Implications of low-density microfilariae carriers in Anopheles transmission areas: molecular forms of Anopheles gambiae and Anopheles funestus populations in perspective
© Kwansa-Bentum et al.; licensee BioMed Central Ltd. 2014
Received: 5 December 2013
Accepted: 27 March 2014
Published: 1 April 2014
Previous studies have shown a general reduction in annual transmission potential (ATP) of Anopheles species after mass drug administration (MDA) in lymphatic filariasis endemic communities. Whereas results obtained from a monitoring programme after three years of MDA revealed a decrease in ATP of Anopheles funestus this was not the same for An. gambiae s.s. in Ghana. In this study, the ability of these vectors in transmitting Wuchereria bancrofti in nine lymphatic filariasis endemic communities in Gomoa District of Ghana after four rounds of MDA with ivermectin and albendazole was investigated.
After mass screening of inhabitants in these communities, twelve consenting volunteers with different intensities of microfilariae (mf) slept under partly opened mosquito nets as sources of mf blood meal. Hourly collection of mosquitoes and finger-pricked blood were taken from 21.00 to 06.00 hours the following day. For each hour, half of the mosquitoes collected were immediately killed and dissected for mf. The remaining half were maintained up to 13 days for parasite maturation. Parasitaemia and infection rates in the mosquitoes were determined by microscopy. The mosquitoes were identified by microscopy and molecular techniques.
A total of 1,083 participants were screened and the overall parasite prevalence was 1.6% with mf intensities ranging from 0 to 59 per 100 μl and geometric mean intensity of 1.1 mf per ml of blood. Of the 564 mosquitoes collected, 350 (62.1%) were Anopheles spp., from which 310 (88.6%) were An. funestus and 32 (9.1%) An. gambiae. Six anopheline mosquitoes (1.7%) were found infected with L1, but no larva was observed in any of the mosquitoes maintained up to 13 days. Molecular studies showed all An. gambiae s.l. to be An. gambiae s.s., of which 21 (70%) were of the M molecular form.
At low-level parasitaemia after 4 rounds of MDA, there was no recovery of infective stage larvae of W. bancrofti in An. funestus s.l. as well as M and S forms of An. gambiae.
Wuchereria bancrofti is one of the three filarial worms responsible for about 90% of all lymphatic filariasis (LF) cases in the world . These parasites are transmitted through the bite of infective mosquitoes of various genera. A competent vector is one that is capable of ingesting microfilariae (mf) from an infected human, supporting their development to the infective stage larvae (L3) and subsequently transmitting them to other uninfected persons. Depending on the vector species, its region of origin and the parasite density ingested, this ability to sustain the maturation and transmission of LF may be enhanced or restricted [2, 3]. Low density mf is defined as the density of circulating mf in a specified blood compartment that cannot be detected in a significant number of instances when commonly used blood sampling techniques are applied in epidemiological studies; thus 4 mf per 20 μl (200 mf per ml) . Zhang et al. in their study reported that between 1.55 and 2.23% prevalence, there was a threshold provided no individual had an mf density of more than 12 mf per 60 μl of blood.
The Global Programme to Eliminate Lymphatic Filariasis (GPELF) was launched in 2000, with the main goal of halting transmission and reducing disability through annual mass drug administration (MDA) to all persons at risk of infection, particularly if the vectors are Anopheles species . The strategy relies on the assumption that if the mf reservoir in the human host is reduced below a certain threshold, transmission of W. bancrofti by anopheline vectors could be interrupted . This is due to the observation that even though Anopheles mosquitoes yield more infective stage larvae than Culex species, the latter is more efficient at ingesting and developing low-density mf (limitation) than the former . Thus Anopheles mosquitoes are presumed to be efficient vectors of LF when the parasite density in the human population is high, a phenomenon known as “facilitation” . This observation has been the source for the heightened interest in the advocacy for the possible elimination of anopheline-transmitted filariasis; however, a study has observed “facilitation” in An. gambiae s.s. and An. arabiensis but not in An. melas in Gambia or An. merus in Tanzania . Additional health benefits of MDA targeting LF is the reduction in soil transmitted helminths and scabies .
