Investigating the Biology and Behavior of Anopheles Squamosus and Its Role in Residual Malaria Transmission in Southern Zambia

INVESTIGATING THE BIOLOGY AND BEHAVIOR OF ANOPHELES SQUAMOSUS AND ITS ROLE IN RESIDUAL MALARIA TRANSMISSION IN SOUTHERN ZAMBIA

By Jordan Hoffman

A thesis submitted to the Johns Hopkins University in conformity with the requirements for the degree of Master of Science.

Baltimore, Maryland April 2019

Abstract

In the last decade, malaria cases in Southern Zambia have declined by 90% due in part to national control efforts. Despite this dramatic reduction, prevalence has remained near 1-2% for the past several years. In 2011, higher than expected rates of anthropophily were observed among Anopheles squamosus in the area, a “zoophilic” species and one that had sporadically been found to contain Plasmodium falciparum sporozoites, indicating the potential importance of secondary vectors. The importance of An. squamosus as a secondary vector was confirmed in 2016 when P. falciparum sporozoites were detected in the species in the region. An. squamosus have been shown thus far to be mainly exophilic and exophagic (feeding and resting outdoors). If this is the case in

Southern Province, new control measures may be necessary for achieving and sustaining malaria elimination. Due to its previously presumed lack of importance to malaria transmission, little is known about the biology or behavior of An. squamosus. This study analyzes mosquito collections from two different collection schemes - one performed as part of a reactive-test-and-treat program, and the second performed along a transect.

Adult mosquitoes were collected using CDC light traps and morphologically identified.

Molecular verification of anopheline species, P. falciparum infectiousness, and blood meal source were determined on samples brought to JHSPH. Household data were incorporated to evaluate associations between household factors and An. squamosus presence, abundance, and behavior. Data from these collections support exophagic and zoophilic behavior by An. squamosus. Although no anthropophily was detected and P. falciparum infectiousness could not be confirmed in any An. squamosus in this study, trends in composite evidence suggest a dominant role of An. squamosus in malaria

ii transmission in the area. The phylogenetic structure generated from the An. squamosus specimens in this study indicate the existence of an An. squamosus species complex, which further emphasizes the importance of molecular identification of vectors to direct vector control efforts. This study confirms that indoor vector control strategies will not be sufficient for elimination of malaria in southern Zambia.

Primary Reader: Dr. Douglas Norris

Secondary Reader: Dr. William Moss

iii

Acknowledgments

All my work on this thesis could not have been achieved without the unwavering support I received from Dr. Douglas Norris. I came into the program with little molecular biology or even laboratory experience, and he made sure I developed the skills I needed to work on this project and to pursue my professional goals. He weathered my bizarre turns in career goals and never lost faith in my ability to accomplish whatever I decided I wanted to do. I cannot stress enough the value of his patience, kindness, and sincerity in my development as a scientist and as a human being. Doug’s leadership was modeled in the laboratory; my experience was shaped tremendously by Christine Jones, Ilinca Ciubotariu, Julia Pringle, Mary Gebhardt, and Giovanna Carpi. They welcomed me into the lab and spent an inordinate amount of time teaching me basic science, answering my endless questions, and reassuring me in the depths of data analysis that it would all work out. They also provided a practically constant supply of tea and chocolate and discussed with equivalent enthusiasm the species of trees we saw on the way to work or the best way to measure tornado incidence across geographic areas. I am incredibly fortunate to have had such a wonderful team supporting me these past two years. In addition to the Norris lab, I would like to thank the ICEMR team at Hopkins, most notably Bill Moss and Tim Shields. Bill was a fantastic travel companion, and his practice of unexpectedly asking for updates during weekly group meetings kept me intimately familiar with my data. It was a joy getting to know him, and I hope someday I can have as full a life as he has (although I would like a bit more sleep). Tim helped me make sense of spatial data, both in the classroom and in meetings, and equipped me with the skills to analyze it independently. He also kept me nourished by alerting me to free food and provided expert advice for traveling west, which I will make use of just as soon as I submit this thesis. Supplementing all this professional support was the company of my peers, most especially that of Hannah MacLeod, Victoria Garcia, Ty Pan, Laura Canaday, and Abeer Sayeed. They gave my life some normalcy, shared in the frustrations of science and life, and, most importantly, made sure I ate regularly. Literally keeping a roof over my head while I pursued this degree, and ultimately who I have to thank for all of this, is my sister, Andrea HoffmAn. She has never stopped believing in me, and it has been so much fun to spend the last few years with her. Finally, I owe so much gratitude to the Macha Research Trust (MRT) team in Zambia for their work with me on this project, along with their help getting me settled in Zambia. Jenny Stevenson for her advice in the set up; Limonty Simubali for leading me through field collections and mosquito morphology; and Harry Hamapumbu for sharing his knowledge of the area and teaching me the ins and outs of study design. The collections used in this study were performed by both the MRT field teams, and none of this project could have been completed without their hard work and advice.

Table of Contents

Sections Page

Title Page ...... i

Abstract ...... ii

Acknowledgments ...... iv

Table of Contents ...... v

List of Tables ...... vi

List of Figures ...... viii

Introduction ...... 1

Research Aims ...... 14

Methods ...... 15

Results ...... 28

Discussion ...... 58

Future Directions ...... 66

Conclusions ...... 67

References...... 68

Appendices of Protocols...... 75

Curriculum Vitae...... 90

List of Tables

Table 1. Summary of trap-nights excluded from analysis under each collection scheme.

Table 2. Outdoor molecular species composition of Anopheles in Collection Scheme I.

Table 3. Indoor molecular species composition of Anopheles in Collection Scheme I.

Table 4. Impact of household factors on indoor presence of An. squamosus at first visit

in Collection Scheme I.

Table 5. Impact of household factors on indoor presence of An. squamosus at all three

visits in Collection Scheme I.

Table 6. Impact of household factors on outdoor presence of An. squamosus at first visit

in Collection Scheme I.

Table 7. Impact of household factors on outdoor presence of An. squamosus at all three

visits in Collection Scheme I.

Table 8. Results of Plasmodium assays for JHSPH samples from Collection Scheme I.

Table 9. Outdoor morphological species composition of Anopheles in Collection Scheme

II.

Table 10. Indoor morphological species composition of Anopheles in Collection Scheme

II.

Table 11. Impact of household factors on indoor presence of An. squamosus in

Collection Scheme II.

Table 12. Impact of household factors on outdoor presence of An. squamosus in

Collection Scheme II.

Table 13. Impact of household factors on outdoor abundance of An. squamosus in

Collection Scheme II.

Table 14. Molecular species composition of Anopheles in JHSPH Samples from

Collection Scheme II.

Table 15. Results of Plasmodium assays for JHSPH samples from Collection Scheme II.

Table 16. Individual sample results of blasting NCBI database with sequences from

JHSPH samples from Collection Scheme I amplifying a fragment larger than

1,000 bp using the ITS2 PCR described in Appendix B.

Table 17. Details of sequenced JHSPH samples.

vii

List of Figures

Figure 1. Images of An. squamosus collected in Macha area in May 2018 by CDC LT.

Figure 2. Distribution of An. squamosus in Africa.

Figure 3. Map of study area delineating households included in this study.

Figure 4. Weekly average rainfall and minimum and maximum temperatures during the

collections analyzed in this study.

Figure 5. Distribution of households included in this study.

Figure 6. Map delineating the households included in the study from Collection Scheme

I: Enhanced Step D.

Figure 7. Distribution of traps under Collection Scheme I.

Figure 8. Distribution of traps under Collection Scheme I that collected An. squamosus.

Figure 9. Results of 12S mtDNA PCR (Appendix G) for female JHSPH samples from

Collection Scheme I.

Figure 10. Map delineating the households included in the study from Collection Scheme

II: Anopheles funestus Transect.

Figure 11. Distribution of traps under Collection Scheme II.

Figure 12. Distribution of traps under Collection Scheme II that collected

An. squamosus.

Figure 13. Results of 12S mtDNA PCR (Appendix G) for JHSPH samples in Collection

Scheme II.

Figure 14. Imaged 2% Agarose Gel (Appendix E) showing 330 base pair fragments

expected for An. squamosus using the PCR described in Appendix B.

viii

Figure 15. Imaged 2% Agarose Gel (Appendix E) showing 330 base pair fragments

expected for An. squamosus, as well as 3 unexpected >1000 base pair

fragments, using the PCR described in Appendix B.

Figure 16. COI Bayesian tree showing clade structure within An. squamosus and genetic

relatedness of An. sp. 15.

Introduction

According to the World Health Organization (WHO), vector-borne diseases are responsible for more than 17% of infectious diseases in humans annually and lead to more than 700,000 deaths per year. The diversity of ecologies for each vector contributes to the widespread occurrence of vector-borne diseases, putting billions of people at risk all over the world. Of the 26 major vector-borne diseases, 11 are transmitted by mosquitoes. Three genera, Aedes, Culex, and Anopheles, transmit most of the mosquito- borne pathogens to humans. Aedes mosquitoes are best known for transmission of arboviruses such as Chikungunya virus, dengue virus, yellow fever virus, and Zika virus.

Culex mosquitoes are notorious for transmitting West Nile virus, both western and eastern equine encephalitis virus, Venezuelan equine encephalitis virus, and filaria, the cause of lymphatic filariasis. Anopheles mosquitoes are also associated with lymphatic filariasis. Anophelines are, however, inextricably responsible for transmission of

Plasmodium, the causative agent of human malaria and the vector-borne disease responsible for the most morbidity and mortality worldwide [1].

The Parasite

Malaria is the name for the human disease caused by protozoans of the genus

Plasmodium. The five species that have been shown to infect humans are Plasmodium malariae, P. ovale, P. knowlesi, P. vivax, and P. falciparum. The latter two are the dominant species worldwide, with P. vivax being prominent in South and Central

America and Asia and P. falciparum being the most common in sub-Saharan Africa [2].

All five species are transmitted by female Anopheles mosquitoes. Though the dominant

vector species varies geographically, the capacity to vector Plasmodium is limited to approximately 80 species out of over 450 species of Anopheles [3, 4].

When an infective mosquito bites an uninfected human host, it deposits the sporozoite stage of the parasite into the host’s skin and blood with its own saliva. The sporozoite then travels through the circulatory system to the liver, where it infects hepatocytes and replicates intracellularly. After proliferation in the liver, the parasite matures to the merozoite stage, exits the liver, and invades erythrocytes. Within the erythrocyte, parasites undergo further development into the ring stage, trophozoite stage, schizont form, and finally gametocyte stage. This final stage, which occurs 7 to 30 days after mosquito bite and infection [5], is infective to new mosquitoes. When a host with a gametocyte-stage infection is fed on by an anopheline, the gametocytes in the blood are picked up by the mosquito during feeding. The gametocytes undergo sexual reproduction within the mosquito midgut, develop into ookinetes, and traverse the midgut epithelium to form an oocyst on the outside of the midgut epithelium. Within that oocyst, they mature into sporozoites, proliferate, exit the oocyst into the hemolymph and invade the mosquito salivary glands to infect the next host. The entire process within the mosquito takes 9 to 18 days [2].

In humans, the disease is characterized by an enlarged spleen, fever, and malaise.

Plasmodium falciparum, the species of interest in this discussion, can also cause severe symptoms such as severe anemia, kidney failure, and neurologic impairment due to its ability to adhere to endothelial cells lining the vasculature [5]. Descriptions like this in ancient records have provided evidence of the parasite’s lengthy relationship with

humans, as well as human’s lengthy efforts to control it [6-8]. Despite this long history, discovery of the mosquito vector did not occur until the end of the 19th century.

The Vector

The Anophelinae subfamily of the Dipteran family Culicidae (mosquitoes) contains three genera; the genus Anopheles is exclusively associated with human

Plasmodium transmission and comprises approximately 450 species. The genus is further divided into seven subgenera, of which most African malaria vectors fall into two:

Anopheles and Cellia. Within each subgenus, there are series; these series are then divided into groups; these groups can then be divided into complexes [9, 10]. The complex phylogeny of Anopheles makes it easy to imagine the diverse ecological preferences and behavioral tendencies of the many species that exist within the genus [4].

Despite these differences, many features of the Anopheles mosquito are shared between all species.

All Anopheles are reproductively obligate blood feeders, therefore, only the females take a blood meal [4]. Though they feed on a variety of hosts, most species have at least regional host preferences. Primary malaria vectors typically feed during the late night hours [4, 11]. Many of the historically dominant malaria vectors are endophagic, prefer feeding indoors, and endophilic, prefer resting indoors [4, 11]. Once a female mosquito has blood fed, she typically must rest in order to digest the blood meal; for most species she is too heavy to fly very far [4]. Once the blood is digested, she has the nutrients she needs to produce eggs. Having already mated, often just after eclosion, she lays her eggs. Although many species prefer to lay eggs in a sunlit pool or slow-moving stream with some vegetation [12], there is great diversity in breeding site preference

among Anopheles [4], with evidence of female anophelines using a wider variety of sites when ideal breeding sites are scarce [13, 14]. After just 2-3 days, the eggs hatch into larvae which feed on detritus, algae, and phytoplankton in the water column [4]. The larvae pupate after approximately 7 days and eclose into adults in another 2-4 days [4].

Although research on mosquito migration patterns is ongoing, it has historically been understood that most anophelines do not fly much farther than 2-3 km from their breeding site [4].

In order for a species of the genus Anopheles to be a vector of human malaria, it must first demonstrate the selection of humans as a host. Many of the dominant vector species, such as An. gambiae s.s. and An. funestus s.s., are anthropophilic, in that they have a preference for human hosts. Other species that feed primarily or solely on animals are considered zoophilic; these species are often considered to be of negligible importance in the malaria transmission cycle, as their biting rates on humans are too low to support parasite transmission even if the species itself is biologically competent for

Plasmodium development. As control methods targeting the behaviors of dominant vectors become implemented successfully, however, it has become increasingly apparent that other anopheline species can sustain malaria transmission by exhibiting behaviors that evade control measures. Identification of these lesser studied species, particularly species difficult to distinguish morphologically such as those within groups or complexes, has been made possible by the development of molecular techniques. Before polymerase chain reactions and sequencing and even the development of morphological keys, the man who discovered the role of mosquitoes in malaria transmission recorded

only two distinguishing features of the mosquitoes in his experiments: “grayish” and

“brownish”[15, 16].

Historical Timeline of Malaria

The earliest evidence of malaria in written record appears in medical texts from ancient China; though there is controversy over the date of their writing, some estimations place it as early as 2697 BCE [6]. The first written recommendation for treatment of the disease also came from China, which after clinical research in the 1970s led to the discovery of artemisinin [7]. Quinine entered European historical record in

1631 CE as a treatment for malaria derived from a tree bark in Peru [8]. The parasite itself was not discovered until 1880, when Charles Laveran found the parasites in a blood smear from a patient at a hospital in Algeria [17]. In 1897, Ronald Ross found the parasites in the midguts of two mosquitoes of a “brown species” [15]. He went on to demonstrate transmission of Plasmodium, referred to as proteosoma, in birds during the deposition of saliva by mosquitoes during biting [16].

Mosquito control efforts accelerated soon after. Dichloro-diphenyl- trichloroethane (DDT), used as an insecticide against lice in the early 1940s, was discovered to have long-lasting insecticidal properties against Anopheles larvae and adults [18]. The marked reduction of malaria in Greece after the sustained campaign of rapid treatment of malaria and the application of DDT motivated the launch of the Global

Malaria Eradication Programme (GMEP) by the World Health Organization (WHO) in

1955 [19]. Guidelines focused on strategies heavily reliant on DDT [19]. In 1969, although over 30 countries had achieved elimination and many others significant reductions in both incidence and mortality, persistence of malaria in many remaining

countries prompted the WHO to lower the bar to control rather than elimination [19].

Without clear guidelines, however, national commitment to control waned, and surveillance captured a notable global increase in malaria-related deaths in the 1970s and

1980s [19, 20].

While some countries continued to progress to elimination at a national level, global efforts stagnated until 1992 [21]. The adoption of the Global Malaria Strategy, with participation from African nations placing a new focus on malaria on the African continent, reestablished the disease as a global priority [21]. With the late 1990s came the proliferation of both international and regional organizations, programs, and commitments to reduce the burden of malaria, once again focusing on the tools of vector control, rapid diagnosis, and treatment [21]. Added to the toolbox this time were insecticide-treated nets (ITNs), additional drugs and diagnostics, and an enormous increase in funding as global support gathered behind the efforts [21]. Since 2000, global malaria cases have declined by 37%, and malaria mortality has declined by 60% [22].

582 million ITNs were distributed in malaria endemic countries from 2014-2016, increasing the household ownership and use from 50% in 2010 to 80% in 2016 [22]. With the tremendous success witnessed in the late 2000s and early 2010s, many countries set goals of elimination within their borders within the next 20 years. One of those countries, with an elimination deadline of 2021, is Zambia.

Malaria in Zambia

Malaria control in Zambia began in the 1940s in the form of annual residual spraying, and it remained effective in some areas until the late 1970s [23]. The disease, a leading cause of death in Zambia, was reduced to an incidence of 121.5 cases per 1000

people per year [23]. In the 1980s, however, a weakening economy and subsequent decentralization of efforts led to a resurgence [24]; in 1999, incidence rates were 321.4 cases per 1000 people per year [23]. In addition to the lapse in vector control, Zambia was relying on single-drug treatment of cases using chloroquine, which contributed to the spread of drug resistance [24].

