INFECTION AND IMMUNITY, Dec. 1975, p. 1450-1456 Copyright (© 1975 American Society for Microbiology
Vol. 12, No. 6 Printed in U.S.A.
Comparison of the Rate of Absorption and Proteolysis of [14C]Choleragen and ["4C]Bovine Serum Albumin in the Rat Jejunum DONALD R. STROMBECK* AND DORIS HARROLD University of California, School of Veterinary Medicine, Department of Medicine, Davis, California 95616 Received for publication 6 May 1975
['4C]choleragen was used to study the rate of disappearance of choleragen enterotoxin from the jejunum of rats. [14C]bovine serum albumin (BSA) was studied in a similar manner. Almost one-third of the labeled toxin had disappeared from the intestine after 6 h. Its rate of disappearance was the same in germfree rats as in conventional rats. The rate of proteolysis of ['4C]choleragen and ['4C]BSA by intestinal mucosal lysosomal enzymes was also studied. Neither was significantly degraded by neutral proteases; however, heat-inactivated toxin was. They were all degraded by acid proteases; however, the rate of BSA proteolysis was only one-third of that of toxin. Soybean trypsin inhibitor had no effect on the in vivo disappearance of toxin nor on the acid proteases. It did inhibit the neutral protease digestion of heat-treated toxin. Apirotinin and protamine inhibited disappearance in loops of gut but had no effect to inhibit degradation rates. Gangliosides inhibited both rates of disappearance and proteolysis of toxin. These agents had some different effects on disappearance rates and proteolysis of BSA. The data indicate that cholera enterotoxin is absorbed by intestinal mucosal cells and is degraded by acid proteases in the cells.
Vibrio cholerae produces an enterotoxin which, when released by the organism, binds rapidly to the intestinal mucosal surface (19). The effect of the toxin is to stimulate the secretion of fluid and the electrolytes. Five minutes after binding of the toxin to the mucosal surface, this secretory response cannot be prevented by washing the surface in an attempt to remove the bound toxin (11). The toxin's effect has been shown to be mediated by stimulation of the enzyme adenyl cyclase to increase intracellular concentrations of adenosine 3',5'-cyclic monophosphate (6, 13, 15). Adenyl cyclase in the intestinal mucosal epithelial cell is located in the plasma membranes of the basal and lateral margins of the cells, and it does not appear to be a normal constituent of the brush border (4, 9). Thus, cholera toxin exerts its effect by either (i) causing an event to occur in the brush border, which leads to increased activity of adenyl cyclase in the basal and lateral cell membranes, or (ii) entering the cell and stimulating adenyl cyclase directly after intracellular transport of the toxin to the basal and lateral membranes. At the present time, there is some question of whether cholera toxin remains entirely on the surface while exerting its effect, or whether it enters the epithelial cells after binding to the surface. An autoradiographic study demonstrated that cholera toxin labeled with
tritium does not enter the mucosal epithelial cell (10), whereas an immunofluorescent study suggested that cholera toxin does enter these cells (5). The purpose of the present investigation was to study the fate of cholera toxin in the intestine, with a more specific goal being to determine whether the toxin is absorbed by the intestinal mucosal cells. It would seem logical to assume that antigenically competent cholera toxin must be absorbed for the production of antibodies that confer immunity to protect against the toxin after a bout of the disease. In the intestine, these specific immunoglobulin A antibodies against cholera toxin are produced by plasma cells in the lamina propria and then are secreted into the lumen of the gut (12).
