Tomato Yellow Leaf Curl Geminivirus (TYLCV-Is) Is Transmitted among Whiteflies (Bemisia tabaci) in a Sex-Related Manner

Abstract

Tomato yellow leaf curl virus (TYLCV) is the name given to a complex of geminiviruses infecting tomato cultures worldwide. TYLCV is transmitted by a single insect species, the whitefly Bemisia tabaci. Herein we show that a TYLCV isolate from Israel (TYLCV-Is) can be transmitted among whiteflies in a sex-dependent manner, in the absence of any other source of virus. TYLCV was transmitted from viruliferous males to females and from viruliferous females to males but not among insects of the same sex. Transmission took place when insects were caged in groups or in couples, in a feeding chamber or on cotton plants, a TYLCV nonhost. The recipient insects were able to efficiently inoculate tomato test plants. Insect-to-insect virus transmission was instrumental in increasing the number of whiteflies capable of infecting tomato test plants in a whitefly population. TYLCV was present in the hemolymph of whiteflies caged with viruliferous insects of the other sex; therefore, the virus follows, at least in part, the circulative pathway associated with acquisition from infected plants. Taken as a whole, these results imply that a plant virus can be sexually transmitted from insect to insect.

Tomato yellow leaf curl virus (TYLCV) is the name given to a complex of genetically different geminiviruses (family Geminiviridae, genus Begomovirus) affecting tomato cultures worldwide (11). TYLCV is transmitted exclusively by the whitefly Bemisia tabaci in a circulative manner (7, 28). Although acquisition and transmission parameters of TYLCV (as well as other begomoviruses) have been studied extensively (1, 5, 22, 25, 31), the association between TYLCV and B. tabaci is still poorly understood. Using a local virus isolate and an insect colony maintained in the laboratory, we have found that the association of TYLCV with its insect vector is suggestive of a pathogen-host interaction. Once acquired, the genome of the virus remains associated with the insect for its entire adult life. This long-term relationship was associated with a decrease in virus transmission efficiency, longevity, and fecundity of the insect (28). The virus was transmitted to the progenies of viruliferous whiteflies for at least two generations (13). In this study we have investigated the question of whether TYLCV can be transmitted from insect to insect in a sex-dependent manner.

Transmission of viruses through the gametes of insects has been documented, especially for Drosophila spp. (3). One of the best-studied virus of this kind is the Drosophila S virus (DSV), a reolike virus (20) which causes developmental malformation (the S phenotype) in the SimES strain of Drosophila simulans (10). Microscopic observations revealed invasion of early differentiating male and female germ cells by DSV particles (21). DSV was transovarially transmitted to some of the progenies. The rate of virus transmission by males was lower than that by females, probably because the proportion of infected spermatozoa was relatively smaller (21). Another well-studied virus is the baculovirus-like gonad-specific virus (GSV) that causes abnormalities in the reproductive systems (the agonadal syndrome) in both infected males and females of two closely related moth species, Helicoverpa zae and H. armigera (26). GSV is confined to the reproductive system. It penetrates the eggs before their chorion is hardened prior to oviposition. The sperm could also carry virions into the egg at the time of fertilization (26). GSV was able to replicate in TN-368 cells in cultures (4).

In this communication, we provide the first report of a plant virus transmitted by viruliferous males to nonviruliferous females and vice versa, in the absence of any other virus source.

MATERIALS AND METHODS

Maintenance of virus cultures, whiteflies, and plants.

Cultures of TYLCV-Is (an isolate from Israel) (24) were maintained in tomato plants (Lycopersicon esculentum cv. Daniella). B. tabaci of the B biotype (9) were reared on cotton plants (Gossypium hirsutum cv. Akala), a TYLCV nonhost, grown in insect-proof wooden cages at 24 to 27°C (31). The colony comprised approximately twice as many females as males at the time the experiments were done.

Acquisition of TYLCV by whiteflies from infected tomato plants and transmission to tomato test plants.

Viruliferous whiteflies were obtained after caging adults, 1 to 2 weeks after emergence, with TYLCV-infected tomato plants at the six- to eight-leaf stage for a 48-h acquisition period. Tomato plants at the four- to six-leaf stage were used for whitefly-mediated inoculations.

Rearing insects on artificial medium.

