Evaluation of the Myocilin (MYOC) Glaucoma Gene in Monkey and Human Steroid-Induced Ocular Hypertension

Abstract

purpose. Glucocorticoid-induced ocular hypertension (the steroid response) may result in optic nerve damage that very closely mimics the pathologic course of primary open angle glaucoma (POAG). In addition, patients with glaucoma and their relatives are much more likely to exhibit the steroid response than unaffected individuals, suggesting a potential link between the steroid response and POAG. Recently, the expression of a gene (MYOC) in the trabecular meshwork was shown to be steroid-induced. MYOC variations thought to be disease-causing also were found in 3% to 5% of POAG cases. The purpose of this study was to determine whether some variations in MYOC might be involved in steroid-induced ocular hypertension.

methods. Seventy human steroid responders and 114 control subjects were screened for variations in the coding sequence and promoter of MYOC. Also, topical doses of dexamethasone (DEX) were administered to cynomolgus monkeys to determine their steroid responsiveness, and the MYOC orthologue was cloned from the cynomolgus monkey.

results. Overall, 109 instances of 20 different sequence variations were identified in the human myocilin gene. However, only four of these (each observed in a single individual) met the study criteria for a possible phenotype-altering variation. Three of these were present in steroid responders and one in a control patient, a distribution that was not statistically significant (P = 0.3). In addition, the allele frequency of a closely flanking marker was compared between the steroid responders and the control subjects, and no evidence for linkage disequilibrium was observed. Reproducible and reversible ocular hypertension was induced in approximately 40% of the monkeys treated with DEX, similar to that seen in man. Ten monkeys were screened for MYOC mutations with single-strand conformation polymorphism (SSCP) analysis. Overall, 37 instances of 13 different sequence variations were observed. Four of these changes met the study criteria for a possible phenotype-altering variation, and these were equally distributed between responder and nonresponder monkeys.

conclusions. This study identified no statistically significant evidence for a link between MYOC mutations and steroid-induced ocular hypertension.

Glucocorticoid-induced ocular hypertension (the steroid response) has long been associated with the pathophysiology of primary open-angle glaucoma (POAG). Several clinical observations suggest that there is a link between glucocorticoids and glaucoma. A subset of the general population (steroid responders), experiences a significant elevation of intraocular pressure (IOP) in response to glucocorticoid administration. ^{1 2 3 4 5} Extended periods of steroid-induced ocular hypertension result in glaucomatous optic atrophy and visual field loss that persist after the drug therapy is discontinued and the IOP has returned to normal. There is also marked similarity between the aqueous outflow obstruction, optic nerve cupping, and visual field deficits associated with glucocorticoid-induced glaucoma and those found in POAG. ^{3 4 5} In addition, the steroid response is much more common in patients with POAG than in the unaffected population, further supporting a connection between steroids and glaucoma. ^{6 7}  

Since Armaly ⁸ and Becker and Chevrette ⁹ first characterized the steroid response in the 1960s, many investigators have studied the nature of this phenomenon in search of clues to the pathophysiology of POAG. Early studies suggested that the steroid response may be determined by the autosomal recessive inheritance of a single gene. Subsequent studies have confirmed the heritability of the steroid response and its connection with glaucoma ^{10 11} ; however, these studies have also suggested that the genetics of this phenomenon are more complex than was initially suspected. ^{12 13}  

There have been attempts to reproduce this steroid-induced ocular hypertension in several different animal models. The topical ocular administration of glucocorticoids leads to ocular hypertension in rabbits ^{14 15} but is often accompanied by systemic effects. Glucocorticoid-induced ocular hypertension also has been reported in cats. ^{16 17} An initial attempt to test steroid responsiveness in a nonhuman primate model was not successful. ¹⁸  