MATERIALS AND METHODS Animals. The stimulation of secretion of intestinal fluid by cholera toxin was studied in the rat by using a method described previously (16). Male Fisher rats, 200 to 300 g, were prepared for experiments by withholding food for 24 h but allowing them to drink a 5% glucose solution ad libitum. Under ether anesthesia, the small intestine was ligated at a point just distal to the ligament of Treitz, and its contents were flushed into the cecum with 10 ml of 0.9% saline, which was introduced with a syringe and 27-gauge needle. After the intestine was emptied, the distal end was ligated. Three 12-cm loops of jejunum were measured and ligated at 1450
VOL. 12, 1975
['4C]CHOLERAGEN AND ['4C]BSA ABSORPTION
1451
each end to form closed loops, which were then mg to 40 Atg of toxin before loop inoculation. This injected with either 40 ,ug of ['4C]choleragen or 40 jig was done to see if the effect of aprotinin to inhibit of ['4C]bovine serum albumin (BSA) in a volume of the disappearance rate of toxin might be a property 0.5 ml. The loops were removed 6 h later and of a basic peptide that is not a protease inhibitor. Gangliosides bind cholera toxin and prevent attachweighed with and without their fluid contents. The rate of disappearance of ['4C]choleragen from ment to the mucosal receptor sites. To study these closed loops of intestine in germfree rats was deter- glycolipids' effects on the disappearance rate of mined to ascertain if the intestinal microflora in ['4C]choleragen, 1 mg of ganglioside (type II, bovine conventional rats can alter choleragen and hence brain, Sigma Chemical Co.) was premixed with 40 change its rate of disappearance. Fisher germfree ,ug of toxin before injection into loops. In another rats were bred and raised using germfree equipment group ganglioside was combined with cerebrosides and germfree methods described previously (17). (18) and added to toxin so that the equivalent of 1 mg The protocol used in the germfree rats was the same of ganglioside was used with 40 ,ug of ['4C]choleragen before injection into loops. as that used for conventional rats. Calculation of disappearance of ['4C]cholerCholera toxin. A highly purified cholera toxin, choleragen lot 1071, was used which was prepared agen and ['4C]BSA. Loop fluid in scintillation vials under contract for the National Institute of Allergy was desiccated in vacuo with P205. Protosol (New and Infectious Diseases by R. A. Finkelstein, Uni- England Nuclear) was used to solubilize loop tissue versity of Texas Southwestern Medical School, Dal- and also the residue remaining after evaporation of las, Tex., according to a procedure described previ- loop fluid. Aquasol (New England Nuclear) was ously (3). It was reconstituted in 0.01 M phosphate- added to the solubilized specimens, and radioactivbuffered saline (PBS) (pH 7.4) to achieve a concentra- ity was determined in a scintillation counter. The tion of 200 ,ug of choleragen/ml. Choleragen was sum of the total radioactivity in the loop fluid and in kindly provided by Carl E. Miller, the Cholera Pro- the tissue at the end of 6 h was used to calculate the gram Officer of the National Institute of Allergy and percentage of injected counts that was left. The counts remaining at 6 h could have represented Infectious Diseases. '4C labeling of choleragen and BSA. Five milli- label present in both intact choleragen and its degragrams of choleragen in 25 ml of PBS was labeled dation products which would result from digestion with 14C after a procedure described previously (16). either in the lumen of the intestine or within the Unlabeled choleragen was first dialyzed against 0.1 cell. Moreover, the radioactivity that had disapM sodium borate, pH 9. It was then subjected to peared from the intestinal loops is also assumed to reductive alkylation by the addition of 20 ,uCi of represent degraded ['4Clcholeragen and possilby in[14C]formaldehyde (specific activity, 59.2 mCi/mmol; tact toxin. Thus, what is being measured in deterNew England Nuclear Corp.), which was followed mining the disappearance rate is the result of both by the addition of 4.0 mg of sodium borohydride. The digestive and absorptive processes in the intestine. reaction was carried out for 30 min at 38 C. The Parts of the study were designed to ascertain the labeled choleragen was dialyzed against PBS (pH relative importance of digestion versus absorption 7.4) until the level of radioactivity in the dialysant in the