Biological Control 174 (2022) 105008

Contents lists available at ScienceDirect

Biological Control
journal homepage: www.elsevier.com/locate/ybcon

Properties of nucleopolyhedrovirus occlusion bodies from living and

virus-killed larvae of Spodoptera frugiperda (Lepidoptera: Noctuidae)
Esbeidy A. Velasco a, b, Cindy S. Molina-Ruíz a, Juan S. Gómez-Díaz a, Trevor Williams a, *
a
Instituto de Ecología AC, Xalapa, Veracruz, Mexico
b
Facultad de Ciencias Biológicas, Universidad Veracruzana, Xalapa, Veracruz, Mexico

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• OB production increased over time.

• Prevalence of immature OBs decreased
markedly over time.
• No differences in viral genome content
of OBs.
• OB pathogencity correlated with preva­
lence of mature OBs.

A R T I C L E I N F O A B S T R A C T

Keywords: Previous studies report that the pathogenicity of nucleopolyhedrovirus occlusion bodies (OBs) is reduced if OBs
SfMNPV are collected from living infected insects compared to virus-killed insects. We compared the production and
OB production characteristics of SfMNPV OBs collected from living and virus-killed fourth instars of Spodoptera frugiperda at 3–4
OB size
days, 5 days and 6–7 days post-inoculation. Cumulative virus-induced mortality increased from 10.5 % at 3 days
OB maturation
Pathogenicity
to 99.6 % at 7 days post-inoculation. The production of OBs/mg larval weight increased significantly between the
Genome copies first and second samples, but did not increase thereafter. The total number of OBs produced in living and virus-
Biological insecticide killed larvae increased at each sample time. The size (cross-sectional area) of OBs increased significantly during
Fall armyworm the experiment, probably due to OB maturation. The median prevalence of immature OBs decreased over time
from 25 to 44 % in the first samples to 9 % in the final sample. OB pathogenicity was compared in second instars
that consumed a discriminating concentration of inoculum (2 × 104 OB/ml). Virus-induced mortality increased
markedly in larvae that consumed OBs collected later (6–7 days) compared to OBs from earlier samples, but was
similar for OBs from living or virus-killed insects at each time point. We conclude that the pathogenicity of OB
samples was sensitive to the prevalence of immature OBs rather than the viral genome content of OBs, which did
not differ over sample times or in OBs from living or virus-killed insects. These findings can inform decisions on
the production of biological insecticides and laboratory studies on the insecticidal properties of OBs.

* Corresponding author at: Instituto de Ecología AC (INECOL), Carretera antigua a Coatepec 351, Xalapa, Veracruz 91073, Mexico.
E-mail address: [email protected] (T. Williams).

https://doi.org/10.1016/j.biocontrol.2022.105008
Received 21 June 2022; Received in revised form 27 July 2022; Accepted 31 July 2022
Available online 2 August 2022
1049-9644/© 2022 Elsevier Inc. All rights reserved.
E.A. Velasco et al. Biological Control 174 (2022) 105008

