Scholarly article on topic 'Transcriptomic analysis across nasal, temporal, and macular regions of human neural retina and RPE/choroid by RNA-Seq'

Transcriptomic analysis across nasal, temporal, and macular regions of human neural retina and RPE/choroid by RNA-Seq Academic research paper on "Clinical medicine"

CC BY-NC-SA
0
0
Share paper
Academic journal
Experimental Eye Research
OECD Field of science
Keywords
{RNA-Seq / Retina / Choroid / RPE / "Gene expression" / "Retinitis pigmentosa" / "Macular degeneration" / Transcriptome}

Abstract of research paper on Clinical medicine, author of scientific article — S. Scott Whitmore, Alex H. Wagner, Adam P. DeLuca, Arlene V. Drack, Edwin M. Stone, et al.

Abstract Proper spatial differentiation of retinal cell types is necessary for normal human vision. Many retinal diseases, such as Best disease and male germ cell associated kinase (MAK)-associated retinitis pigmentosa, preferentially affect distinct topographic regions of the retina. While much is known about the distribution of cell types in the retina, the distribution of molecular components across the posterior pole of the eye has not been well-studied. To investigate regional difference in molecular composition of ocular tissues, we assessed differential gene expression across the temporal, macular, and nasal retina and retinal pigment epithelium (RPE)/choroid of human eyes using RNA-Seq. RNA from temporal, macular, and nasal retina and RPE/choroid from four human donor eyes was extracted, poly-A selected, fragmented, and sequenced as 100 bp read pairs. Digital read files were mapped to the human genome and analyzed for differential expression using the Tuxedo software suite. Retina and RPE/choroid samples were clearly distinguishable at the transcriptome level. Numerous transcription factors were differentially expressed between regions of the retina and RPE/choroid. Photoreceptor-specific genes were enriched in the peripheral samples, while ganglion cell and amacrine cell genes were enriched in the macula. Within the RPE/choroid, RPE-specific genes were upregulated at the periphery while endothelium associated genes were upregulated in the macula. Consistent with previous studies, BEST1 expression was lower in macular than extramacular regions. The MAK gene was expressed at lower levels in macula than in extramacular regions, but did not exhibit a significant difference between nasal and temporal retina. The regional molecular distinction is greatest between macula and periphery and decreases between different peripheral regions within a tissue. Datasets such as these can be used to prioritize candidate genes for possible involvement in retinal diseases with regional phenotypes.

Academic research paper on topic "Transcriptomic analysis across nasal, temporal, and macular regions of human neural retina and RPE/choroid by RNA-Seq"

ELSEVIER

Contents lists available at ScienceDirect

Experimental Eye Research

journal homepage: www.elsevier.com/locate/yexer

EXPERIMENTAL EVE RESEARCH

Transcriptomic analysis across nasal, temporal, and macular regions of human neural retina and RPE/choroid by RNA-Seq

S. Scott Whitmore a'b, Alex H. Wagner a'c, Adam P. DeLuca a'b, Arlene V. Drack a'b, Edwin M. Stone a'b, Budd A. Tucker a'b, Shemin Zeng a'b, Terry A. Braun a'b'c, Robert F. Mullins a'b, Todd E. Scheetz a'b'c' *

a Stephen A. Wynn Institute for Vision Research, The University of Iowa, Iowa City, IA, USA

b Department of Ophthalmology and Visual Sciences, Carver College of Medicine, The University of Iowa, Iowa City, IA, USA c Department of Biomedical Engineering, College of Engineering, The University of Iowa, Iowa City, IA, USA

CrossMark

ARTICLE INFO

Article history:

Received 14 September 2014 Received in revised form 31 October 2014

Accepted in revised form 4 November 2014 Available online 5 November 2014

Keywords:

RNA-Seq

Retina

Choroid

Gene expression Retinitis pigmentosa Macular degeneration Transcriptome

ABSTRACT

Proper spatial differentiation of retinal cell types is necessary for normal human vision. Many retinal diseases, such as Best disease and male germ cell associated kinase (MAK)-associated retinitis pigmentosa, preferentially affect distinct topographic regions of the retina. While much is known about the distribution of cell types in the retina, the distribution of molecular components across the posterior pole of the eye has not been well-studied. To investigate regional difference in molecular composition of ocular tissues, we assessed differential gene expression across the temporal, macular, and nasal retina and retinal pigment epithelium (RPE)/choroid of human eyes using RNA-Seq. RNA from temporal, macular, and nasal retina and RPE/choroid from four human donor eyes was extracted, poly-A selected, fragmented, and sequenced as 100 bp read pairs. Digital read files were mapped to the human genome and analyzed for differential expression using the Tuxedo software suite. Retina and RPE/choroid samples were clearly distinguishable at the transcriptome level. Numerous transcription factors were differentially expressed between regions of the retina and RPE/choroid. Photoreceptor-specific genes were enriched in the peripheral samples, while ganglion cell and amacrine cell genes were enriched in the macula. Within the RPE/choroid, RPE-specific genes were upregulated at the periphery while endothe-lium associated genes were upregulated in the macula. Consistent with previous studies, BEST1 expression was lower in macular than extramacular regions. The MAK gene was expressed at lower levels in macula than in extramacular regions, but did not exhibit a significant difference between nasal and temporal retina. The regional molecular distinction is greatest between macula and periphery and decreases between different peripheral regions within a tissue. Datasets such as these can be used to prioritize candidate genes for possible involvement in retinal diseases with regional phenotypes.

© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-SA

license (http://creativecommons.org/licenses/by-nc-sa/3.0/).

1. Introduction

Human visual activities balance the need for high visual acuity, such as reading, driving, and recognizing faces, with comprehensive peripheral vision. High visual acuity is enabled by the macula, a

* Corresponding author. 375 Newton Road, 3181 MERF, The University of Iowa, Iowa City, IA 52242, USA.

E-mail addresses: steven-whitmore@uiowa.edu (S.S. Whitmore), alex-wagner@ uiowa.edu (A.H. Wagner), adam-deluca@uiowa.edu (A.P. DeLuca), arlene-drack@ uiowa.edu (A.V. Drack), edwin-stone@uiowa.edu (E.M. Stone), budd-tucker@ uiowa.edu (B.A. Tucker), shemin-zeng@uiowa.edu (S. Zeng), terry-braun@uiowa. edu (T.A. Braun), robert-mullins@uiowa.edu (R.F. Mullins), todd-scheetz@uiowa. edu (T.E. Scheetz).

region at the center of the posterior retina with the highest density of cone photoreceptors and ganglion cells. Vision in dim light is made possible by the rod photoreceptor cells which are most concentrated just anterior to the macula.

Many ocular pathologies affect distinct regions of the retina. Macular degenerations, such as age-related macular degeneration, Best disease, Stargardt disease, and North Carolina macular dystrophy, result in loss of photoreceptors and RPE cells in the macula. Other diseases, such as retinitis pigmentosa (RP), cause selective loss of photoreceptors at the periphery of the retina, often sparing central vision until advanced stages of disease. Some diseases manifest further distinction in regional phenotypes. For instance, male germ cell associated kinase (MAK)-related RP, the most common cause of inherited blindness among the Ashkenazi Jewish

http://dx.doi.org/10.1016/j.exer.2014.11.001

0014-4835/© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-SA license (http://creativecommons.org/licenses/by-nc-sa/3.0/).

Table 1

Donor information. Samples marked by asterisks (*

were analyzed for differential expression in Cuffdiff. Library prep involved poly-A selection.

Donor Eye Sex Age Cause of death Death-to-preservation (h) Retina RPE/choroid Library prep.

1* OS F 82 Respiratory failure <4 M, N, T M, N, T stranded

2* OD M 77 Congestive heart failure <5 M, N, T M, N, T stranded

3* OD F 89 Pneumonia 3—4.5 M, N, T M, N, T stranded

4* OS M 91 Pneumonia 5—6 M, N, T M, N, T stranded

5 OD F 91 Unknown <6 M, N, T, S, I — unstranded

Abbreviations: M- macula; N — nasal; T — temporal; S — superior; I — inferior.

tissue & region

macular retina nasal retina A temporal retina D macular RPE/chorold □ nasal RPE/choroid A temporal RPE/choroid

0.0 ■▲■▲J MAJ MAÜÜ Ad AäAÜ Öd ÖA

Fig. 1. Hierarchical clustering on expressed isoforms based on Spearman's correlation.

population, manifests an unusual inferonasal predilection for the photoreceptor cell death (Stone et al., 2011; Tucker et al., 2011).

Classic studies on the distribution of cell types through the neural retina (Curcio and Allen, 1990; Curcio et al., 1990; Jonas et al., 1992), combined with extensive disease phenotyping and detailed analyses of single molecule expression have provided substantial insight into the structural and functional organization of the neural retina, retinal pigment epithelium (RPE), and choroid in different regions of the eye. Previous large-scale studies aimed at differentiating macula from periphery have utilized reverse transcriptase-polymerase chain reaction (RT-PCR) (Kociok and Joussen, 2007), array-based technologies (Ishibashi et al., 2004; Diehn et al., 2005; Bowes Rickman et al., 2006; Radeke et al., 2007; van Soest et al., 2007) and more recently, RNA sequencing (RNA-Seq) (Li et al., 2014). Unlike the former two technologies, RNA-Seq provides not only estimates of gene expression level and isoform abundance, but also captures sequence-level information, potentially uncovering novel exons and other transcriptional events.

3 4 0 1

log10(FPKM+1)

Whitmore et al.

Disease

Glaucoma High myopia Myopia

Refractive error

Fig. 2. Comparison of GWAS associated gene sets with data published by Li et al. (2014). The previously published dataset included sclera in the RPE/choroid punches, likely accounting for the number of genes in RPE/choroid below the diagonal. Spearman rank correlation coefficient shown (r).

Table 2

Top 40 differentially expressed genes between nasal retina vs. macular retina with q-value < 0.001 and absolute log2(fold-change) > 1.

