Article
A multi-cohort genome-wide association study in
African ancestry individuals reveals risk loci for
primary open-angle glaucoma
Graphical abstract
Highlights
d A comprehensive GWAS on African ancestry individuals with

glaucoma was conducted

d 46 risk loci significantly associated with glaucoma were

detected

d Variants in ROCK1P1, ARHGEF12, and DBF4P2

demonstrated likely causal pathophysiology

d Polygenic risk scores derived from African ancestry

individuals show enhanced strength
Verma et al., 2024, Cell 187, 464–480
January 18, 2024 ª 2023 Elsevier Inc.
https://doi.org/10.1016/j.cell.2023.12.006
Authors

Shefali S. Verma, Harini V. Gudiseva,

Venkata R.M. Chavali, ...,

Michael A. Hauser, Marylyn D. Ritchie,

Joan M. O’Brien

Correspondence
joan.o’brien@pennmedicine.upenn.edu

In brief

Glaucoma represents a pressing public

health need among African ancestry

individuals. This study provides novel

insight into the genetic architecture of

glaucoma in this population by identifying

gene variants with pathophysiological

significance.
ll

mailto:joan.o'brien@pennmedicine.upenn.edu
https://doi.org/10.1016/j.cell.2023.12.006
http://crossmark.crossref.org/dialog/?doi=10.1016/j.cell.2023.12.006&domain=pdf


ll
Article

Amulti-cohort genome-wide association study
in African ancestry individuals reveals risk
loci for primary open-angle glaucoma
Shefali S. Verma,1,29 Harini V. Gudiseva,2,29 Venkata R.M. Chavali,2 Rebecca J. Salowe,2 Yuki Bradford,3 Lindsay Guare,3

Anastasia Lucas,3 David W. Collins,2 Vrathasha Vrathasha,2 Rohini M. Nair,2 Sonika Rathi,2 Bingxin Zhao,4 Jie He,2

Roy Lee,2 Selam Zenebe-Gete,2 Anita S. Bowman,2 Caitlin P. McHugh,5 Michael C. Zody,5 Maxwell Pistilli,2

Naira Khachatryan,2 Ebenezer Daniel,2 Windell Murphy,6 Jeffrey Henderer,7 Regeneron Genetics Center,8

Tyler G. Kinzy,9,10 Sudha K. Iyengar,9,10 Neal S. Peachey,10,11 VA Million Veteran Program, Kent D. Taylor,12

(Author list continued on next page)

1Department of Pathology and Laboratory Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA

2Scheie Eye Institute, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
3Department of Genetics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
4Department of Statistics and Data Science, The Wharton School, University of Pennsylvania, Philadelphia, PA, USA
5New York Genome Center, New York, NY, USA
6Windell Murphy, MD, Philadelphia, PA, USA
7Department of Ophthalmology, Lewis Katz School of Medicine, Temple University, Philadelphia, PA, USA
8Tarrytown, NY, USA
9Department of Population and Quantitative Health Sciences, Cleveland Institute for Computational Biology, Case Western Reserve

University, Cleveland, OH, USA
10Louis Stokes Cleveland VA Medical Center, Cleveland, OH, USA
11Cole Eye Institute, Cleveland Clinic, Cleveland, OH, USA
12Department of Pediatrics, The Institute for Translational Genomics and Population Sciences, The Lundquist Institute for Biomedical
Innovation at Harbor-UCLA Medical Center, Torrance, CA, USA
13Viterbi Family Department of Ophthalmology, Shiley Eye Institute, University of California, San Diego, La Jolla, CA, USA
14Department of Ophthalmology and Visual Sciences, Heersink School of Medicine, University of Alabama at Birmingham, Birmingham, AL,

USA

(Affiliations continued on next page)
SUMMARY
Primary open-angle glaucoma (POAG), the leading cause of irreversible blindness worldwide, disproportion-
ately affects individuals of African ancestry. We conducted a genome-wide association study (GWAS) for
POAG in 11,275 individuals of African ancestry (6,003 cases; 5,272 controls). We detected 46 risk loci asso-
ciated with POAG at genome-wide significance. Replication and post-GWAS analyses, including functionally
informed fine-mapping, multiple trait co-localization, and in silico validation, implicated two previously unde-
scribed variants (rs1666698 mapping to DBF4P2; rs34957764 mapping to ROCK1P1) and one previously
associated variant (rs11824032 mapping to ARHGEF12) as likely causal. For individuals of African ancestry,
a polygenic risk score (PRS) for POAG from our mega-analysis (African ancestry individuals) outperformed a
PRS from summary statistics of a much larger GWAS derived from European ancestry individuals. This study
quantifies the genetic architecture similarities and differences between African and non-African ancestry
populations for this blinding disease.
INTRODUCTION

Primary open-angle glaucoma (POAG) is an insidious neurode-

generative disease of the optic nerve (ON) that causes progres-

sive loss of peripheral vision.1 This disease affects 44 million in-

dividuals worldwide, with a projected prevalence of 80 million by
464 Cell 187, 464–480, January 18, 2024 ª 2023 Elsevier Inc.
2040.2 It is estimated that 6 million individuals were bilaterally

blinded by POAG in 2020.3 African ancestry populations world-

wide are disproportionately affected by this disease.3 Individuals

of African ancestry are four to five timesmore likely to be affected

by POAG than individuals of European ancestry4 and up to 15

times more likely to experience vision loss from the disease.5

http://crossmark.crossref.org/dialog/?doi=10.1016/j.cell.2023.12.006&domain=pdf


Xiuqing Guo,12 Yii-Der Ida Chen,12 Linda Zangwill,13 Christopher Girkin,14 Radha Ayyagari,13 Jeffrey Liebmann,15

Chimd M. Chuka-Okosa,16 Susan E. Williams,17 Stephen Akafo,18 Donald L. Budenz,19 Olusola O. Olawoye,20

Michele Ramsay,21 Adeyinka Ashaye,20 Onoja M. Akpa,22 Tin Aung,23 Janey L. Wiggs,24 Ahmara G. Ross,2 Qi N. Cui,2

Victoria Addis,2 Amanda Lehman,2 Eydie Miller-Ellis,2 Prithvi S. Sankar,2 Scott M. Williams,25 Gui-shuang Ying,2

Jessica Cooke Bailey,9,10,28 Jerome I. Rotter,12 Robert Weinreb,13 Chiea Chuen Khor,26 Michael A. Hauser,27

Marylyn D. Ritchie,3,30 and Joan M. O’Brien2,30,31,*
15Department of Ophthalmology, Columbia University Medical Center, Columbia University, New York, NY, USA
16Department of Ophthalmology, University of Nigeria, Ituku, Nigeria
17Division of Ophthalmology, Department of Neurosciences, University of the Witwatersrand, Johannesburg, South Africa
18Unit of Ophthalmology, Department of Surgery, University of Ghana Medical School, Accra, Ghana
19Department of Ophthalmology, University of North Carolina, Chapel Hill, NC, USA
20Department of Ophthalmology, University of Ibadan, Ibadan, Nigeria
21SydneyBrenner Institute forMolecular Bioscience, Faculty of Health Sciences, University of theWitwatersrand, Johannesburg, South Africa
22Department of Epidemiology and Medical Statistics, College of Medicine, University of Ibadan, Ibadan, Nigeria
23Singapore Eye Research Institute, Singapore, Singapore
24Department of Ophthalmology, Massachusetts Eye and Ear, Harvard Medical School, Boston, MA, USA
25Department of Population and Quantitative Health Sciences, Case Western Reserve University, Cleveland, OH, USA
26Genome Institute of Singapore, Singapore, Singapore
27Department of Medicine, Duke University, Durham, NC, USA
28Department of Pharmacology and Toxicology, Center for Health Disparities, Brody School of Medicine. East Carolina University, Greenville,
NC, 27834, USA
29These authors contributed equally
30These authors contributed equally
31Lead contact
*Correspondence: joan.o’brien@pennmedicine.upenn.edu

https://doi.org/10.1016/j.cell.2023.12.006

ll
Article
There are several major risk factors for POAG in all popula-

tions, including advanced age, elevated intraocular pressure

(IOP), and family history of glaucoma, with heritability estimates

ranging from 0.17 to 0.81.6 Elevated IOP is currently the only

targetable component of the disease, and treatments are often

ineffective in halting disease progression.7,8 Additionally, some

patients maintain normal IOP levels (normal-tension glaucoma)

but can still experience disease progression and vision loss.9

This indicates that POAG, a frequently inherited disease, has

additional underlying mechanisms that could be elucidated by

genetic studies.10,11

Despite POAG having a strong inherited component,12–15 un-

derstanding of the disease’s genetics remains incomplete.16 A

total of 174 unique risk loci have been associated with POAG

and related traits through a large-scale meta-analysis17 and

genome-wide association studies (GWASs) in European,18–22

Japanese,23–27 or multi-ancestry populations,22,28,29 with sam-

ple sizes ranging from 387 to 383,500. However, many of these

loci have a reduced or unknown role in individuals of African

ancestry, indicating that global ancestry groups may have differ-

ences in genetic risk and highlighting the need to compare

diverse genetic architectures.29–32

Several studies have investigated glaucoma genetics in multi-

ancestry or African ancestry populations but were often limited

by a small sample size.33,34 One GWAS of continental African

and African American populations identified a previously unde-

scribed candidate locus on the EXOC4 gene, but it did not asso-

ciate in theWest African replication group.31 The African Descent

and Glaucoma Evaluation Study (ADAGES)35 identified an asso-

ciation with advanced POAG and the ENO4 locus.36 Most

recently, a GWAS in the Genetics of Glaucoma in People of Afri-

can Descent (GGLAD) consortium identified a previously unde-
scribed locus in APBB2 as significantly associated with POAG

in individuals of African ancestry, but the index variant

(rs59892895) was not polymorphic in participants of European

or Asian ancestry.37

In this study, we conducted a large GWAS for POAG in African

ancestry individuals from three African population datasets.

We further investigated significant findings through replication

studies in four independent datasets, functional validation

studies, and in silico analysis. We also compared the genetic

architecture of POAG in African and non-African ancestry popu-

lations. Our objective was to identify variants of pathophysiolog-

ical importance to POAG in African ancestry individuals, gaining

insight into the genetics of this blinding familial disease in the

most affected population.

RESULTS

Study datasets
We performed a mega-analysis in a discovery cohort that com-

bined three African ancestry datasets: the ADAGES study (n =

1,999),36 the GGLAD consortium (n = 2,952),37 and the Primary

Open-Angle African American Glaucoma Genetics (POAAGG)

study (n = 6,324)38 (Figure 1). Genotype data were imputed for

each dataset individually using the Trans-Omics for Precision

Medicine (TOPMED) reference panel.39

The results from the discovery analysis were replicated in four

independent datasets of African ancestry individuals: a second

independent dataset from GGLAD (referred to as GGLAD-2),37

All of Us (AOU),40 Penn Medicine BioBank (PMBB),41 and Million

Veteran Program (MVP).42 Additionally, we compared the results

from the discovery mega-analysis with multi-ancestry GWAS re-

sults from the Global Biobank Meta-Analysis Initiative (GBMI)17
Cell 187, 464–480, January 18, 2024 465

mailto:joan.o'brien@pennmedicine.upenn.edu
https://doi.org/10.1016/j.cell.2023.12.006


Figure 1. Study designs and subject characteristics

POAAGG, Primary Open-Angle African American Glaucoma Genetics; GGLAD, Genetics of Glaucoma in People of African Descent; ADAGES, African Descent

andGlaucoma Evaluation Study; AOU, All of Us; GBMI, Global BioBankMeta-Analysis Initiative; EAS, East Asian Ancestry; AFR, African Ancestry; AMR, Admixed

American Ancestry; EUR, European Ancestry; SAS, South Asian Ancestry.

ll
Article
and the National Human Genome Reserach Institute-European

Bioinformatics Institute (NHGRI-EBI) GWAS Catalog.43 Finally,

we created a polygenic risk score (PRS), which was applied to

three independent datasets of African ancestry individuals (Elec-

tronic Medical Records and Genomics [eMERGE], n = 13,493;

GGLAD-2, n = 1,490; PMBB, n = 11,117). The demographics

and designs of the studies used in these analyses are detailed

in Figure 1.