A study in the Bongo district of northern Ghana  indicates a plausible “limitation” in An. gambiae s.l. and/or An. funestus in the transmission of the parasite contrary to other reports . Results from a study in the Gomoa district of southern Ghana also indicated that although transmission potential by An. funestus has decreased significantly after mass chemotherapy with ivermectin and albendazole, there appears to be no change in An. gambiae s.s. in the area (Boakye DA, unpublished data). This suggests that probably not all anophelines exhibit facilitation in their transmission of LF. This work was therefore conducted to determine the roles of these two Anopheles species in the transmission of low level W. bancrofti human mf, since this information is fundamental to the success of GPELF.
Nine LF endemic communities in the Gomoa district of Ghana (between Latitude 5° 24’ - 35’N and Longitude 0° 25’ - 36’W) were selected based on available data on the disease epidemiology in the population and vector species distribution [10–12]. These are Amanful, Ayesuano, Dago, Fawomanye, Hwida, Kyiren, Mampong, Obiri and Okyereko. The district lies in the coastal savannah zone of Ghana and is located 50 km west of Accra, the capital city of Ghana. Average annual rainfall ranges between 760 and 1000 mm, whilst mean annual temperature ranges between 26 and 30°C. The main occupations of the inhabitants are farming and fishing for those living near the shores of the Atlantic Ocean.
Mass screening for microfilariae in the communities
This study was conducted from April to June 2004, the fourth year of MDA with ivermectin and albendazole in these communities. The areas also form part of an on-going annual longitudinal community-based intervention study. Human participation and the mosquito collection were done by cluster sampling method. Mass screening of the study population for mf was done by collecting 100 μl finger-pricked blood from each individual into heparinised capillary tubes and immediately mixing with 900 μl 3% acetic acid. Quantification of parasitaemia used the Sedgwick-Rafter counting chamber method with the compound microscope set at ×100 magnification .
Mosquito collection, maintenance and dissection
After consenting to participate, twelve adult volunteers with varying mf levels slept under partially opened mosquito nets hung over beds in their rooms. At the mid-point of each collection hour, finger-pricked blood was taken and mf density estimated using the same procedure described above. Mosquitoes trapped in the nets were collected each hour from 21.00 hours to 06.00 hours on the next day using an aspirator. About half the number of mosquitoes collected were killed immediately and dissected for ingested mf. The remaining mosquitoes were fed on 10% sugar solution and maintained for up to 13 days in paper-cups at 26-28×C, relative humidity 70-80% and 12-hour photoperiod in the insectary . Mosquitoes that died before day 13 were dissected for developing stages of W. bancrofti, whilst those that survived until the last day were dissected for the presence of infective stage L3 larvae of the parasite.
PCR identification of Anopheles species
Oligonucleotide primer sequences and PCR reaction conditions for species identification
Primer for species ID
Sequence 5’ ⇒ 3’
PCR product size (bp)
Anopheles gambiae s.l. species
93°C 3’ followed by 35 cycles (93°C 30”; 50°C 30”; 72°C 1’); 93°C 30”; 50°C 30”; 72°C 10’
Anopheles gambiae s.s.
Anopheles funestus s.l. species
30 cycles (94°C 30”; 40°C 30”; 72°C 30”); 72°C 10’
Anopheles funestus s.s.
94°C 3’; followed by 35 cycles (94°C 1’; 55°C 1’; 72°C 2’); 94°C 1’; 55°C 1’; 72°C 10”
After gene amplification and digestion, the PCR products were electrophoresed separately in 2% agarose gel. The gel was prepared by adding TAE buffer to the powder, which was placed in a microwave oven (230 V, 50 Hz, 2660 W, 12.0A) for 1 minute to dissolve the solute, and then stained with 0.5 μg/ml Ethidium Bromide. For the electrophoresis, 8 μl of each sample was added to 1 μl of orange G (5X) gel loading dye after placing the solidified gel in 1X TAE buffer in a mini gel system (BIORAD USA). One hundred volts of electric current was passed through it for an hour and the gel photographed over a UV trans-illuminator (UPC, USA) at short wavelength using a Polaroid camera and film type 667 (Polaroid, USA). The sizes of the PCR products were estimated by comparison with the mobility of a 100 base pair molecular weight size marker (Sigma).
