Ocular Phenotyping of Knockout Mice Identifies Genes Associated With Late Adult Retinal Phenotypes

Abstract

Purpose

Analyze phenotypic data from knockout mice with late-adult retinal pathologic phenotypes to identify genes associated with development of adult-onset retinal diseases.

Methods

The International Mouse Phenotyping Consortium (IMPC) database was queried for genes associated with abnormal retinal phenotypes in the late-adult knockout mouse pipeline (49–80 weeks postnatal age). We identified human orthologs and performed protein–protein analysis and biological pathways analysis with known inherited retinal disease (IRD) and age-related macular degeneration (AMD) genes using Search Tool for the Retrieval of Interacting Genes/Proteins (STRING), PLatform for Analysis of single cell Eye in a Disk (PLAE), Protein Analysis Through Evolutionary Relationships (PANTHER), and Kyoto Encyclopedia of Genes and Genomes (KEGG).

Results

Screening of 587 late-adult mouse genes yielded 12 with abnormal retinal phenotypes, which corresponded to 20 human orthologs. Three of the 12 mouse genes and two of the 20 human orthologs were previously implicated in retinal pathology or physiology in a literature review. Although all of the genes demonstrated retinal pathology when deleted from the mouse genome, most do not have established roles in human retinal disease. Furthermore, human protein–protein analysis and biological pathway analysis yielded only a few relationships between the candidate gene list and that of known IRD and AMD genes, suggesting they may represent novel retinal functions.

Conclusions

We identified 12 mouse genes with significant late-adult abnormal retinal pathology, eight of which have not been previously implicated in either mouse or human retinal physiology or pathology. These serve as novel retinal disease gene candidates for late-onset retinal disease.

The genetic basis of retinal biology and photoreceptor function is a fast-moving field of active research. Despite a large body of literature arising from many scientific groups around the globe over the past several decades, the biology of photoreceptors and other retinal cells is incompletely understood. This field of research is relevant to single gene disorders of the retina in human populations. Inherited retinal diseases (IRDs) are a heterogeneous group of disorders, both phenotypically and genetically, which result in degeneration of the retina and vision impairment. The estimated prevalence of IRDs is around 1 in every 2000 to 3000 individuals,^1,2 and they are predominantly monogenic in etiology, causing photoreceptor degeneration or dysfunction. There are currently around 300 genes that have been associated with IRDs.^3,4 Genetic testing can provide a definitive diagnosis and help provide prognostic data for the patient as well as for at-risk family members. Multiple next-generation sequencing (NGS) panels are currently available, including comprehensive and disease-specific panels⁵; however, panel-based NGS still fails to diagnose 30% to 40% of IRDs.^6,7

An incomplete understanding of genetics and the heterogeneity of clinical presentation make IRDs challenging to diagnose, especially because some individuals may have slow-onset disease or a clinical picture that may be confused with other diagnoses such as age-related macular degeneration (AMD) or resolved ocular inflammation. Therefore, the identification of new genes or biological pathways specifically implicated in late-onset IRD is necessary to distinguish phenotypically overlapping entities from one another. Elucidating these genetic underpinnings not only enhances our biological understanding but also improves the accuracy of clinical diagnosis.

Formed in 2011, the International Mouse Phenotyping Consortium (IMPC) is comprised of an international group of research institutions that produce single-gene knockout mouse lines using gene-targeted embryonic stem cells or, more recently, CRISPR/Cas9 methodology.^8,9 The Consortium aims to complete the knockout resource for the human orthologous genome and provide broad-based phenotyping for each knockout line produced. At least 14 mice (a minimum of seven males and seven females), along with matched wild-type controls, are produced for each knockout line. When viable, homozygous mice are phenotyped; however, when homozygous mice are not viable, heterozygous mice are evaluated. All viable mice undergo a uniform regimented battery of phenotyping tests from 4 to 16 weeks postnatal age in the “early adult pipeline,” after which terminal procedures are performed. In a pilot project, 587 mouse strains were also phenotyped between 49 and 80 weeks in the “late-adult pipeline” to identify phenotypes present in the later adult period.

These late adult phenotypes are of great interest for retinal pathologies such as AMD or slow-onset IRDs, which only appear in the late-adult time frame and would not be observed during early adult phenotyping. Here, we report on 12 genes from knockout mouse phenotyping identified through a systematic query of the IMPC database that are associated with late-onset retinal pathology.

