Analyzing human knockouts to validate GPR151 as a therapeutic target for reduction of body mass index.

Novel drug targets for sustained reduction in body mass index (BMI) are needed to curb the epidemic of obesity, which affects 650 million individuals worldwide and is a causal driver of cardiovascular and metabolic disease and mortality. Previous studies reported that the Arg95Ter nonsense variant of GPR151, an orphan G protein-coupled receptor, is associated with reduced BMI and reduced risk of Type 2 Diabetes (T2D). Here, we further investigate GPR151 with the Pakistan Genome Resource (PGR), which is one of the largest exome biobanks of human homozygous loss-of-function carriers (knockouts) in the world. Among PGR participants, we identify eleven GPR151 putative loss-of-function (plof) variants, three of which are present at homozygosity (Arg95Ter, Tyr99Ter, and Phe175LeufsTer7), with a cumulative allele frequency of 2.2%. We confirm these alleles in vitro as loss-of-function. We test if GPR151 plof is associated with BMI, T2D, or other metabolic traits and find that GPR151 deficiency in complete human knockouts is not associated with clinically significant differences in these traits. Relative to Gpr151+/+ mice, Gpr151-/- animals exhibit no difference in body weight on normal chow and higher body weight on a high-fat diet. Together, our findings indicate that GPR151 antagonism is not a compelling therapeutic approach to treatment of obesity.

Introduction

Obesity, defined as a body mass index (BMI) of >30 kg/m2, is a major global health concern. In 2015, 7.1% of global deaths were attributable to high BMI [1]. Predictions estimate that half of the world’s population will be obese by 2030 [2]. By 2030 in the United States alone, 25% of the population may be severely obese, as defined by BMI>35 kg/m2 [3]. Being overweight or obese leads to a steep increase in all-cause mortality [4], largely through increased risk for Type 2 Diabetes (T2D) [5], nonalcoholic fatty liver disease (NAFLD) [6], and cardiovascular disease (CVD) [1]. Reduction in BMI appears to remit many obesity-associated disorders [7].

Given the mortality, morbidity and public cost associated with obesity, there is a strong interest in identifying drug targets for sustained reduction in BMI. To date, two particular therapies that promote body weight reduction also reduce hospitalization and mortality from associated co-morbidities. Glucagon-like peptide-1 (GLP-1) agonists are incretin mimetics that can reduce body weight by up to 12%, improve glycemic status, and reduce cardiovascular events [8]. SGLT2 inhibitors prevent glucose re-uptake in kidneys, reduce body weight by 2 kg, improve glycemic status, reduce cardiovascular events, and improve kidney function [9-14]. Both GLP-1 agonists and SGLT2 inhibitors modify glucose metabolism, nonetheless, their effect on body weight is consistent with the expectation that reduction of BMI is therapeutically beneficial.

Human studies have identified numerous genetic loci associated with BMI [15]. However, many of the genes linked to these loci are either technically challenging to drug or are poorly validated as causal. For example, the fat mass and obesity-associated (FTO) gene encodes an mRNA demethylase [16-17] strongly associated with BMI [18]. However, a direct role for FTO in regulating BMI is unclear and has been called into question by studies suggesting linkage to variants in nearby genes IRX3 and IRX5 [19-21]. Loss-of-function variants in melanocortin 4 receptor (MC4R) are associated with obesity in humans [22-24]. MC4R modulators were often associated with adverse cardiovascular side effects until the identification of setmelanotide, which does not elicit these undesirable effects and was approved in the United States and Europe for treatment of genetic obesity [25]. Identification of additional BMI-associated genes may provide greater insight into the biology of body weight control and yield genes for which therapeutic modulation is tractable.

G protein coupled receptors (GPCRs) are tractable drug targets that have been associated with numerous phenotypes in human genetics studies and in mouse models [26]. GPR151 is a poorly understood, brain-specific GPCR [27] for which loss-of-function has been associated with decreased BMI [28-30]. Additionally, at GPR151, carriage of rare (alternative allele frequency [aaf] < 1%) putative loss-of-function (plof) variants and bioinformatically predicted damaging missense variants have also been associated with a decrease in BMI [30].

To date, homozygous plof carriers (human knockouts) of GPR151 have not been reported in detail. To follow up on the published genetic association, we use the Pakistan Genome Resource (PGR), which is the world’s largest biobank of human homozygous plof carriers (knockouts) identified through whole-exome sequencing of >80,000 participants. Here, we (i) identify homozygous carriers of GPR151 plof variants, including those specific to South Asia, (ii) confirm in vitro that these variants are loss-of-function, (iii) test if GPR151 knockouts are associated with BMI, T2D, or other metabolic traits, and (iv) characterize Gpr151-/- mice for body weight.

Results and discussion

Association of GPR151 plof variants with BMI and cardiometabolic events

The PGR at the Center for Non-Communicable Diseases (CNCD) in Pakistan is a large biobank of highly consanguineous participants. A medical history and numerous clinical measurements including BMI, T2D status, and myocardial infarction (MI) status, are available for most participants. In the PGR, we identified a total of 11 plof variants, with a cumulative allele frequency of 2.2%, including 48 homozygous carriers of three plof variants in GPR151 (S1 Table). Variants with homozygous carriers included Arg95Ter, Tyr99Ter and Phe175LeufsTer7. The latter two variants are highly enriched in South Asia compared to other populations. In UK Biobank, 21 homozygous plof carriers were identified among 281,852 exome-sequenced participants, and in gnomAD 13 homozygous plof carriers were found in non-south-Asian populations [31].

