Effects of environmental conditions on Cryptosporidium oocyst viability: a pilot study

Heidi L. Enemark¹, Cynthia D. Juel¹, Simone Caccìo²

¹Danish Veterinary Institute, Bülowsvej 27, DK-1790 Copenhagen V, Denmark.

²Istituto Superiore di Sanità, Viale Regina Elena 299, 00161 Rome, Italy.

Introduction

Cryptosporidium parvum is a ubiquitous protozoan parasite capable of causing acute self-limiting enteritis or chronic and potentially life-threatening diarrhoea in man as well as animals (Casemore et al., 1997). Infections with Cryptosporidium are initiated by ingestion or inhalation of the oocyst stage, which is unusually resistant to natural stresses and chemical disinfectants (Fair et al., 1997). The infective dose is low, and has been shown to vary from 9 to 1042 oocysts depending on the isolate (Okhuysen et al., 1999). Within the past decades numerous studies have focused on viability and survival of Cryptosporidium oocysts particularly in water, and the resistance of the oocysts to commonly utilized disinfection techniques. For most chemicals, including chlorine in any concentration that can be used to treat drinking water, effective concentrations are generally not practical for disinfection outside the laboratory, and high concentrations that greatly reduce oocyst infectivity are either very expensive or quite toxic (Fayer et al., 1997).

Laboratory studies have attempted to determine the effects of different environmental conditions, and to elucidate survival limits of oocysts to exposure from various physical factors, some of which are listed in table 1. However, although an increasing number of food-associated outbreaks have been documented recently, there remains a paucity of detailed studies of oocyst survival in food.

Food-associated outbreaks of cryptosporidiosis have implicated raw fruits and vegetables (Sterling et al., 1986; Monge & Chinchilla, 1996; Robertson et al., 2002), raw milk, offal, sausage (Casemore et al., 1986; Gelletli et al., 1997), apple cider (Millard et al., 1994), as well as different foods contaminated by food handlers (Besser-Wiek et al., 1996;Quiroz et al., 2000). Although oocysts of Cryptosporidium have been detected in a number of different foods, direct incrimination of food in the transmission of cryptosporidiosis is hampered by the limited numbers of oocysts in suspected food samples, the lack of an enrichment culture for oocyst recovery, and the lack of sensitive detection methods, leading to an underestimation of the incidence (Laberge et al., 1996; Fayer et al., 2000). This absence of adequate detection techniques increases the need for knowledge about oocyst survival in different food products so that the duration and probability of potential threats of infection can be realistically assessed. The present study was undertaken to evaluate the effect of pH, temperature, desiccation, and storage on oocyst survival in milk, apple juice, and water.

Table 1. Resistance of Cryptosporidium parvum oocysts to physical stress and environmental conditions (selected studies).
Stress factor	Conditions	Results	Test	Ref.
Water:
Storage	Room temp., 176 d^a	96% reduced	Ex/dye	Roberteson et al., 1992
Heat^a	Submerged in river
	at ambient temp., 176 d^b	94% reduced	Ex/dye
	4^oC, 35 d^c	38% reduced	Ex/dye
	59.7^oC, 5 min	I	In vivo	Fayer, 1994
	64.2^oC, 5 min	NI	In vivo
	67.5^oC, 1 min	I	In vivo
	72.4^oC, 1 min	NI	In vivo
	55^oC, 30 s	NI	In vivo	Fujino et al., 2002
	60^oC, 15 s	NI	In vivo
	70^oC, 5 s	NI	In vivo
Freezing	-196^oC, 10 min	NI	In vivo	Sherwood et al., 1982
	-20^oC, 3 d	NI	In vivo
	-70^oC, 1 h	NI	In vivo	Fayer & Nerad, 1996
	-20^oC, 8 h; 1 d	I; NI	In vivo
	-15^oC, 24 h; 1 w	I; NI	In vivo
	-10^oC, 1 w	I	In vivo
	Liquid nitrogen	100% reduced	Ex/dyes	Robertson et al., 1992
	-22^oC, ≤32 d	98% reduced	Ex/dyes
	-20^oC, 24 h	92% reduced	Dye	Deng & Cliver, 1999
Drying	Air dried, 2 h	97% reduced	Ex/dyes	Robertson et al., 1992
	Air dried, 4 h	100% reduced	Ex/dyes
Salinity	10-30 ppt., 10^oC, 1-12 w	I	In vivo	Fayer et al., 1998
	10 ppt., 20^oC, 1-12 w	I	In vivo
	20-30 ppt., 20^oC, 12 w	NI	In vivo
Dairy products:
Yogurt production	37^oC, 48 h	40% reduced	Dye	Deng & Cliver, 1999
	37^oC, 48 h + 4^oC, 8 d	42% reduced	Dye
Ice-cream	Mixing and freezing	80% reduced	Dye	Deng & Cliver, 1999
production	Mixing and freezing		Dye
	+ hardening at -20^oC,	100% reduced	Dye
	24 h 72.4^oC, 1 min	NI	In vivo	Fayer, 1994
Heat	71.7^oC, 5 sec,	NI	In vivo	Harp et al., 1996
Faeces:
Storage	Ambient temp., 176 d^d	66% reduced	Ex/dyes	Robertson et al., 1992
	4^oC, 178 d^e	78% reduced	Ex/dyes
	4^oC, 410 d^d	90% reduced	Dye	Jenkins et al., 1997
Freezing	-20^oC, 2, 7 & 30 d	66-88%	In	Kim & Healey, 2001
		reduced	vivo/ex/
Drying	Air dried, 1-4 d		Tis. cult.	Anderson, 1986
		NI	Invivo

