Incorporation of an interferon-β neutralizing antibody assay into routine clinical practice

BACKGROUND

Incorporation of routine clinical testing for neutralizing antibodies (NAbs) to interferon (IFN)-β has remained problematic. With increasing treatment choice for patients, routine NAb testing should be incorporated to aid therapeutic decisions.

OBJECTIVE

We sought to improve interpretation of NAb results by combining the luciferase NAb assay (luciferase gene expression assay under control of interferon-stimulated response element) and in-vivo biomarker (myxovirus A protein, MxA) induction in patients with MS.

METHODS

Blood samples (serum and PAXGene(®) for RNA) were obtained pre-injection and 12 hours post-injection of IFN-β from 144 subjects. Sera were tested for NAbs using the luciferase assay. MxA expression was quantified by real-time polymerase chain reaction (PCR).

RESULTS

26% of samples were NAb positive (titre > 20 NU). There was no difference in NAb titres in the pre- or post-dose sera (p = 0.643). MxA expression was inhibited in a dose-dependent fashion in NAb positive samples. Mean MxA level post-IFN-β: NAb negative 2330 (95% CI 1940-2719), NAb 20-99 NU 1533 (95% CI 741-2324), NAb 100-600 NU 832 (186-1478) and NAb > 600 NU 101 (95% CI 0-224). NAb titre and MxA level correlated strongly: MxA pre- (Spearman r = -0.72, p < 0.0001), MxA post- (Spearman r = -0.79, p < 0.0001) and MxA induction (Spearman r = -0.67, p = 0.0004).

CONCLUSION

A single, 12-hour post-injection sample should be used to test for NAbs using the luciferase assay and IFN-β bioactivity (MxA) in the clinical setting.

Hierarchical Structure and Light Scattering in the Cornea

The complex hierarchical structure of the cornea of the eye as it is known from light and electron-microscopy is reviewed. Microscopy depicts the cornea as a layered structure with each layer parallel to the cornea's surfaces. The middle 90% of the cornea is a connective tissue layer called the stroma, which in turn is comprised of 2 μm thick sheets (lamellae) of collagen fibrils embedded in a ground substance. Fibrils within a given lamella have their axes parallel to one another, while fibril axes in adjacent lamellae make large angles with one another. Predictions of light scattering theory are then used to show how light scattering measurements made on fresh tissue can test these structures for possible artifacts due to preparative procedures. With unpolarized light, the total scattering cross-section depends on the fibrillar ultrastructure, but not on the lamellar structure. Angular scattering depends on both levels of hierarchical organization. The lamellar level of organization is especially significant for the propagation and scattering of polarized light. To date the structural implications of scattering measurements are in accord with structures seen in electron micrographs.

Estimation of pulsatile ocular blood flow from intraocular pressure

A relationship has been derived between intraocular pressure and pulsatile blood flow in the eye. Measurements of intraocular pressure show a time variation that is associated with the pulsatile component of arterial pressure. Experimental results provide a means of transforming intraocular pressure changes into ocular volume changes. The eye is represented by a chamber with elastic walls, a pulsatile incoming flow of incompressible fluid (blood), and a steady outgoing flow of blood. Under these conditions, the rate of pulsatile blood flow through the eye can be approximated from the instantaneous intraocular pressure measurements. Data from a healthy human eye are used to illustrate the analysis.

Humoral immune response to EBV in multiple sclerosis is associated with disease activity on MRI

BACKGROUND

Evidence suggests that Epstein-Barr virus (EBV) plays a role in triggering or perpetuating disease activity in multiple sclerosis (MS).

METHODS

We investigated 100 subjects (50 clinically isolated syndrome [CIS], 25 relapsing-remitting [RR] MS, 25 primary progressive [PP] MS) for 1) evidence of EBV reactivation and 2) disease activity as indicated by serial gadolinium (Gd)-enhanced MRIs over a 5-year period. EBV DNA in blood was quantified by real-time quantitative PCR and EBV serology for anti-Epstein-Barr virus nuclear antigen 1 (EBNA-1) immunoglobulin G (IgG), anti-viral capsid antigen (VCA) IgG, and anti-EBV IgM. Data were analyzed using repeated measures analysis, analysis of variance, and logistic regression analysis.