Symbol Description Chromosome Start Stop Macular Nasal Log2(FC) P-value Q-value

retina (FPKM) retina (FPKM)

HSD17B2 Hydroxysteroid (17-beta) chr16 82,056,731 82,132,139 1.29 9.55 2.88 5.00E-05 2.86E-04

dehydrogenase 2

CYP26B1 Cytochrome P450, family 26, subfamily chr2 72,356,363 72,375,687 0.41 2.77 2.75 5.00E-05 2.86E-04

B, polypeptide 1

COL2A1 Collagen, type II, alpha 1 chr12 48,366,747 48,398,285 1.56 9.98 2.68 5.00E-05 2.86E-04

STEAP4 STEAP family member 4 chr7 87,900,209 87,936,319 0.31 1.89 2.58 5.00E-05 2.86E-04

FOX13 Forkhead box I3 chr2 88,744,347 88,752,197 0.49 2.81 2.51 5.00E-05 2.86E-04

OXTR Oxytocin receptor chr3 8,792,094 8,811,460 0.49 1.91 1.96 5.00E-05 2.86E-04

NPVF Neuropeptide VF precursor chr7 25,264,190 25,268,105 29.12 106.31 1.87 5.00E-05 2.86E-04

Z1C1 Zic family member 1 chr3 147,103,834 147,134,784 1.24 4.25 1.78 5.00E-05 2.86E-04

PRL Prolactin chr6 22,287,472 22,303,082 4.48 15.15 1.76 5.00E-05 2.86E-04

LOXL4 Lysyl oxidase-like 4 chr10 100,007,442 100,028,007 0.98 3.20 1.71 1.00E-04 5.49E-04

FXYD2a FXYD domain containing ion transport chr11 117,690,789 117,748,201 12.05 35.28 1.55 5.00E-05 2.86E-04

KRT39b regulator 2

Keratin 39 chr17 39,113,368 39,143,387 3.43 10.02 1.55 5.00E-05 2.86E-04

1D3 Inhibitor of DNA binding 3, dominant chr1 23,884,420 23,886,285 17.01 48.24 1.50 5.00E-05 2.86E-04

negative helix-loop-helix protein

EGR1 Early growth response 1 chr5 137,801,180 137,805,004 15.34 39.41 1.36 5.00E-05 2.86E-04

HTR1F 5-hydroxytryptamine (serotonin) chr3 87,841,874 88,049,226 0.78 2.00 1.35 5.00E-05 2.86E-04

receptor 1F, G protein-coupled

CSRP2 Cysteine and glycine-rich protein 2 chr12 77,252,493 77,272,799 19.34 48.16 1.32 5.00E-05 2.86E-04

CYR61 Cysteine-rich, angiogenic inducer, 61 chr1 86,046,443 86,049,648 11.76 29.12 1.31 5.00E-05 2.86E-04

LGALS3 Lectin, galactoside-binding, soluble, 3 chr14 55,595,934 55,612,148 16.89 41.31 1.29 5.00E-05 2.86E-04

C4Ac Complement component 4A (Rodgers chr6 31,982,571 32,003,195 1.25 3.02 1.27 1.00E-04 5.49E-04

blood group)

NR4A1 Nuclear receptor subfamily 4, group A, chr12 52,416,615 52,453,292 15.05 36.28 1.27 1.50E-04 8.01E-04

member 1

KCNC2 Potassium voltage-gated channel, chr12 75,433,857 75,603,528 57.75 17.88 -1.69 5.00E-05 2.86E-04

Shaw-related subfamily, member 2

Potentially novel chr6 113,745,551 113,755,041 13.32 4.05 -1.72 5.00E-05 2.86E-04

TRPM2 Transient receptor potential cation chr21 45,770,087 45,869,277 8.60 2.60 -1.72 5.00E-05 2.86E-04

channel, subfamily M, member 2

R1T2 Ras-like without CAAX 2 chr18 40,323,182 40,695,841 13.61 4.11 -1.73 5.00E-05 2.86E-04

1SLR2 Immunoglobulin superfamily chr15 74,418,236 74,429,152 15.76 4.72 -1.74 5.00E-05 2.86E-04

containing leucine-rich repeat 2

AHNAK2 AHNAK nucleoprotein 2 chr14 105,371,613 105,444,694 84.49 25.19 -1.75 1.00E-04 5.49E-04

1RX2 Iroquois homeobox 2 chr5 2,745,916 2,751,769 14.95 4.34 -1.78 5.00E-05 2.86E-04

CPNE9 Copine family member IX chr3 9,745,509 9,771,592 5.89 1.70 -1.79 5.00E-05 2.86E-04

TTC39A Tetratricopeptide repeat domain 39A chr1 51,752,172 51,810,785 22.16 6.24 -1.83 5.00E-05 2.86E-04

HTR1B 5-hydroxytryptamine (serotonin) chr6 78,168,875 78,174,100 3.40 0.95 -1.84 1.50E-04 8.01E-04

receptor 1B, G protein-coupled

PHOSPHO1 Phosphatase, orphan 1 chr17 47,300,731 47,308,128 6.47 1.81 -1.84 5.00E-05 2.86E-04

FABP3 Fatty acid binding protein 3, muscle and chr1 31,769,837 31,846,528 131.10 35.95 -1.87 5.00E-05 2.86E-04

heart (mammary-derived growth

inhibitor)

PTH1R Parathyroid hormone 1 receptor chr3 46,761,072 46,945,351 11.74 3.18 -1.88 1.00E-04 5.49E-04

LCP1 Lymphocyte cytosolic protein 1 (L- chr13 46,700,056 46,785,977 7.48 1.94 -1.95 5.00E-05 2.86E-04

plastin)

SLN Sarcolipin chr11 107,578,100 107,582,787 42.78 10.92 -1.97 5.00E-05 2.86E-04

MCTP1 Multiple C2 domains, transmembrane 1 chr5 94,036,882 94,620,279 3.33 0.84 -2.00 5.00E-05 2.86E-04

MRGPRE MAS-related GPR, member E chr11 3,246,156 3,254,111 12.08 2.58 -2.22 5.00E-05 2.86E-04

RASGRP3 RAS guanyl releasing protein 3 (calcium chr2 33,630,960 33,792,285 5.84 1.09 -2.42 5.00E-05 2.86E-04

and DAG-regulated)

C1orf170 Chromosome 1 open reading frame 170 chr1 901,876 917,673 8.05 1.37 -2.55 5.00E-05 2.86E-04

PRPH Peripherin chr12 49,686,760 49,692,481 55.03 8.39 -2.71 5.00E-05 2.86E-04

a FXYD2, FXYD6, and FXYD6-FXYD2 (readthrough) were reported as a single transcriptional locus by Cufflinks. b KRT39 and KRT40 were reported as a single transcriptional locus by Cufflinks. c C4A and C4B_2 were reported as a single transcriptional locus by Cufflinks.

Here we present the first RNA-Seq dataset to investigate the temporal, macular, and nasal regions of the retina and RPE/choroid along with an additional sample with superior and inferior regions represented. Distinct expression profiles clearly differentiated retina from RPE/choroid. Differences between periphery and macula were more pronounced in the retina than in the RPE/choroid and in both tissues reflected dominant cell type distributions. In the neural retina, nasal and temporal regions were indistinguishable by our criteria.

2. Materials and methods

2.1. Tissue acquisition

Human eyes were obtained through the Iowa Lions Eye Bank after informed consent by family members and in accordance with the tenets of the Declaration of Helsinki. Donor information is presented in Table 1. In four donors, an 8 mm trephine punch was used to remove the macula and a 6 mm punch was used to remove

Table 3

Top 40 differentially expressed genes between temporal retina vs. macular retina with q-value < 0.001 and absolute log2(fold-change) > 1.

Symbol Description Chromosome Start Stop Macular retina (FPKM) Temporal retina (FPKM) Log2(FC) P-value Q-value

MIR1247 MicroRNA 1247 chr14 102,023,301 102,026,828 0.30 4.16 3.77 5.00E-05 2.86E-04

CYP26B1 Cytochrome P450, family 26, subfamily B, chr2 72,356,363 72,375,687 0.41 4.68 3.51 5.00E-05 2.86E-04

polypeptide 1

HSD17B2 Hydroxysteroid (17-beta) dehydrogenase 2 chr16 82,056,731 82,132,139 1.29 10.96 3.08 5.00E-05 2.86E-04

PRL Prolactin chr6 22,287,472 22,303,082 4.48 29.37 2.71 5.00E-05 2.86E-04

COL2A1 Collagen, type II, alpha 1 chr12 48,366,747 48,398,285 1.56 9.42 2.59 5.00E-05 2.86E-04

ZIC1 Zic family member 1 chr3 147,103,834 147,134,784 1.24 6.24 2.33 5.00E-05 2.86E-04

NDUFA4L2 NADH dehydrogenase (ubiquinone) 1 alpha chr12 57,623,324 57,634,949 42.98 162.00 1.91 1.00E-04 5.49E-04

subcomplex, 4-like 2

FAM46B Family with sequence similarity 46, member B chr1 27,331,510 27,339,333 0.96 3.61 1.91 5.00E-05 2.86E-04

LOXL4 Lysyl oxidase-like 4 chr10 100,007,442 100,028,007 0.98 3.65 1.90 1.00E-04 5.49E-04

FXYD2a FXYD domain containing ion transport regulator 2 chr11 117,690,789 117,748,201 12.05 43.78 1.86 5.00E-05 2.86E-04

ATP1A2 ATPase, Na+/K+ transporting, alpha 2 polypeptide chr1 160,085,519 160,113,374 17.41 57.10 1.71 5.00E-05 2.86E-04

SIX2 SIX homeobox 2 chr2 45,232,323 45,236,563 2.44 7.92 1.70 5.00E-05 2.86E-04

GPR124 G protein-coupled receptor 124 chr8 37,654,400 37,707,431 4.99 15.61 1.65 5.00E-05 2.86E-04

ADSSL1 Adenylosuccinate synthase like 1 chr14 105,190,533 105,213,663 5.25 15.63 1.57 5.00E-05 2.86E-04

HTR1F 5-hydroxytryptamine (serotonin) receptor 1F, G chr3 87,841,874 88,049,226 0.78 2.31 1.56 5.00E-05 2.86E-04

protein-coupled

ATOH8 Atonal homolog 8 (Drosophila) chr2 85,978,936 86,018,506 1.13 3.20 1.50 5.00E-05 2.86E-04

NPVF Neuropeptide VF precursor chr7 25,264,190 25,268,105 29.12 81.63 1.49 5.00E-05 2.86E-04

PPP1R3C Protein phosphatase 1, regulatory subunit 3C chr10 93,379,047 93,392,858 14.02 38.45 1.46 5.00E-05 2.86E-04

CLEC4F C-type lectin domain family 4, member F chr2 71,035,775 71,047,732 2.07 5.67 1.45 1.00E-04 5.49E-04

