Factors affecting the transport of volatile fatty acids across rumen epithelium

Transcription

1 Factors affecting the transport of volatile fatty acids across rumen epithelium C. E. STEVENS AND B. K. STETTLER Department of Physiology, New York State Veterinary College, Cornell University, Ithaca, New York STEVENS, C. E., AND B. K. STETTLER. Factors affecting the transport of volatile fatty acids across rumen epithelium. Am. J. Physiol. 2 IO(~) : I g66.-isolated, short-circuited rumen epithelium was used to study the effects of ph, concentration gradient, imidazole vs. bicarbonate-buffer systems, and fatty acid metabolism on the transport of volatile fa tty acid. Transport of acetate a nd butyrate increased with an increase in concentration gradient or a decrease in lumen bath ph, and acetate transport also was greater when a bicarbonate buffer was used in place of imidazole. The two buffer systems also differed in their effect on short-circuit current and, therefore, the active transport of ions. propionate, and butyrate were all metabolized to a conrable extent by the tissue but their metabolism was not a simple function of the amount of fatty acid absorbed. The results, including the effect of anoxia, indicated that this metabolism played a critical role in determining the rate at which these fatty acids were absorbed and transported by rumen epithelium. ruminant forestomach; fatty acid transport; fatty acid metabolism; epithelial transport; epithelial metabolism I T HAS BEEN WELL DEMONSTRATED that acetate, propionate, and butyrate are readily absorbed from the ruminant forestomach and that each is metabolized by forestomach epithelium. Much of the evidence and its historical development have been reviewed by Annison and Lewis (2). Although the rate of fatty acid absorption increases with a decrease in the ph of rumen contents, they are absorbed even from solutions of neutral ph; Dobson (4) has estimated that about 50 % of the acetate absorbed from a neutral solution, placed in the rumen of anesthetized sheep, was absorbed as anion. Therefore transport of a volatile fatty acid across the rumen epithelium could be dependent on I) the electrochemical gradient of the anion, 2) the chemical gradient of the undissociated acid, 3) the relative permeability of the epithelium to the anion and acid forms, and 4) the rate at Received for publication 28 June This investigation was supported by grants from the Agricultural Research Service, U.S. Dept. of Agriculture (Coop. Agreement (45)) and the National Institutes of Health (AM ). which these are metabolized by the epithelial tissue. The concentration gradients of the anion and acid would in turn vary with the ph of rumen contents and, in vivo, with the rate of blood flow to the organ. This is further complicated by the fact that blood flow to the rumen is stimulated to a varying degree by the absorption of acetate, propionate, or butyrate (5, 14). The possibility of active transport must also be conred in the light of Smyth and Taylor s (I 5) evidence for active transport of acetate, propionate, and butyrate across rat intestine. However, any serious conration of active transport by rumen epithelium requires a better understanding of the factors affecting their passive movement across the tissue. This study was conducted in an attempt to assess separately the contribution of some of the above factors to the transport of acetate, propionate, and butyrate across bovine rumen epithelium. METHODS Samples of rumen wall were removed from unanesthetized adult cows which had been previously fasted for I 8 hr in order to decrease the tissue content of volatile fatty acid and ketone bodies. The tissue was obtained via a permanent rumen fistula and the biopsy wound was closed with two lines of suture. The samples were taken from areas of ventral sac or caudodorsal blind sac of the rumen containing papillae approximately o cm in length. This procedure allowed the collection of about nine biopsy samples, at weekly intervals, from each animal. The general procedure used for preparing samples of rumen epithelium for the study of transport has been previously described ( I 6). The epithelium, dissected free of muscle, was divided into a pair of samples and the samples were each placed between a set of tissue chambers to allow their simultaneous study. The tissues were bathed in Krebs-Ringer solution containing 20,000 units potassium penicillin G, I mg streptomycin SO4, and I I.6 mmoles glucose/liter. This was either buffered with imidazole-hcl and gassed with 02, or bicarbonate gassed with g5 % O2 and 5 % CO2. Experiments were conducted in a room held at a temperature of 39 C. 365