When the Roll Back Malaria Partnership began in the late 1990s, Zambia reinvigorated its efforts to control malaria within its borders [24]. Funding from several international donors, including the World Bank and the President’s Malaria Initiative, helped Zambia see tremendous reductions in malaria prevalence over the next decade

[24]. Zambia’s strategy focused on widespread distribution of long-lasting insecticide- treated nets (LLINs) in addition to switching to artemisinin combination therapy (ACT) and the use of rapid diagnostic tests (RDTs) [24]. Following the introduction of ACT in

2004, pediatric admissions to Macha Hospital for malaria in southern Zambia dropped to

50 cases per year from 1,400 in the 2000-2001 rainy season [25]. National prevalence among children under five decreased by half from a baseline of 22% from 2006 to 2008

[24].

These reductions in malaria in Zambia, however, have been regional. Zambia’s

Southern Province, along with its capital Lusaka, have seen most of the reductions over the last two decades, reaching malaria incidence rates of 26.7 cases per 1,000 people and

31.3 cases per 1,000 people, respectively, by 2015. All of the remaining provinces had incidence rates of over 200 cases per 1,000 people in 2015, with many exceeding 300 and

North Western Province as high as 806 [26]. Contributing to these differences are resources, which impacts the pressures the parasite faces, and climate, which shapes the

habitat of the vectors. Zambia recently tested a strategy using community health workers

(CHWs) to perform home visits within their villages starting in Southern Province. This strategy improved detection of malaria, particularly subclinical malaria, resulting in more rapid treatment. The strategy, however, overwhelmed the capacity of rural health centers, and an evaluation by the Southern and Central Africa International Center for Excellence in Malaria Research (ICEMR) found that it only identified 22% of infected individuals

[27]. Perhaps more importantly, the climates of the two provinces are significantly different. Choma District in Southern Province receives the lowest mean annual rainfall in the country [28] and a malaria prevalence of 1-2%. The region experiences three distinct seasons: cool and dry (May-July), hot and dry (August-October), and wet

(November-April) [24], with malaria transmission confined to the wet season. In contrast,

Luapula Province, an area with a malaria prevalence of 50.5% in 2010 despite implementation of control strategies, has year-round breeding sites for Anopheles mosquitoes, contributing to a maintenance of malaria transmission even in the dry season

[24].

In both provinces, the NIH-funded Southern and Central Africa ICEMR is collaborating with local teams to characterize and explain those differences to better inform the National Malaria Elimination Program (NMEP). While the high prevalence in

Luapula Province presents a clear and urgent need for adaptations, sustained low prevalence in Southern Province also poses an obstacle to elimination efforts. Despite a significant decline in malaria prevalence and mortality in Southern Province, prevalence still sits at around 1-2%. Much of this decline was likely due to the introduction of ACT

concurrent with a drought in 2004-2005 that dramatically reduced vector populations

[24].

Persistent Malaria Transmission

Southern Province is not alone; persistence and re-emergence of malaria transmission are significant challenges to control efforts around the world [29, 30]. While in some areas vector control is ineffective due to insufficient resources, in other areas, malaria persists despite sustained and widespread implementation of standard control methods. Cross-border malaria transmission, or even movement within a country, can challenge localized control strategies [31-33]. Single drug treatment of malaria has resulted in the spread of resistant parasites that require combination drug therapies [34].

Anopheline populations and even species have also undergone change in response to prolonged vector control efforts. Resistance to insecticides has been documented in all major Anopheles vector species, some species demonstrating resistance to multiple insecticides simultaneously [22]. Other species have shown a shift in behavior, with higher proportions foraging outdoors or at times when humans are outside of the bed net

[35-37]. On a population level, species composition is changing in many areas with stable vector control, with species that were once considered primary vectors on the decline.

In some cases, as these primary vectors are successfully controlled, the role of other species in malaria transmission in a region gains significance. These species, also known as secondary vectors, species largely considered to play only a minor role in transmission, are increasing in recognition. Like that of more prominent vectors such as

An. arabiensis, the role of secondary vectors in malaria transmission is regionally

variable, with species testing positive for Plasmodium sporozoites and displaying high rates of anthropophily in some areas and testing negative for both sporozoites and human blood meal in others [38-40]. Although secondary vectors are not well researched, they are thought to exist in undefined species complexes. Understanding the ecology and bionomics of these newly recognized vectors and elucidating their genetic structure is increasingly important to reducing and even eliminating malaria transmission on a regional basis. Many of these secondary vectors evade existing indoor-targeted control methods by foraging and resting outdoors, thus maintaining malaria transmission even at low levels and making elimination goals unachievable. Preventing malaria transmission by these secondary species will require study of their foraging and resting behavior, and adjustment to current strategies.

In the Macha area, the anopheline population has yet to exhibit insecticide resistance or behavioral resistance in the primary vector, but it has undergone some shifts in its species composition [41]. Prior to the 2004-2005 drought, the primary vectors were

An. arabiensis and An. funestus s.s. After the drought, An. arabiensis remained; An. funestus, however, seemed to have disappeared [42]. An. arabiensis, while those that exhibit endophily have been found to have a human blood index (HBI) of 0.923 in

Southern Province[43], can exhibit zoophily that reduces their vectorial capacity in comparison with An. funestus [42]. That shift has been hypothesized to have contributed to a decline in malaria prevalence in the region; the persistence, however, may be due in part to the activity of secondary vectors [24]. Several species present in the Macha area have been implicated as secondary vectors in other regions, including An. coustani, An. rufipes, and An. squamosus [44-46]. Although An. rufipes has been demonstrated to be

contributing to malaria transmission in Cameroon, the species appears to be highly zoophilic in the Macha area and thus has not yet been implicated as a secondary vector in the region [44]. In 2010, however, the ICEMR team found An. coustani and An. squamosus to be unexpectedly anthropophilic, with 86% and 65% of blood fed samples testing positive for human DNA, respectively [38]. Although no samples of either species had been found to contain Plasmodium parasites at the time, the level of anthropophily suggested that there was potential risk that either species could be acting as a secondary vector of malaria. In 2016, that risk became a reality for An. squamosus; six samples collected in the Macha area during the 2015 rainy season tested positive for P. falciparum sporozoites using an ELISA for the circumsporozoite protein. Four of those also tested positive for P. falciparum DNA via qPCR [39].

Anopheles squamosus

Although Plasmodium sporozoites have been detected in the salivary glands of

An. squamosus and anthropophilic behavior has been demonstrated in both Zambia and in

Madagascar, the species has not been reported to play an important role in malaria transmission to date [38, 40, 45, 47-49]. Research on the bionomics of An. squamosus, therefore, has been minimal, largely due to its dominant zoophilia and the failure of early dissections to demonstrate sporozoites in the salivary glands [50]. Much of the research in the last fifty years has been on its role in transmission of Rift Valley fever virus

(RVFV), an important virus as it plagues cattle and other agricultural animals in sub-

Saharan Africa. In recent years, however, RVFV has been recognized as an emerging disease among humans, and research on it the field of public health is building [51].

Understanding the ecology and behavior of An. squamosus could thus contribute both to eliminating residual malaria transmission and preventing new arboviral epidemics.

Figure 1. Images of An. squamosus collected in Macha area in May 2018 by CDC light trap. The species is notably dark and hairy. In the image on the left, note the lateral tufts along the abdomen. On the right, note the distinctly shaggy palps.

As a species, An. squamosus was first described in 1901 as “A very pronounced and scaly species, not like any other Anopheles I have seen” [52]. Wellman described it as a “striking object” and “a rather bold anopheline” [53]. As early as 1903, Theobald noted slight morphological differences between mosquitoes caught in different regions that he still recorded as An. squamosus [54], indicating even then the existence of a possible species group or complex. Adults of the species demonstrate regional differences in behavior; while the species is often described as being uncommon in houses, in some areas it is caught more often indoors than the recognized malaria vector in the area [55-

62]. This species has been found to have a broad distribution across sub-Saharan Africa and has even been found on the Arabian Peninsula [63, 64] (Figure 2). This mosquito’s marked absence in indoor resting collections, in addition to an apparent lack of impact of

indoor residual spraying (IRS) on its numbers, indicate that it is a strongly exophilic species [50, 59, 60, 62, 65-67].

Although much of the data supports the strong zoophilic behavior by An. squamosus [3, 13, 61, 68], some collections exhibited more opportunistic foraging behaviors [61, 69] and others high rates of anthropophily [38, 40]. Though 75% of An. squamosus collected in Madagascar in 1953 tested positive for human blood, all dissections then and prior to that were negative for Plasmodium oocysts or sporozoites

[40, 56]. In the 1960s, however, Gillies detected sporozoites in An. squamosus samples, confirming for the first time a potential role for the species in malaria transmission [11,

45]. Despite the development of less labor and time intensive techniques for detecting

Plasmodium in mosquitoes since then, primarily with the use of ELISAs for sporozoite detection and PCR/qPCR for detecting parasite DNA, there were no additional reports of

An. squamosus carrying Plasmodium until 2016, when the Southern and Central Africa

ICEMR team reported six infectious samples in southern Zambia [39].

Figure 2. Distribution of An. squamosus in Africa. Not pictured: Report of An. squamosus on Arabian Peninsula [64].

Research Aims

The goals of this thesis are to evaluate the role of An. squamosus in malaria transmission in southern Zambia in two seasons in 2018, characterize its local behavior and ecology in two distinct study areas, identify genetic structure within the species, and analyze factors associated with the presence of the species at a household, abundance, rate of infectivity, host choice, and foraging behavior.

Methods

Figure 3. Map of study area delineating households included in this study. Households are colored according to the collection scheme under which they were sampled.

Study Area

All households in the study were within an 81 km radius of Macha Research Trust

(MRT), a field station partnered with the Johns Hopkins Malaria Research Institute

(Figure 3). MRT is located at 16.39292° S, 26.79061° E, within Zambia’s Southern

Province in Choma District at an elevation of 1,100 meters above sea level. The ecotype around the field station is primarily miombo woodland, a combination of dry savannah and woodland. The region experiences three seasons: a cool dry season (typically from

May through July), a hot dry season (typically from August through October), and a rainy season (November through April).

Choma District Rainfall and Temperatures During Weeks of Collection 180 40

160 35

140 30 120 25 100 20 80

Rainfall (mm) Rainfall 15 60 10

40 (Celsius) Temperature

20 5

0 0 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 Weeks from December 1, 2017

Rainfall Minimum Temperature Maximum Temperature

Figure 4. Weekly average rainfall and minimum and maximum temperatures during the collections analyzed in this study.

Data Collection

Mosquitoes and associated household data for this study were collected under two different collection schemes. All mosquitoes were collected between the hours of 6 PM and 6 AM using CDC light traps set up either inside a sleeping house near someone sleeping under a bed net or next to an animal pen.

Collection Scheme I: Enhanced Step D

Households in this collection scheme were selected as part of a reactive test and treat program. When an individual came to a clinic and tested positive for malaria, they were invited to participate in the study. If the individual consented to enrollment, the

MRT team traveled to the homestead of that individual, administered a questionnaire, tested every household member for malaria using RDTs, and collected dried blood spots from each member. The team also went to every homestead within 250 meters of the original (index) household and, upon enrollment in the study, tested every member of those households for malaria using RDTs and collected dried blood spots. Entomological collections were also carried out at each enrolled household using CDC light traps [70-

72]. If there was no animal pen at the household, just one trap was placed at that household inside the primary sleeping house. The team returned to each household 30 days and 90 days after the initial visit to repeat data and entomological collections.

All entomological samples in this collection were morphologically identified and molecular species confirmation is complete. These samples were collected from

December 2017 through June 2018. Molecular identifications of all samples were used in analyses. Samples collected in February 2018 were selected for additional analysis in this study, because at the time of sample transport they represented an unprocessed set of specimens collected during peak malaria season that was dominated by An. squamosus.

They were transported to JHSPH in Baltimore, Maryland for completion of morphological identification and additional molecular analysis. These will be referred to as the JHSPH samples from Collection Scheme I for the remainder of this thesis.

Collection Scheme II: Anopheles funestus Transect

Households under this collection scheme were selected by proximity to pre- determined distances along a transect that extended from the Macha catchment area to

Lake Kariba. As this study was designed to capture outdoor and indoor foraging, for a household to be included in the scheme, it had to have an animal pen at which to set a trap. All households with animal pens along the transect were enumerated using satellite imagery. After enumeration, 12 households were randomly selected in each village.

Community health workers were provided with maps indicating each household, and they were tasked with getting consent from the heads of the households. The field team then went out and visited each household, administered a questionnaire, and set up two CDC light traps per household, one inside the primary sleeping house and the second next to an animal pen. The field team also collected information about breeding sites within 500 meters of the household, including visiting those sites, recording their coordinates, and collecting any larvae present.

All entomological samples in this collection were morphologically identified.

These samples were collected in May and June 2018. Morphological identifications of all samples were used in analyses. At the time of sample transport, only the first 1500 samples were morphologically identified. As this study was focused on An. squamosus, all samples that had thus far been morphologically identified as An. squamosus were transported to JHSPH in Baltimore, Maryland for additional molecular analysis. These will be referred to as the JHSPH samples from Collection Scheme II for the remainder of this thesis.

Number of Households Sampled per Week 40

NumberHouseholds of 1 3 6 7 9 10 12 13 14 15 17 18 19 20 21 22 23 24 25 26 27 28 29 Weeks from December 1, 2017

Enhanced Step D Anopheles funestus Transect

Figure 5. Distribution of households included in this study. Bars bordered in black include samples transported to and processed in Baltimore, Maryland.

Sample Processing

Mosquito samples were handled and processed similarly for both collection schemes. After collection, trap cups were placed in a -20C freezer for at least 30 minutes.

Insects were then separated by genus. Anophelines were sorted into 0.5ml tubes with silica beads and cotton or paper towels individually. Culicine mosquitoes were sorted into

0.5ml tubes with silica beads and cotton or paper towels in batches of 10. Counts of anophelines and culicines were recorded by household; anophelines were also distinguished by sex and if visually blooded. All other insects were discarded.

Individual anophelines were morphologically identified by trained staff using a morphological key [73]. Each anopheline was then split into two, one tube containing the head and thorax and a second containing the abdomen; wings and legs were included with the head and thorax except in the case of missing abdomens due to damage during collection. All split mosquitoes were then stored at -80C until molecular processing.

Molecular Analysis

Heads/Thoraces

Head/Thorax samples were stored at -80oC prior to homogenization. They were then homogenized following a protocol adapted from the Malaria Research and

Reference Reagent Resource Center (MR4) Methods in Anopheles Research Manual

(http://www.mr4.org/Publications/MethodsinAnophelesResearch/tabid/336/Default.aspx)

[74] (Appendix H). Each homogenate was divided into two: one to be run on a CSP

ELISA, and the second to be stored for future gDNA extraction. Homogenates were stored at -20oC.

To test for the presence of P. falciparum sporozoites, ELISA homogenates were run on the CSP ELISA in pools of 5 samples per well on a 96-well plate (Appendix H).

The ELISA uses a monoclonal antibody (2A10) that binds to the circumsporozoite protein present on the surface of P. falciparum sporozoites. The 2A10 antibody is specific to P. falciparum when tested with other Plasmodium species, but has demonstrated some cross-reactivity to bovine blood. This ELISA has a limit of detection of 25 ng/ µl of CSP protein. Plates were read on the Aquamax 4000 (Molecular Devices,

San Jose, CA) using an absorbance of 405 nm. Following CDC protocol, wells with an

OD of at least 2x higher than the average of four negative control wells were deemed positive. Samples in pools with at least one positive replicate well were then run individually on a subsequent CSP ELISA.

If a sample had at least one positive replicate well on the individual CSP ELISA, or if the original pool that contained the sample had two positive replicate wells but none of the individual samples in the pool had a positive replicate well, the portion of the

homogenate set aside for DNA extraction was extracted using the Qiagen DNeasy Blood

& Tissue Kit (Qiagen, Hilden, Germany). The extracted sample was then run on an adapted SYBR-based qPCR targeting an 85 base pair (bp) fragment of the P. falciparum- specific single copy lactate dehydrogenase (PfLDH) gene [75]. Primers were as follows:

PfLDH F (5'- ACG ATT TGG CTG GAG CAG AT -3') and PfLDH R (5'- TCT CTA

TTC CAT TCT TTG TCA CTC TTT C -3'). Each reaction contained 8.05 µl of nuclease- free water, 0.225 µl of PfLDH F, 0.225 µl of PfLDH R, 12.5 µl of SyberGreen Master

Mix ABI, and 4 µl of DNA to reach a volume of 25 µl per reaction. Standards were serially diluted 10-fold five times using stock P. falciparum 3D7 DNA. Reactions were loaded onto a 96-well plate, covered with a sealing film, and placed into a StepOne Real-

Time PCR System (Thermo Fisher Scientific, Carlsbad, CA). The qPCR was run with the following settings: 50ºC for 2 minutes, 95ºC for 10 minutes, and 45 cycles of denaturation at 95ºC for 15 seconds and annealing at 60ºC for 1 minute. Samples were considered positive for P. falciparum by qPCR if at least one replicate had a quantity mean exceeding 1 and a melting temperature of 76ºC-77ºC (Appendix I).