Whiteflies collected from the colony were anesthetized by a 10- to 15-min incubation at −10°C. Males and females were separated with a fine paintbrush under the binocular. Whiteflies were introduced in 3-cm-diameter, 4-cm-high black plastic cylinders covered with a layer of stretched Parafilm membrane and standing on a black plastic board. About 0.2 ml of a 15% sucrose solution supplemented with yellow food dye was deposited on the membrane and covered with a second layer of stretched membrane. Insects that died during caging accumulated at the bottom of the cage.

Detection of TYLCV DNA in insects and in plants.

Whitefly and plant DNA was prepared as described elsewhere (12, 13). PCR amplification of TYLCV DNA was performed with primers V61 and C473 (13). The PCR products were subjected to electrophoresis in a 1% agarose gel, stained with ethidium bromide, and photographed. Viral DNA in tomato plants was identified by Southern blot hybridization using the radiolabeled plasmid pTYH20.7 (contains a dimeric copy of the TYLCV genome [24]) as the probe.

Detection of TYLCV CP in insects and in plants.

The virus coat protein (CP) was detected by immunocapture-PCR (18) (in whiteflies and in tomato plants) and by Western blot immunodetection (23) (plants), using an antibody raised against the CP of a TYLCV isolate from the Dominican Republic overexpressed in Escherichia coli (a gift from R. Gilbertson). The buffers used for immunocapture-PCR are described online (www.bioreba.com/list8.html). PCR tubes were filled with 200 μl of antiserum (1:1,000 diluted in coating buffer), incubated for 3 h at 37°C, and washed five times for 5 min each with 200 μl of washing buffer. Whitefly or plant homogenates in 200 μl of extraction buffer were incubated in the coated PCR tubes for 18 h at 4°C. The tubes were then washed five times for 5 min each with 200 μl of washing buffer and dried. PCR amplification of the viral DNA from the virions bound to the antibody-coated tubes was performed with primer pairs V61 and C473 (13). The TYLCV CP was immunodetected in plants by Western blot analysis using the same antiserum (23) and visualized by chemiluminescence (ECL kit; Amersham).

RESULTS

TYLCV DNA and CP are present in whiteflies that have been caged with viruliferous insects of the other sex.

Transmission of TYLCV between whiteflies of the opposite sex was investigated. Twenty viruliferous males and 20 nonviruliferous females were introduced in a feeding cage. The insects settled and started to feed on the artificial medium within 1 to 2 h. Courtship and copulation were observed under the binocular through the translucent membrane. After 48 h, DNA was prepared from each of the surviving females, and the presence of TYLCV DNA was assessed by PCR. Figure 1A shows that viral DNA was found in 10 of the 18 females. In the reciprocal experiment, 20 viruliferous females and 20 nonviruliferous males were caged together. Five of the 18 surviving males contained viral DNA. An identical experiment was performed to investigate whether the virus CP was also transmitted between insects of the opposite sex. Insect extracts were incubated with PCR tubes coated with an antiserum raised against the TYLCV CP, and the DNA of the immunocaptured virions was detected by PCR. Figure 1B shows that four of the nine females tested contained TYLCV CP transmitted by viruliferous males; similarly, three of the nine males tested contained CP transmitted by viruliferous females. TYLCV CP was not detected when PCR tubes were not coated with the antiserum or when the insect extract was omitted (see also Fig. 4). These results indicated that TYLCV was transmitted from viruliferous insects to nonviruliferous insects of the opposite sex, likely in the form of an encapsidated virion.

FIG. 1.

Transmission of TYLCV (DNA and CP) from viruliferous males to females and from viruliferous females to males. (A) Analysis by PCR. Twenty viruliferous males were caged with 20 nonviruliferous females, and vice versa. After 48 h, the DNA prepared from each surviving insect was subjected to PCR using TYLCV DNA-specific primers. The products were subjected to agarose gel electrophoresis and stained. (B) Analysis by immunocapture-PCR. Male and female whiteflies were caged as for panel A. After 48 h, extracts prepared from individual insects (nine males and nine females) were incubated in PCR tubes coated with an antiserum raised against TYLCV CP. The DNA of the virions bound to the antibody was amplified by PCR. The products were subjected to agarose gel electrophoresis and stained. P, plasmid pTYH20.7; M* and F*, viruliferous male and female; M and F, nonviruliferous male and female. Thick arrows, ∼410-bp amplified viral DNA fragments; thin arrow, primers.

