Download Draft protocol: Tephritidae 2010-TPDP

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Community fingerprinting wikipedia , lookup

DNA barcoding wikipedia , lookup

Transcript
Draft protocol: Tephritidae
2010-TPDP-19
Agenda: 13
DRAFT - TPDP
June 11, 2010
Tephritidae - Identification of immature stages of fruit flies of economic importance by
molecular techniques
Note from the IPPC Secretariat: draft for further discussion and guidance by the TPDP.
Issues raised by the author:
“As I have said before I was not able to develop this into various formats like the thrips TPDP. I hope it
provides some useful information for the development of fruit fly diagnostics. I think if we were able to
narrow the taxonomic range of the request it would be more feasible to develop working protocols that can
be evaluated for merit. It would be good to have the molecular TPDP group working with the morphological
group. This would bring synergy into the process”.
Issues raised by the discipline lead:
whether to put in contact the author with the authors who work on Anastrepha and Bactocera, to try to
create interaction and avoid duplication
main protocols considered are of the internal use in the institution and are not published.
few tools to get the goal for different genera.
Perhaps we must think to combine techniques in a key, morphological and molecular as complementary
information.
TPDP needs to give guidance to author on what is possible to consider for a useful DP.
1. Pest information
The majority of fruit fly species in the family Tephritidae are phytophagous and their biology is
closely associated with host plants where gravid females deposit their eggs and the immature stages
feed and develop. The family includes a wide range of feeding strategies regarding host
specialization (e.g., polyphagous, oligophagous, stenophagous) and plant tissue usage (e.g., fruits,
seeds, stems). Tephritid species are present on all continents except Antarctica. Over 300 species
have been recognized as pests because the immature stages cause direct damage to commodities as
they feed. The majority of pest species belong to just five genera (i.e., Anastrepha Schiner,
Rhagoletis Loew, Bactrocera Macquart, Ceratitis MacLeay, and Dacus Fabricius) and two
subfamilies (i.e., Dacinae and Trypetinae). General information on the taxon can be found at
http://www.sel.barc.usda.gov/diptera/tephriti/tephriti.htm.
2. Taxonomic information
The family Tephritidae includes over 4,000 species (and 480 genera) classified into eight
subfamilies. Species lists are reported for each genus in the 1998 Fruit Fly Expert identification
System and Systematic Information database (Edited by F.C. Thompson) and further taxonomic
information can be found at http://www.sel.barc.usda.gov/diptera/tephriti/tephriti.htm.
3. Detection
Tephritid fruit flies have six developmental stages. The stages can be found inside plant tissue,
within soil or around the base of the host, or on plant tissue:
Egg
In plant tissue
Larva (instar 1)
In plant tissue
Larva (instar 2)
In plant tissue
Larva (instar 3)
In plant tissue
Pupa
In soil, in plant tissue, or on plant tissue
Adult
On plant tissue
Draft protocol: Tephritidae
2010-TPDP-19
Agenda: 13
For species of economic importance, the eggs and larvae are located inside either the fruit or tissue
of the commodity. Pupae can be found in association with the plant tissue if pupation occurred
during the transport of the commodity. Pupae can be found inside commodities (e.g., the Bactrocera
oleae in olives). Alternatively, flies can be found pupating in the soil near the host plant. Adults can
be found on the substrates of host plants such as leaves, stems, and fruits, during times of courtship,
feeding, and ovipositing.
Commodities should be screened for evidence of tephritids by looking for oviposition markings and
feeding damage. The size of an immature or mature fly can vary greatly for different species.
In order to rear tephritid fruit flies place the fruit in a box with sand and a paper. Keep fruit
individually into clear plastic cups, sealed with a mesh, with a 5cm bed of white perlite to allow
pupation. This system allows a fast surveillance of infested samples without opening the cups,
avoiding the putative escape of adult flies.
4. Identification
The morphological identification of tephritid fruit flies to the levels of family, genus, and species is
possible using adult material. Sometimes species-level identification is only reliable with sexspecific characters. Characters of the third instar larvae can be used to identify flies to the family- or
genus- level. But these identifications are based on a limited number of descriptions representative
of the family, genera, and species. If these third instar characters can be viewed in the pupal stage,
then diagnostic information can be derived from that developmental stage. Eggs, first instar larvae,
and second instar larvae are not used for identification.