Discovery of known and previously undescribed
POAG loci
In the discoverymega-analysis, all subjects had self-reported Af-

rican ancestry or were recruited from continental Africa, and

51.5% were female (Figure 1). In total, 12,944,299 variants

were included in the discovery analyses (STARMethods). Corre-

lation analyses revealed a generally high positive correlation

(Figure S1A) between mega-analysis and meta-analysis results,

indicating a strong agreement between the two approaches

(Pearson correlation coefficient = 0.86). This finding suggests

that bothmethods yield consistent findings overall. Furthermore,

to investigate the power of the mega-analysis approach, we

focused on variants filtered at a p value threshold of 1 3 10�04

(Figure S1B). Notably, we observed that smaller effect sizes in

our mega-analysis exhibited higher p values, indicating higher

power to detect these variants. These results highlight the

robustness and statistical power of the mega-analysis approach

employed in this study. Moreover, the observation that smaller
466 Cell 187, 464–480, January 18, 2024
effect sizes had higher power in the mega-analysis emphasizes

the advantages of pooling individual-level data when possible

to increase the precision and sensitivity of our analysis.

We identified a total of 1,110 suggestive loci for POAG (p %

2.8 3 10�4) from the mega-analysis (Table S1), of which 46

reached genome-wide significance as shown in Table 1. Only

2 of the 46 significant loci were previously reported in other glau-

coma studies, as shown in the overlap of the Venn diagram in

Figure S2A. Conditional analyses using Genome-wide Complex

Trait Analysis-Conditional and Joint Association Analysis

(GCTA-COJO)44,45 identified 47 conditionally independent asso-

ciations exceeding genome-wide significance (p % 5 3 10�8)

(Table S2). All known and previously undescribed genes for

POAG risk are annotated in Figure 2, with a full breakdown given

in the Figure360 video. The quantile-quantile (QQ)-plot for this

analysis is shown in Figure S2C.

We also replicated 37 (21%) of 174 previously reported POAG

loci (listed in Table S3) at a Bonferroni-corrected significance for

174 loci (0.05/174; p % 2.8 3 10�4). Among these replicated as-

sociations are variants mapping to genes such as ARHGEF12,

CDKN2B-AS1, TMCO1, AFAP1, LMX1B, and many more. Of

note, rs9992186 (chr4: 46875929), mapping nearest to the

APBB2 gene, is in high linkage disequilibrium with rs59892895

(R2 = 0.81), which has previously been associated with POAG

risk in African ancestry individuals.37 In our study, this association

was also identified at suggestive GWAS significance (p value =

2.15 3 10�6, odds ratio [OR] = 1.03), as reported in Table S1.



Table 1. Genome-wide significant loci from the mega-analysis in discovery cohorts

Chr:Position

Lead variant

rsID Nearest gene

Effect

allele

Non-

effect

allele

Allele

frequency OR

Lower limit

95% CI

Upper limit

95% CI p value

1:146895002 rs36172683 LOC124900456 T G 0.79 1.67 1.51 1.86 5.2E�23

1:1557146 rs4320726 SSU72 C T 0.08 1.42 1.25 1.61 3.5E�08

1:192888862 rs1231753 RN7SKP126 G A 0.94 0.61 0.53 0.72 4.4E�10

1:197079478 rs35655718 ASPM T C 0.49 0.81 0.75 0.86 7.9E�10

1:53756898 rs1183394 LOC124904179 A G 0.93 1.63 1.39 1.91 6.2E�10

10:28988113 rs17755250 LOC124403927 A C 0.18 1.33 1.22 1.44 3.6E�12

10:39164667 rs879082727 CHEK2P5 A G 0.79 1.38 1.23 1.56 3.7E�08

10:47323706 rs4922508 GDF2 A G 0.86 0.56 0.49 0.64 8.5E�18

10:48058038 rs1308452687 RNA5SP315 T A 0.83 0.70 0.63 0.78 9.9E�11

11:120354080 rs11824032 ARHGEF12 G A 0.25 1.24 1.16 1.33 6.6E�10

11:121138670 rs34002948 TECTA G A 0.66 1.32 1.23 1.42 2.1E�14

11:50385050 rs117353933 LINC02750 G A 0.05 0.47 0.39 0.58 3.3E�13

11:54708680 rs4611187 OR4A5 T C 0.06 0.60 0.51 0.70 9.6E�11

12:127391021 rs57347531 LINC02375 G T 0.06 0.68 0.60 0.79 3.7E�08

12:128365235 rs74537852 TMEM132C C T 0.08 0.67 0.59 0.75 2.9E�11

12:70810380 rs1040027 PTPRR C T 0.56 1.26 1.18 1.34 2.7E�13

12:7783972 rs6488629 LOC108942766 T C 0.92 1.67 1.43 1.95 4.8E�11

13:23420079 rs4770444 SACS G T 0.92 0.72 0.64 0.81 4.9E�08

14:105708459 rs144145908 IGH C T 0.09 1.50 1.33 1.68 3.7E�12

14:31624055 rs28580449 NUBPL C T 0.38 1.23 1.15 1.31 1.6E�09

14:56502862 rs4584747 LOC101927690 A C 0.53 1.25 1.17 1.34 1.5E�11

15:100159652 rs145914721 ADAMTS17 T C 0.06 0.62 0.54 0.71 4.2E�12

15:25437859 rs77631699 UBE3A A G 0.02 2.67 2.04 3.49 4.2E�13

16:33823269 rs720126 IGHV3OR16-13 C T 0.34 1.34 1.21 1.47 2.4E�09

16:5113716 rs7189371 ENPP7P14 G C 0.88 0.70 0.62 0.78 4.9E�10

17:83183702 rs71193787 LOC101929650 T A 0.30 1.40 1.29 1.51 4.6E�17

18:15232752 rs9963245 BNIP3P3 A G 0.45 1.45 1.33 1.58 1.5E�16

18:93327 rs34957764 ROCK1P1 C T 0.10 1.75 1.97 1.56 4.9E�24

19:6832739 rs73000309 VAV1 T C 0.05 1.63 1.39 1.90 3.6E�10

2:111570643 rs1666698 DBF4P2 T C 0.63 0.80 085 0.74 3.6E�10

2:88849801 rs141848863 LOC108348023 C T 0.06 0.61 0.53 0.71 3.9E�11

20:62675278 rs4809477 SLCO4A1 T C 0.75 0.80 0.75 0.86 1.7E�09

22:19485902 rs5746749 CDC45 A C 0.94 0.66 0.57 0.75 1.2E�09

3:125210558 rs671064 SLC12A8 G A 0.89 1.36 1.22 1.52 2.2E�08

4:189790415 rs2739537 FRG1-DT T C 0.87 1.50 1.32 1.70 4.2E�10

4:42192281 rs139976752 BEND4 C A 0.09 1.40 1.25 1.57 6.0E�09

4:51830626 rs9998449 DCUN1D4 A G 0.64 1.33 1.22 1.46 6.9E�11

6:191736 rs9502903 LOC285766 A G 0.69 1.41 1.30 1.53 2.2E�16

6:3176138 rs369229648 TUBB2BP1 C T 0.03 0.53 0.44 0.65 7.0E�10

6:88968505 rs6939237 RNGTT A G 0.96 0.60 0.50 0.72 8.3E�09

7:4048923 rs6462562 SDK1 A G 0.46 1.22 1.14 1.30 1.8E�09

7:73913547 rs55701129 LOC121740686 T C 0.05 1.52 1.32 1.76 3.8E�09

8:101852942 rs534512 NCALD A G 0.64 1.23 1.14 1.32 1.0E�08

8:12323125 rs34536966 DEFB130A G A 0.73 0.75 0.67 0.83 1.1E�08

8:68235044 rs11556027 PREX2 C T 0.87 1.47 1.32 1.63 3.8E�13

9:131033574 rs3780279 LAMC3 A C 0.93 0.71 0.63 0.81 4.2E�08

ll

Cell 187, 464–480, January 18, 2024 467

Article



Figure 2. Discovery of known and previously undescribed loci from the discovery mega-analysis of African ancestry individuals

For a Figure360 author presentation of this figure, see https://doi.org/10.1016/j.cell.2023.12.006.

The circular plot depicts genes previously associated with POAG (blue) and previously undescribed genes revealed by the discovery mega-analysis (orange). The

inner section (black) shows the mega-analysis Manhattan plot in a circular fashion. The green section demonstrates heterogeneity I2 values from sex-stratified

meta-analyses. The inner scatter plots display the relationship between the absolute beta and minor allele frequency (MAF) among the mega-analysis data,

including African ancestry individuals only and GBMI results for European individuals, East Asian individuals, and cross-ancestry meta-analyses.

See also Tables S1–S4 and S7 and Figures S1, S2, and S4.

ll
Article
One of the previously undiscovered associations identified in

the mega-analysis was the variant rs34957764, which maps to

theRho-associated coiled-coil-containing protein kinase 1 pseu-

dogene1 (ROCK1P1) gene region. Rho-associated coiled-coil ki-
468 Cell 187, 464–480, January 18, 2024
nase, a pseudogene resulting from partial duplication of the

ROCK1 gene, regulates cellular responses such as cell growth,

proliferation, and apoptosis. We also identified a previously un-

described variant (rs4938802) and known variant (rs11824032)

https://doi.org/10.1016/j.cell.2023.12.006


Figure 3. Trait co-localization results

(A) Mega-analysis of discovery cohort, colored by trait GWAS data for baseline CDR, with separation of congruent SNPs (increased expression associated with

increase in the trait) versus incongruent SNPs (increase in expression associated with a decrease in the trait).

(B) Enrichment of SNPs associated with baseline CDR among significant SNPs from the mega-analysis.

(C) P-P plot, with congruent SNPs on top and incongruent SNPs at the bottom.

(D) Violin boxplot, with rs11824032 genotypes on the x axis and baseline CDR on the y axis.

(E) Violin boxplot, with rs10892564 genotypes on the x axis and baseline CDR the on the y axis.

