Vol. 58, No. 11
INFECTION AND IMMUNITY, Nov. 1990, p. 3516-3522
0019-9567/90/113516-07$02.00/0 Copyright C 1990, American Society for Microbiology
Excystation of In Vitro-Derived Giardia lamblia Cysts SHAYNE-EMILE M. BOUCHER AND FRANCES D. GILLIN*
Department of Pathology, H811F, University of California, San Diego Medical Center, 225 Dickinson Street, San Diego, California 92103 Received 16 April 1990/Accepted 10 August 1990
This is the first in-depth analysis of the excystation of Giardia lamblia cysts prepared in vitro. Its goals were both to achieve efficient excystation and to gain insights into this crucial but poorly understood process. To identify the critical elements of excystation, we tested the sequential low-pH induction and protease treatments which had been reported to be important for excystation of fecal cysts. The optimal pH for induction of excystation was 4.0. Emergence was greatly (-10-fold) stimulated by subsequent exposure of in vitro-derived cysts to chymotrypsin, trypsin, or human pancreatic fluid. The stimulatory activity of each was abolished by soybean trypsin inhibitor, demonstrating that the activity of pancreatic fluid was due to these proteases. Excystation of in vitro-derived cysts was -10 to 38%. Although the walls of in vitro-derived cysts were partially digested by protease treatment, trophozoites emerged only from one pole, as observed with fecal cysts. The conditions of encystation also determined the efficiency of excystation. Specifically, encystation in the presence of lactic acid, a major metabolite of colonic bacteria, stimulated excystation approximately fourfold, although it did not increase the total numbers of cysts. These experiments have shown that excystation of in vitro-derived cysts reflects that of cysts purified from human feces in that it is dependent upon conditions which simulate the passage of cysts through the human stomach (low pH) and into the small intestine (pancreatic proteases). elles visible in relief by differential interference contrast microscopy). Some of the cysts (range, -1 to 9.5% of total cysts) completed the life cycle by excysting in vitro, but most of the trophozoites emerged only partially from the cyst wall. Since >90% of the type I cysts were viable (7), as determined by the fluorogenic dye method of Schupp and Erlandsen (19), we hypothesized that the efficiency of excystation could be increased. Therefore, the goals of the present study were to achieve high-efficiency excystation of in vitro-derived G. lamblia cysts and to gain an increased understanding of this important biologic process.
Since excystation of ingested Giardia lamblia cysts is essential to the transmission of disease, it may be considered a virulence factor in human giardiasis. Nonetheless, this process is not well understood, largely because, until recently, the only source of G. lamblia cysts was the feces of infected humans or experimentally infected animals. Excystation of fecal cysts varied from 95% (17; unpublished data). Cysts prepared in vitro have the advantages of being free of fecal contaminants and of more consistent physiologic conditions for studies of excystation. During natural infection, ingested cysts pass through the stomach of the host, where they are exposed to gastric acid (2). Subsequently, cysts enter the duodenum, where the gastric chyme is rapidly neutralized by influxes of bicarbonate (3). It is important that trophozoites not emerge from cysts in the stomach because they would be killed by the acid. Excystation is presumed to occur in the upper small intestine, but the exact site is not known. Descent past the entrance of the common bile duct would expose the cysts to a variety of degradative enzymes and bile salts, which have detergent activity (3, 10). The pioneering studies of fecal cysts by Bingham et al. (1, 2) demonstrated that excystation was dependent upon "those conditions most closely approximating the organism's in vivo environment" (2). Specifically, they showed the importance of exposing cysts to a low pH, which mimics passage through the stomach. Trophozoites emerged when the acid-treated cysts were transferred to a neutral medium, reproducing entry into the duodenum (2). Excystation methods published since then have incorporated other physiologic intestinal factors, including bicarbonate, trypsin (a pancreatic protease), thiol-reducing agents (16, 20), and bile salts (11). In a previous study (7), we demonstrated the production of large numbers (>105/ml) of water-resistant cysts with type I morphology (smooth, oval, phase bright, with cyst organ*
MATERIALS AND METHODS
Materials. Unless otherwise noted, all reagents were obtained from Sigma Chemical Co. (St. Louis, Mo.). Bicarbonate was purchased from Fisher Scientific, and crude trypsin, from U.S. Biochemicals, was the gift of Judith Sauch. Cysts were isolated from the feces of a patient with symptomatic giardiasis by the method of Douglas et al. (5). Human pancreatic fluid. An adult male volunteer was intubated orally with a duodenal tube (AN 20; A. W. Anderson, Santa Monica, Calif.). The tube was allowed to pass through the stomach and into the duodenum under fluoroscopic guidance to an area between the ampulla of Vater (common bile duct) and the beginning of the jejunum. Before samples were collected, all residual pancreatic and duodenal fluids were withdrawn through the duodenal tube and discarded. Basal pancreatic secretions were collected for 15 min. Pancreatic secretions were then stimulated by intravenous injection of 75 U of secretin (Ferring Laboratories, Sufferin, N.Y.). Two sequential 15-min stimulated pancreatic fluid samples were then collected. Parasite cultivation. G. lamblia WB (ATCC 30957) was routinely cultivated in Diamond TYI-S-33 medium (4) with 10% adult bovine serum (Irvine Scientific) and bovine bile (13) but without added iron, vitamins, or antibiotics as previously described (7), with subculturing twice weekly. The same lots of Biosate (BBL) and serum were used
Corresponding author. 3516
VOL. 58, 1990
throughout these experiments, since any change which affected growth tended to decrease encystation. Encystation. (i) Pre-encystation cultures. The condition of the pre-encystation trophozoite monolayer appeared to play an important role in the efficiency of in vitro encystation (7). Cultures grown to the late log phase in growth medium were chilled, inverted 12 times, and counted in a hemacytometer chamber. Trophozoites (5,000 per ml, final concentration) were added to chilled pre-encystation medium. The preencystation medium was freshly prepared TYI-S-33 growth medium (pH 7.1) containing the antibiotics piperacillin (500 Rg/ml; Lederle Laboratories) and amikacin (125 ,ug/ml; Bristol Laboratories) but no bovine bile. Pre-encystation cultures (in 8-ml borosilicate glass screw-capped tubes [13 by 100 mm]) were grown for 3 days at 37.5°C and at 5° from horizontal. Cultures were still in the log phase, and monolayers were -50 to 80% confluent. The tubes were inverted eight times, and the unattached trophozoites and medium were discarded. The attached trophozoite monolayers were refed with fresh encystation medium (see below). (ii) Encystation cultures. Unless otherwise specified, the encystation medium consisted of pre-encystation medium adjusted to pH 7.8 with 1 M NaOH and supplemented with porcine bile (0.25 mg/ml, final concentration) and lactic acid (hemi-calcium salt, 5 mM, final concentration) (7). The medium, bile (10-mg/ml stock solution), and lactic acid (100 mM stock solution) were all freshly prepared and filter sterilized separately. Encysting trophozoites were incubated for 66 h, since initial experiments showed that this length of incubation was optimal for the biologic activity of cysts (unpublished data). Parasites were harvested and waterresistant cysts were isolated as described previously (7), except that the cultures were not chilled. Tubes were inverted (eight times), and nonadherent cells, including cysts, were decanted into 15-ml conical centrifuge tubes, pelleted, washed in 15 ml of double-distilled water, and incubated for 30 to 45 min at room temperature in 15 ml of double-distilled water. Trophozoites and cysts with incomplete walls were lysed by this treatment. Type I and type II water-resistant cysts were collected by low-speed centrifugation (5 min, 135 x g, 4 to 10°C), suspended in the volume (-200 ,ul) of water remaining in the tube, and enumerated in a hemacytometer chamber by differential interference contrast microscopy (7). As defined previously (7), type I cysts have a smooth, oval shape, are somewhat refractile, and have cyst organelles appearing in relief. Cysts not fulfilling these criteria are referred to as type II. Excystation. Unless otherwise specified, excystation was carried out on the same day that in vitro-derived cysts were harvested. Our standard excystation procedure incorporates variations of the two-step method of Rice and Schaefer for fecal cysts (16), based on the results of the experiments presented here. The variables tested are described in each experiment. A low-pH induction solution was prepared fresh by mixing 6.8 ml of Hanks balanced salt solution containing L-cysteine hydrochloride (57 mM) and reduced glutathione (32.5 mM) with 6.8 ml 0.1 M NaHCO3 and 11.3 ml of double-distilled water. The pH of the solution was adjusted to 4.0 with 0.01 N HCl. In the first, or induction step, 50 ,ul of cyst suspension (5 x 103 to 5 x 104 cysts) and 0.55 ml of the low-pH induction solution were mixed in a 1.5-ml Eppendorf tube. The tube was capped, vortexed, and incubated for 30 min in a 37.5°C water bath. The cysts were sedimented for 3 min at 135 x g and room temperature, and the supernatant was aspirated and discarded. In the second, or excystation, step, 1 ml of a prewarmed