Molecular methods have been proposed for identification of fruit flies. These techniques are not
restricted to a particular life stage because they utilize DNA as the diagnostic characters. Similar to
conventional larval identification tools, these molecular tools were developed using a restricted
number of tephritid taxa, geographic samples/collections, and individuals of a species.
Host-use and geographic characters have also been used in fruit fly diagnosis. These characters are
not always reliable, however, because the host range of many species has not been completely
documented (Aluja and Mangan 2008) and movement of associated commodities can confound the
accuracy of geographic records.
It is important to distinguish the process of fruit fly identification based on species descriptions (i.e.,
tools and keys that include all possible taxa in the analysis) versus the process of identification
based on a “limited universe” (e.g., tools and keys only including pest taxa or only including
species documented from a particular geography or country).
Currently, all molecular and larval morphological tools are based on a limited universe. When
identified using a “limited universe” the final identification should be regarded as tentative (the
identified specimen is consistent with the profile of a taxa based on host-use, geography, molecular
characters, etc.).
4.1 Morphological identification of adult flies to family, genus, and species
Reliable identification of tephritid fruit flies to the rank of species is only possible with adult
material. The identification of some species requires analysis of sex-specific characters. When keys
are not available for particular genera, identification must be performed by comparison with the
species description. If adult material is available, morphological identification is the primary
method of identification.
4.1.1 Preparation of samples
2
Draft protocol: Tephritidae
2010-TPDP-19
Agenda: 13
See White and Elson-Harris (1994) for general guidelines. Some taxa such as Bactrocera require
analysis of color patterns on the thorax or abdomen for species-level identification. Preservatives
such as ethanol can distort or remove these diagnostic characters.
4.1.2 Identification of the Family Tephritidae
Identification of adult fruit flies to the family-level is possible using general keys of the dipteran
families (Triplehorn and Johnson 2005). Korneyev (2000) provides morphological descriptions for
Tephritidae, with respect to other taxa in the superfamily Tephritoidea, and the subfamilies and
tribes within the family.
4.1.3 Identification of the major pest genera
Lucid keys + Interactive Keys of Pest Fruit Flies of the World (http://deltaintkey.com/ffl/www/_wintro.htm)
4.1.4 Identification of Anastrepha species
4.1.5 Identification of Rhagoletis species
4.1.6 Identification of Bactrocera species
Lucid key for Bactrocera
4.1.7 Identification of Ceratitis species
Publications by De Meyer
4.1.8 Identification of Dacus species
4.2 Morphological identification of third instar larvae to family and genus
Identification of immature third instar larvae to Tepritidae and to the major pest genera (under a
limited universe) is possible.
4.2.1 Preparation of samples
Tissue fixation by gently boiling in water about 1–2 minutes is adequate. Let cool. Gradually add
alcohol to 50% and 70%. Store in 90% alcohol till needed. See White and Elson-Harris (1994) for
details.
4.2.2 Identification of the Family Tephritidae
e.g.Anastrepha and Bactrocera DP´s (in process)
4.2.3 Identification of the major pest genera
Although keys have been developed to identify fruit flies to the level of genus, these tools include a
limited number of genera based on geographic regions (e.g., Foote et al. 1993) or pest status (e.g.,
White and Elson-Harris 1994) or on both (e.g., Frias et al. 2006, 2008; present a taxonomic key
based on cephalopharingeal skeleton of third instar larva of genera Anastrepha, Bactrocera,
Ceratitis, Rhagoletis and Toxotrypana; or some larval characters differentiating Rhagoletis species
between Neotropical and Nearctic species).
Therefore, these a priori limitations on taxonomic diversity could produce incorrect diagnoses
when applied to species not included in the genera chosen for the key. A genus-level key was
developed by White and Elson-Harris (1994) using characters on the third instar larvae but were
developed using only the pest genera and a limited number of pestiferous species. Subsequently,
research has enhanced those keys and newer versions are incorporated into on online identification
tool called the Interactive Keys of Pest Fruit Flies of the World (Carroll et al. 2004; http://delta3
Draft protocol: Tephritidae
2010-TPDP-19
Agenda: 13
intkey.com/ffl/www/_wintro.htm). The online tools also focus on pest genera, do not include all
species within genera, and include species diagnostic characters that were represented by limited
sample numbers.