Methods

Mouse Background

Each IMPC center adheres to strict ethical review and licensing procedures, following both its regional regulatory guidelines as well as the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines, a checklist of items designed to increase transparency and provide standardization.¹⁰ A minimum of seven male and seven female knockout mice are generated, with standardized procedures performed with regard to phenotyping and eye examination procedures, including an anterior slit-lamp exam, iris/pupil response, and a dilated exam with imaging of abnormal findings as capacity permits. Several different strategies have been employed in generating knockout mice, with increasing use of CRISPR/Cas9. Other strategies that have been used include the “knockout-first” allele design using site-specific recombinases Cre and Flp, as well as the VelociGene null allele design to delete protein coding sequences of a target gene. Phenotyping is conducted at each center, and procedures are assessed routinely to ensure animal welfare.

Statistical Analysis

Statistical analyses were conducted by the IMPC using the OpenStats toolkit in R (R Foundation for Statistical Computing, Vienna, Austria) to provide appropriate data analysis depending on the data type. Fisher's exact tests were used for categorical data, and abnormal phenotypes for a specific trait were compared collectively to the normal phenotype. The IMPC analysis does not take into account whether an abnormality is fully or incompletely penetrant, and the same statistical weight is given whether the abnormal phenotype is present in one or both eyes. Statistical significance was defined as P < 0.0001. If the phenotype was determined to be significant, it was further determined if this effect was the same and present in both sexes or in only one sex. In cases where the center was not able to determine if an abnormal phenotype was present, the mouse was excluded from the analysis.

Data Query

The IMPC database was queried on Data Release 19.0. As of May 9, 2023, it contained 8491 phenotyped genes, 9138 phenotyped mutant lines, and 102,805 phenotype calls, including both early- and late-onset mutant lines. Of these, 587 mutant lines were phenotyped in the late-adult pipeline. Mouse lines with an abnormal retinal phenotype only in the late-adult pipeline were included in this study, whereas knockout lines with abnormal phenotypes in the early and late adult stages were excluded. After an initial automated search, each gene was manually reviewed to confirm an abnormal phenotype was identified. The Retinal Information Network (RetNet) database, which tracks genes causing inherited retinal diseases, was also queried to provide a list of known genes involved in retinal diseases. Additionally, a search was done through the National Human Genome Research Institute and European Bioinformatic Institute (NHGRI-EBI) genome-wide association study (GWAS) catalog to identify all known genes associated with AMD.

Literature Search

A literature search was conducted on PubMed using the gene name as well as the search term “eye” and separately with the search term “retina,” and each article was reviewed for any reported association of the gene being involved in retinal pathophysiology.

Bioinformatic Analysis

Human orthologs of all knockout mouse genes were identified using the GeneCards human gene database by entering the mouse gene and [ortholog] as the search term. The human ortholog was confirmed to match with the mouse gene if the identical name was found in the ortholog section of the queried gene. The human ortholog found set was then used for further analyses (Supplemental Materials). Protein Analysis Through Evolutionary Relationships (PANTHER) and Kyoto Encyclopedia of Genes and Genomes (KEGG) were used to identify biological pathways associated with the human ortholog found set. Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) was used to identify protein–protein interactions between the human ortholog found set and the human RetNet and GWAS gene lists using a confidence cutoff of 0.7. The STRING network was expanded by adding secondary interacting proteins that were not initially identified with a selectivity of interactors of 0.7 to look for additional protein–protein interactions. PLatform for Analysis of single cell Eye in a Disk (PLAE), version 0.94, was used to evaluate gene expression of the human ortholog found set by retina cell type.¹¹

Histopathology

Complete necropsy was done, and all abnormal phenotypic findings were recorded using standardized IMPC Gross Pathology ontology.¹² Tissue samples were then fixed, typically in 10% neutral buffered formalin, sectioned at 5 µm, stained with hematoxylin and eosin, and examined by a veterinary pathologist. Mild to moderate focal retinal dysplasia was considered incidental and attributed to the C57BL/6N substrain, which is known to be homozygous for the rd8 mutation in the Crb1 gene, causing background retinal abnormalities.^13,14

Fundus Photography

At imaging-capable centers, fundus photography was obtained to document fundus abnormalities. In general, mice were anesthetized with intraperitoneal ketamine/midazolam, and their eyes were dilated with topical tropicamide or topical tropicamide and phenylephrine, then lubricated with methylcellulose-containing artificial tears prior to fundus photographs being taken with a MICRON III or IV retinal imaging system (Phoenix-Micron, Inc., Bend, OR, USA). Techniques and protocols varied slightly among imaging centers. Although most centers performed screening using indirect ophthalmoscopy, the Baylor College of Medicine site also performed optical coherence tomography screening.