I = infectious; NI = non-infectious; In vivo testing in mice; Ex = excystation; tis. cult = In vitro testing in tissue culture. ^a Tap water; ^b River water; ^c Sea water; ^d Cow faeces; ^e Human faeces.

Materials and methods

Source and purification of oocysts

C. parvum oocysts used in this study were obtained from Istituto Superiore di Sanità (ISS), Rome, Italy. The strain (Rome isolate), originally isolated from a Danish calf, has been propagated in calves by ISS for several years. Sheather's sugar flotation, discontinuous percoll gradient centrifugation, and repeated washing and centrifugation as described by Peeters & Villacorta (1995) purified the oocysts. The oocysts were suspended in phosphate-buffered saline (PBS, pH 7.2) containing 100 U of penicillin and 100 µg of streptomycin per ml, and stored at 5^oC until use within two months.

Assessment of viability

Oocyst viability was determined using a modification of the method described by Campbell et al. (1992), which depends upon the morphology and inclusion or exclusion of the two fluorogenic vital dyes 4',6-diamidino-2-phenylindole (DAPI) and propidium iodide (PI) by oocysts. Oocyst suspensions (0.1 g) were washed in 1.5 ml water, centrifuged at 3500 x g (10 min. at 5^oC) and the pelleted oocysts preincubated in acidified (pH 2.75) Hank's balanced salt solution (HBSS; Sigma). Following two further washings in HBSS (pH 6.6), 100 µl of oocyst suspension was incubated for 2h with 10 µl DAPI (2 mg/ml in absolute methanol) and PI (1 mg/ml in 0.1 M PBS, pH 7.2) at 37^oC. Subsequently, the samples were washed twice in water, suctioned to 100 µl, transferred to Teflon printed diagnostic slides and dried overnight in the dark.

To facilitate oocyst identification, a direct immunofluorscence assay (IFA) was included: the dried samples were fixed with methanol (50 µl per well), and incubated with 40 µl FITC conjugated anti-Crypto monoclonal antibody (Microgen Bioproducts Ltd., UK) for 1 h at room temperature. Following a wash with 100 µl PBS, 5 µl mounting fluid was added to each well, and the slides sealed with nail polish.

The slides were examined with a standard fluorescence microscope (Leitz, Germany) equipped with an UV filter block (350-nm excitation, 450-nm emission), and a PI filter block (500-nm excitation, 630-nm emission). Oocysts were categorized as dead if they were PI-positive (PI+) or ruptured without internal contents (ghosts), and viable if they were PI-negative (PI-), and either DAPI-positive (DAPI+) or DAPI-negative (DAPI-). Viability was calculated as the percentage of PI- oocysts in a total of 200 counted oocysts.

Matrices investigated

Three different beverages were selected for monitoring of oocyst survival with water serving as control.