Molecular identification of Wuchereria bancrofti larvae in mosquito vectors
After the carcass of infected mosquito was scrapped into 1.5 ml eppendorf tubes, DNeasy Tissue Kit (QIAGEN Inc., USA) was used in the extraction of the parasite’s genomic DNA from animal tissues following the manufacturer’s recommended protocol. After the DNA extraction, aliquots of 5 μl of the filarial DNA extract from the mosquitoes were used as templates for the amplification reaction. The PCR assay was performed using two published specific oligonucleotide primers, NV-1 and NV-2 . The PCR products were electrophoresed in 2% agarose gel as described in the previous section.
Molecular identification of Wuchereria bancrofti microfilariae in human blood
Microfilariae (mf) in human blood samples that were preserved in 3% acetic acid were also characterised after extracting the genomic DNA using the same kit described above. Infected blood samples were amplified and identified using the same procedure described in previous sections.
For the yearly MDA and mass screening for mf prevalence in the communities, oral informed consent was sought from all participants. Subsequently, written consent was obtained from each volunteer who slept under bed nets after the study purpose, procedures, entry and exit criteria were explained to them. All volunteers and the entire community members received that year’s round of MDA immediately after the blood sample collection. The Institutional Review Board of Noguchi Memorial Institute for Medical Research approved the study.
Data were entered into Microsoft Access and analysed for the vector competency of Anopheles spp. in supporting the development of mf to the infective stage larvae. The same software was used to calculate the geometric mean intensity on mf in the human population. One-way analysis of variance (ANOVA) was used to test for the significance of age- and gender-specific variations between the human population and mf, with p value set at 0.05.
Human microfilariae load in the communities after four rounds of MDA
Prevalence of mf and the geometric mean intensity in the study area
Age group (yrs)
mf positive individual (%)
*mf geometric mean intensity
mf positives only
Blood sampling results showing number of people infected with microfilaria of Wuchereria bancrofti in the year 2004 (four years of MDA)
Number positive (mf density/ml of blood)
Geometric mean intensity (mf/ml)
4 (1 – 15.3)
2 (5 – 21)
10 (1 – 155.6)
Mosquito species composition and entomological indices
Entomological indices of the various mosquito species that were caught during the study
Entomological indices (%)
Intensity of infectiond
Number of mosquitoes caught and examined before and after maintenance in the laboratory
No: of mosquitoes
No: of mosquitoes examined immediately after collection
No: of mosquitoes examined from days 1-8 of maintenance
No: of mosquitoes examined from days 9-13 of maintenance
Engorged with blood
Infected (no: mf)
Infected (no: mf)
PCR identification of Anopheles mosquitoes and Wuchereria bancrofti
Of the 32 An. gambiae s.l. collected, 30 were identified as An. gambiae s.s. as they showed the expected diagnostic band size of 390 base pairs. After restriction enzyme treatment with Hha I, M forms remained a single band of 390 bp since there was no digestion. S forms on the other hand resulted in two bands of 110 and 280 bp. Of the An. gambiae s.s digested, 21 (70%) were M forms with 9 (30%) S molecular forms. Among the 310 An. funestus s.l. collected, 286 were identified by PCR; of which 267 (86%) were An. funestus s.s. with diagnostic band sizes of 460 base pairs, and 19 (6%) were An. Leesoni with band sizes of 146 base pairs. The presence of W. bancrofti in 20 infected mosquitoes and 10 human blood samples were confirmed at 188 bp.
An LF-endemic community is said to have low mf density when the density of circulating mf is less than 200 mf per ml of blood, an amount which cannot be detected in a significant number of instances when commonly used blood sampling techniques are employed . Nonetheless, this depends on variables such as volume of blood examined, source of blood sampled (venous or capillary) and method of mf detection that is employed. In this study, 100 μl of finger-prick blood was used, which is an appreciable amount of blood compared to the popular technique for mf detection in routine public health practice of 20 μl finger-prick blood . As such it could be inferred that the mean mf intensity of 1.07 and 79.45 mf per ml of capillary blood in the entire study communities and mf positive individuals respectively were really low in the studied area. This may be due to the 66.6% overall coverage rate in MDA with ivermectin and albendazole for 4 years leading to a reduction in mf densities among the inhabitants (Boakye DA, unpublished data). Evidence from Okyereko supports this view that MDA has been effective; in this study period, 155.6 mf per ml of blood were recorded among mf positive individuals, hitherto the commencement of MDA, as high as 819 mf/ml of blood were recorded among this group .