2 366 C. E. STEVENS AND B. K. STETTLER Imidazole was used in the early experiments for two reasons. It provided an effective buffer at both ph 7.4 and 6.4, maintaining the bathing solutions to within ph units during the 2.5-hr experimental period. It was also used to provide a basis for the effect of adding HCO, and CO2 to the system. Both surfaces of the tissue were bathed with 20 ml of solution. The solution bathing the lumen surface of the tissue usually contained acetate, propionate, or butyrate, added to the Ringer solution as a Na salt in replacement of its equivalent in NaCl, and it was buffered to a ph of either 7.4 or 6.4. The Ringer solution added to the blood of the tissue contained none of the organic salt and was always adjusted to a ph of 7.4. All bathing solutions were adjusted to 300 milliosmols/kg water to closely approximate the mean plasma osmolality of nine of the experimental animals tested. The transepithelial electrical potential was measured at the beginning, middle, and end of the experimental period. At all other times the tissue was maintained in the short-circuited state to remove the effect of transepithelial electrical gradient on ion transport. At the end of an experimental period both bathing solutions were sampled. propionate, and butyrate concentrations were measured as titratable volatile acid, following steam distillation using a Markham still. Acetone plus acetoacetate and total ketone body determinations were made by the method of Procos (I I). The gases used to circulate and oxygenate the baths left the system through an ice-water trap to aid in the recovery of acetone. This procedure allowed a g5 % recovery of acetone solutions added to the chambers and resulted in a small increase in total ketone body recovery during the experiments. Creatinine was added to the bathing solutions 5 min prior to their collection to provide a measure of final volume. Creatinine was assayed by an adaptation of the Jaffe reaction (3). The bathing solutions contained essentially no creatinine prior to its addition at the end of the experiment, and recovery trials gave a I02 rt I. 7 % recovery propionate, and butyrate were present in the solutions in the form of both anions and undissociated fatty acids. However, for simplicity the two forms will be referred to collectively as volatile fatty acid (WA), due to the common usage of this term in transport studies. The above assays allowed calculation of the amount of VFA disappearing from the lumen bath, the amount gained on the blood of the tissue and the accumulation of ketone bodies in the two bathing solutions. In some experiments, total recovery was also measured by collecting the remainder of each bathing solution, rinsing the chambers with Ringer solution and collecting the tissue for extraction with I N H&Q. The tissue content of VFA was determined as the difference between the value for total recovery and the amount present in both bathing solutions prior to their quantitative collection. The rinsing procedure was necessary to remove VFA from the tissue surfaces and the above procedure compensated for the passage of tissue VFA into the rinsing solutions. 24 I2 I I LUMEN BATH: 30m M ACETATE (ph 6.4) 10 - LOSS LUMEN SIDE 9 I\ GAIN BLOOD SIDE MINUTES LUMEN BATH: 30mM BUTYRATE (ph 6.4) MINUTES FIG. I Absorption and transport of acetate and butyrate with time. Ketone body values represent total amount appearing in both bathing solutions. The tissues from all experiments were dried at 80 C to constant weight and the results expressed on the basis of dry tissue weight. This was done to compare the present results with other studies of ketone body production from rumen epithelium. It also allowed a comparison of results expressed per unit dry weight to those expressed per square centimeter of tissue diaphragm separating the chambers. Since the latter measurement did not take the surface area of the papillae into account it was believed that tissue weight might also be a better expression of total surface area. RESULTS The relationship of VFA transport and ketone body production to time was studied by sampling the bathing solutions at 0.5, I.5, and 2.5 hr. The lumen bath was