Due to inconsistent results, a stringent protocol was established to standardize when a sample was considered infectious with P. falciparum. A sample must first have at least one positive replicate well when run individually on CSP ELISA. If both replicate wells are positive for that sample, the sample is considered infectious even if both replicates are negative on qPCR. If only one replicate well is positive for the sample on the individual CSP ELISA, that result must be confirmed by at least one positive replicate on qPCR.

Abdomens

Mosquito abdomens were homogenized individually, and DNA was extracted from each homogenate using a modified salt extraction [76](Appendix A). To confirm species, all samples were first run on a polymerase chain reaction (PCR) targeting the variable internal transcribed spacer 2 (ITS2) region that distinguishes between several species or groups of anophelines (Appendix B). Primers flanked the 5.8S and 28S genes,

ITS2A (5’-TGT GAA CTG CAG GAC ACA T-3’) and ITS2B (5’- TAT GCT TAA ATT

CAG GGG GT -3’) [77]. Due to consistent failure of An. squamosus samples to amplify with these primers [3], two primers were added that targeted a 330 bp fragment of the cytochrome oxidase I (COI) gene specific to An. squamosus among anophelines (Jones, unpublished): SQFor405 (5’- CCA TTT CCA TTA TGT CCT ATC TAT AGG -3’) and

SQRev707 (5’- GGG AAA GCA GGA GTT CGT TGA G- 3’). Each 25 µl reaction contained 2.5 µl of 10X PCR buffer, 200 micromolar of each dNTP, 0.3 µl of each primer (ITS2A, ITS2B, SQFor405, SQRev707), 2.0 units of Taq polymerase, 1.0 µl of

DNA template, and remaining volume with nuclease-free water. Products were amplified under the following thermocycler conditions: 2 minute initial denaturation at 94ºC, 40 cycles of 30 seconds at 94ºC, 30 seconds at 50ºC, and 40 seconds at 72ºC, and a 10 minute final extension at 72ºC. Products were then stored at 4ºC until gel electrophoresis.

All PCR products were run on a 2% agarose gel (Appendix E), electrophoresed at 150V for 45 minutes, and visualized using a Red Imager (ProteinSimple, San Jose).

Samples whose products from the ITS2 PCR described above were around 600 bp, potentially representing a member of the An. gambiae complex, were then run on a

PCR targeting the ribosomal DNA intergenic spacer region designed to distinguish

members of that species complex [78] (Appendix C). This modified PCR utilizes four primers to distinguish between An. gambiae s.s., An. arabiensis, and An. quadriannulatus: UN (5’- GTG TGC CCC TTC CTC GAT GT -3’), GA (5’- CTG GTT

TGG TCG GCA CGT TT -3’), AR (5’- AAG TGT CCT TCT CCA TCC TA -3’), and

QD (5’- CAG ACC AAG ATG GTT AGT AT -3’). Each 25 µl reaction contained 2.5µl of 10X PCR buffer, 200 micromolar of each dNTP, 3.0 µl of the AR and QD primers, 0.5 microliters of the GA primer, and 1.0 µl of the UN primer, 1.5 units of Taq polymerase,

1.0 µl of DNA template, and the remaining volume nuclease-free water. Products were amplified under the following conditions: 2 minute initial denaturation at 94ºC, 30 cycles of 30 seconds at 94ºC, 30 seconds at 50ºC, and 30 seconds at 72ºC, and a 7 minute final extension at 72ºC. Products were then stored at 4ºC until gel electrophoresis. All PCR products were loaded into a 2% agarose gel (Appendix E), electrophoresed at 150V for

45 minutes, and visualized using a Red Imager (ProteinSimple, San Jose).

Fifteen percent of the JHSPH samples from each collection scheme, ensuring spatial and temporal distribution, were run on a PCR targeting the 3’ Barcode of Life

(BOL) portion of the cytochrome oxidase I (COI) gene that can be used for sequencing most anopheline mosquitoes [79] (Appendix D). Five percent of the JHSPH samples were sent for sequencing due to a failure to determine species using PCR methods; the remaining 10% sent for sequencing were to confirm PCR species identifications and for later phylogenetic analysis. Samples that failed to amplify on COI, as well as samples without clear sequencing results, were run once more on ITS2 for sequencing. All PCR products were evaluated on 2% agarose gel (Appendix E) using gel electrophoresis and

imaged with a Red Imager (ProteinSimple, San Jose). Those samples that amplified were then processed further for sequencing as described below.

To determine blood meal source, all samples were run on a PCR targeting the 12S ribosomal RNA gene that signals presence of a blood meal [80] (Appendix F). This PCR utilizes two primers, a universal forward and a universal reverse, to detect vertebrate

DNA within Anopheles mosquitoes: UNFic (5’- GGA TTA GAT ACC CCA CTA TGC -

3’) and UNIRic (5’- GCT GAA GAT GGC GGT ATA TAG -3’). Each 25 µl reaction contained 2.5 µl of 10X PCR buffer, 200 µl of each dNTP, 0.3 µl of the primers, 2.0 units of Taq polymerase, 1.0 µl of DNA template, and the remaining volume nuclease-free water. Genomic DNA was amplified under the following conditions: 5 minute initial denaturation at 94ºC, 35 cycles of 30 seconds at 94ºC, 30 seconds at 51ºC, and 30 seconds at 72ºC, and a 7 minute final extension at 72ºC. Products were then stored at 4ºC until gel electrophoresis. All PCR products were loaded into a 2% agarose gel (Appendix

E), electrophoresed at 150V for 45 minutes, and visualized using a Red Imager

(ProteinSimple, San Jose). As all sources of blood meal amplify the same size of fragment using this PCR (205 bp), all samples that amplified had to be sequenced to identify the host.

Samples that amplified on the COI, ITS2, or 12S PCR meant for sequencing were purified using the QIAquick PCR Purification Kit (QIAGEN, Hilden, Germany).

Concentration was checked using the NanoDrop 2000 (Thermo Scientific, Wilmington).

Purified PCR product and their corresponding primers were diluted to 10ng/µl and 4µM, respectively, and sent to the Johns Hopkins Medical Institutions (JHMI) Synthesis and

Sequencing facility for sequencing. Resulting sequences were downloaded, assembled on

SnapGene Viewer 4.2.4, and pairwise aligned using Geneious Prime 2019.1.0. Pairwise alignments were performed using Geneious Alignment as a global alignment with free end gaps with a 65% similarity cost matrix, gap open penalty of 12, and gap extension penalty of 3. Resulting consensus sequences were blasted on NCBI’s nucleotide program.

Results were recorded into an Excel file with a threshold of at least 95% identity; preference was given to results with the highest identity and query cover.

Phylogenetic Analysis

Reverse complements were generated for all consensus sequences resulting from the pairwise alignments described above in Geneious Prime 2019.1.0

(https://www.geneious.com) to ensure all consensus sequences were ordered from 5’ to

3’. Reference sequences were downloaded from NCBI for each species in the analysis.

These consensus sequences were then multiple aligned with all other sequences of the same region. Multiple alignments were performed using Consensus Align as a global alignment with free end gaps with a 65% similarity cost matrix, gap open penalty of 12, and gap extension penalty of 3. The resulting alignment was then used to generate a

Bayesian phylogenetic tree using the Mr. Bayes plug-in in Geneious Prime. The tree was constructed using a GTR substitution model in combination with gamma rate variation.

Chains were run for 100,000 generations with a sub-sampling frequency of 200 and a burn-in length of 25,000. Branch lengths were unconstrained.

Statistical Analysis

Statistical analysis was performed to test for factors associated with the presence or absence of An. squamosus. Chi-squared tests were performed to test the association of each factor with the presence of An. squamosus in a trap. Because not all households in

Collection Scheme I were visited an equal number of times during the study period, analyses for Collection Scheme I were restricted to data collected on the first visit. This included data from only 20 households and thus 20 trap-nights indoors and 10 households and thus 10 trap-nights outdoors. Additional analyses were performed only on households that had received all three visits. This included data from 10 households and thus 30 trap- nights indoors and 6 households and thus 18 trap-nights outdoors. All households in

Collection Scheme II were included in these analyses. For comparisons of the presence of other anopheline species with the presence of An. squamosus, households not catching any anophelines were excluded. The number of traps excluded for these analyses are summarized in Table 1.

Although the restrictions in Collection Scheme I made the sample size too small for further analysis, data from Collection Scheme II was analyzed to test the association of household factors with the abundance of An. squamosus in an outdoor trap. Only traps catching any An. squamosus were included in this analysis; this included data from 33 trap-nights. The median number of An. squamosus in each trap was calculated. For each factor, Mood’s Median Test was performed to count the number of traps collecting a number of An. squamosus above and below the median in each group (with or without the factor). A Chi-squared test was then performed to test whether the difference in medians was significant.

Table 1. Summary of trap-nights excluded from analysis under each collection scheme. Households with no Anopheles collected were excluded only for analyses regarding presence of other Anopheles species.

Number of Trap-Nights Excluded Reason for Collection Scheme I: Enhanced Step D Collection Scheme exclusion II: Anopheles funestus Transect First Visit All Three Visits Indoor Outdoor Indoor Outdoor Indoor Outdoor No 11 1 18 7 83 41 Anopheles collected

Results

Collection Scheme I: Enhanced Step D

Figure 6. Map delineating the households included in the study from Collection Scheme I: Enhanced Step D. Entomological samples from the highlighted households in February 2018 were transported to JHSPH in Baltimore, Maryland for additional molecular analysis.

Under Collection Scheme I, a total of 24 households were sampled from

December 2017 to June 2018. A total of 96 traps were set over the course of 29 weeks.

Only 20 of those traps collected An. squamosus. The placement of all traps is described in

Figure 7; the placement of traps collecting An. squamosus is described in Figure 8.

Collection Scheme I: Enhanced Step D Distribution of Traps December 2017-June 2018

16, 17%

19, 20%

61, 63%

Sleeping House Goat Pen Cattle Kraal

Figure 7. Distribution of traps under Collection Scheme I. Numbers on pie chart reflect number of that type of trap and the percentage of that type among all traps.

Collection Scheme I: Enhanced Step D Distribution of Traps that Collected An. squamosus

7, 28%

10, 40%

8, 32%

Sleeping House Goat Pen Cattle Kraal

Figure 8. Distribution of traps under Collection Scheme I that collected An. squamosus. Numbers on pie chart reflect number of that type of trap catching An. squamosus and the percentage of that trap among all traps.

Species Composition

Under Collection Scheme I, a total of 943 anophelines were collected from

December 2017-June 2018. Morphological identifications were completed, but those results are not detailed here. All samples in this study from Collection Scheme I were molecularly confirmed, and thus these identities will be used in this study. Molecular species results are summarized in

Table 2 and Table 3. Anopheles squamosus dominated outdoors, representing 40.0% of collections (n=301). Other species collected outdoors included An. rufipes (23.6%, n=178), An. quadriannulatus (9.2%, n=69), An. gambiae s.l. (7.2%, n=54), An. arabiensis

(5.3%, n=40), An. coustani (4.9%, n=37), and An. longipalpis (3.6%, n=27). A remaining

6.2% (n=47) of specimens collected outdoors remained molecularly unidentifiable; of those, 40.4% were morphologically identified as An. squamosus (n=19), 27.7% as An. gambiae s.l. (n=13), 8.5% as An. rufipes (n=4), 4.3% as An. pharoensis (n=2), 4.3% as

An. pretoriensis (n=2), and 2.1% as An. coustani (n=1). Six samples (0.64%) were neither identifiable by morphological or molecular techniques.

Although 63% of traps were set indoors, only 20.2% of anophelines in Collection

Scheme I were collected indoors. Indoors, An. arabiensis dominated, representing 43.2%

(n=82) of collections. An. squamosus, however, was second most abundant, making up

27.9% (n=53) of collections. Other species collected indoors included An. gambiae s.l.

(6.3%, n=12), An. longipalpis (3.7%, n=7), An. coustani (3.2%, n=6), An. quadriannulatus (2.1%, n=4), and An. rufipes (1.6%, n=3). A remaining 12.1% (n=23) of specimens collected indoors remained molecular unidentifiable; of those, all but one

(n=22) were morphologically identified as An. gambiae s.l.; the remaining specimen

(n=1) was morphologically identified as An. squamosus.

Table 2. Molecularly confirmed outdoor species composition of Anopheles in Collection Scheme I over 35 trap-nights. Enhanced Step D Outdoor Molecular Species Composition Species Number Number per % of Total Collected Trap-Night Collection An. arabiensis 40 1.1 5.3 An. coustani 37 1.1 4.9 An. gambiae s.l. 54 1.5 7.2 An. longipalpis 27 0.8 3.6 An. quadriannulatus 69 2.0 9.2 An. rufipes 178 5.1 23.6 An. squamosus 301 8.6 40.0 Unidentified 47 1.3 6.2 Total 753 21.5 100.0

Table 3. Molecularly confirmed indoor species composition of Anopheles in Collection Scheme I over 61 trap-nights.

Enhanced Step D Indoor Molecular Species Composition

Species Number Number per % of Total Collected Trap-Night Collection An. arabiensis 82 1.3 43.2 An. coustani 6 0.1 3.2 An. gambiae s.l. 12 0.2 6.3 An. longipalpis 7 0.1 3.7 An. quadriannulatus 4 0.1 2.1 An. rufipes 3 0.1 1.6 An. squamosus 53 0.9 27.9 Unidentified 23 0.4 12.1 Total 190 3.1 100.0

Several household factors were analyzed for their impact on An. squamosus presence. Because not all households were sampled the same number of times, analyses were performed considering either only data from the first visit at each household or only data from households with all three visits. Within this study, 20 households had indoor traps set at first visit; 10 households had indoor traps at all 3 visits. At first visit, households that had received IRS and malaria mass drug administration (MDA) (n=3) within the last 6 months were more likely to have An. squamosus indoors. Although the difference in proportions of indoor traps catching An. squamosus between those catching

An. arabiensis and those not catching An. arabiensis approaches significance among households that were visited 3 times, only one indoor trap catching any anophelines did not catch An. arabiensis. Results for indoor trap analyses can be found in Table 4 and

Table 5.

The same factors were analyzed for association with presence of An. squamosus outdoors, with the addition of type of animal pen. Within this study, 10 households had outdoor traps set at first visit; 6 households had outdoor traps at all 3 visits. At first visit, outdoor traps catching An. arabiensis were more likely to also catch An. squamosus when compared to traps catching other anophelines. Among outdoor traps that were set 3 times, those catching An. coustani were more likely to also catch An. squamosus when compared to traps catching other anophelines. Households at which all members were protected by an ITN were less likely to catch An. squamosus outdoors. Results for outdoor trap analyses can be found in Table 6 and Table 7. Due to the small sample size of households either at first visit or with all three visits catching An. squamosus, further

analyses examining factors associated with abundance and relative abundance compare with other anophelines could not be performed.

Table 4. Results from Chi-squared tests evaluating impact of household factors in Collection Scheme I on An. squamosus presence indoors at a household with 1 degree of freedom. This analysis is restricted to the first visit at each household, which limited the number of households and trap-nights to 20. *Index refers to household with initial malaria case that prompted clinic visit and thus inclusion in Collection Scheme I. **Refers to the presence of the species. ***Full ITN Coverage refers to all members in the household sleeping under a bed net the night of collection. †Both IRS and Malaria MDA refer to treatment (insecticide application or mass drug administration) within the last 6 months.

Association of Household Factors with Presence of An. squamosus Indoors at First Visit

Collection Scheme I

Factors % of trap- % of trap- % of trap- Chi-squared P-value

nights with nights with nights without Test

factor factor factor catching

catching An. An. squamosus

squamosus

Index* 45.0 22.2 18.2 0.1 0.8

An. rufipes** 11.1 100.0 37.5 0.0 0.9

An. arabiensis** 75.0 50.0 0.0 0.2 0.7

An. gambiae s.l.** 66.7 50.0 33.3 0.2 0.6

An. coustani** 0.0

Thatched Roof 40.0 12.5 25.0 0.0 0.9

Open Eaves 70.0 21.4 16.7 0.1 0.8

Full ITN Coverage*** 75.0 26.7 0.0 0.4 0.5

IRS† 15.0 66.7 11.8 2.0 0.2

Malaria MDA† 15.0 66.7 11.8 2.0 0.2

Table 5. Results from Chi-squared tests evaluating impact of household factors in Collection Scheme I on An. squamosus presence indoors at a household with 1 degree of freedom. This analysis is restricted to households who received 3 visits, which limited the number of households to 10 and trap-nights to 30.

