Water Structure Changes in Dilute Solutions of Physiologically Active Substances
Kenneth S. Marsh, Ki Brissey, Su-il Park
Abstract
100 samples of dilute solutions of physiologically active substances were prepared in water and a vicinal water solution which simulated cellular interfaces. The solutions were then scanned in a medical MRI. The results demonstrate that dilute solutions of these substances yield a wide range of response suggesting that the water protons are experiencing variations in the water environment. The vicinal solutions exhibited more contrast than the corresponding bulk water solutions. The suggestion is that physiologically active substances modify water structure, especially interfacial water, and that this influence on water structure is part of their physiological effect.
Introduction
This experiment is designed to investigate the hypothesis that pharmacological action is accomplished, at least in part, through water structure changes. A necessary (but not sufficient) manifestation of this hypothesis would be observance of experimental changes in measures of water structure through the influence of low concentrations of pharmacologically active substances.
It is known that theorganization of molecules in the water matrix changes in space and time and can be modified through interactions with macromolecules, ions, and other chemical compounds. The exact nature of these changes is not known, and includes numerous forces including hydrogen bonding, van der Waals forces and dipole effects. For lack of a better term, we will employ “electronic” interactions to imply influences of the electron cloud. The existence of the different environments is exhibited in Magnetic Resonance (MR) spectra. T1 relaxation times vary with the special and electronic environment surrounding water protons, and are therefore used to present differences in that environment, both in chart form (nuclear magnetic resonance spectra) and graphic form (MR imaging or MRI). The sum total of all interactions on the water protons effects its ability to realign with the magnetic field after the radio frequency pulse is removed, which presents subtle changes in the water environment with optical density differences in the MRI. We will refer to these different environments collectively as water structure with no implication that we know the configuration of these structures. We will, however, demonstrate that water structure is influenced by very low concentrations of physiologically active substances.
Background
Michael Tracey (1964) determined that pharmacologically active substances which are known to be stimulants acted upon the rheology of wheat dough systems as if additional water had been added to the system during dough development. Depressants acted on the dough as if water had been removed during dough development. These actions were obtained with drugs at physiological levels, i.e. levels below which colligative effects would be anticipated. Tracey defined stimulants as water structure breakers and depressants as water structure makers.
Although not specifically stated in Tracey's study, the wheat gluten experiment represented a system characterized by interfaces. The interfacial effects of water have been well studied by Walter Drost-Hansen who defined a series of anomalous interactions which occur at interfaces. The term "vicinal water" describes water in proximity to a surface, which was found by Drost-Hansen to be any substance with a molecular weight exceeding 3000-5000 Daltons. The occurrence of these anomalous behaviors in the vicinity of a surface, regardless of the surface composition, was described by Drost-Hansen as the paradoxical effect. He went further to suggest that most, if not all, of the water in cells would be sufficiently close to a sufficiently large molecular weight entity to act vicinally.
Plant systems cannot promote survival by a fight or flight response typical of the animal kingdom. As a result, plants must combat environmental and other attacks chemically. Plant systems have evolved a cornucopia of chemical agents which have beneficially impacted their survival. Humans have extracted many plant alkaloids and found them to be physiologically active in animals (including humans) and are the basis for many drugs.
If water structural changes are vital to life processes (as described by Drost-Hansen) and modifications of water structure are implicated in drug action (as suggested by Tracey), then it follows that drugs should exhibit modifications of water structure. This paper is designed to study this premise.
Our intent is to study MR images of very dilute drug concentrations in both standard and vicinal solutions.
Experimental Design
The concept under test is that physiologically active substances in low concentration will demonstrate a change in water structure as measured through MRI. Since the biological systems exist in a vicinal environment, vicinal and bulk water environments will be employed.
Sample Choice
Physiologically active substances are compounds which demonstrate effects on life processes at low concentrations. Stimulants, depressants, humectants, chemotherapeutic agents, vitamin C, hormones, and organic acids were chosen for study. Stimulants and depressants were chosen to expand on the work of M. Tracy. Vitamin C was included because of previous work by the WIS2H Institute (unpublished) which found MRI scan of vitamin C to reveal an unexpected change in water structure as indicated by a reduced T1. Additional organic acids were included to determine if the dramatic response of ascorbic acid is an effect of unusual properties of ascorbic acid itself, or simply a response to characteristics of an organic acid.
Humectants are substances which modify water holding capacity. Humectants were included to obtain an alternate measure of humectant activity through MRI and also to test substances known to have a direct effect on water. The data will also provide a forum for further studies which relate water structure and humectant activity to biological reactions and growth. Since water control has been a traditional means to preserve food products by limiting microbial growth, microbial growth studies are planned.
Chemotherapeutics were included to explore the hypothesis from WIS2H that changes in water structure must precede malignant (or any) growth. A random effect on optical density of the MRI would suggest that water structure influence is not a factor in chemotherapeutic response; any non-random behavior would provide evidence to support the hypothesis, with degree of non-random behavior a factor. Water structural changes are necessary, but not sufficient for proving the hypothesis. Therefore this work is exploratory.
