The Effect of Molecular Weight on Passage of Proteins Through the Blood-Aqueous Barrier

Abstract

Purpose: To determine the effect of molecular weight (MW) on the concentration of plasma-derived proteins in aqueous humor and to estimate the plasma-derived and eye-derived fractions for each protein.

Methods: Aqueous humor and plasma samples were obtained during cataract surgery on an institutional review board–approved protocol. Protein concentrations were determined by ELISA and quantitative antibody microarrays. A total of 93 proteins were studied, with most proteins analyzed using 27 to 116 aqueous and 6 to 30 plasma samples.

Results: Plasma proteins without evidence of intraocular expression by sequence tags were used to fit a logarithmic model relating aqueous-plasma ratio (AH:PL) to MW. The log(AH:PL) appears to be well predicted by the log(MW) (P < 0.0001), with smaller proteins such as cystatin C (13 kDa) having a higher AH:PL (1:6) than larger proteins such as albumin (66 kDa, 1:300) and complement component 5 (188 kDa, 1:2500). The logarithmic model was used to calculate the eye-derived intraocular fraction (IOF) for each protein. Based on the IOF, 66 proteins could be categorized as plasma-derived (IOF<20), whereas 10 proteins were primarily derived from eye tissue (IOF >80), and 17 proteins had contribution from both plasma and eye tissue (IOF 20–80).

Conclusions: Protein concentration of plasma-derived proteins in aqueous is nonlinearly dependent on MW in favor of smaller proteins. Our study demonstrates that for proper interpretation of results, proteomic studies evaluating changes in aqueous humor protein levels should take into account the plasma and eye-derived fractions.

Identification of the protein constituents of body fluid can lead to a better understanding of normal physiology and improve our understanding of the molecular underpinnings of common diseases. The aqueous proteome is of particular interest as it might be altered in many ocular diseases. Previous molecular analysis of the aqueous humor (AH) have mainly focused on the protein composition and the number of proteins in the Human Eye Proteome Project has increased to 827 proteins in 2018.¹ 

To better understand the influence of the AH proteome on ocular physiology and disease, it is critical to differentiate between aqueous proteins derived from ocular structures and plasma-derived proteins that diffuse into AH. Although plasma-derived proteins may exert influence on ocular physiology, their levels within AH are highly dependent on their plasma levels. In contrast, proteins derived from intraocular structures are more likely to be internally regulated and hence are more likely to have direct impact on ocular function or be related to ocular pathophysiology. 

In contrast to plasma, the AH is relatively free of proteins, so as to maintain clarity of vision. Plasma proteins are restricted from entering AH by the blood-aqueous barrier, although a significant portion of AH proteins originates from plasma. The blood-aqueous barrier is formed by the tight junctions of the nonpigmented epithelia of the ciliary body and posterior iris, and the nonfenestrated iris capillaries. Plasma proteins can pass through the fenestrated capillaries of the ciliary body but are prevented from entering the posterior chamber by the ciliary body nonpigmented epithelia. Similarly, plasma proteins circulating through iris capillaries are blocked from entering the iris stroma by the iris capillary vessel walls.^2,3 

However, plasma proteins that pass through the fenestrated capillaries of the ciliary body can diffuse through the ciliary body stroma into the iris stroma at the iris root. Because the anterior surface of the iris is not bound by any nonpenetrable barrier, plasma proteins within the iris stroma can then diffuse into the AH within the anterior chamber.^4,5 

According to Fick's laws of diffusion, small proteins diffuse through media more rapidly than large proteins. Thus, we hypothesize a size-dependent relationship between the concentrations of proteins in AH relative to their plasma concentrations. A similar process has been described for plasma proteins entering the cerebrospinal fluid (CSF) through the blood-CSF barrier.^6,7 Previous proteomics studies on AH^8–18 have provided insight into the protein composition but to the best of our knowledge, a comprehensive analysis of the combined proteome of the AH and plasma has not yet been carried out. To test our hypothesis, we simultaneously determined the concentrations of 93 plasma proteins within plasma and AH and examined the effect of molecular weight (MW) on the ratio of AH concentration to plasma concentration. 