disappearance of labeled choleragen from the was no greater than the background activity. The loops. Germfree rats and proteinase inhibitors were specific activity of the labeled toxin was 7.2 used in some groups to estimate the contribution of mCi/mmol. By inoculation into rat intestinal loops intraluminal digestion of ['4C]choleragen to its rate and into rabbits intradermally, it was verified that of disappearance. Incubation of ['4C]choleragen and [14C]BSA choleragen had not lost any measurable amount of its toxicity in the labeling process. Forty micro- with mucosal cell homogenates. The mucosa was grams of labeled choleragen produced a maximum scraped from the jejunum of fasted rats and homogesecretory response in loops of jejunum, and [14C]- nized in 10 volumes of 0.15 M sucrose and 5 mM choleragen had the same number of bluing doses per ethylenediaminetetraacetate, pH 7.0, with a motormicrogram as unlabeled choleragen. No proteins driven Teflon pestle in a Potter-Elvehjem homogewere added to stabilize the labeled choleragen, nizer. The nuclear and cell debris and heavy mitowhich was stored at 4 C until used. Twenty-five chondria were removed by centrifugation at 4,000 x milligrams of BSA was dissolved in 25 ml of PBS and g for 10 min. The supernatant was centrifuged at labeled in exactly the same manner as choleragen. 55,000 x g for 10 min to yield a pellet containing It was stored frozen until used. The ['4C]albumin light mitochondria and lysosomes. Both the pellet (after resuspension in 10 volumes of buffer) and the had a specific activity of 10.2 mCi/mmol. Modification of [14C]choleragen and ['4C]BSA supernatant were used to study degradation of choldisappearance rate. Heat inactivation of ['4C]cho- eragen and BSA and were assayed for cathepsin (1) leragen was accomplished by heating to 85 C for 10 and N-acetyl glucosaminidase (using p-nitrophenylmin with agitation. It was injected into loops and N-acetyl-,8-D-glucosaminide as substrate in citrate added to mucosal lysosomes to study the effect of buffer, 0.05 M, pH 4.2). Assays for degradation of the denaturation of toxin on absorption and digestion. proteins consisted of 20 Ag of either ['4C]choleragen The protease inhibitors aprotinin (Trasylol) and or ['4C]BSA in either a PBS buffer (pH 7.0) or an soybean trypsin inhibitor were added at levels of acetate buffer in which the pH of the entire incubate 3,000 KIU and 10 mg, respectively, to 40 ,ug of toxin was 4.5. At the onset of incubation 100 p.l of either before injection into loops. In one study, protamine the 55,000 x g pellet suspension or the supernatant sulfate, which is a basic peptide similar in size and from this centrifugation was added, and for the conpK properties to aprotinin, was added at a level of 3 trols, they were added at the end of the incubation.
1452
STROMBECK AND HARROLD
INFECT. IMMUN.
Aprotinin (1,000 KIU), soybean trypsin inhibitor (1 mg), or cerebroganglioside (containing 0.33 mg of ganglioside) was added to the assays at the onset and to the controls at the end of the incubation. They were incubated for 4 h and stopped by addition of ice cold trichloroacetic acid to give a final concentration of 5%. After centrifugation, the supernatant was added to Aquasol and counted in a scintillation counter. A time study of the release of trichloroacetic acid-soluble label over a period of 6 h was done with both choleragen and BSA. Electrophoresis of heat-treated and unaltered [(4C]choleragen. The effects of heat treatment on ['4C]choleragen were determined by comparing the electrophoretic patterns of heated toxin with untreated toxin following methods previously described (16). They were both incubated for 30 min at 37 C in 0.01 M sodium phosphate buffer (pH 7.1) containing 0.5% sodium dodecyl sulfate and 0.5% 2mercaptoethanol, after which electrophoresis was performed in 10% polyacrylamide gels. In another group, the two toxins were prepared in 0.1 M borate buffer (pH 9), and electrophoresis was done on 10% polyacrylamide gels. Protein was detected by staining 2 to 3 h with Coomassie brilliant blue (0.25% in methanol)-water-acetic acid (5:5:1, vol/vol). They were destained in water- acetic acid- methanol (35:3:2, vol/vol) (20). The relative amount of protei'n in each band was estimated by scanning with a Gilford gel densitometer. The gels were sliced and each part was dissolved in 30% H202, after which Aquasol was added, and radioactivity was determined in a scintillation counter.