Results

In total, 12 knockout mouse lines (Table) in the IMPC database had retinal pathology identified only in the late-adult pipeline, where phenotyping was extended from 49 to 80 weeks depending on the center in which the mice were examined. These were Abhd17b, Atp8b1, Calcoco1, Gnb4, Mt1, Nt5dc1, Sorl1, Spink12, Tat, Thsd7a, Tmem51, and Trim39.

Table.

Knockout Mouse Genes With Abnormal Late Retinal Phenotype

Mouse Gene Symbol	Mouse Gene Name	Mouse Genome Informatics ID	Human Ortholog Symbol	Phenotype Summary
Abhd17b	Abhydrolase domain containing 17B	1917816	ABHD17B	Abnormal retina morphology
Atp8b1	ATPase, class I, type 8B, member 1	1859665	ATP8B1	Abnormal retinal blood vessel morphology
Calcoco1	Calcium binding and coiled coil domain 1	1914738	CALCOCO1	Abnormal retina morphology
Gnb4	Guanine nucleotide binding protein (G protein), beta 4	104581	GNB4	Abnormal retina morphology
Mt1	Metallothionein 1	97171	MT1A, MT1B, MT1E, MT1F, MT1G, MT1H, MT1HL1, MT1M, MT1X, MT2A	Abnormal retina inner nuclear layer morphology
Nt5dc1	5′-Nucleotidase domain containing 1	2442446	NT5DC1	Abnormal retina morphology
Sorl1	Sortilin-related receptor, LDLR class A repeats-containing	1202296	SORL1	Abnormal retina morphology
Spink12	Serine peptidase inhibitor, Kazal type 12	1925492	None	Abnormal retina morphology
Tat	Tyrosine aminotransferase	98487	TAT	Abnormal retina morphology, abnormal retina blood vessel morphology
Thsd7a	Thrombospondin, type I, domain containing 7A	2685683	THSD7A	Abnormal retina morphology
Tmem51	Transmembrane protein 51	2384874	TMEM51	Abnormal retina morphology
Trim39	Tripartite motif-containing 39	1890659	TRIM39	Abnormal retina morphology

The identified retinal phenotypes were abnormal retina morphology, abnormal retina inner nuclear layer morphology, and abnormal retina blood vessel/vasculature morphology. Only one zygosity was examined in these genes, with all lines being homozygous knockouts. Notably, many of these genes exhibited sexual dimorphism, with the retinal phenotype achieving significance in either males or females only. The knockout genes that achieved significance in females only were Abhd17b, Calcoco1, Mt1, and Sorl1. The knockout genes that achieved significance in males only were Tat, Thsd7a, Tmem51, Trim39, Nt5dc1, and Spink12, which are all on autosomal chromosomes.

Examples of fundus photographs depicting abnormal retinal morphology (Atp8b1 and Tat) are shown in Figure 1, as well as an example of abnormal retinal histology in Figure 2. As part of this study, further careful examination of Atp8b1 was done retrospectively with the phenotyping site, revealing a subtle abnormal phenotype in the early adult stage that was initially missed. It is unclear if subtle abnormalities may have been identified in the early stage in any of the other 11 genes with careful retrospective examination that would be missed on prospective screening; given this possibility, the gene was included in this study.

Figure 1.

Fundus photograph examples of knockout lines with retinal abnormalities. (A) Tat homozygote at 75 weeks with pigmentary changes throughout the fundus. (B) Atp8b1 homozygote at 58 weeks with numerous yellow spots and a lesion below the optic nerve (arrow). (C) Wild-type C57BL/6NCrl at 58 weeks for comparison. Some yellow spots are seen (arrow) consistent with background rd8 mutation. (D) Wild-type C57BL/6NCrl at 9 weeks.

Figure 2.

Histological example of knockout line with retinal abnormalities. (A) Atp8b1 homozygote knockout at 59 weeks demonstrating numerous cells present in the outer plexiform layer (yellow arrows). (B) Wild-type at 59 weeks showing normal anatomy of the retina.