FIG. 4.

Inoculation of tomato plants by insects caged with viruliferous insects of the other sex. Thirty viruliferous males were mixed with 30 nonviruliferous females. After 48 h, the females were collected and caged with tomato test plants, one insect per plant (upper panel). The reciprocal experiment was conducted (lower panel). Plants were analyzed after 5 weeks. (S) DNA was extracted from each plant and was Southern blot hybridized with a TYLCV-specific radiolabeled DNA probe. ssDNA, TYLCV genomic single-stranded DNA; dsDNA, virus genomic double-stranded replicative form. (P) The same DNA served as template for PCR. (I) Extracts of the same plants were used for immunocapture-PCR. T* and T, infected and noninfected tomato plants; 0A, no antibody; 0E, no plant extract. Thick arrows, ∼410-bp amplified viral DNA fragments; thin arrows, primers.

An experiment similar to that described above was conducted, with the difference that insects were collected after 4 and 8 h of caging. The results summarized in Table 1 indicate that increasing the length of the caging period did not necessarily lead to a linear increase in the rates of TYLCV transmission.

TABLE 1.

Transmission of TYLCV from viruliferous whiteflies to insects of the opposite sex^a

Duration of caging (h)	No. (%) of insects with TYLCV DNA/total after passage from:
	Female to male	Male to female	Male to male	Female to female
4	2/12 (16)	14/24 (58)
8	9/17 (53)	4/13 (30)
48	5/18 (30)	10/18 (56)
			0/20 (0%)	0/20 (0%)

^aViruliferous females were caged with an equal number of nonviruliferous males. After the time indicated, the males were collected and analyzed individually for presence of TYLCV DNA by PCR. The reciprocal experiment was done. Viruliferous whiteflies were also caged with nonviruliferous insects of the same sex. 

Since viruliferous whiteflies could contaminate the artificial medium with virus-containing saliva during feeding (27), we investigated the question of whether this medium could serve as a virus source for nonviruliferous insects. One hundred viruliferous whiteflies were introduced in a cage and allowed to feed on artificial medium. After 48 h, the insects were removed. One hundred nonviruliferous insects were introduced into the cage and allowed to feed on the same medium. The experiment was done in duplicate. PCR analysis indicated that the insects collected after 48 h did not contain viral DNA (Fig. 2). Insects handled similarly were unable to inoculate tomato test plants (9 plants, 10 insects/plant, 48-h inoculation-access period). Therefore, the nonviruliferous whiteflies caged with viruliferous insects of the other sex did not acquire TYLCV by feeding on contaminated medium.

FIG. 2.

Whiteflies do not acquire TYLCV from artificial medium used to feed viruliferous whiteflies. One hundred viruliferous whiteflies were allowed to feed on artificial medium. After 48 h, the insects were removed and replaced with 100 nonviruliferous insects that were allowed to feed on the same medium. After 48 h, the whiteflies were randomly collected in groups of 10 and analyzed for the presence of TYLCV DNA by PCR. P, plasmid pTYH20.7; T*, infected tomato plant; W*, viruliferous whiteflies; F-W, nonviruliferous whiteflies 48 h after feeding on the medium the viruliferous insects fed on; W, nonviruliferous whiteflies from the colony; 0, PCR without DNA. Thick arrow, ∼410-bp amplified viral DNA fragment; thin arrow, primers.

The question of whether TYLCV is transmissible among whiteflies of the same sex was investigated. Twenty viruliferous females and 20 nonviruliferous females were introduced in a feeding cage. The same was done with 20 viruliferous and 20 nonviruliferous males. The viruliferous insects were marked with a tiny blue dot on the dorsal side of the thorax. After 48 h, all of the whiteflies (alive and dead) were analyzed by PCR. In each case, viral DNA was detected in all the 20 dotted viruliferous whiteflies but none of the nonviruliferous insects, establishing that TYLCV was not transmitted among insects of the same sex (Table 1).