RESULTS
Fluid production and disappearance of ['4C]choleragen from loops of jejunum after 6 h. A substantial amount of fluid was produced when ['4C]choleragen was injected into loops of
conventional rats (Table 1). Greater than 30% of the radioactivity placed in the loops was not present 6 h later. The results of injection of labeled toxin into loops of germfree rats produced similar results. The addition of aprotinin to choleragen had no effect on fluid production but reduced the rate of disappearance of choleragen to 40% of that in loops with choleragen alone. Similar results were seen with protamine. Soybean trypsin inhibitor had no effect on fluid production or the disappearance rate of choleragen. Ganglioside and cerebroganglioside inhibited fluid production completely and also caused the rate of choleragen disappearance to fall more than 50%. A very small amount of fluid was produced by heat-inactivated choleragen. This treatment of the toxin caused it to disappear at twice the rate of choleragen. Treatment of heat-inactivated toxin with aprotinin or ganglioside produced similar effects as these compounds had on choleragen. The calculation of disappearance rates in loops with fluid was based on counts in loop fluid which represented 92 to 93% of the total counts at 6 h and on those in loop tissue which amounted to 7 to 8% of the total. These percentages were not significantly different between any of the groups of loops that produced fluid. In loops with no fluid, no attempt was made to flush the mucosal surface and to determine how much radioactivity was bound to or within the tissue and how much was free on the surface. There was no difference in the disappearance rate of ['4C]choleragen between the three loops in each rat. Therefore, the data were pooled for
TABLE 1. Fluid production and disappearance of ["4C]choleragen from loops
Group'
ofijejunum after 6 h
Fluid production (g of fluid secreted/g of inNo. of loops (3/rat) testine [wet wt] per 6 ha)
Disappearance of'
['4C]choleragen maining in 6 (%ha) re-
Choleragen in: Conventional rats Germfree rats
84 24
4.87 ± 0.26 4.96 ± 0.39
69.85 ± 1.10 67.76 ± 2.94
Conventional rats: Choleragen Choleragen + aprotinin Choleragen + soybean trysin inhibitor Choleragen + ganglioside
84 35 29 57
4.87 ± 0.26 5.16 ± 0.26 4.00 ± 0.46 0 0 4.93 ± 0.30 0.61 ± 0.21 0.37 ± 0.11 0
69.85 88.40 71.76 85.05 89.31 83.96 39.19 62.01
Choleragen
+
cerebroganglioside
Choleragen + protamine
Heated choleragen Heated choleragen + aprotinin Heated choleragen + ganglioside a
b
Mean ± standard error of the mean. P < 0.001 compared to choleragen. P < 0.001 compared to heated choleragen.
25
23 24 21 21
± 1.10 ± 1.14 ± 2.24 ± 1.07b ± 1.68 ± 1.93
±1.65
+ 1.06" 52.84 ± 1.16"
VOL. 12, 1975
[14C]CHOLERAGEN AND ['4C]BSA ABSORPTION
each group with no breakdown as to the region of the jejunum from which the loop was taken. Disappearance of ['4CIBSA from loops of jejunum after 6 h. The rate of disappearance of labeled albumin was the greatest in the first part of the jejunum (Table 2). Aprotinin did not change the rate of disappearance, but ganglioside increased it significantly. Release of trichloroacetic acid-soluble '4C from ['4C]choleragen and ["CIBSA after incubation with intestinal mucosal proteases. Proteolytic activity is suggested to have been present in excess in all the lysosomal and 55,000 x g supernatant
1453
assays did not change the rate of appearance of trichloroacetic acid-soluble '"C. The rate of appearance of trichloroacetic acid-soluble '"C label from [''Cicholeragen and ['"C]BSA is depicted in Fig. 1. They are linear for both; however, the rate of label release from BSA was one-third of that from choleragen. The action of the lysosomal fraction of the cells to release '4C from labeled choleragen and BSA is shown in Table 3. A very small amount of degradation occurred at pH 7.0. This activity increased 10-fold at pH 4.5. Neither aprotinin or soybean trypsin inhibitor had any effect on the rate of release of '4C. Cerebroganglioside inhibited the rate of choleragen degradation 40
preparations since doubling enzymes added to the
the amount of lysosomal
30
20 0
10
~~~~~~~~~~~BSA
0
2
4
3
5
6
TIME (HOURS)
FIG. 1. Rate of appearance of trichloroacetic acid (TCA)-soluble '4C from ['C]choleragen and [1 C]BSA during incubation with intestinal mucosal acid proteases. [''Clcholeragen: y 6.568 (x) 0.303, r 0.998, standard error = 0.753. ['4C]BSA: y = 2.108 (X) - 0.898, r = 0.991, standard error = 0.584. =
-
=
TABLE 2. Disappearance of 1'4C BSA from loops ofjejunum after 6 h Groupa Determinants
No. of animals
Disappearance of [4C]BSAb Proximal' Middle Distal
Probability Proximal versus middle Proximal versus distal Proximal versus proximal Middle versus middle Distal versus distal a
I
II
III
11
9
10
41.2 + 6.6 65.3 ± 6.97 71.2 ± 5.81
45.9 + 4.34 68.8 + 3.26 76.6 + 3.97
14.9 + 2.09 37.4 + 6.93 54.3 + 8.83
Vol. 12, No. 6 Printed in U.S.A.