1. Introduction (McCormick, Mexico City) using the droplet feeding method (Hughes
et al., 1986). Larvae that consumed the suspension within 10 min were
The efficient production of baculoviruses is critical to their com­ reared individually at 27 ± 0.5 ◦ C until death and then homogenized in
mercial development as the basis for biological insecticides (Lacey et al., sterile water and filtered through a stainless steel mesh (80 µm pore size)
2015). Advances in cell culture techniques have stimulated interest in to remove debris. The resulting OB suspension was counted in triplicate
the production of baculovirus insecticides in bioreactors, although this using a Neubauer chamber under a phase contrast microscope and
technology faces a variety of technical constraints related to the yield, stored at 4 ◦ C prior to use in experiments.
production costs and stability of virus strains in cell culture (Claus et al., All experimental procedures described in this study were performed
2012; Reid et al., 2016). As a result, current methods for the production under controlled laboratory conditions at 25 ± 1 ◦ C, 70 ± 10 % relative
of virus occlusion bodies (OBs) require the inoculation, rearing and humidity and 14:10 h L:D photoperiod.
harvesting of large numbers of host larvae (Grzywacz and Moore 2017).
The quantity and insecticidal activity of OBs produced in vivo depend on 2.2. Inoculation and sampling of infected insects
a range of variables, including genetic factors such as host and virus
strains, environmental factors such as temperature and numerous Third instars of S. frugiperda were starved overnight and those that
operational factors such as inoculum dose, inoculation method, insect had molted to the fourth instar were selected and inoculated with a
stage, diet, rearing densities and harvesting of infected individuals suspension of 2.1 × 108 OBs/ml using the droplet feeding method. This
(Hunter-Fujita et al., 1998; Shapiro 1986). Optimization of these aspects concentration was expected to infect 99 % of experimental insects based
of OB production are necessary as OB production is the most costly and on the results of preliminary studies, in order to minimize the presence
time-consuming aspect of virus-based insecticide manufacturing of uninfected larvae in samples of experimental insects. Control larvae
(Grzywacz et al., 2014). were inoculated with sucrose solution and food coloring without OBs.
A few studies have reported that harvesting of nucleopolyhedrovirus Groups of 24 inoculated larvae were placed in one of three 24-well tissue
(Baculoviridae: Alphabaculovirus) OBs prior to death of the host results in culture plates containing a piece of diet and were incubated in darkness
OBs with reduced insecticidal activity, usually measured in terms of at 25 ± 0.5 ◦ C in a laboratory incubator (Heratherm, Thermo Scientific,
dose-mortality metrics. Compared to OBs from virus-killed insects, OBs USA). Each plate was then assigned at random to one of three sampling
from living infected larvae were twofold less pathogenic for Spodoptera regimes in which samples were taken at: (i) 3 and 4 days post-
litura nucleopolyhedrovirus (SpliNPV) (Takatsuka et al., 2007), or inoculation, (ii) 5 days post-inoculation, (iii) 6 and 7 days post-
sevenfold less pathogenic for Lymantria dispar multiple nucleopolyhe­ inoculation. At each sample time, larvae were classified as living or
drovirus (LdMNPV) (Shapiro and Bell 1981) or Helicoverpa zea nucle­ dead (unresponsive to the touch of a toothpick), counted, transferred to
opolyhedrovirus (HzNPV) (Ignoffo and Shapiro 1978), whereas no 1.5 ml microcentrifuge tubes (one tube for living larvae, another for
significant difference was observed in the pathogenicity of Spodoptera virus-killed larvae), weighed on a precision balance (Explorer EX124,
exigua multiple nucleopolyhedrovirus OBs from living or dead larvae Ohaus, USA) and frozen at − 20 ◦ C. It is important to note that all tissue
(Smits and Vlak 1988). The production of OBs with reduced insecticidal culture plates were checked on each day of the experiment and virus-
activity in living larvae is an issue of interest as a fraction of the infected killed larvae that had died prior to the assigned sample time were
insects are often harvested prior to death, in order to minimize loss of removed and discarded, so that samples taken at 3–4 days, 5 days or 6–7
OBs due to post-mortem liquefaction of larval corpses (Cherry et al., days post-inoculation comprised only larvae that died (or were alive) on
1997; Grzywacz et al., 1998). The reasons that OBs from living larvae those sample dates and did not include larvae that died at earlier times
are less pathogenic than OBs from virus-killed insects is unclear but may (Fig. 1). This procedure was employed to avoid including insects that
be related to the reduced size or virion content of OBs from living died at early times in samples taken at later sample times. The daily
infected larvae (Behle 2018; Takatsuka et al., 2007). observation of larvae also allowed dead larvae to be collected prior to
The Spodoptera frugiperda multiple nucleopolyhedrovirus liquefaction. The procedure was performed on ten occasions using ten
(SfMNPV) has been used as the basis for biological insecticide prepa­ different batches of insects that represented ten independent experi­
rations against larvae of the fall armyworm, Spodoptera frugiperda mental replicates. Sample sizes varied depending on the number of
(Lepidoptera: Noctuidae) (Barrera et al., 2011; Behle and Popham 2012; virus-killed or living infected insects at each time point (see Results
Cruz et al., 1997; García-Banderas et al., 2020; Haase et al., 2015) that section 3.1).
has spread across Africa, Asia and Oceania (Tay et al., 2022). In the To produce OB suspensions all the larvae from each replicate were
present study we compared the physical and biological characteristics of thawed and transferred to a glass homogenizer. Each microcentrifuge
OBs collected from living and virus-killed larvae of S. frugiperda at tube was washed with 300 µl of water to remove any insect remains and
different intervals post-inoculation, with the aim of identifying changes these washings were transferred to the homogenizer. Larvae were ho­
in OB characteristics during the development of polyhedrosis disease. mogenized in 1.5 ml MilliQ water and the resulting suspension was
These findings are likely to inform decisions on the timing of harvesting filtered through a stainless steel mesh (80 µm pore size) to remove
of virus infected insects during commercial OB production and labora­ debris. Each OB suspension was then centrifuged at 2800 g for 15 mins
tory studies on the insecticidal properties of nucleopolyhedrovirus OBs. and the resulting pellet was resuspended in 1 ml MilliQ water. To
quantify OBs, each sample was diluted in Milli-Q water and triplicate
2. Methods samples were counted in a Neubauer counting chamber at x400
magnification. The OB suspensions were then stored at − 20 ◦ C until
2.1. Insects and virus required.

The laboratory colony of S. frugiperda was started in 2016 from 2.3. Determination of OB size and maturity
larvae collected in maize fields close to Xalapa, Veracruz, Mexico.
Larvae were reared at 26 ± 1 ◦ C on a semi-synthetic diet based on For OB size measurements, 100 µl of OB suspension from each
soybean flour, wheatgerm, yeast, agar, and vitamins (Hunter-Fujita sample was pooled among all ten replicates to produce 1 ml of OB sus­
et al., 1998). For oviposition, groups of adults were placed in paper bags pension that was adjusted to a concentration of 5 × 107 OBs/ml in MilliQ
with continuous access to 10 % sucrose solution on a cotton pad. water. Each suspension represented a pooled sample of between 38 and
A Nicaraguan wild-type isolate of SfMNPV (Simón et al., 2011) was 125 larvae (see Results section 3.1). Each suspension was mixed thor­
amplified by inoculating S. frugiperda fourth instars with a suspension of oughly on a vortex mixer and 20 µl droplets were pipetted on to an
1 × 108 OBs in 10 % sucrose and 0.1 % food coloring solution aluminum stub and allowed to dry. Each sample was then coated with

2
E.A. Velasco et al. Biological Control 174 (2022) 105008

Fig. 1. Experimental design and sampling regime. Larvae of Spodoptera frugiperda were inoculated and incubated prior to sampling at 3–4 days, 5 days or 6–7 days
post-inoculation. Larvae were classified as living or virus-killed at each sample time point. Occlusion bodies (OBs) were harvested and subjected to different analyses
as indicated. A single living larva at 7 days post-inoculation was not included in the study.