To confirm that TYLCV was transmitted during contact of sexual partners, 10 couples of insects (one viruliferous male and one nonviruliferous female) were enclosed in 10 separate cages. After 24 h, the presence of viral DNA was assessed by PCR as in Fig. 1A. Six of the 10 females contained TYLCV DNA. In the reciprocal experiment, 3 of the 10 males caged with viruliferous females contained TYLCV DNA (not shown). When the experiment was performed with five couples of males (one viruliferous and one nonviruliferous) and five couples of females (one viruliferous and one nonviruliferous), none of the nonviruliferous partners, males or females, contained detectable viral DNA (not shown). Taken together, the experiments described above demonstrated that TYLCV could be transmitted between insects during contact of partners of the opposite sex, probably during intercourse.

TYLCV can be transmitted serially among sexual partners.

We have investigated the question of whether TYLCV acquired by a whitefly from a sexual partner could be transferred to a partner of the opposite sex and for how many passages could it be transmitted in this fashion. One hundred viruliferous males were caged with 100 nonviruliferous females. After 24 h, 60 of the surviving females were collected and caged with 100 nonviruliferous males for an additional 24 h and then analyzed for the presence of viral DNA. Of the surviving males, 60 individuals were collected, caged with 100 nonviruliferous females for 24 h, and then analyzed for the presence of viral DNA. The experiment was continued until four passages were completed, starting from the initial viruliferous males. The reciprocal experiment was conducted in a similar manner, starting with 100 viruliferous females caged with 100 nonviruliferous males. The viruliferous males passed the virus to 83% of the females. These females passed it to 39% of the males, who in turn passed it to 33% of the females, who finally passed it to 11% of the males. In the reciprocal experiment, the viruliferous females passed the virus to 67% of the males, who passed it to 39% of the females, who in turn passed it to 22% of the males; finally, the latter passed the virus to 22% of the females.

TYLCV transmission among sexual partners contributes to the spread of the virus in a whitefly population.

The contribution of TYLCV transmission among sexual partners to the increase in the number of viruliferous insects in a whitefly population was investigated. Three viruliferous males and three viruliferous females were mixed with 120 nonviruliferous insects (about 40 males and 80 females) randomly picked from the insect colony. All insects were reared on a cotton plant, a TYLCV nonhost. Groups of whiteflies were collected randomly after 2, 4, 6, and 8 days. For each group, males and females were separated and the presence of TYLCV in each individual was assessed by PCR. After 2 days, none of the eight males and one of the eight females collected contained TYLCV DNA. After 4 days, 2 of the 8 males and 2 of the 20 females sampled contained viral DNA. After 6 days, 1 of the 8 males and 5 of the 22 females collected contained viral DNA. After 8 days, 1 of the 10 males and 8 of the 20 females collected contained viral DNA. Altogether 4 of the 34 males tested contained viral DNA (one more than the input), and 16 of the 70 females tested contained viral DNA (13 more than the input).

In the previous experiment, the whiteflies sampled were subtracted from the population. If a sample contained one or more viruliferous insects, their removal prevented further spread of TYLCV. To obtain an accurate image of the spread of the virus among the whiteflies, three identical populations were established. Each population contained three viruliferous males and three viruliferous females together with 120 nonviruliferous insects (again the ratio of male to female was about 1 to 2). The three populations were reared separately on three cotton plants. The insects of the first population were collected after 3 days. Males and females were separated, and the insects were analyzed for presence of TYLCV DNA by PCR. The second and third populations were similarly processed after 5 and 7 days, respectively. After 3 days, 21% of the males and 33% of the females contained viral DNA. After 5 days, these values increased to 50% of the males and 47% of the females. After 7 days, 69% of the males and 51% of the females contained viral DNA. These results show that TYLCV spreads with time among the insect population; during 7 days the percentage of viruliferous males increased from 7 to 69, and that of viruliferous females increased from 3.7 to 51.

TYLCV acquired by females from viruliferous males can be transovarially transmitted to adult progeny.

We have shown previously that TYLCV can be transmitted to whitefly progeny through the egg (13). We have investigated whether virus acquired by females caged with viruliferous males can be transovarially transmitted. Thirty nonviruliferous females and 30 viruliferous males were caged together with a cotton plant. After 4 days, the insects were removed from the plant and the eggs were allowed to develop. DNA extracted from groups of five adult insects that emerged from the eggs was analyzed by PCR for the presence of TYLCV DNA. Figure 3 shows that viral DNA was present in two of the six groups tested, indicating that TYLCV acquired by whiteflies from viruliferous males was transmitted to eggs. The ability of these insects to infect tomato test plants was examined. In three independent identical experiments, 40 tomato test plants were caged with whiteflies (five insects per plant) for a 48-h inoculation-feeding period. None of the plants developed disease symptoms or contained detectable viral DNA after 8 weeks.