Comparison of the Rate of Absorption and Proteolysis of [14C]Choleragen and ["4C]Bovine Serum Albumin in the Rat Jejunum DONALD R. STROMBECK* AND DORIS HARROLD University of California, School of Veterinary Medicine, Department of Medicine, Davis, California 95616 Received for publication 6 May 1975
['4C]choleragen was used to study the rate of disappearance of choleragen enterotoxin from the jejunum of rats. [14C]bovine serum albumin (BSA) was studied in a similar manner. Almost one-third of the labeled toxin had disappeared from the intestine after 6 h. Its rate of disappearance was the same in germfree rats as in conventional rats. The rate of proteolysis of ['4C]choleragen and ['4C]BSA by intestinal mucosal lysosomal enzymes was also studied. Neither was significantly degraded by neutral proteases; however, heat-inactivated toxin was. They were all degraded by acid proteases; however, the rate of BSA proteolysis was only one-third of that of toxin. Soybean trypsin inhibitor had no effect on the in vivo disappearance of toxin nor on the acid proteases. It did inhibit the neutral protease digestion of heat-treated toxin. Apirotinin and protamine inhibited disappearance in loops of gut but had no effect to inhibit degradation rates. Gangliosides inhibited both rates of disappearance and proteolysis of toxin. These agents had some different effects on disappearance rates and proteolysis of BSA. The data indicate that cholera enterotoxin is absorbed by intestinal mucosal cells and is degraded by acid proteases in the cells.
Vibrio cholerae produces an enterotoxin which, when released by the organism, binds rapidly to the intestinal mucosal surface (19). The effect of the toxin is to stimulate the secretion of fluid and the electrolytes. Five minutes after binding of the toxin to the mucosal surface, this secretory response cannot be prevented by washing the surface in an attempt to remove the bound toxin (11). The toxin's effect has been shown to be mediated by stimulation of the enzyme adenyl cyclase to increase intracellular concentrations of adenosine 3',5'-cyclic monophosphate (6, 13, 15). Adenyl cyclase in the intestinal mucosal epithelial cell is located in the plasma membranes of the basal and lateral margins of the cells, and it does not appear to be a normal constituent of the brush border (4, 9). Thus, cholera toxin exerts its effect by either (i) causing an event to occur in the brush border, which leads to increased activity of adenyl cyclase in the basal and lateral cell membranes, or (ii) entering the cell and stimulating adenyl cyclase directly after intracellular transport of the toxin to the basal and lateral membranes. At the present time, there is some question of whether cholera toxin remains entirely on the surface while exerting its effect, or whether it enters the epithelial cells after binding to the surface. An autoradiographic study demonstrated that cholera toxin labeled with
tritium does not enter the mucosal epithelial cell (10), whereas an immunofluorescent study suggested that cholera toxin does enter these cells (5). The purpose of the present investigation was to study the fate of cholera toxin in the intestine, with a more specific goal being to determine whether the toxin is absorbed by the intestinal mucosal cells. It would seem logical to assume that antigenically competent cholera toxin must be absorbed for the production of antibodies that confer immunity to protect against the toxin after a bout of the disease. In the intestine, these specific immunoglobulin A antibodies against cholera toxin are produced by plasma cells in the lamina propria and then are secreted into the lumen of the gut (12).