See also Table S5.

ll
Article
that map to the ARHGEF12 gene; the rs4938802 variant has

previously been associated with elevated IOP and POAG risk in

European ancestry individuals.46,47 However, the previously

undiscovered locus reported in our mega-analysis is in a

different 250-kb region than the locus reported in the previous

GWAS. Another previously undescribed variant of interest is

rs145914721, mapping near to the ADAMTS17 gene, which en-

codes a member of the ADAMTS protein family. Mutations in

this gene, along with ADAMTS10 and LTBP2, are associated

with Weill-Marchesani-like syndrome, an inherited connective

tissue disease characterized by lens abnormalities and second-

ary glaucoma.48,49 Another previously undescribed variant,

rs6462562, maps to the SDK1 gene, which encodes for an adhe-

sion molecule expressed in the cells that are damaged in POAG,

retinal ganglion cells (RGCs), and in a subset of interneurons. This

variant also promotes synaptic connectivity.50

Differential effects by sex
Weconducted a sex-stratifiedGWAS in the discoverymega-anal-

ysis dataset (Figure S2B) and then meta-analyzed the sex-strati-

fied results to differential effects by sex. We found 37 loci that

reached the genome-wide significance threshold (p % 5.0 3

10�8). (Figure 2 green panel; Table S4.) Filtering for heterogeneity
p value < 0.01 resulted in nine variants corresponding to three loci

that showed evidence of heterogeneity between males and fe-

males. All nine variants in these three loci demonstrated a stronger

effect in females (OR ranging from 1.87 to 1.92) than in males (OR

ranging from 1.16 to 1.46, with lowest heterogeneity p value for

rs1181192 at p = 0.0009). When calculating a Pearson correlation

among effect sizes for all variants at p value < 0.01 inmales and fe-

males, we observed a significant and strong correlation of 0.74

(p < 0.0001). This finding indicates a consistent pattern of genetic

effects across sexes, suggestingminimal influenceof sex-specific

effects. Moreover, the relatively equal sample sizes in males and

females further support the robustnessand reproducibilityof these

associations within our study cohort.

Quantitative trait co-localization analyses
We examined the genetic co-localization51 between SNPs iden-

tified from the discovery mega-analysis and SNPs identified in a

GWAS of POAG endophenotypes from the POAAGG study.38

These endophenotypes included IOP, cup-to-disc ratio (CDR),

mean deviation (MD) from visual fields, and retinal nerve

fiber layer (RNFL) thickness. Lead variant rs11824032 (chr11:

120354080) mapped to the ARHGEF12 gene (PP4 > 0.8). In the

included gene region, there were 2,091 variants considered in
Cell 187, 464–480, January 18, 2024 469



Figure 4. Forest plot showing effect estimates for discovery mega-analyses and replication studies, as well as the pooled effect meta-

analyses for two replicated risk variants

Pooled estimates for odds ratio and 95% confidence intervals of all replication datasets and meta-analyses (discovery + all replication sets) were calculated by

random effect analyses. Results for individual datasets are denoted by rectangles with lines indicating the 95% confidence intervals and black diamonds

indicating the final pooled summary values. All replication datasets include GGLAD-2 + PMBB + AOU + MVP datasets. Meta-analyses p values calculated from

the ForestPM plot package are denoted on the top of each plot.

(A) SNP rs34957764 in the ROCK1P1 gene was tested in all datasets.

(B) SNP rs1666698 in the DBF4P2 gene was tested in all datasets.

GGLAD-2, Genetics of Glaucoma in People of African Descent; PMBB, Penn Medicine BioBank; AOU, All of Us; MVP, Million Veteran Program.

See also Table S6 and Figures S2 and S3.

ll
Article
the co-localization analyses between POAG and baseline CDR

(Table S5). The locus zoom plot for the ARHGEF12 region’s

GWAS results is shown in Figure S2F. Further, we performed lo-

cus quantification to visualize the change in baseline CDR mea-

surements with respect to the genotypes (Figure 3). Lastly, we

conducted a follow-up two sample Mendelian randomization

(MR) analysis at rs11824032 with baseline CDR as the exposure

and POAG as the outcome, which resulted in a significant p

value = 1.33 3 10�9 (beta = 12.02, SE = 1.98).

Replication of POAG associations in African ancestry
datasets and meta-analysis
We performed replication analyses for the 47 risk loci (353 vari-

ants found in most of the replication datasets) identified in the

conditional analyses in the discovery dataset in African ancestry

individuals extracted from four independent datasets (AOU,

GGLAD-2, PMBB, and MVP). Two variants, mapping to the

DBF4 zinc-finger pseudogene 2 (DBF4P2) and ROCK1P1 genes

and validated in independent datasets, were found to have a sig-

nificant p value below the Bonferroni correction threshold (0.001)

of 0.05/47 (Figure 4; Table S6). Locus zoom plots for these repli-

cating genes are shown in Figures S2D and S2E. Variant

rs34957764 mapping to the ROCK1P1 locus on chromosome

18 reached the Bonferroni significant p value in the AOU dataset

(p value = 0.0002) and showed a strong association with POAG in

a meta-analysis of datasets (discovery p value = 4.93 3 10�24

and pooled meta p value = 1.67 3 10�24). Variant rs1666698

mapping to the DBF4P2 gene was marginally significant in the

MVP dataset (p value = 0.0009, discovery p value = 3.59 3

10�10, and pooled meta p value = 1.123 10�10). Both replicated

SNPs showed consistent effect direction across replication

datasets.
470 Cell 187, 464–480, January 18, 2024
We evaluated theminor allele frequency (MAF) of the two repli-

cating variants (rs1666698 and rs34957764) in our study across

multiple datasets. In addition to the discovery and replication da-

tasets, we assessed the MAF in the 1000 Genomes populations,

providing a comprehensive analysis of global allele frequencies.

Figure S3 presents theMAF patterns of the two SNPs in the 1000

Genomes populations and the specific MAF values observed in

each population within the discovery mega-analysis and replica-

tion datasets. rs1666698 exhibited consistent MAF across Afri-

can populations, with MAF values exceeding 0.4. However, in

European and East Asian populations, theMAFwasmuch lower,

approximately 1%. This stark difference in allele frequencies be-

tween African and non-African populations highlights the popu-

lation-specific nature of this variant. rs34957764 displayed a

distinct MAF pattern, with a prevalence of 20% in African popu-

lations compared with an MAF of 0.12 in other populations. This

marked discrepancy in MAF between African and non-African

populations underscores the significant population stratification

and genetic diversity observed for this variant. This examination

of MAF distribution offers valuable insights into the frequency

and distribution of these variants across diverse populations,

enhancing our understanding of their relevance in a global

context.

Cross-ancestry comparisons
We compared the per-allele effect sizes and MAF between both

previously known and undiscovered loci in our dataset of African

ancestry individuals (Figure S4A). Previously undiscovered vari-

ants identified from themega-analysis have largeeffect sizes inAf-

rican ancestry individuals, whereas previously known variants

from non-African populations (i.e., GBMI GWAS on POAG) show

smaller effect sizes. We also calculated the genetic impact



ll
Article
correlation (rgi) using Popcorn,52 which is the correlation coeffi-

cient of the population-specific allele-variance-normalized SNP

effect sizes (Figure S1C). Genetic correlations between ancestries

ranged from rgi = 0.59 (African with East Asian ancestry) to rgi =

0.73 (African with European ancestry), with the highest correlation

among East Asian and European ancestry individuals (rgi = 0.78).

Since the majority of large GWAS for POAG have been per-

formed in non-African ancestry individuals, we next demon-

strated the added value of including African ancestry individuals

when explaining phenotypic variability for POAG (Figure S4B).

GCTA-restricted maximum likelihood (REML) analyses53

showed that the estimate of variance for POAG was 0.0295

(SE = 0.006) for previously known loci but increased to 0.05

when including both previously known loci and undiscovered

loci from our mega-analysis (Figure S4B). Of note, the previously

undiscovered variants from our mega-analysis explained 0.041

variance, so the highest estimate of variance (after suggestive

significant hits) came from the inclusion of both previously

known and undiscovered loci from our African ancestry cohort.

Prioritizing causal variants, genes, and pathways
Functionally informed fine-mapping

Combining annotations using fine-mapping approaches has been

proven to enhance the ability to localize causal variants. Using in-

tegrated regulatory annotations from epigenome integration

across multiple annotation project (EpIMAP),54 we prioritized pu-

tative causal variants identified through our discovery mega-anal-

ysis.55 We performed fine-mapping using SuSIE56 (STAR

Methods), which uses functional enrichment of regulatory regions

to weight the GWAS summary statistics and compute a posterior

probability. This analysis identified 51 credible set pairs with pos-

terior probability >95% for containing a causal variant. There were

four variants thatmapped to enhancer regions among the credible

sets: rs371063473 (LOC101929650), rs4809477 (SLCO4A1),

rs73719555 (LOC285766), and rs73669125 (LMX1B) (Table S7).

Of note, the rs73669125 variant is nearest to the LMX1B gene;

this gene is known to be involved in the development of the ante-

rior segment of the eye.57 Additionally, a mutation in LMX1B is

known to cause nail-patella syndrome (NPS), an inherited devel-

opmental disorder; one of the manifestations of NPS includes

the development of open-angle glaucoma.58–63 The rs73235527

variant maps to the geneBEND4; this gene is associatedwith reti-

nitis pigmentosa 26.64

In our investigation of potential biological insights from the

mega-analysis summary statistics, we conducted pathway ana-

lyses using the multivariate analysis of genomic annotation

(MAGMA) tool, which is seamlessly integrated within the func-

tional mapping and annotation (FUMA) platform.65,66 The

MAGMA gene set pathway analysis allowed us to explore the

enrichment of curated gene sets and Gene Ontology (GO) terms

sourced from MsigDB.67 Following the MAGMA analyses and

correction for multiple testing using the Bonferroni method, we

identified three pathways that exhibited significant enrichment

at a corrected p value threshold (pbon) < 0.05. The specific en-

riched pathways include negative regulation of translation in

response to stress (p = 5.14E�08), cellular response to leucine

starvation (p = 6.52E�08), and negative regulation of CREB tran-

scription factor activity (p = 3.52E�07).
Quantitative expression of POAG-associated genes in

human eye tissues

Thepathogenesis ofPOAG involves the lossofRGCsanddamage

to trabecular meshwork (TM) cells, which can be caused by

increased oxidative stress.68 We induced oxidative stress with

H2O2 inhumanocularcell lines tounderstandgeneexpressionpat-

terns and to determine this stressor’s functional relevance in

POAG.We quantified the expression profiles of the nearest genes

to the three likely causal variants from themega-analysis in human

TM (hTM) cells and RGCs derived from induced pluripotent stem

cells (iPSC-RGCs)69 treated with H2O2. We include a video that

demonstrates the electrophysiological responses of iPSC-RGCs

in cell culture at day 71 obtained using patch-clamp recording in

a whole-cell configuration mode (Video S1). These variants

included the two loci demonstrating replication in independent Af-

rican datasets (rs34957764mapping toROCK1P1 and rs1666698

mapping toDBF4P2) and the loci associatedwith baseline CDR in

genetic co-localization analyses (rs11824032 mapping to ARH-

GEF12). Gene expression in human retinal tissues from normal

and glaucoma donors was also quantified for these three near-

est genes.

When comparing hTMs under oxidative stress to untreated

hTMs, we found significant overexpression of ARHGEF12

(30.2± 1.05, p value = 0.008).ROCK1P1wasmarginally increased

(2.2 ± 0.5), while no expression of DBF4P2 was detected (Fig-

ure 5A). In stressed versus untreated iPSC-RGCs, the ROCK1P1

(2.45 ± 0.45, p value = 0.009) and DBF4P2 (2.15 ± 0.4, p value =

0.02) genes were significantly upregulated. There was an upward

trend in the expression of ARHGEF12 (1.6 ± 0.4) (Figure 5B).

With respect to human retinal tissue samples, there was no signif-

icant difference in the expression of ARHGEF12 (11.4 ± 4.05) and

ROCK1P1 (17.4±3.8) in the retinal tissue isolated fromaPOAGpa-

tient compared with the normal retinal tissue sample; however,

DBF4P2 (198 ± 0.3, p value: 0.0003) was significantly upregulated

in the retinal tissue fromthePOAGpatient (Figure5C). Thepossible

mechanisms involving the variants in the ARHGEF12 and

ROCK1P1 genes and their relation to POAG are shown in

Figure 5D.