GPR151 is expressed from a single exon, and nonsense substitutions are more likely to escape nonsense mediated decay (NMD) in single-exon genes. To determine if truncated proteins are expressed from GPR151 variant transgenes, we transiently transfected HEK293 cells with cDNA expression constructs corresponding to variants for which we identified homozygous plof carriers in PGR. Changes in GPR151 protein sequence may alter epitopes detected by antibodies specific to GPR151 and thus confound detection by western blot. Therefore, constructs were tagged at the N-terminus with an HA epitope tag to directly compare expression of GPR151 variants in vitro. From transfected cells, total cell lysates and isolated membrane extracts were generated and evaluated by western blot (Fig 1). The reference allele (i.e. wild-type [WT]) GPR151 protein expressed at high levels and was detected in both the total lysate and in membrane extracts. In contrast, Arg95Ter and Tyr99Ter were not detectable in either extract. Although, the Phe175LeufsTer7 variant protein was detectable in both the total cell and membrane lysates, migrating at a size consistent with the expected truncation, it expressed at significantly lower levels compared to wild-type, indicating impaired stability. Phe175LeufsTer7 is missing the last three of the protein’s seven transmembrane domains, the entire cytoplasmic tail, and intracellular loop 3, which is typically critical for G protein activity. The severity and diminished expression of this truncation suggest that this variant is a loss-of-function. Our in vitro observations confirm that the homozygous GPR151 plof variants in PGR are loss-of-function alleles.

Fig 1

GPR151 variant proteins are not stably expressed.

Western blot expression of HEK293 cells transfected with pcDNA3.1 plasmids encoding GPR151 variants with N-terminal HA-tag. The expected molecular weight of wild-type (WT) GPR151 is 47 kilodaltons (kDa). Na+/K+ ATPase is shown as a loading control.

To test for associations with BMI, we analyzed plof variants individually and in a gene burden test to increase power. For our gene-burden analyses (cumulative allele frequency [caf] = 2.2%), our study was adequately powered (80% at an α = 0.05) to detect a mean difference of 0.36 kg/m2 of BMI in knockouts compared to non-carriers. Similarly, for Tyr99Ter (aaf = 1.8%), our study was adequately powered (80% at an α = 0.05) to detect a mean difference of 0.40 units of BMI in knockouts compared to non-carriers.

Unlike previous studies of GPR151, our burden tests and individual variant analyses failed to identify statistically significant associations with BMI in PGR (Table 1). Most importantly, homozygous GPR151 knockout did not confer low BMI compared to non-carriers. We analyzed knockouts across all variants versus reference carriers and did not observe significant association for either the gene burden result (knockout n = 38, p = 0.98) or Tyr99X variant (n = 34, p = 0.55). We also performed a sample size-based meta-analysis with UK Biobank GPR151 knockouts (n = 28) reported previously [28-30]. The meta-analyzed p-value remained non-significant (p = 0.67). In PGR, we also tested for associations with other relevant traits including waist-to-hip ratio, cholesterol and triglyceride levels and observed no significant associations or consistent trends (S2 Table).

Table 1

GPR151 associations with BMI.

GRCh38 chr:pos	Reference allele	Alternate allele	HGVSp	Genotype counts (RR|RA|AA)	P-value	Beta [95% CI] kg/m2 (additive)	P-value (knockouts only)	Beta [95% CI] kg/m2 (knockouts only)
5:146515831	G	A	Arg95Ter	27273|55|1	0.82	-0.126 [-1.23–0.98]
5:146515817	G	T	Tyr99Ter	26350|945|34	0.92	0.0131 [-0.24–0.27]	0.55	0.431 [-0.99–1.85]
5:146515587	CTA	C	Phe175LeufsTer7	27206|120|3	0.28	0.406 [-0.32–1.14]
Gene Burden				26150|1141|38	0.73	0.0405 [-0.20–0.28]	0.98	-0.021 [-1.37–1.33]

chr, chromosome; pos, position; HGVSp, Human Genome Variation Society protein level change; R, reference allele; A, alternate allele; kg, kilograms; m, meter; CI, confidence interval

As stated above, in PGR alone or in the combined meta-analyses, we did not observe a clinically meaningful effect on BMI in human knockouts despite a sizeable number of plof homozygous carriers for GPR151. We further examined if the previously reported weak genetic effect on BMI, largely conferred by heterozygous plof carriers, is reproducible. We meta-analyzed our results with summary statistics from the GIANT consortium [32] (Total N = 497,110; African Ancestry = 27,610; Admixed American Ancestry = 10,772; East Asian Ancestry = 8,839; European Ancestry = 449,889). With a significantly larger sample size, we replicated the Arg95Ter association with BMI (p = 6.72E-4; Beta = -0.042 [-0.063 –-0.0171]) with an additive model as used in the published study. There was no evidence of a population or study-specific effect (p-value for heterogeneity = 0.95). Hence, the original additive association between Arg95Ter and BMI, based primarily on heterozygous plof carriers, is reproducible but with a weak effect. Our findings with human knockouts across multiple plof variants indicate that complete absence of GPR151 does not further enhance this weak effect into a therapeutically meaningful reduction in BMI.

Next, we analyzed GPR151 variants to assess reduction in T2D risk (Table 2). For gene-burden analyses, our study was powered to observe an odds ratio of T2D of 0.82 or lower (80% at an α = 0.05). No individual plof variant was associated with a significant change in the risk for T2D. For the plof gene burden we observed a slight increase in T2D risk (P = 0.03, OR = 1.18 [1.02–1.37]), but this increase did not meet the threshold of significance corrected for multiple testing (threshold of p = 0.0083). Similar analyses for MI risk did not yield significant associations (S3 Table).