EXCYSTATION OF G. LAMBLIA
3517
mixture containing 1 mg of chymotrypsin (from bovine pancreas, 3X crystallized, 50 U/mg of protein) in Tyrode salt solution (16), with the pH raised to 8.0 with freshly prepared 7.5% sodium bicarbonate, was added to the cyst pellet. The tube was capped, vortexed, and incubated for 30 min in a 37.5°C water bath. The cysts were sedimented for 3 min at 135 x g and room temperature, and the supernatant was removed to avoid prolonged exposure to chymotrypsin, which was detrimental to the emerging trophozoites. After the excystation step, 100 ,ul of prewarmed TYI-S-33 growth medium was added to the pellet, which was suspended with a Pipetman, to promote trophozoite motility and viability. Twenty microliters of each sample was loaded into a hemacytometer chamber, which was placed in a 37°C humidified chamber. Partially and totally excysted trophozoites and intact cysts were enumerated by differential interference contract microscopy with a 20X objective. Early experiments showed that the percent excystation did not change significantly during the time required to count all the samples (-4 h). The formula of Bingham et al. (1) was used to determine percent excystation. The original formula was as follows: percent excystation = {[(TET/2) + PET]/[(TET/2) + PET + IC]} x 100, where TET is the number of totally excysted trophozoites, PET is the number of partially excysted trophozoites, and IC is the number of intact cysts. PET are motile and active but not completely emerged from the cyst wall. This formula included a twofold factor based on the observation that TET from fecal cysts divided immediately upon emergence, producing two trophozoites (1). However, on the basis of morphologic observations, the division of strain WB trophozoites emerging from cysts prepared in vitro was delayed by several hours. Therefore, for in vitro cysts, we did not divide TET by two, unless the trophozoites had already divided, as determined microscopically. Each figure shown is from a single experiment representative of two to six repetitions with different cyst preparations. Each determination represents an average of four fields with -200 cells per field. The results of each repetition were consistent with those shown in the figures. Significance was determined by Student's t test. RESULTS Roles of acid and protease treatments in excystation. In an initial experiment, we assessed two published methods of excystation. With the method of Rice and Schaefer (16), we observed 13.5% + 0.8% excystation. Moreover, >97% of the excysted trophozoites emerged completely from the cyst wall. In contrast, with the method of Schupp et al. (20), which does not include protease treatment, we observed only 4.3% + 0.8% excystation. Moreover, most (90.1%) of the trophozoites were only partially excysted, meaning that the trophozoites did not emerge completely from the cyst wall or remained partially adherent to the cyst shell (P < 0.0005). Therefore, we investigated the two major parameters of the method of Rice and Schaefer: pH in the first, or induction, step and treatment of cysts with protease in the second, or emergence, step. The studies of Bingham et al. (1, 2) demonstrated the importance of the exposure of fecal cysts to a low pH for subsequent excystation. Therefore, we assessed the effect of exposure to an acidic pH on excystation of in vitro-prepared cysts (Fig. 1). The pH curve was quite broad, with 10 to >35% excystation after exposure to pH 2 to 8, with an optimum at pH 4.0, which was used in subsequent experi-
INFECT. IMMUN.
BOUCHER AND GILLIN
3518
N211
40-
6 ('n
40
110
110
030
]O
i6
D
o qC'J 0 11 0
Cl)
0
(1)0Q
E . 0)
0C] 0
QOL
20-
cxi11
o
_
0-
2
4
I
R 10'
2.0 5
6
7
ments. In other experiments, excystation did not differ greatly at pHs between 2 and 4 but always decreased
gradually with increasing pH above pH 4. Although the proportion of totally excysted trophozoites decreased slightly with increased pH, it remained >80%. Following exposure to gastric acid, ingested cysts pass into the small intestine, where they are exposed to pancreatic proteases at a slightly alkaline pH. This process was modeled by Rice and Schaefer (16), who treated fecal cysts with a crude preparation of trypsin. To elucidate the role of proteases in excystation, we exposed in vitro-derived cysts to crude and purified trypsin preparations and to purified chymotrypsin. The highest level of excystation (19.2% 7.7%) was observed following exposure of in vitro-derived cysts to purified chymotrypsin, as compared with 15.7% + 5.1% with crude trypsin at 5 mg/ml or 6.8% 1.6% with purified trypsin at 1 or 5 mglml (data not shown). Since chymotrypsin was effective at lower concentrations, we used it in all subsequent experiments. We next compared the effects of a low pH and proteolytic treatment on the excystation of cysts isolated (5) from the feces of a patient with symptomatic giardiasis (Fig. 2). Excystation was maximal after exposure to pH 2.0 and almost absent after exposure to pH 6.0. Moreover, chymotrypsin stimulated the excystation of fecal cysts by approximately twofold. The maximal excystation (-38%) of fecal cysts in this experiment was directly comparable to that of the in vitro-derived cysts in Fig. 1, since these experiments were performed at the same time. Since trypsin and chymotrypsin have different substrate specificities, we next tested whether other endoproteases would also stimulate excystation of in vitro-derived cysts. Four relatively nonspecific proteases, proteinase K, subtilisin, thermolysin, and elastase, stimulated excystation as efficiently (14.0% 3.3% to 19.7% 1.6%) as did chymotrypsin (16.5% 1.2%; P > 0.05), but neither the exoproteases carboxypeptidase A and leucine aminopeptidase nor pepsin (incorporated into the low-pH induction step) stimulated excystation (data not shown). Stimulation of excystation of in vitro-derived cysts by pancreatic fluid. Since trypsin and chymotrypsin are secreted into the small intestine by the pancreas, we tested the effects ±
±
4.0
6.0
pH
FIG. 1. Effect of various pHs on the excystation of in vitroderived cysts. Water-resistant cysts were incubated in cysteineglutathione-bicarbonate solution, the pH of which was adjusted with dilute HCI or NaOH. Otherwise, the procedure was as described in Materials and Methods. S.D., Standard deviation of the total percent excystation. The P values are for exposure to the indicated pH, compared with pH 4.0.
±
g 20 a
0O
pH
±
+ CT + CT, boiled
0
D Vv
10-
El
30
C\
CM 0 11 0
2 a
PET TET
FIG. 2. Roles of low pH and chymotrypsin in the excystation of fecal cysts. The percent excystation of fecal cysts can be directly compared with that of in vitro-derived cysts in Fig. 1, since these experiments were carried out at the same time. The second step of excystation was carried out in the presence of a 1-mg/ml concentration of active chymotrypsin (CT) or chymotrypsin which had been inactivated by boiling for 20 min and cooled prior to the experiment. Excystation was significantly decreased following exposure to pH 4 or 6 (P s 0.001) or to boiled chymotrypsin (P < 0.001).
of pancreatic secretions from a healthy human volunteer on excystation. Pancreatic fluids collected from the same subject before and after stimulation with secretin (3) greatly increased excystation (P s 0.002) (Fig. 3). Stimulated pancreatic secretions were more effective at lower concentrations and consistently yielded higher levels of excystation than unstimulated pancreatic secretions. Both stimulated pancreatic fluid and chymotrypsin increased excystation over a >100-fold concentration range. Concentrations of stimulated pancreatic fluid above 2% or of chymotrypsin above 1 mg/ml led to decreased excystation (data not shown). To determine whether the stimulation of excystation by pancreatic fluid was due to chymotrypsin and/or trypsin activity, we assessed the effect of soybean trypsin inhibitor, an inhibitor of both proteases (Fig. 4). Stimulation of excystation by both chymotrypsin and pancreatic fluid was virtually eliminated by soybean trypsin inhibitor (P < 0.01). Effect of cyst storage on biological activity. Since Bingham et al. (1) showed that fecal cysts were able to excyst after >30 days of storage in distilled water at 8°C, we monitored the biologic activity of two preparations of in vitro-derived cysts from the day they were harvested (day 1 in Fig. 5). Biologic activity was detected for up to 26 days of incubation at 4°C, after which excystation dropped below 0.1%. Although the initial percent excystation of the two cyst preparations differed by a factor of approximately three, it was highest on the day of harvest and declined with incubation at 40C. Effect of encystation conditions on the biologic activity of cysts. Using the optimized excystation procedure, we tested the hypothesis that the conditions of encystation determine cyst biologic activity and the idea that metabolites of colonic bacteria may be important for this process. To do this, we assessed the effects of organic acids produced by the colonic bacterial flora (3) on both the numbers of cysts and the efficiency of excystation. None of the organic acids tested increased the numbers of total cysts as compared with porcine bile alone (at pH 7.8), although lactic acid led to slightly increased numbers of type I cysts (Fig. 6A) (7). In contrast, the percent excystation was increased more than fourfold by the inclusion of 5 mM lactic acid in the encystation medium (P = 0.001) (Fig. 6B).