In silico analysis of gene expression in ocular tissues
We assessed the expression levels for the genes that are nearest

to the loci that replicated in independent African ancestry data-

sets (ROCK1P1 andDBF4P1) and were associated with baseline

CDR in genetic co-localization analyses in silico (ARHGEF12),

wherever data were available. We used the most current

publicly available source, the Human Eye Transcriptome Atlas

(HETA).70,71 This database contains the largest number of

distinct ocular tissues, compared with all earlier databases,

and it is the only database employing fresh rather than postmor-

tem tissue for transcriptome analysis, a clear advantage for ac-

curacy of results. As there is no ocular tissue information in Ge-

notype-Tissue Expression (GTEx)72 or Encyclopedia of DNA

Elements (ENCODE)73 databases, we did not utilize these for

determining gene expression.

ARHGEF12 is widely expressed in ocular tissues, including the

ON.ROCK1P1 is expressed in theONand in the peripheral retina

but not in the retinal center. These findings are consistent with

the location of RGCs and their axons. The transcriptional profile
Cell 187, 464–480, January 18, 2024 471



Figure 5. Quantitative gene expression analyses results and proposed mechanisms of genes

Solid arrows indicate the proposed mechanisms of variants identified in our dataset and dashed arrows indicate a mechanism that is known in the literature.

*ROCK1P1 is a pseudogene and is a result of the partial duplication of the ROCK1 gene. A variant that maps to ROCK1P1 was identified in our discovery mega-

analysis and replicated in the AOU dataset.

The quantitative expression profiles of the nearest genes to the three likely causal variants from the mega-analysis (ARHGEF12, ROCK1P1, and DBF4P2) were

analyzed in human eye tissues (hTMs, iPSC-RGCs, and retinal cells) using quantitative RT-PCR, with and without H2O2 treatment.

(A) hTMs were treated with or without 100 mM H2O2.

(B) iPSC-RGCs were treated with or without 650 mM H2O2.

(C) Human retinal tissues from normal and glaucoma donors were used. GAPDH was used as the housekeeping gene. Data are presented as mean ± SEM. The

data were normalized to control cells and analyzed using Student’s t test statistical analysis (*p < 0.05, **p < 0.01, ***p < 0.001).

(D) Schematic representation of RhoA/ROCK downstream signaling pathway involving ARHGEF12 and ROCK1 and leading to glaucoma in individuals of African

ancestry. We propose that ARHGEF12 functions downstream of the TGF-b and RhoA/Rho-associated kinase signaling pathways to activate ROCK1 via RhoA,

which decreases TM plasticity and aqueous humor outflow and increases RGC death and optic nerve atrophy.

ll
Article
of DBF4P2 is not available. The only transcriptional profile avail-

able is for DBF4P1 and this would have uncertain relevance to

the gene of interest.
Polygenic prediction of POAG in African ancestry
Summary statistics from the discovery mega-analysis (African

ancestry individuals) were used to create a PRSMEGA, and

summary statistics from GBMI (European ancestry individuals)

were used to create a PRSGBMI. Each PRS was calculated

using PRS-CS and was applied to three independent data-

sets of African ancestry individuals (eMERGE, n = 13,493;

GGLAD-2, n = 1,490; PMBB, n = 11,117), and the AUC was

compared (Table S8). The highest AUC was achieved in the

GGLAD-2 data (AUC = 0.657) with summary statistics from the

mega-analysis. We also compared individuals identified in the

top 20% of the PRS by both mega-analysis and GBMI reference

data and the remainder of the dataset. The PRSMEGA outper-

forms the PRSGBMI in two out of the three datasets, as shown

in Figure 6 and Table S8.
472 Cell 187, 464–480, January 18, 2024
DISCUSSION

Only two percent of genetic studies have focused on individuals

of African ancestry as of 2019.74 Even diseases that dispropor-

tionately affect African ancestry individuals, such as POAG,

remain understudied in this population. To help address this

disparity, we conducted a very large GWAS on POAG in African

ancestry individuals to date (n = 11,275).We identified 46 risk loci

at genome-wide significance in our discovery cohort. Two of

these loci, mapping to the ROCK1P1 and DBF4P2 genes, repli-

cate in independent African ancestry datasets, and one previ-

ously undescribed locus mapping to the ARHGEF12 gene is

associated with baseline CDR in genetic co-localization ana-

lyses. We show through subsequent functional analyses that

these three loci can be logically implicated in African ancestry

POAG pathogenesis.

This mega-analysis of African ancestry individuals enables the

identification of previously undescribed risk loci for POAG, while

also replicating risk loci across ancestries. The majority of previ-

ously associated loci identified in other ancestral populations do



Figure 6. Performance of PRS generated from summary statistics from GBMI (European individuals) and the discovery mega-analysis

(African individuals) in independent datasets of African ancestry individuals

Summary statistics from GBMI (European individuals) and discovery mega-analysis (African individuals) were each used to create a PRS. Each PRS was

measured in three independent datasets of African ancestry individuals (eMERGE, GGLAD-2, PMBB).

(A) The Nagelkerke R2 is shown on the x axis, and the test datasets are shown on the y axis, with the p value corresponding to the Nagelkerke p value.

(B) The odds ratio (x axis)was calculatedbycomparing individuals in top 20%PRSwith rest of the testdatasets (y axis). Thep value corresponds to regressionp value.

See also Table S8.

ll
Article
not replicate at a conventional genome-wide significance level,

possibly due to genetic heterogeneity across different ancestry

groups. However, 21% replicate at the Bonferroni threshold p

value for 174 previously reported unique loci.

Significant findings from thediscoverymega-analysis replicate

in four independent datasets of African ancestry, chosen due to

their use of the same imputation panel as our study (with the

exception of AOU, which consists of whole-genome sequence

data). Two loci from the discovery mega-analysis replicate within

these African ancestry datasets. The rs34957764 variant map-

ping to the ROCK1P1 locus replicates in the AOU dataset, and

the rs1666698 variant mapping to the DBF4P2 gene replicates

in the MVP dataset and in meta-analyses of all replicated data-

sets (p value = 0.002); both replicated SNPs also showed consis-

tent effect estimates in most replication datasets. The observed

MAF differences between African and non-African populations

for rs1666698 and rs34957764 highlight the influence of popula-

tion-specific genetic factors in shaping allele frequencies. The

substantial variation in MAF suggests potential differences in

the evolutionary history and selective pressures experienced by

these populations. These findings support the importance of

considering population-specific effects when investigating the

functional and clinical implications of genetic variants. Further-

more, the contrasting MAF patterns underscore the need for

careful interpretation of genetic association studies conducted

in diverse populations. The substantial differences in MAF for

these two variants between African and non-African populations

may have implications for disease risk assessment, drug

response, and population-specific genetic studies.

ROCK1P1 and DBF4P2 are both pseudogenes. ROCK1P1 is

located on chromosome 18p11.32 and is a result of the partial
duplication of the ROCK1 gene on chromosome 18q11.1.

ROCK1 is a serine/threonine kinase and a downstream effector

of the Rho GTPase. ROCK1 is a target for Rho kinase inhibitors

to reduce aqueous humor outflow resistance by directly acting

on TM cells and Schlemm’s canal. These inhibitors are used as a

treatment for glaucoma.75–77 Despite a lack of clear evidence on

the strength of the underlyingmechanisms, our analysis highlights

the potential importance of these sites. TheMAF of rs34957764 is

0.25 in African ancestry samples, comparedwith 0.12 in European

ancestry samples. By contrast, the MAF of rs1666698 is substan-

tially higher in African and African American populations (MAF =

0.4) than in Europeans (MAF = 0.02). These findings underscore

the need for additional studies to elucidate the functional signifi-

cance of these SNPs and their potential contribution to disease

susceptibility in diverse populations.

We also identify genetic co-localization between a lead variant

mapping toARHGEF12 and baseline CDR. TheARHGEF12 gene

has previously been associated with elevated IOP in individuals

of European ancestry.46 The genetic co-localization between

rs11824032 and baseline CDR is not previously reported.

Although the significant MR p value provides supporting evi-

dence for a potential causal association, we acknowledge the

limitations of interpreting causality based on a single-SNP MR

test. It is important to consider the complexities of establishing

directionality in such analyses. Furthermore, we acknowledge

that the observed association between rs11824032 and baseline

CDRmay be influenced by other factors specific to this ancestry

group. For example, individuals in this group may experience

earlier and more severe neurodegeneration compared with

less affected ancestry groups, which could also contribute to

the observed association.
Cell 187, 464–480, January 18, 2024 473



ll
Article
Our quantitative real-time PCR data indicate that ROCK1P1

mRNA is significantly overexpressed in iPSC-RGCs under oxida-

tive stressed conditions. In silico analysis demonstrates that

ROCK1P1 is expressed in the ON and peripheral retina, consis-

tent with the location of RGCs and their axons. ARHGEF12 is

overexpressed in hTM cells under stressed conditions and

shows a trend toward overexpression in iPSC-RGCs but fails

to reach significance. In silico analysis confirms that ARHGEF12

is expressed in the ON. There are also upward trends detected in

the expression of the ARHGEF12 and ROCK1P1 genes in the

retinal tissue isolated from a POAG patient, but these expression

profiles are not statistically different when compared with control

retina. The retina is comprised many different layers, and it is

possible that the full retinal biopsy obscures the expression of

genes located in the RGC layer, which is most affected by

glaucoma.

The ARHGEF12 (Rho guanine nucleotide exchange factor

[GEF] 12) gene is located on chromosome 11q23.3; an intronic

variant rs58073046 in this gene has previously been associated

with glaucoma and IOP.46 ARHGEF12 is reported to function

downstream of the RhoA/Rho-associated kinase pathway,

which is known to regulate TM plasticity via actin-myosin inter-

actions.78–80 It is proposed that ARHGEF12 activates RhoA,

which leads to ROCK activation, causing a decrease in aqueous

humor outflow and permeability of Schlemm’s canal cells.81 The

result is a subsequent increase in IOP, contributing to increased

CDR in African ancestry populations. Additionally, the ARH-

GEF12 gene interacts with ABCA1 and is reported to bind

directly to the C terminus of ABCA1 to activate RhoA. Simulta-

neously, ARHGEF12 prevents the degradation of ABCA1 by ex-

tending its half-life.82,83 Variants in ABCA1 have previously been

found to be associated with both POAG and IOP.84,85

RhoA/ROCK and transforming growth factor (TGF)-b

signaling pathways also play a significant role in the pathogen-

esis of POAG.86–88 TGF-b2 is the predominant form of the three

isoforms in the eye.89,90 Analyses of the aqueous humor and

TM cells from glaucoma patients have detected increased

levels of TGF-b2,91 causing high levels of ECM deposition in

TM, ON head, and lamina cribrosa,88,92 leading to the death

of RGCs and atrophy of the ON. This mechanism can possibly

explain the increase in CDR observed in our group of African

ancestry individuals that carry the ARHGEF12 and ROCK1P1

variants.