(I) Apple juice (pasteurised, pH 3.4; Dansk Kernefrugt, Valby, DK).
(II) Skim milk (pasteurised, homogenised, initial pH 6.6, fat 0.1%; Arla Foods amba, Viby, DK).
Fresh milk directly from the cow (un-pasteurised, un-homogenised, initial pH 6.6, fat approximately 4%).
Water (double distillated, pH 6.5).

Environmental conditions investigated

For each of the four fluids, 6.3 x 10⁴oocysts were added to 10 ml tubes containing 6.3 ml of juice, milk and water, respectively, and the effect of pH, temperature and desiccation were investigated.

PH: pH was adjusted to 5, 7 & 9 by addition of 0.1 M sodium hydroxide (NaOH) or hydrochloride acid (HCl), and the solutions stored at 5^oC in the dark.
Temperature: Oocyst solutions were stored in the dark at 5^oC, room temperature (approximately 21^oC) or 37^oC. From one oocyst solution 100 µl aliquots were pipetted into cryotubes and frozen at -18^oC. Before assessment of oocyst viability, the samples were allowed to thaw at room temperature.
Desiccation: Aliquots (100 µl) of the different vehicles (I-IV) were pipetted into Eppendorf tubes, and left to dry at room temperature in the dark. The remaining volume was noted at daily intervals. Oocyst viability of these samples were assessed daily during the first week and then subsequently one week later for those samples still containing live oocysts. Assessment of viability was terminated when the viability was ≤2%. For all other samples oocyst viabilities were assessed at approximately weekly intervals.

Preliminary results

With the exception of oocysts exposed to desiccation, their viability diminished rapidly at first, then more gradually. Although the results have not yet been analysed statistically it was obvious that desiccation, storage at 37^oC and freezing dramatically affected oocyst survival. Storage at room temperature also reduced the oocyst survival time compared to storage at 5^oC, whereas pH variation between 5 and 9 had no significant influence on the survival time (Fig. 1).

pH: Following 36 days of storage, oocyst viability was reduced with 41.2% ± 8.2, 49.2% ± 5.7, and 49.3% ± 5.9 at pH 5, 7, and 9 respectively, i.e. approximately half of those oocysts, which were alive from the initiation of the study (52.5%) had died irrespective of matrix. After 153 days, oocyst survival had decreased with 82.8% ± 5.9, 84.6% ± 1.6, and 79.0% ± 4.1 (pH 5, 7, and 9) corresponding to an oocyst survival of approximately 10% (Fig. 2). There were no obvious differences concerning oocyst survival between water, juice, and milk. Nevertheless, compared to water a tendency towards higher oocyst survival in milk, and lower survival time in juice was seen.

Temperature: At 5^oC oocyst viability was roughly halved following 40 days of storage in water as well as in milk and juice (Fig. 2). A mean of 6.6% ± 3.8 of the oocysts were viable after 153 days corresponding to 12.6% of the initially live oocysts. Viability in raw milk was not assessed later than 55 days post inoculation because reading was severely impeded by excessive bacterial growth as well as decreased efficacy of the immunofluorescence assay. Oocyst viability was roughly halved after one week at 21^oC. Following 49 days, all oocysts in raw as well as in skim milk were dead, whereas 5.8% and 7.8% (corresponding to 11.0% and 14.9% of the initially live oocysts) were able to survive 153 days in water and juice respectively. Although a small number of oocysts survived for a month in juice and raw milk, storage at 37^oC for one week killed >95% of the oocysts irrespective of the vehicle. Approximately 10% of the oocysts suspended in water were able to withstand freezing at -18^oC for 7 days, and an even larger fraction of oocysts (mean 16.8 % ± 3.4) survived in milk and juice at this temperature. Four weeks at -18^oC resulted in 100% death for oocysts in water and juice, while a small proportion (< 2%) was still viable in the milk even after 153 days.

Desiccation. In water and juice the majority (>98%) of the oocysts died within 4 days, whereas a small fraction (< 5%) survived for one week or more in milk (Fig.2, table 2).