4.3 Molecular assays for identifying all life stages of Tephritidae
Molecular diagnostic techniques can be roughly divided into two general categories based on the
type of molecules being characterized: the first is Protein-based techniques including
immunological methods (e.g., ELISA) and Multi-Locus Enzyme Electrophoresis (MLEE; also
called isozyme analysis) and the second is DNA-based techniques. Currently there are no
published studies using immunological technology for fruit fly identification. The MLEE approach
has been used extensively for population analysis (within-species variation) of various fruit flies
and only rarely used for species identification (Berlocher 1980; Yong 1995; Ochando et al. 2003).
These methods are not considered useful for accurate identification of fruit flies to species or higher
levels.
The DNA-based methods of diagnosis can be further categorized based on the data type they
generate. There are many DNA-based methods being developed for molecular diagnostics but only
the commonly used methods relevant to fruit fly systems are included here. All of these methods
exploit differences in the length and/or base composition of DNA sequences that are derived from
different species or individuals. These data-type categories include differential PCR success, size
difference, hybridization (melting) rates based on temperatures, and DNA sequence characters.
These technologies are not mutually exclusive and some diagnostics utilize multiple data types to
generate unique information.
The first type is based on differential success during PCR (i.e., measure presence versus absence of
a product after a chemical reaction). This includes protocols that use conventional PCR, Real
Time-PCR (RT-PCR), and Loop-Mediated Isothermal Amplification (LAMP). DNA variation
between species (e.g., two species could have different nucleotide sequences or the entire genetic
locus is unique to one of the species) can enable PCR primers in a conventional PCR to recognize
one species and not the other. This results in amplified PCR product (called the amplicon) for the
target species and nothing for the non-target species. RT-PCR and LAMP are based on this
principle but can include additional oligonucleotides (i.e., primers and probes) in the protocol.
These additional oligonucleotides are necessary for generating the amplicon and can increase the
specificity of the diagnostic. Rather than quantify a PCR amplicon by running it through an agarose
matrix with a dye that stains the DNA molecules, it is possible to combine dyes or fluorescent
oligonucleotides in the PCR and measure accumulation of amplicon independent of a matrix. This
approach is used in all RT-PCR techniques.
The second major category includes methods that quantify size differences in DNA as they migrate
through a matrix. For example, DNA fragments can be run through agarose gels (called Agarose
Gel Electrophoresis or AGE) to estimate their sizes. If the size of a target gene is different between
two or more species, then conventional PCR will generate products that can be distinguished using
AGE. If the diagnostic marker (i.e., PCR primer set) targets multiple regions of a genome it is
possible to generate a profile comprised of a pool of PCR products of various sizes. This is
observed in Amplified Fragment Length Polymorphisms (AFLP) and RAPD-PCR (Randomly
Amplified Polymorphic DNA) techniques. A similar approach is used to measure variant forms of
target genes using Simple Sequence Repeats (SSRs; also called microsatellite DNA). The most
commonly used technique for fruit fly diagnostics combines PCR with a post-PCR treatment called
a restriction digestion. Similar to how an oligonucleotide (e.g., primer) can be selective (or specific)
for a target DNA sequence, the restriction digestion utilizes an enzyme (nuclease) that recognizes
specific nucleotide sequences in a DNA molecule and cuts the DNA at the recognition site. Even if
the original PCR amplicons from two species are the same length, a restriction digestion of the
4
Draft protocol: Tephritidae
2010-TPDP-19
Agenda: 13
amplicons can generate two different migration profiles for the species when measured using AGE.
This technique is called PCR-RFLP (Restriction Fragment Length Polymorphism).
The third analysis category includes methods that measure differences between species based on
hybridization rates. Even if a common gene such as the cytochrome oxidase subunit I (COI) locus is
present in two species, these copies typically will not be identical in the two species. Differences in
total nucleotide composition (A-T binding pairs and C-G binding pairs) can be measured as
differences in the melting/binding temperatures required for single stranded DNA molecules to bind
with a second strand. This fact is exploited in the melting peak curve analyses used in some RTPCR methods to differentiate different species. In this method the two DNA strands are from the
same target individual (species) and it is the total AT/CG composition that functions as a surrogate
character (temperature or energy required to manipulate the strands) of the species. It is also
possible to measure binding efficiencies of a single stranded DNA molecule to various complement
strands from different species (hetero-duplex values) to determine which strand (and thereby
taxonomic source) is the best match. This technique is applied in DNA oligonucleotide arrays and
is similar to the principle of base pairing between oligonucleotides and template DNA in a
conventional PCR.
