:100; Perilipin 1- ThermoFisher
   #GT2781, 1:100) was overnight at 4 °C followed by three short washes
   with PBS 1X and incubation with the secondary antibody Alexa Fluor-594
   goat anti-rabbit and Alexa Fluor-488 goat anti-mouse (1:500,
   Invitrogen) for 1 h at room temperature in the dark. Two washes with
   PBS 1X were performed and in the last wash, DAPI 1 μg/mL in PBS and was
   applied on the cells for 5 min to stain the nuclei. The coverslips were
   mounted in Prolong Gold (Invitrogen) and were allowed to dry overnight
   prior to imaging. Images were taken with Leica TCS SP8 confocal
   microscope using LAS-X software and 3–4 randomized z-stacks of 7 images
   per stack were taken from each condition. The images were analyzed with
   Fiji/ImageJ2 v. 2.9.0/1.53t and manual blinded cell count was performed
   by three independent investigators on the same images.

2.12. RNA-sequencing

   RNA was isolated on day 7 post differentiation after 7 h of stimulation
   with CL 316,243 or vehicle as described above. The samples underwent
   DNase treatment, and 2 μg of total RNA was submitted to GENEWIZ for
   standard RNA sequencing (Illumina HiSeq). The data were acquired
   demultiplexed in the form of FASTQ files. Alignment and quantification
   of gene expression levels were performed using the DolphinNext RNA-seq
   pipeline (revision 6) [[131]44]. Parameters for the pipeline were set
   to remove Illumina 3' adapter sequences using trimmomatic software
   (v0.39) with seed mismatches set to 2, a palindrome clip threshold of
   30 and a simple clip threshold of 5. The DolphinNext pipeline used RSEM
   (v1.3.1) to align RNA-Seq reads to a mouse reference transcript (using
   the RSEM reference STAR and and Bowtie genomes) to estimate gene
   expression levels. DESeq2 software (v 1.28.1) was then used on these
   expression levels to find differentially expressed genes between two
   groups of samples. We set the parameters test = “LRT” (Likelihood Ratio
   Test), fitType = “parametric”, betaPrior = FALSE, and reduced = ∼1 (to
   compare to the control group). padj was set to 0.05 and a minFC = 1.3
   was specified. Once differentially expressed genes were found, enriched
   pathways were identified using the WebGestalt online tool and a p value
   cut-off of 0.05 [[132][45], [133][46], [134][47], [135][48],
   [136][49]].

2.13. Plasma insulin measurements

   Whole blood was collected from the mouse recipients from their
   submandibular vein under anesthesia which was administered as described
   in 2.1 on week 15 following implantation after an overnight 16-hour
   fasting. The blood was placed in EDTA containing tubes and the plasma
   was collected after a 15-minute centrifugation at 2,000 rcf and 4 °C.
   For insulin measurements, ELISA kit (Crystal Chem) used with 40 μL per
   sample and the assay was read with a Luminex 200 machine (Millipore).
   For the data analysis, the measurement found to be below the detectable
   cut-off, was replaced with 0.09 ng/mL which is the lowest detectable
   value of 0.09 ng/mL by the kit.

2.14. Statistics and reproducibility

   For the statistical analysis of the data, GraphPad Prism 9.5.1 was used
   unless otherwise specified (see 2.12, 2.15) and the statistical methods
   used for each analysis are described in the corresponding figure
   legend. The data are presented as means ± S.E.M. p value < 0.05 was the
   threshold for statistical significance.

2.15. Databases, software, and online tools

   For the mapping of exons on the Nrip1 and Esrra genes we used
   IGV_2.5.3. For the Design of sgRNAs we used a combination of the Broad
   Institute sgRNA designer, CHOPCHOP and the online sgRNA checkers by
   Synthego and IDT. For the design of genomic DNA primers, we used
   MacVector 17.0. For the statistical analysis and data presentation of
   the metabolic cage studies we used the CalR app [[137][50], [138][51],
   [139][52]]. For the design of RT-PCR primers we used Primer Bank
   ([140]https://pga.mgh.harvard.edu/primerbank/). For the browning
   probability potential, the unnormalized mapped read counts were applied
   in the ProFAT online tool [[141]53]. For the transcription factor
   analysis, the online available software ChEA3 was used [[142]54]. For
   the data graphing, we used Prism GraphPad 9 unless otherwise specified.