EXCYSTATION OF G. LAMBLIA
VOL. 58, 1990
3519
t Unstimulated
Stimulated *
a
20 *
*F
T
10 *
None
2
5
10
25
0.02 0.05
0.5
10
2.0
Pancreatic Fluid, %
50
250 1000
Chymotrypsin,
pg/ml FIG. 3. Stimulation of the excystation of in vitro-derived cysts by pancreatic fluid. Unstimulated or stimulated pancreatic fluid or chymotrypsin at the concentrations shown was substituted for the standard 1-mg/ml concentration of chymotrypsin in the second step of excystation. *, Significant increase in percent excystation (P 0.002), compared with the control with no pancreatic fluid or chymotrypsin. -
DISCUSSION Taken together, data from our laboratory (6-9, 24) and other laboratories (e.g., 1, 2, 11, 13, 16, 23) show that conditions encountered by G. lamblia during each step of its natural life cycle are crucial to reproducing that step in vitro. By varying the conditions of encystation (7) and excystation, we have consistently obtained in vitro-derived cysts of G. lamblia with levels of biologic activity comparable to those of fecal cysts. Our experiments, based on earlier work with fecal cysts (2, 16), have permitted us to identify two parameters necessary for the excystation of in vitro-derived cysts. The first, exposure of in vitro-derived cysts to a low pH, which mimics the passage of ingested cysts through the stomach, was shown by Bingham et al. (1, 2) to be crucial for the excystation of fecal cysts. The second, exposure of in vitro-derived cysts to trypsin, as reported by Rice and Schaefer (16) for fecal cysts, or to chymotrypsin, mimics the
passage of cysts into the lower duodenum, where they are bathed in pancreatic secretions containing these proteases (3). At the optimal pH for each, we observed equal excystation of fecal and in vitro-derived cysts. The optimal pH (2.0) for triggering the excystation of fecal cysts was slightly lower than that for triggering the excystation of in vitroderived cysts (4.0). A second difference was that excystation of in vitro-derived cysts was totally dependent upon exposure to proteases, whereas excystation of fecal cysts was depressed only by 50 to 60% in the absence of proteases. Differences between the excystation of in vitro-derived and fecal cysts may simply be related to differences between strains. The WB strain, which typifies the most common group of G. lamblia isolates (14), was isolated in 1979 from a patient infected in Afghanistan, whereas the fecal cysts used in this study were isolated from a patient infected in Kenya. Alternatively, the physical state of the wall of in vitro-
*
20-
*
E3
No Addition 1 mg/ml Chymotrypsin 2% Pancreatic Fluid, stim. 25% Pancreatic Fluid, unstim.
10-
*
* *
n~
I* ~
*
5
0
Soybean Trypsin Inhibitor, mg/ml FIG. 4. Inhibition by soybean trypsin inhibitor of stimulation of the excystation of in vitro-derived cysts by pancreatic fluid or chymotrypsin. Chymotrypsin or pancreatic fluid was preincubated with soybean trypsin inhibitor at the indicated concentrations for 15 min at 37°C prior to being used in the second step of excystation. *, Significant inhibition of excystation by soybean trypsin inhibitor (P < 0.001). **, Significant inhibition (P = 0.006); stim., stimulated; unstim., unstimulated.
3520
INFECT. IMMUN.
BOUCHER AND GILLIN 30
pg20
D ..- --
cyst prep. A cyst prep. B
10
0
1
7
4
10
13
16
19
22
25
28
Day FIG. 5. Effect of storage at 4°C on the excystation of in vitro-derived cysts. Water-resistant cysts were harvested on day 1 and kept in an ice water bath in double-distilled water with the antibiotics piperacillin and amikacin. At the times indicated, samples were removed and excysted by the standard procedure. A and B are two different cyst preparations (cyst prep.). Percent excystation was significantly decreased on and after day 5 (P 0.002) for preparation A and day 9 (P < 0.002) for preparation B. -
derived cysts mlay differ in subtle ways from that of the wall of fecal cysts. This difference may be due to incubation of the latter withiin the fecal mass both before and after passage. For exannple, the cyst wall may be acted upon by bacterial or hos,t enzymes or metabolites.
Exposure of cysts to an acidic pH has been a hallmark of published excystation procedures (1, 2, 11, 17, 18). The pH curve for excystation of in vitro-derived cysts was broad. The percent excystation was maximal at pH 2 to 4 but always declined steadily at pHs above 5. Even so, substantial excystation (25 to 35% of maximal) was observed following exposure to pH 7 to 8. This result may have been due 8- A to prior exposure of cysts during harvesting, incubating, and * Type cysts washing in double-distilled water at a pH of -5.7. Such * Type 11 cysts exposure may trigger low levels of excystation which can be i 6-1 subsequently increased by exposure to a lower pH. This process may explain the incidence of giardiasis in patients in_ Ito with reduced gastric acidity. In the present study, excystation of in vitro-derived cysts 0 was >90% dependent on exposure to proteases in the second 3 * -step. The stimulatory activity of human pancreatic fluid was to trypsin and chymotrypsin, since it was virtually _ _ ~~~~~~~~due __ abolished by soybean trypsin inhibitor, an inhibitor of both 1 mM5mM 1 M 5mM None 1 1 mM 1 mM 5M 5mM proteases. This observation shows that other components in + LA + AA + BA + SA or collected with the pancreatic secretions, such as bile salts * or other enzymes, are not required, although they may have 14 B a contributory role (11). The requirement for proteases was not specific, since all endoproteases tested were effective. 12 * 1 mM The idea that proteins important components of the cyst * 5mM 010 earlier observation that cyst wall IS consistent withareour by polyacrylamide gel electrophoresis separated antigens 8 and transferred to nitrocellulose were digested by trypsin 6 _ ~~~~~~~~~~~~(15). Compared with the walls of fecal cysts, the walls of in C4vitro-derived cysts appeared by light microscopy to be more 2 gtransparent after protease treatment. However, trophozoites y _ always emerged from a pole, as reported earlier for fecal _T L +M cysts (2), suggesting the presence of a protease-sensitive None + LA + SA + BA component or area at the pole of the cyst, possibly produced Organic Acid Addition or unmasked by prior exposure of the cyst to acid and/or FIG. 6. Effect s of organic acids in the encystation medium upon thiol-reducing agents. In contrast, treatment of cysts with numbers of watei r-resistant cysts and subsequent excystation. The purified chitinase did not stimulate excystation (data not
MTN
standard 5 mM lacCtic acid was omitted from the encystation medium and replaced as in dicated with succinic acid (SA), butyric acid (BA), acetic acid (AA), or lactic acid (LA). (A) Numbers of cysts (105/ml). (B) Percent excysstation of each cyst preparation determined by the standard assay. *, Excystation was significantly increased (P = 0.001) following e ncystation in the presence of 5 mM lactic acid.
shown), suggesting that chitin, which was previously re-
ported to be a cyst wall constituent by one group (22) but not by another (12), may not be located at the site of trophozoite emergence.