Association of Household Factors with Presence of An. squamosus Indoors Over Three

Visits

Collection Scheme I

Factor % of trap- % of trap- % of trap- Chi- P-value

nights with nights with nights without squared

factor factor factor catching Test

catching An. An. squamosus

squamosus

Index 80.0 20.8 16.7 0.1 0.8

An. rufipes 16.7 50.0 50.0 0.0 1.0

An. arabiensis 88.9 50.0 0.0 0.9 0.3

An. gambiae s.l. 75.0 44.4 66.7 0.4 0.5

An. coustani 25.0 33.3 55.6 0.4 0.5

Thatched Roof 53.3 18.8 21.4 0.0 0.9

Open Eaves 83.3 24.0 0.0 0.4 0.5

Full ITN 73.3 22.7 12.5 0.0 0.9

Coverage

IRS 10.0 0.0 22.2 0.0 0.9

Malaria MDA 10.0 0.0 22.2 0.0 0.9

Table 6. Results from Chi-squared tests evaluating impact of household factors in Collection Scheme I on An. squamosus presence outdoors at a household with 1 degree of freedom. This analysis is restricted to the first visit at each household, which limited the number of households and trap-nights to 10.

Association of Household Factors with Presence of An. squamosus Outdoors at First Visit

Collection Scheme I

Factors % of trap- % of trap- % of trap- Chi-squared P-value

nights with nights with nights Test

factor factor without

catching factor

An. catching

squamosus An.

squamosus

Index 50.0 20.0 20.0 0.0 1.0

An. rufipes 66.7 50.0 0.0 0.1 0.8

An. arabiensis 40.0 100.0 0.0 1.7 0.2

An. gambiae s.l. 50.0 66.7 0.0 0.8 0.4

An. coustani 0.0

Goat Pen 50.0 20.0 20.0 0.0 1.0

Full ITN Coverage 70.0 14.3 33.3 0.5 0.5

Table 7. Results from Chi-squared tests evaluating impact of household factors in Collection Scheme I on An. squamosus presence outdoors at a household with 1 degree of freedom. This analysis is restricted to households with all 3 visits, which limited the number of households and trap-nights to 18.

Association of Household Factors with Presence of An. squamosus Outdoors Over Three Visits

Collection Scheme I

Factor % of trap-nights % of trap- % of trap- Chi-squared Test P-value

with factor nights with nights

factor without factor

catching An. catching An.

squamosus squamosus

Index 83.3 40.0 66.7 0.0 0.8

An. rufipes 54.5 83.3 60.0 0.0 0.8

An. arabiensis 42.9 100.0 75.0 0.9 0.4

An. gambiae s.l. 63.6 71.4 75.0 0.0 0.9

An. coustani 54.5 100.0 40.0 2.4 0.1

Goat Pen 50.0 55.6 33.3 0.2 0.6

Full ITN 61.1 27.3 71.4 1.8 0.2

Coverage

Host Choice

The host choice of the female anophelines in the JHSPH samples from Collection

Scheme I (47.2%, n=207) was analyzed using a PCR targeting the host 12S mitochondrial DNA (mtDNA). Households included in this analysis are highlighted in

Figure 6. Results of the PCR are summarized in Figure 9. While An. squamosus, An. coustani, An. quadriannulatus, and An. arabiensis were all represented in those with detectable blood meals, only An. arabiensis (n=2) had detectable human blood meals.

Both An. arabiensis positive for human blood meals were collected indoors. Of the An. squamosus samples with detectable blood meals, 39.0% (n=23) had fed on cow and

61.0% (n=36) on goat. None of the An. squamosus collected indoors analyzed (n=20) had detectable blood meal. Although An. squamosus has been shown to take blood from multiple hosts, our methods did not allow for identification of mixed blood meals [81].

Collection Scheme I: Enhanced Step D Blood Meal Results 100% 2 1 36 80% 1 13 23 60% 9 5 116 40% 1 20%

0% Percent Samples of Percent

Species

Undetectable Cow Goat Human

Figure 9. Results of 12S mtDNA PCR (Appendix G) for female JHSPH samples from Collection Scheme I.

Detection of Plasmodium falciparum

Female anophelines in the JHSPH samples from Collection Scheme I (n=207) were tested for the presence of P. falciparum using an ELISA and a qPCR described above. Complete results are reported in Table 8. Samples are recorded as CSP ELISA positive if at least one replicate well was positive. Ten samples, 9 of which were An. squamosus, were found to have at least one positive replicate well. Using the conservative definition of infectiousness requiring either two positive replicate wells or a confirmation with a positive qPCR replicate, however, none of these samples were infectious with P. falciparum.

Table 8. Results of two Plasmodium assays run on female JHSPH samples from Collection Scheme I. The CSP ELISA detects circumsporozoite protein of Plasmodium falciparum and PfLDH qPCR detects P. falciparum DNA in the head and thorax of the mosquito. CSP ELISA positives indicate the absorbance level of at least one replicate well exceeded 2x the average absorbance level of negative controls.

Collection Scheme I: Enhanced Step D Plasmodium Assay Results Molecular Species Number CSP ELISA PfLDH Molecularly Positive qPCR Tested Positive

An. arabiensis 11 1 0 An. coustani 2 0 0 An. quadriannulatus 6 0 0 An. squamosus 175 9 0 Unidentified 13 0 0 Total 207 10 0

Collection Scheme II: Anopheles funestus Transect

Figure 10. Map delineating the households included in the study from Collection Scheme II: Anopheles funestus Transect. Entomological samples from the highlighted households that were morphologically identified as An. squamosus were transported to JHSPH in Baltimore, Maryland for molecular analysis. All the highlighted households were sampled in May 2018.

Under Collection Scheme II, a total of 126 households were sampled in May and

June 2018. A total of 251 traps were set over the course of 6 weeks. Only 39 of those traps collected An. squamosus. The placement of all traps is described in Figure 11; the placement of traps collecting An. squamosus is shown in Figure 12.

Collection Scheme II: An. funestus Transect Distribution of Traps May 10 - June 20 2018

43, 17%

127, 51%

81, 32%

Sleeping House Goat Pen Cattle Kraal

Figure 11. Distribution of traps under Collection Scheme II. Numbers on pie chart reflect number of that type of trap and the percentage of that type among all traps.

Collection Scheme II: An. funestus Transect Distribution of Traps that Collected An. squamosus

6, 15% 10, 26%

23, 59%

Sleeping House Goat Pen Cattle Kraal

Figure 12. Distribution of traps under Collection Scheme II that collected An. squamosus. Numbers on pie chart reflect number of that type of trap catching An. squamosus and the percentage of that trap among all traps.

Species Identification

Under Collection Scheme II, a total of 2,632 anophelines were collected during the months of May and June in 2018. Because molecular species confirmation has only been performed on a subset of the samples (15.8%, n=416), morphological identifications will be detailed here. Morphological species results are summarized in Table 9 and Table

10. Outdoors, Anopheles squamosus represented 7.9% of collections (n=194). Other species included An. rufipes (29.7%, n=731), An. coustani (9.5%, n=235), An. gambiae s.l. (7.1%, n=174), An. funestus s.l. (6.2%, n=152), An. pretoriensis (3.7%, n=92), and

An. longipalpis (2.4%, n=60). An. brunnipes, An. dancalicus, An. hancocki/brohieri, An. machardyi, An. maculipalpis, and An. theileri together made up less than 1% of collections (n=24). A remaining 32.4% (n=799) were unable to be morphologically identified.

Although 51% of traps were set indoors, only 6.4% (n=169) of anophelines in

Collection Scheme II were collected indoors. An. gambiae s.l. dominated indoors, representing 39.0% (n=66) of collections. Other species collected indoors included An. rufipes (13.6%, n=23), An. funestus (4.1%, n=7), An. coustani (3.6%, n=6), An. squamosus (3.6%, n=6), and An. longipalpis (1.8%, n=3). A remaining 33.7% (n=57) were unable to be morphologically identified. The high rate of failure of morphological identification in both indoor and outdoor collections is due in part to Collection Scheme

II capturing less familiar mosquitoes as the collections got farther away from MRT, but it is also a result of a high rate of damage. 95% of unidentified samples were recorded as damaged during collection.

Table 9. Morphologically determined outdoor species composition of Anopheles in Collection Scheme II over 124 trap-nights.

Anopheles funestus Transect Outdoor Morphological Species Composition Morphological Species Number Number per % of Total Collected Trap-Night Collection An. brunnipes 1 0.01 0.04 An. coustani 235 1.9 9.5 An. dancalicus 1 0.01 0.04 An. funestus s.l. 152 1.2 6.2 An. gambiae, s.l. 174 1.4 7.1 An. hancocki/brohieri 1 0.01 0.04 An. longipalpis 60 0.5 2.4 An. machardyi 1 0.01 0.04 An. maculipalpis 15 0.1 0.6 An. pretoriensis 92 0.7 3.7 An. rufipes 731 5.9 29.7 An. squamosus 194 1.6 7.9 An. theileri 5 0.04 0.2 Male 2 0.02 0.1 Unidentified 799 6.4 32.4 Total 2463 19.9 100.0

Table 10. Morphologically determined indoor species composition of Anopheles in Collection Scheme II over 127 trap-nights.

Anopheles funestus Transect Indoor Morphological Species Composition Morphological Species Number Number per % of Total Collected Trap-Night Collection

An. coustani 6 0.05 3.6 An. funestus 7 0.1 4.1 An. gambiae, s.l. 66 0.5 39.0 An. longipalpis 3 0.02 1.8 An. rufipes 23 0.2 13.6 An. squamosus 6 0.05 3.6 Male 1 0.01 0.6 Unidentified 57 0.4 33.7 Total 169 1.3 100.0

Several household factors were analyzed for their impact on An. squamosus presence indoors. Only 6 out of 126 households in Collection Scheme II caught An. squamosus indoors, so these results should be interpreted with caution. Excluding indoor traps that caught no Anopheles, a trap catching An. gambiae s.l. specimens was less likely to also catch An. squamosus. Additionally, households in which the last member entered after 21:00 hours were more likely to catch An. squamosus. Results for indoor trap analyses can be found in Table 11.

Various husehold factors were also analyzed for association with presence of An. squamosus outdoors. All four species tested for association, including An. rufipes, An. gambiae s.l., An. maculipalpis, and An. coustani, were found to be associated with An. squamosus in outdoor traps. Also associated with the presence of An. squamosus were the animal enclosure type, with An. squamosus being more likely to be caught by open kraals than by other enclosure types. Finally, traps at households at which people gathered outside after 19:00 hours were more likely to catch An. squamosus outdoors. Results for outdoor trap analyses can be found in Table 12. Additional analyses were performed to evaluate the association of various household factors with the abundance of An. squamosus caught outdoors.

Table 11. Results from Chi-squared tests evaluating impact of household factors in Collection Scheme II on An. squamosus presence indoors at a household with 1 degree of freedom. This analysis includes all 126 households and thus 126 trap-nights. Anopheline species refer to the presence of the specific species in the trap. **IRS refers to the receipt of IRS within the last 6 months. ***Full ITN Coverage refers to all members in the household sleeping under a bed net the night of collection. ****Entering after 21:00 hours refers to the time the last household member entered the house containing the trap.

Association of Household Factors with Presence of An. squamosus Indoors

Collection Scheme II

Factor % of trap- % of trap- % of trap- Chi- P-value

nights with nights with nights squared

factor factor without Test

catching An. factor

squamosus catching An.

squamosus

An. rufipes 34.9 20.0 10.7 0.1 0.7

An. maculipalpis 0.0

An. gambiae s.l. 39.5 0.0 23.1 2.8 0.1

An. coustani 14.0 16.7 13.5 0.0 0.8

Goat Pen 64.3 4.9 4.4 0.0 0.9

Open Eaves 54.0 2.9 6.9 0.4 0.5

Full ITN Coverage*** 67.5 4.7 4.9 0.0 1.0

IRS** 58.8 6.0 4.3 0.2 0.7

Asbestos Roof 1.6 0.0 4.8 0.1 0.8

Metal Roof 74.6 4.3 6.3 0.2 0.6

Thatched Roof 23.0 3.4 5.2 0.1 0.7

Fired Brick Walls 89.7 5.3 0.0 0.0 0.9

Entering After 21:00 35.7 8.9 2.5 1.4 0.2

Hours****

Table 12. Results from Chi-squared tests evaluating impact of household factors in Collection Scheme II on An. squamosus presence outdoors at a household with 1 degree of freedom. This analysis includes 124 households and thus 124 trap-nights. *Animals Other refers to animals contained in a structure other than a pen, an open kraal, or staked. **Gathering After 19:00 Hours refers to people gathering outside at the household on the night of collection after 19:00 hours, usually to eat.

Association of Household Factors with Presence of An. squamosus Outdoors

Collection Scheme II

Factor % of trap- % of trap- % of trap- Chi- P-value

nights with nights nights squared

factor with without Test

factor factor

catching catching

An. An.

squamosus squamosus

An. rufipes 65.1 50.0 20.7 5.6 0.0

An. maculipalpis 2.4 100.0 38.3 1.1 0.3

An. gambiae s.l. 42.2 57.1 27.1 6.4 0.0

An. coustani 36.1 56.7 30.2 4.6 0.0

Goat Pen 65.3 28.4 23.3 0.2 0.7

Animals in Open 77.4 20.8 46.4 6.0 0.0

Kraal

Gathering After 37.9 36.2 20.8 2.8 0.1

19:00 Hours**

Table 13. Results from Chi-squared tests evaluating impact of household factors in Collection Scheme II on An. squamosus abundance outdoors with 1 degree of freedom. This analysis includes 33 households and thus 33 trap-nights. Median An. squamosus collected at households collecting any An. squamosus was 3.

Association of Household Factors with Abundance of An. squamosus

Outdoors

Collection Scheme II

Factor % of trap- % of Traps % of Traps Chi- P-value

nights Above the Above the squared

with Median Median Test

factor with Factor Without Factor

An. rufipes 75.0 40.0 20.0 0.1 0.8

An. gambiae s.l. 35.0 28.6 38.5 0.2 0.7

An. coustani 20.0 0.0 47.1 1.4 0.2

Goat Pen 70.0 56.5 20.0 2.4 0.1

Gathering After 51.5 47.1 43.8 0.0 0.9

19:00 Hours

Host Choice

Host choice for the JHSPH samples from Collection Scheme II (4.4%, n=116) was analyzed using a PCR targeting the host 12S mtDNA. Households included in this analysis are highlighted in Figure 10. As all these samples were molecularly confirmed, all further discussions of species in this section will be in reference to the molecular species detailed in. Results of the blood meal PCR are summarized in Figure 13. While

An. squamosus, An. coustani, An. maculipalpis, An. rufipes, An. sp. 15/16, and An. arabiensis were all represented in those with detectable blood meals, only An. arabiensis

(n=1) had detectable human blood meal. The An. arabiensis positive for human blood meal was collected indoors. Of the An. squamosus samples with detectable blood meals,

15.0% (n=3) had fed on cow and 85.0% (n=17) on goat; all An. squamosus with detectable blood meals were collected next to a goat pen. Although An. squamosus have previously been shown to take blood from multiple hosts, our methods did not allow for identification of mixed blood meals [81].

Collection Scheme II: Anopheles funestus Transect Blood Meal Results 100% 1 1 1 1 2 3 17 9 80% 3 60% 3 1 68 40% 3 1 20% 2

Percent Percent Samples of 0%

Species

Undetectable Cow Goat Human

Figure 13. Results of 12S mtDNA PCR (Appendix G) for JHSPH samples from Collection Scheme II.

Table 14. Molecularly confirmed species composition of Anopheles in JHSPH Samples from Collection Scheme II. All 116 of these samples were morphologically identified as An. squamosus.

Collection Scheme II: Anopheles funestus Transect JHSPH Samples Molecular Species Results Molecular Species Number of Samples % of Samples

An. arabiensis 1 0.86 An. coustani 1 0.86 An. longipalpis 1 0.86 An. maculipalpis 4 3.45 An. rufipes 6 5.17 An. sp. 15/16 7 6.03 An. squamosus 88 75.86 Unidentified 8 6.90 Total 116 100

Detection of Plasmodium falciparum

JHSPH Samples from Collection Scheme II (n=116) were tested for the presence of P. falciparum using ELISA and a qPCR described above. Complete results are reported in Table 15. Samples are recorded as CSP ELISA positive if at least one replicate well was positive. Using the conservative definition of infectiousness requiring either two positive replicate wells or a confirmation with a positive qPCR replicate, however, none of these samples were infectious with P. falciparum. Although 15 samples in 3 pools tested positive by CSP ELISA, no samples tested positive individually. Two of these positive pools contained only An. squamosus; the final pool contained four An.

squamosus and a single unidentified specimen. All samples that had tested positive in the pooled CSP ELISA were tested for P. falciparum DNA using qPCR. None of the samples tested positive for P. falciparum using qPCR.

Table 15. Results of two Plasmodium assays run on JHSPH samples from May 2018 in Collection Scheme II. The CSP ELISA detects circumsporozoite protein of P. falciparum and PfLDH qPCR detects P. falciparum DNA in the head and thorax of the mosquito. CSP ELISA positives indicate the absorbance level of at least one replicate well exceeded 2x the average absorbance level of negative controls.

Collection Scheme II: Anopheles funestus Transect Plasmodium Assay Results Molecular Species Number CSP ELISA PfLDH qPCR Molecularly Positive Positive Tested An. arabiensis 1 0 0 An. coustani 1 0 0 An. longipalpis 1 0 0 An. maculipalpis 4 0 0 An. rufipes 6 0 0 An. sp. 15/16 6 0 0 An. squamosus 88 0 1 Unidentified 9 0 0 Total 116 0 1

JHSPH Samples from Collection Schemes I & II

Genetic Structure

Samples from both Collection Scheme I (n=212) and Collection Scheme II

(n=116) transported to JHSPH in Baltimore, MD were subjected to additional molecular processing for species analysis. All the JHSPH samples were run on the ITS2 PCR with additional primers targeting a portion of the COI region designed to be specific to An.

squamosus. 14% of all samples amplifying at the appropriate band size were sequenced both to confirm the specificity of the primers and to pursue further genetic analysis.

