In response to:

Hypophosphorylation of Aqueous Humor sCD44 and Primary Open-Angle Glaucoma
Paul A. Knepper, Adam M. Miller, John Choi, Robert D. Wertz, Michael J. Nolan, William Goossens, Susan Whitmer, Beatrice Y. J. T. Yue, Robert Ritch, Jeffrey M. Liebmann, R. Rand Allingham, John R. Samples
Investigative Ophthalmology & Visual Science. August 2005, Vol.46, 2829-2837. doi:https://doi.org/10.1167/iovs.04-1472

LETTER TO THE EDITOR

We read with interest the article by Knepper et al.¹ on "Hypophosphorylation of Aqueous Humor sCD44 and Primary Open Angle Glaucoma." While the findings regarding altered sCD44 in glaucomatous patients are intriguing, we have concerns about the claim made in the article that "increased pressure influenced the HA [hyaluronan] polymer and binding sites." To draw this conclusion, the investigators examined the effects of pressure on the binding of soluble CD44 (sCD44) to hyaluronan (HA), and on the conformation of hyaluronan as visualized by rotary shadowing. They carried out experiments at either atmospheric pressure or at a pressure 40 mmHg above atmospheric, conditions that they refer to as "0 mmHg" and "40 mmHg."

It should be recognized that the pressure applied in these experiments was a hydrostatic pressure applied to molecules in a rigid container, which is a fundamentally different situation from the case of a pressure difference occurring across a deformable substrate. In the case of the eye the relevant pressure difference is the intraocular pressure (IOP), i.e. the difference between pressure inside the eye and atmospheric pressure. In the eye, IOP (or more correctly, the difference between IOP and episcleral venous pressure or between IOP and retrolaminar cerebrospinal fluid pressure) can result in significant mechanical strains on the tissues and macromolecules of the eye (in particular, the trabecular meshwork and the optic nerve), while the same value of IOP acting in a rigid container would have little effect. A simple estimate helps put this distinction into perspective. We know from experimental observations that a pressure difference between IOP and episcleral venous pressure of 15-50 mmHg in live monkey and enucleated human eyes causes the trabecular meshwork to stretch very significantly.² On the other hand, by using published values of the compressibility of water, proteins and polysaccharides,^3,4 we can estimate that a 50 mmHg increase in the hydrostatic pressure applied to these molecules in a rigid container will lead to deformations of these molecules less than 0.005%. In general, hydrostatic pressures applied to molecules in rigid containers result in miniscule deformations unless the pressures are many, many atmospheres. We are aware of no studies in the literature that show that the miniscule deformations expected in the study of Knepper et al. can give rise to significant configurational changes in macromolecules, as is claimed by the authors.

The extraordinary nature of the claim made in this article can be appreciated in another way. We can estimate the free energy change of the hyaluronan due to a hydrostatic pressure increase of 40 mm Hg. Using compressibility values for polysaccharides,^3,4 allowing that the density of hyaluronan is approximately 0.65 g/ml,^5,6 and using a HA molecular weight of 1x10⁶, we can calculate that the work done in compressing the hyaluronan as the hydrostatic pressure increases by 40 mm Hg is roughly 1 calorie/mole. This must equal the change in free energy of the HA molecule. On the other hand, typical binding energies such as would be associated with binding CD44 to HA are on the order of kilocalories/mole, i.e. thousands of times larger. Hence, it is again difficult to understand how this very modest hydrostatic pressure change would have any effect on binding of CD44 to hyaluronic acid. (Note also that the calculation done here is very conservative, since we used a molecular weight of 1x10⁶ for the entire HA molecule whereas only a small part of the HA molecule is actually involved with binding the CD44).

While we find the images and data presented to be interesting, such a remarkable claim about the effects of small changes in hydrostatic pressure on physicochemical properties of macromolecules calls for extraordinarily convincing experimental support. We were not convinced by the data for several reasons. First, it is important to appreciate that the pressure experienced by an object (including the samples used in this study) is the sum of atmospheric pressure (nominally 760 mmHg) and any additional pressure imposed by the investigators. It is not indicated in the article what the atmospheric pressure was on those days that the measurements were made. As the magnitude of the atmospheric pressure can change daily by 10-20 mm Hg (and by as much as 50 mm Hg or more during storms), it is unclear what the total pressure was in the experiments that were being investigated. Were each of the replicate experiments done on days of identical atmospheric pressure?