More recently, the genetic basis and cell biology of the steroid response have been explored using tissue and organ culture models. Cultured trabecular meshwork (TM) cells have been treated with glucocorticoids to mimic conditions of steroid-induced ocular hypertension. Using this system, several steroid-induced changes in TM cells have been identified, including increased deposition of several components of the extracellular matrix ^{19 20 21} and the rearrangement of the TM cytoskeleton in which actin microfilaments become cross linked. ^{22 23} Studies by Polansky et al. ²⁴ and Nguyen et al. ²⁵ have demonstrated that the expression of the protein myocilin (previously known as TIGR and GLC1A) in cultured TM cells is greatly enhanced by treatment with glucocorticoids. ^{24 25} This observation led to the hypothesis that increased expression of this protein is a key step in steroid-induced ocular hypertension. Later reports demonstrated that mutations in the myocilin gene (MYOC) are associated with 3% to 5% of POAG cases. ^{26 27 28 29}  

In this study we investigated the possible association between myocilin expression and the steroid response. Human steroid responders were screened for variations in myocilin gene sequences. In addition, a novel primate model of glucocorticoid-induced ocular hypertension was developed. The steroid response in cynomolgus monkeys was characterized and compared with that of humans and other animal models. Finally, the monkey orthologue to the human myocilin gene was cloned, and both steroid-responding and nonresponding monkeys were screened for MYOC sequence variations. 

Materials and Methods

Generation of Steroid-Induced Ocular Hypertension in Monkeys

Eleven cynomolgus monkeys (9 female, 2 male; average age, 7.5 years; range, 6–12 years) were behaviorally trained for conscious IOP readings. Each monkey received topical ocular drops (10 μl) of 0.1% dexamethasone alcohol (DEX) three times per day for 28 days. IOP readings were taken every 3 to 4 days for 6 weeks (4 weeks during DEX treatment and 2 weeks after discontinuation of DEX administration) using an applanation pneumotonometer (Alcon, Fort Worth, TX). A topical ocular anesthetic (0.1% proparacaine) was administered before taking IOP measurements. Ocular examinations, recording of body weights, and blood tests for liver enzymes and latent viruses were routinely performed during the course of the studies. Steroid responsiveness was defined as a more than 5-mm Hg change in IOP from the pretreatment baseline IOP as previously reported by Becker. ² To test the reproducibility of steroid responsiveness, 10 of the monkeys from the original study were retested 6 months after the first study using the same protocol. One monkey was not eligible for retreatment due to illness at the time of the second study. All animals were managed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Patients

One hundred eighty-four individuals were studied: 70 unrelated steroid responders (45 from Australia and 25 from Iowa), 23 steroid nonresponders from Australia, and 91 control subjects from Iowa. All participants provided informed consent after the nature and consequences of the study were explained, in accordance with the Declaration of Helsinki. Steroid responders included patients who exhibited an elevation of IOP of more than 5 mm Hg after administration of glucocorticoid steroids (prednisolone acetate, DEX, prednisolone phosphate, fluorometholone, betamethasone, or oral prednisolone) for at least 4 weeks or who exhibited glaucomatous optic nerve damage after a prolonged course of oral or topical glucocorticoids. Steroid nonresponders had no elevation of IOP after topical administration of a potent topical glucocorticoid four times a day for at least 1 month. Twenty-four of the 25 Iowan steroid responders and 18 of the 45 Australian steroid responders exhibited glaucomatous optic neuropathy. Finally, the 91 normal control subjects were more than 40 years of age and had no personal or family history of glaucoma. The control subjects were not tested for the steroid response. 