Excystation is the most stringent criterion of cyst biologic
EXCYSTATION OF G. LAMBLIA
VOL. 58, 1990
activity and is extremely sensitive to the conditions of encystation. For example, although encystation in the presence of lactic acid led to relatively small (-20 to 100%) (Fig. 6) (7) increases in type I cysts but not in total cysts, it stimulated excystation strongly (more than fourfold). The idea that lactic acid may have a role in the terminal stages of encystation is consistent with the fact that it is a major metabolite of bacteria in the large intestine. Moreover, the addition of lactic acid to cultures after 24 h in encystation medium did not decrease excystation (data not shown). Our earlier study (7) showed that viability, defined by the fluorogenic dye assay of Schupp and Erlandsen (19), was a less stringent criterion for cyst quality, since viable cysts were not necessarily able to excyst (20). Similarly, fecal cysts retained the ability to exclude eosin (viability) after they lost the ability to excyst (1). In the present study, excystation of different preparations of in vitro-derived cysts using the same encystation and excystation procedures varied from -8 to 38%. This value is well within the broad range (1 to 95%) reported for cysts isolated from human patients (17). Nonetheless, the apparent efficiency of excystation of in vitro-derived cysts may be low because the calculation is based on total numbers of water-resistant cysts, a heterogeneous population of which only -10 to 30% has type I morphology. While -90% of type I cysts are viable (7) and therefore are theoretically capable of excystation, the remaining type II cysts include cysts with a shrunken cytoplasm as well as cysts which appear morphologically less mature. Only -33% of type II cysts were found viable (unpublished data) by the fluorogenic dye method (19). We have not been able to determine whether type I or type II cysts or both excyst, because the excystation procedure obscures cyst morphology. However, on the basis of morphology and viability, type I cysts may be more competent at excystation. The variability in cyst morphology is not an artifact of encystation in vitro, since both type I and type II cysts are isolated from feces (7, 19). Moreover, the numbers of cysts passed by infected humans fluctuate greatly. We have observed cyst numbers from >106/g of stool to barely detectable in the same untreated patient (unpublished data; 21), suggesting that the formation and/or shedding of cysts in vivo is highly sensitive to variations in conditions which have not yet been identified. Our in vitro encystation conditions yield more consistent numbers of biologically active cysts. Our experiments have shown that a low pH and pancreatic proteases which are important for the excystation of fecal cysts also induce high-efficiency excystation of in vitroderived cysts. These observations again underscore the value of mimicking human gastrointestinal tract conditions for completing the life cycle of G. lamblia in vitro. The availability of all stages of the life cycle of G. lamblia will permit new cellular, molecular, and immunologic studies of this important pathogen. ACKNOWLEDGMENTS
We are grateful to our colleagues at the Environmental Protection Agency for suggesting the method of Rice and Schaefer, to D. Reiner for stimulating discussions, to Dan Hogan (Division of Gastroenterology, University of California, San Diego) for the pancreatic fluid, to C. Davis and J. Sauch for critiquing the manuscript, to W. Strum (Scripps Clinic) for patients, and to S. McFarlin for preparing the manuscript. This study was supported by U.S. Environmental Protection Agency cooperative agreement CR 814537 and by Public Health
3521
Service grants AM 35108, AI 24285, and Al 19863 from the National Institutes of Health.
LITERATURE CITED 1. Bingham, A. K., E. L. Jarroll, E. A. Meyer, and S. Radulescu. 1979. Giardia sp.: physical factors of excystation in vitro and excystation vs eosin exclusion as determinants of viability. Exp. Parasitol. 47:284-291. 2. Bingham, A. K., and E. A. Meyer. 1979. Giardia excystation can be induced in vitro in acidic solutions. Nature (London) 227: 301-302. 3. Davenport, H. W. 1977. Physiology of the digestive tract. Year Book Medical Publishers, Chicago. 4. Diamond, L. S., D. Harlow, and C. C. Cunnick. 1978. A new medium for the axenic cultivation of Entamoeba histolytica and other Entamoeba. Trans. R. Soc. Trop. Med. Hyg. 72:431-432. 5. Douglas, H., D. S. Reiner, and F. D. Gilln. 1987. A new method for purification of G. lamblia cysts. Trans. R. Soc. Trop. Med. Hyg. 81:315-316. 6. Gault, M. J., F. D. Gillin, and A. J. Zenian. 1987. Giardia lamblia: human intestinal mucus and epithelial cells stimulate growth in serum-free medium. Exp. Parasitol. 64:664-668. 7. Gillin, F. D., S. E. Boucher, and D. S. Reiner. 1989. Giardia lamblia: the roles of bile, lactic acid, and pH in completion of the life cycle in vitro. Exp. Parasitol. 69:164-174. 8. Gillin, F. D., D. S. Reiner, and S. E. Boucher. 1988. Small intestinal factors promote encystation of Giardia lamblia in vitro. Infect. Immun. 56:705-707. 9. Gillin, F. D., D. S. Reiner, M. J. Gault, H. Douglas, S. Das, A. Wunderlich, and J. Sauch. 1987. Encystation and expression of cyst antigens by Giardia lamblia in vitro. Science 235:10401043. 10. Hofmann, A. F., and A. Roda. 1984. Physicochemical properties of bile acids and their relationship to biological properties: an overview of the problem. J. Lipid Res. 25:1477-1489. 11. Isaac-Renton, J., E. M. Proctor, R. Prameya, and Q. Wong. 1986. A method of excystation and culture of Giardia lamblia. Trans. R. Soc. Trop. Med. Hyg. 80:989. 12. Jarroll, E. L., P. Manning, D. G. Lindmark, J. R. Coggins, and S. L. Erlandsen. 1989. Giardia cyst wall-specific carbohydrate: evidence for the presence of galactosamine. Mol. Biochem. Parasitol. 32:121-132. 13. Keister, D. B. 1983. Axenic culture of Giardia lamblia in TYI-S-33 medium supplemented with bile. Trans. R. Soc. Trop. Med. Hyg. 77:487-488. 14. Nash, T. E., T. McCutchan, D. Keister, J. B. Dame, J. D. Conrad, and F. D. Gillin. 1985. Restriction-endonuclease analysis of DNA from 15 Giardia isolates obtained from humans and animals. J. Infect. Dis. 152:64-73. 15. Reiner, D. S., D. Douglas, and F. D. Gillin. 1989. Identification and localization of cyst-specific antigens of Giardia lamblia. Infect. Immun. 57:963-968. 16. Rice, E. W., and F. W. Schaefer III. 1981. Improved in vitro excystation procedure for Giardia lamblia cysts. J. Clin. Microbiol. 14:709-710. 17. Sauch, J. F. 1988. A new method for excystation of Giardia, p. 261-264. In P. M. Wallis and B. R. Hammond (ed.), Advances in Giardia research. University of Calgary Press, Calgary, Alberta, Canada. 18. Schaefer, F. W., III. 1988. A review of methods that are used to determine Giardia cyst viability, p. 249-254. In P. M. Wallis and B. R. Hammond (ed.), Advances in Giardia research. University of Calgary Press, Calgary, Alberta, Canada. 19. Schupp, D. G., and S. L. Erlandsen. 1987. A new method to determine Giardia cyst viability: correlation of fluorescein diacetate and propidium iodide staining with animal infectivity. Appl. Environ. Microbiol. 53:704-707. 20. Schupp, D. G., M. M. Januschka, L. A. F. Sherlock, H. H. Stibbs, E. A. Meyer, W. J. Bemrick, and S. L. Erlandsen. 1988. Production of viable Giardia cysts in vitro: determination by fluorogenic dye staining, excystation, and animal infectivity in the mouse and Mongolian gerbil. Gastroenterology 95:1-10.
3522
BOUCHER AND GILLIN
21. Stevens, D. P., and F. D. Gillin. 1990. Giardiasis, p. 344-349. In K. S. Warren and A. A. F. Mahmoud (ed.), Tropical and geographical medicine, 2nd ed. McGraw-Hill Book Co., New York. 22. Ward, H. D., J. Alroy, B. J. Lev, G. T. Keusch, and M. E. A. Pereira. 1985. Identification of chitin as a structural component of Giardia cysts. Infect. Immun. 49:629-634.
INFECT. IMMUN.
23. Ward, H. D., B. I. Lev, A. V. Kane, G. T. Keusch, and M. E. A. Pereira. 1987. Identification and characterization of Taglin, a mannose 6-phosphate binding, trypsin-activated lectin from Giardia lamblia. Biochemistry 26:8669-8675. 24. Zenian, A., and F. D. Gillin. 1985. Interactions of Giardia lamblia with human intestinal mucus: enhancement of trophozoite attachment to glass. J. Protozool. 32:664-668.