MATERIALS AND METHODS Animals. The stimulation of secretion of intestinal fluid by cholera toxin was studied in the rat by using a method described previously (16). Male Fisher rats, 200 to 300 g, were prepared for experiments by withholding food for 24 h but allowing them to drink a 5% glucose solution ad libitum. Under ether anesthesia, the small intestine was ligated at a point just distal to the ligament of Treitz, and its contents were flushed into the cecum with 10 ml of 0.9% saline, which was introduced with a syringe and 27-gauge needle. After the intestine was emptied, the distal end was ligated. Three 12-cm loops of jejunum were measured and ligated at 1450
VOL. 12, 1975
['4C]CHOLERAGEN AND ['4C]BSA ABSORPTION
1451
each end to form closed loops, which were then mg to 40 Atg of toxin before loop inoculation. This injected with either 40 ,ug of ['4C]choleragen or 40 jig was done to see if the effect of aprotinin to inhibit of ['4C]bovine serum albumin (BSA) in a volume of the disappearance rate of toxin might be a property 0.5 ml. The loops were removed 6 h later and of a basic peptide that is not a protease inhibitor. Gangliosides bind cholera toxin and prevent attachweighed with and without their fluid contents. The rate of disappearance of ['4C]choleragen from ment to the mucosal receptor sites. To study these closed loops of intestine in germfree rats was deter- glycolipids' effects on the disappearance rate of mined to ascertain if the intestinal microflora in ['4C]choleragen, 1 mg of ganglioside (type II, bovine conventional rats can alter choleragen and hence brain, Sigma Chemical Co.) was premixed with 40 change its rate of disappearance. Fisher germfree ,ug of toxin before injection into loops. In another rats were bred and raised using germfree equipment group ganglioside was combined with cerebrosides and germfree methods described previously (17). (18) and added to toxin so that the equivalent of 1 mg The protocol used in the germfree rats was the same of ganglioside was used with 40 ,ug of ['4C]choleragen before injection into loops. as that used for conventional rats. Calculation of disappearance of ['4C]cholerCholera toxin. A highly purified cholera toxin, choleragen lot 1071, was used which was prepared agen and ['4C]BSA. Loop fluid in scintillation vials under contract for the National Institute of Allergy was desiccated in vacuo with P205. Protosol (New and Infectious Diseases by R. A. Finkelstein, Uni- England Nuclear) was used to solubilize loop tissue versity of Texas Southwestern Medical School, Dal- and also the residue remaining after evaporation of las, Tex., according to a procedure described previ- loop fluid. Aquasol (New England Nuclear) was ously (3). It was reconstituted in 0.01 M phosphate- added to the solubilized specimens, and radioactivbuffered saline (PBS) (pH 7.4) to achieve a concentra- ity was determined in a scintillation counter. The tion of 200 ,ug of choleragen/ml. Choleragen was sum of the total radioactivity in the loop fluid and in kindly provided by Carl E. Miller, the Cholera Pro- the tissue at the end of 6 h was used to calculate the gram Officer of the National Institute of Allergy and percentage of injected counts that was left. The counts remaining at 6 h could have represented Infectious Diseases. '4C labeling of choleragen and BSA. Five milli- label present in both intact choleragen and its degragrams of choleragen in 25 ml of PBS was labeled dation products which would result from digestion with 14C after a procedure described previously (16). either in the lumen of the intestine or within the Unlabeled choleragen was first dialyzed against 0.1 cell. Moreover, the radioactivity that had disapM sodium borate, pH 9. It was then subjected to peared from the intestinal loops is also assumed to reductive alkylation by the addition of 20 ,uCi of represent degraded ['4Clcholeragen and possilby in[14C]formaldehyde (specific activity, 59.2 mCi/mmol; tact toxin. Thus, what is being measured in deterNew England Nuclear Corp.), which was followed mining the disappearance rate is the result of both by the addition of 4.0 mg of sodium borohydride. The digestive and absorptive processes in the intestine. reaction was carried out for 30 min at 38 C. The Parts of the study were designed to ascertain the labeled choleragen was dialyzed against PBS (pH relative importance of digestion versus absorption 7.4) until the level of radioactivity in the dialysant in the