New assessment of Anopheles vector species identification using MALDI‑TOF MS
Loading...
Date
Authors
Nabet, Cecile
Kone, Abdoulaye K.
Dia, Abdoulaye K.
Sylla, Moussa
Gautier, Magali
Yattara, Mohammed
Thera, Mahamadou A.
Faye, Ousmane
Braack, L.E.O.
Manguin, Sylvie
Journal Title
Journal ISSN
Volume Title
Publisher
BioMed Central
Abstract
BACKGROUND : Anopheles species identification is essential for an effective malaria vector control programme. Matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry (MS) has been developed to identify adult Anopheles species, using the legs or the cephalothorax. The protein repertoire from arthropods can vary according to compartment, but there is no general consensus regarding the anatomic part to be used. METHODS : To determine the body part of the Anopheles mosquitoes best suited for the identification of field specimens, a mass spectral library was generated with head, thorax with wings and legs of Anopheles gambiae, Anopheles arabiensis and Anopheles funestus obtained from reference centres. The MSL was evaluated using two independent panels of 52 and 40 An. gambiae field-collected in Mali and Guinea, respectively. Geographic variability was also tested using the panel from Mali and several databases containing added specimens from Mali and Senegal. RESULTS : Using the head and a database without specimens from the same field collection, the proportion of interpretable and correct identifications was significantly higher than using the other body parts at a threshold value of 1.7 (p < 0.0001). The thorax of engorged specimens was negatively impacted by the blood meal after frozen storage. The addition of specimens from Mali into the database significantly improved the results of Mali panel (p < 0.0001), which became comparable between head and legs. With higher identification scores, the using of the head will allow to decrease the number of technical replicates of protein extract per specimen, which represents a significant improvement for routine use of MALDI-TOF MS. CONCLUSIONS : The using of the head of Anopheles may improve the performance of MALDI-TOF MS. Region-specific mass spectrum databases will have to be produced. Further research is needed to improve the standardization in order to share online spectral databases.
Description
Additional file 1: Fig. S1. Maximum likelihood tree of Anopheles ITS2
sequences. Seaview v4 software, Clustal W and phyML tools. Specimens
identified in reference centres are indicated by taxonomic identification
along with GenBank accession number, namely, An. arabiensis (Pretoria
University, South Africa), An. gambiae (IRD Montpellier, France) and An.
funestus (MRTC Bamako, Mali).
Additional file 2: Fig. S2. Impact of the log(score) threshold on MALDITOF MS species identification using panel A from Mali versus database 1 for each body part of Anopheles, n=52. The number of specimens having correct species identification, error of species identification and absence of identification due to an LSV<threshold are shown in different colours for each body part.
Additional file 3: Fig. S3. Composite correlation index (CCI) heat map grid of mass spectrum protein profiles of Anopheles gambiae. Panel A from Mali, n=52 (a). Panel B from Guinea, n=40 (b). Levels of mass spectral reproducibility are indicated in blue and red, revealing incongruence and relatedness between spectra, with a correlation index variation between 0 and 1, respectively. The coloured squares of the central diagonal reflect the degree of reproducibility of each mass spectrum when compared to itself. Around the central diagonal, spectra from various specimens of Anopheles gambiae were compared. The CCI matrix was calculated using MALDI Biotyper v4.1 software with default settings.
Additional file 4: Fig. S4. Impact of body part on identification results using panels A+B versus database 1, n=92. The number of specimens having correct species identification, error of species identification and absence of identification due to an LSV<1.7 are shown in different colours for each body part.
Additional file 5: Fig. S5. Cross-matching between anatomic parts and sex, panel A versus database 1, n=52. The number of specimens of panel A is shown on the vertical axis. Characteristics of the corresponding MSPs of database 1 (anatomic parts, sex and insufficient matching due to LSV <1.7) are shown in different colours.
Additional file 6: Fig. S6. Impact of the association of anatomic parts, panel A versus database 1, n=52. The number of specimens having correct species identification, error of species identification and absence of identification due to an LSV<1.7 are shown in different colours for each body part and association of body parts.
Additional file 7: Fig. S7. Distribution of spectral log(scores) from heads, thoraces and legs. Panel A versus database 1 or database 2, n=52. Violin plots showing the distribution taking into account the densities of the points for the different log(score) values. The median score is represented with dashes, and the quartiles are represented by dashed lines.
Additional file 8: Fig. S8. Identification results, panel A versus database 1, database 2, database 3 or database 4, n=52. Database 1 was created using n=20 non-engorged laboratory-reared Anopheles and field specimens from the collection of reference centres. Database 2 was created by adding 10 Anopheles specimens collected from the field in Mali to database 1. Databases 3 and 4 were created by adding 10 field specimens from Senegal to databases 1 and 2, respectively. The number of specimens having correct species identification, error of species identification and absence of identification due to an LSV<1.7 are shown in different colours for each body part.
