J. Membrane Biol. 64, 217-224 (1982) The Journal of Membrane Biology Correlation between Lipid Partition Coefficients and Surface Permeation in Schistosoma japonicum Eain M. Cornford Southwest Regional Veterans Administration Epilepsy Center, Veterans Administration Medical Center, Wadsworth, Research Service, Veterans Administration Medical Center, Brentwood, Los Angeles, California 90073, and Department of Neurology, University of California, Los Angeles, Los Angeles, California 90024 Summary. Comparison of transintegumental membrane perme- ability and partition coefficients of selected nonelectrolytes was attempted to correlate the parameters of lipid solubility, hy- drophilicity, and membrane permeation in male and female schistosomes (parasites of the portal venous tributaries of man). Surface permation (measured by the triple isotope technique) and octanol/water partition coefficients were determined for 17 com- pounds (acetamide, aminopyrine, antipyrine, benzyl alcohol, bu- tanol, caffeine, ethanol, ethylene glycol, glycerol, inosine, man- nitol, methanol, polyethylene glycol, propylene glycol, sucrose, thiourea, and urea). Linear regression analyses comparing the logarithm of the partition coefficient to transintegumental uptakes indicate a posi- tive correlation in both sexes: R=0.76 (P<0.001) for males, and R=0.77 (P<0.001) for females. Similarly, linear regression ana- lyses comparing hydrogen bond number with the logarithm of tissue uptake index demonstrate a high (negative) correlation in both males (R=-0.85, P<0.001) and females (R=-0.90, P<0.001). The male and female schistosomes showed no statisti- cally significant differences in correlation of these parameters. Surface permeation was the same in male and female schisto- somes, suggesting that male-female variations in integumental uptake rates previously observed may be restricted to metabolites which enter by way of a selective carrier system. Key words schistosomal membrane permeability - partition coefficients (cid:12)9 hydrogen bond number (cid:12)9 syncytial integument Introduction Partition coefficients, in systems such as octanol- water or olive oil-water, have been used frequently for characterization of lipophilic properties of drugs. The manner in which a solute equilibrates between aqueous and nonaqueous phases has been found useful in quantifying substituent effects for linear free energy relationships, which form the basis of quantitative structure-activity relationships em- ployed widely in drug design. Partition coefficients for a large group of nonelectrolytes have been de- rived semiempirically from chemical structures and melting points in a recent study emphasizing the relationships between changing chemical structure (biotransformation) and concomitant alterations in solubility (Yalkowsky & Valvani, 1980). A dis- cussion of the use of partition coefficients in this field may found in the recent review of Hansch and Leo (1979). In the present study, transintegumental uptakes of some seventeen different molecules have been measured in Schistosoma japonicum males and fe- males in an attempt to correlated partition coeffi- cients with penetration of the living schistosomal surface membrane. The rationale for this specific investigation may be attributed to certain unique features of the membrane system. The integumentary surfaces of the male and female schistosome are characterized as a syncytial, rather than epithelial structure, and recent studies suggest male and fe- male worms possess functionally different surfaces (Cornford & Huot, 1981). The external boundary may not be the typical bilaminate unit membrane, but a pentalaminate (Smith, Reynolds &von Lich- tenberg, 1969) or heptalaminate (Hockley & McLaren 1973a) membrane. It has also been de- scribed as a continuous border of stacked unit mem- branes (Lumsden, 1975). This surface furthermore represents the host-parasite interface, and schisto- somes escape recognition by the host immune sys- tem by maintaining host antigens on this surface by (nonspecific) passive adsorption (Smithers, Terry & Hockley, 1969), specific absorption (Kemp etal., 1977), and constitutive host-like antigens ("molec- ular mimicry"; Damian, 1979). Thus the barrier properties of the integument cannot be presumed to be identical to typical vertebrate unit membrane cell surfaces. Ultrastructural studies of the surfaces of S. ja- ponicum males and females (Sakamoto & Ishi, 1977; Voge, Price & Jansma, 1978) indicate sex-specific differences; the surface of the female is smoother and less irregular than that of the male (Sakamoto & Ishi, 1977). For this reason, male and female 0022-2631/82/0064-0217 $01.60 (cid:14)9 1982 Springer-Verlag New York Inc.