RESULTS

All subjects had serologic evidence of previous EBV infection, but no lytic reactivation was detected. Significant differences in EBNA-1 IgG titers were found between subgroups, highest in the RRMS cohort compared with PPMS (p < 0.001) and CIS (p < 0.001). Gd-enhancing lesions on MRI correlated with EBNA-1 IgG (r = 0.33, p < 0.001) and EBNA-1:VCA IgG ratio (r = 0.36, p < 0.001). EBNA-1 IgG also correlated with change in T2 lesion volume (r = 0.27, p = 0.044) and Expanded Disability Status Scale score (r = 0.3, p = 0.035).

CONCLUSIONS

The correlation between elevated Epstein-Barr virus nuclear antigen 1 (EBNA-1) immunoglobulin G (IgG) and gadolinium-enhancing lesions suggests an association between Epstein-Barr virus (EBV) infection and multiple sclerosis (MS) disease activity. The heightened immune response to EBV in MS is specifically related to EBNA-1 IgG, a marker of the latent phase of the virus. The lack of association between acute viral reactivation in the peripheral blood and Gd(+) lesions suggests a limited role of the former in driving disease activity.

Small-angle light scattering and birefringence properties of chick cornea

PURPOSE

Techniques employing polarized light propagation and scattering are useful in examining the cornea's lamellar structure. Recent advances in theoretical methods have significantly increased the ability to relate features of lamellar arrangements to measurements of transmitted polarized light. The chick cornea, because of its hypothesized structure of a gradual helical rotation of lamellar pairs, presents an interesting model for further development of this methodology.

METHODS

Small-angle light scattering (SALS) and polarized transmission measurements were made on 7-week-old chick corneas under conditions that closely approximate the physiological state. Birefringence properties were determined from the transmission measurements and compared to the results of model calculations of polarized light propagating through lamellae organized according to the hypothesized structure for chick cornea.

RESULTS

The I+ small-angle light scattering pattern had 4 cloverleaf lobes aligned with the crossed polarizer and analyzer axes. The lobes disappeared when the transcorneal pressure was increased from zero to 18 mmHg. Retardation measured at 18 mmHg was very small (approximately 0.01 microm).

CONCLUSION

The disappearance of the I+ small-angle light scattering pattern when IOP is increased suggests that the lamellae undulate in their relaxed state and the undulations straighten when IOP is increased. Measured birefringence properties are consistent with the hypothesized lamellar structure.

Polarized light propagation in corneal lamellae

The propagation of polarized light through the cornea is affected by the orientations of the corneal lamellae and by the refractive imbalance between the collagen fibrils and the ground substance. Thus, well-designed measurements and analyses of polarized light propagation through the cornea can be used to obtain information regarding the cornea's lamellar and fibrillar structures. This paper shows that, for the rabbit, measured values of the optical parameters strongly suggest that the distribution of lamellae orientations is not random, but has one (or two) preferred orientation directions. Also, there is considerable evidence that collagen is intrinsically anisotropic. The Weiner formula gives the effective birefringence of an assembly of parallel isotropic fibrils and its generalization to the case of anisotropic fibrils is presented. Finally, calculations based on preferred orientation models having lamellae composed of anisotropic fibrils show that comparison with experimental values can yield structural information.

Spatial mapping of polarized light transmission in the central rabbit cornea

In this paper polarized light transmission measurements are made under conditions that closely approximate the physiological state in order to probe lamellar structure in the central cornea of New Zealand white rabbits. The results are interpreted with the aid of a newly developed theory (published elsewhere) in which the cornea is modeled as stacked birefringent layers corresponding to the lamellae. The theory enables predictions of the statistical properties of lamellar ordering based on characteristics of the transmission of polarized light. The experimental results are consistent with a structure in which a number of lamellae have a fixed azimuthal orientation (i.e. about an axis normal to the corneal surface), whereas the remainder are essentially randomly oriented. Comparisons with the theoretical predictions are consistent with a model in which the preferred direction in the apical region of the cornea does not vary significantly among rabbits; and the preferred lamellar orientation direction determined for the population measured here is very close to that suggested in previous experiments on a smaller number of rabbits. Mapping experiments using a new goniometric holder showed that the preferential order at the apical region is closely preserved throughout the central approximately 4 mm diameter optical zone in individual corneas.