Additional file 9: Fig. S9. Dendrogram of legs mass spectra constructed with specimens of Anopheles gambiae from Kenya, Mali and Guinea (n=15). Specimens from Kenya are laboratory-reared females (mass spectra library). Specimens from Mali and Guinea are field-caught females (panel A and panel B, respectively). The dendrogram was calculated using MALDI Biotyper v4.1 and distance units correspond to relative similarity of mass spectra.
Additional file 10: Fig. S10. Dendrogram of head mass spectra constructed with specimens of Anopheles gambiae from Kenya, Mali and Guinea (n=15). Specimens from Kenya are laboratory-reared females (mass spectra library). Specimens from Mali and Guinea are field-caught females (panel A and panel B, respectively). The dendrogram was calculated using MALDI Biotyper v4.1 and distance units correspond to relative similarity of mass spectra.
Additional file 2: Fig. S2. Impact of the log(score) threshold on MALDITOF MS species identification using panel A from Mali versus database 1 for each body part of Anopheles, n=52. The number of specimens having correct species identification, error of species identification and absence of identification due to an LSV<threshold are shown in different colours for each body part.
Additional file 3: Fig. S3. Composite correlation index (CCI) heat map grid of mass spectrum protein profiles of Anopheles gambiae. Panel A from Mali, n=52 (a). Panel B from Guinea, n=40 (b). Levels of mass spectral reproducibility are indicated in blue and red, revealing incongruence and relatedness between spectra, with a correlation index variation between 0 and 1, respectively. The coloured squares of the central diagonal reflect the degree of reproducibility of each mass spectrum when compared to itself. Around the central diagonal, spectra from various specimens of Anopheles gambiae were compared. The CCI matrix was calculated using MALDI Biotyper v4.1 software with default settings.
Additional file 4: Fig. S4. Impact of body part on identification results using panels A+B versus database 1, n=92. The number of specimens having correct species identification, error of species identification and absence of identification due to an LSV<1.7 are shown in different colours for each body part.
Additional file 5: Fig. S5. Cross-matching between anatomic parts and sex, panel A versus database 1, n=52. The number of specimens of panel A is shown on the vertical axis. Characteristics of the corresponding MSPs of database 1 (anatomic parts, sex and insufficient matching due to LSV <1.7) are shown in different colours.
Additional file 6: Fig. S6. Impact of the association of anatomic parts, panel A versus database 1, n=52. The number of specimens having correct species identification, error of species identification and absence of identification due to an LSV<1.7 are shown in different colours for each body part and association of body parts.
Additional file 7: Fig. S7. Distribution of spectral log(scores) from heads, thoraces and legs. Panel A versus database 1 or database 2, n=52. Violin plots showing the distribution taking into account the densities of the points for the different log(score) values. The median score is represented with dashes, and the quartiles are represented by dashed lines.
Additional file 8: Fig. S8. Identification results, panel A versus database 1, database 2, database 3 or database 4, n=52. Database 1 was created using n=20 non-engorged laboratory-reared Anopheles and field specimens from the collection of reference centres. Database 2 was created by adding 10 Anopheles specimens collected from the field in Mali to database 1. Databases 3 and 4 were created by adding 10 field specimens from Senegal to databases 1 and 2, respectively. The number of specimens having correct species identification, error of species identification and absence of identification due to an LSV<1.7 are shown in different colours for each body part.
Additional file 9: Fig. S9. Dendrogram of legs mass spectra constructed with specimens of Anopheles gambiae from Kenya, Mali and Guinea (n=15). Specimens from Kenya are laboratory-reared females (mass spectra library). Specimens from Mali and Guinea are field-caught females (panel A and panel B, respectively). The dendrogram was calculated using MALDI Biotyper v4.1 and distance units correspond to relative similarity of mass spectra.
Additional file 10: Fig. S10. Dendrogram of head mass spectra constructed with specimens of Anopheles gambiae from Kenya, Mali and Guinea (n=15). Specimens from Kenya are laboratory-reared females (mass spectra library). Specimens from Mali and Guinea are field-caught females (panel A and panel B, respectively). The dendrogram was calculated using MALDI Biotyper v4.1 and distance units correspond to relative similarity of mass spectra.
Keywords
Anopheles, Malaria vectors, Taxonomic identification, Anopheles gambiae, Head, Thorax, Legs, Matrix-assisted laser desorption ionization-time of flight (MALDI-TOF), Mass spectrometry (MS)
Sustainable Development Goals
Citation
Nabet, C., Kone, A.K., Dia, A.K. et al. 2021, 'New assessment of Anopheles vector species identification using MALDI‑TOF MS', Malaria Journal, vol. 20, art. 33, pp. 1-16.