`
`Novo Nordisk Ex. 2030, P. 1
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`218 E.M. Cornford: Surface Permeation in a Schistosome uptakes have been measured independently, employ- ing a method which permits estimation of that pro- portion of test isotope which may be passively car- ried within the folds of the integumentary surface (Cornford & Oldendorf, 1979). Materials and Methods A Japanese strain of Schistosoma japonicum used for this study was obtained from the Lowell Research Foundation (Lowell, Mass.). Schistosomes were removed from the portal veins of Swiss-Webster mice (which had been exposed to about 40 cer- cariae) by perfusion of the liver with 37 ~ Hanks balanced salt solution (20 mM Hepes buffer) using the method of Duvatt and DeWitt (1967). The wet weight of male schistosomes ranged fiom 661-1394pg and 392-698gg for females. Schistosomes were washed with and maintained in RPMI :~ 1640 (Gibco, Santa Clara, Calif.), buffered with 20 mM Hepes (Sigma Chemical Company, St. Louis, Mo.), pH 7.4-7.5, at 37 ~ until they were incubated in the isotope-containing media. The wet weight of male schistosomes ranged from 661-1394 gg and 392-738 lag for females. Octanol/Saline Partition Coefficients Although olive oil/water partition coefficients (PC) correlate well with the octanol/water system (Leo, Hansch & Elkins, 1971) according to the relationship log PC (octanol) = 1.122 +0.857 log PC (olive oil), an octanol/buffered saline partition sys- tems was chosen rather than olive/oil/water because of the variabil- ity in samples of olive oil. An ether/water partition system was not selected because a saturated water solution contains 6% ether at room temperature and the ether phase contains 1% water (Wind- holtz, 1976). Partition coefficients were determined for each compound. About 1-2laCi of the test compound was added to 6ml of schistosome saline. The 6 ml of labeled saline was then added with a needle and syringe to an evacuated glass test tube with a rubber septum (Vacutainer, Becton-Dickinson, Rutherford, N.J.). An equal volume of n-octanol (octyl alcohol, J.T. Baker Chemical Company, Phillipsburg, N.J.) was added and the tube vortexed for one min at high speed and then centrifuged for 10min at approximately 2000 x g to separate the oil and saline phases. A sample of 100 gl was taken from the oil phase and the test tube gently inverted, centrifuged again, and the saline phase sampled (100 gl) and prepared for scintillation counting as described be- low. For each test compound, the first octanol/saline partition was discarded routineiy as a wash because large variations tended to occur in the initial washings. The phase containing the greatest quantity (cpm) of the compound being partitioned was then re- moved and added to a fresh aliquot of the opposite phase, vor- texed, centrifuged, and sampled as above. This procedure was performed on duplicate tubes and repeated until the partition coefficient had stabilized (typically two or three washes). The partition coefficient was then determined as follows: 1~C dpm g-1 octanol Partition coefficient- ~4 C dpm g-1 saline " It is laboratory it has been determined that, for chromatographi- cally pure radioisotopes, there is no change in the calculated partition coefficient from the 2rid, 3rd, 4th, 5th or 6th washings, but the third washing is routinely reported. Radiochemicals The tritiated water and tr aminopyrine, antipyrine, caf- feine, ethanol, mannitol, polyethylene glycol (molwt=4000) and thiourea were obtained from New England Nuclear (Boston, Mass.). The ~4C-labeled acetamide, benzyl alcohol, ethylene glycol, glyce- rol, propylene glycol, and urea were obtained from Roschem (Los Angeles, Calif.), and 14C-labeled inosine and sucrose from Amer- sham (Arlington Heights, Ill.). The specific activities of these compounds (in mCi x mmol 1) were: acetamide, 20; anainopyrine, 93.9; antipyrine, 52.25; benzyl alcohol, 10; caffeine, 48.96; ethanol, 4.4; ethylene glycol, 57; glycerol, 60; inosine, 250; mannitol, 50; methanol, 3.4; polyethylene glycol, 0.16; propylene glycol, 20; sucrose, 600; thiourea, 45; urea, 57.5; and water, 0.18. The 113mIndium generator was purchased from New England Nuclear, Radiopharmaceuticals Division (North Billerica, Mass.). Each 1.0 cc of Indium eluted from the generator was chelated by the addition of 10~tl of sterile disodium edetate (150mgxml-1) so- lution (Endrate, Abbott Laboratories, North Chicago, Ill.). This Indium-EDTA chelate solution was then titrated with 0.5N NaOH to neutrality (about 0.1 ml) and adjusted to pH=7.55 by addition of 100 mM Hepes buffer. Both In-EDTA and In-transfer- fin were used for studies involving sucrose and polyethylene glycol (two compounds which are typically excluded by mem- branes). In the latter instance, the 113mln eluate was mixed di- rectly with an equal votume of dialyzed normal serum. The In- dium then binds to (the large serum protein) transferrin, and this mixture was neutralized and used just as the In-EDTA chelate. Incubations Schistosomal permeability to the various isotopes was studied in a mansonian schistosome saline (E.E. Bueding, personal communi- cation) containing 92 mM NaC1, 4 mM KC1, 1.1 mM CaCI 2, 0.8 mM MgC12, and 20mM Hepes (Sigma Chemical Company, St. Louis, Mo.). The incubations were carried out at a pH=7.5 and a temperature of about 38 ~ The isotopically labeled incubation media typically contained about 1-2 laCi of 14C-labeled test iso- tope, a tenfold greater amount of tritiated water, and 0.2 mCi of 113mIndium-EDTA (prepared as described above) in a total vol- ume of about 1.5 ml of schistosome saline. Groups of five or six male and female schistosomes were removed from the RPMI 4F1640 maintenance medium in poly- styrene baskets (approximately 1 cm diameter) with fine (70 micron mesh) nylon screen (Tetko, Elmsford, N.Y.), bottoms of which allow free movement of medium in and out of the baskets, the worms and baskets were rinsed in unlabeled schistosome saline and incubated in the labeled media for 5 sec. At the end of the incubation, the worms (in the basket) were blotted quickly on an absorbent pad to remove excess medium and then covered with ice cold silicone oil, density =0.96 (Aldrich Chemical Company, Milwaukee, Wisc.), to minimize possible evaporation of tritiated water and/or other volatile isotopes. The oil rinse additionally removes excess isotopic medium from the worm surface by a sheeting action. The schistosomes were then rapidly transferred to scintillation vials, solubilized at room temperature with an or- ganic base (Soluene-100, Packard Instruments, Downers Grove, Ill.) and quickly prepared for scintillation counting as described below. Triplicate 20 lal aliquots of each incubation medium were taken to determine relative isotopic content. Liquid Scintillation Counting The digested schistosome, or the sample dissolved for partition coefficient determination, was mixed with 5 cc of scintillation
`
`Novo Nordisk Ex. 2030, P. 2