3 VOLATILE FATTY ACID TRANSPORT BY RUMEN EPITHELIUM t LUMEN BATH: 30 mm ACETATE (ph 74) TABLE I. Transport and metabolism of acetate and butyrate I VFA Ketone Bodies Initial Concn Aniin Lumen Bat1 mal 1 No. Loss lumen Gain blood Acetone + acetoacetate Total I I MINUTES FIG. 2. Absorption and transport of acetate with time. The acetate and ketone body content of the epithelium was also measured and the ketone body values include that of both the tissue and the bathing solutions. buffered with imidazole at ph 6.4. Figure I summarizes the results from three experiments in which a 3o-mM acetate-ringer solution was used as the lumen bath and three experiments using 30 mm butyrate-ringer. The appearance of acetate and butyrate in the solution bathing the blood of the tissue was nearly linear with time. The associated accumulation of ketone bodies in the bathing solutions was also nearly linear. However, both acetate and butyrate disappeared most rapidly from the lumen bath during the first 54 hr. This was especially evident with acetate. These experiments were repeated on paired tissue samples and both tissue and bathing solutions were analyzed for VFA and ketone bodies. One tissue sample was used to obtain results at 0.5 hr and the other for results at 2.5 hr. The same concentrations of acetate and butyrate were used in the lumen bath but this was buffered at a ph of 7.4. The results of two acetate experiments are shown in Fig. 2. The tissue content of acetate increased during the first 34 hr, then remained relatively constant. This rapid initial uptake by the tissue could partly explain the greater rate of disappearance from the lumen bath during this period. The experiments using butyrate showed that its accumulation in the tissue was similar to that of acetate. Due to the nonlinearity of VFA disappearance from the lumen bath, the results of the following experiments were expressed as values obtained over a 2.5-hr period rather than as rates per hour. Control experiments were conducted on tissues from 13 animals to determine the amounts of titratable volatile acid and ketone bodies which would accumulate in the bathing solutions in the absence of added VFA. At the end of 2.5 hr the lumen bath contained an average of 0.37 =t 0.22 pmoles volatile acid/cm2 tissue and the solution bathing the blood contained an average of 0.22 =t 0.07 pmoles/cm2. The mean value for total ketone body accumulation was o. I 2 pmoles/cm2. Acid extraction of the epithelium immediately after collection from the animal gave mean values of 0.24 pmoles volatile acid/cm2 and 0.03 pmoles ketone bodies/cm2, indicating that some I j3-o so * , *55 ~ pmoles/ioo mg I jmoles/roo mg I I.58 I I.25 I A and B refer to a pair of tissue samples collected and studied simultaneously. The identical experimental results are expressed per cm2 of tissue diaphragm in the upper one-half of the table and per 100 mg dry tissue weight in the lower one-half. Values in this and subsequent tables were obtained from 2x-hr experimental periods. volatile acid and ketone bodies were produced during the experimental incubation. Therefore the data in the following tables were corrected either by use of the mean control values given above or by a control run simultaneously on a sample of the same tissue. Transport and metabolism of acetate and butyrate at a lumen bath ph of 7.4. In the following experiments a 3o-mM acetate or butyrate-ringer solution was placed in the lumen bath. Both bathing solutions were buffered with imidazole at ph 7.4, thus maintaining the acetate and butyrate almost entirely as anions. Table I gives the results of experiments on eight pairs of tissue samples from four animals. The results in the first half of the table are expressed as micromoles per square centimeter of tissue diaphragm, and do not take into account the additional surface area of the papillae. The second half of the table gives the same results in pmoles/~ oo mg dry tissue weight. It can be seen that calculations on the basis of dry weight did not decrease the variability in results from separately collected tissues. Also the method of calcula

