Nuclear polyhedrosis virus of Helicoverpa armigera (HaNPV) from Madurai (south India) was isolated, purified and subjected to electron microscopical studies. The results indicate that the isolate is a multi embedded form of NPV. Further, STEM observations on polyhedra and polyhedral envelope of the HaMNPV are presented. The restriction pattern and genome size of HaMNPV Madurai isolate are compared with HaNPV isolates reported earlier. A dendrogram is presented to show the relatedness among the isolates.

Helicoverpa armigera, a major pest in cotton field, forms two types of polyhedral inclusion bodies (PIBs) (Monroe and McCarthy 1984): (i) single embedded form which has only one nucleocapsid in a viral envelope (SNPV) and (ii) multi embedded form which has more than one nucleocapsid in the viral envelope (MNPV). Patel et al (1968) first reported the occurrence of H. armigera (HaSNPV) in north India and Jacob and Subramanian (1972) reported their occurrence in southern India. Multi embedded form of HaNPV has not so far been detected/reported in India. This communication reports the presence of multi embedded form of HaMNPV in India. Many publications have reported the size of SNPV and MNPV PIBs. However, none of the publications have comprehensively analysed and compared the PIBs of different NPVs. In this paper, the size relatedness of the PIBs was analysed considering the size of PIBs from 23 MNPVs and 16 SNPVs. Jacob and Subramaniam (1972) and Mohan et al (1996) have made electron microscopical observations on the morphology of HaNPV PIBs. Detailed electron microscopic observations of the HaMNPV PIBs are presented in this paper.

Restriction endonuclease analysis (REN) pattern of HaNPV genome has been reported for nine isolates and all the authors have used EcoRI enzyme for restriction (Smith and Summers 1978; Gettig and McCarthy 1982; Mclntosh and Ignoffo 1983; Susan 1983). EcoRI REN pattern of HaMNPV (Madurai isolate) was also studied and comparative analyses of the REN pattern of all the ten isolates are presented.

2. Materials and methods

2.1 Isolation and propagation NPV

PIBs of HaNPV were obtained from the viral infected dead H. armigera larvae collected from cotton fields of Madurai, southern India. This isolate was propagated in the laboratory reared H. armigera larvae by feeding the larvae with artificial diet (Armes et al 1992) containing the suspension of the purified polyhedra (5 × 10⁵ PIBs/ml of diet). The larvae which died due to viral infection were collected and the PIBs were purified from these larvae.

2.2 Purification of PIBs for electron microscopical studies

Following the procedure of Sudhakar et al (1997) the polyhedral inclusion bodies were purified. The samples containing PIBs were subjected to 2% SDS wash twice with vigorous mixing and over night incubation at 4° C. About 2 g of PIBs were further subjected to sucrose cushion (62% wt/vol) ultracentrifugation for 30 min at 17,000 g. The PIBs that formed a band on top of the cushion were collected and again subjected to sucrose cushion (44% wt/vol) centrifugation at 17,000 g for 30 min. The PIB pellet was suspended in 5 ml of water and layered on 10 to 80% (wt/vol) sucrose gradient and spun at 50,000 g for 1 h. The band at the bottom of the gradient containing the PIBs were collected and washed with double distilled (DD) water twice. All the above steps were carried out at 4° C as described by Scharnhort et al (1977). Pure PIBs were suspended in DD water and stored at – 20° C.

2.3 Purification of polyhedral envelope

Following the procedure of Minion et al (1979), polyhedral envelope (PE) was purified. Pure PIBs were incubated in 0·01 M Na₂CO₃ and 0·1 M NaCl (pH 10·5) at 4° C for 4 days with constant gentle shaking. The stages of dissolution were observed during the period and after 4 days, clean PEs were obtained. The PEs were pelleted at 7700 g for 10 min washed with DD water and stored
at – 20° C for electron microscopical (EM) analysis.

2.4 Scanning electron microscopy

The pure PIBs dehydrated in an ethanol series were mounted on a clean glass slide and air dried. The sample mounted on the glass slide was placed on rotating and tilting "omni" table in a sputter coater. Gold palladium wire was loaded in the sputter coater and ignited under high vacuum condition. Gold palladium was allowed to form a uniform layer over the sample. Slowly the vacuum was released. The sample was observed under Joel scanning electron microscope and the image was photographed (Adams and Wilcox 1982).

2.5 Transmission electron microscopy

The PIBs and PEs were mounted on carbon and formvar coated copper grids separately. The grids were stained with 2% filtered uranyl acetate for 1 min. Samples were observed in Joel transmission electron microscope and

photographed. PIB and agar–agar (1%) sandwich was prepared and processed for sectioning following the procedure of Hughes and Addison (1970). Sections of 80 nm thickness were made using Reichert Ultracuts (Type-702501) and mounted on carbon coated grid. The sections were stained with uranyl acetate for 1 min and observed under transmission electron microscope for documentation.