Cloning of the Monkey Myocilin Gene

DNA was extracted from blood samples obtained from five steroid-responding and five non–steroid-responding cynomolgus monkeys. ³⁰ The exons and flanking DNA sequences of the monkey myocilin gene were amplified with the polymerase chain reaction (PCR) using primers specific for the human myocilin gene. Monkey myocilin gene fragments were amplified from 12.5 ng of monkey genomic DNA in a 30-μl reaction containing 4.5 μl 10× PCR buffer (100 mM Tris-HCl [pH 8.3]; 500 mM KCl; 15 mM MgCl₂); 300 μM each dCTP, dATP, dGTP, and dTTP; 9 picomoles of each primer; and 1 unit Taq DNA polymerase. The primers used in these reactions are listed in Table 1 . Samples were denatured for 5 minutes at 94°C and incubated in a DNA thermocycler for 35 cycles at the following temperatures: 94°C for 30 seconds, 55°C for 30 seconds, and 72°C for 30 seconds. PCR products were purified using a commercial kit (QiaQuick Spin kit; Qiagen, Chatsworth, CA) and were sequenced using automated sequencers (model 377; Applied Biosystems, Foster City, CA). Bidirectional sequencing was conducted using standard dideoxynucleotide chemistry. 

SSCP Mutation Detection

The human and monkey DNA samples were screened for MYOC sequence variations using a single-strand conformation polymorphism (SSCP) assay that included the entire coding sequence and the proximal promoter. Sequences upstream from the MYOC coding sequence were also screened. In the humans, 913 bp of upstream sequence was screened, whereas in the monkeys only 555 bp were evaluated, because the 171-bp segment of the monkey promoter that contains a short tandem repeat polymorphism (STRP) was excluded from the assay. Genomic DNA (12.5 ng) from each human subject and each cynomolgus monkey was PCR amplified in a 8.35-μl reaction containing 15 mM Tris-HCl (pH 8.3); 75 mM KCl; 2.25 mM MgCl₂; 300 μM of dCTP, dATP, dGTP, and dTTP; 2.5 picomoles of each primer; and 0.25 units Taq DNA polymerase. Primers used in this assay are shown in Table 1 . Samples were denatured for 5 minutes at 94°C and incubated in a DNA thermocycler for 35 cycles at the following temperatures: 94°C for 30 seconds, 55°C for 30 seconds, and 72°C for 30 seconds. After amplification, 5 μl of stop solution (95% formamide, 10 mM NaOH, 0.05% bromophenol blue, and 0.05% xylene cyanol) was added to each PCR product. The products were then denatured at 94°C for 3 minutes, and 4 μl of each sample was loaded onto 6% (37.5:1 bis) polyacrylamide, 0.5× TBE (44.5 mM Tris, 44.5 mM boric acid, and 0.5 mM EDTA), nondenaturing gels. Products were electrophoresed at room temperature for 3 hours at 25 W. DNA bands were visualized with silver nitrate staining. ³¹ PCR products that produced aberrant SSCP bands were sequenced to reveal the underlying DNA variations. 

Evaluation of MYOC Variations

Our criteria for judging a variation to be potentially involved in the steroid response phenotype (and therefore to be included in the statistical comparisons between steroid responders and control subjects) included alteration of either the charge, size or polarity of the predicted amino acid sequence; alteration of a consensus splice site sequence; or alteration of a putative promoter–enhancer element. Putative enhancer and promoter elements were identified in the sequence upstream of the monkey and human myocilin genes, by using the network application program TESS (http://www.cbil.upenn.edu/tess/index.html; provided free of charge by the University of Pennsylvania, Philadelphia) and the transcription factor binding site data set TRANSFAC v3.2. All comparisons were evaluated for significance using Fisher’s exact test, and all probabilities are two-tailed. This study was designed to test the hypothesis that mutations in the coding region or proximal promoter of the myocilin gene were responsible for a large fraction of the steroid response in humans. It had a power of greater than 95% to detect a statistically significant difference between steroid responders and control subjects if PCR-detectable variations in the myocilin gene were responsible for at least 30% of the steroid response in the patient populations we studied. This power calculation was performed as described by Rosner ³² with the assumption that 35% of the control individuals would have been steroid responders if they had been challenged with the same amount of steroid as the steroid responding group. 

