Diversity of Culicoides in the middle belt of Ghana with Implications on the transmission of Mansonella perstans; a molecular approach

Abstract

Background

Culicoides, also known as biting midges, carry pathogens which include Mansonella perstans. Mansonella perstans is a nematode parasite implicated in a number of disease outcomes. Even though a high prevalence of about 75% M. perstans infection has been recorded in some communities in the middle belt of Ghana, and a wide diversity of Culicoides species has been identified, the exact Culicoides species transmitting M. perstans in Ghana has not yet been deciphered. This study therefore aimed at assessing the species diversity of Culicoides and their role in the transmission of M. perstans in the middle belt of Ghana.

Methods

Culicoides species were sampled from 11 communities in the Asante-Akim North and Sene West districts in the middle belt of Ghana. Centre for Disease Control (CDC) UV light traps, as well as human bait (i.e. human landing catch and engorged catch) methods were used to assess the species abundance and diversity of Culicoides in the study communities in the wet and dry season. A colorimetric Loop-Mediated Isothermal Amplification (LAMP) assay was performed to assess the vector competence of the various Culicoides species.

Results

A total of 4810 Culicoides from 6 species were sampled. These included Culicoides inornatipennis, C. milnei, C. schultzei, C. grahamii, C. neavei, and C. imicola. Culicoides imicola was the most abundant species (56%) followed by C. grahamii (16%). Light traps sampled the most diverse species (6 species). Human landing catch and engorged catch methods identified three anthropophilic species, C. grahamii, C. milnei, and C. inornatipennis, with C. grahamii being the most anthropophilic with a peak biting time between the hours of 5 p.m. to 6 p.m. Generally, there was relatively higher species abundance in the wet than dry season. LAMP assay identified C. grahamii as the potential vector for M. perstans transmission in the middle belt of Ghana.

Conclusions

For the first time, we have demonstrated that C. grahamii is the potential competent vector for M. perstans transmission in the middle belt of Ghana. It is more abundant in the rainy season and has a peak biting time between the hours of 5 and 6 p.m.

Graphical Abstract

Background

Biting midges of the genus Culicoides, family Ceratopogonidae, have a near worldwide distribution with over 1400 species described [1]. These vectors are small, stout hematophagous flies with a distinctive pattern of dotted wings. They are a biting nuisance to people, domestic animals, and wild animals in regions where they are prevalent. Their bites can occasionally cause allergic skin reactions, which result in urticaria in some individuals [2].

Biting midges have been incriminated as the vectors of parasites as well as multiple arboviruses of veterinary and public health importance [3] such as Schmallenberg virus and Bluetongue virus in ruminants [4]. Their occurrence, abundance, and diversity vary among different ecological zones and seasons. The lifestyle of Culicoides is characterized by complete metamorphosis from the egg through larval and pupal stages until the adult stage [5]. The presence of moisture-rich habitats is essential for the development of pupal and larval stages, as such Culicoides are mostly associated with aquatic and semi-aquatic habitats, e.g. mud, marshes, and ponds. Characteristically, they can be found in areas where there are many plantain farms. The decaying leaves and stems of plantain serve as a good breeding ground for these flies. The presence of livestock also plays a crucial role in the distribution of Culicoides [6].

Culicoides spp. are vectors for transmitting the nematode parasite Mansonella perstans [7]. The life cycle of M. perstans is similar to that of other nematode parasites such as Onchocerca volvulus, Loa loa, and Wuchereria bancrofti, where humans are the definitive host [2, 5].

About 114 million people in Africa are estimated to be infected with M. perstans, and over 580 million people globally are considered to be at risk [7]. Despite the wide distribution and high prevalence of M. perstans infection, it is not officially regarded as one of the Neglected Tropical Diseases (NTDs) and has received minimal attention [2]. Unlike other human filarial infections that have well-defined clinical signs and symptoms, such as in W. bancrofti infection and the formation of subcutaneous nodules in O. volvulus infection, M. perstans infection does not present with any distinct or peculiar clinical signs and symptoms [2]. However, M. perstans infection has been shown to modulate the immune system resulting in complications in disease manifestations and recovery [2, 8].