We also introduce the eye-derived intraocular fraction (IOF) as an estimate of the proportion of each protein derived from ocular structures. This allows us to categorize the proteins into three groups: the proteins derived from the surrounding eye tissue, the proteins derived from the plasma, and proteins with contributions from both. 

Methods

This study was approved by the Indiana University Institutional Review Board and followed the tenets of the Declaration of Helsinki. All subjects gave informed consent after explanation of the nature and possible consequences of the study. 

Subjects

Aqueous samples were collected from eyes at the start of cataract surgery. Eligible subjects were older than 40, had no eye disease or disorder except cataract, had consented to undergo cataract surgery, and were willing to participate in this study. Exclusion criterion was prior intraocular surgery of any type, including laser. 

Sample Collection and Clinical Data

AH samples were collected at the start of surgery by inserting a 30-gauge needle through the peripheral cornea and withdrawing approximately 100–250 μL of aqueous into a tuberculin syringe. Alternatively, a paracentesis was created with a 15-degree blade and the aqueous sample collected with a 30-gauge cannula on a tuberculin syringe. The aqueous sample was immediately transferred to a barcode-labeled cryotube, placed on ice, and transported to the Biospecimen Repository where the specimens were stored at −80°C until proteomic analysis. Clinical data and information associated with the AH sample were entered into the caTISSUE suite software that we have adapted for eye disease research. 

Blood samples were collected in EDTA-coated vacutainers from a subset of subjects just before the start of surgery. To obtain platelet-poor plasma, a double spin protocol was used. Blood samples were first centrifuged at 3000g at room temperature for 10 minutes. The plasma was transferred to a sterile 2-mL Eppendorf tube without disturbing the buffy coat and subjected to a second centrifugation step at 12,000g for 10 minutes. The platelet-poor plasma was aliquoted into sterile cryovials and stored at −80°C. Samples were frozen within 2 hours of collection. 

Enzyme-Linked Immunosorbent Assay

Protein concentration in aqueous and plasma samples was measured using commercial ELISA kits (Abcam, Cambridge, UK; Raybiotech, Norcross, GA, USA; LifeSpan BioScience, Seattle, WA, USA) as per the manufacturer's instruction. Standards, AH samples, and diluted plasma samples were added to 96-well plates containing highly purified protein-specific antibodies, incubated, and washed. Protein-specific biotinylated detection antibody was added, incubated, and washed, followed by the addition of an enzyme substrate, then a stop solution. Plates were read at 450 nm with a Tecan GENios Pro Multifunction Microplate Reader (Tecan, Mannedorf, Switzerland). Concentration estimates were found by fitting a four-parameter log-logistic model to a standard curve generated alongside the experimental data using the drc package¹⁹ in R.²⁰ 

Quantitative Antibody Microarray Analysis

The proteins in AH and plasma of patients were analyzed with the glass-slide quantitative antibody microarray platform (Quantibody; Raybiotech), which enables the accurate concentration determination of multiple proteins simultaneously. Proteins were combined in the same experiment based on their similar expected concentration range in AH and plasma. The proteins were screened for cross reactivity by the Raybiotech support staff. Antibodies to 20 to 40 proteins, along with positive and negative controls, were spotted onto a glass slide in quadruplicate. Samples were assayed according to the manufacturer's protocol. 

Briefly, the glass slides were blocked, incubated with 70 μL of AH or plasma overnight at 4°C, washed, and then incubated with the biotin-conjugated antibodies specific for the proteins bound to the antibody on the glass slide. The slides were developed with streptavidin-labeled Cy3 equivalent dye and the signal was quantified using a laser scanner (GenePix 4000A; Molecular Devices Corporation, San Jose, CA, USA) at four different settings of the photomultiplier tube gain. These were combined using a linear regression model, in which the log-observed intensity at each spot was modeled as a function of gain for unsaturated spots with intensities at or above 200. Poor-quality spots were removed from the analysis based on visual inspection. Data were then normalized using the positive controls. Concentration estimates were found by fitting a four-parameter log-logistic model to a standard curve generated alongside the experimental data using the drc¹⁹ package in R.²⁰ 

Protein Data

The theoretical MW of a protein was calculated using the sequence of the circulating protein chain from the UniProt Knowledgebase^21,22 and the Compute pI/Mw tool of the ExPASy bioinformatic resource portal.²³ We conducted literature searches to identify the experimentally determined MW of the circulating form of each protein and the theoretical MW was adjusted when experimental data were available. 