CD4 CD4 molecule chr12 6,898,637 6,929,977 5.55 14.99 1.43 5.00E-05 2.86E-04

POU4F2 POU class 4 homeobox 2 chr4 147,560,044 147,563,623 11.13 1.98 -2.49 5.00E-05 2.86E-04

NEFM Neurofilament, medium polypeptide chr8 24,771,273 24,776,612 544.29 94.76 -2.52 5.00E-05 2.86E-04

SLC17A6 Solute carrier family 17 (vesicular glutamate chr11 22,355,132 22,401,046 43.08 7.44 -2.53 5.00E-05 2.86E-04

transporter), member 6

TTC39A Tetratricopeptide repeat domain 39A chr1 51,752,172 51,810,785 22.16 3.76 -2.56 5.00E-05 2.86E-04

ISLR2 Immunoglobulin superfamily containing leucine- chr15 74,418,236 74,429,152 15.76 2.61 -2.59 5.00E-05 2.86E-04

rich repeat 2

POU4F1 POU class 4 homeobox 1 chr13 78,628,989 79,233,323 23.65 3.84 -2.62 5.00E-05 2.86E-04

AHNAK2 AHNAK nucleoprotein 2 chr14 105,371,613 105,444,694 84.49 13.65 -2.63 5.00E-05 2.86E-04

IRX1 Iroquois homeobox 1 chr5 3,596,054 3,601,517 5.55 0.88 -2.66 5.00E-05 2.86E-04

IRX2 Iroquois homeobox 2 chr5 2,745,916 2,751,769 14.95 2.34 -2.67 5.00E-05 2.86E-04

HTR1B 5-hydroxytryptamine (serotonin) receptor 1B, G chr6 78,168,875 78,174,100 3.40 0.51 -2.74 5.00E-05 2.86E-04

protein-coupled

MCTP1 Multiple C2 domains, transmembrane 1 chr5 94,036,882 94,620,279 3.33 0.50 -2.75 5.00E-05 2.86E-04

Potentially novel chr6 113,745,551 113,755,041 13.32 1.90 -2.81 5.00E-05 2.86E-04

CPNE9 Copine family member IX chr3 9,745,509 9,771,592 5.89 0.83 -2.83 5.00E-05 2.86E-04

MRGPRE MAS-related GPR, member E chr11 3,246,156 3,254,111 12.08 1.55 -2.96 5.00E-05 2.86E-04

ABO ABO blood group (transferase A, alpha 1-3-N- chr9 136,099,931 136,151,454 3.68 0.44 -3.07 5.00E-05 2.86E-04

acetylgalactosaminyltransferase; transferase B,

alpha 1-3-galactosyltransferase)

Potentially novel chr12 108,203,132 108,258,620 3.78 0.45 -3.08 5.00E-05 2.86E-04

RASGRP3 RAS guanyl releasing protein 3 (calcium and DAG- chr2 33,630,960 33,792,285 5.84 0.62 -3.24 5.00E-05 2.86E-04

regulated)

C3orf55 Chromosome 3 open reading frame 55 chr3 157,260,745 157,395,552 6.11 0.60 -3.34 5.00E-05 2.86E-04

PRPH Peripherin chr12 49,686,760 49,692,481 55.03 5.00 -3.46 5.00E-05 2.86E-04

C1orf170 Chromosome 1 open reading frame 170 chr1 901,876 917,673 8.05 0.66 -3.60 5.00E-05 2.86E-04

a FXYD2, FXYD6, and FXYD6-FXYD2 (readthrough) were reported as a single transcriptional locus by Cufflinks.

temporal and nasal punches. In one additional sample, a 4 mm punch was used to dissect macular, nasal, temporal, superior and inferior punches (Braun et al., 2013). Neural retina and RPE/choroid were separated, flash frozen in liquid nitrogen, and stored at -80 ° C. All samples used in this study were preserved in liquid nitrogen within 6 h of death.

2.2. RNA sequencing

RNA was extracted from frozen tissue punches using a Qiagen RNeasy Mini Prep Kit (Qiagen, Valencio, CA) and kept frozen prior to sequencing. The four temporal-macula-nasal sets were prepared for paired-end sequencing using an Illumina TruSeq Stranded mRNA Sample Prep Kit and sequenced on the Illumina platform in the Genomics Division of the Iowa Institute of Human Genetics.

Additional paired-end sequencing of one temporal-macula-nasal-superior-inferior sample was performed on the Illumina platform at HudsonAlpha Institute for Biotechnology (Huntsville, AL).

2.3. Bioinformatic analysis

Sequenced reads were mapped to human genome build hg19 using TopHat2 (ver. 2.0.11; (Kim et al., 2013)), transcript structure and abundance were estimated using Cufflinks (ver. 2.1.1; (Trapnell et al., 2010)), and differential expression analysis was performed using Cuffdiff (ver. 2.1.1; (Trapnell et al., 2013)). Quality control analysis was performed using RNA-SeQC (ver. 1.1.7; (DeLuca et al., 2012)). The cummeRbund package (ver. 2.4.1; (Trapnell et al., 2012)) for R (ver. 3.0.2) was used for data visualization. Differential expression analysis was performed for the four donor samples

Table 4

Top 40 differentially expressed genes between nasal RPE/choroid vs. macular RPE/choroid with q-value < 0.001 and absolute log2(fold-change) > 1.

Symbol Description Chromosome Start Stop Macular RPE/ choroid (FPKM) Nasal RPE/ choroid (FPKM) Log2(FC) P-value Q-value

SCG5 Secretogranin V (7B2 protein) chr15 32,933,869 32,989,298 8.80 117.88 3.74 5.00E-05 2.86E-04

TFP12 Tissue factor pathway inhibitor 2 chr7 93,514,708 93,545,731 1.87 23.83 3.67 5.00E-05 2.86E-04

FH1T Fragile histidine triad chr3 59,733,016 61,237,133 14.79 115.57 2.97 5.00E-05 2.86E-04

RHO Rhodopsin chr3 129,247,481 129,254,187 9.49 73.90 2.96 5.00E-05 2.86E-04

MT1G Metallothionein 1G chr16 56,700,645 56,701,980 61.09 468.12 2.94 5.00E-05 2.86E-04

GNAT1 Guanine nucleotide binding protein (G chr3 50,229,042 50,235,129 2.34 16.38 2.81 5.00E-05 2.86E-04

protein), alpha transducing activity

polypeptide 1

A1PL1 Aryl hydrocarbon receptor interacting chr17 6,327,058 6,338,519 2.44 16.67 2.77 5.00E-05 2.86E-04

protein-like 1

TF Transferrin chr3 133,426,380 133,497,850 6.84 46.00 2.75 5.00E-05 2.86E-04

Potentially chr2 106,544,607 106,554,533 4.52 28.58 2.66 5.00E-05 2.86E-04

PRSS33 Protease, serine, 33 chr16 2,833,938 2,837,578 1.83 11.43 2.64 5.00E-05 2.86E-04

COL9A2 Collagen, type 1X, alpha 2 chr1 40,766,086 40,782,939 3.20 19.90 2.64 5.00E-05 2.86E-04

ATP10Ba ATPase, class V, type 10B chr5 159,990,126 160,365,633 31.64 191.17 2.60 5.00E-05 2.86E-04

RCVRN Recoverin chr17 9,801,026 9,808,684 5.57 31.88 2.52 5.00E-05 2.86E-04

SAG S-antigen; retina and pineal gland chr2 234,210,734 234,255,703 9.56 53.93 2.50 5.00E-05 2.86E-04

(arrestin)

KCNV2 Potassium channel, subfamily V, chr9 2,717,525 2,730,037 1.88 10.23 2.44 5.00E-05 2.86E-04

member 2

SLC4A10 Solute carrier family 4, sodium chr2 162,480,844 162,841,786 1.33 6.97 2.39 5.00E-05 2.86E-04

bicarbonate transporter, member 10

ALDH1A3 Aldehyde dehydrogenase 1 family, chr15 101,420,008 101,456,897 18.69 96.70 2.37 5.00E-05 2.86E-04

member A3

ECEL1 Endothelin converting enzyme-like 1 chr2 233,344,536 233,362,207 1.43 6.55 2.20 5.00E-05 2.86E-04

SFRP1 Secreted frizzled-related protein 1 chr8 41,119,475 41,166,990 13.15 60.45 2.20 5.00E-05 2.86E-04

AKR7A3b Aldo-keto reductase family 7, member chr1 19,592,475 19,616,124 0.98 4.06 2.06 1.50E-04 8.01E-04

A3 (aflatoxin aldehyde reductase)

NRN1 Neuritin 1 chr6 5,998,227 6,008,078 11.86 4.36 -1.44 5.00E-05 2.86E-04

PR1MA1 Proline rich membrane anchor 1 chr14 94,184,643 94,255,079 11.83 4.13 -1.52 5.00E-05 2.86E-04

CXCL14 Chemokine (C-X-C motif) ligand 14 chr5 134,906,370 134,914,969 21.66 7.49 -1.53 5.00E-05 2.86E-04

SCN7A Sodium channel, voltage-gated, type chr2 167,260,082 167,350,946 9.90 3.33 -1.57 5.00E-05 2.86E-04

V11, alpha subunit

C1QL1 Complement component 1, q chr17 43,036,339 43,046,605 11.94 3.90 -1.62 1.50E-04 8.01E-04

subcomponent-like 1

TRDN Triadin chr6 123,537,304 123,958,601 24.71 7.99 -1.63 5.00E-05 2.86E-04

PKHD1L1 Polycystic kidney and hepatic disease 1 chr8 110,374,702 110,549,447 6.81 2.19 -1.64 5.00E-05 2.86E-04

(autosomal recessive)-like 1

GABRE Gamma-aminobutyric acid (GABA) A chrX 151,121,595 151,143,206 6.17 1.98 -1.64 5.00E-05 2.86E-04

receptor, epsilon

RGS7BP Regulator of G-protein signaling 7 chr5 63,801,626 63,910,486 6.04 1.92 -1.65 5.00E-05 2.86E-04

binding protein

EPHA3 EPH receptor A3 chr3 89,156,673 89,531,390 11.95 3.72 -1.69 5.00E-05 2.86E-04

CRHBP Corticotropin releasing hormone chr5 76,248,679 76,276,815 6.58 1.97 -1.74 5.00E-05 2.86E-04

binding protein

CCL14c Chemokine (C-C motif) ligand 14 chr17 34,310,691 34,329,084 208.78 62.35 -1.74 5.00E-05 2.86E-04