The set of 12 knockout mouse genes was converted to the corresponding 20 human orthologous genes, with one mouse gene (Mt1) having 10 human orthologs and another gene (Spink12) lacking a known human ortholog. A literature search of these 20 human orthologs revealed that two genes (MT2Aand SORL1) have been previously associated with human retinal physiology or pathology,^15,16 and three genes (Sorl1, Tat, and Gnb4) have been implicated in mouse retinal physiology or pathology.^17–19 Notably, polymorphisms in MT2A have been associated with AMD in Northern Spanish patients.¹⁶ Eight mouse genes corresponding to seven genes in the human ortholog found set (Spink12 has no known human ortholog) have not been previously associated with human retinal pathophysiology, nor have their corresponding mouse genes been linked to any significant retinal processes. STRING analysis with a selectivity of interactors of 0.7 was used to identify potential protein-protein interactions within the set of 20 human orthologs. No interactions were predicted (Supplementary Fig. S1), except for the human orthologs of Mt1 as a self-interacting protein (Fig. 3A).

Figure 3.

STRING analysis of protein–protein interactions. (A) Interaction between MT1 candidate gene orthologs (purple). (B) Interaction between GNB4 (1) (purple) with RetNet database genes (green). (C) Interaction between TAT (2) and SORL1 (3) candidate genes (purple) with AMD GWAS genes (blue). Organism set to Homo sapiens (settings: network type = full STRING network, selectivity of interactors = 0.7). Nodes corresponding to genes for which there were PubMed reports of implication in retinal physiology or pathology in human, mouse or both are circled respectively in gray, blue, or red.

Each gene and its human ortholog were then queried in the PLAE ocular meta-atlas, a resource for single-cell transcriptomics from ocular cell types in several species. Of the 12 knockout mouse genes, eight (67%) were expressed in the retina (Abhd17b, Calcoco1, Gnb4, Mt1, Sorl1, Thsd7a, Tmem51, and Trim39). Seven (58%) were expressed in the retinal pigmented epithelium (RPE; Abhd17b, Calcoco1, Mt1, Sorl1, Thsd7a, Tmem51, and Trim39); all genes that were expressed in the RPE were also expressed in retina except Gnb4, which was only expressed in the retina and not the RPE. Of the 20 human orthologs, 14 (70%) were expressed in the retina and RPE (Calcoco1, Gnb4, Mt1a, Mt1e, Mt1f, Mt1g, Mt1h, Mt1h, Mt1x, Mt2a, Nt5dc1, Sorl1, Thsd7a, and Tmem51), and the remaining six were not expressed in either tissue. The expression of three genes (Atp8b1, Tat, and Spink12) was not detected in either mouse or human retina or RPE.

The 20 human orthologs were compared with 264 genes known to cause inherited retinal disease, including macular dystrophies and retinitis pigmentosa, using the RetNet database and analyzed using STRING. No large clusters of protein interactions were identified between the RetNet dataset and the human ortholog found set (Supplementary Fig. S2), even with an additional 20 closest interactors. Notably, even among the RetNet genes themselves, many were singletons and did not have known interactors within the database. There was one knockout gene that did interact with the RetNet dataset (Fig. 3B): GNB4, a constituent of heterotrimeric G proteins.

To determine if the genes identified in the late-adult pipeline may be involved in retinal diseases of aging, we used similar bioinformatic strategies to assess relationships with loci implicated in AMD. A list of 123 unique AMD-associated genes was obtained through the NHGRI-EBI GWAS catalog using AMD as the “trait.” STRING analysis was used to look for protein–protein interactions between the 20 human orthologs and these 123 GWAS AMD genes (Supplementary Fig. S3). No large clusters of protein interactions were identified, including when an additional 10 interactors were added. Two knockout genes interacted with the GWAS AMD genes: sortilin-related receptor 1 (SORL1), a gene coding for a receptor involved in intracellular sorting of protein which is primarily associated with Alzheimer disease,²⁰ and tyrosine aminotransferase (TAT), which is involved in the tyrosine breakdown pathway (Fig. 3C).

To determine which molecular pathways were involved in the 20 genes in the human ortholog found set, we used the PANTHER and KEGG bioinformatics tools. PANTHER analysis revealed that many of the genes had no assigned category, whereas the remaining genes were involved in a wide array of pathways, such as the PI3 kinase pathway and Wnt signaling pathway (Fig. 4). Notably, of the eight novel mouse genes corresponding to seven human orthologs with no prior retinal disease association, none had annotations in PANTHER for known pathways. An additional literature review was conducted to evaluate known functions or pathways of these genes. Atp8b1 is involved in the transfer of phospholipids, with mutations known to cause congenital cholestasis and hearing loss.²¹ Other functions of the novel genes include regulation of N-Ras palmitate turnover (Abhd17b),²² autophagy of the endoplasmic reticulum and Golgi (Calcoco1),²³ and neuronal apoptosis (Trim39).²⁴ Some genes have also been implicated in various disease processes, such as chronic obstructive pulmonary disease (Nt5dc1),²⁵ membranous nephropathy,²⁶ and laryngeal squamous cell carcinoma.²⁷

Figure 4.