gold–palladium in a Quorum Q150R sputter-coater and observed at sample with a negative control were included in each analysis and the
x10,000 magnification in a FEI Quanta 250 FEG scanning electron mi­ standard curve (range 5 × 10-3 ng to 5 × 10-7 ng viral DNA) was per­
croscope (SEM) at an accelerating voltage of 10 kV. Images of OBs formed in triplicate to determine the efficiency of each reaction. The
(~100 OBs/image) were taken at 5–7 different points within each qPCR protocol consisted of an initial denaturation step at 95 ◦ C for 3 min
sample at a fixed resolution of 1536 × 1103 pixels, 71 dpi. The cross- followed by 45 amplification cycles of 95 ◦ C for 15 s and 60 ◦ C for 30 s
sectional area of an average (±SE) of 438 ± 73 OBs from living and and, to determine the melting curve, a cycle of 30 s with increments of
dead larvae was then estimated with reference to the 20 µm scale bar 0.5 ◦ C between 60 ◦ C and 95 ◦ C. Data acquisition and processing was
using the ImageJ program (https://imagej.net) as described previously performed using MxPro software (Stratagene, La Jolla, USA).
(Ramírez-Arias et al., 2019). The prevalence of mature and immature
OBs was calculated by counting each type of OB in 3–5 photomicrograph 2.5. Insecticidal activity of OBs
images according to the presence or absence of the polyhedron enve­
lope. Mature OBs have a smooth sealed exterior surface whereas The mortality response of S. frugiperda larvae to a discriminating
immature OBs have an obvious pitted and irregular surface and lack the concentration of OBs was determined using pooled OB samples from the
polyhedron envelope (Sajjan and Hinchigeri 2016). ten replicates of each treatment. For this, groups of 24 larvae were
starved overnight as they molted to the second instar and, using the
2.4. Quantification of viral genomes droplet feeding method, they were then inoculated with 2 × 104 OBs/ml
that was previous determined to be the 50 % lethal concentration (LC50)
Samples of 2 × 108 OBs in 300 µl sterile water were obtained for all in second instars from our laboratory colony (Ramírez-Arias et al.,
treatments and replicates. Virions were released from the OBs by adding 2019). Control larvae consumed sucrose and food coloring solution
100 µl of 0.5 M Na2CO3 and 100 µl of 10 % SDS for 10 min at 55 ◦ C. The without OBs. Insects that drank the suspension in 10 mins were indi­
pH was adjusted by addition of 60 µl 0.1 M HCl. Debris was pelleted by vidually transferred to the wells of a tissue culture plate with diet,
centrifugation at 4200 g for 5 min and was discarded, whereas the maintained in darkness at 25 ± 0.5 ◦ C and checked daily until death or
virion-containing supernatant was incubated with 12.5 µl proteinase K pupation. The bioassay was performed on three occasions using different
(20 mg/ml) at 55 ◦ C for one hour. Each sample was then treated with 20 batches of insects.
µl RNase solution, followed by 200 µl 96 % ethanol and cell lysis solution
(PureLink Genomic DNA kit, Invitrogen, Waltham, MA). The sample was 2.6. Statistical analyses
then loaded on to a spin column and centrifuged at 10,000 g, washed
twice and DNA was eluted in 50 µl of elution buffer following the The experiment was originally designed as a two-way ANOVA with
manufacturer’s protocol. The concentration of DNA in each sample was infection status (living infected insects vs virus-killed insects) and
then determined by calculating the average reading from triplicate sample time as factors. However, only one larva was alive in the 6–7 day
samples in a spectrophotometer at 260 nm wavelength (BioSpec-Nano, sample which resulted in an unbalanced design. Consequently, the
Shimadzu, Japan). remaining treatment combinations were treated as five separate groups
Quantitative PCR (qPCR) based on SYBR fluorescence was performed and subjected to Kruskal-Wallis analysis followed by Dwass-Steel-
in 96-well plates in a Mx3005P qPCR System (Stratagene, La Jolla, USA). Critchlow-Fligner (DSCF) pairwise comparisons (Hollander and
The primers were designed based on the polyhedrin gene sequence and Sethuraman 2015). The results of these analyses are shown as median
had been used in a previous qPCR study (Ramírez-Arías et al., 2019). values and the corresponding asymmetrical interquartile range (IQR).
The forward primer (5′ -GAACCTTCACTCTGAGTACACGCAC) and All analyses were performed using the R-based software Jamovi v.2.3.0
reverse primer (5′ -AGACGATGGGTTTGTAGAAGTTCTCC) amplified an (Jamovi 2022). OB pathogenicity was determined by analyzing larvae
82 bp fragment of the SfMNPV polyhedrin gene. Amplifications were mortality in bioassays by fitting a generalized linear model (GLM) with a
performed in a volume of 10 μl, comprising 5 μl of iQ SYBR Green binomial error distribution in GLIM 4 (Numerical Algorithms Group
Supermix (Bio-Rad), 3.6 μl of sterile Milli-Q water, 0.2 μl of each of the 1993). The SE values of binomially distributed mortality data are
primers (forward and reverse, 10 mM) and 1 µl of a 1/1000 dilution of asymmetrical. There was no evidence of overdispersion in the mortality
DNA (between 5 × 10-4 and 5 × 10-5 ng). Three reactions for each data.

3
E.A. Velasco et al. Biological Control 174 (2022) 105008

3. Results among groups (Kruskal-Wallis H = 34.9; d.f. = 4; P < 0.001). Living

larvae were significantly heavier than virus killed larvae at the same
3.1. Larval mortality time point (Fig. 3A). The median weight of virus-killed larvae increased
by almost fourfold during the experiment.
Following inoculation of larvae, cumulative mortality due to lethal The production of OBs/mg of insect body weight differed signifi­
polyhedrosis disease increased from 10.5 % (asymmetrical upper - lower cantly among experimental groups (H = 29.3; d.f. = 4; P < 0.001). The
range of SE: 1.1–1.2 %) at 3 days post-inoculation to 99.6 % (range of SE:
0.8–0.2 %) at 7 days post-infection (Fig. 2A). A total of 15 larvae (2 %)
died from bacterial infection, confirmed by microscopic examination of
Giemsa smears, which were discarded and were not considered in cu­
mulative mortality calculations. Across all ten replicates, the number of
sampled insects decreased over time due to mortality at earlier time
points (Fig. 2B). A total of 101 virus-killed larvae and 125 living infected
larvae were collected and analyzed in the 3–4 day sample, representing
an average (±SE) of 10.1 ± 1.1 virus-killed insects and 12.5 ± 0.9 living
insects from each replicate. The 5 day sample consisted of 80 virus-killed
larvae (average 8.0 ± 1.1 larvae/replicate) and 38 living larvae (average
3.8 ± 1.1 larvae/replicate). Finally, 44 virus-killed larvae (4.4 ± 0.8
larvae/replicate) and a single living larva comprised the sample taken at
6–7 days. The single living larva in the 6–7 day sample was removed
from the study and is not considered further. No virus induced mortality
was observed in control larvae that were discarded after 7 days post-
inoculation.