Linkage Disequilibrium Experiments

Human steroid responders, nonresponders, and normal control subjects were genotyped at the MY5 STRP located 341 bp upstream of the MYOC coding sequence. Using 12.5 ng of each patient’s DNA, the MY5 marker was PCR amplified (as described earlier), electrophoresed on 6% polyacrylamide, 1× TBE, and 7-M urea gels at 65 W for approximately 3 hours, and stained with silver nitrate. Marker alleles were identified by observing the electrophoretic mobility of the PCR products. Allele frequencies were compared using Fisher’s exact test. 

Results

Glucocorticoid-Induced Ocular Hypertension in Monkeys

Steroid-induced ocular hypertension developed in 5 of 11 cynomolgus monkeys (45%) treated with DEX for 4 weeks. The IOP elevation was progressive over the 4 weeks of treatment, and the average IOP increase for the steroid responder group was 10.6 ± 3.6 mm Hg after 28 days of DEX administration (Fig. 1 , Table 2 ). In contrast, the six nonresponder monkeys had IOP elevations of only 2.3 ± 1.7 mm Hg during this period. The DEX-induced ocular hypertension in the steroid responder group was reversible. The IOP gradually returned to baseline pressures two weeks after discontinuation of topical ocular DEX administration (Fig. 1) . 

The response of these monkeys to DEX-induced ocular hypertension was reproducible. In a second 4-week topical ocular DEX administration study, 10 of the same monkeys were retested for steroid responsiveness. All 4 of the initial steroid responders developed DEX-induced ocular hypertension in the second study with an average increase in IOP of 8.1 ± 2.6 mm Hg (Fig. 2 , Table 2 ). Similarly, all six of the monkeys that were categorized as steroid nonresponders in the initial study remained nonresponders in the subsequent study. The nonresponder group had IOP changes that averaged 2.3 ± 3.0 mm Hg. The reproducibility of the response to topical ocular DEX administration in individual monkeys in the two studies is shown in Table 2 . The reversibility of the steroid-induced ocular hypertension again was seen in this second study, with IOPs in the responder group returning to pretreatment baselines 2 weeks after discontinuation of DEX administration(Fig. 2) . 

Monkey Myocilin Gene Structure

An orthologue to the human myocilin gene was amplified by PCR from cynomolgus monkey genomic DNA using oligonucleotide primers specific to the human gene. Comparison of the monkey myocilin gene sequence with the previously cloned human and mouse myocilin genes revealed that the proximal promoter and coding sequence of the monkey gene is highly homologous to human and mouse myocilin genes (Fig. 3A ). In addition, the locations of the introns in the myocilin gene are completely conserved among human, monkey, and mouse (Fig. 3B) . Alignment of the proteins predicted by the myocilin genes also reveals a high degree of homology. The monkey and human genes predict proteins that are 97% identical (Fig. 3C) . All myocilin genes encode proteins that contain a putative secretory signal sequence, a leucine zipper domain, two hydrophobic regions, and several potential phosphorylation and glycosylation sites. 

The putative translation start site of the monkey myocilin gene is the same as that of the mouse and cow genes but differs from the human myocilin gene. The largest open reading frame in human MYOC cDNA begins with the first ATG codon and is in frame with a second ATG 42 bp downstream. The monkey, mouse, and cow myocilin genes do not possess the sequence corresponding to the first ATG in the human myocilin gene (Figs. 3A 3C) . 

Monkey MYOC and the Steroid Response

Ten cynomolgus monkeys (five steroid responders and five nonresponders) were screened for MYOC sequence variations using SSCP analysis. The entire coding sequence and 555 bp of the first 726-bp upstream sequence of the monkey MYOC promoter were screened. Thirty-seven instances of 13 different sequence variations were detected (Table 3) . Four of these sequence variations met our criteria for potential phenotype-altering sequence changes. However, these were found to be equally distributed between steroid responding and nonresponding animals. Specifically, one steroid responder was homozygous for a newly created TGT3 site and heterozygous for an Arg160Ser coding sequence change. A second responder was heterozygous for a newly created E1AF site. Similarly, one nonresponder was heterozygous for both a newly created TGT3 site and a Glu218Lys coding sequence change, whereas a second nonresponder was heterozygous only for the latter variation. 