CPAMD8 C3 and PZP-like, alpha-2-macroglobulin chr19 17,003,760 17,137,625 12.55 3.69 -1.77 5.00E-05 2.86E-04

domain containing 8

AQP7P1 Aquaporin 7 pseudogene 1 chr9 67,269,284 67,289,625 14.91 4.37 -1.77 5.00E-05 2.86E-04

L0C100507387 chr5 175,542,798 175,554,408 11.01 2.77 -1.99 5.00E-05 2.86E-04

POSTN Periostin, osteoblast specific factor chr13 38,136,718 38,172,981 5.91 1.45 -2.02 5.00E-05 2.86E-04

NRXN1 Neurexin 1 chr2 50,145,624 51,259,674 4.18 1.00 -2.06 5.00E-05 2.86E-04

SULT1E1 Sulfotransferase family 1E, estrogen- chr4 70,706,929 70,725,870 8.30 1.82 -2.19 5.00E-05 2.86E-04

preferring, member 1

WFDC1 WAP four-disulfide core domain 1 chr16 84,303,639 84,363,457 154.28 33.71 -2.19 5.00E-05 2.86E-04

TBX15 T-box 15 chr1 119,425,665 119,543,988 5.11 1.08 -2.24 5.00E-05 2.86E-04

a ATP10B and LOC285629 were reported as a single transcriptional locus by Cufflinks.

b AKR7A3 and AKR7L were reported as a single transcriptional locus by Cufflinks.

c CCL14, CCL15, CCL15-CCL14 (readthrough) were reported as a single transcriptional locus by Cufflinks.

with similar processing (macula, temporal, and nasal retina and RPE/choroid). Spearman's rank correlation coefficient as implemented in R was used for all correlations. Genes were deemed differentially expressed if the absolute value of the log2 (fold change) was >1, the q-value < 0.001, and the FPKM (Fragments Per Kilobase of transcript per Million mapped reads) values of either group compared were all > 1.

3. Results

We performed 100 bp paired-end RNA-Seq on retina and RPE/ choroid from the temporal, macular, and nasal regions of four clinically normal human donor eyes (Table 1). Hierarchical clustering of samples showed clear separation of neural retina from RPE/choroid samples and evidence of donor effect, i.e., samples derived from the same donor tend to cluster together more often than not (Fig. 1) (Rouhani et al., 2014).

98 S.S. Whitmore et al. / Experimental Eye Research 129 (2014) 93-106

Table 5

Top 40 differentially expressed genes between temporal RPE/choroid vs. macular RPE/choroid with q-value < 0.001 and absolute log2(fold-change) > 1.

Symbol Description Chromosome Start Stop Macular RPE/choroid (FPKM) Temporal RPE/choroid (FPKM) Log2(FC) P-value Q-value

TFPI2 Tissue factor pathway inhibitor 2 chr7 93,514,708 93,545,731 1.87 82.08 5.45 5.00E-05 2.86E-04

SCG5 Secretogranin V (7B2 protein) chr15 32,933,869 32,989,298 8.80 118.78 3.75 5.00E-05 2.86E-04

FHIT Fragile histidine triad chr3 59,733,016 61,237,133 14.79 194.50 3.72 5.00E-05 2.86E-04

COL9A2 Collagen, type IX, alpha 2 chr1 40,766,086 40,782,939 3.20 25.78 3.01 5.00E-05 2.86E-04

ALDH1A3 Aldehyde dehydrogenase 1 family, member A3 chr15 101,420,008 101,456,897 18.69 139.08 2.90 5.00E-05 2.86E-04

SFRP1 Secreted frizzled-related protein 1 chr8 41,119,475 41,166,990 13.15 91.25 2.79 5.00E-05 2.86E-04

Potentially novel chr2 106,544,607 106,554,533 4.52 28.78 2.67 5.00E-05 2.86E-04

ATP10Ba ATPase, class V, type 10B chr5 159,990,126 160,365,633 31.64 181.87 2.52 5.00E-05 2.86E-04

PRSS33 Protease, serine, 33 chr16 2,833,938 2,837,578 1.83 9.95 2.44 5.00E-05 2.86E-04

ECEL1 Endothelin converting enzyme-like 1 chr2 233,344,536 233,362,207 1.43 7.16 2.33 5.00E-05 2.86E-04

VIP Vasoactive intestinal peptide chr6 153,054,002 153,080,902 1.91 9.35 2.29 5.00E-05 2.86E-04

MT1G Metallothionein 1G chr16 56,700,645 56,701,980 61.09 287.71 2.24 5.00E-05 2.86E-04

PCOLCE2 Procollagen C-endopeptidase enhancer 2 chr3 142,536,701 142,608,045 2.57 12.07 2.23 5.00E-05 2.86E-04

SFRP4 Secreted frizzled-related protein 4 chr7 37,945,534 37,956,525 30.86 125.85 2.03 5.00E-05 2.86E-04

DCN Decorin chr12 91,539,034 91,576,806 102.16 403.72 1.98 5.00E-05 2.86E-04

TMEM151A Transmembrane protein 151A chr11 66,059,344 66,068,063 1.65 6.39 1.95 5.00E-05 2.86E-04

RD3 Retinal degeneration 3 chr1 211,649,102 211,667,124 1.39 5.29 1.93 5.00E-05 2.86E-04

FABP4 Fatty acid binding protein 4, adipocyte chr8 82,390,731 82,395,473 8.71 32.59 1.90 5.00E-05 2.86E-04

FXYD3 FXYD domain containing ion transport regulator 3 chr19 35,606,414 35,615,228 24.91 88.14 1.82 5.00E-05 2.86E-04

PKP1 Plakophilin 1 chr1 201,252,579 201,302,121 3.18 10.93 1.78 5.00E-05 2.86E-04

chr4 145,664,157 145,666,550 10.80 3.89 -1.47 5.00E-05 2.86E-04

AQP1b Aquaporin 1 (Colton blood group) chr7 30,791,750 30,965,131 142.63 50.74 -1.49 5.00E-05 2.86E-04

NMNAT2 Nicotinamide nucleotide adenylyltransferase 2 chr1 183,217,073 183,387,634 6.38 2.19 -1.54 5.00E-05 2.86E-04

KCNAB1 Potassium voltage-gated channel, shaker-related subfamily, beta member 1 chr3 155,838,336 156,256,927 11.14 3.80 -1.55 5.00E-05 2.86E-04

IGJ Immunoglobulin J polypeptide, linker protein for immunoglobulin alpha and mu polypeptides chr4 71,521,257 71,532,652 70.78 23.84 -1.57 5.00E-05 2.86E-04

ITGA8 Integrin, alpha 8 chr10 15,555,950 15,762,323 45.26 15.18 -1.58 5.00E-05 2.86E-04

POSTN Periostin, osteoblast specific factor chr13 38,136,718 38,172,981 5.91 1.94 -1.60 5.00E-05 2.86E-04

LOC100507387 Non-coding RNA chr5 175,542,798 175,554,408 11.01 3.40 -1.70 5.00E-05 2.86E-04

RGS7BP Regulator of G-protein signaling 7 binding protein chr5 63,801,626 63,910,486 6.04 1.82 -1.73 5.00E-05 2.86E-04

TBX15 T-box15 chr1 119,425,665 119,543,988 5.11 1.53 -1.74 5.00E-05 2.86E-04

CPAMD8 C3 and PZP-like, alpha-2-macroglobulin domain containing 8 chr19 17,003,760 17,137,625 12.55 3.68 -1.77 5.00E-05 2.86E-04

PKD1L2 Polycystic kidney disease 1-like 2 chr16 81,134,483 81,253,975 5.61 1.64 -1.78 5.00E-05 2.86E-04

SULT1E1 Sulfotransferase family 1E, estrogen-preferring, member 1 chr4 70,706,929 70,725,870 8.30 2.35 -1.82 5.00E-05 2.86E-04

TMEM132C Transmembrane protein 132C chr12 128,751,947 129,192,460 8.78 2.35 -1.90 5.00E-05 2.86E-04

EPHA3 EPH receptor A3 chr3 89,156,673 89,531,390 11.95 2.77 -2.11 5.00E-05 2.86E-04

CRHBP Corticotropin releasing hormone binding protein chr5 76,248,679 76,276,815 6.58 1.45 -2.18 5.00E-05 2.86E-04

CXCL14 Chemokine (C-X-C motif) ligand 14 chr5 134,906,370 134,914,969 21.66 4.45 -2.28 5.00E-05 2.86E-04

CALCB Calcitonin-related polypeptide beta chr11 15,095,145 15,100,177 7.76 0.99 -2.97 5.00E-05 2.86E-04

WFDC1 WAP four-disulfide core domain 1 chr16 84,303,639 84,363,457 154.28 14.90 -3.37 5.00E-05 2.86E-04

TRDN Triadin chr6 123,537,304 123,958,601 24.71 2.33 -3.41 5.00E-05 2.86E-04

a ATP10B and LOC285629 were reported as a single transcriptional locus by Cufflinks.

b AQP1, FAM188B, INMT, and INMT-FAM188B (readthrough) were reported as a single transcriptional locus by Cufflinks.

Table 6

All differentially expressed genes between temporal RPE/choroid vs. nasal RPE/choroid with q-value < 0.001 and absolute log2(fold-change) > 1.

Symbol Description Chromosome Start Stop Nasal RPE/ Temporal RPE/ Log2(FC) P-value Q-value

choroid (FPKM) choroid (FPKM)

MPZ Myelin protein zero chr1 161,274,410 161,279,793 5.28 40.31 2.93 5.00E-05 2.86E-04

VIP Vasoactive intestinal peptide chr6 153,054,002 153,080,902 2.62 9.35 1.84 5.00E-05 2.86E-04

SCN7A Sodium channel, voltage-gated, type VII, alpha chr2 167,260,082 167,350,946 3.33 6.72 1.01 1.50E-04 8.01E-04

subunit

WFDC1 WAP four-disulfide core domain 1 chr16 84,303,639 84,363,457 33.71 14.90 -1.18 5.00E-05 2.86E-04

HBA2 Hemoglobin, alpha 2 chr16 222,845 223,709 139.35 56.67 -1.30 5.00E-05 2.86E-04

TRDN triadin chr6 123,537,304 123,958,601 7.99 2.33 -1.78 5.00E-05 2.86E-04

TULP1 Tubby like protein 1 chr6 35,465,650 35,480,663 7.68 2.21 -1.80 5.00E-05 2.86E-04

TF Transferrin chr3 133,426,380 133,497,850 46.00 12.52 -1.88 1.00E-04 5.49E-04

GUCA1B Guanylate cyclase activator 1B (retina) chr6 42,151,021 42,168,689 7.48 1.62 -2.21 5.00E-05 2.86E-04

RCVRN Recoverin chr17 9,801,026 9,808,684 31.88 4.18 -2.93 5.00E-05 2.86E-04

SAG S-antigen; retina and pineal gland (arrestin) chr2 234,210,734 234,255,703 53.93 6.09 -3.15 5.00E-05 2.86E-04

GNAT1 Guanine nucleotide binding protein (G protein), chr3 50,229,042 50,235,129 16.38 1.54 -3.41 5.00E-05 2.86E-04

alpha transducing activity polypeptide 1

AIPL1 Aryl hydrocarbon receptor interacting proteinlike 1 chr17 6,327,058 6,338,519 16.67 1.53 -3.44 5.00E-05 2.86E-04

RHO Rhodopsin chr3 129,247,481 129,254,187 73.90 6.61 -3.48 5.00E-05 2.86E-04

Table 7

Top 40 differentially expressed genes between macular RPE/choroid vs. macular retina with q-value < 0.001 and absolute log2(fold-change) > 1.