Identification of SLF1 as a new copper homeostasis gene involved in copper sulfide mineralization in Saccharomyces cerevisiae

In Saccharomyces cerevisiae, at least 12 genes are important for cells to propagate in medium containing elevated concentrations of copper salts (J. Welch, S. Fogel, C. Buchman, and M. Karin, EMBO J. 8:255-260, 1989). Complementation studies were carried out on a copper-sensitive mutation (cup14) from this group. A new yeast gene, designated SLF1, was identified as a multicopy suppressor of the cup14 mutation. Slf1 is important for the physiological process of copper sulfide (CuS) mineralization on the surface of cells cultured in medium containing copper salts. CuS mineralization causes the cells to turn brown. Disruption of SLF1, which is located close to the telomere region of chromosome IV, leads to limited copper sensitivity, and the resulting cells lack the normal brownish coloration when grown in CuSO4-containing medium. Overproduction of Slf1 in wild-type cells confers superresistance to CuSO4 and enhances the coloration of cells cultured in the presence of CuSO4. Upon addition of KCN to Cu-grown cells, the brownish coloration was bleached instantly, and copper ions were solubilized. These data are consistent with Slf1-dependent accumulation of CuS complexes on the cell surface. Disruption of SFL1 also results in loss of the ability of yeast cells to deplete Cu but not Cd ions from the growth medium, whereas overexpression enhances Ca depletion ability and the resulting deposition of CuS particles. It is proposed that Slfl participates in a copper homeostasis pathway, distinct from the Cup1 detoxification system, that leads to sulfide generation and CuS biomineralization on the cell surface. This process may coordinate with the Cup1 pathway at different copper concentrations to prevent copper-induced toxicity.

A numerical test of the normal incidence uniaxial model of corneal birefringence

The suggestion that the central cornea can be modeled as a uniaxial birefringent material with its optic axis normal to the surface is explicitly tested by numerical calculations. A theoretical framework is presented to model the corneal stroma as a series of stacked, uniaxial birefringent layers (lamellae). Calculations are then made of the transmission of normally incident, linearly polarized light through model systems having various azimuthal orientations of the layers, motivated by the suggestion of an overall "random" organization of the stromal lamellae. It is concluded that the uniaxial description, and the assumptions upon which that description is based, do not hold for the cornea. In particular, the calculations are in agreement with recent experiments in which one always observes a non-zero cross-polarized transmission (hence birefringence) at normal incidence.

Identification of the Zn(II) site in the copper-responsive yeast transcription factor, AMT1: a conserved Zn module

The N-terminal metal-binding domains of the copper-activated yeast transcription factors, ACE1 and AMT1, bind to specific DNA sequences in a Cu-dependent fashion. Recombinant AMT1 and ACE1 metal-binding domains are isolated as Cu4Zn1-protein complexes. Site-directed mutagenesis of AMT1 was used in this study to map the ligands of the Cu(I) and Zn(II) ions. The results are consistent with the N-terminal halves of AMT1 and ACE1 consisting of two independent submodules, one binding a single Zn(II) ion and the second binding the tetracopper cluster. The basis of this conclusion is, first, that mutations of two cysteinyl codons and a histidyl codon in the first 42 residues of AMT1 do not alter DNA binding. In contrast, serine substitutions at four cysteine positions at codons 43, 61, 90, and 98 abolish DNA binding. We demonstrated previously that population of the Zn(II) site in AMT1 does not alter the ability of the protein to bind DNA but bound Cu(I) ions are essential for DNA binding [Thorvaldsen, J. L., et al. (1994) Biochemistry 33, 9566-9577]. Second, mutations in the N-terminal 42 residue segment reduce the Zn(II) content of purified mutant AMT1 molecules. Third, a synthetic peptide consisting of the N-terminal 42 residues in AMT1 forms a stable Zn(II) complex and substitution with Co(II) reveals an electronic spectrum identical to that of the Co-substituted intact Cu4AMT1 protein. 113Cd(II) NMR studies reveal that the divalent metal site consists of ligands provided by three cysteinyl thiolates and a single histydyl imidazole. The sequence homology between AMT1, ACE1, and MAC1 in the N-terminal 42 residues suggests that ACE1 and MAC1 will, likewise, contain N-terminal Zn modules. A 42-residue ACE1 synthetic peptide gives identical metal binding properties to the corresponding AMT1 synthetic peptide. Thus, AMT1 and likely ACE1 consist of two contiguous modules, residues 1-42 forming an independent Zn(II) module and residues 43-110 enfolding a tetracopper cluster.