This study demonstrates that the inclusion of African ancestry

individuals in several analyses improves the prediction of POAG

risk for these individuals. First, a comparison of per-allele effect

sizes demonstrates that previously undiscovered variants from

the mega-analysis have large effect sizes in African ancestry in-

dividuals, whereas previously reported variants (mainly from

European individuals) tend to have smaller effect sizes that point

to the polygenic tail of POAG risk.93 Additionally, in a cross-

ancestry genetic correlation analysis, we demonstrate lower ge-

netic correlation among African and East Asian individuals (59%)

and African and European individuals (73%), compared with Eu-

ropean and East Asian individuals (78%). It is certainly possible

that the smaller effect sizes in our study are not due to differ-

ences in biology but rather to the Winner’s Curse,94 whereby

the initial discovery (in Europeans) is an overestimate of the
474 Cell 187, 464–480, January 18, 2024
effect sizes. Additional studies should be conducted in large

global populations to obtain more estimates of the effect size

of these variants. We also show that the inclusion of both previ-

ously known and undiscovered loci from our African ancestry

cohort helps to increase the estimate of variance for POAG

from GCTA-REML analyses. Finally, we demonstrate that the

PRSMEGA (African ancestry) showed an improved prediction of

disease risk over the PRSGBMI (European ancestry) in two out

of three African ancestry cohorts, despite the much smaller

size of the PRSMEGA dataset. Notably, the PRSMEGA demon-

strated an improved ability to discern individuals in the top 20th

percentile of PRS distribution, suggesting the clinical utility of

the PRS in African ancestry individuals.

This study has several major strengths, beginning with the

implementation of data integration strategies to combine geno-

type data from 11,275 African ancestry individuals into a mega-

analysis dataset, allowing us to achieve substantial statistical

power. Several previous studies have shown that mega-ana-

lyses combining individual-level data are superior to meta-ana-

lyses, which rely on summary statistics from numerous hetero-

geneous samples.95,96 Our findings further support the existing

literature, as we demonstrate a strong correlation (r = 0.86) be-

tween the mega-analysis and meta-analysis results. Moreover,

our results suggest that the mega-analysis approach, which

combines individual-level data, exhibits higher statistical po-

wer, particularly for detecting smaller-effect-sized variants.

Although mega-analyses may not always be practical due to

constraints such as data availability and other logistical chal-

lenges, our study was fortunate to have comprehensive individ-

ual-level data across multiple datasets. This approach allows

us to pool the row-level data, thereby increasing the statistical

power of our analysis and facilitating more precise estimates of

the effects under investigation. Another strength of our study is

that the POAAGG study carefully ascertained quantitative

phenotypic traits from each subject, which is crucial to prevent

residual confounding effects of unmeasured phenotypes within

association studies. This rich dataset allows us to study and

identify significant differences in quantitative trait measure-

ments for genotypes from our mega-analyses in a case-only

analysis. Finally, we employ multi-factorial strategies to

improve the detection of potential causal variants, genes, and

biological pathways that contribute to POAG pathophysiology.

By incorporating functionally informed fine-mapping methods,

co-localization of GWAS signals from binary and quantitative

phenotypes, and in silico validation methods, we are able to

prioritize variants and systemically characterize genes in path-

ways to elucidate POAG mechanisms.

In conclusion, this study greatly expands our knowledge of the

genetic landscape of POAG in African ancestry individuals.

Though the increased burden of POAG in this ancestry group

was identified more than three decades ago, new generations

continue to experience premature vision loss from this familial

disease. A critical barrier to progress has been the lack of genetic

studies of POAG in African ancestry individuals. Genetic studies

are needed to identify novel targets for screening and therapeu-

tic intervention, as all treatment options target only one disease

mechanism, elevated IOP, with limited success.97 This study

represents a crucial first step toward addressing this disparity



ll
Article
by conducting the largest-ever GWAS on POAG in African

ancestry individuals. Althoughmore experimental evidence is ul-

timately required to pinpoint causal mechanisms and generate

predictive tools for early screening of POAG, many useful in-

sights can be drawn from our study. Our work is an important

step toward achieving future goals, including defining subgroups

of disease that aid in early detection, providing the capability

for early screening within families, and discovering targetable

pathways for personalized therapeutic interventions. Future

studies can also help to determine whether African ancestry in-

dividuals have different responses to treatments, such as

recently approved ROCK inhibitors. POAG is the leading cause

of irreversible blindness in theworld today, and familial blindness

perpetuates increased morbidity, poverty, and mortality across

generations. The lack of genetic studies in the most affected Af-

rican ancestry population is both a failure of science and a failure

of our moral obligation to address systemic racism prevalently

visible today.

Limitations of the study
A limitation of this study is the relatively small sample size of our

African ancestry discovery cohort. Although this may be the

largest African ancestry GWAS for POAG to date, it is still smaller

than many GWAS in datasets of European ancestry individuals.

Nevertheless, the African ancestry PRSMEGA for POAG risk strat-

ification still shows improvement over European ancestry

PRSGBMI, which was drawn from a much larger dataset

(GBMI). Recruiting more individuals of African ancestry should

be a priority for developing amore refined PRSwith clinical utility

for POAG diagnosis in African ancestry populations; this is

imperative because 50% of patients with POAG are unaware

that they are afflicted with this blinding disease. A second limita-

tion of the study is posed by the remarkable diversity of African

ancestry populations, which are more dissimilar than any other

ancestry group worldwide.98 This diversity could provide one

reason for lack of replication of many loci in our replication data-

sets. Additionally, many of the replication datasets relied on

diagnosis code-based case/control definitions for POAG, which

could also be responsible for the lack of replication of many sig-

nals. These datasets were also characterized by imbalanced

case/control ratios, which may introduce some degree of uncer-

tainty or misclassification. Although we estimate effect sizes in

these independent replication datasets and perform meta-ana-

lyses to enhance robustness, it remains essential to interpret

the observed effect sizes with caution, due to the potential

impact of these limitations on the precision and accuracy of re-

sults. Another possible limitation of our study is the potential

impact of study source as a confounding factor. Although we fol-

lowed the imputation strategy shown to be robust in previous

studies,99,100 study source can still represent a potential source

of bias.96 However, it is worth noting that in these analyses, using

study source as a covariate was challenging due to the uneven

distribution of cases and controls among the three datasets in

the discovery analyses. Although the GGLAD and POAAGG da-

tasets have fairly equal divisions between cases and controls,

the ADAGES study has a very high distribution of cases (92%),

which can cause the ADAGES batch to be highly correlated

with case status. This would result in multicollinearity issues in
the sensitivity analyses. We also do not believe that including

the study source as a covariate in sensitivity analyses would sub-

stantially alter our main findings. Nonetheless, we acknowledge

the potential impact of the study source on our results and have

taken steps to account for potential confounding factors through

estimating the inflation factor and performing conditional ana-

lyses and replication analyses. We performed sensitivity ana-

lyses that compare the results of meta-analyses of the discovery

datasets and mega-analyses, providing additional insights into

the robustness of our findings. Finally, although we used today’s

industry-standard analyses for imputation, GWAS analysis, and

PRS analysis, it will be important to continue to develop new

methodologies that incorporate local ancestry into the analyses,

which we anticipate will further improve the identification of both

ancestry-specific association signals as well as those that are

relevant across multiple ancestry groups.

STAR+METHODS

Detailed methods are provided in the online version of this paper

and include the following:

d KEY RESOURCES TABLE

d RESOURCE AVAILABILITY
B Lead contact

B Materials availability

B Data and code availability

d EXPERIMENTAL MODEL AND STUDY PARTICIPANT

DETAILS

B Human Participants

B Cells Lines

d METHOD DETAILS

B Study Design and Participants

B Quality Control and Imputation

B Discovery Mega-Analysis

B Sensitivity Analyses

B Identifying Known Risk Loci and Associations

B Sex-Stratified Analysis

B Trait Co-localization Analysis

B Replication and Joint Meta-Analysis

B Heritability Estimation

B Cross-Ancestry Comparison

B Functionally Informed Fine-Mapping

B Pathway Analyses

B Evaluating Oxidative Stress in TM cells and

iPSC-RGCs

B In Silico Analyses for Gene Function

B Polygenic Risk Scores (PRS)

d QUANTIFICATION AND STATISTICAL ANALYSIS

SUPPLEMENTAL INFORMATION

Supplemental information can be found online at https://doi.org/10.1016/j.cell.

2023.12.006.

ACKNOWLEDGMENTS

We greatly appreciate the work of the Clinical Trial Coordinators in the

POAAGG study, who made it possible to enroll more than 10,200 individuals
Cell 187, 464–480, January 18, 2024 475

https://doi.org/10.1016/j.cell.2023.12.006
https://doi.org/10.1016/j.cell.2023.12.006


ll
Article
of African ancestry in Philadelphia. We acknowledge the Penn Medicine

BioBank (PMBB) for providing data and thank the patients of Penn Medicine

who consented to participate in this research program. We would also like

to thank the Penn Medicine BioBank team and Regeneron Genetics Center

for providing genetic variant data for analysis. The PMBB is approved under

IRB protocol# 813913 and supported by Perelman School of Medicine at Uni-

versity of Pennsylvania, a gift from the Smilow family, and the National Center

for Advancing Translational Sciences of the National Institutes of Health under

CTSA award number UL1TR001878. We are grateful to the VA VINCI and

GENISIS support and MVP Core Statistical Analysis teams. We would like to

thank Dr. Pieter Bonnemaijer for attempting replication of our signals in glau-

coma patients from the African descent study (GIGA) dataset. We would

also like to thank Dr. Tess Cherlin for support in creating Figure 2, Dr. Theodore

Drivas for helping with Figure 3, Sergei Nikonov for helping create Video S1,

and Yan Zhu for uploading data to Mendeley Data. Support for title page cre-

ation and format was provided by AuthorArranger, a tool developed at the Na-

tional Cancer Institute. All groups appreciate the critical contribution made by

the enrollees. The Primary Open-Angle African American Glaucoma Genetics

(POAAGG) study was supported by the National Eye Institute, Bethesda,

Maryland (grant #1R01EY023557-01) and the Vision Research Core Grant

(P30 EY001583). Funds also come from the F.M. Kirby Foundation, Research

to Prevent Blindness, The UPenn Hospital Board of Women Visitors, and The

Paul and Evanina Bell Mackall Foundation Trust. Support also came from Re-

generon Genetics Center, the Ophthalmology Department at the Perelman

School of Medicine, the Genetics Department at the Perelman School of Med-

icine, and the VA Hospital in Philadelphia, PA. The Genetics of Glaucoma in

People of African Descent (GGLAD) study was funded by the National Eye

Institute and National Human Genome Research Center (U54HG009826).

ADAGES3-Genetics was supported by R01EY023704, and other support

included R01EY011008, R01EY019869, R01 EY021818, P30 EY022589, T32

EY026590, and an unrestricted grant from Research to Prevent Blindness

(New York, NY). J.I.R. was supported in part by the National Center for

Advancing Translational Sciences, CTSI grant UL1TR001881, the National

Institute of Diabetes and Digestive and Kidney Disease Diabetes Research

Center (DRC) grant DK063491 to the Southern California Diabetes Endocri-

nology Research, and contract R01EY023704. Infrastructure for the

CHARGE Consortium was supported in part by the National Heart, Lung,

and Blood Institute (NHLBI) grant R01HL105756. MVP research is based on

data from the Million Veteran Program, Office of Research and Development,

Veterans Health Administration and was supported by award I01 BX004557.

This publication does not represent the views of the Department of Veteran Af-

fairs or the United States Government. This work was also funded by the

Cleveland Institute for Computational Biology, NIH Core grants (P30

EY025585 and P30 EY011373) and unrestricted grants from Research to Pre-

vent Blindness to Case Western Reserve University (CWRU) and Cleveland

Clinic Lerner College of Medicine of CWRU. J.C.B. and T.G.K. are also sup-

ported by NIH R01 EY033829. The sponsor or funding organization had no

role in the design or conduct of this research.