3.2. OB production

The median weight of larvae in each sample varied significantly

Fig. 3. Larval weight and production of occlusion bodies (OBs). (A) Median
Fig. 2. Larval mortality and larval sampling.(A) The cumulative mortality of weight of living and virus-killed larvae at each sample time point. (B) Median
larvae during the course of the experiment due to lethal polyhedrosis disease. logarithm of OB production/mg of larval weight. (C) Median logarithm of the
Vertical bars indicate SE. (B) Numbers of living and virus-killed larvae sampled number of OBs produced per larva. Vertical bars indicate the asymetrical
at each time point. Numbers within each column indicate total number of interquartile range. Columns headed by identical letters did not differ signifi­
larvae collected of each type. cantly (Kruskal-Wallis/DSCF pairwise comparisons; P > 0.05).

4
E.A. Velasco et al. Biological Control 174 (2022) 105008

number of OBs/mg larval weight in living infected insects was lower

than measured in virus-killed insects, although this difference was only
significant in the 3–4 day sample as increased variation was observed in
the sample taken at 5 days post-inoculation (Fig. 3B). The difference
between the OBs/mg values at 3–4 days and 5 days was borderline
significant in living insects (P = 0.055). In virus-killed larvae OBs/mg
values increased significantly between the first and second samples, but
did not increase in the final sample taken at 6–7 days post-inoculation
(Fig. 3B).
The total number of OBs produced in each virus-killed larva
increased significantly at each sample time (H = 34.9; d.f. = 4; P <
0.001), reflecting the growth of larvae during the infection period
(Fig. 3C). The median number of OBs/larva increased from 9.3 × 107
OBs in larvae killed at 3–4 days post-inoculation to 7.2 × 108 OBs in
larvae killed at 6–7 days post-inoculation. Interestingly, the number of
OBs recovered from living infected larvae was similar to the numbers
from insects that died from polyhedrosis disease at the same time point
(Fig. 3C), suggesting that insect death was not solely dependent on total
OB load.

3.3. Determination of OB size and maturity

The cross-sectional area of OBs varied from 0.111 µm2 to a maximum

of 6.226 µm2 across a total of 2192 OBs measured. The cross-sectional
area varied significantly among the five experimental groups (H =
264.9; d.f. = 4; P < 0.001). The median cross-sectional area of OBs in
virus-killed insects did not differ for samples taken at 3–4 days and 5
days post-inoculation, but increased significantly in OBs sampled at 6–7
days post-inoculation (Fig. 4A). The smallest OBs were collected from
living larvae at 3–4 days post-inoculation but the size of these OBs
increased significantly in the sample taken from living larvae at 5 days
post-inoculation (Fig. 4A).
The prevalence of mature and immature OBs was determined by
examination of scanning electron photomicrographs (Fig. 4B). The me­
dian prevalence of immature OBs collected from virus-killed larvae
decreased steadily from 25.5 % (IQR: 22.6–33.0) in the 3–4 day sample, Fig. 4. Size of OBs harvested from living and virus-killed insects at each sample
to 8.9 % (IQR: 2.4–12.3) in the sample collected at 6–7 days post- time point. (A) Median cross-sectional area of OBs calculated from scanning
inoculation (H = 39.2; d.f. = 4; P < 0.001). The prevalence of imma­ electron photomicrographs. (B) Median percentage of immature OBs present in
ture OBs was markedly higher in living larvae collected at 3–4 days post- samples. Vertical bars indicate the asymetrical interquartile range. Columns
inoculation (44.2 %, IQR: 36.1–55.1) compared to virus-killed insects at headed by identical letters did not differ significantly (Kruskal-Wallis/DSCF
the same time point, but this difference decreased and was not signifi­ pairwise comparisons; P > 0.05).
cant in the sample taken at 5 days post-inoculation (Fig. 4B).
experimental groups (GLM: χ2 = 73.48; d.f. = 4; P < 0.001). Although all
3.4. Quantification of viral genomes insects received the same concentration of inoculum, virus-induced
mortality increased from ~ 3 % in insects that consumed OBs from
The qPCR analysis of DNA extracted from samples of 2 × 108 OBs living and virus-killed insects collected at 3–4 days post-inoculation to
from each replicate resulted in amplification with a mean ± SD Cq value 52 % (range of SE: 6.0–5.9) in insects that consumed OBs collected from
of 22.2 ± 1.7 across all samples and an efficiency of 105–106 %, within virus-killed at 6–7 days post-inoculation, with intermediate values (~16
the acceptable limits for qPCR studies (Taylor et al., 2019). The median %) in the samples from living and virus-killed insects taken at 5 days
copy number varied from 1.92 × 109 copies in the samples of OB from post-inoculation (Fig. 6). No virus-induced mortality was observed in
virus-killed insects collected at 5 days post-inoculation to 2.39 × 109 control larvae.
copies in the OB samples collected from virus-killed larvae collected at
3–4 days post-inoculation (Fig. 5). The qPCR measurements of copy 4. Discussion
number varied considerably within each time point but did not vary
significantly among time points or between OB samples from living and A systematic examination of the production and characteristics of
virus-killed larvae (H = 1.60; d.f. = 4; P = 0.806). To confirm these OBs harvested from living and virus-killed larvae of S. frugiperda at three
findings the qPCR assay was repeated with similar non-significant time points along the cumulative mortality curve resulted in findings
variation among samples and time points (H = 2.07; d.f. = 4; P = that have applications in the production of biological insecticides and in
0.724; data not shown). the study of the insecticidal phenotype of nucleopolyhedrovirus OBs in
the laboratory.
3.5. Insecticidal activity of OBs The study was conducted by inoculating fourth instars with a high
concentration of inoculum (2.1 × 108 OBs/ml) to ensure that virtually
The biological activity of OBs was determined by inoculating all the experimental insects acquired a lethal infection. As a result, it was
S. frugiperda second instars with 2 × 104 OB/ml previously estimated to not necessary to analyze living larvae individually at each sample time
result in 50 % mortality (Fig. 6). Virus-induced mortality varied signif­ to determine their infection status prior to the analysis of their OB load,
icantly among larvae inoculated with OBs from one of the five which would have been very laborious given that we sampled a total of