Human MYOC and the Steroid Response

Seventy steroid responders, 23 steroid nonresponders, and 91 control subjects were screened for MYOC promoter and coding sequence variations. Overall, 109 instances of 20 different MYOC sequence changes were identified (Table 4) . Only four of these (each observed in a single individual) met our criteria for potential phenotype-altering sequence variations. One promoter sequence change (−153 bp T→C), which alters a putative Sp1 enhancer site from one consensus sequence (CCCAGCCTC) to another (CCAGCCCC), was present in a single steroid responder. Two coding sequence changes (ARG82CYS and GLN368STOP) were each identified in single steroid responders. Both of these changes have been described as glaucoma-causing changes. ^{27 28} A final amino acid–altering variation was present in a single control individual. The distribution of these four variations between steroid responders and control subjects was not statistically significant (P = 0.3). If MYOC sequence variations of this type were responsible for 30% or more of the steroid response phenotype in the population we examined, this study would have had greater than 95% power to detect it. 

Linkage Disequilibrium Experiments

Seventy steroid responders, 91 normal control subjects, and 23 steroid nonresponders were genotyped at the STRP MY5 located 341 bp upstream of the MYOC coding sequence. The allele frequencies of MY5 observed in the steroid responder group (Table 5) were not significantly different from the other groups (P > 0.05). 

Discussion

A novel monkey model of the human steroid response was developed as part of this study. The ocular hypertension observed in DEX-treated cynomolgus monkeys shares many characteristics with ocular hypertension in man. The fraction of monkeys (45%) that exhibited the steroid response was similar to the frequency of the steroid response in humans (40%). ^{1 2} In both organisms, a gradual and reversible increase in IOP is observed during administration of DEX. A previous study failed to show glucocorticoid-induced ocular hypertension in cynomolgus monkeys. ¹⁸ There are several possible explanations for this discrepancy. The earlier study measured the aqueous outflow facility of anesthetized monkeys at the very end of the glucocorticoid treatment regimen and found no differences between the average outflow facilities of the steroid-treated or placebo-treated groups. In the present study, we evaluated steroid responsiveness in monkeys in a manner similar to that used in the previously published human studies. The IOP was measured before, during, and after glucocorticoid administration in conscious animals, and the IOP response in monkeys was markedly similar to that seen in humans. In addition, the cynomolgus monkey myocilin gene was cloned and found to be highly homologous (97%) to the human myocilin gene. These similarities suggest that the cynomolgus monkey model system is well suited for studying the steroid response in humans and exploring the role of myocilin in this phenomenon. 

Both human and monkey steroid responders were examined for variations in the myocilin gene that might be associated with the steroid response. Screening human steroid responders for variations in MYOC identified no sequence changes present at a rate significantly higher than in the control populations. In fact, only 3 of the 70 steroid responders were found to harbor sequence variations in the myocilin gene that met our criteria for potential phenotype altering changes. Two of these variations (GLN368STOP and ARG82CYS) were each identified in steroid responders with positive personal or family histories of glaucoma. The GLN368STOP variation was found in a single steroid responder, who is also affected with POAG and has a family history of glaucoma. The ARG82CYS variation was identified in a single steroid responder who has no other signs of glaucoma. This patient’s brother and a son have mild glaucomatous disc changes and also harbor the ARG82CYS mutation. Both of these variations have been characterized as glaucoma-causing changes. ^{27 28} The third variation (153 bp C→T) was located in a putative enhancer site (Sp1) in the MYOC promoter. This variation, also identified in a single steroid responder, converts one Sp1 consensus site to another Sp1 consensus site. It is possible that in this single patient, the enhancer Sp1 has an increased affinity for the altered Sp1 site and, therefore, causes a pathologic increase in the expression of MYOC. However, the rarity of this variation (<2%), provides little support for the hypothesis that MYOC promoter variations are a common cause of steroid-induced hypertension. 