4 368 C. E. STEVENS AND B. K. STETTLER tion had little effect on comparisons involving the use of paired tissue samples, since these differed little in weight. The tissue diaphragm separating the chambers had an area of 7.1 cm2 and the mean dry weight of separately collected tissues was o. I 76 & g. However, the difference between the mean values of 65 tissue pairs was only g and statistically insignificant (P > 0.7). Since the use of dry weight seemed to offer no advantage and prevented extraction of the tissue, the diaphragm area was used below as a basis for comparison. The results in Table I show relatively good duplicability between pairs of tissue samples, collected and studied simultaneously. They also show that, whereas the loss of butyrate from the lumen bath was approximately four times that of acetate, only one-half as much butyrate appeared on the blood of the tissue. This could be largely accounted for by the much greater production of ketone bodies from butyrate. The experimental procedure, described above, was used as a basis for all later comparisons. The reasons for choosing imidazole as a buffer were given previously. The use of ph 7.4 for both bathing solutions simplified the system and resulted in almost complete dissociation of the fatty acids. VFA concentrations of 30 mmoles/ liter were used to provide the condition in which each fatty acid was present at about one-third the total VFA concentration of normal rumen contents. A series of I 2 acetate and I 2 butyrate experiments were carried out under these conditions but on separately collected tissues. values and their standard deviations were calculated per square centimeter of tissue. These showed that during the experimental period 5.4 & I. I pmoles acetate disappeared from the lumen bath, & 0.27 pmoles acetate were transported to the opposing bath, and 0.38 =t o. I 4 pmoles ketone bodies accumulated in the two bathing solutions. During this same period of time I 6. I & 3.9 pmoles butyrate disappeared from the lumen bath and pmoles appeared in the opposing bath, with the associated accumulation of 8.5 st I.7,umoles ketone bodies. Table 2 gives the distribution of ketone bodies between the two bathing solutions under these experimental conditions. The distribution of total ketone bodies showed more variability in the acetate experiments but no preferential accumulation with either fatty acid present. The acetoacetate-acetone fraction accounted for about 50 % of the total in the acetate experiments, and about 70 % of the total when butyrate was studied. However, there was a preferential accumulation of this fraction in the lumen bath of both experimental series. Similar experiments were repeated on tissue pairs from each of two animals to determine the effects of anoxia. The solutions bathing one tissue sample were gassed with 02, as before, while those bathing its paired sample were gassed with N2 (Table 3). Gassing with N2 TABLE 2. Ketone bodv accumulation J - - in bathing solutions Initial Concn in No. of Lumen Bath Tissues 30 mm 30 mm Percentage of Total Ketone Bodies in Bathing Solutions % Appearing blood 57*13 46&8 % Appearing as AcZAct acetone ~~~ Values are mean percentage %I SD. L~i~~n Blood Total 82&18 36&18 55&20 88&7 55&8 72%7 TABLE 3. Efect of anoxia on acetate and butyrate transport and metabolism Initial Concn in Lumen Bath 30 mm 02 (PH 7*4) (PH 30 mm 7-4-j 1 Gas N2 02 N2 - Animal 1 VFA, Loss lumen pmoles/cm2 Gai; jjood rota1 Ketone Bodies, j.4moles/cm II.4 I.13 7 * Paired tissue samples from each animal were used to compare effect of gassing bathing solutions with 02 vs. N2. Bathing solutions were buffered with imidazole. VFA which appeared on the blood was increased. The most marked effects were seen with butyrate. Effect of PI-f on transport and ketone body production. The same general procedure was used to study the effect of ph. The lumen bath, containing a 3o-mM acetate, propionate, or butyrate solution, was buffered at ph 7.4 for one tissue of each pair, and at ph 6.4 for the other. Table 4 gives the results of these experiments and it can be seen that at the lower ph each VFA showed an increase in both disappearance from lumen bath and transport to the blood. However, VFA absorption from the lumen bath was increased to a greater extent than transport, especially in the case of acetate. This suggested that there was an increase in either the tissue storage or metabolism of acetate. Extrapolation of the acetate absorption curves (Figs. I and 2) back to zero shows that an additional 3,umoles acetate may have accumulated in the tissue at ph 6.4. This would account for less than one-half of the additional unaccountable acetate in two of the three experiments, indicating that acetate metabolism had also increased. The production of ketone bodies from acetate and butyrate appeared to decrease with the lowering of lumen bath ph while insignificant amounts of ketone bodies were measured at either ph in the propionate experiments. Therefore, an increase in acetate metabolism could not be explained on the basis of ketone body production.