5
E.A. Velasco et al. Biological Control 174 (2022) 105008

replicate, compared to samples taken at earlier time points. The OB load

per mg of larval weight also increased significantly over time (Fig. 2B).
Studies with other nucleopolyhedroviruses have noted an increase in OB
production per mg larval weight over time (Kumar et al., 2005; Takat­
suka et al., 2007), and a 4-fold to 6-fold higher production of OBs per mg
larval weight in virus killed larvae compared to living infected insects in
the case of the H. zea/HzNPV pathosystem (Ignoffo and Shapiro 1978).
Scanning electron microscope examination revealed that OBs varied
widely in size (range 0.111–6.226 µm2 cross-sectional area). The size of
OBs tended to increase at later sampling times and the prevalence of
immature OBs decreased markedly. Indeed, the decrease in the preva­
lence of immature OBs over time was evident by simple visual inspection
of electron microscope images. Interestingly, samples of SeMNPV OBs
from living larva also contained many small OBs (Smits 1987) and an
increase in the size of SpliNPV OBs at successive sample times was also
reported by Takatsuka et al. (2007). In the present study, the increase in
OB size at later sample times was likely due to the OB maturation pro­
cess in which immature OBs acquire a smooth sealed outer envelope of
multilayered polyhedron envelope protein (PEP), in a process that is
Fig. 5. Logarithm of SfMNPV genome copy number calculated by qPCR facilitated by a vermiform complex of P10 protein and electron dense
amplification of a polh gene fragment on DNA extracted from samples of 2 × spacers in the nucleus of the infected cell (Graves et al., 2019). Various
108 OBs from living insects (pale boxes) and virus-killed larvae (dark boxes) at functions have been assigned to the polyhedron envelope include pre­
each sample time point. Horizontal lines indicate medians, boxes indicate venting the loss of ODVs from the polyhedrin matrix, preventing the
interquartile range, vertical bars indicate range, dots indicate individual data
fusion of OBs during their development in the cell nucleus, and
points from replicates within each range. P value indicates the result of a
increasing the physical stability of OBs in the environment outside of the
Kruskal-Wallis test.
host (Gross et al., 1994; Sajjan and Hinchigeri 2016). As the polyhedron
envelope is sensitive to sodium dodecyl sulfate (Lua et al., 2003), we
avoided the use of this or other detergents during sample preparation,
except when performing DNA extraction.
As OBs differed markedly in size and maturity, we suspected that
they may differ in ODV content. To examine this, we determined the
number of genomes occluded by OBs using qPCR targeted at the poly­
hedrin gene. All the genotypic variants present in the SfMNPV-NIC
isolate are polyhedrin positive (Simón et al., 2004), so we were confi­
dent that amplification based on this gene would provide an accurate
estimate of genome copies in each OB sample. Contrary to our suppo­
sition, the OB samples had a similar viral genome content, irrespective of
the time point or status (living/dead) of the host. This suggests that there
may be a threshold number of ODVs in the nucleus around which the OB
forms and that the size of the OB that develops subsequently does not
reflect the ODV content that it occludes. Previous statements that OB
size is positively correlated with the number of ODVs that each OB oc­
cludes (Takatsuka et al., 2007) finds no support in the cited literature
(Allaway 1983; Mazzone and McCarthy 1981). However, as we did not
employ cell culture assay techniques or transmission electron micro­
Fig. 6. Mean mortality of Spodoptera frugiperda larvae inoculated with a scopy, we cannot rule out the possibility that different OB samples had
discriminating concentration of 2 × 104 OBs/ml from living or virus-killed different numbers of ODVs, for example if the ratio of single nucleo­
insects at each sample time point. The inoculum was expected to result in 50 capsid to multinucleocapsid ODVs changed over time but the number of
% larval mortality. Vertical bars indicate SE. Columns headed by identical genome copies remained stable at different sample times. This seems
letters did not differ significantly (GLM, P > 0.05).
unlikely however, as the distribution of nucleocapsids among ODVs
seems to be a characteristic of each nucleopolyhedrovirus strain (Fuji­
164 living larvae distributed across the ten replicates of the study. In moto et al., 2017; Rohrmann 2014) and appears to be genetically
reality, in large-scale virus production facilities, the inoculum concen­ determined as several genes influence the number of nucleocapsids
tration has to be optimized to produce the highest number of highly packaged within ODVs (Beperet et al., 2013; Li et al., 2014; Yu et al.,
insecticidal OBs from each batch of inoculated larvae. This usually re­ 2009), or the number of ODVs occluded within OBs (Li et al., 2014;
quires a trade-off, as high inoculum concentration will infect a large Simón et al., 2008a; Yang et al., 2014). The identity of the host cell can
fraction of the larvae, but very high inoculum concentration will shorten also influence nucleocapsid aggregation within ODVs (Chaeychomsri
larval survival time and reduce the overall harvest of OBs (Behle 2018; et al., 2018; Xu et al., 2012).
Grzywacz and Moore 2017). OB samples were used to inoculate larvae at a standard discrimina­
In the present study, both living and virus-killed larvae harvested at tory concentration (2 × 104 OB/ml) that was expected to result in 50 %
3–4 days post-inoculation produced fewer OBs than larvae at later time lethal infection in S. frugiperda second instars (Ramírez-Arias et al.,
points, whereas at 5 days and 6–7 days post-inoculation the production 2019). The pathogenicity of OBs from each time point showed marked
of OBs per larva increased significantly in virus-killed insects (Fig. 2C). differences across sample times but did not differ significantly between
Clearly, OB numbers at later time points reflect a combination of OBs from living or virus-killed insects at a given time point (Fig. 6). It
increased host resources due to larval growth during the systemic phase was previously suggested that differences in OB pathogenicity were due
of the infection and the extended period for which the virus can to variation in the number of ODVs occluded (Behle 2018). Our qPCR