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`E.M. Cornford: Surface Permeation in a Schistosome fluid (Instagel, Packard Instruments, Downers Grove, Iti.), and the pH of the resulting solution was brought to near neutrality by addition of 37.5 gl of glacial acetic acid to the scintillation vial. The Indium was then counted (1-min counts) in a Packard Tri- carb 3390 scintillation counter with the windows set as described by Oldendorf and Szabo (1976). Days later, after most of the Indium (T1/2 = 100 min) had decayed, the vials were recounted for determination of 3H and 14C content. The net Indium counts were obtained by subtraction and appropriate decay correction (each successive vial is decay-corrected for the elapsed counting time; i.e., 1.0min plus the time required for the machine to change samples, 0.375min. The 14C and tritium counts per minute (cpm) were then converted to disintegrations per minute (dpm) by cubic regression analysis and correction for quench as described previously (Cornford & Oldendorf, 1979). Tissue Uptake Indices The schistosome transintegumental influx, or Tissue Uptake In- dex (TUI), of the nonelectrolytes examined was determined by the equation: TU I ( O~o)= [~_FI4C/aH mediaW~ ~ 13mIn/3H 113mln/3H mediajW~ x 100%. In this equation the minuend, or the uncorrected TUI, represents the total radioactivity present both within the worm and carried on the external integument. The subtrahend, or IUI (Indium- EDTA Uptake Index), is an estimate of the test substance which is carried on the outer surface of the integument, since Indium- EDTA is excluded by membranes. The resulting subtraction yields a TUI value which has been corrected for extra-integumental isotope (Cornford & Oldendorf, 1979). Analysis of Data All data are presented in the form of a mean (x) and standard deviation (SD) unless otherwise specified. Linear regression ana- Table 1. Octanol/saline partition coefficients 219 lyses and other calculations were performed on a Hewlitt- Packard model 9820 programmable calculator. Results Partition CoeJficients Octanol/saline partition coefficients determined for 17 test compounds are presented in Table 1. The range of these partition coefticients extends from a low of 0.00003 for polyethylene glycol (the most hydrophilic compound tested) to a high of 12.53 for benzyl alcohol (the most lipophilic of the group). These compounds additionally represent a variety of molecular weights and possess variable hydrogen bonding potential. The partition coefficients could be serially ranked in the following sequence: benzyl alcohol > butanol > aminopyrine > antipyrine > caffeine > ethanol > methanol > propylene > glycol > thiourea > acetamide > ethylene glycol > urea > glycerol > inosine > mannitol > suc- rose > polyethylene glycol. Transintegumental Uptakes The transintegumental uptakes for S. japonicum male and female schistosomes are indicated in Fig. 1. For both sexes the compound which showed the highest TUI was benzyl alcohol (86.5 +6.9 % for males, 80.5 _+12.2% for females). Butanol, aminopyrine, and caffeine were also high uptake (ca. 70 %) compounds, Compound Molecular Hydrogen Partition Logarithm of Octanol/ weight bond number coefficient partition coefficient water PC Benzyl aIcohol 108.13 2 12.53 1.096 1.i Butanol 74.12 2 8.50 0.929 0.89 Aminopyrine 231.29 1 7.34 0.865 0.8, 1.0 Antipyrine 188.22 1 2.003 0.302 0.28 Caffeine 194.19 2 1.036 0.0154 - Ethanol 48.07 2 0.664 - 0.1778 - 0.31 Methanol 32.04 2 0.300 - 0.523 - 0.77 Propylene glycol 76.09 4 0. i 19 - 0.924 - 0.92, - 1.7 Yhiourea 76.12 5 0.112 - 0.951 - 1.14 Acetamide 59.07 3 0.082 - 1.086 - 1.15, - 1.26 Ethylene glycol 62.07 4 0.046 - 1.337 - 1.93 Urea 26.64 5 0.029 - 1.538 - 2.11 Glycerol 92.09 6 0.017 - 1.770 - 1.76 Inosine 268.23 8 0.0083 -2.081 -2.08 Mannitol 182.17 12 0.0078 - 2.108 - Sucrose 342.3 16 0.00058 - 3.24 - 3.7 Polyethylene glycol 4000 (ca.) - 0.00003 -4.5 - Hydrogen bond numbers were determined according to Stein (i967). Octanol/water partition coefficients (PC) are log values from Hansch and Leo (1979) or estimates derived from other solvent systems obtained from Dr. Corwin Hansch (personal communication). These constants are in good agreement with values obtained for octanol/schistosome saline partitions in the present study.