5 VOLATILE FATTY ACID TRANSPORT BY RUMEN EPITHELIUM 369 TABLE 4. Effect of ph on acetate and butyrate transport and metabolism Initial Concn in Lumen Bath Propionate, Animal No. VFA, pmoles/cm2 Ketone Bodies, ymoles/cm2 Loss lumen PH 7.4 PH II.o 12.8 >H 7.4 ph *3* l Acetone + acetoacetate PH ?H 6.4 PH?H Total o-53 o *94 T ph values refer to that of lumen bath and comparisons were made using paired tissue samples. Bathing solutions were buffered with imidazole. E$ect of using a bicarbonate b@er. Table 5 compares the results of a bicarbonate-co2 buffered system with results using imidazole-hcl as a buffer. Measurements of bathing solution ph showed that both buffers provided an essentially constant and equivalent ph throughout the experimental period. Total recovery of VFA and ketone bodies was attempted by assaying the tissue as well as the bathing solutions, and the values for ketone bodies in Table 5 include the tissue content. The absorption and transport of acetate was conrably greater in the bicarbonate system, although this was accompanied by a decrease in ketone body production. The two buffer systems showed no consistent difference in their effect on butyrate transport or metabolism. The percentage of absorbed VFA recovered, as VFA in the tissue and opposing bath and as ketone bodies, was determined. The conversion to ketone bodies assumed the utilization of 2 moles acetate and I mole butyrate/ mole ketone body. About 85 % of the absorbed acetate was recovered as VFA and ketone bodies in the system buffered with imidazole compared to about a 75 % recovery using the bicarbonate buffer. Approximately go % of the butyrate absorbed from the lumen bath was recovered as VFA plus ketone bodies using either buffer system. Imidazole also had a characteristic effect on the shortcircuit current. Earlier studies (16) showed that when rumen epithelium was bathed on both s with identical solutions of bicarbonate-ringer solution, the current required to maintain a short-circuited condition decreased almost linearly with time, at a rate roughly parallel to that of active Na transport. Figure 3 shows the short-circuit current recorded during two series of experiments. One series compares the two buffer systems with no VFA in the lumen bath and, therefore, identical solutions bathing both tissue surfaces. The other compares the two buffer systems under the conditions described in Table 5. It can be seen that the bicarbonate system showed a relatively linear decrease in shortcircuit current either in the presence or absence of acetate. However, the tissues buffered with imidazole consistently recorded a marked initial drop in shortcircuit current, usually followed by a rise after about I hr. When the two buffer systems were tested on paired tissue samples the respective short-circuit currents were approximately equal after 2 hr. These results indicated that active ion transport was also affected by the change in buffer systems. This was not due to differences in the Na or Cl concentrations of the two buffer systems since additional experiments correcting these differences with MgC12 did not change the results. E$ect of increased concentration gradient. The results of increasing the lumen bath concentration of acetate, propionate, or butyrate to go mmoles/liter are given in Table 6. Bicarbonate-Ringer solution was used and each experiment was duplicated on paired tissue samples to help compensate for the increased error in the measurement of absorption from the more highly concentrated lumen bath. If the results of the acetate and butyrate experiments in this table are compared with the results in the bicarbonate columns of Table 5 it can be seen that a threefold increase in the concentration of VFA in the lumen bath was associated with a marked increase in absorption and transport. The ketone body values in Table 5 include the amount found in the tissue, but if this is subtracted the amounts in the bathing solutions averaged 0.37,umoles/cm2 for acetate and g. I 7 pmoles/ cm2 for butyrate. This indicated a greater ketone body production with increased absorption of acetate, but not with the increased butyrate absorption. DISCUSSION The transport of volatile fatty acids across the rumen epithelium was studied in the absence of a transepithelial electrical potential gradient. Under this condition passive transport of both the undissociated and the dissociated forms of a fatty acid would be a direct function of their respective concentration gradients across the tissue. The relative rate at which each would be transported down a given concentration gradient would then depend on the differences in tissue permeability to the two forms of acid. However, due to tissue metabolism of fatty acid, the critical concentration gradients would be those between the site of metabolism and the solutions bathing the two surfaces of the epithelium. Since the rate of metabolism would affect the intracellular concentration of VFA, an increased rate of fatty acid metabolism would increase its rate of absorption by the tissue while decreasing its rate of transport across the tissue. An inhibition or saturation of the enzyme systems responsible for this metabolism would have the opposite effect.