Statistical Analysis

For each protein, we modeled the concentration on the log scale using a mixed effects model with fluid (AH or plasma) as a predictor. Subjects were treated as random effects to incorporate correlation between fluids from the same patient. The variance was estimated separately for each fluid. From each analysis, we obtained the concentration ratio between aqueous and plasma (AH:PL), the standard errors associated with each estimate, and the correlation across the two fluids. All analyses were conducted using the nlme²⁴ and multcomp²⁵ packages in R. 

To predict the diffusion component of AH proteins, we first selected a subset of high-abundance plasma proteins measured by ELISA with no evidence of intraocular expression by sequence tags (EST) within the ciliary body,²⁶ iris,²⁷ cornea,²⁸ anterior segment, trabecular meshwork,²⁹ lens,³⁰ retina,³¹ and RPE/choroid³² from the NEIBank sequence tag analysis project.^33–35 We then modeled AH:PL ratio, on the log scale as a function of the MW, also on a log scale:  

\(\def\upalpha{\unicode[Times]{x3B1}}\)\(\def\upbeta{\unicode[Times]{x3B2}}\)\(\def\upgamma{\unicode[Times]{x3B3}}\)\(\def\updelta{\unicode[Times]{x3B4}}\)\(\def\upvarepsilon{\unicode[Times]{x3B5}}\)\(\def\upzeta{\unicode[Times]{x3B6}}\)\(\def\upeta{\unicode[Times]{x3B7}}\)\(\def\uptheta{\unicode[Times]{x3B8}}\)\(\def\upiota{\unicode[Times]{x3B9}}\)\(\def\upkappa{\unicode[Times]{x3BA}}\)\(\def\uplambda{\unicode[Times]{x3BB}}\)\(\def\upmu{\unicode[Times]{x3BC}}\)\(\def\upnu{\unicode[Times]{x3BD}}\)\(\def\upxi{\unicode[Times]{x3BE}}\)\(\def\upomicron{\unicode[Times]{x3BF}}\)\(\def\uppi{\unicode[Times]{x3C0}}\)\(\def\uprho{\unicode[Times]{x3C1}}\)\(\def\upsigma{\unicode[Times]{x3C3}}\)\(\def\uptau{\unicode[Times]{x3C4}}\)\(\def\upupsilon{\unicode[Times]{x3C5}}\)\(\def\upphi{\unicode[Times]{x3C6}}\)\(\def\upchi{\unicode[Times]{x3C7}}\)\(\def\uppsy{\unicode[Times]{x3C8}}\)\(\def\upomega{\unicode[Times]{x3C9}}\)\(\def\bialpha{\boldsymbol{\alpha}}\)\(\def\bibeta{\boldsymbol{\beta}}\)\(\def\bigamma{\boldsymbol{\gamma}}\)\(\def\bidelta{\boldsymbol{\delta}}\)\(\def\bivarepsilon{\boldsymbol{\varepsilon}}\)\(\def\bizeta{\boldsymbol{\zeta}}\)\(\def\bieta{\boldsymbol{\eta}}\)\(\def\bitheta{\boldsymbol{\theta}}\)\(\def\biiota{\boldsymbol{\iota}}\)\(\def\bikappa{\boldsymbol{\kappa}}\)\(\def\bilambda{\boldsymbol{\lambda}}\)\(\def\bimu{\boldsymbol{\mu}}\)\(\def\binu{\boldsymbol{\nu}}\)\(\def\bixi{\boldsymbol{\xi}}\)\(\def\biomicron{\boldsymbol{\micron}}\)\(\def\bipi{\boldsymbol{\pi}}\)\(\def\birho{\boldsymbol{\rho}}\)\(\def\bisigma{\boldsymbol{\sigma}}\)\(\def\bitau{\boldsymbol{\tau}}\)\(\def\biupsilon{\boldsymbol{\upsilon}}\)\(\def\biphi{\boldsymbol{\phi}}\)\(\def\bichi{\boldsymbol{\chi}}\)\(\def\bipsy{\boldsymbol{\psy}}\)\(\def\biomega{\boldsymbol{\omega}}\)\(\def\bupalpha{\unicode[Times]{x1D6C2}}\)\(\def\bupbeta{\unicode[Times]{x1D6C3}}\)\(\def\bupgamma{\unicode[Times]{x1D6C4}}\)\(\def\bupdelta{\unicode[Times]{x1D6C5}}\)\(\def\bupepsilon{\unicode[Times]{x1D6C6}}\)\(\def\bupvarepsilon{\unicode[Times]{x1D6DC}}\)\(\def\bupzeta{\unicode[Times]{x1D6C7}}\)\(\def\bupeta{\unicode[Times]{x1D6C8}}\)\(\def\buptheta{\unicode[Times]{x1D6C9}}\)\(\def\bupiota{\unicode[Times]{x1D6CA}}\)\(\def\bupkappa{\unicode[Times]{x1D6CB}}\)\(\def\buplambda{\unicode[Times]{x1D6CC}}\)\(\def\bupmu{\unicode[Times]{x1D6CD}}\)\(\def\bupnu{\unicode[Times]{x1D6CE}}\)\(\def\bupxi{\unicode[Times]{x1D6CF}}\)\(\def\bupomicron{\unicode[Times]{x1D6D0}}\)\(\def\buppi{\unicode[Times]{x1D6D1}}\)\(\def\buprho{\unicode[Times]{x1D6D2}}\)\(\def\bupsigma{\unicode[Times]{x1D6D4}}\)\(\def\buptau{\unicode[Times]{x1D6D5}}\)\(\def\bupupsilon{\unicode[Times]{x1D6D6}}\)\(\def\bupphi{\unicode[Times]{x1D6D7}}\)\(\def\bupchi{\unicode[Times]{x1D6D8}}\)\(\def\buppsy{\unicode[Times]{x1D6D9}}\)\(\def\bupomega{\unicode[Times]{x1D6DA}}\)\(\def\bupvartheta{\unicode[Times]{x1D6DD}}\)\(\def\bGamma{\bf{\Gamma}}\)\(\def\bDelta{\bf{\Delta}}\)\(\def\bTheta{\bf{\Theta}}\)\(\def\bLambda{\bf{\Lambda}}\)\(\def\bXi{\bf{\Xi}}\)\(\def\bPi{\bf{\Pi}}\)\(\def\bSigma{\bf{\Sigma}}\)\(\def\bUpsilon{\bf{\Upsilon}}\)\(\def\bPhi{\bf{\Phi}}\)\(\def\bPsi{\bf{\Psi}}\)\(\def\bOmega{\bf{\Omega}}\)\(\def\iGamma{\unicode[Times]{x1D6E4}}\)\(\def\iDelta{\unicode[Times]{x1D6E5}}\)\(\def\iTheta{\unicode[Times]{x1D6E9}}\)\(\def\iLambda{\unicode[Times]{x1D6EC}}\)\(\def\iXi{\unicode[Times]{x1D6EF}}\)\(\def\iPi{\unicode[Times]{x1D6F1}}\)\(\def\iSigma{\unicode[Times]{x1D6F4}}\)\(\def\iUpsilon{\unicode[Times]{x1D6F6}}\)\(\def\iPhi{\unicode[Times]{x1D6F7}}\)\(\def\iPsi{\unicode[Times]{x1D6F9}}\)\(\def\iOmega{\unicode[Times]{x1D6FA}}\)\(\def\biGamma{\unicode[Times]{x1D71E}}\)\(\def\biDelta{\unicode[Times]{x1D71F}}\)\(\def\biTheta{\unicode[Times]{x1D723}}\)\(\def\biLambda{\unicode[Times]{x1D726}}\)\(\def\biXi{\unicode[Times]{x1D729}}\)\(\def\biPi{\unicode[Times]{x1D72B}}\)\(\def\biSigma{\unicode[Times]{x1D72E}}\)\(\def\biUpsilon{\unicode[Times]{x1D730}}\)\(\def\biPhi{\unicode[Times]{x1D731}}\)\(\def\biPsi{\unicode[Times]{x1D733}}\)\(\def\biOmega{\unicode[Times]{x1D734}}\)\begin{equation}{\rm{lo}}{{\rm{g}}_{10}}{\rm{(}}AH:PL{\rm{)}} = a + b \times {\rm{lo}}{{\rm{g}}_{10}}{\rm{(MW)}} + \epsilon \end{equation}

 

The data were fit using a weighted mixed effects model in which proteins were treated as random effects, and each protein ratio was weighted by its standard error using the metafor³⁶ package in R. Using this initial model, we calculated a 95% prediction interval for all proteins based on the MW. 