Of the 163 samples morphologically identified as An. squamosus in the JHSPH samples from Collection Scheme I, 153 (93.9%) were molecularly confirmed. 25 samples molecularly identified as An. squamosus (11.8% of 212) were morphologically identified as other species or had failed to be identified morphologically. 14% (n=25) of samples amplifying at the expected 330 bp band on the nested PCR (n=178) (Appendix B), displayed on a gel in Figure 14Figure 14, were sequenced using the COI Barcode of Life

PCR described in Appendix D. All 25 samples matched available sequences for An. squamosus with >95% identity. Sequences from this data were submitted to GenBank; accession numbers can be found in Table 17. Nine of these sequences were included in the phylogenetic tree in Figure 16.

Of the 116 samples morphologically identified as An. squamosus in the JHSPH samples from Collection Scheme II, 88 (75.9%) were molecularly confirmed. As morphological identification as An. squamosus was a criterion for transport to JHSPH, there were no JHSPH samples from Collection Scheme II not morphologically identified as An. squamosus. 23.9% (n=21) of samples amplifying at the expected 330 bp “An. squamosus” band on the ITS2 nested PCR (n=88) (Appendix B), displayed on a gel in

Figure 15, were sequenced using the COI Barcode of Life PCR described in Appendix D.

All 21 samples matched available sequences of An. squamosus with >95% identity.

Sequences from this data were submitted to GenBank; accession numbers can be found in

Table 17. Thirteen of these sequences were included in the phylogenetic tree in Figure

16.

Figure 14. Imaged 2% Agarose Gel (Appendix E) showing 330 bp fragments expected for An. squamosus using the PCR described in Appendix B. All samples on this gel are JHSPH samples from Collection Scheme I.

Figure 15. Imaged 2% Agarose Gel (Appendix E) showing 330 bp fragments expected for An. squamosus, as well as 3 unexpected >1000 base pair fragments, using the PCR described in Appendix B. All samples on this gel are JHSPH samples from Collection Scheme II.

Seven of the JHSPH samples from Collection Scheme II amplified a fragment that appeared larger than 1,000 bp for the ITS2 PCR, a band size as yet undescribed in the protocol. These same samples amplified in the absence of the “An. squamosus” primers suggesting that these samples are different from most An. squamosus. Five samples were sequenced for the ITS2 region; 4 samples, two of which were also sequenced for the ITS2 region, were sequenced at the Barcode of Life COI PCR target. Despite having band sizes

> 1,000 bp, consensus sequences were considerably smaller due to poor quality reads at each end. After alignment, the samples were blasted against the NCBI database. The

available sequence with the highest similarity to the ITS2 sequences was An. sp. 16, a species described from the Kenyan highlands, with identities ranging from 77% to 100%

[49]. The available sequence with the highest similarity to the COI sequences was An. sp.

15, another species described in the same paper, with all identities reaching 100% [49].

These results are summarized in Table 16. Several of these sequences were submitted to

GenBank and their accession numbers can be found in Table 17.

Table 16. Individual sample results of blasting NCBI database with sequences from JHSPH samples from Collection Scheme I amplifying a fragment larger than 1,000 bp using the ITS2 PCR described in Appendix B. Fragment size refers to the number of base pairs used in the search of NCBI.

Collection Scheme II: Anopheles funestus Transect Sequencing Results for ITS2 Unknowns Fragment Species % % Query Accession Size Match Identity Cover Number ITS2 PCR FLMa00017 573 bp An. sp. 16 94 90 KJ522828.1 FLMa01134 770 bp An. sp. 16 77 80 KJ522828.1 FLMa01136 755 bp An. sp. 16 100 97 KJ522828.1 FLMa01285 901 bp An. sp. 16 96 75 KJ522828.1 FLMa01408 518 bp An. sp. 16 100 98 KJ522828.1

COI PCR FLMa00017 599 bp An. sp. 15 100 98 KJ522843.1 FLMa01285 560 bp An. sp. 15 100 95 KJ522843.1 FLMa01287 558 bp An. sp. 15 100 97 KJ522843.1 FLMa01407 606 bp An. sp. 15 100 98 KJ522843.1

Construction of a Bayesian phylogenetic tree using the COI Barcode of Life region of the mitochondrial genome in a representative subset of JHSPH samples from both collection schemes demonstrated the existence of two clades among An. squamosus specimens with 100 percent support. Two samples that had tested potentially positive for

P. falciparum were split between clades, with one sample being present in each. Samples from February and May are approximately evenly distributed between the two clades.

Within Clade I, 25% of samples fed on goat and 25% fed on cow; within Clade II, 23% fed on goat and 15% fed on cattle. Although there is a difference between percentage feeding on goat versus cow in Clade II, the sample size is too small to detect whether that difference is significant. Indoor-caught An. squamosus were present in both clades, indicating clade structure has no influence on foraging location preferences.

Clade I

Clade II

Figure 16. Cytochrome oxidase subunit I (COI) Bayesian tree. Bootstrap probabilities of branches are displayed at the nodes. Highlighted are two definitive An. squamosus clades in addition to a third group identified only as An. sp. 15. Samples bolded and in red tested potentially positive for P. falciparum.

Table 17. GenBank accession numbers and month and location of collection for all specimens sequenced in this study.

Details of Anopheline Specimens Sequenced Specimen Species Accession Month of ID Number Collection IcMa0238 An. arabiensis MK776730 February FLMa00483 An. coustani MK776731 May FLMa00475 An. maculipalpis MK776733 May FLMa00485 An. maculipalpis MK776734 May FLMa00431 An. rufipes MK776735 May FLMa00456 An. rufipes MK776736 May FLMa00017 An. sp. 15 MK776737 May FLMa01287 An. sp. 15 MK776738 May FLMa01407 An. sp. 15 MK776739 May FLMa00018 An. squamosus MK776740 May FLMa00433 An. squamosus MK776741 May FLMa00455 An. squamosus MK776742 May FLMa00465 An. squamosus MK776743 May FLMa00660 An. squamosus MK776744 May FLMa00788 An. squamosus MK776745 May FLMa00922 An. squamosus MK776746 May FLMa00932 An. squamosus MK776747 May FLMa00971 An. squamosus MK776748 May FLMa01130 An. squamosus MK776749 May FLMa01277 An. squamosus MK776750 May IcMa0040 An. squamosus MK776751 February IcMa0062 An. squamosus MK776752 February IcMa0077 An. squamosus MK776753 February IcMa0097 An. squamosus MK776754 February IcMa0177 An. squamosus MK776755 February IcMa0179 An. squamosus MK776756 February IcMa0221 An. squamosus MK776757 February IcMa0232 An. squamosus MK776758 February IcMa0249 An. squamosus MK776759 February

Discussion

This study evaluated An. squamosus abundance, foraging behavior, P. falciparum infectivity rates, and genetics in an area of residual transmission in southern Zambia. An. squamosus has previously been found to demonstrate both high anthropophily and carriage of P. falciparum sporozoites in the area, implicating it as a secondary vector in the area [38, 39]. Although no specimens in this study can be considered infectious for P. falciparum under the conservative definition established here, evidence strongly suggests the presence of P. falciparum sporozoites in An. squamosus specimens. Nine An. squamosus specimens had a single positive replicate well on CSP ELISA in Collection

Scheme I; 14 An. squamosus were contained within pools with positive replicate wells on

CSP ELISA in Collection Scheme II. An. squamosus was the dominant species among these potential positives in both collection schemes. These results further emphasize the importance of research that increase our understanding of and surveillance capacity for secondary vectors.

An. squamosus was the most abundant species outdoors and second most abundant indoors in Collection Scheme I, making up 40.0% of CDC light trap collections at a collection rate of 8.6 per trap-night outdoors and 27.9% of CDC light trap collections at a collection rate of 0.9 per trap-night indoors. Previous collections using UV light traps and barrier screens outdoors reported An. squamosus comprising 40% of collections [39]; in another collection using human landing catches indoors and outdoors in addition to

CDC light traps found An. squamosus comprised 26% of collections overall [38]. The continued dominance of An. squamosus in the area throughout the rainy season, given its past reports of high anthropophily and infectiousness with P. falciparum, highlight its 58

potential to maintain malaria transmission in the absence of appropriate vector control strategies. This is especially true considering that these collections were conducted at houses as part of Step D, a reactive test and treat program for detection of and clearance of residual malaria. It is impossible to know, given the existing data, whether these transmission foci were a result of An. squamosus or that these imported transmission foci make the presence of An. squamosus of greater risk.

In Collection Scheme II, which did not begin until May 2018, An. squamosus was less dominant; it only comprised 7.9% of collections at a rate of 1.6 mosquitoes per trap- night outdoors and 3.6% of collections at a rate of 0.05 mosquitoes per trap-night indoors. This difference from Collection Scheme I could be temporal or spatial, or it could simply be due to the different sampling designs. There is data that suggests the abundance of the species coincides with the rainy season [57]. Its continued presence, however, as well as its positive pools by CSP ELISA, indicate it could continue to play a role in malaria transmission even as the rainy season comes to an end.

Although the sample size for Collection Scheme I analysis was small and thus all results should be interpreted with caution, in both collection schemes, An. squamosus were more likely to be found at households catching An. coustani outdoors. As both An. coustani and An. squamosus have been found to contain sporozoites and exhibit anthropophily in the area [38, 39], this warrants further evaluation of those households to determine if they could be potential hot spots for malaria transmission. In contrast to

Collection Scheme I, An. squamosus in Collection Scheme II were more likely to be caught outdoors when each species analyzed was present, which reflects the differences in relative abundance of An. squamosus between the collections.

Although the indoor sample size of An. squamosus for Collection Scheme II was small and thus results should be interpreted with caution, An. squamosus were less likely to be found in a household where An. gambiae s.l. were found. As 10 times as many An. gambiae s.l. were caught indoors, this could represent competition; An. squamosus could be less likely to forage indoors when they are less abundant than other more endophagic species. Another association with indoor An. squamosus, although still limited in its interpretation, is with the last household member entering the household after 21:00 hours. This could indicate an initiation of indoor foraging around 21:00 hours, but it will need to be investigated further with a collection design that captures the timing of biting.

Outdoors, in addition to the association of An. squamosus with other species in the analysis, An. squamosus were less likely to be found in an open kraal than in other animal structures and more likely to be found when people were gathered outside after

19:00 hours. The association with animal enclosures other than open kraals, which could include roofed pens, open shelters, or animals staked under trees, is surprising, especially considering the association was maintained even when traps not catching any anophelines were excluded. As open kraals represented more than 75% of the structures, however, the association will have to be evaluated further with a more equal distribution of structure type. The association of An. squamosus being caught in a trap near animal pens more often when humans are gathered outside after 19:00 hours is also curious. Although it could indicate the initiation of foraging on humans outdoors at 19:00 hours, it is important to note that no anthropophily was detected in the JHSPH samples. It is also possible that humans gathering increased generic cues like CO2 and heat that attracted

more zoophilic mosquitoes or that humans returning home meant animals also returned home from foraging in the bush.

Although no association was found between the presence of An. squamosus in a trap and the placement of the trap next to a goat pen, traps near goat pens tended to collect a higher number of An. squamosus in Collection Scheme II. This could indicate a preference of An. squamosus for goat as a host, particularly when paired with data from blood meal analysis on the JHSPH samples from both collection schemes showing a higher proportion of An. squamosus samples feeding on goat than on cow. An. squamosus abundance in Collection Scheme II was also associated, although this time negatively, with the presence of An. coustani. This is despite the presence of An. squamosus being positively associated with An. coustani in both collection schemes. The presence of both associations could reflect similar ecological preferences and thus competition, causing the species to be present in the same areas but with one species, in this case An. coustani, establishing dominance. As the sample size of households collecting An. squamosus indoors was so small in Collection Scheme II, indoor abundance could not be evaluated in this study.

Although the preference for exophagy is reflected in the literature, An. squamosus has been found to forage indoors at low rates. In previous studies in the study area, An. squamosus made up 10-20% of indoor collections depending on trapping method [38].

An. squamosus were found indoors in both collection schemes in this study, albeit in a higher abundance in Collection Scheme I. In Collection Scheme I, An. squamosus made up nearly 30% of indoor collections; in Collection Scheme II, An. squamosus made up only 3.6% of indoor collections. There were no factors found to be associated with

endophagy in either collection scheme, indicating there may be other factors not captured in the data or that the associations require a larger sample to detect.

Despite evidence of significant anthropophily [38, 40], human DNA was not detected in the blood meals of any An. squamosus analyzed in this study. This may be in part due to the limitations of the PCR used as it cannot detect mixed blood meals and only detects the single most prevalent host DNA; in Madagascar, an area where significant rates of anthropophily has at times been reported, there are studies showing a high proportion of An. squamosus samples containing blood meals from multiple hosts

[40, 81]. The rates of anthropophily can also vary significantly through time, as An. squamosus was reported to be primarily anthropophilic in the 1950s after a history of zoophily and has since been reported zoophilic once more [13, 40, 68]. In 2016, when P. falciparum sporozoites were detected in An. squamosus in the Macha area, none of the samples in that study contained detectable human blood meal [39].

As non-primates are not reservoirs of P. falciparum and there are no primates in the Macha area, it can reasonably be concluded that samples testing positive for P. falciparum had at one point fed on a human host. This would indicate that, despite methods of blood meal detection not capturing any anthropophily in either the data reported in 2016 or the samples from 2018 used in this study, the P. falciparum-infected samples are themselves evidence of anthropophily for An. squamosus certainly in 2016 and potentially in 2018 [39]. Shifts in host choice will require further investigation, but the data strongly suggests that An. squamosus is primarily zoophilic with opportunistic feeding on humans.

Although no samples were conclusively positive for P. falciparum in this study, it is comparable to other low transmission settings [82, 83]. Most notable in this study is the dominance of An. squamosus among those samples potentially positive by CSP ELISA.

Of ten specimens with detectable P. falciparum sporozoites in Collection Scheme I, 9 were molecularly confirmed as An. squamosus. Although no samples tested positive on

CSP ELISA individually among the samples tested in Collection Scheme II, there were three pools with detectable P. falciparum sporozoites. Two of those pools contained only

An. squamosus; the third pool contained four An. squamosus and a single unidentified specimen. As zoophilic species have been found to result in false positives by CSP

ELISA, particularly those feeding on cows, these results should be interpreted with some caution [84, 85]. It is hypothesized that the antibody used in the CSP ELISA cross-react with proteins in bovine blood that may have contaminated the thorax of mosquitoes blood-fed on cattle. The data from this study suggest, however, that while that risk of false positives is present, it does not appear to account for all the CSP ELISA positives.

In Collection Scheme I, only 4 out of the 9 An. squamosus that tested positive for P. falciparum by CSP ELISA had detectable blood meal, and only 3 had detectable bovine blood meal. In Collection Scheme II, only 2 of the 15 samples in the pools positive for P. falciparum by CSP ELISA had detectable blood meal, and both contained goat blood meal.

Unfortunately, due to the presence of the single unidentified sample in the third pool positive by CSP ELISA from Collection Scheme II, we cannot attribute the positivity of that pool to An. squamosus with confidence. Although the specimen was morphologically identified as An. squamosus, it was not confirmed molecularly. This

highlights the importance of molecular confirmation of species. Despite all samples from both collection schemes being morphologically identified by the same entomological team, the positive predictive value (PPV) of the morphological identification of An. squamosus varied between collection schemes. Because PPV could only be calculated for those specimens that were molecularly confirmed, these calculations have been limited to the JHSPH samples. In the JHSPH samples from Collection Scheme I, in which An. squamosus made up 84.0% of the samples (n=178), the positive predictive value of an

An. squamosus morphological identification was 93.9%. In contrast, in the JHSPH samples from Collection Scheme II, in which An. squamosus made up only 22.4% of the samples (n=87), the positive predictive value was 75.9%. It appears as though species diversity impacts the accuracy of the morphological identification. Accurate species identification is critical for assigning vector status to the right species and thus deploying appropriate control strategies. As P. falciparum-positive pools of An. squamosus samples were found even when An. squamosus was less abundant, the importance of molecular identification of Anopheles species has been reinforced by this study.

In addition to molecular confirmation of species, this study also performed phylogenetic analysis on a selection of samples from both collection schemes. The resulting topology reinforces the previously described hypothesis that An. squamosus is actually a species complex [38, 39]. Although largely neglected in the literature, it is known that An. squamosus is morphologically indistinguishable from An. cydippis in the adult stage [52, 73]. Larvae of these two species, however, can be easily distinguished. In what little literature An. cydippis is mentioned, it is often referred to as a “variety” of An. squamosus [86]. Mass spectrometry has revealed detectable differences between them,

but there have been no studies thus far investigating the genetic differences between the two species [87].