Numerical modeling of the cornea's lamellar structure and birefringence properties

A model of the cornea's lamellar structure is proposed that is capable of explaining experimental results obtained for the transmission of normal-incidence polarized light through rabbit and bovine cornea. The model consists of a large number of planar lamellae, each approximated as a uniaxial birefringent layer, stacked one upon another with various angular orientations. Polarized light transmission through the composite system is modeled theoretically by use of the Jones matrix formalism. The light transmission is calculated numerically for a large number of model lamellae arrangements, each generated from a statistical description, and histograms are constructed of various properties of the light transmission, including the minimum and maximum cross-polarized output intensities. It is demonstrated that various structural and optical parameters of the lamellae arrangements of actual corneas may be estimated by comparison of the calculations with detailed experimental data. Certain characteristics of the histograms are identified that permit a clear distinction between random and partially ordered systems. Comparisons with previously published experimental data provide strong evidence that the lamellae orientations are not entirely random, but rather a significant fraction are oriented in a fixed, preferred direction.

Ultrastructure in anterior and posterior stroma of perfused human and rabbit corneas. Relation to transparency

PURPOSE

The authors sought to discover whether there are differences in the degree of spatial order in the fibrillar ultrastructure between anterior and posterior stroma.

METHODS

Human corneas were obtained from eye bank eyes. Although they had been classified as normal, some swelling remained after 3 hours of deturgescence. Freshly excised, unswollen rabbit corneas also were used. Image analysis methods were applied to transmission electron micrographs of the anterior, middle, and posterior stroma of these corneas to determine the positions and radii of fibrils, the fraction of total area occupied by fibrils, and the fibril number density. Results were used to calculate the interference factor that appears in the direct summation of the fields for light scattering theory and to estimate the total scattering cross-section per fibril. The interference factor is a measure of the spatial order in the positions and sizes of the fibrils.

RESULTS

Electron micrographs showed anterior-posterior variations in size and number density of fibrils. The interference factor at wavelengths of visible light was lower in posterior stroma than in anterior stroma for humans and rabbits. In some instances in humans, the anterior interference factor was characteristic of mildly swollen cornea. When averaged for the electron micrographs analyzed, the anterior stroma was predicted to scatter approximately twice as much light per unit depth as the posterior stroma in humans (at any given wavelength) and approximately three times as much in rabbits.

CONCLUSIONS

Calculations of the interference factor showed that there were differences in the anterior-posterior spatial ordering of fibrils. In human corneas, the differences could have been caused by intrinsic in vivo differences between anterior and posterior stroma; however, possible anterior-posterior variations in swelling between the two regions in vitro also could have affected the results.

Validity of pulsatile ocular blood flow measurements

The determination of average net pulsatile ocular blood flow form measurements of the intraocular pressure (IOP) pulse is based upon 1) an accurate measurement of IOP and its time variation; 2) a knowledge of the relation between the volume of the living eye and its IOP; 3) a physical model for the flow of blood through the eye: and 4) a concept of steady venous outflow from the eye. Each of these premises needs to be examined. The present analysis assesses the validity of the pneumatic tonometer for measuring the pressure in a flowing column of gas that is directed toward a thin membrane that is in contact with the surface of the cornea. The pressure of this stream of gas exerts a force against the cornea that depresses the corneal surface against the opposing force of the IOP. The balance of these forces and the resultant effect on the tonometer pressure sensor is described by the theories of elasticity, thermodynamics and fluid mechanics. In this treatment, differential equations are solved for the elastic deflections of the cornea subject to the opposing intraocular and tonometer pressures. This permits the pressure in the chamber of the pneumatic tonometer to be related through first principles to the IOP. The response of the tonometer pressure sensor to a range of IOPs (5-60 mmHg) is obtained, as well as the dynamical response of the tonometer to IOP oscillations. The conclusion is that the pneumatic tonometer provides a high fidelity, noninvasive measure of the IOP and its time variation.(ABSTRACT TRUNCATED AT 250 WORDS)

Image processing of electron micrographs for light scattering calculations

Theoretical light scattering calculations, when applied to models based on structures seen in electron micrographs of corneal stroma, require knowledge of the constituent fibril positions and radii. Obtaining this information manually is a difficult and time-consuming task. In order to facilitate this problem, we present a simple and flexible computer algorithm that allows the process to be automated using a Macintosh computer. The accuracy of the results is checked by comparing light scattering calculations using fibril positions and radii found manually with those found by the computer. The results show that the computer algorithm is a viable and accurate means of obtaining the needed data.