The fourth analysis category can be broadly defined as DNA sequence-based analysis. This
requires that a gene (or genetic locus) that is common to multiple species be sequenced (i.e., the
sequential arrangement of DNA bases – A, C, G, and T, in that copy of the gene is read) from
multiple species. These DNA sequences derived from different species can be compared using
statistics or character (nucleotide base) matches to determine if one species exhibits a unique
genetic type useful for diagnosis. If the selected target gene is common to many species, then this
method enables a single protocol to be used to diagnose a target species from a large taxonomic
universe (i.e., the number/diversity of species included in a diagnostic). DNA barcoding is a term
applied to the use of a standard genetic marker (i.e., a gene common to a large taxonomic universe)
to analyze a “global” database of DNA sequences (i.e., barcodes). The use of the term global is
relative because no gene can be applied as a diagnostic for all organisms. For animals, a section of
the COI gene has been proposed as the official barcode marker. Even though the COI barcode can
be applied to all insects, it is not informative as a diagnostic for all species.
4.3.1 Defining the Taxonomic and Geographic Scope of Fruit Fly Molecular Diagnostic
Protocols
Molecular diagnostic tools developed to aid in the identification of fruit flies require explicit rules
regarding the scope of the tool. All molecular diagnostics are based on an observed correlation
between a marker and a taxonomic lineage or ecological grouping. This enables the marker to
function as a surrogate for the various characters that traditionally define the lineage (e.g., as
defined in the species description used for nomenclature). The technology and protocols used to
identify and describe surrogate molecular markers are commonly included in scientific literature for
molecular diagnostics but the stability of the marker itself is not consistently characterized in these
publications.
The stability of the marker can be measured according to the intended taxonomic and geographic
scope of the tool. For example, a conventional PCR diagnostic developed to distinguish between
two species implicitly requires that only those two species be included in the diagnostic tool. It is
possible that additional species (i.e., a larger taxonomic universe) could share the profile of the
target pest. Consequently the marker may not be taxonomically stable and the scope of the tool
should be described in the protocol. This is also true of geographic sampling. The development of a
diagnostic tool requires screening of many specimens to ensure that the marker is stable for each
species. When the tool is developed using limited sample sizes it is not possible to properly estimate
5
Draft protocol: Tephritidae
2010-TPDP-19
Agenda: 13
the variation of the marker in the species or even the variation of a marker for a geographic
population of that species.
If a tool has not been tested on adequate samples from the species range it is important to stipulate
in the protocol what geographic regions comprise the scope of the tool. The practice of stating that
the diagnostic marker was developed on collections from one region does not provide a positive
statement of how the tool should be applied. The user may not understand the true geographic
distribution of the species. Likewise, the user may not understand the possible taxonomic universe
for a fruit fly diagnostic because this universe can be delimited using insect taxonomy, geography,
and host plant taxonomy.
In the examples considered below, limitations caused by scope are used to evaluate utility of the
published tools. In general, if the overall genetic variation of a species has only been estimated for a
geographically restricted population, then the tool should not be applied outside that region. It is
important to note that variability in a species could be greatest at the geographic extremes of its
current distribution or at a localized area representative of its ancestral range. Without prior
knowledge it is difficult to predict how best to sample the species.
4.3.2 Techniques applied to fruit fly diagnostics
To date published methods used to identify fruit flies include several PCR-based techniques: PCRRFLP (Restriction Fragment Length Polymorphism), Real Time-PCR (RT-PCR), DNA Sequencing
(e.g., DNA barcoding), Amplified Fragment Length Polymorphisms (AFLP), Simple Sequence
Repeats (SSRs; also called microsatellite DNA), RAPD-PCR (Randomly Amplified Polymorphic
DNA), PCR-AGE (Agarose Gel Electrophoresis), Loop-Mediated Isothermal Amplification
(LAMP), and DNA oligonucleotide arrays.
Of these methods, the RAPD, AFLP, oligonucleotide array, and LAMP techniques have been used
rarely to study fruit fly diagnostics (Sonvico et al. 1996; Kakuoli-Duarte et al. 2001; Naeole and
Haymer 2003; Ochando et al. 2003; Huang et al. 2009). These techniques were developed using
very limited taxon and geographic sampling and should be regarded as proof of concept studies for
their respective technologies. These methods require further validation studies to document
repeatability and specificity. In general, they have not been developed into functional diagnostic
protocols for routine diagnostic use.