FIG. 3.

TYLCV acquired by female whiteflies from viruliferous males can be transovarially transmitted to progeny. Thirty nonviruliferous females and 30 viruliferous males were caged together with a cotton plant. After 4 days, the insects were removed from the plant and the eggs were allowed to develop. DNA extracted from six groups of five adult progeny was analyzed by PCR for the presence of TYLCV DNA. P, plasmid pTYH20.7; W*, viruliferous whiteflies; W, nonviruliferous whiteflies from the colony. Thick arrow: ∼410-bp amplified viral DNA fragment; thin arrow, primers.

Whiteflies that acquired TYLCV from insects of the other sex are able to transmit the virus to tomato test plants.

We examined the ability of insects that had acquired TYLCV from their sexual partner to transmit the virus to tomato test plants. Thirty viruliferous males were mixed with 30 nonviruliferous females. After 2 days, the females were collected and caged with tomato test plants, one insect per plant. Five weeks thereafter, 10 of the 29 plants showed typical disease symptoms. The reciprocal experiment was conducted where 30 viruliferous females were mixed with 30 nonviruliferous males. Five of the 21 plants caged with males presented symptoms. Figure 4 shows the analysis of nine plants from each of the two sets of experiments. For each tomato plant, the presence of the viral DNA was assessed by Southern blot hybridization and by PCR, and the CP was detected by immunocapture-PCR. The results provided by each of the three detection methods were concordant. Five of the nine plants were infected by the females previously caged with the viruliferous males. Two of the nine plants were infected by males previously caged with viruliferous females. Southern blot analyses showed the genomic single-stranded DNA and its double-stranded replicative form, typical of infected plants. Immunocapture-PCR demonstrated that the plant extracts contained encapsidated virions that bound to the CP antibody. The encapsidated DNA subsequently served as template for PCR. A positive signal was obtained only when both the antibody and the infected plant extract were involved in the procedure. The presence of the CP in the infected plants was confirmed by Western blot immunodetection. Figure 5 shows that the ∼30-kDa CP was conspicuous in those plants that were shown to be infected by the other methods (Fig. 4). These results demonstrated that the virus is transmitted among sexual partners, likely in the form of an infectious encapsidated virion, and that it is in sufficient amounts to produce a systemic disease in tomato.

FIG. 5.

Western blot immunodetection of TYLCV CP in plants infected by whiteflies that acquired the virus from viruliferous insects of the other sex. The plants analyzed are those found to be infected in Fig. 4. Extracts of plants infected by females caged with viruliferous males (M* to F) and by males caged with viruliferous females (F* to M) were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The blotted proteins were reacted with an antiserum raised against TYLCV CP. CP is indicated by an arrow. T* and T, infected and noninfected tomato plants.

Spread of TYLCV among sexual partners leads to an increase in the ability of a whitefly populations to transmit TYLCV to tomato test plants.

We have investigated the question of whether sex-mediated transmission of TYLCV leads to an increase in inoculativity of a whitefly population. Two hundred whiteflies were caged with six couples of viruliferous whiteflies (six males and six females). After 48 h, the insects alive were collected and randomly divided into groups of three insects each. Each group was caged with a tomato test plant (58 plants altogether) for a 72-h inoculation-feeding period. Four weeks thereafter, infection was determined by Southern blot hybridization as in Fig. 4 (Table 2, experiment 1). In a similar experiment, three couples of viruliferous whiteflies (three males and three females) were caged with 200 nonviruliferous insects. The whiteflies were divided into groups of five insects, and each group was caged one of 30 tomato test plants (Table 2, experiment 2).

TABLE 2.

Contribution of whitefly-to-whitefly TYLCV transmission to the ability of insect populations to inoculate TYLCV^a

Expt	n	No. of infected plants
		Expected	Obtained
1	58	12	22
2	30	6	11

^aIn experiment 1, 200 whiteflies were mixed with six couples of viruliferous insects. After 48 h, groups of insects were collected randomly in groups of three, and each group was caged for 72 h with one tomato test plant. In experiment 2, the nonviruliferous insects were mixed with three couples of viruliferous whiteflies, and groups of five insects were caged with tomato test plants. Infection was assessed by Southern blot hybridization after 4 weeks. 