Table 2

GPR151 association with T2D.

GRCh38 chr:pos	Reference allele	Alternate allele	HGVSp	Genotypes cases (RR|RA|AA)	Genotypes controls (RR|RA|AA)	P-value	OR [95% CI]	P-value (knockouts only)	OR [95% CI] (knockouts only)
5:146515831	G	A	Arg95Ter	6531|30|0	33106|54|2	0.24	1.85 [0.67–5.09]
5:146515817	G	T	Tyr99Ter	6295|265|1	32008|1115|39	0.08	1.15 [0.98–1.35]	0.99	0.99 [0.34–2.87]
5:146515587	CTA	C	Phe175LeufsTer7	6532|29|0	33003|153|6	0.87	1.03 [0.68–1.59]
Gene Burden				6234|326|1	31757|1358|47	0.03	1.18 [1.02–1.37]	0.49	1.60 [0.42–6.0]

chr, chromosome; pos, position; HGVSp, Human Genome Variation Society protein level change; R, reference allele; A, alternate allele; OR, odds ratio, CI, confidence interval

In total, complete loss of GPR151 function in human knockouts was not associated with clinically meaningful changes in BMI or other traits related to obesity or metabolism.

Body weight in Gpr151 mice

To determine if GPR151 plays a role in body weight regulation in mice, we generated Gpr151 knockout (Gpr151) mice using CRISPR-Cas9. In human, pig, and mouse, GPR151 is expressed primarily in the brain [27]. In mouse, Gpr151 is expressed in the habenula, which is located in the dorsal thalamus of the brain [33]. In humans, GPR151 is expressed in the midbrain [27]. In mice, we evaluated Gpr151 expression by in situ hybridization (ISH). In wild-type (Gpr151) mice, Gpr151 mRNA expression was primarily observed in the habenular nucleus in the brain (Fig 2A), consistent with prior reports [33], and in mucosal cells of the ilium and jejunum (Fig 2C). Gpr151 mRNA expression was not detected in Gpr151 mice (Fig 2B and 2D), confirming loss of expression in knockout animals.

Fig 2

Gpr151 mice do not express Gpr151 mRNA.

(A and B) Sections of mouse brain containing the medial habenula (MHb) and lateral habenula (LHb) from Gpr151 and Gpr151 mice, respectively, stained with a riboprobe for Gpr151. (C and D) Sections of mouse intestine containing the ileum and jejunum from Gpr151 and Gpr151 mice, respectively, stained with a riboprobe for Gpr151. Black arrows indicate cells containing Gpr151 mRNA. Inset shows higher magnification of boxed region.

Next, we compared the body weights of Gpr151 and Gpr151 on standard chow and a high-fat diet. On a standard chow diet, no difference in body weights was observed between Gpr151 and Gpr151 control mice of either sex (Fig 3A and 3C). Gpr151 male mice that were fed a high-fat diet for 12 weeks weighed ~16% more than Gpr151 controls (Fig 3A). Food intake was similar between the two groups suggesting an alternative mechanism for diet-induced weight gain in Gpr151 male mice (Fig 3B). Female Gpr151 mice that were fed a high-fat diet gained weight at the same rate as Gpr151 counterparts, suggesting a sex-based difference in the response to an obesogenic diet (Fig 3C).

Fig 3

Male Gpr151 mice gain weight on high-fat diet (HFD).

(A) Body weights of male Gpr151+/+ and Gpr151 mice on a standard chow diet (chow) and high-fat diet (HFD). Data are presented as ± standard error of the mean (SEM). * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, repeated measures, two-way ANOVA followed by post-hoc analysis using Sidak’s multiple comparisons test. (B) Cumulative food intake in kcal (kilocalories) of male Gpr151+/+ and Gpr151 mice on a standard chow diet and high-fat diet. Data are presented as ± SEM. (C) Body weights of female Gpr151+/+ and Gpr151 mice on a standard chow diet and high-fat diet. Data are presented as ± SEM.

In summary, complete loss of function of GPR151 is not associated with a clinically meaningful change (i.e., > 5% change) in BMI. We identified homozygous carriers of the previously published Arg95Ter plof and additional South Asia-specific plof variants in PGR, confirmed that plof variants are unstable in vitro, and observed no statistically significant reduction in BMI in either heterozygous or homozygous carriers. The body weights of male and female Gpr151 mice were indistinguishable from Gpr151 control mice on a standard chow diet and were elevated in male Gpr151 mice on a high-fat diet without a corresponding increase in food intake. The preclinical model data indicate that the lack of association with BMI is generalizable rather than a human population-specific phenomenon. Our results highlight the importance of taking into account the effect estimates and directionality of multiple loss-of-function variants when prioritizing GWAS results for functional follow-up. In aggregate, loss of GPR151 does not affect BMI in human knockouts in a clinically meaningful way and GPR151 antagonism is likely not a compelling therapeutic strategy for BMI reduction or T2D remission in humans.

Materials and methods

Ethics statement

The Institutional Review Board (IRB) at the Center for Non-Communicable Diseases (IRB: 00007048, IORG0005843, FWAS00014490) approved the study. All participants gave written informed consent.