2.6 Viral DNA isolation and restriction analysis

Purified PIBs were lysed in 0·1 M Na₂CO₃ and incubated with proteinase K (100 µg/ml) at 50° C for over night (O’Reilly et al 1992). Viral DNA was isolated following phenol : chloroform : isoamylalcohol, and chloroform : isoamylalcohol extraction (Sambrook et al 1989). The viral DNA was precipitated with 2 volumes of ethanol, washed with 70% ethanol twice and suspended in Tris EDTA (TE) (pH 8). Purified HaMNPV DNA was digested with EcoRI restriction endonuclease. HindIII and EcoRI restriction fragments of DNA have 3¢ recessed ends, which were labelled as follows: the restricted DNA fragments were extracted twice with phenol : chloroform, ethanol precipitated, washed with 70% ethanol and suspended in TE pH 8. Two µg of DNA samples were end filled in the presence of dNTPs (1 mM each except dATP), [³²P]dATP (2 µCi) with Klenow (1·5 U) at room temperature for 30 min. The end-filled DNA samples were freed of unincorporated label as follows: the samples were incubated at – 70ºC for 2 h in the presence of 2 M ammonium acetate and 5 µg yeast tRNA with 2 volumes of ethanol. After spinning the samples at 10,080 g for 10 min, the supernatant was discarded. The pellet containing the DNA was washed twice with 70% ethanol. The samples were air dried and suspended in minimum volume of TE (pH 8).

The end labelled DNA samples were resolved in 0·9% agarose gel and also in 8% acrylamide gel in order to detect all restriction fragments. The gels were dried, exposed to X-ray film and developed. Molecular size was determined for all the fragments using the marker (l /HindIII) and the computer program called DNASIS (1988).

2.7 Construction of dendrogram

EcoRI restriction pattern of ten different isolates of HaNPV DNA have already been published (Smith and Summer 1978; Gettig and McCarthy 1982; McIntosh and Ignoffo 1983; Susan 1983). The REN pattern of these ten HaNPV DNA varied from one another and the fragment size was expressed in Dalton in these isolates. For comparing the genome of HaNPV isolates, the restriction fragments size were converted into kbp from Dalton using the standard value 660 Dalton/bp. Schematic representation of EcoRI REN pattern of DNA of all HaNPV isolates were drawn, keeping l DNA digested with HindIII as marker. From the schematic representation, a number of shared fragments between two isolates were identified. Normally identical fragments between two isolates are called as shared fragments. The shared fragments between the isolates were identified as follows:

(i) EcoRI REN patterns of HaNPV isolate were studied by five different authors. Gettig and McCarthy (1982) observed the REN pattern for six different isolates under identical electrophoretic conditions. The remaining four REN pattern were observed by four different authors independently. The electrophoretic conditions, gel length and gel percentage varied in the four independent analysis. Since the REN pattern was observed in different electrophoretic conditions, it was necessary to narrow down the difference in the electrophoretic separation of fragment to compare the REN pattern of isolates. Relaxation of 200 bp was given to fix the shared fragments i.e., the fragments were considered identical even if the size was 200 bp plus or minus. This relaxation was necessary to compare the independently resolved DNA of isolates and such relaxation was not given for the comparison of the isolates that were resolved in the same gel.

(ii) Fragments smaller than 1·4 kbp were not considered for comparison because in two isolates, smaller fragments (> 1·4 kbp) were not detected. Comparison of smaller fragments in the other isolates may introduce error in computing P value. P value reveals the similarities between the isolates while generating a simple distance matrix. Higher P value indicates lower similarity between the respective isolates and lower P value means higher similarity between them. Zero P value indicates identical REN pattern of the isolates. Comparing the number of shared fragments of two isolates, the sequence divergence value was calculated following Nei and Li (1979, equation 20 and figure 1). Likewise, each isolate was compared with all other isolates. A simple distance matrix of all P values were obtained and using Kitsch pro-gram (Felsenstein 1986) a dendrogram was constructed showing the relatedness among isolates.

3. Results and discussion

3.1 Scanning electron microscopical studies of PIBs

Majority of the PIBs of H. armigera MNPV (present isolate) are spherical while some of them are irregular in shape (figure 1A). The size ranged from 0·6 µm to 2·3 µm averaging to 1·35 µm. Jacob and Subramanian (1972) have reported that the size of PIBs of HaSNPV (north India isolate) ranged from 0·5 to 2·5 µm.