`
`Novo Nordisk Ex. 2030, P. 3
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`220 E.M. Cornford: Surface Permeation in a Schistosome SUCROSE POLYETHYLENE GLYCOL MANNITOL 1NOSINE THIOUREA METHANOL UREA GLYCEROL ANTIPYRINE ACETAMIDE ETHANOL ETHYLENE GLYCOL PROPYLENE GLYCOL AMINOPYRINE CAFFEINE BUTANOL BENZYL ALCOHOL ~=- 1 I ' (cid:12)9 l i i ,~2%"s I (cid:12)9 ' I I i I ' ' I i ~~1 ' f - ' I I I ~,.:.:.:.:-I , ' 1 r-'-i MALE FEMALE I I I I I I ! l l 0 20 40 60 80 TISSUE UPTAKE iNDEX (%) I IOO Fig. 1. Transintegumental uptakes (TUI) for the seventeen measured compounds in Schistosoma japonicum. Bars indicate means and lines indicate standard deviations of each compound(cid:12)9 Male values are indicated with open bars, and female values are cross-hatched. The sample size ranged from 5-16 schistosomes of each sex per compound and schistosomes were studied 37-71 days post-infection Table 2. Comparison of integumental permeability in copulating and separated male and female Schistosoma japonicum Tissue uptake index (~) Compound Unpaired Paired Paired Unpaired male male female female Antipyrine 55.4_+3.8 53.7_+6.8 51.7_+4.3 52.8_+5.5 Acetamide 52.6_+3.7 55.8+4.0 52.8_+6.1 57.0+7.3 Ethylene glycol 69.1 -+7.4 67.4 _+9.2 60.1 _+6.7 62.9 _+8.3 Urea 30.1_+4.3 31.7_+5.7 33.7-+8.8 31.5+_6.6 Sample size= 3-7 for each mean _+sin and slight male-female differences were not statisti- cally significant. The two compounds showing the lowest uptakes in both sexes were polyethylene gly- col and sucrose, both compounds being well below background level (<2G) for the method. (A mean negative TUI of -3.2~o was obtained for poly- ethylene glycol, presumably because greater quan- tities of 113rain than 14C were washed off the surface with the oil rinse. A numerical zero uptake is re- ported.) When the mean of the male and female uptakes are considered. TUI's can be serially ranked: benzyl alcohol > aminopyrine > butanol > caffeine > ethylene glycol > antipyrine > pro- pylene glycol > acetamide > ethanol > glycerol > urea > methanol > thiourea > inosine > man- nitol > sucrose > polyethylene glycol. This se- quence is similar, but not identical, to the serial ranking of partition coefficients. Uptakes were compared in copulating and un- paired (apart) males and females in a small group of metabolites (Table 2). These studies confirm the sim- ilar permeabilities of the male and female integu- ment for these compounds and indicate that no changes in transintegumental permeability to these compounds can be attributed to the copulating state. In contrast, preliminary data from this laboratory suggests that glucose uptake (at a 20gM concen- tration) is lower in unpaired S.japonicum females (TUI =18.2 _+ 3.9 G) than paired females (22.2_+ 3.5}/0), and relatively higher in copulating males (TUI=26.1 _+2.8~) than in separated males (24.1 _+ 4.5; n=6 for each mean). Correlation between Partition Coefficients and Transintegumental Uptakes The semi-logarithmic correlation between partition coefficients and transintegumental uptakes in male schistosomes can be seen in Fig. 2. Linear regression analysis of these parameters indicated a high cor- relation (R=0.76) which was statistically significant (P<0.001). The same parameters were analyzed in S.japonicum females (Fig. 3). These data similarly demonstrate positive, statistically significant corre-
`
`Novo Nordisk Ex. 2030, P. 4
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`E.M. Cornford: Surface Permeation in a Schistosome 221 ~8 i i 70_ 60- (cid:12)9 50_ x 40_ _z 3O- 20- r In l-- I- oT BENZYL . ALCOHOL ETHYLENE AMINOPYRINE GLYCOL ! CAFFE,NET / ~BUTANOL ACETAMIDE r l/~ / J. GLYCEROLT it ~ ~ANTIP'YRINE PROPYLENE'~'2"" ; ETHANOL o UREA T J" J" THIOUREA ~OL,~THYLENE / _T'INO~,N~ GLYCOL / g I / , N,TOL - S~CRO~E .~.