Some Culicoides species are anthropophilic (attracted to humans) whereas others are not. Moreover, not all anthropophilic species are competent vectors of M. perstans. In East Africa, the taxonomy of Culicoides has been investigated by Khamala & Kettle, who identified 61 species that did not contribute to the transmission of M. perstans [9]. Different geographical locations have reported diverse Culicoides species but very few studies have identified vectors of M. perstans in endemic areas. Culicoides milnei is implicated in the transmission of M. perstans in the southwest region of Cameroon [10], and C. grahamii is confirmed as the vector in Congo [11]. However, the vector for M. perstans transmission in Ghana remains unknown even though high infection prevalence has been reported [12]

Previously, identification of M. perstans vectors was by detection of infective larvae in the female adult Culicoides upon dissection by microscopy. This requires great expertise as a result of the morphological similarities of different microfilariae species. Molecular (DNA-based) techniques, particularly PCR for filarial parasite detection in both the human host and vectors, are sensitive and specific, and useful in M. perstans epidemiological surveys [2]. A number of isothermal amplification methods targeting DNA have been developed, which offer appreciable advantages over PCR [13]. Of these, loop-mediated isothermal amplification (LAMP) is a widely adopted approach. Its ability to be conducted at a stable temperature, sensitivity and specificity, visual detection formats without the need for sophisticated equipment, and good performance, even using crude DNA, offers considerable advantages over PCR [14].

We carried out an entomological survey to assess the species diversity of Culicoides, established their role in the transmission of M. perstans in the middle belt of Ghana, and sought to identify the potential vector for its transmission using a loop-mediated isothermal amplification assay.

Methods

Study design

The study was conducted in 11 communities within the middle belt of Ghana. Seasonal collections of Culicoides were carried out from June 2020 to October 2020 for the wet season and November 2020 to April 2021 for the dry season. Culicoides were collected during the 1st week in each month. Centre for Disease Control (CDC) light traps were used to assess the species diversity of Culicoides present in the study communities. Human landing catch (HLC) was used to identify anthropophilic Culicoides species. Engorged Culicoides collected by drop trap using a known microfilaria-positive donor as a bait were used to assess vector competence. DNA was extracted from all Culicoides collected, and the LAMP assay was performed to assess vector competence of M. perstans transmission by Culicoides.

Collection of adult Culicoides species using CDC miniature UV light traps

Here, each community was divided into four quadrant centers. A CDC New Standard Miniature light trap (John W. Hock Company, Gainesville, FL, USA), was mounted in each quadrant center, near human habitations. Collections were made overnight (6 p.m. to 6 a.m.) each sampling day. Light traps were mounted at the same spot in the wet and dry seasons. Attracted by UV light emitted by the trap, Culicoides species and other flies were trapped in a Petri dish containing 80% alcohol placed in the suspended trap. The trapped flies were transferred into labelled 50-ml Falcon tubes containing 80% alcohol and placed in a cold box for transportation to the laboratory for morphological identification. The number of each Culicoides species sampled was recorded after morphological identification.

Human landing catches (HLC)

To determine anthropophilic Culicoides species, flies were collected using the human landing catch technique in all study communities. HLC was carried out in the evening from 4 to 7 p.m. each sampling day. Sampling was done by four well-trained collectors dressed in protective clothing against midges. They were positioned in four randomly selected houses and provided with torches to aid collection in darkness. Female Culicoides seeking blood meals were directly aspirated as soon as they landed on collectors and were then transferred into hourly labelled netted plastic cups and transported to the laboratory for morphological identification and further assessment. The number of Culicoides collected each hour was recorded to assess the peak biting time of the various anthropophilic Culicoides species.

Collection of engorged Culicoides from known M. perstans-positive volunteer using a drop trap

To elucidate the role of anthropophilic species in M. perstans transmission (vector competence assessment), engorged catch method was deployed. The principle underlying this technique stems from the transmission cycle of M. perstans. This method involved an M. perstans microfilaremic donor who acted as bait. Collections were made from 6 p.m. to 6 a.m. each day for 4 consecutive nights. The M. perstans microfilaremic volunteer sat under a netting cage trap, with the net raised about 200 cm above the ground. Upon exposure, Culicoides were attracted, and cage netting was lowered after the volunteer was exposed for about 10 min to trap the attracted Culicoides. After about 10 min of lowering the cage (which is the estimated time for Culicoides to be fully engorged), the flies were aspirated from the net and blown into labelled 50 ml Falcon tubes and transported to the laboratory for morphological identification and further assessment.