Symbol Description Chromosome Start Stop Macular retina (FPKM) Macular RPE/ choroid (FPKM) Log2(FC) P-value Q-value

Potentially novel chr14 106,512,071 106,518,924 0.00 322.59 Inf 5.00E-05 2.86E-04

PRRX2 Paired related homeobox 2 chr9 132,427,911 132,484,952 0.00 60.99 Inf 5.00E-05 2.86E-04

PITX1 Paired-like homeodomain 1 chr5 134,363,350 134,370,461 0.00 25.51 Inf 5.00E-05 2.86E-04

ABCB5 ATP-binding cassette, sub- chr7 20,655,244 20,796,642 0.00 21.83 Inf 5.00E-05 2.86E-04

family B (MDR/TAP), member 5

CCL23 Chemokine (C—C motif) ligand T2 chr17 34,340,096 34,345,026 0.00 21.39 Inf 5.00E-05 2.86E-04

CPXM1 23 Carboxypeptidase X (M14 chr20 2,774,714 2,781,292 0.00 19.60 Inf 5.00E-05 2.86E-04

family), member 1

Potentially novel chr14 106,573,231 106,791,526 0.00 18.70 Inf 5.00E-05 2.86E-04

LOC340357 Long non-coding RNA chr8 12,623,570 12,675,830 0.00 14.19 Inf 5.00E-05 2.86E-04

Potentially novel chr2 72,375,757 72,376,046 0.00 14.09 Inf 5.00E-05 2.86E-04

GZMK Granzyme K (granzyme 3; chr5 54,320,106 54,329,960 0.00 9.88 Inf 5.00E-05 2.86E-04

tryptase II)

Potentially novel chr11 23,099,882 23,100,568 0.00 8.98 Inf 5.00E-05 2.86E-04

SULT1E1 Sulfotransferase family 1E, chr4 70,706,929 70,725,870 0.00 8.30 Inf 5.00E-05 2.86E-04

estrogen-preferring, member 1

CD3E CD3e molecule, epsilon (CD3- chr11 118,175,294 118,186,890 0.00 7.73 Inf 5.00E-05 2.86E-04

TCR complex)

Potentially novel chr9 69,616,397 69,650,111 0.00 6.93 Inf 5.00E-05 2.86E-04

Potentially novel chr12 85,380,571 85,386,590 0.00 6.24 Inf 5.00E-05 2.86E-04

KCNK17 Potassium channel, subfamily chr6 39,266,776 39,282,237 0.00 5.88 Inf 5.00E-05 2.86E-04

K, member 17

Potentially novel chr5 39,891,795 40,053,420 0.00 4.94 Inf 5.00E-05 2.86E-04

LINC00226 Long intergenic non-protein chr14 106,573,231 106,791,526 0.00 4.93 Inf 5.00E-05 2.86E-04

coding RNA 226

Potentially novel chr5 134,374,392 134,375,704 0.00 4.50 Inf 5.00E-05 2.86E-04

CTSG Cathepsin G chr14 25,042,723 25,045,466 0.00 4.32 Inf 5.00E-05 2.86E-04

Potentially novel chr16 84,273,764 84,281,615 3.58 0.00 -Inf 5.00E-05 2.86E-04

DEFB131 Defensin, beta 131 chr4 9,446,259 9,452,240 3.61 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr4 147,559,182 147,559,943 3.65 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr14 48,702,743 48,793,893 3.69 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr7 21,181,327 21,253,288 3.82 0.00 -Inf 5.00E-05 2.86E-04

LRTM2 Leucine-rich repeats and chr12 1,901,122 2,032,895 4.46 0.00 -Inf 1.00E-04 5.49E-04

transmembrane domains 2

Potentially novel chr3 76,359,774 76,360,809 4.76 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr12 55,403,375 55,409,332 5.28 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr20 61,785,077 61,787,572 5.67 0.00 -Inf 5.00E-05 2.86E-04

DSCR8 Down syndrome critical region chr21 39,493,544 39,528,605 5.99 0.00 -Inf 5.00E-05 2.86E-04

gene 8

TRPC7 Transient receptor potential chr5 135,548,424 135,732,845 7.25 0.00 -Inf 5.00E-05 2.86E-04

cation channel, subfamily C,

member 7

CCDC172 Coiled-coil domain containing chr10 118,083,939 118,139,551 7.56 0.00 -Inf 5.00E-05 2.86E-04

172 Potentially novel chr8 55,506,068 55,508,509 8.16 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr1 23,280,951 23,299,340 9.02 0.00 -Inf 5.00E-05 2.86E-04

FEZF2 FEZ family zinc finger 2 chr3 62,355,295 62,360,692 9.07 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr18 5,847,207 5,876,306 9.92 0.00 -Inf 5.00E-05 2.86E-04

POU4F2 POU class 4 homeobox 2 chr4 147,560,044 147,563,623 11.13 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr5 178,422,323 178,423,333 12.42 0.00 -Inf 5.00E-05 2.86E-04

CACNG3 Calcium channel, voltage- chr16 24,266,873 24,373,737 16.08 0.00 -Inf 5.00E-05 2.86E-04

dependent, gamma subunit 3

Potentially novel chr12 9,727,536 9,728,249 30.87 0.00 -Inf 5.00E-05 2.86E-04

3.1. Comparison with previously published data

We compared our dataset to a recently published table of FPKM (Fragments Per Kilobase of transcript per Million mapped reads) values derived from macular and peripheral retina and RPE/ choroid/sclera for 91 genes implicated by genome-wide association studies (GWAS) in various eye diseases (Li et al., 2014). To compare the previous results to ours, we averaged our nasal and temporal FPKM values together for one "peripheral" value (Fig. 2). Where Cufflinks matched two separate loci to a single gene symbol, we took the higher of the two values. While the values for retina roughly follow the diagonal, FPKM values were consistently higher for genes in our RPE/choroid samples compared to the RPE/

choroid/sclera samples of Li et al., likely reflecting the cellular paucity of the sclera. The obvious exception is COLWA1, which shows higher expression in peripheral RPE/choroid/sclera than in peripheral RPE/choroid alone. In addition, TIMP3, reported by Li et al. as a predominantly retinal transcript, showed robust expression in all RPE/choroid samples in the current report (Fig. 2).

3.2. Regional gene expression by tissue

To evaluate gene expression across regions of the retina and RPE/choroid, we used Cuffdiff to perform pairwise tests between groups of punches (Supplemental Table 1). When comparing nasal

Table 8

Top 40 differentially expressed genes between nasal RPE/choroid vs. nasal retina with q-value < 0.001 and absolute log2(fold-change) > 1.

Symbol Description Chromosome Start Stop Nasal retina (FPKM) Nasal RPE/choroid (FPKM) Log2(FC) P-value Q-value

Potentially novel chr14 106,512,071 106,518,924 0.00 142.27 Inf 5.00E-05 2.86E-04

P1TX1 Paired-like homeodomain 1 chr5 134,363,350 134,370,461 0.00 19.53 1nf 5.00E-05 2.86E-04

L0C340357 Long non-coding RNA chr8 12,623,570 12,675,830 0.00 10.88 1nf 5.00E-05 2.86E-04

GZMK Granzyme K (granzyme 3; chr5 54,320,106 54,329,960 0.00 8.65 1nf 5.00E-05 2.86E-04

tryptase II)

KCNK17 Potassium channel, subfamily chr6 39,266,776 39,282,237 0.00 8.45 1nf 5.00E-05 2.86E-04

K, member 17

CCL23 Chemokine (C—C motif) ligand T2 chr17 34,340,096 34,345,026 0.00 7.03 1nf 5.00E-05 2.86E-04

23 Potentially novel chr5 39,891,795 40,053,420 0.00 6.18 1nf 5.00E-05 2.86E-04

TEX41 Testis expressed 41 (non- chr2 145,425,533 146,021,001 0.00 5.35 1nf 5.00E-05 2.86E-04

protein coding)

Potentially novel chr14 22,320,484 22,323,708 0.00 4.66 1nf 5.00E-05 2.86E-04

Potentially novel chr5 134,374,392 134,375,704 0.00 3.67 1nf 5.00E-05 2.86E-04

KLRB1 Killer cell lectin-like receptor chr12 9,747,869 9,760,497 0.00 3.53 1nf 5.00E-05 2.86E-04

subfamily B, member 1

CCL13 Chemokine (C—C motif) ligand chr17 32,683,470 32,685,629 0.00 3.52 1nf 5.00E-05 2.86E-04

13 Potentially novel chr18 3,466,243 3,478,970 0.00 2.99 1nf 5.00E-05 2.86E-04

LMX1B LIM homeobox transcription chr9 129,376,673 129,463,311 0.00 2.93 1nf 5.00E-05 2.86E-04

factor 1, beta

SLAMF7 SLAM family member 7 chr1 160,709,032 160,725,021 0.00 1.57 1nf 5.00E-05 2.86E-04

ZG16B Zymogen granule protein 16B chr16 2,880,172 2,882,285 0.00 1.53 1nf 5.00E-05 2.86E-04

SAA1 Serum amyloid A1 chrll 18,287,807 18,291,523 0.45 2255.58 12.30 5.00E-05 2.86E-04

PLA2G2A Phospholipase A2, group IIA chrl 20,301,923 20,306,932 0.27 553.86 11.01 5.00E-05 2.86E-04