INFECTION AND IMMUNITY, Nov. 1990, p. 3516-3522
0019-9567/90/113516-07$02.00/0 Copyright C 1990, American Society for Microbiology
Excystation of In Vitro-Derived Giardia lamblia Cysts SHAYNE-EMILE M. BOUCHER AND FRANCES D. GILLIN*
Department of Pathology, H811F, University of California, San Diego Medical Center, 225 Dickinson Street, San Diego, California 92103 Received 16 April 1990/Accepted 10 August 1990
This is the first in-depth analysis of the excystation of Giardia lamblia cysts prepared in vitro. Its goals were both to achieve efficient excystation and to gain insights into this crucial but poorly understood process. To identify the critical elements of excystation, we tested the sequential low-pH induction and protease treatments which had been reported to be important for excystation of fecal cysts. The optimal pH for induction of excystation was 4.0. Emergence was greatly (-10-fold) stimulated by subsequent exposure of in vitro-derived cysts to chymotrypsin, trypsin, or human pancreatic fluid. The stimulatory activity of each was abolished by soybean trypsin inhibitor, demonstrating that the activity of pancreatic fluid was due to these proteases. Excystation of in vitro-derived cysts was -10 to 38%. Although the walls of in vitro-derived cysts were partially digested by protease treatment, trophozoites emerged only from one pole, as observed with fecal cysts. The conditions of encystation also determined the efficiency of excystation. Specifically, encystation in the presence of lactic acid, a major metabolite of colonic bacteria, stimulated excystation approximately fourfold, although it did not increase the total numbers of cysts. These experiments have shown that excystation of in vitro-derived cysts reflects that of cysts purified from human feces in that it is dependent upon conditions which simulate the passage of cysts through the human stomach (low pH) and into the small intestine (pancreatic proteases). elles visible in relief by differential interference contrast microscopy). Some of the cysts (range, -1 to 9.5% of total cysts) completed the life cycle by excysting in vitro, but most of the trophozoites emerged only partially from the cyst wall. Since >90% of the type I cysts were viable (7), as determined by the fluorogenic dye method of Schupp and Erlandsen (19), we hypothesized that the efficiency of excystation could be increased. Therefore, the goals of the present study were to achieve high-efficiency excystation of in vitro-derived G. lamblia cysts and to gain an increased understanding of this important biologic process.
Since excystation of ingested Giardia lamblia cysts is essential to the transmission of disease, it may be considered a virulence factor in human giardiasis. Nonetheless, this process is not well understood, largely because, until recently, the only source of G. lamblia cysts was the feces of infected humans or experimentally infected animals. Excystation of fecal cysts varied from 95% (17; unpublished data). Cysts prepared in vitro have the advantages of being free of fecal contaminants and of more consistent physiologic conditions for studies of excystation. During natural infection, ingested cysts pass through the stomach of the host, where they are exposed to gastric acid (2). Subsequently, cysts enter the duodenum, where the gastric chyme is rapidly neutralized by influxes of bicarbonate (3). It is important that trophozoites not emerge from cysts in the stomach because they would be killed by the acid. Excystation is presumed to occur in the upper small intestine, but the exact site is not known. Descent past the entrance of the common bile duct would expose the cysts to a variety of degradative enzymes and bile salts, which have detergent activity (3, 10). The pioneering studies of fecal cysts by Bingham et al. (1, 2) demonstrated that excystation was dependent upon "those conditions most closely approximating the organism's in vivo environment" (2). Specifically, they showed the importance of exposing cysts to a low pH, which mimics passage through the stomach. Trophozoites emerged when the acid-treated cysts were transferred to a neutral medium, reproducing entry into the duodenum (2). Excystation methods published since then have incorporated other physiologic intestinal factors, including bicarbonate, trypsin (a pancreatic protease), thiol-reducing agents (16, 20), and bile salts (11). In a previous study (7), we demonstrated the production of large numbers (>105/ml) of water-resistant cysts with type I morphology (smooth, oval, phase bright, with cyst organ*
MATERIALS AND METHODS
Materials. Unless otherwise noted, all reagents were obtained from Sigma Chemical Co. (St. Louis, Mo.). Bicarbonate was purchased from Fisher Scientific, and crude trypsin, from U.S. Biochemicals, was the gift of Judith Sauch. Cysts were isolated from the feces of a patient with symptomatic giardiasis by the method of Douglas et al. (5). Human pancreatic fluid. An adult male volunteer was intubated orally with a duodenal tube (AN 20; A. W. Anderson, Santa Monica, Calif.). The tube was allowed to pass through the stomach and into the duodenum under fluoroscopic guidance to an area between the ampulla of Vater (common bile duct) and the beginning of the jejunum. Before samples were collected, all residual pancreatic and duodenal fluids were withdrawn through the duodenal tube and discarded. Basal pancreatic secretions were collected for 15 min. Pancreatic secretions were then stimulated by intravenous injection of 75 U of secretin (Ferring Laboratories, Sufferin, N.Y.). Two sequential 15-min stimulated pancreatic fluid samples were then collected. Parasite cultivation. G. lamblia WB (ATCC 30957) was routinely cultivated in Diamond TYI-S-33 medium (4) with 10% adult bovine serum (Irvine Scientific) and bovine bile (13) but without added iron, vitamins, or antibiotics as previously described (7), with subculturing twice weekly. The same lots of Biosate (BBL) and serum were used
Corresponding author. 3516
VOL. 58, 1990
throughout these experiments, since any change which affected growth tended to decrease encystation. Encystation. (i) Pre-encystation cultures. The condition of the pre-encystation trophozoite monolayer appeared to play an important role in the efficiency of in vitro encystation (7). Cultures grown to the late log phase in growth medium were chilled, inverted 12 times, and counted in a hemacytometer chamber. Trophozoites (5,000 per ml, final concentration) were added to chilled pre-encystation medium. The preencystation medium was freshly prepared TYI-S-33 growth medium (pH 7.1) containing the antibiotics piperacillin (500 Rg/ml; Lederle Laboratories) and amikacin (125 ,ug/ml; Bristol Laboratories) but no bovine bile. Pre-encystation cultures (in 8-ml borosilicate glass screw-capped tubes [13 by 100 mm]) were grown for 3 days at 37.5°C and at 5° from horizontal. Cultures were still in the log phase, and monolayers were -50 to 80% confluent. The tubes were inverted eight times, and the unattached trophozoites and medium were discarded. The attached trophozoite monolayers were refed with fresh encystation medium (see below). (ii) Encystation cultures. Unless otherwise specified, the encystation medium consisted of pre-encystation medium adjusted to pH 7.8 with 1 M NaOH and supplemented with porcine bile (0.25 mg/ml, final concentration) and lactic acid (hemi-calcium salt, 5 mM, final concentration) (7). The medium, bile (10-mg/ml stock solution), and lactic acid (100 mM stock solution) were all freshly prepared and filter sterilized separately. Encysting trophozoites were incubated for 66 h, since initial experiments showed that this length of incubation was optimal for the biologic activity of cysts (unpublished data). Parasites were harvested and waterresistant cysts were isolated as described previously (7), except that the cultures were not chilled. Tubes were inverted (eight times), and nonadherent cells, including cysts, were decanted into 15-ml conical centrifuge tubes, pelleted, washed in 15 ml of double-distilled water, and incubated for 30 to 45 min at room temperature in 15 ml of double-distilled water. Trophozoites and cysts with incomplete walls were lysed by this treatment. Type I and type II water-resistant cysts were collected by low-speed centrifugation (5 min, 135 x g, 4 to 10°C), suspended in the volume (-200 ,ul) of water remaining in the tube, and enumerated in a hemacytometer chamber by differential interference contrast microscopy (7). As defined previously (7), type I cysts have a smooth, oval shape, are somewhat refractile, and have cyst organelles appearing in relief. Cysts not fulfilling these criteria are referred to as type II. Excystation. Unless otherwise specified, excystation was carried out on the same day that in vitro-derived cysts were harvested. Our standard excystation procedure incorporates variations of the two-step method of Rice and Schaefer for fecal cysts (16), based on the results of the experiments presented here. The variables tested are described in each experiment. A low-pH induction solution was prepared fresh by mixing 6.8 ml of Hanks balanced salt solution containing L-cysteine hydrochloride (57 mM) and reduced glutathione (32.5 mM) with 6.8 ml 0.1 M NaHCO3 and 11.3 ml of double-distilled water. The pH of the solution was adjusted to 4.0 with 0.01 N HCl. In the first, or induction step, 50 ,ul of cyst suspension (5 x 103 to 5 x 104 cysts) and 0.55 ml of the low-pH induction solution were mixed in a 1.5-ml Eppendorf tube. The tube was capped, vortexed, and incubated for 30 min in a 37.5°C water bath. The cysts were sedimented for 3 min at 135 x g and room temperature, and the supernatant was aspirated and discarded. In the second, or excystation, step, 1 ml of a prewarmed