TYLCV DNA is present in the hemolymph of whiteflies that acquired TYLCV from insects of the other sex.

The previous experiments show that insects that have acquired TYLCV from whiteflies of the other sex are able to infect tomato test plants. Therefore, the virus has to follow, at least in part, the path inherent to acquisition from plants. One of the crucial steps in the virus circulative path is the passage from the digestive track through the hemolymph on the way to the salivary gland (15). Therefore, we investigated the presence of TYLCV in the hemolymph of insects caged with viruliferous insects of the other sex. Three groups of 20 viruliferous males were caged with 20 nonviruliferous females. After 4, 8, and 24 h of caging, hemolymph was collected from groups of five females and analyzed by PCR for the presence of TYLCV DNA. The reciprocal experiment was conducted. The results presented in Fig. 6 show that the viral DNA was not detected in samples collected after 4 h but was conspicuous in the hemolymph of insects collected after 8 and 24 h, males as well as females. The time course was compared with the velocity of translocation of TYLCV acquired by female whiteflies from infected tomato plants. After various acquisition access periods (AAP), the midgut (after flushing with sterile water) and the hemolymph of single insects were subjected to PCR. Figure 6 shows that TYLCV was detected in the midgut after a 1-h (but not 0.5-h) AAP and in the hemolymph 30 min thereafter. Therefore, TYLCV reaches the hemolymph much faster when it is acquired from plants than when it is acquired from another insect (approximately 1.5 h versus more than 4 h).

FIG. 6.

Presence of TYLCV DNA in the hemolymph of whiteflies caged with viruliferous insects of the other sex. (Left) Three groups of 20 viruliferous males were caged with 20 nonviruliferous females, and three groups of viruliferous females were caged with nonviruliferous males. Hemolymph was collected from groups of five nonviruliferous male (M) and female (F) insects and analyzed by PCR for the presence of TYLCV DNA after 4, 8, and 24 h of caging. P, plasmid pTYH20.7; H, hemolymph from five nonviruliferous insects from the insect colony. Thick arrow, ∼410-bp amplified viral DNA fragment; thin arrow, primers. (Right) Female whiteflies were caged with infected tomato plants. After the time indicated (hours), the midgut was dissected, flushed with sterile distilled water, and subjected as is to PCR. Hemolymph was collected and subjected to PCR. Two insects were analyzed for each time point.

DISCUSSION

Male and female whiteflies caged together on artificial diet or on cotton plants were often seen as copulating couples soon after they settled. If one of the partners was viruliferous, it could transmit TYLCV to its sexual partner. TYLCV was transmitted from viruliferous males to females and from viruliferous females to males but not among insects of the same sex (Fig. 1; Table 1). Transmission occurred when insects were caged in groups or in couples. Transmission took place when the insects were let to feed on artificial diet in a chamber or when they were caged with cotton plants, a TYLCV nonhost. Both viral DNA and CP were detected in the recipient whiteflies, strongly indicating that the insects acquired encapsidated virions (Fig. 1). The recipient insects were able to efficiently inoculate tomato test plants (Fig. 4). These plants contained the virus genomic DNA and its replicative form as well as the virus CP (Fig. 4 and 5) and presented the symptoms of a systemic infection. Therefore, the virus whiteflies acquired from sexual partners had all the infectious properties characteristic of TYLCV virions ingested from infected tomato plants. Insect-to-insect transmission was instrumental in increasing the number of whiteflies able to infect tomato test plants (Table 2). TYLCV was present in the hemolymph of whiteflies that acquired the virus from sexual partners (Fig. 6), indicating that the virus follows, at least in part, the circular path inherent to acquisition from plants. TYLCV reached the haemolymph more than 4 h after the whiteflies have been caged with viruliferous insects of the other sex; in comparison, the virus was found in the hemolymph of insects caged with infected tomato plants after 1.5 h. In these experiments, it is likely that the hemolymph was not contaminated by virus from other organs (mainly from the digestive tract). Indeed, TYLCV could be detected in the midgut of insects feeding on infected tomato plants at the time (after a 1-h AAP) when the hemolymph was devoid of detectable virus (Fig. 6).