The topology of the tree in Figure 16 reveals structural features that provide evidence for the existence of an An. squamosus species complex. The first evidence is the defined clade structure among specimens recognized as An. squamosus. This has been demonstrated in previous literature from Zambia [38, 39]. The second is the close genetic relationship of An. sp. 15 to specimens recognized as An. squamosus. All An. sp. 15 specimens in this study, as well as in a second study in Nchelenge in northern Zambia, were morphologically identified as An. squamosus [80]. However, An. sp. 15 and An. squamosus partition molecularly. Phylogenetic comparison using COI sequences (Figure

16) illustrates that An. sp. 15 clusters with An. squamosus but clearly falls as a strongly supported independent clade. In addition, although An. squamosus fails to amplify using standard ITS2 primers, specimens identified as An. sp. 15 produce a consistently large band outside the scope described for this molecular tool. Although the sample size in our combined studies is small (n=10), these collective data suggest that An. sp. 15 may be a cryptic sibling species of An. squamosus. As all our samples were collected as adults, and as there are no known molecular tools nor sequence data for differentiating An. squamosus and An. cydippis, it is impossible to know with certainty whether An. cydippis is represented in our data.

Whether the samples collected in this study are An. squamosus or An. cydippis, the data demonstrate that members of an An. squamosus complex may not only be playing a role in malaria transmission in southern Zambia but may indeed be playing the dominant role in malaria transmission in pre-elimination southern Zambia. Although the

clear preference for zoophily is demonstrated in this study, the dominance of An. squamosus in samples potentially positive for P. falciparum relative to the other species in this study is concerning. According to Tantely regarding the transmission of Rift

Valley fever virus, “Transmission of the virus from vectors to humans is possible even if the human is the least attractive for all vector species” [68]. This is especially true considering both the relative abundance of An. squamosus found in Collection Scheme I and the strong exophagy exhibited by An. squamosus. The strong zoophily, exophagy, and exophily of An. squamosus suggest that the traditional methods of vector control such as IRS and ITN use will likely not be effective against this secondary vector. The WHO does not currently recommend any outdoor control strategies due to the lack of convincing evidence of their efficacy [88]. If the global community fails to develop effective outdoor vector control strategies, and if Zambia and other regions in sub-

Saharan Africa do not adapt their vector control strategies to include such outdoor control methods, malaria elimination will not be achieved.

Future Directions

As this study had several limitations, there is plenty of room for future research on An. squamosus. First and foremost, although Collection Scheme I had the benefit of representing vector populations and behavior during peak malaria season, the study design significantly limited the sample size that could be used for analysis. The study also failed to capture resting behavior; collection of outdoor resting mosquitoes would help elucidate where An. squamosus are resting and thus how to target them with control efforts. Indoor resting mosquitoes are targeted with IRS; if An. squamosus is resting 66

outdoors as suspected, identifying where could help decide if spraying those locations would be safe and effective. Outdoor collections of foraging adults using human bait, using human landing catches (HLCs), light traps set near outdoor gathering places, or artificial bait, would reveal if An. squamosus exhibit anthropophily outdoors. Use of an assay that can capture mixed blood meals may increase detection of anthropophily in An. squamosus. Improved sensitivity and specificity of assays to detect P. falciparum, particularly infectious P. falciparum, are critical for establishing the role of vector species with confidence. Larval sampling could help to address the An. squamosus and An. cydippis difference, as they are distinguishable at the larval stage, in addition to revealing any breeding site preferences. Finally, extending the data to include longer periods of random sampling would allow for additional ecological and temporal analysis.

Conclusions

The data in this study confirm the role of An. squamosus in residual malaria transmission in southern Zambia. An. squamosus was abundant during the typical malaria transmission season, increasing the significance of its role as a secondary vector. An. squamosus appear to be preferentially exophagic, although specimens were caught indoors in both collections. While none of the An. squamosus in this study had detectable human blood meals, An. squamosus in both February and May 2018 were potentially positive for P. falciparum. Phylogenetic structure reveals two strongly supported clades among specimens identified as An. squamosus, in addition to a third closely related clade currently identified as An. sp. 15. Together, these data contribute to the growing body of evidence of a cryptic species complex involved in residual malaria transmission. 67

References

1. Vector-borne diseases [http://www.who.int/news-room/fact-sheets/detail/vector- borne-diseases] 2. About Malaria: Biology [https://www.cdc.gov/malaria/about/biology/index.html] 3. Norris LC, Norris DE: Phylogeny of anopheline (Diptera: Culicidae) species in southern Africa, based on nuclear and mitochondrial genes. Journal of Vector Ecology 2015, 40:16-27. 4. Gilles HM, Warrell DA: Bruce-Chwatt's Essential Malariology. Third edn. New York: Oxford University Press; 1993. 5. About Malaria: Disease [https://www.cdc.gov/malaria/about/disease.html] 6. Veith I: Huang ti nei ching su wen: The Yellow Emperor's classic of internal medicine. Berkeley, CA: University of California Press; 1966. 7. Faurant C: From bark to weed: The history of artemisinin. Parasite 2011, 18:215-218. 8. Rodriguez FM: Precisiones sobre la historia de la quina. Reumatologica Clinica 2007, 3:151-196. 9. Harbach RE: The classification of genus Anopheles (Diptera: Culicidae): a working hypothesis of phylogenetic relationships. Bulletin of Entomological Research 2004, 94:537-553. 10. Krzywinski J, Besansky NJ: Molecular systematics of Anopheles: from subgenera to subpopulations. Annu Rev Entomol 2003, 48:111-139. 11. Gillies MT, DeMeillon B: The Anophelinae South of the Sahara (Ethiopian Zoogeographical Region). Johannesburg: South African Institute for Medical Research; 1968. 12. Bates M: The natural history of mosquitoes. New York: The Macmillan Company; 1949. 13. Tantely ML, Le Goff G, Boyer S, Fontenille D: An updated checklist of mosquito species (Diptera: Culicidae) from Madagascar. Parasite 2016, 23. 14. Leeson H: Preliminary Notes on the Anophelini of Southern Rhodesia. South African Journal of Science 1927, 24. 15. Ross R: On Some Peculiar Pigmented Cells Found in Two Mosquitoes Fed on Malarial Blood. The British Medical Journal 1897, 2:1786-1788. 16. Ross R: Infection of Birds with Proteosoma by the Bites of Infected Mosquitoes. The Indian Medical Gazette 1899, 34:1-3.

17. Laveran A: Un nouveau parasite trouve dans le sang de malades atteints de fievre palustre. Origine parasitaire des accidents de l'impaludisme. Bulletins et mémoires de la Société Médicale des Hôpitaux de Paris 1881, 17:158-164. 18. Cushing EC: History of Entomology in World War II. Washington D.C.: Smithsonian Institution; 1957. 19. Feachem RGA, Phillips AA, Targett GA: Shrinking the malaria map: A prospectus on malaria elimination. San Francisco: The Global Health Group, Global Health Sciences, University of California, San Francisco; 2009. 20. Snow RW, Trape JF, Marsh K: The past, present, and future of childhood malaria mortality in Africa. Trends in Parasitology 2001, 17:593-597. 21. Roll Back Malaria Partnership: A decade of partnership and results. Geneva2011. 22. Dhiman S: Are malaria elimination efforts on right track? An analysis of gains achieved and challenges ahead. Infectious Diseases of Poverty 2019, 8. 23. Sharp B, van Wyk P, Sikasote JB, Banda P, Kleinschmidt I: Malaria control by residual insecticide spraying in Chingola and Chililabombwe, Copperbelt Province, Zambia. Trop Med Int Health 2002, 7:732-736. 24. Mharakurwa S, Thuma PE, Norris DE, Mulenga M, Chalwe V, Chipeta J, Munyati S, Mutambu S, Mason PR: Malaria epidemiology and control in Southern Africa. Acta Tropica 2012, 121:202-206. 25. Moss WJ, Hamapumbu H, Kobayashi T, Shields T, Kamanga A, Clennon J, Mharakurwa S, Thuma PE, Glass G: Use of remote sensing to identify spatial risk factors for malaria in a region of declining transmission: A cross- sectional and longitudinal community survey. Malaria Journal 2011, 10. 26. National Malaria Elimination Centre: National Malaria Elimination Strategic Plan 2017-2021: Moving from accelerated burden reduction to malaria elimination in Zambia. (Health Mo ed. Lusaka, Zambia: Government of Zambia; 2017. 27. Searle KM, Hamapumbu H, Lubinda J, Shields TM, Pinchoff J, Kobayashi T, Stevenson JC, Bridges DJ, Larsen DA, Thuma PE, Moss WJ: Evaluation of the operational challenges in implementing reactive screen-and-treat and implications of reactive case detection strategies for malaria elimination in a region of low transmission in southern Zambia. Malaria Journal 2016, 15. 28. Baumle R, Nkhoma J, Silembo O: Hydrogeological Map of Zambia, Southern Province: Map Sheet Kafue Flats and Southern Tributaries 1:250,000. . (Affairs DoW ed. Hannover, Germany: Ministry of Lands, Energy and Water Development, Zambia, Federal Institute for Geosciences and Natural Resources; 2007.

29. Sturrock HJW, Roberts KW, Wegbreit J, Ohrt C, Gosling RD: Tackling Imported Malaria: An Elimination Endgame. American Journal of Tropical Medicine and Hygiene 2015, 93:139-144. 30. Killeen GF: Characterizing, controlling and eliminating residual malaria transmission. Malaria Journal 2014, 13:330. 31. Hiwat H, Martinez-Lopez B, Cairo H, Hardjopawiro L, Boerleider A, Duarte E, Yadon Z: Malaria epidemiology in Suriname from 2000 to 2016: trends, opportunities and challenges for elimination. Malaria Journal 2018, 17. 32. Wangdi K, Gatton M, Kelly G, Clements A: Cross-border malaria: a major obstacle for malaria elimination. Advances in Parasitology 2015, 89:79-107. 33. Pizzitutti F, Mena C, Feingold B, Pan W: Modeling asymptomatic infections and work-related human circulation as drivers of unstable malaria transmission in low-prevalence areas: a study in the Northern Peruvian Amazon. Acta Tropica 2019. 34. Plowe CV: The evolution of drug-resistant malaria. Transactions of The Royal Society of Tropical Medicine and Hygiene 2009, 103:S11-S14. 35. Moiroux N, Gomez MB, Pennetier C, Elanga E, Djenontin A, Chandre F, Djegbe I, Guis H, Corbel V: Changes in Anopheles funestus biting behavior following universal coverage of long-lasting insecticidal nets in Benin. Journal of Infectious Diseases 2012, 206:1622-1629. 36. Moiroux N, Damien GB, Egrot M, Djenontin A, Chandre F, Corbel V, Killeen GF, Pennetier C: Human exposure to early morning Anopheles funestus biting behavior and personal protection provided by long-lasting insecticidal nets. PLoS One 2014, 9:e104967. 37. Russell TL, Govella NJ, Azizi S, Drakeley CJ, Kachur SP, Killeen GF: Increased proportions of outdoor feeding among residual malaria vector populations following increased use of insecticide-treated nets in rural Tanzania. Malaria Journal 2011, 10:80. 38. Fornadel CM, Norris LC, Franco V, Norris DE: Unexpected Anthropophily in the Potential Secondary Malaria Vectors Anopheles coustani s.l. and Anopheles squamosus in Macha, Zambia. Vector-Borne and Zoonotic Diseases 2011, 11:1173-1179. 39. Stevenson JC, Simubali L, Mbambara S, Musonda M, Mweetwa S, Mudenda T, Pringle JC, Jones CM, Norris DE: Detection of Plasmodium falciparum Infection in Anopheles squamosus (Diptera: Culicidae) in an Area Targeted for Malaria Elimination, Southern Zambia. Journal of Medical Entomology 2016, 53:1482-1487. 40. Lacan A: Anophelines of the Plateau Region of Madagascar in 1952. Memoires de l'Institut Scientifique de Madagascar 1953, 4:503-519.

41. Norris LC, Norris DE: Heterogeneity and changes in inequality of malaria risk after introduction of insecticide-treated bed nets in Macha, Zambia. The American Journal of Tropical Medicine & Hygiene 2013, 88:710-717. 42. Kent RJ, Thuma PE, Mharakurwa S, Norris DE: Seasonality, blood feeding behavior, and transmission of Plasmodium falciparum by Anopheles arabiensis after an extended drought in southern Zambia. American Journal of Tropical Medicine and Hygiene 2007, 76:267-274. 43. Norris D: Human Blood Index of Indoor-resting Anopheles arabiensis. (Hoffman J ed.2019. 44. Tabue RN, Awono-Ambene P, Etang J, Atangana J, Antonio-Nkondjio C, Toto JC, Patchoke S, Leke RGF, Fondjo E, Mnzava AP, et al: Role of Anopheles (Cellia) rufipes (Gough, 1910) and other local anophelines in human malaria transmission in the northern savannah of Cameroon: a cross-sectional survey. Parasites & Vectors 2017, 10. 45. Gillies MT: The Role of Secondary Vectors of Malaria in North-East Tanganyika. Trans R Soc Trop Med Hyg 1964, 58:154-158. 46. Nepomichene TN, Tata E, Boyer S: Malaria case in Madagascar, probable implication of a new vector, Anopheles coustani. Malaria Journal 2015, 14:475. 47. De Meillon B: The Anophelini of the Ethiopian geographical region. Publications of the South African Institute for Medical Research Johannesburg 1947:238-241. 48. Lobo NF, Laurent BS, Sikaala CH, Hamainza B, Chanda J, Chinula D, Krishnankutty SM, Mueller JD, Deason NA, Hoang QT, et al: Unexpected diversity of Anopheles species in Eastern Zambia: implications for evaluating vector behavior and interventions using molecular tools. Sci Rep 2015, 5:17952. 49. St Laurent B, Cooke M, Krishnankutty SM, Asih P, Mueller JD, Kahindi S, Ayoma E, Oriango RM, Thumloup J, Drakeley C, et al: Molecular Characterization Reveals Diverse and Unknown Malaria Vectors in the Western Kenyan Highlands. The American Journal of Tropical Medicine & Hygiene 2016, 94:327-335. 50. Mansfield-Aders W: Notes on Malaria and Filariasis in the Zanzibar Protectorate. Transactions of the Royal Society of Tropical Medicine and Hygiene 1927, 21. 51. Fawzy M, Helmy YA: The One Health Approach is Necessary for the Control of Rift Valley Fever Infections in Egypt: A Comprehensive Review. Viruses 2019, 11. 52. Theobald FV: A Monograph of the Culicidae or Mosquitoes. London: British Museum of Natural History; 1901.

53. Wellman FC: Notes on the Common Mosquitoes of Bihe and Bailundo Districts, Portuguese West Africa. Journal of Infectious Diseases 1905, 2:627- 631. 54. Theobald FV: A Monograph of the Culicidae or Mosquitoes. London: British Museum of Natural History; 1903. 55. Hicks E: The transmission of Wuchereria bancrofti in Sierra Leone. Annals of Tropical Medicine & Parasitology 1932, 26:407-422. 56. Monier H: Faits nouveaux concernant l'épidémiologie du paludisme à Tananarive. Bulletin de la Société de Pathologie Exotique 1935, 28. 57. Lewis DJ: The egg of Anopheles squamosus Theobald. East African Medical Journal 1946, 23. 58. Colbourne M, Wright F: Malaria in the Gold Coast. West African Medical Journal 1955, 4. 59. Barbier D, Moreau JP, Radanielina R: Epidemiologic study among companies of the armed-forces of Southeastern Madagascar Malagasy-Republic: Part III. Archives de l'Institut Pasteur de Madagascar 1975, 44:127-130. 60. Nigatu W, Petros B, Lulu M, Adugna N, Wirtz R: Species composition, feeding and resting behaviour of the common anthropophilic anopheline mosquitoes in relation to malaria transmission in Gambella, south west Ethiopia. International Journal of Tropical Insect Science 1994, 15:371-377. 61. Diagne N, Fontenille D, Konate L, Faye O, Lamizana MT, Legros F, Molez JF, Trape JF: The Anopheles of Senegal - an Annotated and Illustrated Check List. Bulletin de la Societe de Pathologie Exotique 1994, 87:267-277. 62. Degefa T, Zeynudin A, Godesso A, Michael YH, Eba K, Zemene E, Emana D, Birlie B, Tushune K, Yewhalaw D: Malaria incidence and assessment of entomological indices among resettled communities in Ethiopia: a longitudinal study. Malaria Journal 2015, 14:24. 63. Kyalo D, Amratia P, Mundia CW, Mbogo CM, Coetzee M, Snow RW: A geo- coded inventory of anophelines in the Afrotropical Region south of the Sahara: 1898-2016. Wellcome Open Research 2017, 2:57. 64. Kouznetsov R, Organization WH: Distribution of anophelines in the Yemen Arab Republic and its relation to malaria. Geneva: World Health Organization; 1976. 65. Jadin J: Malaria Control by Residual DDT Spraying in the Astrida Region, Ruanda-Urundl. In Annales de la Societe Belge de Medecine TropicaleSociete Belge de Medecine Tropicale; 1951: 631-651. 66. Lumsden W: Probable insect vectors of yellow fever virus, from monkey to man, in Bwamba County, Uganda. Bulletin of Entomological Research 1951, 42:317-330.