A PCR-AGE approach to fruit fly diagnostics using length variation in the internal transcribed
spacer 1 region (ITS-1) has been applied to Ceratitis identification (Douglas and Haymer 2001;
Barr et al. 2006). But recent systematic analyses by Virgilio et al. (2008) using the ITS-1 marker
raises questions regarding the applicability of the previously published tools because of multiple
copy number of the locus within fruit flies.
The SSRs and RT-PCR techniques are relatively new approaches to fruit fly diagnostics (Baliraine
at al. 2003; Baliraine et al. 2004; Yu et al. 2004; Yu et al. 2005). In theory, the SSR approach can
provide species-level identifications if alleles are unique and fixed to a particular species or if the
cumulative frequency of alleles can statistically assign a fly to a species. These approaches require
good geographic sampling and currently published studies have analyzed limited geographies of the
fruit fly species. The RT-PCR technique is commonly used for molecular diagnostics of animal and
plant pathogens. It is most appropriate when the number of possible species to identify is small (i.e.,
a limited universe) because the system either requires development of species-specific probes or
uses annealing temperature values (i.e., melting curve analysis) that are prone to homoplasy.
The most commonly published technique for tephritid molecular diagnostics is PCR-RFLP
(Armstrong et al. 1997; Armstrong and Cameron 2000; Muraji and Nakahara 2002; Salazar et al.
6
Draft protocol: Tephritidae
2010-TPDP-19
Agenda: 13
2002; Barr et al. 2006). This approach analyzes a common genetic locus from fruit flies via PCR
and then diagnostic differences are observed after the PCR product is digested with restriction
enzymes that can recognize base differences among species. The RFLP approach does not observe
all of the differences present in a DNA sequence (Barr 2009). An alternative approach that
maximizes the information of a locus is to analyze the DNA sequence of the amplified PCR
product. This technique is called DNA barcoding when a commonly used locus is used for a large
group of taxa. Despite the large number of DNA sequences generated for fruit fly systematic and
population genetic studies, the use of the sequence data for barcode diagnostics has been published
infrequently for fruit flies (Armstrong and Ball 2005; Asokan et al. 2007). Armstrong and Ball
(2005) compared the utility of barcodes to RFLP data for Bactrocera species but did not release the
data or develop it into an operational diagnostic tool. Asokan et al. (2007) has reported barcode
sequences for three Bactrocera species but did not provide sufficient samples sizes to evaluate its
utility. Recent studies of Ceratitis and Dacus suggest that the barcode approach does not work for
all fruit fly species (Virgilio et al. 2008; Virgilio et al. 2009).
Despite the interest in developing molecular diagnostic tools for immature stages all of the
published protocols are developed using samples from a restricted taxonomic and geographic range.
Therefore, application of these techniques requires that the diagnostician explicitly state the
assumptions made in the interpretation of the diagnostic analysis: (1) The specimen to be analyzed
must be one of the species included in the diagnostic tool; (2) The genetic variation of the species is
uniform across the species’ geographic and host range (i.e. no population structure). Given these
assumptions, it is difficult to regard most molecular diagnoses as species-level identifications.
Terminology such as “the immature specimen is consistent with variation documented in the
species” would be more appropriate.
4.3.3 Preparation of immature samples for DNA analysis
It should be noted that samples to be used for morphological analysis prior to DNA analysis may
require boiling steps that can have a negative effect on DNA quality.
The effects of boiling a larva have not been published in peer-reviewed journals but several labs
have successfully analyzed DNA from boiled larvae using standard protocols developed for nonboiled sampels. Several protocols are in use for serial analysis of samples by boiling in water (or in
TNE buffer) the larva specimens at 65-100ºC up to 15 min (15’ at 65ºC in 100ul followed by
adding 100ul more & vortexing, and 5’ at 100ºC in a thermal cycler). The boiled media is
transferred to another vial and stored at 4ºC for short term (up to one week). PCR is performed
directly with 1-5 ul. The specimen is then stored in 70% EtOH for further studies. Assuming
samples are to be used for DNA analysis the specimen should be stored in a preservative such as
>95% non-denatured ethanol or propylene glycol. Freezing specimens reduces the degradation of
DNA but is not necessary if the samples are to be analyzed within a short period (e.g., one month).