Molecular pathways of 20 candidate genes using the PANTHER tool, with each slice representing the proportion of the specific molecular pathway among the 20 genes.

KEGG Mapper was also used to identify molecular pathways in which the knockout genes are involved. A wide variety of pathways were identified in which various genes are known to participate, but no single pathway included more than one gene from the human ortholog found set. When KEGG Mapper was used to assess the human ortholog found set in the context of known IRD genes from RetNet, several overlapping pathways were identified: apelin signaling, cholinergic synapse, glutamatergic synapse, GABAergic synapse, pathways in cancer, Ras signaling pathway, chemokine signaling pathway, Kaposi sarcoma-associated herpesvirus infection, retrograde endocannabinoid signaling, PI3K/Akt pathway, and cytomegalovirus infection.

Discussion

The genetics of retinal biology is a complex area of research. As a result of this complexity, our understanding of the genetic causes of retinal pathology is incomplete, particularly in slow-onset retinal diseases and for diseases of aging. A large number of unique variants and alleles are implicated in retinal diseases. This is further complicated by heterogeneous phenotypes, diverse inheritance patterns, and overlapping causative genes.²⁸ An estimated one out of three people worldwide carries at least one pathologic mutation in an established retinal disease gene.^29,30 Identification of the causative genes involved is crucial for prognostication and to advance gene-specific therapies, as demonstrated by the success of gene therapy for RPE65 Leber's congenital amaurosis.³¹

Late-onset IRDs present an additional challenge, as non-specific or heterogeneous visual symptoms and clinical findings may lead to delayed or incorrect diagnoses and cause confusion with more common diagnoses such as AMD. For example, late-onset retinal degeneration is an autosomal dominant IRD that has several features, including subretinal deposits and choroidal neovascularization, that can mimic AMD.

In contrast, AMD is the most common cause of irreversible blindness in the developed world, and it is projected that close to 300 million people worldwide will be affected by 2040.³² In contrast to monogenic IRDs, AMD has a multifactorial etiology, including age, genetics, and environmental factors. Genes associated with members of the complement cascade (C2, C3, CFB, and CFH), as well as extracellular matrix regulation and mitochondrial oxidative stress, have been implicated in AMD pathogenesis through GWAS studies.³³ This has led to the exploration and approval of recent AMD drugs avacincaptad pegol (iZERVAY; Iveric Bio, Parsippany, NJ, USA) and pegcetacoplan (SYFOVRE, Apellis Pharmaceuticals, Waltham, MA, USA), both of which inhibit the complement system and slow progression of macular atrophy. These examples emphasize how genetic understanding of retinal pathobiology leads to development of therapeutic medications that prevent vision loss.

The majority of the genes identified in this study were expressed in the retina and retinal pigment epithelium. However, a few of the genes were not expressed in either of these tissues. This suggests that some gene mutations may have downstream effects in other tissues. For example, TAT is primarily expressed in the liver, but a deficiency in TAT leads to tyrosinemia, which has systemic and ocular effects due to the accumulation of tyrosine and its byproducts. However, there are no known retinal findings associated with tyrosinemia reported in the literature.

This study is limited in its ability to identify potential biological pathways implicated in AMD and late-onset IRDs. The IMPC late-adult pipeline comprises only about 5% of the knockout mice, and the criteria for selected genes varied across each IMPC center, sometimes based on prior GWAS data, early adult data, or other interests unrelated to the eye. This is relevant, as most IRDs are non-syndromic and are isolated to retina-specific genes.²⁸ If more data become available, they may reveal yet to be discovered pathways relevant to the retina. Furthermore, Although the IMPC phenotyping process assesses retinal morphology, confirmatory molecular biological testing in retina or any other tissue, as well as routine histopathology, electroretinography, and optomotor testing to measure retinal/visual function, are beyond the scope of the Consortium.

An additional limitation is the confounding influence of the homozygous rd8 alleles (Crb1 gene) in the background of these knockout mice (C57BL/6N background), which causes its own slow retinal degeneration. This rd8 mutation may result in late-adult retinal phenotypic findings hits that could be a result of interaction between the Crb1 gene and the targeted deletion.