~ -4 -~ ; ', ' 1.5 LOG (PARTITION COEFFICIENT) Fig. 2. Correlation between the TUI and the logarithm of the partition coefficient for male Schistosoma japonicum. Points indicate mean values and bars indicate standard deviations of the TUI's for the compounds studied. Males displayed a high correlation between these parameters (R = 0.76), which was statistically significant (P <0.001). The slope (_+sE) of the regression line is 14.l _+ 1.3, the intercept = 57.1. A lesser correlation (R = 0.63, P < 0.01) was observed in comparing TUI vs. log (partition coefficient)/(square root of mol wt) as employed by Bissonette et al. (1979) and Levin (1980), perhaps suggesting that, for the compounds herein studied, molecular weights do not make a major contribution to schistosomal uptake =,9~ l ,.=, ,o I %1. i i'~ 50 ,E 40_ x _z 30_ 20_ '" I0_ m F- I _ I /r~l u A,CO,-,o, ETH gNET TZ T I/ I ;UTANoL / l,&7"'N' GLYCEROL! OU~REA T ......... ///~ I ..... ! l ~'ETHANOL / !INOSINE I THIOUREA P 0 L~IE~GHoYLL E N E///// MANNITOLOT[ 1 s OSE I I I I I I -4 -3 -2 -I 0 I LOG (PARTITION cOEFFICIENT) Fig. 3. Linear regression analysis of TUI vs. the logarithm of the partition coefficient for female schistosomes. As in Fig. 2, points indicate means and bars indicate standard deviations of the female transintegumental uptakes. The correlation between these parameters in Schistosoma japonicum females was also positive and quite high (R = 0.77) and showed statistical significance (P <0.001). The transintegu- mental uptakes and the slope (_+SE) of the regression line (15.8 _+ 1.4, intercept =59.4] are not significantly different from the values seen in the male worms despite rather marked sexual differences in ultrastructure. As with the males, a lesser correlation (R=0.48) was observed contrasting TUI with log (PC)/(square root mol wt), but this was not statistically significant (P<0.1). Thus for the range of molecular weights examined (30-350) this parameter does not appear to make a major contribution to membrane permeability as indicated by the TUI method lation (R=0.77; P<0.001) between log partition coefficient and transintegumental uptake. Since the uptakes of the nonelectrolytes tested were similar in male and female schistosomes, with the possible exception of glycerol (Fig. 1), it follows that the slopes of the regression lines for males (14.06_+1.28; y intercept=57.1) and females (15.81 __ 1.4; y intercept = 59.4) do not differ significantly. Data in Table 1 demonstrate that the hydrogen bond number, an estimator of hydrophilic proper-
`
`Novo Nordisk Ex. 2030, P. 5
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`222 E.M. Cornford: Surface Permeation in a Schistosome ;I _~ 2 -- 1.8 z g g if} < ?, BENZYL ALCOHOL AMINOPYRINE/'~ BUTANOL [ o~'~ CAFFEINE "~L.~///'/ o ETHYLENE GLYCOL o ~o ACETAMIDE o GLYCEROL [ o ~'~L"-~~ ~ PR 0 PY k E N E ANTIPYRINE -ETHANOL ~ GLYCOL o METHANOL THIOUREA o INOBINE MANNITOL I I I HYDROGEN BOND NUMBER Fig. 4. Correlation between the logarithm of Tissue Uptake Index and hydrogen bond number in male Schistosoma japonicum. Hydrogen bond numbers, an indicator of relative hydrophilic properties, are indicated in Table 1. The high (negative) correlation (R= -0.85, P < 0.001) indicates a greater permeability of hydrophobic molecules and is consistent with the concept illustrated in Fig. 1. The relationship (slope = -0.086, intercept= 1.95) is a higher correlation than seen in contrasting log (TUI) x (square root of tool wt) vs. hydrogen bond number (R = 0.71, P <0.01; slope = - 0.076, intercept = 2.89) c ~l.5- z g 0.5 z q BENZYL ALCOHOL BUTANOL AMINOPYRINEII "~J //CAFFEINE ETHYLENE GLYCOL oJl PROPYLENE GLYCOL O O c-~ \"'~A&TAM'DE ANTIPYRINE ETHANOL~ oGLYCEROL "--&UREA o METHANOL o~ ~ Fig. 5. Hydrogen bond number compared with the THIOURE~ ~ logarithm of Tissue Uptake Index in female Schistosoma japonicum. As demonstrated previously INOSINE ~ for males, a similarly high negative correlation (R = -0.90, P<0.001) is observed (slope= -0.11, intercept = 2.02) in females. This relationship ~ represents a higher degree of correlation than seen in comparing log (TUI) x (square root of tool wt) vs. hydrogen bond number (R = -0.80, P < 0.001 : slope = -0.097, intercept =2.95), just as has been demonstrated for males (Fig. 3) i , J ~ i ~ , i I i I i I 2 3 4 5 6 7 8 9 I0 II 12 HYDROGEN BOND NUMBER ties, varies inversely with the measured octanol/- saline partition coefficient (which increases in lip- ophilic compounds). A high (negative) correlation is