Variant quality control (QC) and annotation

This study included a subset of 30,833 individuals with Whole Exome Sequencing and 9,292 individuals with Whole Genome Sequencing from the Pakistan Genome Resource (PGR). Samples were sequenced at an average of 30X coverage. Samples with low allele balance for a variant (< 0.2) or low depth (< 10) were set to missing and variants that had a missingness rate > 5% were removed. We also removed variants failing VQSR filters or failing visual validation on IGV. Variants were annotated using Variant Effect Predictor [34] based on the Ensembl101 gene model. For human GPR151, we used Ensembl Transcript ENST00000311104, which is the only annotated transcript for this gene. Variants annotated as frameshift, stop gained, splice acceptor and splice donor variants are considered plof variants. The three reported homozygous GPR151 variants (Arg95Ter, Tyr99Ter, and Phe175LeufsTer7) have a call rate of 1 (i.e. zero missingness). Additionally, we filtered out plofs annotated as ‘low confidence’ according to the filtering criteria in LOFTEE [31]. Cumulative allele frequency (caf) was calculated as described [35].

Case classification

Patients were categorized as T2D cases if they satisfied any one of the following criteria: (1) Physician diagnosis at a diabetes clinic, (2) HbA1c > 6.5%, (3) use of glucose lowering medication or (4) fasting glucose > 126 mg/dl. An age of first diagnosis >22 years was used to exclude type 1 diabetes as much as possible. Patients were categorized as having had an MI as described previously [36].

Statistical analysis

Associations with BMI (kg/m2), cholesterol (mg/dl), triglycerides (mg/dl) and waist-to-hip ratio were analyzed using multivariate linear regression adjusting for age, sex, age2 and top 5 genetic principal components (PCs). Associations with T2D and MI were analyzed using logistic regression, with Firth correction as implemented in glow [37]. The genomes and exomes datasets were analyzed separately and the summary statistics were meta-analyzed using inverse variance weighted meta-analysis as implemented in METAL [38]. Power calculations were performed using Quanto v1.2 [39]. To meta-analyze our results with the GIANT Consortium, we first transformed BMI values using rank-based inverse normalization. METAL was then used to perform inverse variance weighted meta-analysis.

In vitro expression and western blot

Human codon optimized cDNAs, corresponding to GPR151 reference (wild-type; NCBI reference sequence NP_919227.2) or to nonsense mutant constructs, were cloned in pcDNA3.1(+) mammalian expression vectors with hemagglutinin (HA) epitope tags (YPYDVPDYA) appended to the amino termini. Transient transfection of adherent HEK293 cells was performed using Lipofectamine 2000 (Invitrogen) in a 6-well plate format according to manufacturer’s instructions. Cells were harvested 48–72 hours post-transfection by scraping, washed with phosphate-buffered saline (PBS), and pelleted by centrifugation at 300xg for 5 min. Whole-cell lysates were prepared from half of each sample by sodium dodecyl sulfate (SDS) extraction. Cells were resuspended in PBS containing 2.5% (weight/volume) SDS, samples were incubated at 4°C for 10 minutes with end-over-end rotation, and insoluble material was removed by centrifugation at 16,000xg for 15 minutes. Membrane fractions were isolated from remaining cell sample using the Mem-PER Plus Membrane Protein Extraction Kit (ThermoScientific) according to manufacturer’s instructions. Whole-cell lysates and isolated membrane fractions were analyzed by SDS polyacrylamide gel electrophoresis (PAGE) and Western Blot against the HA epitope to detect GPR151 expression. The following antibodies were used: anti-HA monoclonal antibody (Clone 2–2.2.14, Invitrogen 26183); anti-Na+/K+ ATPase alpha-1 antibody (clone C464.6, Sigma 05–369); and anti-mouse IgG HRP-conjugated antibody (R&D Systems HAF007).

Generation and phenotyping of Gpr151 mice

Mice lacking the Gpr151 gene were generated using the CRISPR-Cas9 system. All animal protocols were reviewed and approved by the Novartis Institutional Animal Care and Use Committee. The entire coding sequence of Gpr151 is contained on a single exon (Ensembl gene ID# ENSMUSG00000042816). Two single guide RNA (sgRNA) sequences targeting sites just upstream of the translation start codon in exon 1 (ATCAAGCTCCTCCCTGCAGA) and within the 3’ untranslated region (3’ UTR) (TCATCAATATTGCTAAGCAG) were synthesized as crRNAs for Alt-R CRISPR-Cas9 system (Integrated DNA Technologies, Coralville, IA). A ribonucleoprotein mixture of the two crRNAs complexed with tracrRNA (Integrated DNA Technologies) and Cas9 protein (PNA Bio Inc, Newbury Park, CA) was electroporated into fertilized C57BL/6J embryos. The embryos were then implanted into pseudopregnant recipients. DNA lysates were prepared from tail biopsies of F0 generation pups using KAPA Mouse Genotyping Kit according to the manufacturer’s instructions (Kapa BioSystems, Cat# KK7302). Mice were genotyped by polymerase chain reaction (PCR) using the following primers: For1 (5’-ACTTACAGACACTGTGAACAGC-3’) anneals to sequence upstream of Gpr151 exon 1, For2 (5’-TGGCTCCCAGAGTGGATAGC-3’) anneals to sequence within exon 1, and Rev1 (5’-TGCCTTTCTACTTACCAGGTTC-3’) anneals to sequence downstream of the Cas9 cut site within the 3’ UTR. For2 and Rev1 amplify a product of 614 bp corresponding to the wild-type allele, and For1 and Rev1 amplify a product of ~233 bp corresponding to the null allele. PCR conditions were as follows: denaturation at 95°C for 3 min, 35 cycles of 15 sec at 95°C, 15 sec at 60°C, and 30 sec at 72°C, then 5 min at 72°C. F0 founders were bred to C57BL/6J mice for germline transmission of mutant alleles. The null allele of the F1 founder line selected was confirmed by Sanger sequencing (GeneWiz, South Plainfield, NJ) to have a deletion of 1343 bp between the expected Cas9 cleavage sites. Heterozygous mice were interbred to generate homozygous offspring for studies.