AUTHOR CONTRIBUTIONS

Conceptualization: S.S.V., H.V.G., M.D.R., and J.M.O.; methodology: S.S.V.,

H.V.G., V.R.M.C., M.D.R., and J.M.O.; software: S.S.V., H.V.G., Y.B., L.G.,

A.L., M.D.R., and J.M.O.; validation: S.S.V., H.V.G., B.Z., A.S.B., C.P.M.,

M.C.Z., S.M.W., J.C.B., J.I.R., R.W., C.C.K., M.A.H., M.D.R., and J.M.O.;

formal analysis: S.S.V., H.V.G., Y.B., M.P., G.-s.Y., M.D.R., and J.M.O.; inves-

tigation: S.S.V., H.V.G., V.R.M.C., R.J.S., D.W.C., V.V., R.M.N., S.R., J.H., R.L.,

S.Z.-G., A.S.B., N.K., E.D., W.M., J.H., Regeneron Genetics Center, T.G.K.,

S.K.I., N.S.P., VA Million Veteran Program, K.D.T., X.G., Y.-D.I.C., L.Z., C.G.,

R.A., J.L.W., C.M.C.-O., S.E.W., S.A., D.L.B., O.O.O., M.R., A.A., O.M.A.,

T.A., J.L.W., A.G.R., Q.N.C., V.A., A.L., E.M.-E., P.S.S., J.C.B., J.I.R., R.W.,

C.C.K., M.A.H., M.D.R., and J.M.O.; resources: M.D.R. and J.M.O.; data cura-

tion: R.L., E.D., M.D.R., and J.M.O.; writing – original draft: S.S.V., H.V.G.,

V.R.M.C., and R.J.S.; writing – review & editing: all authors; visualization:

S.S.V., H.V.G., V.R.M.C., Y.B., V.V., M.D.R., and J.M.O.; supervision: M.D.R.

and J.M.O.; project administration: S.S.V., H.V.G., R.J.S., M.D.R., and

J.M.O.; funding acquisition: M.D.R. and J.M.O.
476 Cell 187, 464–480, January 18, 2024
DECLARATION OF INTERESTS

J.M.O. is a member of the scientific advisory board of Life Biosciences and a

paid consultant of Atheneum Partners, Cerner Enviza (includes Kantar

Health), and Calico. A.G.R. holds intellectual property for the use of gene

therapy to treat glaucoma. E.M.-E. is a scientific advisor for Avisi and a

paid consultant of Aerie Pharmaceuticals, Allergan, Eyenovia, and Thea

Pharma. J.L. receives instrument support from Carl Zeiss Meditech, Inc.,

and Heidelberg Engineering, GmBH; receives research support from Novar-

tis, Inc.; and is a paid consultant at Thea, Inc., Alcon Laboratories, Inc., John-

son & Johnson, Inc., Abbvie, Inc., Carl Zeiss Meditech, Inc., Genetech, Inc.,

and ONL Therapeutics, Inc.

INCLUSION AND DIVERSITY

Weworked to ensure gender balance in the recruitment of human subjects.We

worked to ensure ethnic or other types of diversity in the recruitment of human

subjects. We worked to ensure that the study questionnaires were prepared in

an inclusive way. We worked to ensure diversity in experimental samples

through the selection of the cell lines. We worked to ensure diversity in exper-

imental samples through the selection of the genomic datasets. One ormore of

the authors of this paper self-identifies as an underrepresented ethnic minority

in their field of research or within their geographical location. One or more of

the authors of this paper self-identifies as a gender minority in their field of

research. One or more of the authors of this paper self-identifies as a member

of the LGBTQIA+ community. One or more of the authors of this paper self-

identifies as living with a disability. One or more of the authors of this paper

received support from a program designed to increase minority representation

in their field of research. While citing references scientifically relevant for this

work, we also actively worked to promote gender balance in our reference

list. We avoided ‘‘helicopter science’’ practices by including the participating

local contributors from the region where we conducted the research as au-

thors on the paper.

Received: April 5, 2023

Revised: July 24, 2023

Accepted: December 4, 2023

Published: January 18, 2024

REFERENCES

1. Weinreb, R.N., Leung, C.K., Crowston, J.G., Medeiros, F.A., Friedman,

D.S., Wiggs, J.L., andMartin, K.R. (2016). Primary open-angle glaucoma.

Nat. Rev. Dis. Primers 2, 16067.

2. Tham, Y.C., Li, X., Wong, T.Y., Quigley, H.A., Aung, T., and Cheng, C.Y.

(2014). Global prevalence of glaucoma and projections of glaucoma

burden through 2040: a systematic review and meta-analysis. Ophthal-

mology 121, 2081–2090.

3. Quigley, H.A., and Broman, A.T. (2006). The number of people with glau-

coma worldwide in 2010 and 2020. Br. J. Ophthalmol. 90, 262–267.

4. Tielsch, J.M., Katz, J., Sommer, A., Quigley, H.A., and Javitt, J.C. (1994).

Family history and risk of primary open angle glaucoma. The Baltimore

eye survey. Arch. Ophthalmol. 112, 69–73.

5. Muñoz, B., West, S.K., Rubin, G.S., Schein, O.D., Quigley, H.A., Bressler,

S.B., and Bandeen-Roche, K. (2000). Causes of blindness and visual

impairment in a population of older Americans: the Salisbury Eye Evalu-

ation Study. Arch. Ophthalmol. 118, 819–825.

6. Janssen, S.F., Gorgels, T.G., Ramdas, W.D., Klaver, C.C., van Duijn,

C.M., Jansonius, N.M., and Bergen, A.A. (2013). The vast complexity of

primary open angle glaucoma: disease genes, risks, molecular mecha-

nisms and pathobiology. Prog. Retin. Eye Res. 37, 31–67.

7. Prum, B.E., Jr., Rosenberg, L.F., Gedde, S.J., Mansberger, S.L., Stein,

J.D., Moroi, S.E., Herndon, L.W., Jr., Lim, M.C., and Williams, R.D.

(2016). Primary open-angle glaucoma preferred practice pattern� guide-

lines. Ophthalmology 123, P41–P111.

http://refhub.elsevier.com/S0092-8674(23)01338-7/sref1
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref1
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref1
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref2
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref2
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref2
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref2
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref3
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref3
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref4
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref4
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref4
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref5
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref5
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref5
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref5
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref6
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref6
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref6
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref6
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref7
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref7
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref7
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref7


ll
Article
8. Heijl, A., Bengtsson, B., Hyman, L., and Leske,M.C.; EarlyManifest Glau-

coma Trial Group (2009). Natural history of open-angle glaucoma.

Ophthalmology 116, 2271–2276.

9. Anderson, D.R., Drance, S.M., and Schulzer, M.; Collaborative Normal-

Tension Glaucoma StudyGroup (2001). Natural history of normal-tension

glaucoma. Ophthalmology 108, 247–253.

10. Heijl, A., Leske,M.C., Bengtsson, B., Hyman, L., Bengtsson, B., and Hus-

sein, M.; Early Manifest Glaucoma Trial Group (2002). Reduction of intra-

ocular pressure and glaucoma progression: results from the Early Mani-

fest Glaucoma Trial. Arch. Ophthalmol. 120, 1268–1279.

11. Gordon, M.O., Beiser, J.A., Brandt, J.D., Heuer, D.K., Higginbotham,

E.J., Johnson, C.A., Keltner, J.L., Miller, J.P., Parrish, R.K., 2nd, Wilson,

M.R., et al. (2002). The Ocular Hypertension Treatment Study: baseline

factors that predict the onset of primary open-angle glaucoma. Arch.

Ophthalmol. 120. 714-20; discussion 829.

12. Leske, M.C., Wu, S.Y., Hennis, A., Honkanen, R., and Nemesure, B.;

Bess Study Group (2008). Risk factors for incident open-angle glaucoma:

the Barbados Eye Studies. Ophthalmology 115, 85–93.

13. O’Brien, J.M., Salowe, R.J., Fertig, R., Salinas, J., Pistilli, M., Sankar,

P.S., Miller-Ellis, E., Lehman, A., Murphy, W.H.A., Homsher, M., et al.

(2018). Family history in the primary open-angle African American glau-

coma genetics study cohort. Am. J. Ophthalmol. 192, 239–247.

14. Teikari, J.M. (1987). Genetic factors in open-angle (simple and capsular)

glaucoma. A population-based twin study. Acta Ophthalmol. (Copenh)

65, 715–720.

15. Gottfredsdottir, M.S., Sverrisson, T., Musch, D.C., and Stefansson, E.

(1999). Chronic open-angle glaucoma and associated ophthalmic find-

ings in monozygotic twins and their spouses in Iceland. J. Glaucoma 8,

134–139.

16. Liu, Y., and Allingham, R.R. (2017). Major review: molecular genetics of

primary open-angle glaucoma. Exp. Eye Res. 160, 62–84.

17. Zhou, W., Kanai, M., Wu, K.H., Rasheed, H., Tsuo, K., Hirbo, J.B., Wang,

Y., Bhattacharya, A., Zhao, H., Namba, S., et al. (2022). Global biobank

Meta-analysis Initiative: powering genetic discovery across human dis-

ease. Cell Genomics 2, 100192.

18. Burdon, K.P., Macgregor, S., Hewitt, A.W., Sharma, S., Chidlow, G.,

Mills, R.A., Danoy, P., Casson, R., Viswanathan, A.C., Liu, J.Z., et al.

(2011). Genome-wide association study identifies susceptibility loci for

open angle glaucoma at TMCO1 and CDKN2B-AS1. Nat. Genet. 43,

574–578.

19. Gibson, J., Griffiths, H., De Salvo, G., Cole, M., Jacob, A., Macleod, A.,

Yang, Y., Menon, G., Cree, A., Ennis, S., et al. (2012). Genome-wide as-

sociation study of primary open angle glaucoma risk and quantitative

traits. Mol. Vis. 18, 1083–1092.

20. Thorleifsson, G., Magnusson, K.P., Sulem, P., Walters, G.B., Gudbjarts-

son, D.F., Stefansson, H., Jonsson, T., Jonasdottir, A., Jonasdottir, A.,

Stefansdottir, G., et al. (2007). Common sequence variants in the

LOXL1 gene confer susceptibility to exfoliation glaucoma. Science 317,

1397–1400.

21. Thorleifsson, G., Walters, G.B., Hewitt, A.W., Masson, G., Helgason, A.,

DeWan, A., Sigurdsson, A., Jonasdottir, A., Gudjonsson, S.A., Magnus-

son, K.P., et al. (2010). Common variants near CAV1 and CAV2 are asso-

ciated with primary open-angle glaucoma. Nat. Genet. 42, 906–909.

22. Wiggs, J.L., Yaspan, B.L., Hauser, M.A., Kang, J.H., Allingham, R.R., Ol-

son, L.M., Abdrabou, W., Fan, B.J., Wang, D.Y., Brodeur, W., et al.

(2012). Common variants at 9p21 and 8q22 are associated with

increased susceptibility to optic nerve degeneration in glaucoma. PLoS

Genet. 8, e1002654.

23. Nakano, M., Ikeda, Y., Taniguchi, T., Yagi, T., Fuwa, M., Omi, N., Tokuda,

Y., Tanaka, M., Yoshii, K., Kageyama, M., et al. (2009). Three susceptible

loci associated with primary open-angle glaucoma identified by genome-

wide association study in a Japanese population. Proc. Natl. Acad. Sci.