Tables 1 and 2 present the size of PIBs measured so far including the present observation. The mean size was 1·58 ±  0·39 µm for MNPV (n = 23) and 1·02 ±  0·40 µm for SNPV (n = 16). Critical analysis of tables 1 and 2 reveal that 87% of MNPV polyhedra are above 1·3 µm and 81·3% of SNPV PIBs are below 1·3 µm in size. Leaving few exceptions, it is possible to determine the nature of polyhedra without cross-section based on the size of PIBs. If the mean size of PIBs exceed 1·3 µm, the NPV is likely to be a multi embedded form.

3.2 TEM of PE

Figure 1B shows the negatively stained PIB of HaNPV. At high magnification the negatively stained PIB samples precisely revealed the PE. It is loosely bound around the polyhedron (figure 1C). After the addition of 0·01 M Na₂CO₃ solution, the polyhedra lost its shape immediately and wriggles appeared on its surface (figure 2B and C). The intact PE obtained (figure 2D) after 4 days of incubation in the alkali appeared like a collapsed folded bag.

3.3 Cross-section of PIBs

Figure 3 shows the cross-section of polyhedron of HaMNPV. In the boundary of polyhedron, the PE appears as an electron dense layer. The distance between the PE and polyhedral crystalline matrix is not uniform around the polyhedron. The cross-section clearly shows the multi embedded nature of the HaNPV isolated from Madurai. In fact, this is the first report on the occurrence of multi embedded form of HaNPV in India. The cross-section of the PIB is also helpful to calculate the total number of virus particle per polyhedron (Allaway 1983). Following the protocol of Allaway (1983), the polyhedra volume and number of virion per polyhedron were calculated. The mean volume of PIB is 1·29 ±  0·09 µm³ and number of nucleocapsid per polyhedron is 113 ±  7·4.

3.4 REN pattern and comparative analysis of HaNPV genome

HaMNPV DNA (Madurai isolate) was digested with EcoRI, end labelled and resolved in 0·9% agarose gel (figure 4; lane 2). Smaller fragments were resolved in 8% acrylamide gel (figure 4; lane 4). EcoRI restriction of HaMNPV genome generated 27 fragments and the size ranged from 15·5 to 0·23 kbp. The total size of the genome was 115·85 kbp.

Different authors have reported EcoRI REN pattern for number of HaNPV genome isolates. The first report of the REN pattern was published in 1978 by Smith and Summers (1978) and they have resolved up to 1·4 kbp fragments. The next report of the REN pattern of 6 different isolates of HaNPV was reported by Gettig and McCarthy (1982) and they have resolved up to 0·65 kbp. In 1983, Susan resolved up to a smallest fragment of 0·57 kbp for an isolate from Newzeland; however McIntosh and Ignoffo (1983) resolved the fragment only up to 1·1 kbp. In the present study, the restriction fragments were identified by using agarose gel, acry-
lamide gel electrophoresis and end labelling techniques. These techniques are sensitive and hence the fragments up to 0·26 kbp are resolved clearly.

Figure 5 shows schematic representation of EcoRI REN pattern of HaNPV isolates reported so far. Using the formula derived by Nei and Li (1979), the sequence divergence (P value) was calculated (considering the REN pattern reported in figure 5). The P values are presented in table 3 in the form of simple distance matrix. From the matrix data and schematic representation, the following points can be discussed.

(i) In the schematic representation, lanes B to E are for the multi embedded NPVs and the remaining five lanes are for single embedded NPVs. All the MNPV isolates have one or more fragments above 15·5 kbp; whereas, none of the SNPV isolates have fragments larger than 15·0 kbp. It may be due to total genome size i.e., MNPV genome is larger than that of SNPVs (figure 5).

(ii) From the schematic representation, it is observed that none of the two isolates have identical REN pattern. The variations among the isolates could be mainly due to any one or more of the following: (a) insertion of host DNA into viral genome, (b) duplication of viral genome, (c) point mutation (Brown et al 1985), and (d) deletion of a part of viral sequence (Carstens 1982; Kumar and Miller 1987; O’Reilly and Miller 1990; Kool et al 1991; Ohkawa et al 1994; Hawtin et al 1995; Heldens et al 1996).

(iii) The variations in the REN pattern among the SNPV are comparatively low in comparison with MNPV. For example, the maximum divergence among the isolates of SNPV was 1·76 P value, while it was 2·2 among the MNPVs. Morethan 50% individual have the P value of above 0·1 in MNPV whereas 80% individuals have the P value below 0·1 in SNPV.

We thank Dr S J Singh (Indian Institute of Horticultural Research, Bangalore), Ms S Shanthi (Indian Institute of Technology, Madras) for their help in the EM studies. SS is grateful to the Council of Scientific and Industrial Research, New Delhi and SM to the Department of Biotechnology, New Delhi for financial support.

References

MS received 14 January 1999; accepted 13 May 1999

Corresponding editor: Seyed E Hasnain