Experimental and Applied Acarology 25: 947–955, 2001. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
Presence of chitinase in adult Varroa destructor, an ectoparasitic mite of Apis mellifera MARC COLIN 1,*, MARC TCHAMITCHIAN 2, JEAN-MARC BONMATIN 3 and SYLVIE DI PASQUALE 1 1 UMR INRA-UAPV “Ecologie des Invertébrés”, Site Agroparc, Domaine Saint-Paul, Avignon Cedex 9, 84914, France; 2Station de Bioclimatologie, Site Agroparc, Domaine Saint-Paul, Avignon Cedex 9, 84914, France; 3CNRS, Centre de Biophysique Moléculaire, Orléans Cedex 2, 45071, France; *Author for correspondence (e-mail:
[email protected]; fax: 00 33 (0)4 32 72 26 02)
Received 30 July 2001; accepted in revised form 28 January 2002
Key words: ApiZym, Chitinase, Honey bee, Parasitic action, Varroa destructor Abstract. The enzyme spectrum of an ectoparasitic mite of the honeybee, Varroa destructor (Anderson and Trueman) was studied using a semi-quantitative method, especially designed for complex samples which have not been purified. Exopeptidases and phosphatases are shown present. A chitinase and enzymes able to transform  carbohydrates are also present with a large range in the intensity of the reaction. The role of the chitinase can be related to the supply of nutritional needs or/and the piercing and sucking behaviour of the adult parasite. Chitinase activity could be one factor influencing the balance between the parasite and its host.
Introduction The ectoparasitic mite Varroa jacobsoni was first described by Oudemans as an obligate parasite of Apis cerana, a social bee confined in Asia. The genus Varroa first became associated with Apis mellifera L. in the 1950’s, but the species was improperly confused with V. jacobsoni. According to Anderson and Trueman (2000), the genus Varroa present on A. cerana, enclosed not one single species but a complex of at least two species, the second one being V. destructor, which had switched to A. mellifera. Before any efficient control methods appeared, V. destructor spread as an epizootic out of Asia, where lives its primary host Apis cerana. Although the efficiency of treatments and their tolerability for bees have recently been improved, it is still considered as a major pest of the honeybee A. mellifera. The seriousness of the disease is owing to the fact that the mite parasitizes both adult bees and brood. Mite reproduction is in close synchronisation with the life cycle of the honeybee. Adult mated females settle on worker and drone adult bees for at least 48 h. Afterwards, they enter the worker or drone brood cell, respectively 6 or 24 h before capping. Egg laying of the mite begins at the bee prepupal stage. Hatching, nymphal stages and mating of young adults occur before emer-
948 gence of the bee imago. The best reproductive success of the mite is an average of 2.2 fertile daughters for one mother mite reproducing in the drone brood. The relationships between host and mite have mainly been studied from a behavioural perspective, and behavioural resistance to the mite has been shown. Several resistance mechanisms have been evidenced: – cleaning behaviour of the bee (Fries et al. 1996), – uncapping and removal behavior of the infested cells (Correa-Marques and de Jong 1998; Guerra et al. 2000; Vandame et al. 2000), – attractiveness of the bee larvae, – unfertility of the female mite (Rosenkranz and Engels 1994; Medina Medina and Martin 1999). Little is known at the molecular level concerning the host-parasite relationships. The pathogenicity of some agents, for example fungi, appears to be correlated with enzymes released by the parasite promoting its growth and reproduction. In the same way, to ensure its living on adult bee, its development or its reproduction on brood, the female V. destructor can release enzymes when sucking the bee haemolymph. The nutritional needs of V. destructor during reproduction has been the subject of controversy. Based on immunochemical techniques, Tewarson and Engels (1982) asserted that undigested host proteins were present both in the parasite haemolymph and in the egg vitellus. Subsequently Tewarson and Jany (1982) noticed a low proteolytic activity in the mite. These findings were not consistent with the results of the immunochemical study of Dandeu et al. (1991) who reported on the absence of intact host proteins in the parasite. Accompanying the presence of the parasite on its host, four new host proteins were revealed using similar techniques. One additional argument in favour of the degradation of the host proteins is given by Schatton-Gadelmeyer and Engels (1988) who noticed a drastic decrease in the arylphorin titer in the parasitised bee pupae. Bee arylphorin is an hexamer with 74 kDa subunits covalently bound to carbohydrate; these proteins are source of amino acids for the production of adult cuticle (Ryan et al. 1984; Telfer and Kunkel 1991; Danty et al. 1998). It is possible that the parasite secretes enzymes that interfere with host cuticle synthesis and the ability of the cuticle to take up nitrogenous compounds (Mira 2000). Chitinolytic and proteolytic enzymes could also be produced by the parasite, which might help to pierce the bee cuticle or to maintain an opening at the parasite feeding site (Donzé and Guérin 1994). Chitinase also could have a detrimental effect on the peritrophic membrane of the bee if chitin is an abundant constituant of this structure (Tellam et al. 1999). Chitinases have been found in the venom of insects, and in many parasites such as nematodes, fungi and bacteria (Kramer and Muthukrishnan 1997). Chitinases (E.C. 3.2.1.14) are enzymes with a specific hydrolytic activity directed towards chitin, a linear polysaccharide of  -(1,4) -linked -2-acetamido-2-deoxy-D-glucose (N-acetylglucosamine). The aim of this work is to better understand the V. destructor-A. mellifera relationships at the molecular level. Specifically, we have addressed chitinolytic and proteolytic activities, associated with V. destructor.