At five weeks of age, male Gpr151+/+ and Gpr151-/- littermates were split from group housing to individual housing to monitor body weight and food consumption. Female mice remained group housed for body weight studies. At five weeks of age, all animals were provided either a standard chow diet (Purina Picolab 5053) or a high-fat diet deriving 60% kcal from fat (Research Diets D12492i), with ad libitum access to food and water. Body weights (male and female) and food intake (males only) were measured 1–2 times per week for 12 weeks following exposure to high-fat diet. Animals were maintained on a 12-hour light/dark cycle. The study size is shown in Fig 3. One cohort was run for the study. In Fig 3B, one animal in the Gpr151 HFD group was excluded from analysis because food intake data past week 8 was lost. Raw data are shown in S4 Table.

After 12 weeks of study, the mice were euthanized and brains collected to confirm Gpr151 genotype by in situ hybridization. Samples from three animals of each genotype were used. Whole brains were fixed for 48 hours in 10% neutral buffered formalin and later embedded in paraffin. Four microns sections of each brain were collected. Staining was performed on the Leica Bond RX automated staining platform using the RNAscope 2.5 LSx Reagent Kit (ACDBio, Bio-Techne– 322440) with a mouse Gpr151 probe (ACDBio, Bio-Techne—317328) following the standard RNAscope Assay [40].

List of GPR151 plof variants identified in 30,833 exomes and 9,292 genomes

(DOCX)

Click here for additional data file.

GPR151 association with waist-to-hip ratio and lipid-related biomarkers

(DOCX)

Click here for additional data file.

GPR151 association with MI

(DOCX)

Click here for additional data file.

Underlying numerical data for Fig 3

(XLSX)

Click here for additional data file.

27 Nov 2021

Dear Dr Gurtan,

Thank you very much for submitting your Research Article entitled 'Analyzing human knockouts to validate GPR151 as a therapeutic target for reduction of body mass index' to PLOS Genetics.

The manuscript was fully evaluated at the editorial level and by independent peer reviewers. The reviewers appreciated the attention to an important problem, but raised some substantial concerns about the current manuscript. Based on the reviews, we will not be able to accept this version of the manuscript, but we would be willing to review a much-revised version. We cannot, of course, promise publication at that time.

Should you decide to revise the manuscript for further consideration here, your revisions should address the specific points made by each reviewer. We will also require a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript.

If you decide to revise the manuscript for further consideration at PLOS Genetics, please aim to resubmit within the next 60 days, unless it will take extra time to address the concerns of the reviewers, in which case we would appreciate an expected resubmission date by email to plosgenetics@plos.org.

If present, accompanying reviewer attachments are included with this email; please notify the journal office if any appear to be missing. They will also be available for download from the link below. You can use this link to log into the system when you are ready to submit a revised version, having first consulted our Submission Checklist.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please be aware that our data availability policy requires that all numerical data underlying graphs or summary statistics are included with the submission, and you will need to provide this upon resubmission if not already present. In addition, we do not permit the inclusion of phrases such as "data not shown" or "unpublished results" in manuscripts. All points should be backed up by data provided with the submission.

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool.  PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

PLOS has incorporated Similarity Check, powered by iThenticate, into its journal-wide submission system in order to screen submitted content for originality before publication. Each PLOS journal undertakes screening on a proportion of submitted articles. You will be contacted if needed following the screening process.

To resubmit, use the link below and 'Revise Submission' in the 'Submissions Needing Revision' folder.

[LINK]

We are sorry that we cannot be more positive about your manuscript at this stage. Please do not hesitate to contact us if you have any concerns or questions.

Yours sincerely,

Giles S. H. Yeo

Guest Editor

PLOS Genetics

Gregory Barsh

Editor-in-Chief

PLOS Genetics

It is clear that all three reviewers have found the work of interest. However, they (in particular the second reviewer) raise some important issues, that I agree should be addressed, that would improve this manuscript. This piece of work, if true, is an important 'negative', and should hopefully set a precedent for future similar studies. So I believe that the additional rigorous analyses suggested should be undertaken.

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #1: GPR151 is a gene implicated as a GWAS hit in obesity and considered as a possible drug target for lowering BMI. In previous reports, a nonsense variant was associated with lower BMI, suggesting inhibition could be therapeutically beneficial. The authors provide multiple lines of evidence that, unfortunately, show that this gene is not a good drug target for BMI. The truncating variants considered here really do abolish protein expression in cell culture, and yet, association of truncating variants with lower BMI does not replicate in a Pakistani population, no obvious difference in BMI is observed even in homozygous knockouts in this cohort enriched for autozygosity, and no obvious difference in BMI is observed in knockout mice. Indeed, a couple of observations — higher BMI in male mice fed a high fat diet, and nominally (P=0.03) higher T2D risk in people with truncating variants — point in the opposite direction.

This is a careful and thorough paper on an important topic. Functional analyses of GWAS hits like these are vitally important and the results reported here will probably save millions of dollars in wasted preclinical drug discovery efforts. The science is well done and paper well-written, I have no major issues with anything here. A few minor points follow:

In the introductory paragraph beginning “G protein coupled receptors (GPCRs) are attractive drug targets…” it might be nice to give some background for the reader on anything known about GPR151 in particular. I looked it up in GTEx and was surprised to see it expressed at appreciable levels only in brain. Perhaps none of this matters since you found it’s not associated with BMI anyway, but for me as an outsider who has never thought about this gene or this phenotype before, any background you could give would be helpful.