USA 106, 12838–12842.
24. Nakano, M., Ikeda, Y., Tokuda, Y., Fuwa, M., Omi, N., Ueno, M., Imai, K.,

Adachi, H., Kageyama, M., Mori, K., et al. (2012). Common variants in

CDKN2B-AS1 associated with optic-nerve vulnerability of glaucoma

identified by genome-wide association studies in Japanese. PLoS One

7, e33389.

25. Osman, W., Low, S.K., Takahashi, A., Kubo, M., and Nakamura, Y.

(2012). A genome-wide association study in the Japanese population

confirms 9p21 and 14q23 as susceptibility loci for primary open angle

glaucoma. Hum. Mol. Genet. 21, 2836–2842.

26. Takamoto, M., Kaburaki, T., Mabuchi, A., Araie, M., Amano, S., Aihara,

M., Tomidokoro, A., Iwase, A., Mabuchi, F., Kashiwagi, K., et al. (2012).

Common variants on chromosome 9p21 are associated with normal ten-

sion glaucoma. PLoS One 7, e40107.

27. Inoko, A., H., Ota, M., Mizuki, N., and Bahram, S. (2010). Genome-wide

association study of normal tension glaucoma: common variants in

SRBD1 and ELOVL5 contribute to disease susceptibility. Ophthalmology

117, 1331–1338.e5.

28. Bailey, J.N., Loomis, S.J., Kang, J.H., Allingham, R.R., Gharahkhani, P.,

Khor, C.C., Burdon, K.P., Aschard, H., Chasman, D.I., Igo, R.P., Jr., et al.

(2016). Genome-wide association analysis identifies TXNRD2, ATXN2

and FOXC1 as susceptibility loci for primary open-angle glaucoma.

Nat. Genet. 48, 189–194.

29. Gharahkhani, P., Jorgenson, E., Hysi, P., Khawaja, A.P., Pendergrass, S.,

Han, X., Ong, J.S., Hewitt, A.W., Segrè, A.V., Rouhana, J.M., et al. (2021).

Genome-wide meta-analysis identifies 127 open-angle glaucoma loci

with consistent effect across ancestries. Nat. Commun. 12, 1258.

30. Hoffmann, T.J., Tang, H., Thornton, T.A., Caan, B., Haan, M., Millen, A.E.,

Thomas, F., and Risch, N. (2014). Genome-wide association and admix-

ture analysis of glaucoma in the Women’s Health Initiative. Hum. Mol.

Genet. 23, 6634–6643.

31. Bonnemaijer, P.W.M., Iglesias, A.I., Nadkarni, G.N., Sanyiwa, A.J., Has-

san, H.G., Cook, C., GIGA Study Group, Simcoe, M., Taylor, K.D., Schur-

mann, C., et al. (2018). Genome-wide association study of primary open-

angle glaucoma in continental and admixed African populations. Hum.

Genet. 137, 847–862.

32. Liu, Y., Hauser, M.A., Akafo, S.K., Qin, X., Miura, S., Gibson, J.R.,

Wheeler, J., Gaasterland, D.E., Challa, P., Herndon, L.W., et al. (2013).

Investigation of known genetic risk factors for primary open angle glau-

coma in two populations of African ancestry. Invest. Ophthalmol. Vis.

Sci. 54, 6248–6254.

33. Choquet, H., Paylakhi, S., Kneeland, S.C., Thai, K.K., Hoffmann, T.J., Yin,

J., Kvale, M.N., Banda, Y., Tolman, N.G., Williams, P.A., et al. (2018). A

multiethnic genome-wide association study of primary open-angle glau-

coma identifies novel risk loci. Nat. Commun. 9, 2278.

34. Gharahkhani, P., Burdon, K.P., Cooke Bailey, J.N., Hewitt, A.W., Law,

M.H., Pasquale, L.R., Kang, J.H., Haines, J.L., Souzeau, E., Zhou, T.,

et al. (2018). Analysis combining correlated glaucoma traits identifies

five new risk loci for open-angle glaucoma. Sci. Rep. 8, 3124.

35. Zangwill, L.M., Ayyagari, R., Liebmann, J.M., Girkin, C.A., Feldman, R.,

Dubiner, H., Dirkes, K.A., Holmann, M., Williams-Steppe, E., Hammel,

N., et al. (2019). The African Descent and Glaucoma Evaluation Study

(ADAGES) III: Contribution of genotype to glaucoma phenotype in African

Americans: study design and baseline data. Ophthalmology 126,

156–170.

36. Taylor, K.D., Guo, X., Zangwill, L.M., Liebmann, J.M., Girkin, C.A., Feld-

man, R.M., Dubiner, H., Hai, Y., Samuels, B.C., Panarelli, J.F., et al.

(2019). Genetic architecture of primary open-angle glaucoma in individ-

uals of African descent: the African descent and glaucoma evaluation

Study III. Ophthalmology 126, 38–48.

37. Genetics of Glaucoma in People of African Descent (GGLAD) Con-

sortium, Hauser, M.A., Allingham, R.R., Aung, T., Van Der Heide, C.J.,

Taylor, K.D., Rotter, J.I., Wang, S.J., Bonnemaijer, P.W.M., Williams,

S.E., et al. (2019). Association of genetic variants with primary open-
Cell 187, 464–480, January 18, 2024 477

http://refhub.elsevier.com/S0092-8674(23)01338-7/sref8
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref8
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref8
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref9
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref9
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref9
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref10
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref10
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref10
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref10
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref11
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref11
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref11
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref11
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref11
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref12
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref12
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref12
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref13
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref13
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref13
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref13
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref14
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref14
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref14
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref15
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref15
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref15
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref15
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref16
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref16
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref17
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref17
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref17
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref17
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref18
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref18
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref18
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref18
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref18
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref19
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref19
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref19
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref19
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref20
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref20
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref20
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref20
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref20
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref21
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref21
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref21
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref21
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref22
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref22
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref22
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref22
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref22
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref23
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref23
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref23
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref23
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref23
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref24
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref24
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref24
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref24
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref24
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref25
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref25
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref25
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref25
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref26
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref26
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref26
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref26
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref27
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref27
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref27
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref27
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref28
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref28
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref28
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref28
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref28
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref29
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref29
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref29
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref29
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref30
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref30
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref30
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref30
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref31
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref31
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref31
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref31
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref31
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref32
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref32
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref32
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref32
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref32
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref33
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref33
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref33
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref33
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref34
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref34
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref34
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref34
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref35
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref35
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref35
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref35
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref35
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref35
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref36
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref36
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref36
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref36
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref36
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref37
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref37
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref37
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref37


ll
Article
angle glaucoma among individuals with African Ancestry. JAMA 322,

1682–1691.

38. Charlson, E.S., Sankar, P.S., Miller-Ellis, E., Regina, M., Fertig, R., Sali-

nas, J., Pistilli, M., Salowe, R.J., Rhodes, A.L., Merritt, W.T., 3rd., et al.

(2015). The primary open-angle African American glaucoma genetics

study: baseline demographics. Ophthalmology 122, 711–720.

39. Das, S., Forer, L., Schönherr, S., Sidore, C., Locke, A.E., Kwong, A.,

Vrieze, S.I., Chew, E.Y., Levy, S., McGue, M., et al. (2016). Next-genera-

tion genotype imputation service and methods. Nat. Genet. 48,

1284–1287.

40. All of Us Research Program Investigators, Denny, J.C., Rutter, J.L., Gold-

stein, D.B., Philippakis, A., Smoller, J.W., Jenkins, G., and Dishman, E.

(2019). The ‘‘all of us’’ research program. N. Engl. J. Med. 381, 668–676.

41. Lau-Min, K.S., Asher, S.B., Chen, J., Domchek, S.M., Feldman, M., Joffe,

S., Landgraf, J., Speare, V., Varughese, L.A., Tuteja, S., et al. (2021).

Real-world integration of genomic data into the electronic health record:

the PennChart Genomics Initiative. Genet. Med. 23, 603–605.

42. Gaziano, J.M., Concato, J., Brophy, M., Fiore, L., Pyarajan, S., Breeling,

J., Whitbourne, S., Deen, J., Shannon, C., Humphries, D., et al. (2016).

Million Veteran Program: a mega-biobank to study genetic influences

on health and disease. J. Clin. Epidemiol. 70, 214–223.

43. Welter, D., MacArthur, J., Morales, J., Burdett, T., Hall, P., Junkins, H.,

Klemm, A., Flicek, P., Manolio, T., Hindorff, L., et al. (2014). The NHGRI

GWAS Catalog, a curated resource of SNP-trait associations. Nucleic

Acids Res. 42, D1001–D1006.

44. Yang, J., Lee, S.H., Goddard, M.E., and Visscher, P.M. (2011). GCTA: a

tool for genome-wide complex trait analysis. Am. J. Hum. Genet.

88, 76–82.

45. Yang, J., Ferreira, T., Morris, A.P., Medland, S.E., Genetic Investigation of

ANthropometric Traits (GIANT) Consortium; DIAbetes Genetics Replica-

tion AndMeta-analysis (DIAGRAM) Consortium, Madden, P.A.F., Health,

A.C., Martin, N.G., Montgomery, G.W., et al. (2012). Conditional and joint

multiple-SNP analysis of GWAS summary statistics identifies additional

variants influencing complex traits. Nat. Genet. 44, 369–375.

46. Springelkamp, H., Iglesias, A.I., Cuellar-Partida, G., Amin, N., Burdon,

K.P., van Leeuwen, E.M., Gharahkhani, P., Mishra, A., van der Lee,

S.J., Hewitt, A.W., et al. (2015). ARHGEF12 influences the risk of glau-

coma by increasing intraocular pressure. Hum. Mol. Genet. 24,

2689–2699.

47. Gharahkhani, P., Burdon, K.P., Fogarty, R., Sharma, S., Hewitt, A.W.,

Martin, S., Law, M.H., Cremin, K., Bailey, J.N.C., Loomis, S.J., et al.

(2014). Common variants near ABCA1, AFAP1 and GMDS confer risk

of primary open-angle glaucoma. Nat. Genet. 46, 1120–1125.

48. Morales, J., Al-Sharif, L., Khalil, D.S., Shinwari, J.M.A., Bavi, P., Al-Mah-

rouqi, R.A., Al-Rajhi, A., Alkuraya, F.S., Meyer, B.F., and Al Tassan, N.

(2009). Homozygous mutations in ADAMTS10 and ADAMTS17 cause

lenticular myopia, ectopia lentis, glaucoma, spherophakia, and short

stature. Am. J. Hum. Genet. 85, 558–568.

49. Marzin, P., Cormier-Daire, V., and Tsilou, E. (1993). Weill-Marchesani

syndrome. In GeneReviews, M.P. Adam, G.M. Mirzaa, R.A. Pagon, S.E.

Wallace, L.J.H. Bean, K.W. Gripp, and A. Amemiya, eds. (University of

Washington).

50. Yamagata, M., and Sanes, J.R. (2018). Expression and roles of the immu-

noglobulin superfamily recognition molecule Sidekick1 in mouse retina.

Front. Mol. Neurosci. 11, 485.

51. Foley, C.N., Staley, J.R., Breen, P.G., Sun, B.B., Kirk, P.D.W., Burgess,

S., and Howson, J.M.M. (2021). A fast and efficient colocalization algo-

rithm for identifying shared genetic risk factors across multiple traits.