(platelets, synovial fluid)

ABCB5 ATP-binding cassette, sub- chr7 20,655,244 20,796,642 0.01 18.83 10.65 1.50E-04 8.01E-04

family B (MDR/TAP), member 5

PLVAP Plasmalemma vesicle chr19 17,462,251 17,488,137 0.09 122.71 10.47 5.00E-05 2.86E-04

associated protein

KCNC2 Potassium voltage-gated chr12 75,433,857 75,603,528 17.88 0.18 -6.60 5.00E-05 2.86E-04

channel, Shaw-related

subfamily, member 2

NEFM Neurofilament, medium chr8 24,771,273 24,776,612 169.15 1.71 -6.63 5.00E-05 2.86E-04

polypeptide

CNGA3 Cyclic nucleotide gated channel chr2 98,703,594 99,016,789 4.16 0.04 -6.67 5.00E-05 2.86E-04

alpha 3

RASGRF1 Ras protein-specific guanine chr15 79,252,288 79,383,265 14.36 0.14 -6.69 5.00E-05 2.86E-04

nucleotide-releasing factor 1

CACNG5 Calcium channel, voltage- chr17 64,831,229 64,890,871 19.03 0.18 -6.72 5.00E-05 2.86E-04

dependent, gamma subunit 5

AGBL4 ATP/GTP binding protein-like 4 chr1 48,998,369 50,489,626 2.06 0.02 -6.76 5.00E-05 2.86E-04

Potentially novel chr10 3,675,719 3,690,589 7.93 0.05 -7.41 5.00E-05 2.86E-04

CRYAA Crystallin, alpha A chr21 44,570,160 44,592,913 318.89 1.65 -7.59 5.00E-05 2.86E-04

C6orf7 Chromosome 6 open reading chr6 80,472,023 80,580,137 2.24 0.01 -8.22 5.00E-05 2.86E-04

frame 7

Potentially novel chr6 89,881,239 89,881,908 1.30 0.00 -1nf 5.00E-05 2.86E-04

LEMD1 LEM domain containing 1 chr1 205,342,379 205,436,588 1.54 0.00 -1nf 5.00E-05 2.86E-04

Potentially novel chr9 32,865,800 32,909,556 2.00 0.00 -1nf 5.00E-05 2.86E-04

Potentially novel chr17 5,000,474 5,000,915 2.59 0.00 -1nf 5.00E-05 2.86E-04

Potentially novel chrll 3,599,999 3,602,427 2.82 0.00 -1nf 5.00E-05 2.86E-04

Potentially novel chr1 10,873,924 10,874,785 3.12 0.00 -1nf 5.00E-05 2.86E-04

Potentially novel chr16 84,273,764 84,281,615 3.15 0.00 -1nf 5.00E-05 2.86E-04

Potentially novel chr1 91,183,265 91,183,539 3.39 0.00 -1nf 5.00E-05 2.86E-04

C1orf141 Chromosome 1 open reading chr1 67,557,494 67,607,567 3.55 0.00 -1nf 5.00E-05 2.86E-04

frame 141

Potentially novel chr10 81,657,061 81,657,275 3.67 0.00 -1nf 5.00E-05 2.86E-04

DEFB131 Defensin, beta 131 chr4 9,446,259 9,452,240 4.55 0.00 -1nf 5.00E-05 2.86E-04

retina vs. macular retina, 30 genes showed increased expression in nasal retina and 128 genes showed decreased expression (Table 2). When comparing temporal retina vs. macular retina, 37 genes showed increased expression in temporal retina and 323 genes showed decreased expression (Table 3). Notably, of the 40 top differences in expression between nasal retina and macular retina (Table 2), 21 of the same genes were in the list of top 40 differences between temporal retina and macular retina (Table 3). Interestingly, no genes were differentially expressed between temporal and

nasal retina, suggesting that these two regions of the retina are molecularly indistinguishable using these criteria.

In the nasal vs. macular RPE/choroid comparison, 81 genes showed increased expression in the nasal RPE/choroid and 39 genes showed decreased expression (Table 4). When comparing temporal vs. macular RPE/choroid, 70 genes were increased in the temporal RPE/choroid and 44 genes were decreased (Table 5). In contrast to the retina, in which no gene expression differences between different peripheral regions met our criteria for

Table 9

Top 40 differentially expressed genes between temporal RPE/choroid vs. temporal retina with q-value < 0.001 and absolute log2(fold-change) > 1.

Symbol Description Chromosome Start Stop Temporal retina (FPKM) Temporal RPE/ choroid (FPKM) Log2(FC) P-value Q-value

Potentially novel chr14 106,512,071 106,518,924 0.00 192.09 Inf 5.00E-05 2.86E-04

PITX1 Paired-like homeodomain 1 chr5 134,363,350 134,370,461 0.00 19.83 Inf 5.00E-05 2.86E-04

LOC34Ö357 Long non-coding RNA chr8 12,623,570 12,675,830 0.00 14.87 Inf 5.00E-05 2.86E-04

PAX3 Paired box 3 chr2 223,064,605 223,169,936 0.00 13.21 Inf 5.00E-05 2.86E-04

AKR1B10 Aldo-keto reductase family 1, member B10 (aldose reductase) chr7 134,200,788 134,226,166 0.00 8.67 Inf 5.00E-05 2.86E-04

GZMK Granzyme K (granzyme 3; tryptase II) chr5 54,320,106 54,329,960 0.00 8.60 Inf 5.00E-05 2.86E-04

Potentially novel chr1 170,632,134 170,633,074 0.00 8.30 Inf 5.00E-05 2.86E-04

Potentially novel chr5 39,891,795 40,053,420 0.00 7.70 Inf 5.00E-05 2.86E-04

FOXD3 Forkhead box D3 chr1 63,786,575 63,790,797 0.00 6.14 Inf 5.00E-05 2.86E-04

TEX41 Testis expressed 41 (non-protein coding) chr2 145,425,533 146,021,001 0.00 5.47 Inf 5.00E-05 2.86E-04

Potentially novel chr11 23,099,882 23,100,568 0.00 5.43 Inf 5.00E-05 2.86E-04

CLEC10A C-type lectin domain family 10, member A chr17 6,977,855 6,983,826 0.00 4.53 Inf 5.00E-05 2.86E-04

LCK LCK proto-oncogene, Src family tyrosine kinase chr1 32,716,839 32,751,766 0.00 3.57 Inf 5.00E-05 2.86E-04

LMX1B LIM homeobox transcription factor 1, beta chr9 129,376,673 129,463,311 0.00 3.50 Inf 5.00E-05 2.86E-04

Potentially novel chr9 69,616,397 69,650,111 0.00 3.46 Inf 5.00E-05 2.86E-04

S1PR5 Sphingosine-1-phosphate receptor 5 chr19 10,623,417 10,628,668 0.00 3.27 Inf 5.00E-05 2.86E-04

CLIC3 Chloride intracellular channel 3 chr9 139,889,059 139,891,024 0.00 3.18 Inf 5.00E-05 2.86E-04

Potentially novel chr5 134,374,392 134,375,704 0.00 3.17 Inf 5.00E-05 2.86E-04

Potentially novel chr18 3,466,243 3,478,970 0.00 2.72 Inf 5.00E-05 2.86E-04

CCL13 Chemokine (C—C motif) ligand 13 chr17 32,683,470 32,685,629 0.00 2.56 Inf 5.00E-05 2.86E-04

Potentially novel chr11 117,673,186 117,673,498 3.81 0.00 -Inf 5.00E-05 2.86E-04

C1orf141 Chromosome 1 open reading frame 141 chr1 67,557,494 67,607,567 3.83 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr18 76,736,580 76,739,475 4.33 0.00 -Inf 5.00E-05 2.86E-04

NAT16 N-acetyltransferase 16 (GCN5-related, putative) chr7 100,813,667 100,823,557 4.36 0.00 -Inf 5.00E-05 2.86E-04

C16orf11 Chromosome 16 open reading frame 11 chr16 610,179 615,529 4.39 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr12 55,403,375 55,409,332 4.56 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr1 10,873,924 10,874,785 4.76 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr17 77,818,917 77,825,465 4.77 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr3 192,862,226 192,894,687 5.02 0.00 -Inf 5.00E-05 2.86E-04

DEFB131 Defensin, beta 131 chr4 9,446,259 9,452,240 5.46 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr7 127,116,864 127,146,567 6.15 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr7 21,181,327 21,253,288 6.30 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr1 23,280,951 23,299,340 6.93 0.00 -Inf 5.00E-05 2.86E-04

MIR124-3 MicroRNA 124-3 chr20 61,808,787 61,813,324 6.97 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr20 61,785,077 61,787,572 9.70 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr8 55,506,068 55,508,509 11.21 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr5 178,422,323 178,423,333 11.40 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr9 2,734,043 2,734,455 16.22 0.00 -Inf 5.00E-05 2.86E-04

DEFB119 Defensin, beta 119 chr20 29,964,965 29,978,452 56.91 0.00 -Inf 5.00E-05 2.86E-04

Potentially novel chr12 9,727,536 9,728,249 58.97 0.00 -Inf 5.00E-05 2.86E-04

differential expression, three genes (MPZ, VIP, and SCN7A) were increased in the nasal RPE/choroid and 11 were decreased when comparing temporal vs. nasal RPE/choroid (Table 6). Higher expression of neuroretinal genes (e.g., RHO, RCVRN, SAG) in the nasal RPE/choroid group as compared to the macular (Table 4) or temporal (Table 6) RPE/choroid groups is traceable to the slight but detectable neuroretinal contamination of the nasal RPE/ choroid from donor 2.

We also performed comparisons between tissues within a region. We observed 2747 genes with increased expression and 2053 genes with decreased expression when comparing macular RPE/ choroid vs. macular retina (Table 7). We observed 2842 genes with increased expression and 1935 genes with decreased expression when comparing nasal RPE/choroid vs. nasal retina (Table 8). We observed 2762 genes with increased expression and 1909 genes with decreased expression when comparing temporal RPE/choroid

nasal vs. macula

nasal vs. macula temporal vs. macula

down up

nasal vs. macula temporal vs. macula

up f 15 ______' 18 down

/ 30 1 \ 21 \

51 23 / /

Fig. 3. Intersection between differentially expressed genes sets. (A) Regional pairwise comparisons in the retina. (B) Regional pairwise comparisons in the RPE/choroid. (C) RPE/ choroid compared to retina for each region without respect to direction of fold change.

vs. temporal retina (Table 9). Fig. 3 shows the overlaps between differentially expressed genes in these comparisons.