6 370 C. E. STEVENS AND B. K. STETTLER TABLE 5. Comparison of imidasole and bicarbonate bu$er systems VFA, pmoles/cm2!ni- Initial Concn in ma1 Lumen Bath Loss lumen Sain blood 30 mm 30 mm (PH 7-4) - NO. IO II I4 Ketone Bodies, pmoles/cm2 Imid. Bicarb. Imid. Bicarb. Imid. 1 Bicarb I I.02 O o * I *57 O I B-73 I I I II.g Paired tissue samples from each animal were used to compare effect of imidazole vs. bicarbonate. Ketone body values include the amounts found in both the bathing solutions and tissue. f 60- ;; i I Lumen Bath : 30m M Acetate eluded that these results could be explained by inhibition of an active transport of butyrate into the tissue and by the high intracellular concentration of butyrate which would result from inhibiting its metabolism. However, the results could be explained by the inhibition of butyrate metabolism alone, without postulating a system for active transport. In the present experiments anoxia had a much smaller effect on the absorption and transport of acetate. Since acetate was also metabolized less than butyrate in the presence of 02 the effects of anoxia on both fatty acids may have been simply due to an inhibition of their metabolism. Anoxia may have also affected permeability but it is at least interesting to note the similarity in the permeability of anoxic, nonmetabolizing rumen epithelium to the two fatty acids when these were largely in their dissociated form at ph 7.4. As noted in a preliminary study (I 7), the absorption and transport of VFA increased with either an increase in the total fatty acid concentration of the lumen bath or a decrease in lumen bath ph. The increase in lumen bath VFA concentration, from 30 to go mmoles/liter, resulted in a threefold increase in both the dissociated and undissociated acid. Lowering the ph of the lumen bath to 6.4, with no change in total VFA concentration, resulted in a IO-fold increase in the concentration of undissociated fatty acid with only about a 2 % decrease in the concentration of the dissociated form. Therefore, increased absorption and transport at the lower ph could result from both the increased transepithelial concentration gradient of the undissociated acid and a greater tissue permeability to this more lipid-soluble form Although the two procedures for changing the con- Minutes Minutes FIG. 3. Relationship of short-circuit current to time using imidazole vs. bicarbonate to buffer bathing solutions to ph 7.4. Graph on left gives results from control studies on 4 tissues collected on separate days. Graph on right shows the results, with the addition of acetate to the lumen bath, from 3 sets of paired tissue samples. All three of the fatty acids studied were metabolized to an important extent following their absorption by the tissue. When a 3o-mM solution of VFA, buffered with imidazole at ph 7.4, was placed in the lumen bath, an average of about 75 % of the butyrate and 30% of the acetate absorbed by the tissue was converted to ketone bodies. No ketone body production was measurable when propionate was used. Conversely, the amounts of VFA apparently metabolized to products other than ketone bodies were as follows: butyrate I o %, acetate I 5 %, and propionate %. Butyrate was absorbed much more rapidly than acetate from lumen bath solutions buffered at a ph of 7.4, but it was also metabolized more rapidly so that less was transported across the tissue. When metabolism was prevented by anoxia the absorption of both butyrate and acetate decreased, whereas the transport of both increased. This effect of anoxia on butyrate transfer was previously noted by Hird and Weidemann (9). They con- 90 centration gradient were conducted with the use of different buffer systems they appeared to produce a number of similar effects. The increase in lumen bath concentration of either total or undissociated butyric acid increased butyrate absorption about I.5 times but resulted in a three- to fourfold increase in butyrate transport. Therefore, it appeared that the enzyme systems responsible for butyrate metabolism were already saturated at the lower rates of butyrate absorption. This was also indicated by the observation that increased butyrate absorption did not result in a greater production of ketone bodies. A similar increase in lumen bath concentrations of total or undissociated acetic acid usually resulted in about a twofold increase in acetate absorption. Acetate transport increased to about the same extent with a rise in total fatty acid concentration, but showed less of an increase when the ph of the lumen bath was lowered. Ketone body production did increase with an increase in the total acetate concentration of the lumen bath. Although there was no apparent increase in ketone body production at the lower ph, the marked discrepancy noted between the amounts of acetate absorbed and transported suggested an increase in the rate of metabolism by other pathways. The substitution of bicarbonate buffer for imidazole resulted in a marked increase in acetate absorption and transport. Ash and Dobson (3) described a definite rela-

7 VQLATILE FATTY ACID TRANSPORT BY RUMEN EPITHELIUM 37= TABLE 6. Transport and metabolism with go m&t concentrations of VFA in the lumen bath Initial Concn in Lumen Bath (PH Propionate, go rnm T-4) (PH 7*4) go mm go mm Animal No. VFA, pmoles/cm2 Loss lumen Gain blood 20.1 II IO ~ B A B --- I7*39*739*3I mg A l-* j I c-54 I 7, Ketone Bodies, pmoles/cmz 1 ketone + acetoacetate B 3.43 Total I.I8 I.05 I 055 I r A and B refer to paired samples of tissue. Bathing solutions were buffered with bicarbonate B I.13 I.23 I.21 I.1g tionship between the absorption of acetate from the sheep rumen and the appearance of bicarbonate in the rumen contents. They concluded that this was due to the conversion of acetate to the more readily absorbed acetic acid, with carbonic acid serving as the H donor. The present study offers no direct evidence on this hypothesis. Just prior to this writing Alonso, Rynes, and Harris (I) presented evidence which demonstrated that imidazole inhibits H secretion by the gastric mucosa of the frog without inhibiting the transport of Cl. They also found that imidazole inhibited the active transport of Na across the toad urinary bladder. It was concluded that the above effects were not related to buffering capacity and may have been due to inactivation of cyclic AMP or interference with oxidative phosphorylation. The marked drop in short-circuit current seen when imidazole was used in the present study could have resulted from inhibition of the active transport of Na. The lower rate of acetate absorption and transport in the presence of imidazole may also have been due to some form of imidazole inhibition. However, imidazole did not inhibit ketone body production or produce a permanent inhibition of short-circuit current. The short-circuit cur rent returned to a level similar to that recorded in a paired tissue buffered with bicarbonate. This could be attributed to a tissue production of CO2 and would argue that the difference in current seen earlier in the experiment may have been related to a low HCOs concentration or pco2 rather than the presence of imidazole. The high percentage of acetoacetate produced during butyrate metabolism agrees well with the results Sutton and co-workers (18) have reported using calf rumen epithelium, and those of Hird and Weidemann (9) using the rumen epithelium of sheep. Measurements of total ketone bodies indicated that these were distributed about equally between the solutions bathing the two surfaces of the tissue. This also agrees with the results Hird and Weidemann obtained when they used epithelial samples with long papillae. However, the acetoacetate-acetone fraction was distributed predominately to the lumen of the tissue in the present study. Pennington (IO) found ketone body concentrations of 70 and 140 pmoles, respectively, in two solutions of butyrate 30 min after they were placed in the rumen of a sheep. This would indicate that an important percentage of the ketone bodies produced