apparent when hydrogen bond number is plotted as a function of the log TUI in both males (R =-0.85, P<0.001; Fig. 4) and females (R= -0.90, P<0.001; Fig. 5). This confirms the concept that lipophilic properties greatly influence physiological exchanges at the schistosome surface (Figs. 2 and 3) and in- dicates a distinct similarity between the male and female integuments with respect to penetration of nonelectrolytes. Attempts were made to correct measured TUI's for the varying size of the molecules studied. Pre- vious workers have, for example, multiplied perme- ability (Stein, 1967, p. 77) or divided the partition coefficient (Levin, 1980) by the square root of the molecular weight. In the present study a higher cor- relation was always obtained when data were not expressed as a function of molecular weight (see Figs. 2-5). This suggests that for the compounds herein studied, and the range of molecular weights (30-350) examined, penetration of the schistosome surface is not greatly influenced by molecular weight. Discussion Overton's rule, which states that in a homologous series of molecules increasing the number of CI-I 3
`
`Novo Nordisk Ex. 2030, P. 6
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`E.M. Cornford: Surface Permeation in a Schistosome 223 groups increases membrane permeability, suggested a relationship between lipophilicity and membrane permeation (see Naccache & Sha'afi, 1973). Begin- ning with the classical studies (see Collander, 1954; Stein, 1967; and the review of Diamond & Wright, 1969), several investigators have corroborated the relationship between lipophilicity and membrane permeability in various systems. Wright and Dia- mond (1969) demonstrated a correlation between partition coefficients, in both an ether/water and an olive oil/water system, and the permeabilities to sev- eral molecules of the gallbladder membrane in rab- bits. Smulders and Wright (1971) confirmed the re- lationship in rabbit gallbladder epithelium using an isobutanol/water partitioning system. Both Savitz and Solomon (1971) and Naccache and Sha'afi (1973) were able to show that partition coefficients of compounds in an ether/water system are pro- portional to their permeabilities in human red blood cell membranes. Oldendorf (1974) showed that lipid solubility is important in determining a molecule's permeability to the blood-brain barrier in vivo, and this concept was recently corroborated in a study of several brain tumor chemotherapeutic agents (Levin, 1980). Bindslev and Wright (1976) confirmed this relationship for the toad urinary bladder, and Bissonette et al. (1979) showed that the relationship holds for both guinea pig and sheep placental mem- branes. Partitioning characteristics and penetration of alcohols have also been correlated in the rabbit intestinal mucosa (Sallee & Dietschy, 1973) and rat adipocytes (Sherrill & Dietschy, 1975). Thus the cor- relation between these parameters has been estab- lished in the (above mentioned) vertebrate and plant (Collander, 1954; Stein, 1967) cell systems. The dis- tinct similarity in hydrophobic surface properties here- in demonstrated for male and female schistosomes is interesting because their integuments are so dif- ferent ultrastructurally (Sakamoto & Ishi, 1977), and the glycogen/protein ratio in males (0.5) is consider- ably larger than that of females (0.2) in this species (Cornford & Huot, 1981). These data suggest the external unit membranes of males and females possess functionally similar permeability properties, and the great differences in surface topography (vi- sualized by scanning electron microscopy) are re- lated to some surface function other than lipid me- diated uptake. The unusual accumulation of host antigens (Smithers etal., 1969; Kemp etal., 1977) and constitutive host-like antigens (Damian, 1979), which have been demonstrated in S. mansoni and are presumably also present in S. japonicum, do not mod- ify lipid mediated membrane permeability in the in vitro system utilized in this study. If these absorbed host antigens do not inhibit membrane penetration of compounds such as antischistosomal drugs, then the