Various observations have been made regarding mf prevalence and intensities in study populations [9, 11, 12, 22]. These could be attributed in part to the occupational activities of inhabitants of the study areas as well as the biting pattern of the local anopheline vectors, which are presently known to be the main vectors of LF in Ghana [10–12]. Studies on the relationship between mf density in blood meals and the percentage of Anopheles mosquitoes that ingest mf have not provided consistent results [3, 9, 22–25]. Southgate and Bryan  showed that although many of the mf ingested by Anopheles vectors are damaged by the mosquito’s foregut armature, the proportion of mf destroyed does not depend on the number of mf ingested and varies between members of the An. gambiae complex and An. funestus. It is therefore not proper to extend findings from a given area to the other even for the same species. Additionally, other anatomical structures and immune factors other than the foregut armature could be modulating mf density following ingestion. Further studies are thus required to provide more understanding into the vector-parasite relationships.
Indeed the significance of distinctly different host-parasite relationships lies in the importance of low-density mf in sustaining transmission in various endemic areas with different genera of mosquito vectors . As hypothesised, the theory of “limitation” allows transmission to occur and build-up when most infected human hosts have low mf densities, whereas situations of well marked “facilitation” will give rise to transmission thresholds below which transmission will ultimately cease leading to parasite elimination from the human population. However, such predictions of parasite extinction or parasite resurgence can only be made with confidence when characteristics of the local vector-mf relationship are well understood .
As part of our study, we described the circadian pattern of mf periodicity in southern Ghana. Based on hourly examination of twelve volunteers for nine hours, we observed that mf concentration in peripheral blood followed a wave-like concentration peaking around 01.00 hours, which was similar to other findings [12, 27, 28]. This interesting behavioural pattern of mf is said to be the parasite’s response to oxygen tension, which is high in peripheral circulation at night due to the low human activity at this time of the day [29, 30].
In their study, Dzodzomenyo M. et al.  observed An. funestus to be the most abundant mosquito species in the early dry season while An. gambiae was predominant in the wet season. Our study was conducted in March, which is the peak of the dry season in Ghana and thus may contribute to the low number of An. gambiae that were captured. Studies show that the M and S forms of Anopheles gambiae s.s. do occur in sympatry in southern Ghana . Our study revealed that most of the Anopheles gambiae s.s. were M form, which has a remarkable ecological flexibility and is known to prevail in inundated areas where dry season breeding opportunities exist . Further studies could look at the role of these molecular forms of Anopheles gambiae s.s. in transmission of W. bancrofti following MDA.
After 4 rounds of mass drug administration, parasitaemia was brought to a low level in the study communities. Low levels of circulating microfilariae in the inhabitants might have contributed to the no recovery of infective stage larvae of W. bancrofti in An. funestus s.l. as well as M and S forms of An. gambiae. Although the mosquito numbers were low, a further study is recommended to ascertain this observation.
We acknowledge the technical contributions of Sampson Otoo and Philip Doku. We thank the chiefs and elders of the study communities, and all persons who provided blood samples without whom this work would not have seen the light of day. We thank all members of Parasitology Department (NMIMR), and appreciate the support of Professor Alexander Nyarko Director, NMIMR. This work was supported by WHO/ TDR Research Grant to DAB (WHO/TDR grant No. A00638).