Once collected, Culicoides were kept alive for 8 days before laboratory experiments.

Morphological identification of Culicoides species

Morphological identification was done by examination of wing pigmentation pattern under a dissecting microscope. In cases where wing pigmentation was not enough, other morphological features such as genitalia, maxillary palps, and inter-ocular space were used [15, 16].

Culicoides DNA extraction and quantification

After morphological identification, Culicoides were pooled into groups of 100 prior to DNA extraction. Each pool consisted of the same species of Culicoides. Machery Nagel Bioanalysis Nucleospin Tissue Kit was used in the extraction and purification of DNA from the Culicoides. Samples were completely homogenized using a MagNalyser. Afterwards, 180 µl of Lysis Buffer T1 and 25 µl of proteinase K were added to homogenize the samples and incubated at 56 °C overnight for pre-lysis; 200 µl of Lysis Buffer B3 was added and incubated at 70 °C for 10 min to achieve complete lysis of the chitinous exoskeleton and other proteins. Subsequently, 210 µl of ethanol was added to the lysate and thoroughly vortexed to adjust the DNA binding conditions and to precipitate the DNA. The resulting solution was pipetted into Nucleospin tissue columns placed in a 2-ml collection tube and centrifuged at 11,000 g for 1 min. Two washes with wash buffers BW and B5, respectively, were performed to get rid of unwanted dissolved cellular components, after which the spin column membranes were dried at 11,000 g for 1 min. The spin columns were placed in new 1.5-ml sterile Eppendorf tubes; 50 µl of Buffer BE (elution buffer) was added to the dried spin columns, incubated at room temperature for 1 min, and centrifuged at 11,000 g to elute DNA bound to the nucleospin tissue columns. Elution was repeated to achieve maximum yield. After extraction of Culicoides DNA, the concentration and purity of the DNA was measured using DeNovix NanoDrop.

Detection of M. perstans infection in Culicoides species using colorimetric Mp419 LAMP assay.

The LAMP assay was carried out with a primer set as shown in Additional file 1: Table S1. Working solutions of 10× primer mixes were prepared from primer stock as described in Additional file 2: Table S2. The reaction was carried out in a total volume of 20 μl (18 μl of the reaction master mix and 2 μl of the DNA template) in polymerase chain reaction (PCR) micro-tubes (Additional file 3: Table S3). Amplifications were performed using an Applied Biosystems GeneAmp^® PCR System 9700 as all reactions were incubated at 63 °C (isothermal condition) for up to 40 min. A sample was considered positive for M. perstans DNA if an obvious colour change from pink to yellow was observed by two independent assessors, while for the negative samples there was no colour change in the phenol red colour.

Results

A total of 4810 Culicoides comprising six different species (C. imicola, C. grahamii, C. neavei, C. schultzei, C. inornatipennis, and C. milnei) were collected in the 11 study communities. All Afrisere and Dukusen in the Asante-Akim North District recorded the highest abundance of Culicoides accounting for 17.2 and 13.7%, respectively, with Bebuso, recording the least Culicoides abundance (1.5%) (Table 1).

Table 1 Relative abundance of Culicoides species in study communities

Of the 4810 Culicoides, 95% of the collections were from the light traps whereas < 1% was sampled using the drop trap. The light trap collected the highest diversity of Culicoides (6 species) compared to the HLC (2 species) and drop trap (3 species).

Culicoides imicola was the most abundant species (56.2%) and C. inornatipennis was the least (1.3%) (Table 2).

Table 2 Culicoides species collected by light traps, human landing catch (HLC), and engorged catch

The highest abundance of Culicoides was observed in the wet season except for C. neavei, which had higher abundance in the dry season (Fig. 2). Asante Akim North district had the highest number of Culicoides (6 species) and was the only district that recorded C. inornatipennis (Fig. 1).