Effects of metallothionein on the observed copper distribution in cell extracts

Systematic studies have been undertaken to compare the effects of cell lysis and chromatography conditions on the observed distribution of Cu amongst Cu-binding proteins in cultured cells. The variables included rate of centrifugation, presence or absence of non-ionic detergent, and presence or absence of dithiothreitol. The application of an improved FPLC gel filtration system has permitted us to examine the effects of the addition of exogenous metallothionein (MT) to cell extracts. When the cell extract contains low levels of endogenous MT, the addition of MT in the presence of dithiothreitol causes a shift of copper to the MT peak. High levels of MT can therefore remove copper from other Cu-binding ligands during cell homogenization, hence producing artifactual Cu distribution results. The use of an anaerobic buffer system has greatly reduced the observed level of Cu exchange, and has allowed comparison of Cu distribution in normal cells and cells from patients with Menkes' disease.

Cornea epithelial damage thresholds in rabbits exposed to Tm:YAG laser radiation at 2.02 microns

We have determined exposure conditions for minimal damage to the corneal epithelium in rabbit using a continuous wave Tm: YAG laser operating in the TEM00 mode at incident powers between 80 and 450 mW. The 1/e beam diameter was 0.94 mm and exposure durations for threshold damage ranged from 4.3 to 0.08 sec. Calculated temperature increases on the beam axis 10 microns beneath the surface at the measured thresholds were essentially constant and averaged 44 degrees C. This is basically the same temperature increase found for threshold CO2 laser damage and suggests that the critical temperature damage model, which correlates CO2 laser damage, can predict damage thresholds for mid-infrared laser radiation. We also showed that reliable thresholds can be determined in freshly enucleated eyes, thus opening up the possibility of using available laser sources in laboratories not equipped and approved for animal experiments to determine damage thresholds.

Epithelial damage in rabbit corneas exposed to CO2 laser radiation

Corneal injury thresholds are determined for conditions not previously explored for CO2 laser radiation, including multiple-pulse exposures and a systematic investigation of the effect of beam diameter on single-pulse damage thresholds. Multiple-pulse exposures from pulse trains up to 999 pulses, having pulse repetition frequencies between 1 and 100 Hz and individual pulse durations between 10(-3) and 0.5 s, were explored. Damage thresholds are discussed in terms of an approximate critical temperature model, the damage integral model and other empirical correlations. Single-pulse exposures are accurately correlated by an empirical critical temperature model in which the critical temperatures have a weak dependence on exposure duration. However, certain aspects of the single-pulse damage data led us to propose a new thermal damage model that incorporates an endothermic phase transition as the damage mechanism. This physical model accurately correlates single-pulse damage for exposures between 10(-3) and approximately 10 s.

Effects of fibril orientations on light scattering in the cornea

It is generally assumed that the collagen fibrils in the stroma are the primary scatters of light in the nearly transparent cornea of the eye. We derive a scaling relationship between scattering angle and light wavelength that should hold if this hypothesis is valid. The derivation accounts for the cornea's layered nature and the azimuthal orientations of the fibrils in the different layers. The fibrils are treated as finite-length cylinders, and the scaling relation is obtained in both the far- and the intermediate-field zones. The predicted relationship is verified experimentally for normal-thickness rabbit corneas.

Stromal damage in rabbit corneas exposed to CO2 laser radiation

Threshold damage to the cornea from CO2 lasers is confined to the epithelium. Exposures well above the threshold for epithelial damage produce bowl-shaped stromal wounds. Light and electron microscopy and slit-lamp photographs all show a sharp demarcation between the damaged and undamaged regions 48 hr after exposure. The micrographs also show that the damaged region is accellular. Calculations of the expected temperature increases combined with analyses of slit lamp photographs show that the wound boundary corresponds to a surface of equal peak temperature increase. Comparisons with epithelial and endothelial damage conditions suggest that stromal, endothelial and epithelial cells have essentially the same thermal damage mechanism.