EXCYSTATION OF G. LAMBLIA
3517
mixture containing 1 mg of chymotrypsin (from bovine pancreas, 3X crystallized, 50 U/mg of protein) in Tyrode salt solution (16), with the pH raised to 8.0 with freshly prepared 7.5% sodium bicarbonate, was added to the cyst pellet. The tube was capped, vortexed, and incubated for 30 min in a 37.5°C water bath. The cysts were sedimented for 3 min at 135 x g and room temperature, and the supernatant was removed to avoid prolonged exposure to chymotrypsin, which was detrimental to the emerging trophozoites. After the excystation step, 100 ,ul of prewarmed TYI-S-33 growth medium was added to the pellet, which was suspended with a Pipetman, to promote trophozoite motility and viability. Twenty microliters of each sample was loaded into a hemacytometer chamber, which was placed in a 37°C humidified chamber. Partially and totally excysted trophozoites and intact cysts were enumerated by differential interference contract microscopy with a 20X objective. Early experiments showed that the percent excystation did not change significantly during the time required to count all the samples (-4 h). The formula of Bingham et al. (1) was used to determine percent excystation. The original formula was as follows: percent excystation = {[(TET/2) + PET]/[(TET/2) + PET + IC]} x 100, where TET is the number of totally excysted trophozoites, PET is the number of partially excysted trophozoites, and IC is the number of intact cysts. PET are motile and active but not completely emerged from the cyst wall. This formula included a twofold factor based on the observation that TET from fecal cysts divided immediately upon emergence, producing two trophozoites (1). However, on the basis of morphologic observations, the division of strain WB trophozoites emerging from cysts prepared in vitro was delayed by several hours. Therefore, for in vitro cysts, we did not divide TET by two, unless the trophozoites had already divided, as determined microscopically. Each figure shown is from a single experiment representative of two to six repetitions with different cyst preparations. Each determination represents an average of four fields with -200 cells per field. The results of each repetition were consistent with those shown in the figures. Significance was determined by Student's t test. RESULTS Roles of acid and protease treatments in excystation. In an initial experiment, we assessed two published methods of excystation. With the method of Rice and Schaefer (16), we observed 13.5% + 0.8% excystation. Moreover, >97% of the excysted trophozoites emerged completely from the cyst wall. In contrast, with the method of Schupp et al. (20), which does not include protease treatment, we observed only 4.3% + 0.8% excystation. Moreover, most (90.1%) of the trophozoites were only partially excysted, meaning that the trophozoites did not emerge completely from the cyst wall or remained partially adherent to the cyst shell (P < 0.0005). Therefore, we investigated the two major parameters of the method of Rice and Schaefer: pH in the first, or induction, step and treatment of cysts with protease in the second, or emergence, step. The studies of Bingham et al. (1, 2) demonstrated the importance of the exposure of fecal cysts to a low pH for subsequent excystation. Therefore, we assessed the effect of exposure to an acidic pH on excystation of in vitro-prepared cysts (Fig. 1). The pH curve was quite broad, with 10 to >35% excystation after exposure to pH 2 to 8, with an optimum at pH 4.0, which was used in subsequent experi-
INFECT. IMMUN.
BOUCHER AND GILLIN
3518
N211
40-
6 ('n
40
110
110
030
]O
i6
D
o qC'J 0 11 0
Cl)
0
(1)0Q
E . 0)
0C] 0
QOL
20-
cxi11
o
_
0-
2
4
I
R 10'
2.0 5
6
7
ments. In other experiments, excystation did not differ greatly at pHs between 2 and 4 but always decreased
gradually with increasing pH above pH 4. Although the proportion of totally excysted trophozoites decreased slightly with increased pH, it remained >80%. Following exposure to gastric acid, ingested cysts pass into the small intestine, where they are exposed to pancreatic proteases at a slightly alkaline pH. This process was modeled by Rice and Schaefer (16), who treated fecal cysts with a crude preparation of trypsin. To elucidate the role of proteases in excystation, we exposed in vitro-derived cysts to crude and purified trypsin preparations and to purified chymotrypsin. The highest level of excystation (19.2% 7.7%) was observed following exposure of in vitro-derived cysts to purified chymotrypsin, as compared with 15.7% + 5.1% with crude trypsin at 5 mg/ml or 6.8% 1.6% with purified trypsin at 1 or 5 mglml (data not shown). Since chymotrypsin was effective at lower concentrations, we used it in all subsequent experiments. We next compared the effects of a low pH and proteolytic treatment on the excystation of cysts isolated (5) from the feces of a patient with symptomatic giardiasis (Fig. 2). Excystation was maximal after exposure to pH 2.0 and almost absent after exposure to pH 6.0. Moreover, chymotrypsin stimulated the excystation of fecal cysts by approximately twofold. The maximal excystation (-38%) of fecal cysts in this experiment was directly comparable to that of the in vitro-derived cysts in Fig. 1, since these experiments were performed at the same time. Since trypsin and chymotrypsin have different substrate specificities, we next tested whether other endoproteases would also stimulate excystation of in vitro-derived cysts. Four relatively nonspecific proteases, proteinase K, subtilisin, thermolysin, and elastase, stimulated excystation as efficiently (14.0% 3.3% to 19.7% 1.6%) as did chymotrypsin (16.5% 1.2%; P > 0.05), but neither the exoproteases carboxypeptidase A and leucine aminopeptidase nor pepsin (incorporated into the low-pH induction step) stimulated excystation (data not shown). Stimulation of excystation of in vitro-derived cysts by pancreatic fluid. Since trypsin and chymotrypsin are secreted into the small intestine by the pancreas, we tested the effects ±
±
4.0
6.0
pH
FIG. 1. Effect of various pHs on the excystation of in vitroderived cysts. Water-resistant cysts were incubated in cysteineglutathione-bicarbonate solution, the pH of which was adjusted with dilute HCI or NaOH. Otherwise, the procedure was as described in Materials and Methods. S.D., Standard deviation of the total percent excystation. The P values are for exposure to the indicated pH, compared with pH 4.0.
±
g 20 a
0O
pH
±
+ CT + CT, boiled
0
D Vv
10-
El
30
C\
CM 0 11 0
2 a
PET TET
FIG. 2. Roles of low pH and chymotrypsin in the excystation of fecal cysts. The percent excystation of fecal cysts can be directly compared with that of in vitro-derived cysts in Fig. 1, since these experiments were carried out at the same time. The second step of excystation was carried out in the presence of a 1-mg/ml concentration of active chymotrypsin (CT) or chymotrypsin which had been inactivated by boiling for 20 min and cooled prior to the experiment. Excystation was significantly decreased following exposure to pH 4 or 6 (P s 0.001) or to boiled chymotrypsin (P < 0.001).
of pancreatic secretions from a healthy human volunteer on excystation. Pancreatic fluids collected from the same subject before and after stimulation with secretin (3) greatly increased excystation (P s 0.002) (Fig. 3). Stimulated pancreatic secretions were more effective at lower concentrations and consistently yielded higher levels of excystation than unstimulated pancreatic secretions. Both stimulated pancreatic fluid and chymotrypsin increased excystation over a >100-fold concentration range. Concentrations of stimulated pancreatic fluid above 2% or of chymotrypsin above 1 mg/ml led to decreased excystation (data not shown). To determine whether the stimulation of excystation by pancreatic fluid was due to chymotrypsin and/or trypsin activity, we assessed the effect of soybean trypsin inhibitor, an inhibitor of both proteases (Fig. 4). Stimulation of excystation by both chymotrypsin and pancreatic fluid was virtually eliminated by soybean trypsin inhibitor (P < 0.01). Effect of cyst storage on biological activity. Since Bingham et al. (1) showed that fecal cysts were able to excyst after >30 days of storage in distilled water at 8°C, we monitored the biologic activity of two preparations of in vitro-derived cysts from the day they were harvested (day 1 in Fig. 5). Biologic activity was detected for up to 26 days of incubation at 4°C, after which excystation dropped below 0.1%. Although the initial percent excystation of the two cyst preparations differed by a factor of approximately three, it was highest on the day of harvest and declined with incubation at 40C. Effect of encystation conditions on the biologic activity of cysts. Using the optimized excystation procedure, we tested the hypothesis that the conditions of encystation determine cyst biologic activity and the idea that metabolites of colonic bacteria may be important for this process. To do this, we assessed the effects of organic acids produced by the colonic bacterial flora (3) on both the numbers of cysts and the efficiency of excystation. None of the organic acids tested increased the numbers of total cysts as compared with porcine bile alone (at pH 7.8), although lactic acid led to slightly increased numbers of type I cysts (Fig. 6A) (7). In contrast, the percent excystation was increased more than fourfold by the inclusion of 5 mM lactic acid in the encystation medium (P = 0.001) (Fig. 6B).