in rumen epithelium may normally pass into the rumen contents. Since there appears to be a similar production of ketone bodies by rumen and omasal epithelium (8, IO), it is also interesting to note the results of Roe, Bergman, and Kon (I 3), showing that acetoacetate accounted for about one-half the total ketone bodies entering the portal vein of normally fed sheep. This percentage, which is much lower than that calculated in most in vitro studies, agrees with the percentage found on the blood in the present study. This may indicate that a preferential loss of acetoacetate into the lumen also occurs in vivo. The production of small amounts of VFA by the tissue or rumen bacteria required a correction of measurements. Interconversion of fatty acids is also a possible source of error in the interpretation of results. Interconversion has been demonstrated in the normal rumen contents of sheep (6, 7), presumably a result of bacterial metabolism, but Ramsey and co-workers (I 2, I g) have also presented evidence that rumen epithelium can convert butyrate to acetate. In the present study it was assumed that the procedures used to rinse the tissues of bacteria and inhibit their growth prevented any significant degree of interconversion, although the method used for VFA measurement did not allow a test of this assumption. The authors gratefully acknowledge the technical assistance of D. B. Harrington and T. R. Pescod. REFERENCES I. ALONSO, D., R. RYNES, AND J. B. HARRIS. Effect of imidazoles on active transport by gastric mucosa and urinary bladder. Am. J. Physiol. 208: I Igo, ANNISON, E. F., AND D. LEWIS. Metabolism in the Rumen. London: Wiley, ASH, R. W., AND A. DOBSON. The effect of absorption on the acidity of rumen contents. J. Physiol., London 169 : 39-6 I, I DOBSON, A. Active transport through the epithelium of the reticula-rumen sac. J. Physiol., London 146 : I, DOBSON, A., AND A. T. PHILLIPSON. The influence of the contents of the rumen and of adrenalin upon its blood supply. J. Physiol., London 133 : P, GRAY, F. V., G. B. JONES, AND A. F. PILGRIM. The rates of production of volatile fatty acid in the rumen. Australian J. Agri. Res. I I : , GRAY, F. V., A. F. PILGRIM, H. J. RODDA, AND R. A. WELLER. Fermentation in the rumen of the sheep. IV. The nature and

8 372 origin of the volatile fatty acids in the rumen of the sheep. J. Exptl. BioZ. 2g : 57-65, HIRD, F. J., AND R. J. SYMONS. The mode of formation of ketone bodies from butyrate by tissues from the rumen and omasum of sheep. Biochim. Biophys. Acta 46 : , I 961. g. HIRD, F. J., AND M. J. WEIDEMANN. Transport and metabolism of butyrate by isolated rumen epithelium. Biochem. J. 92: I964. IO. PENNINGTON, R. J. The metabolism of short-chain fatty acids in the sheep. Biochem. J. 51: , I I. PROCOS, J. Modifications of the spectrophotometric determination of ketone bodies in blood enabling the total recovery of betahydroxybutyric acid. Clin. Chem. 7 : , RAMSEY, H. A., AND C. L. DAVIS. Metabolism of n-butyrate by the adult goat. J. Dairy Sci. 48 : , ROE, W. E., E. N. BERGMAN, AND K. KON. Absorption of ke- I4* r5* 16. 7* 18. =9* C. E. STEVENS AND B. K. STETTLER tone bodies and other metabolites via the portal blood of sheep. Am. J. Vet. Res. In press. SELLERS, A. F., C. E. STEVENS, A. DOBSON, AND F. D. MCLEOD. Arterial blood flow to the ruminant stomach. Am. J. Physiol. 207: , =964* SMYTH, D. H., AND C. B. TAYLOR. Intestinal transfer of shortchain fatty acids in vitro. J. Physiol., London 141 : 73-80, I 958. STEVENS, C. E. Transport of sodium and chloride by the isolated rumen epithelium. Am. J. Physiol. 206 : I 099-I I 05, STEVENS, C. E., AND B. K. STETTLER. Transport and metabolism of acetate, propionate and butyrate by bovine rumen epithelium. Federation Proc. 24: 276, SUTTON, J. D., A. D. MCGILLIARD, M. RICHARD, AND N. L. JACKSON. Functional development of rumen mucosa. II. Metabolic activity. J. Dairy Sci. 46 : , TAYLOR, T. A., AND H. A. RAMSEY. Metabolism of n-butyrate bv the rumen ebithelium. J. Dairv Sci. ~8: 5ov ~6~.