949 Material and methods V. destructor mites Adult female mites were obtained from 20 colonies belonging to 10 different bee yards of several regions of France. From one bee colony, we constituted a pool of 30 adult females taken from capped drone brood if present, from worker brood if not. The precise feeding status of the mites cannot be known because they feed 1 to 0.1 times per h, depending on oogenesis. Mites were washed by dipping them into distilled water for two h. Afterwards, 20 live mites were crushed in a Potter grinder filled with 2 ml distilled water to constitute the ground mite sample. API ZYM method API ZYM, purchased from Biomérieux SA Lyon France, is a semi-quantitative micromethod used to examine enzymatic activities. The method has been developed to allow enzymatic determinations from non purified extracts. It provides a spectrum of 19 enzymatic determinations, which can be further improved by spectrophotometric and/or electrophoretic techniques. One microtube contains one enzymatic substrate and suitable buffer. This method is currently used for biotyping micro-organisms such as bacteria, fungi and protozoans (Bianchi-Salvadori et al. 1995; Jones et al. 2000). Each microtube was filled with 65 microlitre of the ground mite sample, equivalent to 0.65 Varroa. Afterwards, the strip was incubated for four h at 34 °C, which is bee incubation temperature. Just after incubation, two reagents were added to each microtube and the intensity of the resulting colour in the microtube was visually estimated against a colour chart. Reference enzymes In addition to the chitinase, two purified enzymes of pathological importance are tested by the API ZYM method: lyzozyme (EC 3.2.1.17) and collagenase (EC 3.4.24.3). The enzymes were purchased from Sigma. Statistical analysis The results of the API ZYM method are integer colour grades ranging from 0 to 5. The 19 variables (colour grades for each enzyme) were first plotted using the face method (Chernoff 1973) to discover any similarities between the enzymatic responses across the mite samples (not shown). This method represents each multivariate observation as a human face, associating each variable of an observation with a feature of the face (area, shape, angle of eye brow, length of the nose, etc. . .). It allows a graphical representation of high dimension data (up to 15). Grouping observations is then equivalent to grouping the faces by types of expression. To assert the existence of groups of enzymatic responses which appeared with the faces
950 Table 1. Enzymes revealed in Varroa destructor by the ApiZym method Reference enzymes collagenase lysozyme chitinase
Varroa destructor Nb posit.x Min-Max 100/Tot
colour*
sample nb Alkaline phosphatase Esterase lipase (C8) Esterase (C4) Lipase (C14) Leucine arylamidase Valine arylamidase Cystine arylamidase Trypsin ␣ Chymotrypsin Acid phosphatase Naphtol-AS-BI-Phosphohydrolase ␣ Galactosidase  Galactosidase  Glucuronidase ␣ Glucosidase  Glucosidase N-acetyl- gluc.aminidase ␣ Mannosidase ␣ Fucosidase
0 0 0 0 0 0 0 Posit. 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Posit. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Posit. 0 0
50 70 30 0 70 0 0 0 0 95 75 0 100 80 20 95 100 0 75
0 to 4 0 to 3 0 to 2 0 to 3
0 to 5 0 to 5 1 0 0 0 1
to to to to to
4 3 1 3 5
0 to 2
*The intensity of the visually estimated coloration coded as 1 corresponds to the liberation of 5 substrat nanomoles, 2 to 10 nanomoles, 3 to 20 nanomoles, 4 to 30 nanomoles, 5 greater than 40 nanomoles
plot, a cluster analysis has been performed on these data, where the variables are the 19 enzymatic responses, and where the cases are the different mites tested. The distances between variables have been computed using the Chi-squared distance.
Results Reference enzymes The collagenase and lysozyme enzymes are not revealed by the API ZYM strip. The positive reaction noticed in the trypsin microtube for the collagenase is due to an impurity mentioned by the purchaser (Table 1). For the 20 V. destructor populations, the percentage with/without one of the 19 enzymes and the range of the observed colour grades for each revealed enzyme are given in Table 1 and Figure 1.