- Michael E, Bundy DA, Grenfell BT: Re-assessing the global prevalence and distribution of lymphatic filariasis. Parasitol. 1996, 112: 409-428. 10.1017/S0031182000066646.View ArticleGoogle Scholar
- Crans WJ: Experimental infection of Anopheles gambiae and Culex quinquefasciatus pipiens fatigans with Wuchereria bancrofti in coastal East Africa. J Med Entomol. 1973, 10: 189-193.View ArticlePubMedGoogle Scholar
- McGreevy PB, Kolstrup N, Tao J, McGreevy MM, Marshall TF: Ingestion and development of Wuchereria bancrofti in Culex quinquefasciatus, Anopheles gambiae and Aedes aegypti after feeding on humans with varying densities of microfilariae in Tanzania. Trans R Soc Trop Med Hyg. 1982, 76: 288-296. 10.1016/0035-9203(82)90170-5.View ArticlePubMedGoogle Scholar
- Southgate BA: The significance of low density microfilareamia in the transmission of lymphatic filarial parasites. J Trop Med Hyg. 1992, 95: 79-86.PubMedGoogle Scholar
- Zhang SQ, Zhang QJ, Cheng F, Wang LL, Pen GP: Threshold of transmission of Brugia malayi by Anopheles sinensis. J Trop Med Hyg. 1991, 94: 245-250.PubMedGoogle Scholar
- Yamey G: Global alliance launches plan to eliminate lymphatic filariasis. BMJ. 2000, 320: 269-10.1136/bmj.320.7230.269.PubMed CentralView ArticlePubMedGoogle Scholar
- Webber RH: Eradication of Wuchereria bancrofti through vector control. Trans R Soc Trop Med Hyg. 1991, 73: 722-724.View ArticleGoogle Scholar
- Mohammed KA, Deb RM, Stanton MC, Molyneux DH: Soil transmitted helminths and scabies in Zanzibar, Tanzania following mass drug administration for lymphatic filariasis – a rapid assessment methodology to assess impact. Parasit Vectors. 2012, 5: 299-10.1186/1756-3305-5-299.PubMed CentralView ArticlePubMedGoogle Scholar
- Boakye DA, Wilson MD, Appawu MA, Gyapong J: Vector competence for Wuchereria bancrofti of the Anopheles populations in the Bongo District of Ghana. Ann Trop Med Parasitol. 2004, 98: 501-508. 10.1179/000349804225003514.View ArticlePubMedGoogle Scholar
- Appawu MA, Baffoe-Wilmot A, Afari EA, Nkrumah FK, Petrarca V: Species composition and inversion polymorphism of the Anopheles gambiae complex in some sites of Ghana, West Africa. Acta Trop. 1994, 56: 15-23. 10.1016/0001-706X(94)90036-1.View ArticlePubMedGoogle Scholar
- Dunyo SK, Appawu M, Nkrumah FK, Baffoe-Wilmot A, Pedersen EM, Simonsen PE: Lymphatic filariasis on the coast of Ghana. Trans R Soc Trop Med Hyg. 1996, 90: 634-638. 10.1016/S0035-9203(96)90414-9.View ArticlePubMedGoogle Scholar
- Dzodzomenyo M, Dunyo SK, Ahorlu CK, Coker WZ, Appawu MA, Pedersen EM, Simonsen PE: Bancroftian filariasis in an irrigated project community in southern Ghana. Trop Med Int Health. 1999, 4: 13-18. 10.1046/j.1365-3156.1999.00354.x.View ArticlePubMedGoogle Scholar
- McMahon JE, Marshall TF d C, Vaughan JP, Abaru DE: Bancroftian filariasis: a comparison of microfilariae counting techniques using counting chamber, standard slide and membrane (Nucleopore) filtration. Ann Trop Med Parasitol. 1979, 73: 457-464.PubMedGoogle Scholar
- Janousek TE, Lowrie RC: Vector competency of Culex quinquefasciatus (Haitian strain) following infection with Wuchereria bancrofti. Trans R Soc Trop Med Hyg. 1989, 83: 679-680. 10.1016/0035-9203(89)90395-7.View ArticlePubMedGoogle Scholar
- Collins FH, Mendez MA, Rasmussen MO, Mehaffey PC, Besansky NJ, Finnerty V: A ribosomal RNA gene probe differentiates member species of the Anopheles gambiae complex. Am J Trop Med Hyg. 1987, 37: 37-41.PubMedGoogle Scholar
- Scott JA, Brogdon WG, Collins FH: Identification of single specimens of the Anopheles gambiae complex by the polymerase chain reaction. Am J Trop Med Hyg. 1993, 49: 520-529.PubMedGoogle Scholar