To distinguish plasma-derived proteins from predominately eye-derived proteins, the IOF was calculated for each protein as  

\begin{equation}{\rm{IOF}} = \max \left( {0,1 - {{10}^{{\rm{upper\ bound\ }}-{\rm{\ observed}}}}} \right) \times 100,\end{equation}

as described by Reiber for CSF.^37,38 A final model was produced by also including proteins measured by quantitative antibody microarray that had an IOF of 0, no evidence of ESTs in ocular tissue, and minimal heterogeneity in the reported MW of the circulating plasma protein in the literature. This model was used to produce the final estimate of the IOF for all proteins.  

Results

Samples were withdrawn from patients who ranged in age from 50 to 80. Table 1 shows the demographics of the AH and plasma samples used for the various protein analyses. Sample size for each protein measured is summarized in Table 2 for AH and plasma. 

Table 1

 

Patient Demographics of Aqueous and Plasma Samples

 

Table 2

 

Protein Concentrations in AH and Plasma

 

Analysis of AH Proteome by ELISA

To better understand the relationship between MW and AH concentration of plasma proteins, we sought to quantitate the concentrations of proteins that had been previously found by liquid chromatography tandem mass spectrometry in both aqueous^8–18 and plasma.^39–41 Forty proteins were selected based on the availability of high-sensitivity ELISA needed to detect the expected low concentration in AH. For 3 of the 40 proteins, the concentrations in AH were too low to be adequately quantified. The concentrations of proteins in AH versus plasma are shown in Figure 1. Relative to albumin, the primary protein component of both aqueous and plasma, low-MW proteins tend to be overrepresented in aqueous, whereas high-MW proteins tend to be underrepresented. 

Figure 1

 

Effect of MW on the AH versus plasma protein concentration. The mean aqueous and plasma concentrations of 37 proteins as measured by ELISA are plotted. The area of the bubble represents the MW of the protein. The dashed line indicates the AH:PL of albumin (1:300). Proteins above the line are overrepresented (higher AH:PL) in AH relative to albumin, whereas proteins below the line are underrepresented (lower AH:PL) in AH relative to albumin. Proteins primarily derived from plasma are shown in brown (eye-derived IOF 0–20), proteins primarily derived from eye tissue are shown in blue (IOF 80–100), and proteins originating from both plasma and eye tissues are shown in green (IOF 20–80). Plasma-derived proteins smaller than albumin have a higher AH:PL than albumin (above the dashed line), whereas larger plasma-derived proteins have a lower AH:PL than albumin (below the dashed line).

After determining the concentration of the 37 proteins, the ratio of AH concentration to plasma concentration (AH:PL) was calculated. As shown in Figure 2, smaller proteins such as cystatin C (CYTC, MW ∼13 kDa) tend to have a higher AH:PL ratio (1:5.6) than larger proteins such as albumin (ALBU, MW ∼66 kDa), which had an AH:PL ratio of 1:300, and complement component 5 (CO5, MW ∼188 kDa), which had an AH:PL ratio of 1:2500. The plot of AH:PL versus MW is shown in Figure 2. The AH:PL of the 37 proteins is listed in Table 2. 

Figure 2

 

The AH:PL as a function of MW. The AH:PL ratio for 93 proteins measured in both fluids. Proteins represented by red squares were used to construct the model. Proteins represented by blue circles illustrate the dependence of the IOF on MW. The curve is logarithmic as a function of MW (A), and linear when plotting MW on a log scale (B). The shaded regions show the 95% prediction interval.

Analysis of AH Proteome by Quantitative Antibody Microarray

To identify additional proteins that are quantifiable in aqueous, we screened 157 proteins by quantitative glass-slide antibody microarrays. Of the 157 proteins analyzed in initial screening assays, 101 proteins were not further studied: 42 proteins were not detectable in AH, 38 proteins were at the lower detection limit, and 21 proteins were in the linear range but were excluded for various reasons such as cross reactivity and plasma dilution requirements out of the range of the other proteins. 