Table 2. Effect of desiccation on survival of *Cryptosporidium parvum* oocysts in water, apple juice and milk.
Time (d)	Water		Apple juice		Raw milk		Skim milk
	% Viable oocysts	Vol. (µl)	% Viable oocysts	Vol. (µl)	% Viable oocysts	Vol (µl)	% Viable oocysts	Vol (µl)
1	46.5	75	42.5	75	20.5	90	48.0	70
2	46.5	0	20.5	25	18.2	fat plug	9.5	35
3	5.5	0	2.0	20	-	-	-	-
4	0.5	0	-	-	-	-	-	-
5	-	-	-	-	7.7	0	7.5	0
6	-	-	-	-	7.0	0	7.0	0
7	-	-	-	-	4.3	0	3.0	0
14	-	-	-	-	1.5	0	0	0

- = Not done

Figure 1: Effect of pH, temperature and desiccation on survival of Cryptosporidium oocysts in water

Discussion

The pilot study presented here has primarily been used to gain experiences with techniques to be used in future studies of oocyst survival in a herd environment i.e. in water, milk, urine and faeces. Initially, it was also planned to compare oocyst survival of C. parvum genotype I (syn. C. hominis) and II. Unfortunately these plans were changed, as it was not possible to obtain viable human oocysts in adequate numbers for such studies.

Generally, the viability of C. parvum can be estimated by in vitro excystation, vital staining or infectivity (Black et al., 1996). However, Jenkins et al. (1997) implied that vital staining is only a measurement of oocyst wall permeability, which appears to be correlated to the ability of oocysts to excyst and also to infect animals. Other experiments have shown that excystation as well as vital staining tend to overestimate oocyst viability (Black et al., 1996; Bukhari et al., 2000). Ongoing studies at our laboratory have therefore incorporated a combination of vital staining and in vivo infectivity studies using the infant mouse model (Finch et al., 1993) for a more precise evaluation of C. parvum infectivity.

Some serious limitations of the current study deserve mention. First, the viability of the oocysts at inoculation was relatively low, 52.5% compared to 80-90% in most other studies (Whitmore & Robertson, 1995; Merry et al., 1997; Ding & Clever, 1999). Secondly, the oocyst recovery rate was not determined, thus it was not possible to evaluate whether the viability rate might have been biased by large oocyst decay. These problems have been overcome in ongoing studies in which fresh bovine derived oocysts propagated in a calf at DVI are used.

Although preliminary, experiences from the present study have shown that:

pH variation within the limits 5 to 9 apparently does not affect oocyst survival
>95% of the oocysts were killed at 37^oC for one week
storage at room temperature resulted in a quicker die off compared to storage at 5^oC
a small fraction of the oocysts suspended in water and juice survived at room temperature for as long as 153 days, whereas milk apparently provided some protection against freezing, allowing approximately 2% of the oocysts to survive at -18^oC for 153 days
although drying resulted in the fastest death, slow drying in a minute volume of fluid appeared to extend the oocyst survival time

In correlation to water treatment processes, Robertson et al. (1992) suggested that both high (10.5) and low (1.5) pH have significant impact on oocyst viability. In contrast, studies by Jenkins et al. (1998); Höglund & Stenström (1999) suggested that pH alone does not have an effect on oocyst viability. Jenkins et al. (1998) demonstrated that even low concentrations of ammonia (0.007 M) significantly decreased the viablilty of oocysts after 24 hours of exposure, whereas exposure to pH levels between 7 to 11, corresponding to those associated with the ammonia concentrations, showed minimal effects of alkaline pH alone on oocyst viability. In accordance with these latter results, an absence of any pH effect was seen in the present study. This may indicate that C. parvum oocysts are able to survive various pH values in food products, however further studies determining the exact survival limits are needed.

It is well known that freezing at temperatures ≤70^oC is lethal to oocysts (Robertson et al., 1992; Fayer & Nerad, 1996), and that oocysts can withstand temperatures at or above -20^oC for extended periods (Sherwood et al., 1982; Robertson et al., 1992; Fayer & Nerad, 1996). In accordance with these observations, we found that a number of oocysts were able to survive for up to 3 weeks in all the matrices investigated, whereas milk appeared to provide a cryoprotective environment enabling even longer survival. Likewise, cryoprotective qualities have been demonstrated in faeces. A study of oocysts preserved in faeces showed that 12-34% of the oocysts retained their infectivity for mice when stored at -20^oC for 2-30 days (Kim & Healey, 2001). Nevertheless, oocysts inoculated into ice-cream mix were not unable to survive the production of ice-cream in a study by Deng & Cliver (1999).