3. Results

3.1. NRIP1KO adipocytes improve insulin sensitivity, adiposity, and energy
expenditure in recipient mice on HFD

   To study the systemic metabolic effects of Nrip1 disruption in
   adipocytes, we used CRISPR RNP delivery [[143]15] to disrupt Nrip1 in
   murine preadipocytes which were then differentiated into mature NRIP1KO
   or NTC adipocytes, respectively. Recipient mice received implants of
   NRIP1KO or NTC adipocytes in the subcutaneous subscapular space
   bilaterally while another age and gender-matched group remained without
   implants (NT). All mice underwent baseline body mass and fasting
   glucose tolerance tests prior to randomization and implantation and
   these measurements were repeated every 3 weeks. Six weeks following
   implantation, the mice were placed on HFD to induce glucose intolerance
   and they were followed up for a total of 16 weeks ([144]Figure 1A).
   While on chow diet, no differences were observed among groups in
   glucose tolerance.

   After 6 and 9 weeks on HFD, the NRIP1KO implant recipient mice
   exhibited significantly improved glucose tolerance compared to the NTC
   implanted mice and the NT control mice ([145]Figure 1B), consistent
   with our previous observations [[146]15]. There was a transient
   decrease in total body mass of the NRIP1KO implanted mice, however, by
   the end of the follow-up period there were no significant differences
   in the total body mass ([147]Figure 1C). At 16 weeks after implantation
   (10 weeks following HFD feeding), fasting plasma insulin of the mice
   implanted with NRIP1KO adipocytes was significantly lower
   ([148]Figure 1D) and insulin sensitivity as detected by insulin
   tolerance tests was improved compared to the control groups with a 14%
   reduction of the area under the curve ([149]Figure 1E, F). These data
   are consistent with the concept that implanted NRIP1KO adipocytes
   confer enhanced insulin sensitivity of liver or skeletal muscle or
   both. Since there was no weight loss at the end of the study, we
   suggest that the increased glucose uptake and the secreted factors from
   the NRIP1KO adipocytes [[150]8,[151]11,[152]15,[153]18] may be the
   causes of this beneficial effect.

   To further examine the physiological effects of the NRIP1KO adipocytes
   after implantation, we performed metabolic cage studies. Strikingly,
   despite the absence of total body mass effects, the mice that received
   NRIP1KO implants had increased lean mass and decreased fat mass without
   differences in food uptake or physical activity ([154]Figure 1G-I).
   After a 3-day follow-up, oxygen consumption and energy expenditure were
   increased in the NRIP1KO implant recipients both during the light and
   dark cycles ([155]Figures 1J-L). Interestingly, no differences were
   observed between the NT and the NTC implant recipients
   ([156]Supplementary Fig. 1), suggesting that the described effects are
   solely attributable to the consequences of NRIP1 protein depletion in
   the implanted adipocytes rather than the implants themselves. Taken
   together, NRIP1KO adipocyte implants protect against development of
   insulin resistance in response to HFD, increase lean to fat mass ratio
   and enhance energy expenditure in mice. The mechanisms that mediate
   these effects are unknown, but NRIP1KO adipocytes themselves show a
   marked increased uncoupled respiration and energy expenditure
   [[157]15,[158]17]. Whether this cell autonomous increase in oxygen
   consumption is enough to contribute to the increase in systemic energy
   expenditure in NRIP1KO implanted mice is unclear and will be a topic
   for future investigation.

3.2. NRIP1 depletion induces thermogenic programming independent of the cAMP
pathway

   In murine adipocytes the β[3]-adrenergic receptor in response to
   agonist CL 316,243 ([159]Figure 2A) acts via the heterotrimeric protein
   Gsα,β,γ to activate its effector adenylyl cyclase, producing cAMP from
   ATP [[160]9,[161]27,[162]55,[163]56]. Our RNA sequencing data showed
   that in Nrip1-depleted adipocytes, Gnas mRNA encoding Gsα protein is
   upregulated [[164]15,[165]57] ([166]Supplementary Fig. 2), a finding
   that was confirmed by RT-PCR ([167]Figure 2C). Based on this data, we
   investigated the hypothesis that the UCP1 upregulation in response to
   NRIP1 depletion is partly due to enhancement of the cAMP pathway in
   adipocytes. To test this hypothesis, we isolated SVF from iWAT of
   cAdGsaΚΟ mice and cAdGsa^Fl/Fl mice. We then transfected the
   preadipocytes with NTC or Nrip1 targeting CRISPR RNP to create single
   and double knock-out cells for NRIP1KO and GsαKO. ([168]Figure 2B,
   [169]Supplementary Fig. 2). The system was validated by stimulation
   with CL 316,243 and dibutyryl-cAMP, confirming the lack and presence of
   PKA substrate phosphorylation and UCP1 response, respectively
   ([170]Supplementary Fig. 2).

Figure 2.

   [171]Figure 2
   [172]Open in a new tab

   The adipocyte browning phenotype induced by NRIP1KO in vitro is
   independent of cAMP signaling. A. cAMP activation by β adrenergic
   agonists can be blocked by depletion of the GTPase Gsα in adipocytes
   which is normally activated by endogenous or chemical ligands such as
   norepinephrine or CL 316,243 (see Introduction and references for