Transmission of TYLCV by feeding on contaminated diet was excluded. We showed previously that whiteflies did not acquire virus while feeding on cotton plants on which viruliferous insects were previously reared (13). Similarly we have ruled out the possibility that whiteflies acquired virus while feeding on artificial diet contaminated by the saliva of viruliferous insects (Fig. 2). Such a transmission was conceivable in the light of a report indicating that the squash leaf curl virus, another whitefly-transmitted geminivirus, was detected by PCR in the diet of viruliferous insects (27). In our case, TYLCV DNA was undetectable by PCR in the medium viruliferous whitefly fed on for up to 48 h (not shown). Moreover, insects that fed for 48 h on a diet used to feed viruliferous whiteflies for 48 h did not contain detectable amounts of TYLCV DNA (Fig. 2) and were unable to infect tomato test plants.

Taken collectively, and unless some unknown mechanism of virus transmission exists, our results could be explained only by a sexually related transmission of TYLCV. On one hand males are contaminated by viruliferous whiteflies and on the other hand males contaminate females. We do not know how TYLCV is transmitted among sexual partners and whether the insect gonads are infected. Immunolocalization of the virus may reveal the presence of TYLCV in the germ cells. Whitefly copulation and fertilization of eggs have not been well described. We can only infer from what is known from other insects (6). The fact that virus is found in the hemolymph of recipient males and females more than 4 h after the start of sexual contact (Fig. 6) and the fact that progeny of these females contain virus (Fig. 3) points to several possible modes of transfer. In case of sperm cells and or the seminal liquid of viruliferous insects contain TYLCV, which is by no means proven, the virus is transmitted to the female during copulation. The sperm migrates to the spermatheca, where it is used by the female to fertilize eggs, which in some cases occurs several days after insemination or not at all (29). Fertilized eggs may be infected and progeny may contain TYLCV as we indeed observed (Fig. 3), although these insects were unable to infect tomato test plants, either because the amount of virus was insufficient or because the virus did not reach the salivary glands, or both. We may also encounter a situation in which some of the sperm is injected into the haemocoel as described for some insects (16). Fertilization is then ensured when the sperm adjacent to the lowest follicle penetrate the follicular epithelium by pinocytosis and may enter the oocytes (30). The remaining sperm disperses in the hemolymph, where it may be digested by blood cells or by phagocytes (19). Such a situation would explain the presence of virus in the hemolymph of female whiteflies after copulation (Fig. 6). Similarly, we do not know how the viruliferous female contaminates the male. Also in this case, the virus is found in the male hemolymph, indicating that contamination occurs through the body fluids. During copulation, hemolymph of male and female may mix, thereby favoring the passage of TYLCV from one individual to the other. Indeed, whiteflies have an open blood system with the hemolymph occupying the general body cavity, circulating between the various organs, bathing them directly, and providing them with viral nutrients (6). The hemolymph is also instrumental in the circulative transmission of plant viruses such as TYLCV and other begomoviruses. When the insect feeds on an infected plant, virions are ingested along with phloem sap through the stylets and enter the esophagus, the filter chamber, and the anterior portion of the midgut. Particles translocate from the gut lumen to the hemolymph (15, 17). Virions that reach the salivary glands are excreted with the saliva during feeding (14). Virus that reached the hemolymph following sexual activities will have to be translocated to the salivary glands, likely as an infectious encapsidated virion (Fig. 1 and 4).

Males and females acquire TYLCV from their sexual counterparts with comparable frequency. Likewise, the virus spread in the insect male and female populations with similar velocity. Moreover, the ability of the insects that acquired the virus to infect plants was similar for males and females. The later point was different from the frequency of transmission by male and female whiteflies that acquired virus from infected plant. In this case, females transmitted TYLCV about 5 times more efficiently than males (8). Therefore, it seems that the path of virus circulation in the insect body once transmitted by the other sex is equally efficient in both males and females.

We do not know whether the phenomenon described here is general or is specific for our insect colony. This is a colony established from insects collected in the field about 11 years ago. Many years of rearing may have rendered the colony fragile and prone to infection by geminiviruses. It has been shown for another virus-host system (sigma virus of Drosophila melanogaster) that the efficiency of virus transmission is greatly increased in stabilized insect strain compared to nonstabilized strains (2).