Corneal small-angle light-scattering theory: wavy fibril models

Small-angle light-scattering (SALS) measurements of the cornea together with electron micrographs of the corneal stroma suggest that the waviness in the stromal collagen fibrils of corneas fixed at zero pressure is the structural feature responsible for the cross-polarized SALS patterns. This paper derives and discusses a Born approximation to the parallel- and cross-polarized SALS patterns expected from lamellae of long, thin, optically anisotropic wavy fibrils whose axes are parallel to each other and are spatially distributed about one another in a quasi-ordered fashion. The predicted scattered intensity depends on three factors: (1) the fibrils within a given lamella wave in unison, which produces scattering that is characteristic of a wavy sheet (as opposed to that characteristic of an isolated wavy fibril); (2) the undulations lead to a diffraction condition for determining the dependence of scattered intensity on scattering angle; (3) the relative orientations of fibril axes in different lamellae and the intrinsic electric susceptibility of a fibril determine the dependence of scattered intensity on azimuthal angle. The patterns predicted for anisotropic fibrils with a random distribution of lamella orientations or with distributions of lamella orientations that have one or two preferred directions superimposed upon a random background agree with the qualitative features of the experimental patterns observed with rabbit corneas. Experimental evidence in support of the distributions with preferred orientations is discussed.

On the interpretation of depth dependent light scattering measurements in normal corneas

In recent studies other investigatiors have presented traces of "light scattering" versus depth into the cornea. The present investigation demonstrates that these traces do not measure the actual light scatter intensity and that a proper interpretation of such traces shows that most of the scattering is from within the stroma.

On corneal transparency and its loss with swelling

The cornea is the clear front covering of the eye through which we see and is composed of collagen fibrils embedded in an optically homogeneous ground substance. It has long been recognized that these fibrils scatter light and that transparency results from interference effects due to an ordering in the spatial arrangement of the fibrils about one another. The nature of this ordering and of its disruption in abnormal corneas is of great current interest. The present study reviews experimental light scattering and electron microscopic evidence relevant to this problem. This evidence suggests that theoretical attempts to model and understand corneal transparency and its loss during swelling, in terms of ultrastructure, must account for a short-ranged ordering of fibrils in normal healthy corneas and for the formation of regions void of fibrils in swollen corneas.

Wave-length dependencies of light scattering in normal and cold swollen rabbit corneas and their structural implications

1. The studies described herein involve the use of light scattering measurements to characterize the ultrastructural arrangement of the constituent collagen fibrils in rabbit corneal stromas.2. Theoretical light scattering techniques for calculating the scattering to be expected from the structures revealed by electron micrographs are discussed, and comparison with the experimental light scattering tests the validity of these structures.3. The wave-length dependence of light transmission and of angular light scattering from normal corneas is in agreement with the short range ordering of collagen fibrils depicted in electron micrographs.4. The transmission measurements on oedematous rabbit corneas indicate that transmission decreases linearly with the ratio of thickness to normal thickness.5. The wave-length dependence of transmission through cold swollen corneas indicates that the increased scattering is caused by large inhomogeneities in the ultrastructure. Electron micrographs do, indeed, reveal the presence of such inhomogeneities in the form of large regions completely devoid of fibrils.

The transparency of the mammalian cornea

1. A theoretical and experimental analysis of the relationship of the corneal stromal ultrastructure with light transmission has been made in an attempt to resolve recent contradictory explanations of corneal transparency.2. The spatial distribution of collagen fibrils in electronmicrographs of rabbit corneal stroma has been analysed in terms of a radial distribution function. The results indicate the presence of local order extending to at least 200 nm from individual fibrils.3. The observed spatial distribution of the collagen fibrils was used as a basis to compare the theoretically derived and the experimentally determined values of light transmission. It has been found that the transparency of the normal cornea may be explained by the quasi-random structure revealed by the electronmicroscope.4. Histograms of the collagen fibril diameter in normal rabbit corneal stroma revealed the range to be 12.5-32.5 nm and the mean value to be approximately 20 +/- 1.5 nm. Corneal swelling did not change the collagen fibril diameter significantly.5. It is concluded that the size and distribution of collagen fibrils revealed in electronmicrographs are consistent with the observed transparency of normal stromas.6. A marked heterogeneity in the spatial distribution of collagen fibrils was found in the swollen cornea. This is qualitatively consistent with the observed decrease in transparency.