3.3. Regional expression differences for specific retinal cell types

To explore regional differences in cell types in the retina, we mapped 92 genes identified by Siegert and coworkers as cell type specific in mouse retina (Siegert et al., 2012) to 78 human genes in our dataset (we required genes with human homologs to be expressed above FPKM 0 in at least half of our human retina samples). As indicated in Fig. 4, all four macula samples cluster together, while nasal and temporal samples are intermixed. All photore-ceptor genes are enriched in the peripheral samples, whereas several genes of ganglion cell and amacrine cells are interleaved in a macula-enriched cluster.

3.4. RPE-specific and endothelium associated gene expression in the RPE/choroid

To compare cell populations in the RPE/choroid, we selected predefined RPE-specific and endothelium associated genes sets (Whitmore et al., 2013) and assessed trends in expression across temporal, macular, and nasal RPE/choroid punches (Fig. 5). The RPE-specific set shows an overall lower expression in the macula while the endothelium associated set shows higher expression in the macula.

3.5. Intradonor variation in five regions of the retina

In a separate experiment, we took retinal punches from macula, superior, inferior, temporal, and nasal regions of a single human donor eye and performed pairwise correlation for all genes expressed above FPKM 1 in a least one punch. Macula as compared to any of the peripheral punches showed the greatest dispersion

rTU ^rHr^

cell type

amacrine

■ bipolar cell

■ ganglion cell

■ horizontal cell

■ microglia

■ photoreceptor

aaaaaaaaaaaa EEEEEEEEEEEE 'S'S'S'S'S'S'S'S'S'S'S'S

(ECECECErccECECECECECECE

-5-5-5-580000888 mmmm eeee

Fig. 4. Clustering of cell-type specific genes in the retina. Lighter shading indicates higher expression while darker shading indicates lower expression (FPKM values after row scaling). Hierarchical clustering was performed with complete linking on Spearman rank correlation distance.

from the diagonal (r from 0.97 to 0.98; Fig. 6). All peripheral vs. peripheral comparisons were also highly correlated, (r = 0.99), with modest dispersion from the diagonal in the superior vs. nasal and nasal vs. inferior comparisons.

4. Discussion

Here we present the first RNA-Seq dataset to separately investigate gene expression in the macula, nasal, and temporal human retina and RPE/choroid. While intriguing differences between the peripheral and the macular tissues were found, we did not observe significant differences between nasal and temporal retina. Only a few genes were differentially expressed between nasal and temporal RPE/choroid. Thus, the use of either nasal or temporal retina as general surrogates of "peripheral" retina seems warranted based on these data.

4.1. Limitations of the current study

RNA-Seq provides a wealth of data from each sample but interpreting these data remains challenging. The most commonly used measurement of expression, fragments per kilobase mapped per million reads (FPKM), should not be interpreted as a straightforward reflection of the number of RNA molecules within a given

cell or tissue punch. Moreover, as any RNA-Seq derived expression estimate is fundamentally based on digital read counts, Gaussian models used for microarray analysis do not apply. To reduce false positives, we selected stringent thresholds for calling a gene as differentially expressed, likely missing some biologically relevant true positive events. Several neural retina genes, including RHO, RCVRN, and SAG, showed increased expression in the nasal RPE/ choroid. This is likely due to incomplete separation of neural retina from the RPE during dissection, a problem widely reported in previous studies of ocular gene expression. Moreover, a larger sample size may provide enough power to detect transcriptome-level difference between nasal and temporal retina. However, despite our sample size, our data showed strong rank correlations with published FPKM values of peripheral and macular retina and RPE/choroid/sclera (Li et al., 2014).

4.2. Molecular determinants of the macula

Identifying the molecular determinants of the macular-peripheral distinction remains an open challenge in vision research. We found several transcription factors that were differentially expressed in the macula and periphery. For instance, FOXI3 was upregulated in the nasal retina compared to the macular retina, whereas IRX2 was downregulated in this comparison. SIX2 was upregulated in the temporal retina compared to the macular retina, whereas POU4F1, POU4F2, IRX1, and IRX2 were downregulated in this comparison. POU4F1 and its target, RIT2 (Zhang et al., 2013), are enriched in retinal ganglion cells (Kim et al., 2006), and both genes were significantly enriched in our macular retina samples. RIT2 is also expressed in the mouse inner nuclear layer (Zhang et al., 2013).

4.3. Implications for vasculature in the retina and RPE/choroid

We observed decreased expression of prolactin (PRL) in the macular neuroretina as compared to either peripheral region. Prolactin is anti-angiogenic in the retina (Aranda et al., 2005), and systemic prolactin levels correlate with diabetic retinopathy (Arnold et al., 2010). Previous studies in rats and green monkeys have shown that PRL is expressed throughout the cell layers of the neural retina, and its receptor, PRLR, is expressed in photoreceptor nuclei, inner nuclear layer, and ganglion cell layer (Rivera et al., 2008). To our knowledge, this is the first dataset showing macular-peripheral differential expression of PRL in the retina. Distribution of PRL and other regulators of angiogenesis may help to explain the tendency of choroidal neovascular membranes to grow towards the fovea (Klein et al., 1989; de Jong, 2006), particularly if the neovascular membrane had breached the RPE or if RPE barrier function was compromised.

Of our predefined endothelium associated gene set, MGP (matrix Gla protein) showed the highest expression in the RPE/choroid. MGP may prevent calcification in the choroidal stroma or in Bruch's membrane (Booij et al., 2010). Recently, Gonzalez and coworkers used the MGP promoter to drive expression of b-galactosidase in a gene vector in anterior pole cultures of human donor eyes (Gonzalez et al., 2004). They observed localized expression in the trabecular meshwork (Gonzalez et al., 2004). However, they did not examine the tissues of the posterior pole. The Ocular Tissue Database (Wagner et al., 2013) entry for MGP (available at https:// genome.uiowa.edu/otdb/) shows relatively high expression in the RPE/choroid, on par with the trabecular meshwork, ciliary body, and iris, and low expression in the retina, consistent with our findings. Further study will be needed to determine the extent to which MGP contributes to Bruch's membrane and/or choroidal elasticity.

RPE-specific genes

endothelium associated genes

temporal macula

temporal macula

Fig. 5. Expression of RPE-specific and endothelium associated gene sets across regions of the RPE/choroid.

4.4. Comparison with previously published data

We also observed robust expression of TIMP3 in our dataset, compared to FPKM values of 0 in the RPE/choroid in that of Li et al. These authors noted that expression of TIMP3 was almost entirely limited to the retina. This result may be a side effect of the algorithm used by Cufflinks. When Cufflinks estimates FPKM values, it first computes "fragment bundles." Extremely high bundle values can cause difficulty for the algorithm; thus, Cufflinks labels such genes as "HIDATA" and does not estimate FPKMs. In our initial analysis, we failed to detect RHO in both nasal and temporal retina. However, inspection of individual alignment files for these samples showed robust expression, with read counts in excess of 40,000 across the RHO transcript in temporal samples (Supplemental Figure S1). Further investigation revealed that identified expression of TTR also marked by the "HIDATA" criterion in the nasal RPE/ choroid. Setting the "-max_bundle_frags" parameter for Cufflinks and Cuffdiff to an arbitrarily high value (e.g., 20,000,000) removed the "HIDATA" label, resulting in FPKM estimates for RHO and TTR in all groups.

4.5. Utility of ophthalmic expression datasets

Linkage studies have identified many disease loci in families affected by macular dystrophies or other eye diseases (see examples at RetNet [https://sph.uth.edu/retnet/] and Online Mendelian Inheritance in Man [http://omim.org/]). Causative mutations have yet to be pinpointed within several of these loci, which cover megabases of the genome and contain dozens of genes. Multi-tissue regional ophthalmic datasets, such as we present here, may help reduce the search space for candidate genes, especially for

diseases with region-specific or tissue-specific phenotypes. Moreover, as more Mendelian diseases with coding mutations are solved, identifying silent mutations and changes in regulatory regions is the remaining frontier for molecular genetics. For instance, in 2002 Schulz and coworkers examined NPVF(formerly C7orf9) for coding sequence changes that could explain Cystoid Macular Edema (chr7:7,300,000-28,800,000; OMIM 153880). They found no deleterious SNPs in the coding sequence of NPVF, but they acknowledged that the promoter sequence may harbor the causative mutation (Schulz et al., 2002). In our data, we observe significant macular enrichment of NPVF as compared to both nasal and temporal retina, suggesting that promoter changes could alter the region-specific expression of NPVF in the retina.

MAK-related RP (OMIM 614181) tends to spare the temporal retina while damaging the inferior and nasal retina (Stone et al., 2011; Tucker et al., 2011). While we observed lower expression of MAK in the macula (Fig. 7A), similar to many other photoreceptor cell specific genes (Fig. 4), differences in peripheral regions of the retina were not detected. This suggests that the regional degeneration observed in MAK-related RP is likely due to the distribution of substrates or factors regulating MAK function rather than differences in gene expression across retinal topography.

BEST1, the gene mutated in Best vitelliform macular dystrophy (OMIM 153700), has previously been shown to be expressed at lower levels in the macular RPE/choroid than in the periphery by immunohistochemistry, quantitative PCR, and proteomics (Mullins et al., 2007; Skeie and Mahajan, 2014). We observed a similar pattern of expression in our data (Figs. 5 and 7B). However, comparable profiles were seen for many other RPE-specific genes, including RPE65, a gene previously shown to have relatively equal abundance between macular and peripheral RPE/choroid (Kociok

024 024 024 024 024

log10(FPKM+1)

Fig. 6. Intradonor region variability across five regions of the retina.

A. MAK

B. BEST1

temporal retina

macular retina

nasal retina

50004000 3000200010000-

temporal macular nasal RPE/choroid RPE/choroid RPE/choroid

Fig. 7. Expression of MAK (retina) and BESTI (RPE/choroid) across regions. Mean indicated by ( x ). Error bars denote 95% confidence intervals for abundance as calculated by Cuffdiff.

and Joussen, 2007). One implication of these findings, as depicted in Fig. 5, is that normalization should be performed using cell-type specific genes with relatively consistent expression across anatomic regions.