I kept looking for evidence that the protein is not expressed in the KO mice. Then I realized that the use of HA tag in the in vitro experiments and the use of a riboprobe in the mouse sections must be because there simply exists no good antibody for this protein. If so, this fact might be stated up front to avoid the reader wondering. A quick google search indicates some vendors claim to sell GPR151 antibodies; if the authors already tested these and found them not to work, that info might be useful to some readers.

I found myself wondering how the originally reported association could so spectacularly fail to replicate. Two possibilities came to mind. One, the UKBB association was right at the edge of significance (4.9e-8 in Emdin 2018, 5.7e-9 in Akbari 2021) — was it just a false positive? Two, is there any chance that there is some real effect on BMI but only through an interaction with age or some other variable? How does the median age in this cohort compare to that in UKBB? Might the KO mice have ended up with lower BMI at some age greater than 17 weeks? I agree with the authors that the data already presented here are pretty much sufficient to kill any interest in this gene as drug target, but some discussion/caveats on this topic might be nice to have.

Eric Vallabh Minikel

November 8, 2021

Reviewer #2: In this manuscript the authors were investigating whether they could replicate previous associations between rare putative loss of function (pLOF) variants in GPR151 and obesity/BMI, and type 2 diabetes risk. To do this they analysed sequence data from the Pakistan Genome Resource PGR, including 30,833 individuals with whole-exome data and 9,292 with whole genome sequence data. The authors identified three pLOF, the previously described Arg95Ter, and two additional variants Tyr99Ter and Phe175LeufsTer7, they tested the effect of these variants individually and in a gene burden test for association with BMI and a number of additional related traits (5 additional in total). The authors also investigated in vitro the effect of these variants on expression and created a knockout mouse which they studied under chow and high fat diet conditions. Overall the authors conclude from their analyses that there is no compelling evidence that targeting GPR151 with antagonists would be an effective approach for obesity treatment.

Overall the manuscript is clear, concise and makes an interesting point. However there are some areas of concern:

1. Abstract – The authors state “Moreover, loss of GPR151 confers a nominally significant increase in risk of T2D (odds ratio = 1.2, p value = 0.03). Relative to wild-type mice, Gpr151-/- animals exhibit no difference in body weight on normal chow, and higher body weight on a high-fat diet, consistent with the findings in humans.” Firstly, the authors state the association with increased risk of T2D is nominally significant but this does not take into account the number of tests done (BMI, T2D, cholesterol, triglycerides, waist-hip ratio, MI) so I think this result as stated risks over-interpreting the data (additional comments on T2D and other analysis in other points below). Secondly, the authors state that knockout mice have higher body weight on a high-fat diet, and that this is consistent with human findings. But this is confusing as the initial human data suggested that lof variants in GPR151 lowered BMI in humans and the authors here with their own data show that they do not see any evidence for differences in BMI in humans with lof, so unclear what is meant by being consistent with findings in humans.

2. Background/ context of the findings - Some of the introduction is missing out key recent papers such as Sobreira et al., 2021 relating to the effects at the FTO locus in nearby loci. Also the statement regarding melanocortin 4 receptor agonists is somewhat misleading. Though many of these compounds do have undesirable cardiovascular effects this is not true of all, and so the phrase should be re-stated. Setmelanotide seems to be well tolerated with minimal side effects, this is published in e.g. Clement et al 2018; Haw et al., 2020 and indeed the review cited by the authors Yeo et al 2021 shows this clearly in table 3. I think it is still valid to state additional body weight reduction drugs are needed but important to ensure the statements made are correct.

3. An important point which the authors do not make is that even in the original publications describing an association between pLOF variant Arg95Ter, the effect on BMI was very modest (−0.36 kg/m2). So arguably, the expectation would be that this effect would be too modest to have meaningful clinical impact, although I acknowledge the previous papers were mostly focused on additive effects. However, there are 20 homozygous carriers for this variant in biobank (and only one is the authors data) and the authors could easily investigate this and combine it in meta-analysis with their own data, which I do think is warranted to gain clarity as to the effect of this variant on BMI.

4. Critically, I think there are two questions that the authors need to consider separately:

a. Is there evidence for replication/lack of replication of an association between the previous pLOF variant (Arg95Ter) at this locus and reduced BMI? I note that here the authors show that their results are still consistent (CIs overlap) with a protective effect of this variant on BMI (their numbers here are smaller than previously published so lack power). I would suggest a meta-analysis of this variant across all available cohorts with this data is warranted as the data shown here do not provide evidence “against” this association. The authors might consider including data from non overlapping previous datasets and if possible the Genes and Health initiative as another effort enriched for autozygous individuals.

b. Is there evidence that complete loss of function at this locus will have a clinically meaningful effect on BMI in humans?

The two questions are not exactly identical because although the authors do not have evidence for a clinically meaningful reduction in BMI in their population, their data do not refute an association between the Arg95Ter and reduced BMI. Indeed their CIs overlap previous effect estimates with larger sample sizes. So I think a meta-analysis across all datasets with this variant is warranted to try and establish whether the original association stands, or not.

Regarding the second point one might argue even if the effects replicate the effect on BMI overall is modest. Again I think given the available data in UK biobank including additional homozygous pLOF carriers this point would be best addressed by meta-analysing the results across all possible datasets the authors can access, they clearly have access to UK biobank so this should be straightforward. I’d suggest if possible including data from Genes and Health would also be interesting and add value. Importantly if the desire is to include only data from null alleles it would be critical to ensure incomplete loss of function variants are not included in the burden test.