EXCYSTATION OF G. LAMBLIA
VOL. 58, 1990
3519
t Unstimulated
Stimulated *
a
20 *
*F
T
10 *
None
2
5
10
25
0.02 0.05
0.5
10
2.0
Pancreatic Fluid, %
50
250 1000
Chymotrypsin,
pg/ml FIG. 3. Stimulation of the excystation of in vitro-derived cysts by pancreatic fluid. Unstimulated or stimulated pancreatic fluid or chymotrypsin at the concentrations shown was substituted for the standard 1-mg/ml concentration of chymotrypsin in the second step of excystation. *, Significant increase in percent excystation (P 0.002), compared with the control with no pancreatic fluid or chymotrypsin. -
DISCUSSION Taken together, data from our laboratory (6-9, 24) and other laboratories (e.g., 1, 2, 11, 13, 16, 23) show that conditions encountered by G. lamblia during each step of its natural life cycle are crucial to reproducing that step in vitro. By varying the conditions of encystation (7) and excystation, we have consistently obtained in vitro-derived cysts of G. lamblia with levels of biologic activity comparable to those of fecal cysts. Our experiments, based on earlier work with fecal cysts (2, 16), have permitted us to identify two parameters necessary for the excystation of in vitro-derived cysts. The first, exposure of in vitro-derived cysts to a low pH, which mimics the passage of ingested cysts through the stomach, was shown by Bingham et al. (1, 2) to be crucial for the excystation of fecal cysts. The second, exposure of in vitro-derived cysts to trypsin, as reported by Rice and Schaefer (16) for fecal cysts, or to chymotrypsin, mimics the
passage of cysts into the lower duodenum, where they are bathed in pancreatic secretions containing these proteases (3). At the optimal pH for each, we observed equal excystation of fecal and in vitro-derived cysts. The optimal pH (2.0) for triggering the excystation of fecal cysts was slightly lower than that for triggering the excystation of in vitroderived cysts (4.0). A second difference was that excystation of in vitro-derived cysts was totally dependent upon exposure to proteases, whereas excystation of fecal cysts was depressed only by 50 to 60% in the absence of proteases. Differences between the excystation of in vitro-derived and fecal cysts may simply be related to differences between strains. The WB strain, which typifies the most common group of G. lamblia isolates (14), was isolated in 1979 from a patient infected in Afghanistan, whereas the fecal cysts used in this study were isolated from a patient infected in Kenya. Alternatively, the physical state of the wall of in vitro-
*
20-
*
E3
No Addition 1 mg/ml Chymotrypsin 2% Pancreatic Fluid, stim. 25% Pancreatic Fluid, unstim.
10-
*
* *
n~
I* ~
*
5
0
Soybean Trypsin Inhibitor, mg/ml FIG. 4. Inhibition by soybean trypsin inhibitor of stimulation of the excystation of in vitro-derived cysts by pancreatic fluid or chymotrypsin. Chymotrypsin or pancreatic fluid was preincubated with soybean trypsin inhibitor at the indicated concentrations for 15 min at 37°C prior to being used in the second step of excystation. *, Significant inhibition of excystation by soybean trypsin inhibitor (P < 0.001). **, Significant inhibition (P = 0.006); stim., stimulated; unstim., unstimulated.
3520
INFECT. IMMUN.
BOUCHER AND GILLIN 30
pg20
D ..- --
cyst prep. A cyst prep. B
10
0
1
7
4
10
13
16
19
22
25
28
Day FIG. 5. Effect of storage at 4°C on the excystation of in vitro-derived cysts. Water-resistant cysts were harvested on day 1 and kept in an ice water bath in double-distilled water with the antibiotics piperacillin and amikacin. At the times indicated, samples were removed and excysted by the standard procedure. A and B are two different cyst preparations (cyst prep.). Percent excystation was significantly decreased on and after day 5 (P 0.002) for preparation A and day 9 (P < 0.002) for preparation B. -
derived cysts mlay differ in subtle ways from that of the wall of fecal cysts. This difference may be due to incubation of the latter withiin the fecal mass both before and after passage. For exannple, the cyst wall may be acted upon by bacterial or hos,t enzymes or metabolites.
Exposure of cysts to an acidic pH has been a hallmark of published excystation procedures (1, 2, 11, 17, 18). The pH curve for excystation of in vitro-derived cysts was broad. The percent excystation was maximal at pH 2 to 4 but always declined steadily at pHs above 5. Even so, substantial excystation (25 to 35% of maximal) was observed following exposure to pH 7 to 8. This result may have been due 8- A to prior exposure of cysts during harvesting, incubating, and * Type cysts washing in double-distilled water at a pH of -5.7. Such * Type 11 cysts exposure may trigger low levels of excystation which can be i 6-1 subsequently increased by exposure to a lower pH. This process may explain the incidence of giardiasis in patients in_ Ito with reduced gastric acidity. In the present study, excystation of in vitro-derived cysts 0 was >90% dependent on exposure to proteases in the second 3 * -step. The stimulatory activity of human pancreatic fluid was to trypsin and chymotrypsin, since it was virtually _ _ ~~~~~~~~due __ abolished by soybean trypsin inhibitor, an inhibitor of both 1 mM5mM 1 M 5mM None 1 1 mM 1 mM 5M 5mM proteases. This observation shows that other components in + LA + AA + BA + SA or collected with the pancreatic secretions, such as bile salts * or other enzymes, are not required, although they may have 14 B a contributory role (11). The requirement for proteases was not specific, since all endoproteases tested were effective. 12 * 1 mM The idea that proteins important components of the cyst * 5mM 010 earlier observation that cyst wall IS consistent withareour by polyacrylamide gel electrophoresis separated antigens 8 and transferred to nitrocellulose were digested by trypsin 6 _ ~~~~~~~~~~~~(15). Compared with the walls of fecal cysts, the walls of in C4vitro-derived cysts appeared by light microscopy to be more 2 gtransparent after protease treatment. However, trophozoites y _ always emerged from a pole, as reported earlier for fecal _T L +M cysts (2), suggesting the presence of a protease-sensitive None + LA + SA + BA component or area at the pole of the cyst, possibly produced Organic Acid Addition or unmasked by prior exposure of the cyst to acid and/or FIG. 6. Effect s of organic acids in the encystation medium upon thiol-reducing agents. In contrast, treatment of cysts with numbers of watei r-resistant cysts and subsequent excystation. The purified chitinase did not stimulate excystation (data not
MTN
standard 5 mM lacCtic acid was omitted from the encystation medium and replaced as in dicated with succinic acid (SA), butyric acid (BA), acetic acid (AA), or lactic acid (LA). (A) Numbers of cysts (105/ml). (B) Percent excysstation of each cyst preparation determined by the standard assay. *, Excystation was significantly increased (P = 0.001) following e ncystation in the presence of 5 mM lactic acid.
shown), suggesting that chitin, which was previously re-
ported to be a cyst wall constituent by one group (22) but not by another (12), may not be located at the site of trophozoite emergence.