The remaining 56 proteins were further studied using larger sample sizes (up to 116 aqueous samples and 14 plasma samples) with quantitative antibody microarrays. Proteins were assigned to quantitative antibody microarray group 1 (Q1: 35 proteins) and group 2 (Q2: 21 proteins) based on dilution and protein cross reactivity. The concentration and AH:PLs of all 56 proteins measured by quantitative antibody microarray are included in Table 2. 

Modeling AH:PL as Function of MW

Of the 37 proteins measured by ELISA, 17 that had no evidence of intraocular expression by EST were used to fit the model. Using this model, we calculated the IOF for all 93 proteins, which included the 37 proteins measured by ELISA and the 56 proteins measured by the quantitative antibody microarray (Table 2). Of the 56 proteins measured by quantitative antibody microarray, 10 proteins with an IOF of zero and no evidence of intraocular expression by ESTs were used to refine the model. The Notes column in Table 2 shows the proteins that were used to create (M1) and refine (M2) the model. The plot of AH:PL versus MW of all proteins is shown in Figure 2. Based on the IOF, the 93 proteins were categorized into three groups: proteins that are primarily derived from plasma (IOF <20; 66 proteins), proteins for which eye tissue is the principal source of synthesis (IOF >80; 10 proteins), and proteins with contribution from both eye tissue and plasma (IOF 20–80; 17 proteins). The data are shown in Figure 1 and summarized in Table 2. 

The dependence of the IOF on MW is illustrated in Figure 2B for four proteins with a similar AH:PL ratio: C-X-C motif chemokine 14 (CXL14, 1:2.1), metalloproteinase inhibitor 2 (TIMP2, 1:3), Wnt inhibitory factor 1 (WIF1, 1:2.3), and vascular endothelial growth factor (VEGF) receptor 1 (VGFR1, 1:2.5). As the MW increases from 9 kDa for CXL14, to 22 kDa for TIMP2, to 38 kDa for WIF1, and to 110 kDa for VGFR1, the IOF increases from 0% for CXL14 to 53% for TIMP2, to 87% for WIF1 and to 98% for VGFR1. The data are shown in Figure 2 and summarized in Table 2. 

Correlation Between Aqueous and Plasma Concentrations of Plasma-Derived Proteins

For aqueous and plasma samples that were obtained simultaneously from individual subjects, the correlation coefficient were estimated from the AH:PL ratio models for each protein.⁴² Using those proteins derived from plasma with an IOF of zero, we found that low-MW proteins (less than 50 kDa) exhibited high correlation between aqueous and plasma concentrations. With increasing MW, the correlation between aqueous and plasma concentrations decreased to essentially zero for proteins with a MW of greater than 150 kDa (Fig. 3). 

Figure 3

 

Correlation between AH and plasma as a function of MW. For plasma-derived proteins with no evidence of ocular production, the correlation between AH and plasma tends to be high when the MW is low (< 50 kDa), decreasing when the MW is between 50 and 150 kDa, and approximately 0 for MWs above 150 kDa. The shaded region gives a 95% confidence interval for the smoothed curve.

Rank Order of Proteins in AH Compared With Plasma

The concentration of individual proteins in the AH and plasma is listed in Table 2 in the order of decreasing concentration in AH. The examination of the rank order of the most abundant AH proteins shows that there is an order of magnitude difference between the most abundant protein, albumin, and the rest of the AH proteins. The fundamental difference in the rank order between aqueous and plasma is the relative increase of plasma-derived proteins with lower MW and the relatively high abundance of ocular-derived proteins in AH. This shift is highlighted in Table 3 for two examples of plasma-derived proteins with a relatively low MW, CYTC, and beta 2 microglobulin (B2MG). The three examples of ocular-derived proteins, pigment epithelium-derived factor (PEDF), Dickkopf-related protein 3 (DKK3), and osteopontin (OSTP), all with IOF greater than 95%, highlight the relatively high contribution of proteins that are synthesized by ocular tissue. 

Table 3

 

Comparative Rank of Proteins in AH and Plasma

 

The concept of the percent transfer, that is, how much of the total amount of plasma protein passes into AH (percent transfer = AH concentration / plasma concentration × 100), is shown in Table 3 for the four plasma-derived proteins. 