Cortisone and progesterone were included because hormones have greater influence (in terms of concentration effects) than most drugs, and also influence water balance in the body at times of stress.
Solution Preparation
Much literature suggests that interfacial water behaves differently from bulk water (Drost-Hansen & Clegg, 1979; Franks, 1972 – 1979; Pollack, 2001). Therefore, for the purposes of this study, both water and interfacial water systems were chosen. The interfacial system was modeled by employing a polymer system with a molecular weight greater than 10K Daltons, which was suggested by W. Drost-Hansen () as above the limit for "vicinal" or interfacial differences. Polyvinylalcohol (PVOh) was chosen because it is water soluble, contains a relatively simple and non-reactive chemical structure, and possesses a sufficiently large molecular weight. The molecular weight of synthetic polymers, unlike proteins, is not a specific value but a range. DuPont Elvanol, Grade 90-50 was chosen for this study. Elvano 90-50 was found to posses a molecular weight of 29,700 Daltons by the numerical method (which emphasizes the lower MW components) and 55,800 Daltons by the weight method (which emphasizes the higher MW components), which is considerably larger than the minimum required for vicinal effects. The concentration of PVOh was calculated to allow 50 water molecules for each repeating unit of the polymer matrix, specifically -(CH2-CHOH)-. This structure possesses a molecular weight of 44, so 44 grams of PVOh were used for each 18 MW * 50 water layers or 900 grams of water. The water in this polymer solution, therefore, will all be statistically well within 50 molecular distances of a polymer chain.
Physiological Dosage
Each drug was prepared in standard dosage forms and equi-molar concentrations into both water and PVOh solutions. The "standard" dosages were obtained through dosage recommendations in product literature or through the Physicians Desk Reference (PDR). These dosages are typically the available pill sizes for the drugs. The dosages were converted into concentrations through the following procedure. Dosages are typically calculated in terms of mg/kg body weight. For convenience, manufacturers prepare standard dosages by calculating the dosage required for an average person, typically a 150 pound male. An average built 150 pound male has a blood volume of 4.77 liters (Luisada, 1959). With a simplifying, working assumption in which the active ingredient is uniformly dispersed in the bloodstream (and not in tissue), we applied the pill dosage into 4.77 liters for our “physiological” concentrations. It is recognized that some drugs are metabolized into an active form. Such mechanisms are not incorporated into this study.
The next step was to choose a mean concentration, with the absolute realization that no such meaningful average drug dosages exists. Different drugs act at different concentrations, and any particular value would be a gross over-simplification of drug action. Furthermore, the above concentrations are based upon dosage, not molecular weight. However, the purpose of this study is to compare action on water structure systems, as defined by MRI, and equal concentrations are desired for direct comparison. Molar concentrations were calculated for each test substance and a median value was chosen for the experimental protocol and rounded to concentration of 0.0005M.
The resulting protocol consisted of 19 physiologically active compounds, each in standard dose and a concentration of 0.0005 molar, and each in water and PVOh solution. Some compounds were tested with various concentrations to represent different dosage levels. Pure water and PVOh solutions were incorporated into the test. A total of 100 samples plus four index controls (distilled water) were scanned.
Many of the compounds were available in pure form (sugars, caffeine, organic acids, glycerine). Analgesics and pharmaceuticals were requested in pure form. Those which could not be obtained pure (Vioxx, Phenobarbital, some chemotherapeutic agents) were used in pill form. The pill was ground in a morter and pestle, and a proportionate quantity was weighted with the same ratio of dosage to pill weight in order to compensate for the weight of the binders.
Experimental Procedure
A stock “vicinal” solutions was prepared by dissolving 44 grams of Elvanol 90-50 into 900 ml of distilled water. Distilled water from the same source was use for the “bulk” water samples. Appropriate weights of each active substance were added into both distilled water and vicinal water to prepare the solutions of the correct concentrations. Each test solution was placed into a 50 ml graduated conical polypropylene tube with cap (Becton Dickinson Bluemax 2098) in an expanded polystyrene rack. Each rack held 25 tubes. To assure that the films were not inadvertently reversed during viewing, a 1/2" diameter polypropylene test tube was filled with distilled water and placed in one corner of the rack. This tube provided a second distilled water standard and served as an index for reading the films.
The racks were placed on an expanded LDPE base (Dow Ethafoam 220), which was cut to a rectangular block of 10.5 by 6.125 by 3 inches high. The bottom long edges were optimally radiused to 1.25 inches to allow the base and rack to correctly position the sample tubes across the horizontal diameter of the Atlas Head/Neck Vascular Phased Array Coil. However, truncating the edge with a straight 45 cut, 1.25 inches horizontally and vertically from the bottom edge was proven adequate. The entire rack system was scanned in a Polaris 1.0T MRI Magnetic Imaging system through the cooperation of the Department of Radiology at Oconee Memorial Hospital. Four sets of 25 samples were tested.