5. Conclusion

We have presented a comprehensive, transcriptome-wide view of gene expression across the neural retina and RPE/choroid in three regions of the retina. Gene expression in the retina correlates with the anatomic distribution of rod photoreceptors and ganglion cells in the macula and periphery. RPE-specific and endothelium associated gene expression in the RPE/choroid appear inversely related, with RPE-specific genes expressed at a lower level in the macula than at the periphery, whereas endothelium associated genes show slightly higher expression in the macula than in the periphery. Additional RNA-Seq datasets for ophthalmic tissues will facilitate more complex analyses, including mapping retina-specific transcriptional regulatory circuits and identifying lowly expressed exons.

Both analyses will facilitate the search for novel disease causing mutations. For example, we previously used the five region RNA-Seq data to identify rare alternative exons in ABCA4, screened a cohort of patients with clinical evidence of ABCA4 disease for variants in these exons, and identified novel splice-site mutations not found in controls (Braun et al., 2013). To advance similar studies, we are making our samples available on the database of Genotypes and Phenotypes (dbGAP).

Acknowledgments

This project was supported by NIH Health Grants EY023187 (TES), EY024605 (RFM), EY016822 (EMS), and DP2-0D007483-01 (BAT), the Howard Hughes Medical Institute (EMS), The Elmer and Sylvia Sramek Charitable Foundation (RFM/BAT), the Hansjoerg EJW Kolder MD, PhD Professorship for Best Disease (RFM), and Vision for Tomorrow. This research was supported in part through computational resources provided by The University of Iowa, Iowa City, Iowa.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.exer.2014.11.001.

References

Aranda, J., Rivera, J.C., Jeziorski, M.C., Riesgo-Escovar, J., Nava, G., Lopez-Barrera, F., et al., 2005 Aug. Prolactins are natural inhibitors of angiogenesis in the retina. Investigative Ophthalmol. Vis. Sci. 46 (8), 2947-2953. Arnold, E., Rivera, J.C., Thebault, S., Moreno-Paramo, D., Quiroz-Mercado, H., Quintanar-Stéphano, A., et al., 2010 Dec. High levels of serum prolactin protect against diabetic retinopathy by increasing ocular vasoinhibins. Diabetes 59 (12), 3192-3197. PMCID: PMC2992782. Booij, J.C., Baas, D.C., Beisekeeva, J., Gorgels, T.G.M.F., Bergen, A.A.B., 2010 Jan. The

dynamic nature of Bruch's membrane. Prog. Retin. Eye Res. 29 (1), 1-18. Bowes Rickman, C., Ebright, J.N., Zavodni, Z.J., Yu, L., Wang, T., Daiger, S.P., et al., 2006 Jun. Defining the human macula transcriptome and candidate retinal disease genes using EyeSAGE. Investigative Ophthalmol. Vis. Sci. 47 (6), 2305-2316. PMCID: PMC2813776. Braun, T.A., Mullins, R.F., Wagner, A.H., Andorf, J.L., Johnston, R.M., Bakall, B.B., et al., 2013. Non-exomic and synonymous variants in ABCA4 are an important cause of Stargardt disease. Hum. Mol. Genet. 22 (25), 5136-5145. Curcio, C.A., Allen, K.A., 1990 Oct 1. Topography of ganglion cells in human retina.

J. Comp. Neurology 300 (1), 5-25. Curcio, C.A., Sloan, K.R., Kalina, R.E., Hendrickson, A.E., 1990 Feb 22. Human

photoreceptor topography. J. Comp. Neurology 292 (4), 497-523. de Jong, P.T.V.M., 2006 Oct 5. Age-related macular degeneration. N. Engl. J. Med. 355 (14), 1474-1485.

DeLuca, D.S., Levin, J.Z., Sivachenko, A., Fennell, T., Nazaire, M.-D., Williams, C., et al.,

2012 Jun 1. RNA-SeQC: RNA-seq metrics for quality control and process optimization. Bioinformatics 28 (11), 1530-1532. PMCID: PMC3356847.

Diehn, J.J., Diehn, M., Marmor, M.F., Brown, P.O., 2005. Differential gene expression in anatomical compartments of the human eye. Genome Biol. 6 (9), R74. PMCID: PMC1242209.

Gonzalez, P., Caballero, M., Liton, P.B., Stamer, W.D., Epstein, D.L., 2004 May. Expression analysis of the matrix GLA protein and VE-cadherin gene promoters in the outflow pathway. Investigative Ophthalmol. Vis. Sci. 45 (5), 1389-1395.

Ishibashi, K., Tian, J., Handa, J.T., 2004 Sep. Similarity of mRNA phenotypes of morphologically normal macular and peripheral retinal pigment epithelial cells in older human eyes. Investigative Ophthalmol. Vis. Sci. 45 (9), 3291-3301.

Jonas, J.B., Schneider, U., Naumann, G.O., 1992. Count and density of human retinal photoreceptors. Graefes Arch. Clin. Exp. Ophthalmol. 230 (6), 505-510.

Kim, C.Y., Kuehn, M.H., Clark, A.F., Kwon, Y.H., 2006. Gene expression profile of the adult human retinal ganglion cell layer. Mol. Vis. 12,1640-1648.

Kim, D., Pertea, G., Trapnell, C., Pimentel, H., Kelley, R., Salzberg, S.L., 2013 Apr 25. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. 14 (4), R36.

Klein, M.L., Jorizzo, P.A., Watzke, R.C., 1989 Sep. Growth features of choroidal neo-vascular membranes in age-related macular degeneration. Ophthalmology 96 (9), 1416-1419 discussion1420-1.

Kociok, N., Joussen, A.M., 2007 Jan. Varied expression of functionally important genes of RPE and choroid in the macula and in the periphery of normal human eyes. Graefes Arch. Clin. Exp. Ophthalmol. 245 (1), 101-113.

Li, M., Jia, C., Kazmierkiewicz, K.L., Bowman, A.S., Tian, L., Liu, Y., et al., 2014. Comprehensive analysis of gene expression in human retina and supporting tissues. Hum. Mol. Genet. 23 (15), 4001-4014.

Mullins, R.F., Kuehn, M.H., Faidley, E.A., Syed, N.A., Stone, E.M., 2007 Jul. Differential macular and peripheral expression of bestrophin in human eyes and its implication for Best disease. Investigative Ophthalmol. Vis. Sci. 48 (7), 3372-3380.

Radeke, M.J., Peterson, K.E., Johnson, L.V., Anderson, D.H., 2007 Sep. Disease susceptibility of the human macula: differential gene transcription in the retinal pigmented epithelium/choroid. Exp. Eye Res. 85 (3), 366-380.

Rivera, J.C., Aranda, J., Riesgo, J., Nava, G., Thebault, S., López-Barrera, F., et al., 2008 Feb. Expression and cellular localization of prolactin and the prolactin receptor in mammalian retina. Exp. Eye Res. 86 (2), 314-321.

Rouhani, F., Kumasaka, N., de Brito, M.C., Bradley, A., Vallier, L., Gaffney, D., 2014 Jun. Genetic background drives transcriptional variation in human induced plurip-otent stem cells. PLoS Genet. 10 (6), e1004432. PMCID: PMC4046971.

Schulz, H.L., Stoehr, H., White, K., van Driel, M.A., Hoyng, C.B., Cremers, F., et al., 2002 Mar 19. Genomic structure and assessment of the retinally expressed RFamide-related peptide gene in dominant cystoid macular dystrophy. Mol. Vis. 8, 67-71.

Siegert, S., Cabuy, E., Scherf, B.G., Kohler, H., Panda, S., Le, Y.-Z., et al., 2012 Mar. Transcriptional code and disease map for adult retinal cell types. Nat. Neurosci. 15 (3), 487-495. S1-2.

Skeie, J.M., Mahajan, V.B., 2014 Jul 24. Proteomic landscape of the human choroid-retinal pigment epithelial complex. JAMA Ophthalmol. (in press).

Stone, E.M., Luo, X., Hóon, E., Lam, B.L., Weleber, R.G., Halder, J.A., et al., 2011 Dec. Autosomal recessive retinitis pigmentosa caused by mutations in the MAK gene. Investigative Ophthalmol. Vis. Sci. 52 (13), 9665-9673. PMCID: PMC3341124.

Trapnell, C., Hendrickson, D.G., Sauvageau, M., Goff, L., Rinn, J.L., Pachter, L., 2013. Differential analysis of gene regulation at transcript resolution with RNA-seq. Nat. Biotechnol. 31 (1), 46-53.

Trapnell, C., Roberts, A., Goff, L., Pertea, G., Kim, D., Kelley, D.R., et al., 2012. Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nat. Protoc. 7 (3), 562-578. PMCID: PMC3334321.

Trapnell, C., Williams, B.A., Pertea, G., Mortazavi, A., Kwan, G., van Baren, M.J., et al., 2010 May. Transcript assembly and quantification by RNA-Seq reveals unan-notated transcripts and isoform switching during cell differentiation. Nat. Biotechnol. 28 (5), 511-515. PMCID: PMC3146043.

Tucker, B.A., Scheetz, T.E., Mullins, R.F., DeLuca, A.P., Hoffmann, J.M., Johnston, R.M., et al., 2011 Aug 23. Exome sequencing and analysis of induced pluripotent stem cells identify the cilia-related gene male germ cell-associated kinase (MAK) as a cause of retinitis pigmentosa. Proc. Natl. Acad. Sci. 108 (34), E569-E576. PMCID: PMC3161526.

van Soest, S.S., de Wit, G.M.J., Essing, A.H.W., Brink ten, J.B., Kamphuis, W., de Jong, P.T.V.M., et al., 2007. Comparison of human retinal pigment epithelium gene expression in macula and periphery highlights potential topographic differences in Bruch's membrane. Mol. Vis. 13,1608-1617.

Wagner, A.H., Anand, V.N., Wang, W.-H., Chatterton, J.E., Sun, D., Shepard, A.R., et al.,

2013 Jun. Exon-level expression profiling of ocular tissues. Exp. Eye Res. 111, 105-111. PMCID: PMC3664108.

Whitmore, S.S., Braun, T.A., Skeie, J.M., Haas, C.M., Sohn, E.H., Stone, E.M., et al., 2013. Altered gene expression in dry age-related macular degeneration suggests early loss of choroidal endothelial cells. Mol. Vis. 19, 2274-2297. PMCID: PMC3834599.

Zhang, L., Wahlin, K., Li, Y., Masuda, T., Yang, Z., Zack, D.J., et al., 2013. R1T2, a neuron-specific small guanosine triphosphatase, is expressed in retinal neuronal cells and its promoter is modulated by the POU4 transcription factors. Mol. Vis. 19, 1371-1386. PMCID: PMC3692409.