5. Looking at the data in Table 1, the second termination variant also has an effect size point estimate that is consistent with lower BMI for homozygous carriers. So I think there is a real question whether the frameshift variant which occurs much later in the protein is fundamentally different. Indeed the author’s data show that this variant is expressed although at low levels. The authors conclude this variant is loss of function because of its lower levels of expression but two bands are clearly seen so the variant is expressed, and there is no in vitro functional data to support the statement that this is a complete loss of function variant. Given this, it would be good to see the gene burden test results removing this variant from the burden test.

6. The association with T2D is nominal only and not adjusted for the different tests and phenotypes looked at, so I think interpretation needs to take this into account. Specifically, and if the authors remove the frameshift variant it looks like their CIs overlap the previous estimates for a protective effect? Instead of a straight power calculation for what the authors are powered to detect, I would prefer to see a power calculation for what effect sizes the authors are powered to rule out? Again I think combining this new data with previously published data in meta-analysis would increase power and provide more clarity as to what the data are showing.

7. The mouse data suggest a possible sexual dimorphism in the phenotype of the knockout mice, have the authors analysed the human data stratified by sex? I think despite smaller numbers and loss of power it would be interesting to check whether there is any evidence from human data for different variant effects between the sexes.

8. The manuscript is missing a discussion on how the authors interpret their results in light of previous association results with reduced BMI and obesity at this gene in much larger sample sizes (including comparable numbers for some homozygous individuals and variants)? Specifically, in three different cohorts a burden of pLOF and missense predicted deleterious variants associated with reduced BMI. How do the authors interpret their data in light of previous findings? I think a meta-analysis across available datasets may help provide further clarity here.

9. Data availability: I could not find a specific data availability statement in the manuscript aside from a “no-some restrictions will apply” in the box at the front. Please clarify exactly what data will be available and how, for example is the mouse line available from somewhere, will the summary statistics for all GPR151 variants and associated phenotypes analysed in the manuscript be available somewhere? Although the authors mention all associated data is within the manuscript this is not really the case as full genotype counts for ref/ref ref/alt and alt/alt and corresponding phenotypes are not given for every phenotype tested. If some data have restricted access please explain what data cannot be made available and why.

Minor issues:

1. In all the tables, for clarity it would be helpful to see the N total in cases / controls or in the entire test data not just hets and hom carriers, or better still the number of each genotype class in cases and controls separately.

2. There is no author summary provided.

3. Figure 2 - I suggest modifying Figure 2 title to better represent entire multipanel figure or pulling out the weight curves into separate figure. Would encourage authors to change the colour from male and female mice away from stereotypes of blue and pink.

4. Figure 2D the riboprobe for GPR151 intestine data in wild-type is not particularly obvious and its detection seems a little subjective.

5. Methods: “We obtained a list of high-quality protein coding transcripts with annotated start and stop codons.” Please clarify where this was obtained from or how exactly is a high-quality protein coding transcript defined?

6. Methods: Case classification, T2D cases “1) Documented history of diabetes” , please specify what documented history of diabetes means? Also, how was type 1 diabetes, excluded?

7. Methods: adjusted for top 5 principle components, why 5?

8. Unclear whether genomes and exomes were treated the same or whether there was any adjustment for batch effect in the analysis? It looks like the data were analysed separately and then meta-analysed but it would be good to make this a bit clearer.

9. Methods : somewhat unclear the authors state that “At six-weeks of age, Gpr151-/- mice and wild type littermates were individually housed for body weight and food consumption measurements and provided either a standard chow diet (Purina Picolab 5053) or high fat diet in which 60% kcal is derived from fat (Research Diets D12492i) with ad libitum access to water.” But in the following sentence it is stated that female mice were group housed, so were they group housed from the outset and then from six-weeks the male Kos and wild-type only were individually housed?

Reviewer #3: The present article by Gurtan et al. used the Pakistan Genome Resource, which includes high rate of human homozygous loss-of-function (KO) due to high consanguinity, to validate GPR151 as a potential drug target. Despite accurate statistical power, the authors did not find any significant association between three loss-of-function GPR151 variants and BMI. The authors also investigated mouse models deleted for Gpr151 and found that these mice had no difference in body weight in normal chow.

This article is well written and the reviewer believes that these negative results are of great interest for the community (it is crucial to publish negative results).

The reviewer has the following comments:

- If the reviewer is right, the authors actually combined carriers of heterozygous and homozygous LOF GPR151 variants for their burden analysis, while the abstract and introduction mainly tackled “human homozygous loss-of-function”. It would be important to also analyze carriers of homozygous LOF GPR151 variants only (and remove the carriers of heterozygous variants). Furthermore, to enhance the statistical power of this analysis, the authors could combine the carriers of the three LOF variants (at homozygous state).

- The mean depth of coverage of GPR151 in the Pakistan Genome Resource should be provided, as well as the genotyping success rate for the three variants.

- The authors should also analyse the effect of LOF variants (in homozygous carriers) on obesity risk (obese participants versus normal weight participants)

**********

Have all data underlying the figures and results presented in the manuscript been provided?

Large-scale datasets should be made available via a public repository as described in the PLOS Genetics data availability policy, and numerical data that underlies graphs or summary statistics should be provided in spreadsheet form as supporting information.

Reviewer #1: No: There is currently no "data availability" statement in the manuscript

Reviewer #2: Yes

Reviewer #3: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Eric Vallabh Minikel

Reviewer #2: No

Reviewer #3: Yes: Amelie Bonnefond

29 Jan 2022

Submitted filename: Gurtan et al, response to reviewers.pdf

Click here for additional data file.