Pupae
Larvae
eggs
4.3.4 Molecular Identification of the Family Tephritidae
No molecular tools have been developed for the purpose of identifying flies to the family-level.
Various molecular systematic studies have been conducted at higher taxonomic levels (e.g.,
Tephritoidea) that include tephritid representatives (Han et al. 2002; Han and Ro 2005). But despite
testing relationships of the different dipteran families, these studies provide no clear diagnostic rule
for identification of flies to Tephritidae using DNA.
Recent molecular studies using DNA barcoding technology have accumulated DNA sequence
information from different families within Tephritoidea. Although the technology should be
applicable to the problem of family-level identification, these growing barcode databases have not
7
Draft protocol: Tephritidae
2010-TPDP-19
Agenda: 13
been peer-reviewed for taxonomic accuracy, systematic content, or the process of diagnosis and
interpretation.
4.3.5 Molecular identification of the major pest genera
No general tool has been developed specifically to distinguish among the five pest genera. (No tool
can currently distinguish the large number of genera within Tephritidae.) Representative species of
Anastrepha, Rhagoletis, Ceratitis, and Dacus have been included in PCR-RFLP tools that were
designed to identify various Bactrocera species (Armstrong et al. 1997; Armstrong and Cameron
2000; McKenzie et al. 1999). These non-Bactrocera genera, however, were represented by less than
nine species per genus and no clear pattern for discriminating the genera was reported.
Consequently, interpretation of genus-level diagnosis is not currently possible with PCR-RFLP
techniques.
Molecular phylogenetic studies have generated DNA sequence data sets with the potential to
distinguish among the five pest genera. Extension of a DNA barcode approach to diagnosis of the
pest genera is possible by using a locus that is commonly used for phylogenetic analyses.
Unfortunately, most of the published studies have used different loci: The 3′ region of the COI
gene: Anastrepha (Smith-Caldas et al. 2001; Boykin et al. 2006), Bactrocera (Jamnongluk et al.
2003a,b), Ceratitis (Barr and McPheron 2006; Virgilio et al. 2008); The 5′ end of the COI gene:
Bactrocera (Armstrong and Ball 2005), Dacus (Virgilio et al. 2009), The COII gene: Bactrocera
(Smith et al. 2003), Rhagoletis (Smith and Bush 1997), Various tepritid genera (Han and Ro 2009);
Ribosomal (rRNA) genes: Bactrocera (Muraji and Nakahara 2001), Rhagoletis (McPheron and Han
1997), Anastrepha (McPheron et al. 2000) and various tephritid genera (Han and McPheron 1997;
Han and McPheron 2000, Han and Ro 2009).
Recent DNA barcode programs have selected the 5′ end of the COI gene as a common marker for
future tephritid diagnostic studies. Despite generation of new barcode data, these DNA sequences
have not been evaluated for diagnostic information or peer reviewed. Consequently, there are no
standard, routine molecular tools for distinguishing the five major pest genera.
4.3.6 Molecular identification of Anastrepha species
The genus Anastrepha includes over 200 species and molecular systematic studies have focused on
a very limited number of these species (i.e., primarily the fraterculus group; Smith-Caldas et al.
2001). Molecular diagnostic tools using PCR-RFLP have only analyzed four (Armstrong et al.
1997), six (Armstrong and Cameron 2000) and seven (McKenzie et al. 1999) Anastrepha species.
Evidence for cryptic species within at least two Anastrepha pest species (i.e., A. fraterculus and A.
obliqua; Silva and Barr 2008) raises questions regarding sufficient sampling of the species included
in the PCR-RFLP tools because they were derived from single geographic collection sites. Although
DNA sequences reported by Smith-Caldas et al. (2001) and Boykin et al. (2006) provide the best
published data set for Anastrepha pest identification, these data have not been formally developed
into a diagnostic tool. Consequently, there are no standard, routine molecular tools for
distinguishing among Anastrepha species.
4.3.7 Molecular identification of Rhagoletis species
A PCR-RFLP technique for discriminating among four native Chilean Rhagoletis species (i.e., R.
nova, R. conversa, R. penela, and R. tomatis) has been developed (Salazar et al. 2002). This tool
included multiple collections of the four species but is only appropriate for analyses of flies that
were collected from Chile or Chilean commodities. In addition, Rhagoletis completa and R.
pomonella were included in the PCR-RFLP tool developed by Armstrong et al. (1997). The limited
number of Rhagoletis samples reported in Armstrong et al. (1997) is not representative of the
geographic distribution of these species within North America and will likely underestimate the
intraspecific diversity of the species.