Nat. Commun. 12, 764.

52. Brown, B.C., Asian Genetic Epidemiology Network Type 2Diabetes Con-

sortium, Ye, C.J., Price, A.L., and Zaitlen, N. (2016). Transethnic Genetic-

Correlation Estimates from Summary Statistics. Am. J. Hum. Genet.

99, 76–88.
478 Cell 187, 464–480, January 18, 2024
53. Lee, S.H., Wray, N.R., Goddard, M.E., and Visscher, P.M. (2011). Esti-

mating missing heritability for disease from genome-wide association

studies. Am. J. Hum. Genet. 88, 294–305.

54. Li, D., Hsu, S., Purushotham, D., Sears, R.L., and Wang, T. (2019).

WashU epigenome Browser update 2019. Nucleic Acids Res. 47,

W158–W165.

55. Boix, C.A., James, B.T., Park, Y.P., Meuleman, W., and Kellis, M. (2021).

Regulatory genomic circuitry of human disease loci by integrative epige-

nomics. Nature 590, 300–307.

56. Weissbrod, O., Hormozdiari, F., Benner, C., Cui, R., Ulirsch, J., Gazal, S.,

Schoech, A.P., van de Geijn, B., Reshef, Y., Márquez-Luna, C., et al.

(2020). Functionally informed fine-mapping and polygenic localization

of complex trait heritability. Nat. Genet. 52, 1355–1363.

57. Pressman, C.L., Chen, H., and Johnson, R.L. (2000). LMX1B, a LIM ho-

meodomain class transcription factor, is necessary for normal develop-

ment of multiple tissues in the anterior segment of the murine eye. Gen-

esis 26, 15–25.

58. Ham, J.H., Shin, S.J., Joo, K.R., Park, S.M., Sung, H.Y., Kim, J.S., Choi,

J.S., Choi, Y.J., Song, H.C., and Choi, E.J. (2009). A synonymous genetic

alteration of LMX1B in a family with nail-patella syndrome. Korean J.

Intern. Med. 24, 274–278.

59. Haro, E., Petit, F., Pira, C.U., Spady, C.D., Lucas-Toca, S., Yorozuya, L.I.,

Gray, A.L., Escande, F., Jourdain, A.S., Nguyen, A., et al. (2021). Identi-

fication of limb-specific Lmx1b auto-regulatory modules with nail-patella

syndrome pathogenicity. Nat. Commun. 12, 5533.

60. Mimiwati, Z., Mackey, D.A., Craig, J.E., Mackinnon, J.R., Rait, J.L., Lie-

belt, J.E., Ayala-Lugo, R., Vollrath, D., and Richards, J.E. (2006). Nail-pa-

tella syndrome and its association with glaucoma: a review of eight fam-

ilies. Br. J. Ophthalmol. 90, 1505–1509.

61. Yan, X., Lin, J., Wang, Y., Xuan, J., Yu, P., Guo, T., and Jin, F. (2019). A

novel small deletion of LMX1B in a large Chinese family with nail-patella

syndrome. BMC Med. Genet. 20, 71.

62. Vollrath, D., Jaramillo-Babb, V.L., Clough,M.V., McIntosh, I., Scott, K.M.,

Lichter, P.R., and Richards, J.E. (1998). Loss-of-functionmutations in the

LIM-homeodomain gene, LMX1B, in nail-patella syndrome. Hum. Mol.

Genet. 7, 1091–1098.

63. Lichter, P.R., Richards, J.E., Downs, C.A., Stringham, H.M., Boehnke,

M., and Farley, F.A. (1997). Cosegregation of open-angle glaucoma

and the nail-patella syndrome. Am. J. Ophthalmol. 124, 506–515.

64. Idaghdour, Y., Czika,W., Shianna, K.V., Lee, S.H., Visscher, P.M., Martin,

H.C., Miclaus, K., Jadallah, S.J., Goldstein, D.B., Wolfinger, R.D., et al.

(2010). Geographical genomics of human leukocyte gene expression

variation in southern Morocco. Nat. Genet. 42, 62–67.

65. Watanabe, K., Taskesen, E., van Bochoven, A., and Posthuma, D. (2017).

Functional mapping and annotation of genetic associations with FUMA.

Nat. Commun. 8, 1826.

66. de Leeuw, C.A., Mooij, J.M., Heskes, T., and Posthuma, D. (2015).

MAGMA: generalized gene-set analysis of GWAS data. PLoS Comput.

Biol. 11, e1004219.

67. Liberzon, A., Birger, C., Thorvaldsdóttir, H., Ghandi, M., Mesirov, J.P.,

and Tamayo, P. (2015). The Molecular Signatures Database (MSigDB)

hallmark gene set collection. Cell. Syst. 1, 417–425.

68. Agarwal, R., Gupta, S.K., Agarwal, P., Saxena, R., and Agrawal, S.S.

(2009). Current concepts in the pathophysiology of glaucoma. Indian J.

Ophthalmol. 57, 257–266.

69. Chavali, V.R.M., Haider, N., Rathi, S., Alapati, T., He, J., Gill, K., Nikonov,

S., Nikonov, R., Duong, T.T., McDougald, D.S., et al. (2020). Dual SMAD

inhibition and Wnt inhibition enhances the differentiation of induced

pluripotent stem cells into retinal ganglion cells. Sci. Rep. 10, 11828.

70. Wolf, J., Boneva, S., Schlecht, A., Lapp, T., Auw-Haedrich, C., Lagrèze,

W., Agostini, H., Reinhard, T., Schlunck, G., and Lange, C. (2022). The

Human Eye Transcriptome Atlas: a searchable comparative

http://refhub.elsevier.com/S0092-8674(23)01338-7/sref37
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref37
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref38
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref38
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref38
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref38
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref39
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref39
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref39
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref39
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref40
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref40
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref40
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref40
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref40
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref41
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref41
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref41
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref41
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref42
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref42
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref42
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref42
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref43
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref43
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref43
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref43
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref44
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref44
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref44
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref45
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref45
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref45
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref45
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref45
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref45
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref46
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref46
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref46
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref46
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref46
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref47
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref47
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref47
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref47
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref48
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref48
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref48
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref48
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref48
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref49
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref49
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref49
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref49
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref50
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref50
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref50
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref51
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref51
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref51
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref51
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref52
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref52
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref52
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref52
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref53
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref53
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref53
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref54
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref54
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref54
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref55
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref55
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref55
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref56
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref56
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref56
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref56
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref57
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref57
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref57
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref57
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref58
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref58
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref58
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref58
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref59
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref59
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref59
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref59
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref60
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref60
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref60
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref60
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref61
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref61
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref61
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref62
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref62
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref62
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref62
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref63
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref63
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref63
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref64
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref64
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref64
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref64
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref65
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref65
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref65
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref66
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref66
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref66
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref67
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref67
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref67
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref68
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref68
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref68
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref69
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref69
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref69
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref69
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref70
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref70
http://refhub.elsevier.com/S0092-8674(23)01338-7/sref70


ll
Article
transcriptome database for healthy and diseased human eye tissue. Ge-

nomics 114, 110286.

71. Wolf, J., Lapp, T., Reinhard, T., Agostini, H., Schlunck, G., and Lange, C.

(2022). Web-based gene expression analysis-paving the way to decode

healthy and diseased ocular tissue. Ophthalmologie 119, 929–936.

72. GTEx Consortium (2013). The Genotype-Tissue Expression (GTEx) proj-

ect. Nat. Genet. 45, 580–585.

73. ENCODE Project Consortium (2012). An integrated encyclopedia of DNA

elements in the human genome. Nature 489, 57–74.

74. Sirugo, G.,Williams, S.M., and Tishkoff, S.A. (2019). Themissing diversity

in human genetic studies. Cell 177, 26–31.

75. Rao, V.P., and Epstein, D.L. (2007). Rho GTPase/Rho kinase inhibition as

a novel target for the treatment of glaucoma. BioDrugs 21, 167–177.

76. Wang, J., Liu, X., and Zhong, Y. (2013). Rho/Rho-associated kinase

pathway in glaucoma (Review). Int. J. Oncol. 43, 1357–1367.

77. Tanna, A.P., and Johnson, M. (2018). Rho kinase inhibitors as a novel

treatment for glaucoma and ocular hypertension. Ophthalmology 125,

1741–1756.

78. Pattabiraman, P.P., Maddala, R., and Rao, P.V. (2014). Regulation of

plasticity and fibrogenic activity of trabecular meshwork cells by Rho

GTPase signaling. J. Cell. Physiol. 229, 927–942.

79. Fukata, Y., Amano, M., and Kaibuchi, K. (2001). Rho-Rho-kinase

pathway in smooth muscle contraction and cytoskeletal reorganization

of non-muscle cells. Trends Pharmacol. Sci. 22, 32–39.

80. Honjo, M., Tanihara, H., Inatani, M., Kido, N., Sawamura, T., Yue, B.Y.,

Narumiya, S., and Honda, Y. (2001). Effects of rho-associated protein ki-

nase inhibitor Y-27632 on intraocular pressure and outflow facility.

Invest. Ophthalmol. Vis. Sci. 42, 137–144.

81. Kumar, J., and Epstein, D.L. (2011). Rho GTPase-mediated cytoskeletal

organization in Schlemm’s canal cells play a critical role in the regulation

of aqueous humor outflow facility. J. Cell. Biochem. 112, 600–606.

82. Lessey-Morillon, E.C., Osborne, L.D., Monaghan-Benson, E., Guilluy, C.,

O’Brien, E.T., Superfine, R., and Burridge, K. (2014). The RhoA guanine

nucleotide exchange factor, LARG, mediates ICAM-1-dependent me-

chanotransduction in endothelial cells to stimulate transendothelial

migration. J. Immunol. 192, 3390–3398.

83. Okuhira, K., Fitzgerald, M.L., Tamehiro, N., Ohoka, N., Suzuki, K., Sa-

wada, J., Naito, M., and Nishimaki-Mogami, T. (2010). Binding of PDZ-

RhoGEF to ATP-binding cassette transporter A1 (ABCA1) induces

cholesterol efflux through RhoA activation and prevention of transporter

degradation. J. Biol. Chem. 285, 16369–16377.

84. Trivli, A., Zervou, M.I., Goulielmos, G.N., Spandidos, D.A., and Detorakis,

E.T. (2020). Primary open angle glaucoma genetics: the common variants

and their clinical associations (Review). Mol. Med. Rep. 22, 1103–1110.

85. Iglesias, A.I., Springelkamp, H., Ramdas, W.D., Klaver, C.C.W., Willem-

sen, R., and van Duijn, C.M. (2015). Genes, pathways, and animal models

in primary open-angle glaucoma. Eye (Lond.) 29, 1285–1298.

86. Fuchshofer, R., and Tamm, E.R. (2012). The role of TGF-b in the patho-

genesis of primary open-angle glaucoma. Cell Tissue Res. 347, 279–290.

87. Wordinger, R.J., Sharma, T., and Clark, A.F. (2014). The role of TGF-b2

and bone morphogenetic proteins in the trabecular meshwork and glau-

coma. J. Ocul. Pharmacol. Ther. 30, 154–162.

88. Fl