67. Nepomichene TNJJ, Elissa N, Cardinale E, Boyer S: Species Diversity, Abundance, and Host Preferences of Mosquitoes (Diptera: Culicidae) in Two Different Ecotypes of Madagascar With Recent RVFV Transmission. Journal of Medical Entomology 2015, 52:962-969. 68. Tantely ML, Rakotoniaina JC, Tata E, Andrianaivolambo L, Razafindrasata F, Fontenille D, Elissa N: Biology of Mosquitoes That Are Potential Vectors of Rift Valley Fever Virus in Different Biotopes of the Central Highlands of Madagascar. Journal of Medical Entomology 2013, 50:603-610. 69. Gillies M: Observations on nulliparous and parous rates in some common East African mosquitoes. Annals of Tropical Medicine & Parasitology 1963, 57:435-442. 70. Beier JC: Vector incrimination and entomological inoculation rates. Methods in Molecular Medicine 2002, 72:3-11. 71. Fornadel CM, Norris LC, Norris DE: Centers for Disease Control light traps for monitoring Anopheles arabiensis human biting rates in an area with low vector density and high insecticide-treated bed net use. American Journal of Tropical Medicine and Hygiene 2010, 83:838-842. 72. Sudia WD, Chamberlain RW: Battery-operated light trap, an improved model. Journal of the American Mosquito Control Association 1988, 4:536-538. 73. Gillies T, Coetzee M: A supplement to the anophelinae of Africa South of the Sahara: Afrotropical Region. Johannesburg, South Africa: South African Institute for Medical Research; 1987. 74. Barillas-Mury C: Personal communication. 75. Parr J, Belson C, Patel J, Hoffman I, Kamthunzi P, Martinson F, Tegha G, Thengolose I, Drakeley C, Meshnick S, et al: Estimation of Plasmodium falciparum Transmission Intensity in Lilongwe, Malawi, by Microscopy, Rapid Diagnostic Testing, and Nucleic Acid Detection. American Journal of Tropical Medicine and Hygiene 2016, 95:373-377. 76. Norris DE, Shurtleff AC, Toure YT, Lanzaro GC: Microsatellite DNA polymorphism and heterozygosity among field and laboratory populations of Anopheles gambiae s.s. (Diptera: Culicidae). Journal of Medical Entomology 2001, 38:336-340. 77. Koekemoer LL, Kamau L, Hunt RH, Coetzee M: A cocktail polymerase chain reaction assay to identify members of the Anopheles funestus (Diptera: Culicidae) group. American Journal of Tropical Medicine and Hygiene 2002, 6:804-811. 78. Scott JA, Brogdon WG, Collins FH: Identification of single specimens of the Anopheles gambiae complex by the polymerase chain reaction. American Journal of Tropical Medicine and Hygiene 1993, 49:520-529.

79. Lobo NF, Laurent BS, Sikaala CH, Hamainza B, Chanda J, Chinula D, Krishnankutty SM, Mueller JD, Deason NA, Hoang QT, et al: Unexpected diversity of Anopheles species in Eastern Zambia: implications for evaluating vector behavior and interventions using molecular tools. Scientific Reports 2015. 80. Ciubotariu I: Personal communication. 2018. 81. Tedrow RE, Ratovonjato J, Walker ED, Ratsimbasoa AC, Zimmerman PA: A Novel Assay for Simultaneous Assessment of Mammalian Host Blood, Mosquito Species, and Plasmodium spp. in the Medically Important Anopheles Mosquitoes of Madagascar. The American Journal of Tropical Medicine and Hygiene 2019, 100:544-551. 82. Degefa T, Yewhalaw D, Zhou G, Lee M-c, Atieli H, Githeko AK, Yan G: Indoor and outdoor malaria vector surveillance in western Kenya: implications for better understanding of residual transmission. Malaria Journal 2017, 16:443. 83. Kilama M, Smith DL, Hutchinson R, Kigozi R, Yeka A, Lavoy G, Kamya MR, Staedke SG, Donnelly MJ, Drakeley C, et al: Estimating the annual entomological inoculation rate for Plasmodium falciparum transmitted by Anopheles gambiae s.l. using three sampling methods in three sites in Uganda. Malaria Journal 2014, 13. 84. Bashar K, Tuno N, Ahmed TU, Howlader AJ: False positivity of circumsporozoite protein (CSP)-ELISA in zoophilic anophelines in Bangladesh. Acta Tropica 2013, 125:220-225. 85. Durnez L, Van Bortel W, Denis L, Roelants P, Veracx A, Trung H: False positive circumsporozoite protein ELISA: A challenge for the estimation of the entomological inoculation rate of malaria and for vector incrimination. Malaria Journal 2011, 10. 86. Lips M: Congo (Ex-Belgian) Anophelines. 7. Some Species of the Cellia and Neocellia Groups. Rivista di Parassitologia Roma 1962, 23:107-134. 87. Raharimalala F, Andrianinarivomanana T, Rakotondrasoa A, Collard J, Boyer S: Usefulness and accuracy of MALDI‐TOF mass spectrometry as a supplementary tool to identify mosquito vector species and to invest in development of international database. Medical and Veterinary Entomology 2017, 31:289-298. 88. World Health Organization: Guidelines for malaria vector control. Geneva2019. 89. Parr JB, Belson C, Patel JC, Hoffman IF, Kamthunzi P, Martinson F, Tegha G, Thengolose I, Drakeley C, Meshnick SR, et al: Estimation of Plasmodium falciparum Transmission Intensity in Lilongwe, Malawi, by Microscopy, Rapid Diagnostic Testing, and Nucleic Acid Detection. The American Journal of Tropical Medicine and Hygiene 2016, 95:373-377.

Appendices of Protocols

The following pages contain laboratory protocols used in the completion of this thesis.

Appendix A Marriott DNA Extraction Procedure Materials: Bender Buffer 0.1 M NaCl (5 mL from a 1M stock solution—need to make this stock solution) 0.2 M sucrose (3.42 grams) 0.1 M Tris-HCl (5 mL from a 1M stock) 0.05 M EDTA pH 9.1 (5 mL from a 0.5M stock) 0.5% SDS in DEPC water (0.25 mL from a 0.1M stock) For 50 mL Bender Buffer: Add 3.42 grams dry sucrose to a 50 mL conical tube. Add the proper amounts of the other ingredients, listed above in parentheses. Fill to a final volume with DEPC water. Be sure to add SDS last, after mixing, otherwise the detergent will foam. Filter-sterilize with a 0.2 micron filter before using. Store at room temperature. To make 1 M NaCl stock solution, add 2.9 grams dry NaCl into 50 mL HPLC H2O and vortex. Stock solutions of the liquid reagents should come in the molar concentrations listed. 8M Postassium acetate To make an 8 molar stock solution, add 19.63 grams into 25 mL HPLC H2O. Store at 4°C. Extraction Protocol: 1. If specimens are dry, rehydrate them in a 1.5 mL microfuge tube containing 20 µl HPLC H2O for 10 minutes. If specimens are frozen, begin the procedure from Step 2. 2. Add 100 µl Bender Buffer directly into the tube with the specimen and homogenize until there are no recognizable mosquito parts. Place used pestle in 1M NaOH. 3. Incubate homogenized samples at 65°C for 1 hour. 4. Add 15 µl cold 8M potassium acetate to each sample. Mix gently and incubate on ice for 45 minutes. (Procedure may be stopped here overnight.) 5. Spin samples in a microcentrifuge (14,000 rpm) for 10 minutes, and then transfer the supernatant to a new 1.5 mL microfuge tube. 6. Add 300 µl 100% ethanol (2X volume) to each supernatant to precipitate DNA. Mix well by inverting the tube. Incubate samples at room temperature for 5 minutes. 7. Centrifuge samples (14,000 rpm) for 15 minutes. Following this spin there should be a small pellet of DNA at the bottom of the tube. 8. Carefully remove the supernatant and discard it, leaving the pellet behind in the tube. Let the pellets dry completely before resuspending—residual ethanol can interfere with PCR later. 9. Resuspend pellets in 50 µl HPLC H2O for head/thorax or abdomen extractions (100 µl for whole mosquitoes). Ideally, store overnight at 4°C before use. Store DNA permanently at -20°C. Pestle washing: To prevent DNA contamination in PCR-based analyses, pestles should be soaked in 1M NaOH after use. They should then be washed in soapy water, rinsed off in distilled water, and autoclaved before they are used again.

Appendix B ITS2 rDNA PCR Modified by Christine Jones, Norris Laboratory JHSPH 04/2018

This PCR is very robust and therefore can be used to check the quality of DNA extractions. It targets the ITS2 region of nuclear rDNA and produces amplicons of varying sizes depending on mosquito species. It can be used in tandem with the Funestus PCR to identify ambiguous samples. Because ITSA binds to the conserved 5.8S rDNA and ITS2B binds to the 28S rDNA, this PCR can be used to sequence samples from almost any anopheline mosquito for species identification. ITS2B1, a novel, alternate primer, binds slightly downstream from ITS2B and produces a slightly larger amplicon that can be used to sequence through the entire ITS2.

Expected product sizes for different mosquito species: Funestus group: An. leesoni ~520 bp An. rivulorum and rivulorum-like ~520 bp An. parensis ~ 620 bp An. longipalpis ~620 bp and ~900 bp An. vaneedeni ~ 830 bp An. funestus and funestus-like ~850 bp

Other species: An. rufipes, maculipalpis, and pretoriensis ~500 bp An. theileri ~ 520 bp An. gambiae complex ~600 bp An. coustani ~620 bp An. squamosus with SQFor/Rev ~300 bp

Primers: ITS2A: 5’- TGT GAA CTG CAG GAC ACA T -3’ ITS2B: 5’- TAT GCT TAA ATT CAG GGG GT -3’ ITS2B1: 5’- GTC CCT ACG TGC TGA GCT TC -3’ SQFor405: 5’- CCA TTT CCA TTA TGT CCT ATC TAT AGG -3’ SQRev707: 5’- GGG AAA GCA GGA GTT CGT TGA G- 3’

Note: Only the ITS2B and ITS2B1 primers work well for sequencing.

PCR Program: (ITS2) 1. 94ºC 2 min 2. 94ºC 30 sec 3. 50ºC 30 sec 4. 72ºC 40 sec 5. Go to step 2 39x 6. 72ºC 10 min 7. 4ºC forever

Reaction Mixture: 25 L 10X 2.5 µL dNTPs 2.5 mM 2.0 µL (final conc. 200 M each)

ITS2A 0.3 µL (30 pmol) ITS2B 0.3 µL (30 pmol) Taq 2.0 U dH20 fill to 25 μL

Use 1 µL of template DNA.

Reference: Koekemoer, L.L., L. Kamau, R.H. Hunt, M. Coetzee. 2002. A cocktail polymerase chain reaction assay to identify members of the Anopheles funestus (Diptera: Culicidae) group. Am. J. Trop. Med. Hyg. 6(6): 804-811.

Appendix C Differentiation of the Anopheles gambiae complex by PCR

This PCR uses 4 primers that in combination produce three differentially-sized amplicons of the ribosomal DNA spacer region of An. gambiae complex mosquitoes. The expected product sizes are as follows: An. gambiae s.s. (~390 bp), An. arabiensis (~315 bp), and An. quadriannulaus (~150 bp).

Primers: UN: 5’- GTG TGC CCC TTC CTC GAT GT -3’ GA: 5’- CTG GTT TGG TCG GCA CGT TT -3’ AR: 5’- AAG TGT CCT TCT CCA TCC TA -3’ QD: 5’- CAG ACC AAG ATG GTT AGT AT -3’

PCR Program: (SCOTT) 1. 94ºC 2 min 2. 94ºC 30 sec 3. 50ºC 30 sec 4. 72ºC 30 sec 5. Go to step 2 29x 6. 72ºC 7 min 7. 4ºC forever

Reaction Mixture: 25 L 20 L 12 L 10X 2.5 L 2.0 µL 1.25 L dNTPs 2.5 mM 2.0 L 1.6 µL 1.0 L (final conc. 200 M each) AR 3.0 L 2.4 µL 1.5 L (150 pmol) QD 3.0 L 2.4 µL 1.5 L (150 pmol) GA 0.5 L 0.4 µL 0.25 L (25 pmol) UN 1.0 L 0.8 µL 0.5 L (50 pmol) Taq 1.5 U 1.2 U 0.9 U dH20 fill to total reaction mix volume

Use between 0.5 and 1 L of template DNA.

Reference: Scott, J.A., W.G. Brogdon and F.H. Collins. 1993. Identification of single specimens of the Anopheles gambiae complex by the polymerase chain reaction. Am. J. Trop. Med. Hyg. 49(4): 520-529.

Appendix D Cytochrome Oxidase subunit I (COI) mitochondrial PCR

This PCR targets the 3’ portion of the cytochrome oxidase I gene (bp 2121-2998) and can be used for sequencing most anopheline mosquito species for phylogeny building. Because it is a mitochondrial gene and has high copy number, it is fairly robust. The amplicon target size is 877 bp. Adapted from Lobo et al. 2015.

Primers: LCO 1490: 5’-GGT CAA CAA ATC ATA AAG ATA TTG G-3’ HCO 2198: 5’-TAA ACT TCA GGG TGA CCA AAA AAT CA-3’

PCR Program: (COILobo) 1. 94ºC 5 min 2. 94ºC 40 sec 3. 45ºC 1 min 5 cycles 4. 72ºC 1.5 min 5. 94ºC 40 sec 6. 51ºC 1 min 30 cycles 7. 72ºC 1.5 min 8. 72ºC 5 min 9. 4ºC forever

Reaction Mixture: 25 µL 10X 2.5 µL dNTPs 2.5 mM 2.0 µL (final conc. 200 M each) LCO 1490 0.3 µL (30 pmol) HCO 2198 0.3 µL (30 pmol) Taq 2.0 U dH20 fill to 25 μl

Use 1.0 μl DNA template.

Reference: Lobo, N. F., B. S. Laurent, C. H. Sikaala, B. Hamainza, J. Chanda, D. Chinula, S. M. Krishnankutty, J. D. Mueller, N. A. Deason, Q. T. Hoang, H. L. Boldt, J. Thumloup, J. Stevenson, A. Seyoum, and F. H. Collins. 2015. Unexpected diversity of Anopheles species in Eastern Zambia: implications for evaluating vector behavior and interventions using molecular tools. Scientific Reports 5:17952.

Appendix E Preparation of 2% Agarose Gel Johns Hopkins Bloomberg School of Public Health Laboratory of Dr. Douglas Norris

I. Initial Preparation a. Add 4.0 g of LE Quick Dissolve Agarose to a 500 ml Erlenmeyer flask. b. Add 200 ml of 10% TBE to the 500 ml Erlenmeyer flask. c. Microwave for 2:00 minutes. You do not need to cover the flask with plastic wrap. d. Leave to cool on the counter for 13:00 minutes. During that time, prepare the mold and comb for the gel. e. Some TBE may evaporate; if necessary, top up with 10% TBE to 200 ml. f. Add 6 μl of ethidium bromide and swirl to mix. g. Pour the solution into the mold prepared in step d) and leave for at least 20 minutes to set. The gel can be left up to an hour at this point or wrapped in plastic wrap and placed in the fridge overnight if necessary.

II. Storage a. After running and visualizing the gel, return it to the Erlenmeyer flask in which it was prepared. b. Seal the top of the flask in parafilm. c. Store on the bench at room temperature and labeled with the gel concentration, your initials, and the date of preparation.

III. Re-Use a. Remove the parafilm cover from the top of the flask. b. Microwave for 2:00 minutes. c. Leave to cool on the counter for 13:00 minutes. During that time, prepare the mold and comb for the gel. d. Some TBE may evaporate; if necessary, top up with 10% TBE to 200 ml. e. Add 3 μl of ethidium bromide and swirl to mix. f. Pour the solution into the mold prepared in step c) and leave for at least 20 minutes to set. The gel can be left up to an hour at this point or wrapped in plastic wrap and placed in the fridge overnight if necessary. g. The gel can be re-used in this way until the band quality noticeably diminishes.

Appendix F Mammalian host blood meal species identification by PCR and sequencing*

This PCR diagnostic identifies whether mosquitoes have taken blood meals. Currently, one product of size ~205bp for any blood meal is amplified from the 12S ribosomal RNA gene. This protocol has been tested on human, cow, dog, goat, and pig DNA extracted from blood or tissues. The PCR products can then be run on a 2% agarose gel.

Primers: UNIFic: 5’- GGA TTA GAT ACC CCA CTA TGC -3’ UNIRic: 5’- GCT GAA GAT GGC GGT ATA TAG -3’

PCR Program: (12Sic) 1. 94ºC 5 min 2. 94ºC 30 sec 3. 48ºC 30 sec 4. 72ºC 30 sec 5. Go to step 2 35x 6. 72ºC 7 min 7. 4ºC forever

Reaction Mixture: 25 L 10X Buffer 2.5 µL dNTPs 2.5 mM 2.0 µL (final conc. 100 M each) UNIFic 0.3 µL (30 pmol of each primer) UNIRic 0.3 µL Taq 2.0 U dH20 fill to 25 µL

Use 1 L of template DNA (from abdomen extraction eluted in 50 L dH20).