Fig. 1

Abundance and diversity of Culicoides in the Asante Akim North and Sene West districts

Fig. 2

Seasonal variation of Culicoides abundance

Culicoides grahamii was the most abundant anthropophilic species (66%). Both C. grahamii and C. milnei exhibited similarities in their biting pattern, showing a biting peak at 4–5 pm in the Sene West and Asante Akim North districts. Of the three anthropophilic species, C. milnei showed highest biting activity at 6 to 7 p.m. and the lowest biting activity at 4 to 5 p.m. (Fig. 3).

Fig. 3

Biting patterns of anthropophilic Culicoides species in the Asante Akim North (A) and Sene West (B) districts

After the LAMP reaction, two pools of C. grahamii tested positive for M. perstans (showing a colour change from pink to yellow) (Fig. 4). These were observed in reaction tubes 14 (C. grahamii collected by light trap) and reaction tube 48 (C. grahamii collected by engorged catch) (Fig. 4).

Fig. 4

LAMP Assay reaction tubes before (A) and after reactions (B). Culicoides neavei light trap and HLC collections (1–6), C. milnei light trap and HLC collections (7–9), C. grahamii light trap and HLC collections (10–16), C. schultzei light trap collections (17–18), C. imicola light trap collections (19–45), C. inornatipennis light trap and HLC collections (46), C. milnei drop trap collections (47), C. grahamii drop trap collections (48), Pos ( +) positive control, Neg (−) negative control, pink colour indicates negative and yellow indicates positive

Discussion

Using colorimetric loop-mediated isothermal amplification assay [17], this study has successfully identified C. grahamii as the potential vector for M. perstans transmission in the middle belt of Ghana. A similar finding was made in the southwest region of Cameroon where C. grahamii was identified as a potential competent vector of M. perstans [18] even though C. milnei had been identified in earlier studies [10] to be the main vector in said region. This indicates that C. grahamii is susceptible to M. perstans invasion.

Six Culicoides species were identified in the surveyed communities in the current study, and all six Culicoides species were also identified by Debrah et al. [12], who were the first to comprehensively investigate Culicoides diversity and the burden of M. perstans infection in the middle belt of Ghana. Whereas Debrah et al.’s. [12] attempt to implicate specific Culicoides species as the vectors of M. perstans using microscopy was inconclusive, C. grahamii has proven to be effective in the carriage and transmission of M. perstans in the current study.

The engorged catch (using an overnight drop trap) as an experimental model to assess vector competence of anthropophilic Culicoides species was effective as only competent vectors are able to pick up M. perstans microfilariae from infected individuals and subsequently transmit it to another person through a blood meal. It was therefore not surprising that C. grahamii and C. milnei, being vectors identified in Cameroon, were the only species sampled by this method [10]. In addition to the drop trap, the detection of M. perstans infection in light trap collections (naturally infected Culicoides) is indicative of the Mp 419 LAMP assay’s ability to detect natural infection as well as experimental infection in Culicoides. This is important because it demonstrates the assay’s possible usage as a screening tool for M. perstans infection without having to go through experimental infection, which is time-consuming and possibly costly.

Several studies rely on the dissection of Culicoides post-infection with M. perstans [10, 19] to identify vectors with infective larvae (L3). This method requires high expertise as Culicoides species are dissected to identify M. perstans infective larvae, which may easily be confused with other microfilaria species because of their morphological similarities. In the present study, we relied on a LAMP to detect M. perstans in the Culicoides as used by Poole et al. [17]. As an improvement in existing M. perstans diagnostic techniques mainly microscopy and PCR, Poole et al. [17] developed a LAMP assay that offered considerable advantages such as increased specificity, faster detection time, and requirement of less expensive equipment for performance. It also offers several alternate and easy ways of visualizing results as compared to PCR. With a limit of detection of 0.1 pg (equivalent to 1/1000th) fragment of an M. perstans microfilaria, and a specificity of 100%, it can detect M. perstans in the blood of infected patients and also identify M. perstans in infected Culicoides species [17]. New diagnostic tools with improved diagnostic capacity that are field-friendly and useful in resource-limited settings are needed to improve investigations of NTDs to achieve the Sustainable Development Goals.

In a defined geographical area, the Culicoides abundance and diversity strongly depend on the availability and type of breeding sites [10]. Communities in the Asante Akim North district on average recorded a higher species diversity and abundance than the Sene West district.