Discussion

Although the protein component of AH is very dilute compared with that of plasma, most of the protein mass in AH is thought to arise from plasma. However, AH is not a simple diffusate of plasma, because it has both qualitative and quantitative differences in protein content relative to plasma. 

By analyzing the MW of plasma-derived proteins in AH, we found that lower-MW proteins tended to penetrate AH more readily than higher-MW proteins. Indeed, AH concentrations of plasma-derived proteins are a function of MW. The logarithmic function is consistent with the model in which plasma proteins pass into AH by diffusing through ciliary body and iris stroma along a concentration gradient.^4,5,43 Our study adds to the mounting evidence that the blood-aqueous barrier, as it pertains to the anterior chamber, is best thought of as a size-dependent diffusional gradient.^4,5,43 A similar model has been proposed for entry of plasma proteins into CSF through the blood-CSF barrier.^6,7 

Introducing the AH:PL versus MW logarithmic function enables the estimation of the AH level of a plasma protein and the calculation of the contribution of intraocular tissues to the total aqueous concentration of the protein. By taking the IOF of an aqueous protein into account, one may be able to assess the protein's relative importance to ocular physiology and the pathogenesis of various eye diseases. Proteins with high IOF are more likely to be derived from the corneal endothelium, trabecular meshwork, iris, lens, and ciliary body, and represent disease biomarkers. These proteins may serve a potential target for further studies that look for differences in protein concentration in AH in various pathologic ocular conditions. In addition, our findings could be used to help understand what occurs in conditions associated with the breakdown of the blood-aqueous barrier. The most abundant proteins in AH, albumin and IgG, are plasma derived and have a low IOF. Although purely plasma-derived proteins may be important for normal ocular physiology, their aqueous concentrations are directly dependent on their plasma concentration and MW and thus their ocular function is regulated by their systemic levels. 

It is important to note some limitations building our model. We assume that a protein is essentially plasma-derived if it is a known high-abundance plasma protein and lacks a known EST from intraocular tissues. The catalog of genes identified by EST sequencing of a cDNA library reflects a random sample of the mRNA present in the cell. For abundantly expressed genes, the library provides a good indication of the gene transcription in the tissue. However, less transcriptionally active genes might be missed during the generation of the library. Furthermore, transcript levels do not necessarily reflect protein levels.⁴⁴ Hence, we may have missed the intraocular contribution to the aqueous concentration for some low-abundance proteins. However, given the overall curve fitting, we think the intraocular contribution for the proteins used to build the model is likely negligible. 

Another limitation of our targeted approach is that the analysis is restricted to proteins that are available for ELISA and/or quantitative antibody microarrays. Therefore, only a very limited number of proteins in AH could be quantified. Although our study was thorough, the 93 proteins we quantified in both the AH and plasma are only a fraction of the aqueous proteome. 

Between aqueous and plasma samples collected concurrently, we found strong correlations in concentrations for small proteins but poor correlations for large proteins. This suggests that small proteins diffuse rapidly enough into AH to quantitatively reflect the plasma concentrations. In contrast, larger proteins diffuse far more slowly into AH so that by the time the proteins reach the AH, the plasma concentrations have changed sufficiently such that there is no longer correlation between the corresponding concentrations. 

In conclusion, to our knowledge, this study is the first to simultaneously measure the protein concentration in aqueous and plasma in a large number of proteins. Our comprehensive analysis, demonstrating the logarithmic relationship of the MW and the AH:PL, enables an estimate of the contribution of intraocular tissues to the total aqueous concentration of a protein. Moreover, taking the IOF of proteins into account may help guide future studies of AH proteomics by providing potential ocular-derived protein targets that are relevant to ocular physiology and disease. 

Acknowledgments

The authors thank the contributions of Louis B. Cantor, Yara Catoira-Boyle, Rudy Yung, and Shailaja Valluri for obtaining aqueous samples during cataract surgery. 

Supported by the BrightFocus Foundation, Clarksburg, MD (SR), and the American Glaucoma Society, San Francisco, CA (DW). 

Disclosure: S. Ragg, RayBIotech (F); M. Key, None; F. Rankin, None; D. WuDunn, None