6
E.A. Velasco et al. Biological Control 174 (2022) 105008

findings provided no evidence for this, whereas there was a closer cor­ Declaration of Competing Interest
relation between OB pathogenicity and the prevalence of mature OBs in
the inoculum. This suggests that OB maturation is associated with pro­ The authors declare that they have no known competing financial
cesses that influence key aspects of the primary infection process in the interests or personal relationships that could have appeared to influence
insect midgut. One possible mechanism is the incorporation of alkaline the work reported in this paper.
proteases into OBs during maturation. These proteases markedly facili­
tate OB degradation and ODV release in the alkaline conditions of the Data availability
lepidopteran midgut (Mazzone and McCarthy 1981; Rohrmann, 2019).
An OB endogenous alkaline protease is also involved in the activation of Data will be made available on request.
P74 that forms part of the ODV per os infection factor (PIF) complex that
is essential for virus entry into midgut epithelial cells (Peng et al., 2011). Acknowledgements
An additional candidate for maturation-mediated increases in OB
pathogenicity could be an ortholog of the fusolin-like GP37 protein that We thank our INECOL colleagues Greta H. Rosas for electron mi­
is associated with OBs (Vialard et al., 1990), and that has chitin binding croscopy, Adrián J. Enríquez for valuable advice on qPCR studies and
activity that may affect the integrity of the peritrophic matrix through Lorena Hernández for insect rearing.
which ODVs must pass to access midgut cells (Erlandson et al., 2019).
It would be interesting to verify these findings in other lepidopteran- References
nucleopolyhedrovirus pathosystems and begin to explore the reasons
behind such large variation in OB pathogenicity over time. Although the Allaway, G.P., 1983. Virus particle packaging in baculovirus and cytoplasmic
polyhedrosis virus inclusion bodies. J. Invertebr. Pathol. 42 (3), 357–368.
genotypic variants present in the SfMNPV-NIC isolate differ in their Barrera, G., Simón, O., Villamizar, L., Williams, T., Caballero, P., 2011. Spodoptera
individual pathogenicity and speed-of-kill characteristics (Simón et al., frugiperda multiple nucleopolyhedrovirus as a potential biological insecticide:
2005, 2008b), the possibility that OBs from early time points comprise genetic and phenotypic comparison of field isolates from Colombia. Biol. Control 58,
113–120.
different proportions of genotypic variants with a different insecticidal Behle, R.W., 2018. In vivo production of Agrotis ipsilon nucleopolyhedrovirus for
phenotype than OBs produced at later times is considered to be highly quantity and quality. J. Econ. Entomol. 111, 101–107.
unlikely. This is because the high inoculum concentration used would Behle, R.W., Popham, H.J., 2012. Laboratory and field evaluations of the efficacy of a
fast-killing baculovirus isolate from Spodoptera frugiperda. J. Invertebr. Pathol. 109,
have ensured the transmission of most if not all variants to the experi­ 194–200.
mental insects (Clavijo et al., 2010) and once infection has been ach­ Beperet, I., Barrera, G., Simón, O., Williams, T., López-Ferber, M., Gasmi, L., Herrero, S.,
ieved, the relative prevalence of variants within the host does not Caballero, P., 2013. The sf32 unique gene of Spodoptera frugiperda multiple
nucleopolyhedrovirus (SfMNPV) is a non-essential gene that could be involved in
change significantly over the course of each infection (Simón et al.,
nucleocapsid organization in occlusion-derived virions. PLoS One 8, e77683.
2008b). Chaeychomsri, S., Chaeychomsri, W., Siruntawineti, J., Ikeda, M., Kobayashi, M., 2018.
These findings have clear implications both for the commercial Replication and occlusion body formation of Spodoptera exigua multicapsid
production and the study of alphabaculoviruses as the basis for biolog­ nucleopolyhedrovirus in a homologous cell line. J. Adv. Agric. Technol. 5 (3),
236–244.
ical insecticides. First, for mass production processes it is clear that OB Cherry, A.J., Parnell, M.A., Grzywacz, D., Jones, K.A., 1997. The optimization of in vivo
harvesting should be delayed as long as possible before larval liquefi­ nuclear polyhedrosis virus production in Spodoptera exempta (Walker) and
cation seriously reduces the final yield of OBs (Cherry et al., 1997; Spodoptera exigua (Hübner). J. Invertebr. Pathol. 70 (1), 50–58.