Only one promoter sequence variation (−83 C→T) was commonly observed. This variation does not alter any known promoter or enhancer binding sites and is found at similar frequencies in steroid responders, nonresponders (Table 4) , POAG patients, and control subjects. ²⁷ There is no evidence from this study or from the literature that the −83 C→T variation is associated with either the steroid response or POAG. There is no statistically significant evidence to link any of the identified MYOC variations with the steroid response. Further, the majority (96%) of the steroid responders examined in this study harbored no variations that met our criteria for potential involvement in the steroid response phenotype, clearly indicating that variations in the MYOC coding sequence and proximal promoter are not a common cause of this phenomenon. 

The relationship between the myocilin gene and the steroid response was further explored using the cynomolgus monkey model. The many similarities between the cynomolgus monkey and man, including ocular anatomy, genetic code, and characteristics of the steroid response, suggest that conclusions drawn from monkey studies accurately reflect features of the human steroid response. Screening steroid responder monkeys for MYOC variations identified 13 sequence variations (Table 3) . When these variations were analyzed individually or as a group, there was no statistically significant association with the steroid response. As observed in the human studies, MYOC mutations do not appear to be a common cause of the steroid response in monkeys. 

SSCP analysis of the coding and proximal promoter sequences of any gene would not be expected to identify all disease-causing sequence variations. Therefore, the myocilin gene was further evaluated by looking for linkage disequilibrium between the steroid response phenotype and a STRP closely flanking the myocilin gene. The human steroid responders and control subjects were genotyped at the STRP marker MY5, which is located 341 bp upstream of the MYOC coding sequence. Allele frequencies of all groups (Table 5) were not significantly different (P > 0.05). That there is no linkage disequilibrium suggests that a single ancestral MYOC variation (e.g., further upstream from the portion of the promoter we evaluated) is not a common cause of the steroid response in the populations that we examined. 

The normal function of myocilin and the mechanism by which mutations in MYOC cause glaucoma are unknown. However, because MYOC was originally isolated as a steroid-induced gene, it was a plausible hypothesis that variations in the MYOC coding sequences or proximal promoter could be involved in the development of steroid-induced glaucoma. The findings of this study strongly suggest that this is not the case. 

 Supported by National Institutes of Health Grant EY10564, the Carver Charitable Trust, the Grousbeck Family Foundation, the Glaucoma Research Foundation, and unrestricted grants to the University of Iowa from Research to Prevent Blindness.

 Submitted for publication April 25, 2000; revised August 25, 2000; accepted September 8, 2000.

 Commercial relationships policy: F, P (JHF, WLMA, EMS); P (VCS); E (AFC, MM, LT); N (JEC, GRS, DAM).

 

Table 1.

 