suggestion that antibody to specific schistosome antigens might be useful in selectively delivering a drug to the schistosomal surface, may be worthy of further study. The present studies also indicate that for anti- schistosomal drugs which gain access to the blood fluke via a lipid mediated mechanism, relatively equal amounts would be expected to be delivered to the male and female schistosome. Thus if such a lipophilic drug had a differential effect upon one sex it seems probable that some biochemical difference in the male and female schistosome would best ex- plain differential drug sensitivity. Presumably such a drug would either have a greater binding affinity for the protein or lipid constituents of one sex, or the drug would affect a biochemical pathway which operated differently in that one sex. However, identi- fication and characterization of any male or female specific proteins which possess specific affinity for antischistosomal agents has not been achieved to date. In situations where the administered drug might be quite lipophilic, such antischistosomal me- tabolites may show logarithmic reductions in lipid solubility (e.g., through systemic biotransformation) before any measurable decrease in schistosomal up- take would be observed. Certain compounds (e.g., ethylene glycol and gly- cerol, Fig. 2) have disproportionately high uptakes in relation to their measured partition coefficients. These deviations from Overton's rule have been ob- served for ethylene glycol in studies of rabbit gall- bladder (Wright & Diamond, 1969), red cells (Nac- cache & Sha'afi, 1973), dog lung (Perl, Silverman, Delea & Chinard, 1976), and the placenta (Bisson- ette et al., 1979). Selective carrier-mediated uptake of glycerol has also been described in tapeworm in- tegument (Uglem, Pappas & Read, 1974), so there are previous data suggesting that glycerol and glycol uptakes might not be easily predicted from their lipid solubilities. Traditionally, the apparent "siev- ing" of small hydrophilic solutes has been ascribed to the presence of aqueous pores, which select among the nonelectrolytes according to size, and size is a function of molecular weight. Lieb and Stein (1971) challenged the pore theory, alternatively suggesting the increased permeability of smaller hy- drophilic molecules was a natural consequence of the bilayer structure. Their suggestion that mem- brane properties should be considered more to a rubber-like polymer than an oil layer was not cor- roborated by Finkelstein (1976). It has been further indicated that recent attempts to address this con- troversy have diverted attention from the physiologi- cally relevant correlations of lipid/water partitioning
`
`Novo Nordisk Ex. 2030, P. 7
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`224 E.M. Cornford: Surface Permeation in a Schistosome and permeability, and "obscured the simplicity and usefulness of Overton's rule" (Orbach & Finkelstein, 1980). With respect to the present study, the data do not provide any insight pertinent to the theory of water movement across this membrane system, but do suggest the practical use of water as a reference molecule in vitro studies of this type, as previously demonstrated in vivo (Oldendorf, 1974; Oldendorf & Szabo, 1976). Support from NIH grant No. AI 15692 and the Veterans Admin- istration is gratefully acknowledged. Materials used in this study were provided by an NIAID supply contract (~AI 72524) and the assistance of Drs. K.O. Phifer and J.I. Bruce in making S.japonicum infections available is appreciated. I thank Dr. Jared Diamond for helpful discussions and Cyn- thia Diep, Marie Huot, Kriston Landon, and Gary Rowley for their assistance in execution of this study. References Bindslev, N., Wright, E.M. 1976. Effect of temperature on non- electrolyte permeation across the toad urinary bladder. J. Membrane Biol. 29:265-288 Bissonnette, J.M., Cronan, J.Z., Richards, L.L., Wickham, W.K. 1979. Placental transfer of water and nonelectrolytes during a single circulatory passage. Am. J. Physiol. 236:C47-C52 Collander, R. 1954. The permeability of Nitella cells to nonelec- trolytes. Physiol. Plant