disappearance of labeled choleragen from the was no greater than the background activity. The loops. Germfree rats and proteinase inhibitors were specific activity of the labeled toxin was 7.2 used in some groups to estimate the contribution of mCi/mmol. By inoculation into rat intestinal loops intraluminal digestion of ['4C]choleragen to its rate and into rabbits intradermally, it was verified that of disappearance. Incubation of ['4C]choleragen and [14C]BSA choleragen had not lost any measurable amount of its toxicity in the labeling process. Forty micro- with mucosal cell homogenates. The mucosa was grams of labeled choleragen produced a maximum scraped from the jejunum of fasted rats and homogesecretory response in loops of jejunum, and [14C]- nized in 10 volumes of 0.15 M sucrose and 5 mM choleragen had the same number of bluing doses per ethylenediaminetetraacetate, pH 7.0, with a motormicrogram as unlabeled choleragen. No proteins driven Teflon pestle in a Potter-Elvehjem homogewere added to stabilize the labeled choleragen, nizer. The nuclear and cell debris and heavy mitowhich was stored at 4 C until used. Twenty-five chondria were removed by centrifugation at 4,000 x milligrams of BSA was dissolved in 25 ml of PBS and g for 10 min. The supernatant was centrifuged at labeled in exactly the same manner as choleragen. 55,000 x g for 10 min to yield a pellet containing It was stored frozen until used. The ['4C]albumin light mitochondria and lysosomes. Both the pellet (after resuspension in 10 volumes of buffer) and the had a specific activity of 10.2 mCi/mmol. Modification of [14C]choleragen and ['4C]BSA supernatant were used to study degradation of choldisappearance rate. Heat inactivation of ['4C]cho- eragen and BSA and were assayed for cathepsin (1) leragen was accomplished by heating to 85 C for 10 and N-acetyl glucosaminidase (using p-nitrophenylmin with agitation. It was injected into loops and N-acetyl-,8-D-glucosaminide as substrate in citrate added to mucosal lysosomes to study the effect of buffer, 0.05 M, pH 4.2). Assays for degradation of the denaturation of toxin on absorption and digestion. proteins consisted of 20 Ag of either ['4C]choleragen The protease inhibitors aprotinin (Trasylol) and or ['4C]BSA in either a PBS buffer (pH 7.0) or an soybean trypsin inhibitor were added at levels of acetate buffer in which the pH of the entire incubate 3,000 KIU and 10 mg, respectively, to 40 ,ug of toxin was 4.5. At the onset of incubation 100 p.l of either before injection into loops. In one study, protamine the 55,000 x g pellet suspension or the supernatant sulfate, which is a basic peptide similar in size and from this centrifugation was added, and for the conpK properties to aprotinin, was added at a level of 3 trols, they were added at the end of the incubation.
1452
STROMBECK AND HARROLD
INFECT. IMMUN.
Aprotinin (1,000 KIU), soybean trypsin inhibitor (1 mg), or cerebroganglioside (containing 0.33 mg of ganglioside) was added to the assays at the onset and to the controls at the end of the incubation. They were incubated for 4 h and stopped by addition of ice cold trichloroacetic acid to give a final concentration of 5%. After centrifugation, the supernatant was added to Aquasol and counted in a scintillation counter. A time study of the release of trichloroacetic acid-soluble label over a period of 6 h was done with both choleragen and BSA. Electrophoresis of heat-treated and unaltered [(4C]choleragen. The effects of heat treatment on ['4C]choleragen were determined by comparing the electrophoretic patterns of heated toxin with untreated toxin following methods previously described (16). They were both incubated for 30 min at 37 C in 0.01 M sodium phosphate buffer (pH 7.1) containing 0.5% sodium dodecyl sulfate and 0.5% 2mercaptoethanol, after which electrophoresis was performed in 10% polyacrylamide gels. In another group, the two toxins were prepared in 0.1 M borate buffer (pH 9), and electrophoresis was done on 10% polyacrylamide gels. Protein was detected by staining 2 to 3 h with Coomassie brilliant blue (0.25% in methanol)-water-acetic acid (5:5:1, vol/vol). They were destained in water- acetic acid- methanol (35:3:2, vol/vol) (20). The relative amount of protei'n in each band was estimated by scanning with a Gilford gel densitometer. The gels were sliced and each part was dissolved in 30% H202, after which Aquasol was added, and radioactivity was determined in a scintillation counter.