Claus, J.D., Gioria, V.V., Miceloud, G.A., Visnovsky, G., 2012. Production of insecticidal
Grzywacz et al., 1998). If possible, OBs should not be harvested from baculoviruses in insect cell cultures: potential and limitations. In: Soloneski, S.,
living infected larvae, but it is usually difficult to separate living insects Larramendy, M. (Eds.), Insecticides – Basic and Other Applications. InTechOpen,
for later harvesting for logistical reasons and the costs associated with London, pp. 127–152.
Clavijo, G., Williams, T., Muñoz, D., Caballero, P., López-Ferber, M., 2010. Mixed
increased handling of insects (Grzywacz and Moore, 2017). Second, in genotype transmission bodies and virions contribute to the maintenance of diversity
laboratories involved in the study and selection of alphabaculovirus in an insect virus. Proc. R. Soc. B: Biol. Sci. 277, 943–951.
isolates with high potential as biological insecticides it is clear that the Cruz, I., Figueiredo, M.L., Valicente, F.H., Oliveira, A.C., 1997. Application rate trials
with a nuclear polyhedrosis virus to control Spodoptera frugiperda (Smith) on maize.
procedures involving collecting OBs from virus-killed insects need to be An. Soc. Entomol. Brasil 26, 145–152.
standardized so that comparisons of OBs from different virus strains are Erlandson, M.A., Toprak, U., Hegedus, D.D., 2019. Role of the peritrophic matrix in
performed at similar points on the cumulative mortality curve. This insect-pathogen interactions. J. Insect Physiol. 117, 103894.
Fujimoto, S., Kokusho, R., Kakemizu, H., Izaku, T., Katsuma, S., Iwashita, Y.,
should ensure that the OB preparations under study have similar pro­
Kawasaki, H., Iwanaga, M., 2017. Characterization of a Bombyx mori
portions of mature and immature OBs and therefore differences in OB nucleopolyhedrovirus variant isolated in Laos. J. Insect Biotechnol. Sericol. 86,
potency will reflect the phenotype of the strains being compared rather 85–94.
than the time at which the OBs were harvested. García-Banderas, D., Tamayo-Mejía, F., Pineda, S., de la Rosa, J.I.F., Lasa, R.,
Chavarrieta-Yáñez, J.M., Gervasio-Rosas, E., Zamora-Avilés, N., Morales, S.I.,
Ramos-Ortiz, S., Valle, J., Martínez-Castillo, A.M., 2020. Biological characterization
Funding of two Spodoptera frugiperda nucleopolyhedrovirus isolates from Mexico and
evaluation of one isolate in a small-scale field trial. Biol. Control 149, 104316.
Graves, L.P., Hughes, L.C., Irons, S.L., Possee, R.D., King, L.A., 2019. In cultured cells the
This research did not receive any specific grant from funding baculovirus P10 protein forms two independent intracellular structures that play
agencies in the public, commercial, or not-for-profit sectors. EAV separate roles in occlusion body maturation and their release by nuclear
received an Ayudante de Investigador nivel III scholarship from CON­ disintegration. PLoS Pathog. 15 (6), e1007827.
Gross, C.H., Russell, R.L., Rohrmann, G.F., 1994. Orgyia pseudotsugata baculovirus p10
ACYT, Mexico. and polyhedron envelope protein genes: analysis of their relative expression levels
and role in polyhedron structure. J. Gen. Virol. 75, 1115–1123.
CRediT authorship contribution statement Grzywacz, D., Moore, S., 2017. Production, formulation, and bioassay of baculoviruses
for pest control. In: Lacey, L.A. (Ed.), Microbial Control of Insect and Mite Pests.
Academic Press, San Diego, CA, pp. 109–124.
Esbeidy A. Velasco: Investigation, Methodology, Data curation, Grzywacz, D., Jones, K.A., Moawad, G., Cherry, A., 1998. The in vivo production of
Formal analysis, Writing – original draft. Cindy S. Molina-Ruíz: Spodoptera littoralis nuclear polyhedrosis virus. J. Virol. Meth. 71, 115–122.
Grzywacz, D., Moore, D., Rabindra, R.J., 2014. Mass production of entomopathogens in
Investigation, Data curation, Writing – review & editing. Juan S.
less industrialized countries. In: Morales-Ramos, J.A., Rojas, M.G., Shapiro-Ilan, D.I.
Gómez-Díaz: Investigation, Resources. Trevor Williams: Conceptuali­ (Eds.), Mass Production of Beneficial Organisms. Elsevier, Amsterdam, Netherlands,
zation, Methodology, Formal analysis, Writing – original draft, Writing – pp. 519–553.
review & editing, Supervision, Visualization, Funding acquisition. Haase, S., Sciocco-Cap, A., Romanowski, V., 2015. Baculovirus insecticides in Latin
America: historical overview, current status and future perspectives. Viruses 7,
2230–2267.