8
Draft protocol: Tephritidae
2010-TPDP-19
Agenda: 13
The only published DNA sequences available for Rhagoletis identification were generated for
phylogenetic study (McPheron and Han 1997). The reported DNA sequences used for phylogenetic
analysis do not include estimations of intraspecific variation across the species range. Consequently,
with the exception of four native Chilean species, there are no standard, routine molecular tools for
distinguishing among Rhagoletis species.
4.3.8 Molecular identification of Bactrocera species
Several techniques for discriminating among select Bactrocera species have been reported using
PCR-RFLP of the 18S+ITS1 locus (Armstrong et al. 1997, Armstrong and Cameron 2000), the
ITS1 and ITS2 loci (McKenzie et al. 1999), and mitochondrial DNA regions including parts of the
ribosomal RNA genes (Nakahara et al. 2000, Nakahara et al. 2001, Nakahara et al. 2002, Muraji
and Nakahara 2002). These protocols include data bases generated from multiple species and
multiple individuals per species for some taxa. The McKenzie et al. (1999) publication is a nonpeer reviewed report by CSIRO Entomology that describes the protocol used by New Zealand to
identify intercepted flies. It describes controls and interpretation procedures for a database of 79
species of flies (62 Bactrocera) but includes diagnostic information that may not be representative
of some of the species because of limited geographic sampling. The Muraji and Nakahara (2002)
peer-reviewed, publication reports an identification tool based on 18 Bactrocera species and 83
individuals. It can discriminate amongst 16 of the species but B. carambolae and B. papayae cannot
be distinguished from each other.
The relative performance of a DNA barcode for Bactrocera, in relation to RFLP methods, was
reported by Armstrong and Ball (2005). This study did not describe the process for performing
identifications and was a proof of concept approach to Bactrocera diagnostics. [Recently, MAF,
New Zealand has been developing this technique into a formal tool – we need to obtain information
on this implementation of barcodes.] Bactrocera DNA sequences that were generated and released
in the Muraji and Nakahara (2001) could be used in lieu of the PCR-RFLP technique, but it has not
been formally developed into a protocol.
Additional identification tools using RT-PCR have been developed to discriminate B. latirfons from
other Bactrocera species (Yu et al. 2004) and to discriminate between B. philippinensis and B.
occipitalis (Yu et al. 2005). These studies report primer sets for each of the three species that should
only recognize the COI sequence of the target species. The RT-PCR is analyzed using melting
peaks based on SYBR Green I binding to the PCR product. The B. latifrons protocol was developed
using sample sizes between one and three and single collection locations for the nine Bactrocera
species. The authors indicate that additional B. latifrons specimens from Thailand and China
(Hongkong and Yunnan) were tested but that additional samples are needed to confirm specificity
of the protocol. The B. philippinensis and B. occipitalis samples were collected from a single
commodity (i.e., mango) at a plant quarantine laboratory of the Shenzhen (China) inspection
station. These tools require further experiments to test specificity and develop positive and negative
controls prior to routine use.
4.3.9 Molecular identification of Ceratitis species
There are many molecular diagnostic protocols that include the pest species Ceratitis capitata. Most
of these are developed to discriminate C. capitata from another Ceratitis species, C. rosa,
(McKenzie et al. 1999; Douglas and Haymer 2001; Kakouli-Duarte et al. 2001) or from more
distant taxa such as Bactrocera species (Armstrong et al. 1997; Armstrong and Cameron 2000;
Huang et al. 2009). Only two published studies include multiple Ceratitis species in the diagnostic
process: A SSR approach that includes C. capitata, C. rosa, C. fasciventris, and C. cosyra
(Baliraine et al. 2003) and a PCR-RFLP approach that includes 25 Ceratitis species (Barr et al.
2006).