951
Figure 1. Percentage of the colour grades per revealed enzyme
Comparison of the enzymatic profiles and activities According to the cluster analysis (Figure 2), three groups can be distinguished: (A) enzymes never revealed: lipase (C14), valine arylamidase, cystine arylamidase, trypsin, ␣ chymotrypsin, ␣ galactosidase, ␣ mannosidase (B) enzymes not always revealed: esterase (C4), alkaline phosphatase, ␣ fucosidase, ␣ glucosidase (C) enzymes generally present with high activity: N-acetyl--glucosaminidase (chitinase), leucine arylamidase, acid phosphatase,  galactosidase,  glucosidase,  glucuronidase, esterase lipase (C8), naphtol hydrolase. Although the method used is not quantitatively precise, large variations between samples were observed in chitinase activity (2 to 40 or more nanomoles of substrate degraded).
Discussion Significance of the groups In general each group consists of enzymes belonging to the same metabolic pathway. The range of the intensities in a given group can be interpreted following two ways: 1. Most of the enzymes are not revealed, only one is seldom revealed: – if the corresponding colour grade is high, then the enzyme is of great importance when distinguishing parasite strains, – if the color grade is 1, then the enzyme activity
952
Figure 2. Classification of enzymes using the distances obtained by cluster analysis: – group A: enzymes never revealed – group B: enzymes not always revealed – group C: enzymes generally present with high activity
might be to low for the API ZYM method to detect it. 2. All the enzymes are revealed, implying that the activity of some of them may be related to the intensity of the parasite pathogenicity, for instance the chitinase. Meaning of the profile Digestive enzymes Although endopeptidases were often associated with the insect midgut (Jimenez and Gilliam 1989; Valaitis 1995), neither trypsin-like, nor chymoptrypsin-like activities were revealed with the API ZYM test on V. destructor, which is congruent with the previous literature (Tewarson and Jany 1982). Concerning the exopeptidases, carboxypeptidase A and alanine aminopeptidase were detected by these same authors. In the present work, we noticed a leucine aminopeptidase. To complete the screening, other aminopeptidases have to be checked, especially those releasing tyrosin and phenylalanine, which are of importance for exocuticle sclerotization. However these enzymatic activities could help the parasite to provide for the supply of amino acids (Telfer and Kunkel 1991; Crailsheim and Leonard 1997) by degrading such honeybee proteins as arylphorin. The presence and the activity of the acid phosphatase and the phosphohydrolase is in relation with an active cellular metabolism.
953 Presence of the chitinase and enzymes in relation with the glycoprotein metabolism Chitinase is present in all samples of female adult mites and generally very active. The enzymes necessary for the complete chitin digestion such as chitinase and  glucosidase are present.  galactosidase,  glucosidase,  glucuronidase, ␣ fucosidase can be related to the glycoprotein metabolism (Stryer 1992). Pathogenicity of Varroa destructor V. destructor possesses enzymes that can injure the cuticle (chitinase and proteases), but the absence of endoproteases would mean that the bee cuticle piercing is not possible by mere enzymatic action. However some of these enzymes could counteract wound responses to maintain a puncture site open for up to ten days (Lackie 1988). Viruses carried by the parasite directly enter the bee haemolymph, the barrier with the external surroundings being destroyed. Therefore the incidence of viral diseases is higher in parasitized bee colonies. Because of the rapidity needed to ensure its reproduction, the parasite cuticle must be synthetised faster than that of the bee. It would reinforce the hypothesis that host cuticle, or host cuticle precursors such as arylphorin, can be digested and assimilated as amino acids or  carbohydrates by the mother mites or by the nymphal stages of the mite. Consequences on the host-parasite relationships When the equilibrium between A. mellifera and V. destructor is realised in resistant strains of bees, the chitinase activity has to be precisely measured to verify if the enzyme is less active or inhibited by specific molecules (Blattner et al. 1996), allowing the closing of the mite feeding site. This suggestion has to be verified particularly when the percentage of unfertile parasites is very high. According to Anderson and Trueman (2000), differences also could exist at the enzyme level, between the two haplotypes of V. destructor (V. jacobsoni in the previous litterature) infesting A. mellifera.
Acknowledgements This work was supported by 97/1221 E.C. funds
References Anderson D.L. and Trueman J.W.H. 2000. Varroa jacobsoni (Acari: Varroidae) is more than one species. Exp. Applied Acarol. 24: 165–169.