In the present study, the effect of drying was studied in tubes containing 100 µl of fluid. The samples were allowed to dry slowly at room temperature contrary to other studies in which desiccation were examined by spreading oocyst solutions on to glass slides or other surfaces (Robertson et al., 1992; Deng & Cliver, 1999). The slower rate of drying allowing the oocysts to adapt more gradually to the environmental change, may explain why we found that small fractions of the oocysts were able to survive up to 10-12 days in milk after the samples had dried out completely.

Although a significant proportion of the oocysts were killed in all environments over the 6-months period of investigation, small fractions of the oocysts were resistant against the various selective pressures. Like all other C. parvum isolates, the isolate population analysed in the present study is not clonal, and therefore likely to include multiple subpopulations, the relative abundance of which may change in response to the host or environmental conditions as was shown in a study by Rochelle et al. (2000) who revealed extensive intra-isolate heterogeneity. A subject for further studies is therefore whether those oocysts, which are able to survive for extended periods in a harsh environment, are genetically different from those oocysts that die more rapidly. However, such studies require a technique capable of separating live oocysts from dead oocysts, which, to our knowledge, is not yet available.

The ability of C. parvum to tolerate pH fluctuations and survive for prolonged periods in different environments may be of importance to food safety, and demonstrate the need for further studies of oocyst survival in various matrices.

References

Anderson, B. (1986). Effect of drying on the infectivity of cryptosporidia-laden calf feces for 3- to 7-day-old mice. Am. J. Vet. Res. 47, 2272-2273.

Besser-Wiek, J.K., Forfang, J., Hedberg, C.W. et al. (1996). Foodborne outbreak of diarrheal illness associated with Cryptosporidium parvum - Minesota 1995. Morbid. Mortal. Wkly. Rep. 45, 783

Black, E.K., Finch, G.R., Taghi-Kilani, R., Belosevic, M. (1996). Comparison of assays for Cryptosporidium parvum oocysts viability after chemical disinfection. FEMS Microbiol. Lett. 135, 187-189.

Bukhari, Z., Marshall, M.M., Korich, D.G. et al. (2000). Comparison of Cryptosporidium parvum viability and infectivity assays following ozone treatment of oocysts. Appl. Environ. Microbiol. 66, 2972-2980.

Campbell, A.T., Robertson, L.J., Smith, H.V. (1992). Viability of Cryptosporidium parvum oocysts: correlation of in vitro excystation with inclusion or exclusion of fluorogenic vital dyes. Appl Environ Microbiol. 58: 11, 3488-3493

Casemore, D.P., Jessop, E.C., Douce, D., Jackson, F.B. (1986). Cryptosporidium plus Campylobacter: an outbreak in a semi-rural population. J. Hyg. Camb. 96, 95-106.

Casemore, D.P., Wright, S.E., Coop, R.L. (1997). Cryptosporidiosis - human and animal epidemiology. In: Cryptosporidium and cryptosporidiosis (ed. Fayer, R.), CRC Press, Boca Raton, FL, pp. 65-92.

Deng, M.Q., Cliver, D.O. (1999). Cryptosporidium parvum studies with dairy products.

Int. J.Food Microbiol. 46, 113-121.

Fayer, R. (1994). Effect of high temperature on infectivity of Cryptosporidium parvum oocysts in water. App. Environ. Microbiol. 60, 2732-2735.

Fayer, R., Graczyk, T.K., Lewis, E.J. et al. (1998). Survival of infectious Cryptosporidium parvum oocysts in seawater and eastern oysters (Crassostrea virginica) in the Chesapeake Bay. Appl Environ Microbiol. 64: 3, 1070-1074.

Fayer R., Morgan, U., Upton, S.J. (2000). Epidemiology of Cryptosporidium: transmission, detection and identification. Int. J. Parasitol. 30, 1305-1322.

Fayer, R., Nerad, T. (1996). Effects of low temperatures on viability of Cryptosporidium parvum oocysts. Appl. Environ. Microbiol. 62, 1431-1433.

Fayer, R., Speer, C.A., Dubey, J.P. (1997). The general biology of Cryptosporidium. In: Cryptosporidium and cryptosporidiosis (ed. Fayer, R.), CRC Press, Boca Raton, FL, pp. 1-41.

Finch, G.R., Daniels, C.W., Black, E.K. et al. (1993). Dose response of Cryptosporidium parvum in outbred neonatal CD-1 mice. Appl. Environ. Microbiol. 59, 3661-3665.