- Ramzy RM, Farid HA, Kamal IH, Ghada HI, Zakariah SM, Rifky F, Weil GJ, Williams SA, Gad AM: A polymerase chain reaction–based assay for detection of Wuchereria bancrofti in human blood and Culex pipiens. Trans R Soc Trop Med Hyg. 1997, 91: 156-160. 10.1016/S0035-9203(97)90205-4.View ArticlePubMedGoogle Scholar
- Fanello C, Santolamazza F, della Torre A: Simultaneous identification of species and molecular forms of the Anapheles gambiae complex by PCR-RFLP. Med Vet Entomol. 2002, 16: 461-464. 10.1046/j.1365-2915.2002.00393.x.View ArticlePubMedGoogle Scholar
- Brito AC, Williams P, Fontes G, Rocha EMM: A comparison of two Brazilian populations of Culex quinquefasciatus (Say, 1823) from endemic and non-endemic areas to infection with Wuchereria bancrofti (Cobbold, 1877). Mem Inst Oswaldo Cruz. 1997, 92: 33-36. 10.1590/S0074-02761997000100007.View ArticlePubMedGoogle Scholar
- Kartman L: Suggestions concerning an index of experimental filarial infection in mosquitoes. Am J Trop Med Hyg. 1954, 3: 329-337.PubMedGoogle Scholar
- Ramachandran CP: A guide to methods and techniques in Filariasis Investigations. Filar Res Off Inst Med Res, Kuala Lumpur. 1970, 39pp-Google Scholar
- Coulibaly YI, Dembele B, Diallo AA, Kristensen S, Konate S, Dolo H, Dicko I, Sangare MB, Keita F, Boatin BA, Traore AK, Nutman TB, Klion AD, Touré YT, Traore SF: Wuchereria bancrofti transmission pattern in southern Mali prior to and following the institution of mass drug administration. Parasit Vectors. 2013, 6: 247-10.1186/1756-3305-6-247.PubMed CentralView ArticlePubMedGoogle Scholar
- Bryan JH, McMahon P, Barnes A: Factors affecting transmission of Wuchereria bancrofti by anopheline mosquitoes. 3. Uptake and damage to ingested microfilariae by An. gambiae, An. arabiensis, An. merus and An. funestus in East Africa. Trans R Soc Trop Med Hyg. 1990, 84: 265-268. 10.1016/0035-9203(90)90281-I.View ArticlePubMedGoogle Scholar
- Bryan JH, Southgate BA: Factors affecting transmission of Wuchereria bancrofti by anopheles mosquitoes. 1. Uptake of microfilariae. Trans R Soc Trop Med Hyg. 1988, 82: 128-137. 10.1016/0035-9203(88)90286-6.View ArticlePubMedGoogle Scholar
- Bryan JH, Southgate BA: Factors affecting transmission of Wuchereria bancrofti by anopheles mosquitoes. 2. Damage to ingested microfilariae by mosquito foregut armatures and development of filarial larvae in mosquitoes. Trans R Soc Trop Med Hyg. 1988, 82: 138-145. 10.1016/0035-9203(88)90288-X.View ArticlePubMedGoogle Scholar
- Southgate BA, Bryan JH: Factors affecting transmission of Wuchereria bancrofti by anopheline mosquitoes. 4. Facilitation, limitation, proportionality and their epidemiological significance. Trans R Soc Trop Med Hyg. 1992, 86: 523-530. 10.1016/0035-9203(92)90096-U.View ArticlePubMedGoogle Scholar
- Tanaka H: Periodicity of microfilariae of human filariasis analysed by a trigonometric method (Aikat and Das). Jpn J Exp Med. 1981, 51: 97-103.PubMedGoogle Scholar
- Gatika SM, Fugimaki Y, Njuguna MN, Gachihi GS, Mbugua JM: The microfilarial periodicity pattern of Wuchereria bancrofti in Kenya. J Trop Med Hyg. 1994, 97: 60-64.PubMedGoogle Scholar
- Hawking F, Pattanayak S, Sharma HL: The periodicity of microfilariae XI. The effect of body temperature and other stimuli upon the cycles of Wuchereria bancrofti, Brugia malayi, B. ceylonensis and Dirofilaria repens. Trans R Soc Trop Med Hyg. 1966, 60: 496-513.Google Scholar
- Denham DA, McGreevy PB: Brugian filariasis: epidemiological and experimental studies. Adv Parasitol. 1977, 15: 243-309.View ArticlePubMedGoogle Scholar
- Yawson AE, McCall PJ, Wilson MD, Donnelly MJ: Species abundance and insecticide resistance of Anopheles gambiae in selected areas of Ghana and Burkina Faso. Med Vet Entomol. 2004, 18: 372-377. 10.1111/j.0269-283X.2004.00519.x.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.