Afrisere and Abutantri recorded the highest Culicoides abundance, which could be attributed to the presence of favourable vegetation and breeding sites that enhances Culicoides larval development. The presence of plantations, livestock, and relatively thicker trees and bushes in these communities is known to provide suitable habitats for the vectors [15]. Bebuso in the Asante Akim North district recorded the fewest Culicoides. This could be attributed to the absence of livestock (specifically cattle, sheep, goats, and wild game), which serves as a main blood meal source, and their moist dung, which also provides a fertile breeding ground for some species. This is in line with the findings of Kameke et al. [11] who reported a very high abundance of Culicoides in livestock stables compared to regions farther from stables.

The biting patterns of anthropophilic species were determined by HLC. This is because trap collections are reported to be inaccurate in estimating the biting rate of Culicoides species [19]. HLC identified C. grahamii, C. milnei, and C. inornatipennis, which suggests these species are preferentially anthropophilic. Culicoides grahamii and C. milnei, being the most anthropophilic, accounted for 66% and 26% of the entire anthropophilic species, respectively. Wanji et al. [10], in Cameroon, also identified these three species, among others, as anthropophilic. In the present study, C. inornatipennis was the least anthropophilic species, which does not agree with earlier findings by Debrah et al. [12], who reported C. inornatipennis as the only anthropophilic species, and was also abundant in the Asante-Akyim North district. Over the years, climatic changes, evolving and emerging farming practices (such as the use of insecticides and weedicides), asnd urbanization may have adversely affected the abundance of C. inornatipennis. Notably, the most anthropophilic species (C. grahamii and C. milnei) were present in both districts. Culicoides grahamii and C. milnei exhibited similar biting patterns, with a peak biting time between the hours of 5 and 6 p.m., indicating that the time of maximum human-vector contact occurs around 5 to 6 p.m. Between the hours of 6 and 7 p.m., C. milnei showed the highest biting activity among all the anthropophilic species. This finding agrees with that of Wanji et al. [10] who reported that C. milnei is essentially a nocturnal species. Regarding the biting pattern of C. grahamii, our finding is in contrast to that of Hopkins [20] since he found this species to be only diurnal. This may be a result of differences in diurnal temperatures as these differences have been shown to influence periodicity or flight activity [21]. Increasing the time frame of human collections may provide a better overview of the biting patterns of the different Culicoides identified by HLC.

UV light traps collected approximately 95% (4557 out of 4810) of all the Culicoides recorded, and this included all six species identified, namely C. inornatipennis, C. imicola, C. grahamii, C. milnei, C. neavei, and C. schultzei. The high numbers collected using the light trap indicated that the UV light emitted by the trap serves as a great attractant to the midges as they not only follow any light source but also the kind of light emitted. Therefore, UV light traps are recommended for large-scale entomological surveys [21]. Culicoides imicola was the most abundant species in this study and a similar observation was made by Debrah et al. [12]. This may be due to the presence of livestock, the known preferred animal host for C. imicola [19] in most of the study communities. Despite the similarities in findings, the species diversity in the middle belt of Ghana as reported in the present study differs from the findings of Debrah et al. [12] who reported the presence of Culicoides accraensis and C. fulvithorax, which were not identified in this study. This is suggestive of either a drastic reduction in the abundance of these species due to the unfavourable climatic changes over the years or probably the unavailability of their preferred host(s).

Regarding seasonal variation in Culicoides abundance, there was a higher abundance of Culicoides in the rainy season, as similarly reported by Silva & Carvalho [22] and Debrah et al. [12], than the dry season. The excessive rainfall, which kept the soil and other plant matter moist, provided suitable conditions for Culicoides larval development and might have contributed to the higher abundance of the Culicoides species in the rainy season. The trend of higher Culicoides abundance in the rainy season was observed for all Culicoides species except C. neavei, which suggests that arid conditions enhance the growth of C. neavei pupal and larval stages [23]. Also, it is generally observed that there is a higher occurrence of bush burning in the dry season in Ghana. The destruction of vegetation covers and moist breeding sites might have contributed to the low abundance of the Culicoides species captured in the dry season in both districts.

Conclusions

In conclusion, we have identified, for the first time to our knowledge, Culicoides grahamii as a potential vector for M. perstans in the middle belt of Ghana.