Primers Used in the MYOC SSCP Assay

	Forward Primer	Reverse Primer	Size (bp)
Promoter
P5	TTGCAGATACGTTGTAAGTGAAAT	TCTTCTCAGAAAGATGTTTTCAAAT	190
P4	TTTCATTATCATTTGTTTCCTTTG	TTTGGGAGAACTTTCTAATTTCA	188
P3	ACATTGACATTGGTGCCTGA	TAACAGCCAGCCAGAACACA	164
P2	CCTGGAGCCTGGTAGGGT	TGGTGTGCTGATTTCAACAAG	194
P-CA	TGAGTTTGCAGAGTGAATGGA	ACGCTGCCAGCAAGATTC	241
P1	GGGTGCATAAATTGGGATGT	ATGCCCGAGCTCCAGAGAG	175
Exon
1-A	GGCTGGCTCCCCAGTATATA	ACAGCTGGCATCTCAGGC	174
1-B	ACGTTGCTCCAGCTTTGG	GATGACTGACATGGCCTGG	196
1-C	AGTGGCCGATGCCAGTATAC	CTGGTCCAAGGTCAATTGGT	189
1-D	AGGCCATGTCAGTCATCCAT	TCTCTGGTTTGGGTTTCCAG	214
1-E	TGACCTTGGACCAGGCTG	CCTGGCCAGATTCTCATTTT	200
1-F	TGGAGGAAGAGAAGAAGCGA	CTGCTGAACTCAGAGTCCCC	187
2	AACATAGTCAATCCTTGGGCC	TAAAGACCATGTGGGCACAA	234
3-A	TTATGGATTAAGTGGTGCTTC	ATTCTCCACGTGGTCTCCTG	177
3-B	AAGCCCACCTACCCCTACAC	AATAGAGGCTCCCCGAGTACA	184
3-C	ATACTGCCTAGGCCACTGGA	CAATGTCCGTGTAGCCACC	190
3-D	TGGCTACCACGGACAGTTC	CATTGGCGACTGACTGCTTA	197
3-E	GAACTCGAACAAACCTGGGA	CATGCTGCTGTACTTATAGCGG	195
3-F	AGCAAGACCCTGACCATCC	AGCATCTCCTTCTGCCATTG	179

Figure 1.

 

IOP time course of cynomolgus monkeys during first DEX treatment. Eleven monkeys were treated with topical glucocorticoids (10 μl of 0.1% DEX) three times per day for 28 days. IOP measurements were recorded every 3 to 4 days during 4 weeks of glucocorticoid treatment and for an additional 2 weeks. An elevation of IOP in excess of 5 mm Hg was observed in five monkeys that were considered steroid responders. (•) IOP of steroid responders at each time point; (○) IOP of the six nonresponding monkeys. Data are mean ± SD.

Table 2.

 

Reproducibility of DEX-Induced Ocular Hypertension in Cynomolgus Monkeys

Monkey	Steroid Responsiveness	Change in IOP (mm Hg)*
				Test 1	Test 2
1317	NR	0.5	1.5
1318	NR	3	1
1319	NR	3.5	1
1320	NR	2.5	−1.5
1322	NR	0	1.5
548	NR	1.5	1
1315	R	6	6.5
1316	R	14.5	8.5
1321	R	9.5	—
385	R	14	6
564	R	9	7.5

*  Measured over a 4-week period. NR, nonresponder; R, responder.

Figure 2.

 

IOP time course of cynomolgus monkeys during the second DEX treatment. After a 6-month washout period, 10 of the 11 monkeys were re-treated with the same regimen of DEX. Elevated IOP was observed in the same monkeys that had steroid-induced elevations in IOP during the first course of glucocorticoids. (•) IOP of steroid responders at each time point; (○) IOP of the six nonresponders. Data are mean ± SD.

Figure 3.

 

Myocilin gene sequence. (A) Aligned DNA sequences immediately upstream of the coding sequences of the human, cynomolgus monkey, and mouse myocilin genes. Identical DNA sequences are shaded gray, and gaps in the sequence needed to maintain alignment are represented by dashes. The putative start codon (ATG) is shown for each gene at the 3′ end of the upstream sequence. (B) Exon–intron borders of the monkey myocilin gene. The coding sequence is represented by capital letters, and the intron sequence is represented by lowercase letters. (C) Aligned coding sequences of the human, cynomolgus monkey, cow, and mouse myocilin genes. Amino acids that are conserved among the proteins predicted by these myocilin genes are represented by the one-letter codes beneath the sequence alignment. Dashes indicate amino acids that are not conserved among the predicted proteins. The nucleotide numbering corresponds to the human MYOC sequence.

Table 3.