The first set was scanned for T1 weighed, T2 weighed, Proton Density weighted, and dual T2 /Proton Density weighted responses. The greatest contrast differences between samples was observed with the T1 weighted protocol so T1 was chosen for the study. The differences in T1 were reflected as differences in optical density of the individual images.
The resulting cross-sectional circles from each tube on the film (coronal slice, representing each sample) were evaluated for optical density. Two densitometers were tried, but it was found that the fine structure in the images yielded unacceptable variance in the densitometer readings. An integration over the image was required, but not available through the optical system of either densitometer. Visual inspections were therefore performed in which relative optical densities were evaluated before the identity of the samples was known to the evaluator. Comparative densities were recorded on a chart which included circles representing each sample, and a place to record the image and plate number from the MRI film.
A quantitative evaluation of optical densities was obtained through a colorimeter (Minolta Chroma Meter CR-3000), which provided a larger sample area, and therefore integrated the density for each sample. The "brightness" was obtained through the L-value of the CIE L* Color Space.
The MRI scans of the samples exhibited a non-uniform optical density even though the samples were homogeneous. In order to investigate if the cause was a dielectric effect of polypropylene, four samples of PVOh were incorporated into the fourth experimental set in glass rather than polypropylene holders.
Results
The results of the MRI scans are presented in Figures 1 - 4, with the densitometer values for each sample presented in Tables 1 - 4. Controls of pure distilled water and PVOh solution are included. Table 5 presents the four scans in a single table, ranked by increasing image density. The 100 samples scanned represented concentrations of physiological substances from 5.8 x 10-6 to 2.3 x 10-3 Molar and exhibited optical densities ranging from 66.47 to 26.75, with higher densities representing a darker image.
Discussion
The optical density, both visually (Figures 1 - 4) and numerically (Tables 1 - 4) illustrate that relaxation times of water protons are influenced by low concentrations of pharmacologically active substances. Observations must be taken as tendencies rather than absolute comparisons for a number of reasons. First the field density in the MRI is not uniform. The center of the coil shows brighter images than the periphery. With medical images, the contrast relationships provide valuable information for diagnoses, but the optical densities vary across the film for similar water environments. Therefore, our quantification of optical densities must be evaluated in this context. Direct T1 measurements could have been obtained on a NMR spectrophotometer, which was not available for this study. Proper shimming of the magnet would also reduce the varience.
Set 1
The T1 weighted images exhibited larger contrast between the samples than T2 weighted or proton weighted images. With T1 weighting, the brightest images appeared to be vitamin C in PVOh for the 250, 500 and 1000 concentrations. The vitamin C 1000 sample appeared light, but darker than the other vitamin C in PVOh. The next noticeable group was tartaric acid in PVOh (both 500 and 1000) and caffeine in PVOh (both 50 and 200). The next darker group is citric acid 60, 250, 500 in PVOh, with the citric acid 1000 again slightly darker. Next is vitamin C in water , then citric acid in water, caffeine 200 in water, tartaric 500 and 1000 in water, and final solution is caffeine 50 in water. The pure PVOh sample appears approximately equivalent to vitamin C 500 in water. Water in pure form was darker, but more so with the sample in the smaller tube near the extent of the field. Pure water, as expected, exhibited a longer relaxation time, suggesting that samples are essentially contributing some structure in the water environment..
Ascorbic acid seemed to have a stronger effect on water structure than tartaric and citric acids. This is unlikely to be strictly a manifestation of properties of an organic acid because the Pka values are comparable. Molecular weights are comparable (ascorbic acid, citric acid and tartaric acid have molecular weights of 176.12, 192.12 and 150.09 respectively) but the configuration of the ascorbic acid appears capable of a stronger influence on the surrounding water. This supports the hypothesis of a relationship between water structure and physiological activity.
Our studies exhibited action of caffeine which agrees with Tracey’s work, i.e. demonstrated evidence of a water structure maker. Although the Phenobarbital was considerably less active in the MRI, it still reduced the optical density in the MRI compared with pure water. This finding does not appear to agree with Tracey’s observation on first look. Three factors complicate the observation. First, the caffeine was available in pure form but the phenobarbital was only available in pill form, with the standard binders and sugars. These additional components could alter the observations. Secondly, water samples in different locations exhibited different responses because of field effects in the MRI. One water sample was more active than the Phenobarbital; one was less active. Thirdly, the observed differences should be related to vicinal water. Although Tracey did not state this directly, most or all of the water in a wheat gluten dough systems would be expected to be vicinal. Therefore, the standard for comparison should be with vicinal water, not bulk water.) Hazlewood et al. (1974) reported that at least three fractions of water exist in cells, all of which exhibit faster relaxation times than bulk water. Another interpretation, therefore, is that wtare-structure breaking and making is relative to a value other than that of bulk water – further support for the vicinal water explanation.