This study demonstrated relationships between known AMD-associated genes and established IRD genes for several of the 20 human orthologs in the found set. Although mouse models have been useful in understanding the pathophysiology of retinal degenerations such as in AMD,^34–37 there are important species-specific differences between mouse and human eye biology. Unlike in the human retina, the mouse retina lacks a macula, and there are species differences in photoreceptor density in the central retina.³⁸ The pathologies identified in this study were also primarily retinal in nature, whereas the pathogenesis of AMD appears to originate from the RPE–Bruch's membrane complex. Also, given that C57BL/6 mice have a life expectancy of approximately 120 weeks, the 49- to 80-week “late adult” stage in mice used in this study corresponds to an age of roughly 32 to 52 years in humans, but human AMD typically begins at around age 50 years and becomes severe around age 70. Therefore, the age interval studied may not be ideal to detect AMD-related genes. Additionally, vivarium lighting conditions in mouse facilities are generally lower than those in human daily life, leading to a lower cumulative exposure to damaging ultraviolet light. Furthermore, the absolute age of photoreceptors in aged mice and aged humans differs significantly, which may impact the genesis of AMD. There may also be differences in the aging process in the retina of mice versus primates, such as the degree of inflammation, photoreceptor loss and decrease in ATP production with age³⁹; this suggests that genetic mutations may play differing roles in the pathophysiology of aging in the retina between species.

In summary, the 12 mouse genes (corresponding to 20 human orthologs) with significant late-adult abnormal retinal pathology may prove to be relevant in human retinal disease. Eight of these genes have not been previously implicated in retinal physiology or pathology in either mice or humans, making them exciting avenues for further research. The pathways in which these candidate genes are involved may provide new insights into the pathophysiology of slow-onset retinal diseases and help inform the direction of future therapies. These pathways may include the beta-adrenergic receptor,⁴⁰ dopaminergic receptor,⁴¹ and histamine receptor⁴² mediated signaling, as demonstrated in this study. Additionally, because several of the novel genes have limited data regarding their biological function and pathways, they may represent important yet to be-discovered pathways and functions in the retina. Given that a significant proportion of presumed IRD patients do not receive a molecular diagnosis using currently available tools, we advocate including these candidate genes in IRD genetic testing panels to improve diagnostic rates.

Appendix. Appendix. The International Mouse Phenotyping Consortium (IMPC)

The following authors are IMPC members who contributed to this study: Elif Acar, Antonio Aguilar-Pimentel, Elizabeth Axton, Shinya Ayabe, Abdel Ayadi, Lore Becker, Alexandr Bezginov, Marie-Christine Birling, Jin Woong Bok, Vivian Bradaschia, Julia Calzada-Wack, Adam Caulder, Linda Chan, Audrey E. Christiansen, Dave Clary, James Cleak, Gemma Codner, Amie Creighton, Patricia da Silva Buttkus, Nathalia Dragano, Kyle Duffin, Qing Fan-Lan, Martin Fray, Tamio Furuse, Xiang Gao, Wendy Gardiner, Lillian Garrett, Angelina Gaspero, Marina Gertsenstein, Isabelle Goncalves, Leslie Goodwin, Kristin Nicole Grimsrud, Alain Guimond, Sabine Hölter-Koch, Joanne H. Hsu, Mizuho Iwama, Lois Kelsey, Chang-Hoon Kim, Kyoungmi Kim, Markus Kraiger, Tatsuya Kushida, Denise Lanza, Valerie Laurin, Sophie Leblanc, Ho Lee, Christoph Lengger, Stefanie Leuchtenberger, Lauri G. Lintott, Jimmy Liu, Yang Liu, Isabel Lorenzo, Aline Lux-Simonet, Susan Marschall, Matthew McKay, Matthew McKenzie, David Miller, Ikuo Miura, Saori Mizuno, Toshiaki Nakashiba, Ki Taek Nam, Clare Norris, Yuichi Obata, Manuela Österreicher, Kristina Palmer, Guillaume Pavlovic, Kevin Peterson, Dawei Qu, Tara Rasmussen, Birgit Rathkolb, Kyle Roberton, Adrian Sanz Moreno, Claudia Seisenberger, Audrie Seluke, Xueyuan Shang, Hirotoshi Shibuya, Gillian Sleep, Nadine Spielmann, Claudia Stöger, Toyoyuki Takada, Nobuhiko Tanaka, Lydia Teboul, Todd Tolentino, Igor Vukobradovic, Hongyu Wang, Brandon Willis, Joshua Wood, Xia Xue Catherine Xu, Ikuko Yamada, Jing Zhao.