 

Monkey MYOC Variations

Sequence Variation	Effect*	Responders (n = 5)	Nonresponders (n = 5)
−618 G→A	Creates a TGT3 site	1, †	1
−585 T→C	None	2	2
−522/523 16 bp INS	Creates an E1AF site	1	3
−508 G→A	None	2	2
−79 del 1 bp	None	2	0
PRO26LEU	None	3	2
PRO115GLY	None	1	2
LEU139LEU	None	2	3
ARG160SER	Change in Charge	1	0
Intron 1 C→T 42 bp 3′ exon 1	None	1	0
GLU218LYS	Change in Charge	0	2
SER231ASN	None	0	2
GLY268GLY	None	0	1

*  None indicates that a sequence variation failed to alter the charge, polarity, or size of the predicted myocilin protein or the putative enhancer and promoter sequences of the myocilin gene.

†  This monkey was homozygous for the −618 G→A variation.

Table 4.

 

Human MYOC Varations

A. Sequence Changes Affecting the Promoter or Predicted Protein Sequence
Sequence Variation	Effect	Responders (%) (n = 70)	Nonresponders (%) (n = 23)	Control Subjects (%) (n = 91)
−224 bp 5′ to exon 1 (T→C)	None	3 (4.3)	3 (13), †	7 (7.7), ‡
−190 bp 5′ to exon 1 (G→T)	None	0	1 (4.3)	0
−153 bp 5′ to exon 1 (T→C)	Altered Sp1 site	1 (1.4)	0	0
−126 bp 5′ to exon 1 (T→C)	None	2 (2.9)	4 (17)	6 (6.6)
−83 bp 5′ to exon 1 (A→G)	None	11 (16)	3 (13)	21 (23)
CYS9SER	None	0	0	1 (1.1)
ASN73SER	None	0	0	1 (1.1)
ARG76LYS	None	2 (2.9)	0	17 (19)
ARG82CYS	Change in Charge	1 (1.4)	0	0
SER203PHE	Change in Polarity	0	0	1 (1.1)
GLN368STOP	Premature termination	1 (1.4)	0	0
LYS398ARG	None	0	0	4 (4.4)
A→G 3 bp 3′ to exon 2	None	1 (1.4)	0	0

B. Synonymous Codon Changes
Sequence Variation	Responders (%) (n = 70)	Nonresponders (%) (n = 23)	Control Subjects (%) (n = 91)
;l>PRO13PRO	0	0	1 (1.1)
LEU159LEU	0	0	1 (1.1)
LYS216LYS	0	1 (4.3), §	0
THR285THR	0	0	1 (1.1)
THR325THR	0	0	1 (1.1)
TYR347TYR	2 (2.9)	3 (13)	7 (7.7)
ALA488ALA	0	1 (4.3)	0

*  None indicates that a sequence variation failed to alter the charge, polarity, or size of the predicted myocilin protein or the putative enhancer and promoter sequences of the myocilin gene.

†  One was homozygous T→C, whereas two were heterozygous.

‡  Four were homozygous T→C, whereas three were heterozygous.

§  Homozygous T→C.

Table 5.

 

Allele Frequencies of MY5

Allele	Nonresponders	Normal Controls	Responders
		(n = 23)	(n = 91)	(n = 70)
1	12 (26)	45 (25)	29 (21)
2	11 (24)	36 (20)	24 (17)
3	21 (46)	98 (54)	87 (62)
4	2 (4.3)	1 (0.5)	0
5	0	1 (0.5)	0
6	0	1 (0.5)	0

 Data are number of subjects (percentage of total group).

The authors thank the patients for their participation in this study; Jeaneen Andorf, Gretel Beck, Robyn Hockey, Heidi Haines, Luan Streb, Christine Taylor, and Kimberlie Vandenburgh, for excellent technical assistance; and Robert Buttery, Melinda Cain, Michael Coote, Terry Couper, Danielle Healey, J. Lynch, Paul McCartney, and Julian Rait for help in identifying the subjects that were included in the study.