9
Draft protocol: Tephritidae
2010-TPDP-19
Agenda: 13
The Baliraine et al. (2003) study identified SSR frequency differences that could help identify the
four different species, but analysis of these differences and the process of interpreting the results
were not explicitly stated in the study. The C. capitata and C. fasciventris individuals used in the
study were derived from a single lab colony and may not be representative of variation in these
species. The SSR technology was applied to the study of C. rosa and C. fasciventris population
genetics (Baliraine et al. 2004) but molecular systematic studies and crossing experiments suggest
that these taxa are members of the species complex that also includes C. anonae (Barr and
McPheron 2006; Barr et al. 2006; Virgilio et al. 2008; Erbout 2008). Diagnostic tools previously
developed for these three species (Baliraine et al. 2002; Barr et al. 2006) should not be regarded as
reliable because geographically associated variation was not adequately sampled from these species
(Virgilio et al. 2008).
Separation of C. capitata from C. rosa is possible by using a PCR-AGE protocol to observe ITS-1
length variation between species (Douglas and Haymer 2001; Barr et al. 2006), but under certain
conditions additional banding patterns can be observed for C. rosa (Virgilio et al. 2008). Based on
systematic studies, C. capitata and C. rosa are not close relatives (Barr and McPheron 2006; Barr
and Wiegmann 2009) and discrimination between these two species for intercepted individuals is
based on host-sue information (i.e. both are polyphagous feeders with wide host ranges). Other
Ceratitis species, however, are also polyphagous and/or pests of similar commodities: for example,
C. fasciventris, C. anonae, and C. cosyra. Barr et al. (2006) provides additional information for
distinguishing C. capitata from its closest relative C. caetrata using the same ITS-1 PCR-AGE
maker system. This provides additional information regarding specificity of the diagnostic for its
use to diagnose C. capitata.
The Ceratitis PCR-RFLP method was developed using field collected individuals from Africa. The
majority of samples were made in Kenya and the tool assumes that variation in Kenya is
representative of other regions. This is a reasonable assumption for many Ceratitis species, such as
C. capitata, because several species evolved in that region of Africa (Barr 2009) and should exhibit
the greatest level of diversity around Kenya. The method can diagnose most of the species included
in the tool. The PCR-RFLP method cannot, however, discriminate among the three species in the C.
fasciventris-C. anonae-C. rosa species complex or between C. capitata and its close relative C.
caetrata. Rules for interpretation of the method are indicated in the publication.
Analysis of the Barr et al. (2006) data using DNA sequences can also be performed because, for
several pest species in the study, multiple individuals were analyzed. These sequences were not
developed into a DNA barcode protocol of the genus. The lab of M. De Meyer (Belgium) has also
developed many DNA barcode records for afrcian tephritis including many Ceratitis species. These
data have been deposited in GenBank and BOLD but have not been formally published or
developed into a diagnostic protocol.
4.3.10 Molecular identification of Dacus species
The RCP-RFLP methods of analysis developed for New Zealand biosecurity (Armstrong et al.
1997; Armstrong and Cameron 2000; McKenzie et al. 1999) included between one and nine Dacus
species in the tools. These species were represented by limited sample numbers and geographic
collection locations. The most extensive molecular data set for Dacus was a recent molecular
phylogenetic study by Virgilio et al. (2009). The data was not formally analyzed for diagnostic
utility of the DNA sequences but several species exhibited high intraspecific variation precluding
identification using a DNA barcode of the locus. Consequently, there are no standard, routine
molecular tools for distinguishing among Dacus species.
Conclusions
10
Draft protocol: Tephritidae
2010-TPDP-19
Agenda: 13
Excluding techniques developed for population genetics and invasion pathways, several techniques
could be useful for identification of fruit flies to the species level. All of these have limitations
based on geographic scope or taxonomic coverage. Published techniques that would be useful to
explore include Salazar et al. (2002) (geographically limited to Chilean flies and taxonomically
limited to a few Rhagoletis species), Muraji and Nakahara (2002) (limited to sample sizes and focus
on economically important Bactrocera species), and Barr et al. (2006) (but limited to Ceratitis
species in eastern Africa). Regarding unpublished data sets there are interesting approaches using
DNA barcodes for Bactrocera (contact V. Hernandez, MAF) that are based on the research of
Armstrong and Ball (2005) and unpublished work by N. Barr (USDA) to convert the Ceratitis
samples reported by Barr et al. (2006) into a DNA barcode protocol. A similar approach to
diagnostics using a DNA barcode data set for African fruit fly species is being studied by M. De
Meyer (Belgium). These DNA barcode approaches still require evaluation, verification, and
possibly validation.
This draft has been prepared by Norman Barr and improved by the comments of Beatriz SabaterMuñoz.
11