954 Bianchi-Salvadori B., Camascella P. and Cislaghi S. 1995. Rapid enzymatic method for biotyping and control of lactic acid bacteria used in the production of yogurt and some cheeses. Int. J. Food Microbiol. 27: 253–261. Blattner R., Furneaux R.H. and Lynch G.P. 1996. Synthesis of allosamidin analogues. Carbohydr. Res. 294: 29–40. Brasch J. and Zaldua M. 1994. Enzymes patterns of Dermatophytes. Mycoses 37: 11–16. Chernoff H. 1973. The use of Faces to represent points in k-dimensional space graphically. J. Am. Statist. Assoc. 68: 361–368. Correa-Marques M.H. and de Jong D. 1998. Uncapping of worker bee brood, a component of the hygienic behavior of Africanized honey bees against the mite Varroa jacobsoni. Apidologie 29: 283– 289. Crailsheim K. and Leonard B. 1997. Amino acids in worker bee haemolymph. Amino Acids 13: 141– 153. Dandeu J.P., Lux M., Colin M.E., Rabillon J. and David B. 1991. Etude immuno-chimique de l’hémolymphe d’ouvrière adulte (Apis mellifera L.) saine ou infestée par Varroa jacobsoni. Apidologie 22: 37–42. Danty E., Arnold G., Burmester T., Huet J.C., Huet D., Pernollet J.C. et al. 1998. Identification and developmental profiles of hexamerins in antenna and haemolymph of the honeybee, Apis mellifera. Insect Biochem. Mol. Biol. 28: 387–397. Donzé G. and Guérin P. 1994. Behavioural attributes and parental care of Varroa mites parasitising honey bee brood. Behav. Ecol. Sociobiol. 34: 305–319. Fries I., Huazhen W., Wei S. and Jin C.S. 1996. Grooming behaviour and damaged mites (Varroa jacobsoni) in Apis cerana cerana and Apis mellifera ligustica. Apidologie 27: 3–11. Guerra J.C., Goncalves L.S. and de Jong D. 2000. Africanized honey bees (Apis mellifera L.) are more efficient at removing worker brood artificially infested with the parasitic mite Varroa jacobsoni than are Italian bees or Italian/Africanized hybrids. Genetics Molec. Biol. 23: 89–92. Jimenez D.R. and Gilliam M. 1989. Age-related changes in midgut ultrastructure andd trypsin activity in the honey bee. Apidologie 20: 287–303. Jones J.B., Bouzar H., Stall R.E., Almira E.C., Roberts P.D., Bowen B.W. et al. 2000. Systematic analysis of xanthomonads (Xanthomonas spp.) associated with pepper and tomato lesions. Int. J. System. Evol. Microbiol 50: 1211–1219. Kramer K.J. and Muthukrishnan S. 1997. Insect chitinases: molecular biology and potential use as biopesticides. Insect Biochem. Molec. Biol. 27: 887–900. Lackie A.M. 1988. Immune mechanisms in insects Parasitol. Today 4: 98–105. Medina Medina L. and Martin S.J. 1999. A comparative study of Varroa jacobsoni reproduction in worker cells of honey bees (Apis mellifera) in England and Africanized bees in Yucatan, Mexico. Exp. Appl. Acarol. 23: 659–667. Mira A. 2000. Exuviae eating: a nitrogen meal? J. Insect Physiol. 46: 605–610. Rosenkranz P. and Engels W. 1994. Unfertility of Varroa jacobsoni females after invasion into Apis mellifera worker brood as a tolerance factor against varroatosis. Apidologie 25: 402–411. Ryan R.O., Schmidt J.O. and Law J.H. 1984. Arylphorin from the haemolymph of the larval honey bee Apis mellifera. Insect Biochem. 14: 515–520. Schatton-Gadelmeyer and Engels W. 1988. Hämolymphproteine und Körpergewicht frischgeschlüpfter Bienen-Arbeiterinnen nach unterschiedlich starker Parasitierung durch Brutmilben (Apis mellifera/ Varroa jacobsoni). Entomol. Gener. 14: 93–101. Stryer L. 1992. La Biochimie. Flammarion, Paris, 1088 pp. Telfer W.H. and Kunkel J.G. 1991. The function and evolution of insect storage hexamers. Annu. Rev. Entomol. 36: 205–228. Tellam R.L., Wijfells G. and Willadsen P. 1999. Peritrophic matrix proteins. Insect Biochem. Mol. Biol. 29: 87–101. Tewarson N. and Engels W. 1982. Undigested uptake of non-host proteins by Varroa jacobsoni. J. Apic. Res. 21: 222–225.
955 Tewarson N. and Jany K.D. 1982. Determination of proteolytic activity in Varroa jacobsoni an ectoparasitic hemophagous mite of honey bees. Apidologie 13: 383–389. Valaitis A.P. 1995. Gypsy moth midgut proteinases: purification and characterisation of luminal trypsin, elastase and the brush border membrane leucine aminopeptidase. Insect Biochem. Molec. Biol. 25: 139–149. Vandame R., Colin M.E., Morand S. and Otero Colina G. 2000. Level of compatibility in a new hostparasite association: Varroa jacobsoni-Apis mellifera. Can. J. Zool. 78: 2038–2044.