13 Feb 2022

Dear Dr Gurtan,

We are pleased to inform you that your manuscript entitled "Analyzing human knockouts to validate GPR151 as a therapeutic target for reduction of body mass index" has been editorially accepted for publication in PLOS Genetics. Congratulations!

Before your submission can be formally accepted and sent to production you will need to complete our formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Please note: the accept date on your published article will reflect the date of this provisional acceptance, but your manuscript will not be scheduled for publication until the required changes have been made.

Once your paper is formally accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you’ve already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosgenetics@plos.org.

In the meantime, please log into Editorial Manager at https://www.editorialmanager.com/pgenetics/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production and billing process. Note that PLOS requires an ORCID iD for all corresponding authors. Therefore, please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field.  This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager.

If you have a press-related query, or would like to know about making your underlying data available (as you will be aware, this is required for publication), please see the end of this email. If your institution or institutions have a press office, please notify them about your upcoming article at this point, to enable them to help maximise its impact. Inform journal staff as soon as possible if you are preparing a press release for your article and need a publication date.

Thank you again for supporting open-access publishing; we are looking forward to publishing your work in PLOS Genetics!

Yours sincerely,

Giles S. H. Yeo

Guest Editor

PLOS Genetics

Gregory Barsh

Editor-in-Chief

PLOS Genetics

www.plosgenetics.org

Twitter: @PLOSGenetics

----------------------------------------------------

Comments from the reviewers (if applicable):

The reviewers are all now enthusiastic about accepting this manuscript for publication, and I agree. Please just note the remaining minor comments from reviewer #2.

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #1: My concerns have been adequately addressed.

Reviewer #2: The authors have addressed my concerns and I think the manuscript is considerably improved. However there are a few minor issues that still merit stating clearly for accuracy.

1. The authors have predicted a full loss of function for Phe175LfsTer7 based on computational prediction but the experimental in vitro evidence does not provide proof that this variant is a complete loss of function. Since no functional readouts aside from protein expression were done, I think the wording should be clear on this point. For clarity the authors should amend the sentence “The severity and diminished expression of this truncation indicate that this variant is a loss-of-function.” To say “suggest” instead of “indicate”. I agree that given the nature of the predicted truncation this is very likely to be loss of function but in theory it could be dominant negative so I think “suggest” would be more appropriate wording.

2. A minor point is that a diabetes age of diagnosis after age 22 is not a guarantee that this does include a small proportion of patients with type 1 diabetes. See for example, doi: 10.1016/S2213-8587(17)30362-5.

Reviewer #3: The authors accurately answered to my comments

**********

Have all data underlying the figures and results presented in the manuscript been provided?

Large-scale datasets should be made available via a public repository as described in the PLOS Genetics data availability policy, and numerical data that underlies graphs or summary statistics should be provided in spreadsheet form as supporting information.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: None

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Eric Vallabh Minikel

Reviewer #2: No

Reviewer #3: Yes: Amélie Bonnefond

----------------------------------------------------

Data Deposition

If you have submitted a Research Article or Front Matter that has associated data that are not suitable for deposition in a subject-specific public repository (such as GenBank or ArrayExpress), one way to make that data available is to deposit it in the Dryad Digital Repository. As you may recall, we ask all authors to agree to make data available; this is one way to achieve that. A full list of recommended repositories can be found on our website.

The following link will take you to the Dryad record for your article, so you won't have to re‐enter its bibliographic information, and can upload your files directly:

http://datadryad.org/submit?journalID=pgenetics&manu=PGENETICS-D-21-01413R1

More information about depositing data in Dryad is available at http://www.datadryad.org/depositing. If you experience any difficulties in submitting your data, please contact help@datadryad.org for support.

Additionally, please be aware that our data availability policy requires that all numerical data underlying display items are included with the submission, and you will need to provide this before we can formally accept your manuscript, if not already present.

----------------------------------------------------

Press Queries

If you or your institution will be preparing press materials for this manuscript, or if you need to know your paper's publication date for media purposes, please inform the journal staff as soon as possible so that your submission can be scheduled accordingly. Your manuscript will remain under a strict press embargo until the publication date and time. This means an early version of your manuscript will not be published ahead of your final version. PLOS Genetics may also choose to issue a press release for your article. If there's anything the journal should know or you'd like more information, please get in touch via plosgenetics@plos.org.

31 Mar 2022

PGENETICS-D-21-01413R1

Analyzing human knockouts to validate GPR151 as a therapeutic target for reduction of body mass index

Dear Dr Gurtan,

We are pleased to inform you that your manuscript entitled "Analyzing human knockouts to validate GPR151 as a therapeutic target for reduction of body mass index" has been formally accepted for publication in PLOS Genetics! Your manuscript is now with our production department and you will be notified of the publication date in due course.

The corresponding author will soon be receiving a typeset proof for review, to ensure errors have not been introduced during production. Please review the PDF proof of your manuscript carefully, as this is the last chance to correct any errors. Please note that major changes, or those which affect the scientific understanding of the work, will likely cause delays to the publication date of your manuscript.

Soon after your final files are uploaded, unless you have opted out or your manuscript is a front-matter piece, the early version of your manuscript will be published online. The date of the early version will be your article's publication date. The final article will be published to the same URL, and all versions of the paper will be accessible to readers.

Thank you again for supporting PLOS Genetics and open-access publishing. We are looking forward to publishing your work!

With kind regards,

Katalin Szabo

PLOS Genetics

On behalf of:

The PLOS Genetics Team

Carlyle House, Carlyle Road, Cambridge CB4 3DN | United Kingdom

plosgenetics@plos.org | +44 (0) 1223-442823

plosgenetics.org | Twitter: @PLOSGenetics