Excystation is the most stringent criterion of cyst biologic
EXCYSTATION OF G. LAMBLIA
VOL. 58, 1990
activity and is extremely sensitive to the conditions of encystation. For example, although encystation in the presence of lactic acid led to relatively small (-20 to 100%) (Fig. 6) (7) increases in type I cysts but not in total cysts, it stimulated excystation strongly (more than fourfold). The idea that lactic acid may have a role in the terminal stages of encystation is consistent with the fact that it is a major metabolite of bacteria in the large intestine. Moreover, the addition of lactic acid to cultures after 24 h in encystation medium did not decrease excystation (data not shown). Our earlier study (7) showed that viability, defined by the fluorogenic dye assay of Schupp and Erlandsen (19), was a less stringent criterion for cyst quality, since viable cysts were not necessarily able to excyst (20). Similarly, fecal cysts retained the ability to exclude eosin (viability) after they lost the ability to excyst (1). In the present study, excystation of different preparations of in vitro-derived cysts using the same encystation and excystation procedures varied from -8 to 38%. This value is well within the broad range (1 to 95%) reported for cysts isolated from human patients (17). Nonetheless, the apparent efficiency of excystation of in vitro-derived cysts may be low because the calculation is based on total numbers of water-resistant cysts, a heterogeneous population of which only -10 to 30% has type I morphology. While -90% of type I cysts are viable (7) and therefore are theoretically capable of excystation, the remaining type II cysts include cysts with a shrunken cytoplasm as well as cysts which appear morphologically less mature. Only -33% of type II cysts were found viable (unpublished data) by the fluorogenic dye method (19). We have not been able to determine whether type I or type II cysts or both excyst, because the excystation procedure obscures cyst morphology. However, on the basis of morphology and viability, type I cysts may be more competent at excystation. The variability in cyst morphology is not an artifact of encystation in vitro, since both type I and type II cysts are isolated from feces (7, 19). Moreover, the numbers of cysts passed by infected humans fluctuate greatly. We have observed cyst numbers from >106/g of stool to barely detectable in the same untreated patient (unpublished data; 21), suggesting that the formation and/or shedding of cysts in vivo is highly sensitive to variations in conditions which have not yet been identified. Our in vitro encystation conditions yield more consistent numbers of biologically active cysts. Our experiments have shown that a low pH and pancreatic proteases which are important for the excystation of fecal cysts also induce high-efficiency excystation of in vitroderived cysts. These observations again underscore the value of mimicking human gastrointestinal tract conditions for completing the life cycle of G. lamblia in vitro. The availability of all stages of the life cycle of G. lamblia will permit new cellular, molecular, and immunologic studies of this important pathogen. ACKNOWLEDGMENTS
We are grateful to our colleagues at the Environmental Protection Agency for suggesting the method of Rice and Schaefer, to D. Reiner for stimulating discussions, to Dan Hogan (Division of Gastroenterology, University of California, San Diego) for the pancreatic fluid, to C. Davis and J. Sauch for critiquing the manuscript, to W. Strum (Scripps Clinic) for patients, and to S. McFarlin for preparing the manuscript. This study was supported by U.S. Environmental Protection Agency cooperative agreement CR 814537 and by Public Health
3521
Service grants AM 35108, AI 24285, and Al 19863 from the National Institutes of Health.
LITERATURE CITED 1. Bingham, A. K., E. L. Jarroll, E. A. Meyer, and S. Radulescu. 1979. Giardia sp.: physical factors of excystation in vitro and excystation vs eosin exclusion as determinants of viability. Exp. Parasitol. 47:284-291. 2. Bingham, A. K., and E. A. Meyer. 1979. Giardia excystation can be induced in vitro in acidic solutions. Nature (London) 227: 301-302. 3. Davenport, H. W. 1977. Physiology of the digestive tract. Year Book Medical Publishers, Chicago. 4. Diamond, L. S., D. Harlow, and C. C. Cunnick. 1978. A new medium for the axenic cultivation of Entamoeba histolytica and other Entamoeba. Trans. R. Soc. Trop. Med. Hyg. 72:431-432. 5. Douglas, H., D. S. Reiner, and F. D. Gilln. 1987. A new method for purification of G. lamblia cysts. Trans. R. Soc. Trop. Med. Hyg. 81:315-316. 6. Gault, M. J., F. D. Gillin, and A. J. Zenian. 1987. Giardia lamblia: human intestinal mucus and epithelial cells stimulate growth in serum-free medium. Exp. Parasitol. 64:664-668. 7. Gillin, F. D., S. E. Boucher, and D. S. Reiner. 1989. Giardia lamblia: the roles of bile, lactic acid, and pH in completion of the life cycle in vitro. Exp. Parasitol. 69:164-174. 8. Gillin, F. D., D. S. Reiner, and S. E. Boucher. 1988. Small intestinal factors promote encystation of Giardia lamblia in vitro. Infect. Immun. 56:705-707. 9. Gillin, F. D., D. S. Reiner, M. J. Gault, H. Douglas, S. Das, A. Wunderlich, and J. Sauch. 1987. Encystation and expression of cyst antigens by Giardia lamblia in vitro. Science 235:10401043. 10. Hofmann, A. F., and A. Roda. 1984. Physicochemical properties of bile acids and their relationship to biological properties: an overview of the problem. J. Lipid Res. 25:1477-1489. 11. Isaac-Renton, J., E. M. Proctor, R. Prameya, and Q. Wong. 1986. A method of excystation and culture of Giardia lamblia. Trans. R. Soc. Trop. Med. Hyg. 80:989. 12. Jarroll, E. L., P. Manning, D. G. Lindmark, J. R. Coggins, and S. L. Erlandsen. 1989. Giardia cyst wall-specific carbohydrate: evidence for the presence of galactosamine. Mol. Biochem. Parasitol. 32:121-132. 13. Keister, D. B. 1983. Axenic culture of Giardia lamblia in TYI-S-33 medium supplemented with bile. Trans. R. Soc. Trop. Med. Hyg. 77:487-488. 14. Nash, T. E., T. McCutchan, D. Keister, J. B. Dame, J. D. Conrad, and F. D. Gillin. 1985. Restriction-endonuclease analysis of DNA from 15 Giardia isolates obtained from humans and animals. J. Infect. Dis. 152:64-73. 15. Reiner, D. S., D. Douglas, and F. D. Gillin. 1989. Identification and localization of cyst-specific antigens of Giardia lamblia. Infect. Immun. 57:963-968. 16. Rice, E. W., and F. W. Schaefer III. 1981. Improved in vitro excystation procedure for Giardia lamblia cysts. J. Clin. Microbiol. 14:709-710. 17. Sauch, J. F. 1988. A new method for excystation of Giardia, p. 261-264. In P. M. Wallis and B. R. Hammond (ed.), Advances in Giardia research. University of Calgary Press, Calgary, Alberta, Canada. 18. Schaefer, F. W., III. 1988. A review of methods that are used to determine Giardia cyst viability, p. 249-254. In P. M. Wallis and B. R. Hammond (ed.), Advances in Giardia research. University of Calgary Press, Calgary, Alberta, Canada. 19. Schupp, D. G., and S. L. Erlandsen. 1987. A new method to determine Giardia cyst viability: correlation of fluorescein diacetate and propidium iodide staining with animal infectivity. Appl. Environ. Microbiol. 53:704-707. 20. Schupp, D. G., M. M. Januschka, L. A. F. Sherlock, H. H. Stibbs, E. A. Meyer, W. J. Bemrick, and S. L. Erlandsen. 1988. Production of viable Giardia cysts in vitro: determination by fluorogenic dye staining, excystation, and animal infectivity in the mouse and Mongolian gerbil. Gastroenterology 95:1-10.
3522
BOUCHER AND GILLIN
21. Stevens, D. P., and F. D. Gillin. 1990. Giardiasis, p. 344-349. In K. S. Warren and A. A. F. Mahmoud (ed.), Tropical and geographical medicine, 2nd ed. McGraw-Hill Book Co., New York. 22. Ward, H. D., J. Alroy, B. J. Lev, G. T. Keusch, and M. E. A. Pereira. 1985. Identification of chitin as a structural component of Giardia cysts. Infect. Immun. 49:629-634.
INFECT. IMMUN.
23. Ward, H. D., B. I. Lev, A. V. Kane, G. T. Keusch, and M. E. A. Pereira. 1987. Identification and characterization of Taglin, a mannose 6-phosphate binding, trypsin-activated lectin from Giardia lamblia. Biochemistry 26:8669-8675. 24. Zenian, A., and F. D. Gillin. 1985. Interactions of Giardia lamblia with human intestinal mucus: enhancement of trophozoite attachment to glass. J. Protozool. 32:664-668.