Also of concern was the lack of reported statistical variability for the data presented in Figure 6 or for the images presented in Figure 7. Repeatability is an important issue in light of the discussions about atmospheric pressure changes above.

It is also worth pointing out that what might appears to be a fairly large difference in pressures (0 mmHg vs. 40 mmHg) is actually a difference of only 5% in total pressure (nominally 760 mmHg vs. 800 mmHg). This is rather modest compared to the pressure variations associated with traveling between locations having different elevations. For example, in Denver (Denver International Airport elevation 5,431 ft [1,655 m]), the ambient pressure is approximately 620 mmHg, while in Sedom, Israel (elevation 1275 ft [390 m] below sea level) the ambient pressure is about 795 mmHg. Is it reasonable to expect that molecular binding processes vary significantly depending on the elevation of the city that one lives in? Perhaps more tellingly, if absolute pressure rather than IOP mattered in glaucoma, then one might expect Denver to be glaucoma-free, which it is not.

Can physiological systems sense and respond to pressures? Absolutely, but they invariably do so by transducing the pressure-induced stretching of a deformable substrate. An example is the carotid baroreceptor that responds to stretching of the carotid artery wall by changes in systemic blood pressure. The details of the pressurizing system used by Knepper et al. were only briefly described, but no such deformable substrate seems to have been included. Unfortunately, changing total pressure can change other experimental variables (e.g. evaporation rate in the rotary shadowing preparations and dissolved gas concentration in samples) that could have an effect on the results. It is hence possible that some of the observed differences between the "0 mmHg" and "40 mmHg" states are secondary effects. But it seems most unlikely that total pressure can explain the effects reported in Figures 6 and 7 of this manuscript.

Finally, we raise one other aspect regarding the reported toxicity of sCD44 that we found puzzling. We note that the concentrations of sCD44 found to be toxic to TM cells in this study, and a previous study by this group,^1,7 were 450-fold smaller than the concentration that human vascular endothelial cells are normally exposed to from blood plasma.⁷ It would have been useful to have used vascular endothelial cells as a control to substantiate this toxicity observation.

C Ross Ethier¹
Mark Johnson²

¹Institute of Biomaterials and Biomedical Engineering, University of Toronto, Ontario, Canada
² Department of Biomedical Engineering, Northwestern University, Evanston, Illinois, US

References

1. Knepper PA, Miller AM, Choi J, Wertz RD, Nolan MJ, Goossens W, Whitmer S, Yue BYJT, Ritch R, Liebmann JM, Allingham RR, Samples JR. Hypophosphorylation of aqueous humor sCD44 and primary open-angle glaucoma. Invest Ophthalmol Vis Sci. 2005;46:2829-2837.
2. Johnstone MA, Grant WM. Pressure dependent changes in the structures of the aqueous outflow system of human and monkey eyes. Am J Ophthalmol. 1973;75:365-383.
3. Heremans K. High pressure effects on proteins and other biomolecules. Annu Rev Biophys Bioeng. 1982;11:1-21.
4. Gekko K, Noguchi H. Hydration behavior of ionic dextran derivatives. Macromolecules. 1974;7:224-229.
5. Varga L. Studies of hyaluronic acid prepared from the vitreoius body. J Biol Chem. 1955;217:651-658.
6. Silpananta P, Dunstone J, Ogston AG. Fractionation of a hyaluronic acid preparation in a density gradient: some properties of the hyaluronic acid. Biochemical J. 1968;109:43-50.
7. Choi J, Miller AM, Nolan MJ, Yue BY, Thotz ST, Clark AF, Agarwal N, Knepper PA. Soluble CD44 is cytotoxic to trabecular meshwork and retinal ganglion cells in vitro. Invest Ophthalmol Vis Sci. 2005;46:214-222.