RESULTS
Fluid production and disappearance of ['4C]choleragen from loops of jejunum after 6 h. A substantial amount of fluid was produced when ['4C]choleragen was injected into loops of
conventional rats (Table 1). Greater than 30% of the radioactivity placed in the loops was not present 6 h later. The results of injection of labeled toxin into loops of germfree rats produced similar results. The addition of aprotinin to choleragen had no effect on fluid production but reduced the rate of disappearance of choleragen to 40% of that in loops with choleragen alone. Similar results were seen with protamine. Soybean trypsin inhibitor had no effect on fluid production or the disappearance rate of choleragen. Ganglioside and cerebroganglioside inhibited fluid production completely and also caused the rate of choleragen disappearance to fall more than 50%. A very small amount of fluid was produced by heat-inactivated choleragen. This treatment of the toxin caused it to disappear at twice the rate of choleragen. Treatment of heat-inactivated toxin with aprotinin or ganglioside produced similar effects as these compounds had on choleragen. The calculation of disappearance rates in loops with fluid was based on counts in loop fluid which represented 92 to 93% of the total counts at 6 h and on those in loop tissue which amounted to 7 to 8% of the total. These percentages were not significantly different between any of the groups of loops that produced fluid. In loops with no fluid, no attempt was made to flush the mucosal surface and to determine how much radioactivity was bound to or within the tissue and how much was free on the surface. There was no difference in the disappearance rate of ['4C]choleragen between the three loops in each rat. Therefore, the data were pooled for
TABLE 1. Fluid production and disappearance of ["4C]choleragen from loops
Group'
ofijejunum after 6 h
Fluid production (g of fluid secreted/g of inNo. of loops (3/rat) testine [wet wt] per 6 ha)
Disappearance of'
['4C]choleragen maining in 6 (%ha) re-
Choleragen in: Conventional rats Germfree rats
84 24
4.87 ± 0.26 4.96 ± 0.39
69.85 ± 1.10 67.76 ± 2.94
Conventional rats: Choleragen Choleragen + aprotinin Choleragen + soybean trysin inhibitor Choleragen + ganglioside
84 35 29 57
4.87 ± 0.26 5.16 ± 0.26 4.00 ± 0.46 0 0 4.93 ± 0.30 0.61 ± 0.21 0.37 ± 0.11 0
69.85 88.40 71.76 85.05 89.31 83.96 39.19 62.01
Choleragen
+
cerebroganglioside
Choleragen + protamine
Heated choleragen Heated choleragen + aprotinin Heated choleragen + ganglioside a
b
Mean ± standard error of the mean. P < 0.001 compared to choleragen. P < 0.001 compared to heated choleragen.
25
23 24 21 21
± 1.10 ± 1.14 ± 2.24 ± 1.07b ± 1.68 ± 1.93
±1.65
+ 1.06" 52.84 ± 1.16"
VOL. 12, 1975
[14C]CHOLERAGEN AND ['4C]BSA ABSORPTION
each group with no breakdown as to the region of the jejunum from which the loop was taken. Disappearance of ['4CIBSA from loops of jejunum after 6 h. The rate of disappearance of labeled albumin was the greatest in the first part of the jejunum (Table 2). Aprotinin did not change the rate of disappearance, but ganglioside increased it significantly. Release of trichloroacetic acid-soluble '4C from ['4C]choleragen and ["CIBSA after incubation with intestinal mucosal proteases. Proteolytic activity is suggested to have been present in excess in all the lysosomal and 55,000 x g supernatant
1453
assays did not change the rate of appearance of trichloroacetic acid-soluble '"C. The rate of appearance of trichloroacetic acid-soluble '"C label from [''Cicholeragen and ['"C]BSA is depicted in Fig. 1. They are linear for both; however, the rate of label release from BSA was one-third of that from choleragen. The action of the lysosomal fraction of the cells to release '4C from labeled choleragen and BSA is shown in Table 3. A very small amount of degradation occurred at pH 7.0. This activity increased 10-fold at pH 4.5. Neither aprotinin or soybean trypsin inhibitor had any effect on the rate of release of '4C. Cerebroganglioside inhibited the rate of choleragen degradation 40
preparations since doubling enzymes added to the
the amount of lysosomal
30
20 0
10
~~~~~~~~~~~BSA
0
2
4
3
5
6
TIME (HOURS)
FIG. 1. Rate of appearance of trichloroacetic acid (TCA)-soluble '4C from ['C]choleragen and [1 C]BSA during incubation with intestinal mucosal acid proteases. [''Clcholeragen: y 6.568 (x) 0.303, r 0.998, standard error = 0.753. ['4C]BSA: y = 2.108 (X) - 0.898, r = 0.991, standard error = 0.584. =
-
=
TABLE 2. Disappearance of 1'4C BSA from loops ofjejunum after 6 h Groupa Determinants
No. of animals
Disappearance of [4C]BSAb Proximal' Middle Distal
Probability Proximal versus middle Proximal versus distal Proximal versus proximal Middle versus middle Distal versus distal a
I
II
III
11
9
10
41.2 + 6.6 65.3 ± 6.97 71.2 ± 5.81
45.9 + 4.34 68.8 + 3.26 76.6 + 3.97
14.9 + 2.09 37.4 + 6.93 54.3 + 8.83