7
E.A. Velasco et al. Biological Control 174 (2022) 105008

Hollander, M., Sethuraman, J., 2015. Nonparametric statistics: rank-based methods. In: Shapiro, M., 1986. In vivo production of baculoviruses. In: Granados, R.R., Federici, B.A.
Wright, J.D. (Ed.), International Encyclopedia of the Social and Behavioral Sciences, (Eds.), The Biology of Baculoviruses, vol. II. CRC Press, Boca Raton, FL, pp. 31–62.
second ed. Elsevier, Oxford, UK, pp. 891–897. Simón, O., Williams, T., López-Ferber, M., Caballero, P., 2004. Genetic structure of a
Hughes, P.V., Van Beek, N.A.M., Wood, H.A., 1986. A modified droplet feeding method Spodoptera frugiperda nucleopolyhedrovirus population: high prevalence of
for rapid assay of Bacillus thuringiensis and baculoviruses in noctuid larvae. deletion genotypes. Appl. Environ. Microbiol. 70, 5579–5588.
J. Invertebr. Pathol. 48 (2), 187–192. Simón, O., Williams, T., López-Ferber, M., Caballero, P., 2005. Functional importance of
Hunter-Fuijita, F.R., Entwistle, P.F., Evans, H.F., Crook, N.E., 1998. Insect Viruses and deletion mutant genotypes in an insect nucleopolyhedrovirus population. Appl.
Pest Management. Wiley, Chichester, UK, p. 620. Environ. Microbiol. 71, 4254–4262.
Ignoffo, C.M., Shapiro, M., 1978. Characteristics of baculovirus preparations processed Simón, O., Williams, T., Cabrera-Asensio, A., Ros, S., Gaya, A., Caballero, P., Possee, R.
from living and dead larvae. J. Econ. Entomol. 71 (2), 186–188. D., 2008a. Sf29 gene of Spodoptera frugiperda multiple nucleopolyhedrovirus is a
Jamovi, 2022. Jamovi version 2.3.0. Available from: https://www.jamovi.org. viral factor that determines the number of virions in occlusion bodies. J. Virol. 82,
Kumar, C.M.S., Sathiah, N., Rabindra, R.J., 2005. Optimizing the time of harvest of 7897–7904.
nucleopolyhedrovirus infected Spodoptera litura (Fabricius) larvae under in vivo Simón, O., Williams, T., López-Ferber, M., Taulemesse, J.M., Caballero, P., 2008b.
production systems. Curr. Sci. 88, 1682–1684. Population genetic structure determines the speed of kill and occlusion body
Lacey, L.A., Grzywacz, D., Shapiro-Ilan, D.I., Frutos, R., Brownbridge, M., Goettel, M.S., prodction in Spodoptera frugiperda multiple nucleopolyhedrovirus. Biol. Control 44,
2015. Insect pathogens as biological control agents: back to the future. J. Invertebr. 321–330.
Pathol. 132, 1–41. Simón, O., Palma, L., Beperet, I., Muñoz, D., López-Ferber, M., Caballero, P., Williams, T.,
Li, S.N., Wang, J.Y., Yuan, M.J., Yang, K., 2014. Disruption of the baculovirus core gene 2011. Sequence comparison between three geographically distinct Spodoptera
ac78 results in decreased production of multiple nucleocapsid-enveloped occlusion- frugiperda multiple nucleopolyhedrovirus isolates: detecting positively selected
derived virions and the failure of primary infection in vivo. Virus Res. 191, 70–82. genes. J. Invertebr. Pathol. 107, 33–42.
Lua, L.H., Nielsen, L.K., Reid, S., 2003. Sensitivity of Helicoverpa armigera Smits, P.H., 1987. Nuclear polyhedrosis virus as biological control agent of Spodoptera
nucleopolyhedrovirus polyhedra to sodium dodecyl sulfate. Biol. Control 26, 57–67. exigua. Wageningen Agricultural University, Netherlands, p. 127 pp.. PhD thesis.
Mazzone, H.M., McCarthy, W.J., 1981. Gypsy moth nuclear polyhedrosis virus: Smits, P.H., Vlak, J.M., 1988. Quantitative and qualitative aspects in the production of a
biochemistry and biophysics. In: Doane, C.C., McManus, M.L. (Eds.), The Gypsy nuclear polyhedrosis virus in Spodoptera exigua larvae. Ann. Appl. Biol. 112,
Moth: Research Toward Integrated Pest Management, Technical Bulletin 1584. US 249–257.
Department of Agriculture, Washington, DC, pp. 487–495. Takatsuka, J., Okuno, S., Ishii, T., Nakai, M., Kunimi, Y., 2007. Productivity and quality
Numerical Algorithms Group, 1993. The GLIM System: Release 4 Manual. Clarendon of polyhedral occlusion bodies of a nucleopolyhedrovirus harvested from Spodoptera
Press, Oxford, UK. litura (Lepidoptera: Noctuidae) larvae. Appl. Entomol. Zool. 42 (1), 21–26.
Peng, K., van Lent, J.W., Vlak, J.M., Hu, Z., van Oers, M.M., 2011. In situ cleavage of Tay, W.T., Rane, R.V., Padovan, A., Walsh, T.K., Elfekih, S., Downes, S., Nam, K.,
baculovirus occlusion-derived virus receptor binding protein P74 in the peroral d’Alençon, E., Zhang, J., Wu, Y., Nègre, N., Kriticos, D.J., Czepak, C., Otim, M.H.,
infectivity complex. J. Virol. 85 (20), 10710–10718. Gordon, K.H.J., 2022. Global population genomic signature of Spodoptera frugiperda
Ramírez-Arias, F.G., Lasa, R., Murillo, R., Navarro-de-la-Fuente, L., Mercado, G., (fall armyworm) supports complex introduction events across the Old World. Comm.
Williams, T., 2019. Post-mortem incubation influences occlusion body production in Biol. 5, 297.
nucleopolyhedrovirus-infected larvae of Spodoptera frugiperda. Biol. Control 135, Taylor, S.C., Nadeau, K., Abbasi, M., Lachance, C., Nguyen, M., Fenrich, J., 2019. The
33–40. ultimate qPCR experiment: producing publication quality, reproducible data the first
Reid, S., Chan, L.C.L., Matindoost, L., Pushparajan, C., Visnovsky, G., 2016. Cell culture time. Trends Biotechnol. 37 (7), 761–774.
for production of insecticidal viruses. In: Glare, T., Moran-Diez, M. (Eds.), Microbial- Vialard, J.E., Yuen, L., Richardson, C.D., 1990. Identification and characterization of a
Based Biopesticides. Methods in Molecular Biology. Humana Press, New York, baculovirus occlusion body glycoprotein which resembles spheroidin, an
pp. 95–117. https://doi.org/10.1007/978-1-4939-6367-6_9. entomopoxvirus protein. J. Virol. 64, 5804–5811.
Rohrmann, G.F., 2014. Baculovirus nucleocapsid aggregation (MNPV vs SNPV): an Xu, Y.P., Gu, L.Z., Lou, Y.H., Cheng, R.L., Xu, H.J., Wang, W.B., Zhang, C.X., 2012.
evolutionary strategy, or a product of replication conditions? Virus Genes 49, A baculovirus isolated from wild silkworm encompasses the host ranges of Bombyx
351–357. mori nucleopolyhedrosis virus and Autographa californica multiple
Rohrmann, G.F., 2019. Baculovirus Molecular Biology [Internet]. 4th edition. National nucleopolyhedrovirus in cultured cells. J. Gen. Virol. 93 (11), 2480–2489.
Center for Biotechnology Information (US), Bethesda, MD. Available from: https:// Yang, M., Wang, S., Yue, X.L., Li, L.L., 2014. Autographa californica multiple
www.ncbi.nlm.nih.gov/books/NBK543458/. nucleopolyhedrovirus orf132 encodes a nucleocapsid-associated protein required for
Sajjan, D.B., Hinchigeri, S.B., 2016. Structural organization of baculovirus occlusion budded-virus and multiply enveloped occlusion-derived virus production. J. Virol.
bodies and protective role of multilayered polyhedron envelope protein. Food 88, 12586–12598.
Environ. Virol. 8, 86–100. Yu, I.L., Bray, D., Lin, Y.C., Lung, O., 2009. Autographa californica multiple
Shapiro, M., Bell, R.A., 1981. Biological activity of Lymantria dispar nucleopolyhedrosis nucleopolyhedrovirus ORF 23 null mutant produces occlusion-derived virions with
virus from living and virus-killed larvae. Ann. Entomol. Soc. Am. 74, 27–28. fewer nucleocapsids. J. Gen. Virol. 90, 1499–1504.