Fujino, T., Matsui, T., Kobayashi, F. et al. (2002). The effect of heating against Cryptosporidium oocysts. J. Vet. Med. Sci. 64, 199-200.

Gelletli, R., Stuart, J., Soltano, N. et al. (1997). Cryptosporidiosis associated with school milk. The Lancet 350, 1005-1006.

Harp, J.A., Fayer, R., Pesch, B.A., Jackson, G.J. (1996). Effect of pasteurisation on infectivity of Cryptosporidium parvum oocysts in water and milk. Appl. Environ. Microbiol. 62, 2866-2868.

Höglund, C.E., Stenström, T.A.B. (1999). Survival of Cryptosporidium parvum oocysts in source separated human urine. Can. J. Microbiol. 45, 740-746.

Jenkins, M.B., Anguish, L.J., Bowman, D.D. et al., (1997). Assessment of dye permeability assay for determination of inactivation rates of Cryptosporidium parvum oocysts. Appl. Environ. Microbiol. 63, 3844-3850.

Jenkins, M.B., Dwight, D.B., Ghiorse, W.C. (1998). Inactivation of Cryptosporidium parvum oocysts by ammonia. Appl. Environ. Microbiol. 64, 784-788.

Kim, H.C., Healey, M.C. (2001). Infectivity of Cryptosporidium parvum following cryopreservation. J. Parasitol. 87, 1191-1194.

Laberge, I., Griffiths, M.W., Griffiths, M.W. (1996). Prevalence, detection and control of Cryptosporidium parvum in food. Int. J. Food Microbiol. 32, 1-26.

Merry, R.J., Mawdsley, L.J., Brooks, A.E., Davies, D.R. (1997). Viability of Cryptosporidium parvum during ensilage of perennial ryegrass. J. Appl. Microbiol., 82, 115-120.

Millard, P.S., Gensheimer; K.F., Addiss, D.G. et al. (1994). An outbreak of cryptosporidiosis from fresh-pressed apple cider. J. Am. Vet. Med. Assoc. 272, 1592-1596.

Monge, B.P., Chinchilla, M. (1996). Presence of Cryptosporidium oocysts in fresh vegetables. J. Food Microbiol. 59, 202-203.

Okhuysen, P.C., Chappell, C.L., Crabb, J.H. et al. (1999). Virulence of three distinct Cryptosporidium parvum isolates for healthy adults. J Infect Dis. 180: 4, 1275-1281

Peeters, J.E., Villacorta, I. (1995). Cryptosporidium. In: Biotechnology - Guidelines on techniques in coccidiosis research (ed. Eckert, J., Braun, R., Shirley, M.W., Coudert, P.), Office for official Publications of the European Communities, Luxemburg, pp. 202-240.

Quiroz E.S., Bern C., MacArthur, J.R., et al. (2000). An outbreak of cryptosporidiosis linked to a foodhandler. J. Infect. Dis. 181, 695-700.

Robertson, L.J., Campbell, A.T., Smith, H.V. (1992). Survival of Cryptosporidium parvum oocysts under various environmental pressures. Appl. Environ. Microbiol. 58, 3494-3500.

Robertson, L.J., Johannessen, G.S., Gjerde, B.K., Loncarevic, S. (2002). Microbiological analysis of seed sprouts in Norway. Int. J. Food Microbiol. 75, 119-126.

Rochelle, P.A., De Leon, R., Atwill, E.R. (2000). Intra-isolate heterogeneity and reproducibility of PCR-based genotyping of Cryptosporidium parvum using the ß-tubulin gene. Quant. Microbiol. 2, 87-101.

Sherwood, D., Angus, K.W., Snodgrass, D.R., Tzipori, S. (1982). Experimental cryptosporidiosis in laboratory mice. Infect. Immun. 38, 471-475.

Sterling C.R., Seegar, K., Sinclair, N.A. (1986). Cryptosporidium as a causative agent of travellers diarrhoea. J. Infect. Dis. 153, 380-381.

Whitmore, T.N., Robertson, L.J. (1995). The effect of sewage sludge treatment processes on oocysts of Cryptosporidium parvum. J. Appl. Bacteriol. 78, 34-38.

[Contents] [Previous] [Next]