*The remaining product of this PCR can then be sent for Sanger Sequencing to elucidate from what animal and species the mosquito took a bloodmeal.

Appendix G Mammalian host blood meal species identification by PCR and sequencing* *Updated 2/8/19 JP

This PCR diagnostic identifies whether mosquitoes have taken vertebrate blood meals. Currently, one product of size ~205bp for a vertebrate blood meal (note that this is NOT a pan-vertebrate PCR, but does amplify for thousands of vertebrate species including goat, human, dog, cow, and pig) is amplified from the 12S ribosomal RNA gene. This protocol has been tested on human, cow, dog, goat, and pig DNA extracted from blood or tissues. The PCR products can then be run on a 2% agarose gel. The addition of the 1712_Rev primer yields a human-specific band (572 bp) in addition to the 205 bp vertebrate band in blood meals containing human blood.

Primers: UNIFic: 5’- GGA TTA GAT ACC CCA CTA TGC -3’ UNIRic: 5’- GCT GAA GAT GGC GGT ATA TAG -3’ 1712_REV: 5’- CTC CTA AGT GTA AGT TGG GT -3’

PCR Program: (12Sic) 1. 94ºC 5 min 2. 94ºC 30 sec 3. 51ºC 30 sec 4. 72ºC 30 sec 5. Go to step 2 35x 6. 72ºC 7 min 7. 4ºC forever

Reaction Mixture: 25 L 10X Buffer 2.5 µL dNTPs 2.5 mM 2.0 µL (final conc. 100 M each) UNIFic 0.3 µL (30 pmol of each primer) UNIRic 0.3 µL 1712_REV 0.3 µL Taq 2.0 U dH20 fill to 25 µL

Use 1 L of template DNA (from abdomen extraction eluted in 50 L dH20).

*The remaining product of this PCR can then be sent for Sanger Sequencing to elucidate from what animal and species the mosquito took a bloodmeal.

Appendix H CSP (Circumsporozoite protein) ELISA This assay detects Plasmodium falciparum CSP protein in mosquito samples. CSP is only expressed during the sporozoite stage of malaria development, so this assay detects only sporozoite-positive mosquitoes, which are capable of transmitting malaria. The monoclonal capture antibody nonspecifically binds to the ELISA plate, after which the addition of blocking buffer prevents nonspecific binding of other proteins. After the addition of mosquito homogenate, the capture antibody binds to CSP and holds it during subsequent wash steps. After the monoclonal antibody is added, it also binds CSP and remains after washing. This antibody is conjugated to a peroxidase which catalyzes ABTS indicator solution, turning the solution green, while negative samples remain uncolored. Adapted from the Malaria Research and Reference Reagent Resource Center (MR4) Methods in Anopheles Research Manual, available at http://www.mr4.org/Publications/MethodsinAnophelesResearch/tabid/336/Default.aspx Limited amounts of Plasmodium falciparum positive controls, capture antibodies, and conjugated antibodies are available free of cost through the MR4 website ((MR #890)).

Materials PBS (phosphate buffered saline, available from MMI Dept.) BSA (bovine serum albumin) (A7906) Casein (Sigma C7078) Phenol red (Sigma P4758) IGEPAL CA-630 (Sigma I3021) Nonidet P-40 (----) NaOH (------) HCl (------) Tween (Fisher BP337) P.f. capture MAb (MR #890) P.f. conjugate MAb (MR #890) P.f. CSP positive control (MR #890) Glycerol (Sigma G6279) ABTS solution (Kirkegaard Perry) 10% SDS (sodium dodecyl sulfate) (Gibco #15553-035) 96-well U bottom vinyl ELISA plates (Corning #2797)

Solutions

BSA Blocking Buffer (BBB): 250 mL 250 mL PBS 2.5 g BSA 1.25 g casein 50 µl 0.1 g/mL phenol red stock

Add 250 mL PBS to casein and BSA along with a stir bar. Stir ~>3 hours on stir plate until dissolved. Store overnight or up to 5 days at 4°C or freeze for future use. Store BSA at 4°C.

BBB: IG-630 (mosquito grinding buffer): 5 mL 5 mL BB 25 µl IGEPAL CA-630 detergent

Boiled Casein Blocking Buffer (CBB): 250 mL 1.25 g casein 25 mL NaOH (0.1M) 225 mL PBS ~5 mL HCl (1 M)

Suspend 1.25 g casein in 25 mL 0.1M NaOH and bring to a boil while stirring on a hot plate. After casein dissolves, slowly add 225 mL PBS, allow to cool, and then adjust pH to 7.4 with HCl. Store aliquots at -20C

CBB-Nonidet P-40 (CBB-NP40) 5 uL NP-40 per each 100 uL CBB

Add NP-40 to CBB and mix thoroughly by vortexing. Make fresh daily.

PBS: Tween (wash buffer): 500 mL 500 mL PBS 0.25 mL Tween

MAb (monoclonal antibody) stock Dissolve lyophilized antibody in 1:1 dH2O: glycerol, following instructions on the bottle. Store antibody at -20°C. Make the following antibody dilutions immediately prior to use: Capture antibody: 40 µl stock in 5 mL PBS—this is enough for one 96-well plate Conjugated antibody: 10 µl stock in 5 mL BBB—this is enough for one 96-well plate

P.f. positive control stock Resuspend Plasmodium falciparum CSP protein in 250 µl BBB (vial I) Take 10 µl from vial I, dissolve in 990 µl BBB (vial II, 100x dilution) Take 10 µl from vial II, dissolve in 990 µl BBB for working stock (vial III, 10,000x dilution) For the positive control serial dilution, add 100 µl from vial III to a plate well. Transfer 50 µl of this to the next well down, mix well with 50 µl BBB. Using a new pipet tip, transfer 50 µl to the next well down, mix well with 50 µl BBB, etc., resulting in 1X, 2X, 4X, 8X, 16X, 32X, 64X, and 128X positive control dilutions.

Mosquito homogenate Grind each whole mosquito in 50 µl CBB:NP40 with sterile pestle. Rinse pestle with 125

µl CBB. Transfer 100 uL of this homogenate to a new tube (this 100 uL will be used for gDNA extraction). To the remaining homogenate, add 75 uL of CBB-NP40 for a total of 150 uL of ELISA homogenate. ELISA homogenates can be prepared in advance and stored at -20°C (or -70C for long-term).

Negative controls Homogenize uninfected colony mosquitoes as above for negative controls.

ABTS solution Pour 1-component solution into 15 mL conical, ~< 10 mL total per 96-well plate. Store at 4°C, throw away remaining solution after assay is finished.

Stop Solution 1% SDS (1 mL 10% SDS in 9 mL dH2O for one 96-well plate)

Plate Setup 1 2 3 4 5 6 7 8 9 10 11 12 A neg (+) 1x (+) 1x (+) 1x 1 1 9 9 17 17 25 25 B neg (+) 2x (+) 2x (+) 2x 2 2 10 10 18 18 26 26 C neg (+) 4x (+) 4x (+) 4x 3 3 11 11 19 19 27 27 D neg (+) 8x (+) 8x (+) 8x 4 4 12 12 20 20 28 28 E 33 (+) 16x (+) 16x (+) 16x 5 5 13 13 21 21 29 29 F 33 (+) 32x (+) 32x (+) 32x 6 6 14 14 22 22 30 30 G 34 (+) 64x (+) 64x (+) 64x 7 7 15 15 23 23 31 31 H 34 (+) (+) (+) 8 8 16 16 24 24 32 32 128x 128x 128x

ELISA Protocol Note: All incubations are carried out at room temperature. 1. Add 50 µl capture MAb solution to each well (40 µl MAb in 5 mL PBS). Cover and incubate overnight.

2. Remove solution by knocking plates upside-down. Fill wells with BBB (~220-250 µl) and incubate for 1 hour.

3. Remove solution and add 50 µl mosquito homogenate, positive controls, and negative controls to their respective wells. Run all mosquito samples in duplicate. Add 50 µl BBB to any empty wells. Incubate for 2 hours.

4. During the 2 hour incubation: - Prepare the ABTS solution - Dilute the conjugate MAb in BBB as described above (10 µl MAb in 5 mL BBB). - Confirm enzyme activity by mixing 5 µl conjugate MAb with 100 µl ABTS. A dark green color should begin developing within a few minutes.

5. Remove mosquito homogenate. Wash plate 7 times with PBS-Tween using a plate washer.

6. Add 50 µl conjugate MAb to each well, incubate for 1 hour.

7. Remove conjugate MAb, wash 7 times with PBS-Tween.

8. Add 100 µl ABTS solution to each well and incubate for 60 minutes.

9. Add 100 µl Stop Solution to each well and read plate absorbance at 405 nm.

10. The absorbance cut-off for positive samples in 2X the average absorbance of the negative controls.

DNA Extraction from CSP ELISA Homogenate 1. Transfer 100 µl of CSP ELISA homogenate to clean 1.5 mL microcentrifuge tube (this occurs during homogenization for ELISA).

2. Follow Qiagen DNeasy Blood and Tissue Kit instructions for tissue samples, with the following modifications: a. Step 1a: Incubate at 56C for 2.5 hr. b. Step 2: Incubate samples at 56C for 10 min after adding 200 uL Buffer AL (even though they are technically not blood samples). c. Add 30 sec to all centrifugation steps (i.e. 1.5 min instead of 1 min) as the centrifuge takes a little while to get up to speed and also to ramp down. d. Elute DNA in 100 uL of Buffer AE. Final centrifugation at 8,000 rpm.

Appendix I Ciubotariu, March 2019

Protocol for Pfldh qPCR* The Pfldh qPCR is a real-time PCR assay that is used to detect the falciparum-specific single copy lactate dehydrogenase (pfldh) gene. This qPCR assay targets an 85bp fragment of the gene using SYBR chemistry and was originally tested on dried blood spots for which it was found to be a more sensitive diagnostic tool to detect malaria when compared to microscopy and RDT. It has now been expanded to detect parasite load from genomic DNA extracted from mosquito thoraces. Because this is a SYBR- based detection assay, it may generate false positive signals from any double-stranded DNA, thus it is important to confirm the expected melting temperature (Tm) of 76-77°C for any positive signal.

Primers: Pfldh F: 5'- ACG ATT TGG CTG GAG CAG AT -3' Pfldh R: 5'- TCT CTA TTC CAT TCT TTG TCA CTC TTT C -3'

Instructions:

1) Prepare the Excel file representation of the plate to ensure correct volume of reagents prior to preparation: always prepare 20% more for mixture of water, primers, and SYBR Green PCR Master Mix - catalog #4309155) (see below for #Y samples).

2) Remove the following from -20° freezer and place in an ice bucket: a. tubes containing mosquito thoraces extracted DNA that will be used in current plate b. Pfldh forward and reverse primers c. dH2O 3) Create standards through five 10-fold serial dilutions of stock P. falciparum 3D7 DNA following the dilutions show below (these calculations were done using C1V1 formula). Make sure stock solution is vortexed lightly and spun down, and repeat for each standard. Place standard dilutions in ice bucket.

4) Lightly vortex tubes with DNA and positive controls and spin down. 5) Combine all other components into a MasterMix tube (dH2O, forward and reverse primers, and cold SYBR Green mixture). Use a pipette to mix MasterMix and then transfer it into a plastic reservoir. 6) Load 21μL of Mastermix into each well (as needed) of the 96 well plate using a multichannel pipette. In remaining wells, load 4μL of sample DNA or standard dilution or controls into respective wells as assigned in Excel diagram 7) Cover the 96-well plate with the sealing film and place plate into a StepOne Real- Time PCR System (use same machine each time if possible - #2) with the following settings (saved as Pfldh.template on machine): a. 50°C for 2 minutes b. 95°C for 10 minutes 45 cycles of: c. Denaturation at 95°C for 15 seconds d. Annealing at 60°C for 1 minute

*Adapted from [89]

J. Jordan Hoffman

Home Address: Office Address: 128 N Collington Avenue Johns Hopkins Bloomberg School of Baltimore, Maryland 21231 Public Health Cell: (443) 500-6626 615 N. Wolfe St., E3402 Email: [email protected] Baltimore, MD 21205

Education Expected May 2019 Master of Science (ScM) Department of Molecular Microbiology and Immunology Johns Hopkins Bloomberg School of Public Health, Baltimore, MD Thesis: Investigating the behavior and ecology of a secondary vector in southern Zambia (Dr. Douglas Norris)

May 2016 Bachelor of Science (BS) College of Agriculture, Food, and Natural Resources University of Missouri, Columbia, MO Sustainable Agriculture

May 2016 Bachelor of Arts (BA) College of Arts and Science University of Missouri, Columbia, MO Interdisciplinary Studies with emphases in Peace Studies, Women’s & Gender Studies, and Health Sciences

Certificates May 2016 Multicultural Certificate, University of Missouri, Columbia, MO Digital Global Studies Certificate, University of Missouri, Columbia, MO Peace Corps Preparation Certificate, University of Missouri, Columbia, MO

Research JAn. 2018-present Master’s Student, Department of Molecular Microbiology and Immunology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD Supervisor: Douglas Norris, PhD • Collected anopheline mosquitoes in the field in southern Zambia • Speciated captured mosquitoes using morphology • Confirm mosquito species and identify bloodmeal sources by isolating DNA and performing polymerase chain reactions • Perform ELISAs and qPCRs to detect Plasmodium in mosquitoes • Build phylogenetic tree of mosquito species • Analyze data spatially using ArcGIS and R software • Conduct literature review and compile results in a thesis • Troubleshoot and optimize molecular assays with team • Coordinate with international team members •Present work to department in final defense

Nov. 2017-JAn. 2018 Graduate Researcher, Department of Molecular Microbiology and Immunology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD Supervisor: Photini Sinnis, PhD • Dissected Anopheline mosquitoes to remove salivary glands for Plasmodium sporozoite isolation and quantification • Performed ELISAs with live sporozoites to measure sporozoite motility under various conditions • Visualized ELISAs under a fluorescence microscope to compare quantity and quality of sporozoite trails • Presented results to students and faculty in departmental research forum

May 2016-Aug. 2016 Research Technician, Department of Plant Sciences, University of Missouri, Columbia, MO • Maintained Drosophila melanogaster and Drosophila suzukii colonies • Conducted chemical ecology experiments with Drosophila suzukii testing their preference for volatile chemicals • Assisted with watering and planting for drought experiments involving wheat and aphid feeding behavior • Sorted and identified prairie insect species to compare restored and native prairie fauna

Experience JAn. 2019-Mar. 2019 Teaching Assistant, Vector Biology and Vector-Borne Disease, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD • Set up class sessions for Vector Biology graduate level course • Prepare and present exam review sessions for class of 35 students • Write and grade midterm and final exams • Answer student questions and concerns

May 2018-Dec. 2018 Emergency Room Advocate, TurnAround Inc., Baltimore, MD • Responded to sexual assault, domestic violence, or human trafficking cases in the hospital emergency department • Connected clients with counseling, legal advocacy, and shelter resources as appropriate • Explained resources and options for next steps to client, including hospital treatment and legal proceedings • Advocated for client’s rights and needs within complex hierarchies and between disparate staff

Nov. 2017-Dec. 2017 Course Coordinator, YO! Baltimore Reproductive Health, Baltimore, MD • Taught six-week reproductive health course to 18-24 year olds once a week • Developed and adapted lesson plans concerning sexually transmitted infections and pregnancy • Compiled curricula into digital and physical folders for future use

Oct. 2017-Dec.2017 Teacher, Community Adolescent Sexual Education, Patterson Park Public Charter School, Baltimore, MD • Taught eight-week sexual education curriculum to a class of twenty seventh graders once a week • Adjusted lesson plans as necessary and kept provided resources up to date

Feb. 2017-Aug. 2017 Americorps Member/Production Assistant, Civic Works’ Real Food Farm, Baltimore, MD • Planted, cared for, and harvested vegetables and fruits on an urban farm • Operated farm-owned vehicles for supply pick-up and produce delivery

• Built and repaired farm high tunnels, greenhouses, and sheds • Trained volunteer groups in soil preparation, planting, weeding, and harvesting • Prepared and led workshop teaching community members about vermicomposting • Implemented integrated pest management practices in preparation for and in response to pest issues • Performed physical labor in all weather conditions, including temperature extremes, snow, rain, and hail

Feb. 2016-Mar. 2016 Teaching Assistant, Medical Microbiology and Immunology Lab, University of Missouri, Columbia, MO • Taught a group of eleven students lab techniques and concepts in coordination with their lecture class • Graded lab notebooks and exercises for student feedback • Helped prepare students for their lab and lecture exams with study guides and reviews each lab

Aug. 2015-May 2016 Student Manager, Campus Dining Services Plaza 900, Columbia, MO • Streamlined training and evaluation record keeping for all student employees • Developed & implemented training program for student supervisors • Created new discipline system for student employees • Interviewed applicants and completed the hiring process for those selected

Scholarships & Awards Aug. 2012-May 2016 Bright Flight Scholarship Aug. 2012-May 2016 Curators Scholar Aug. 2012-May 2016 National Merit Scholar Aug. 2012-Dec. 2012 Virginia Booth Memorial Scholarship Aug. 2012-Dec. 2012 Albert Holman Memorial Scholarship Aug. 2013-May 2014 C R Johnston-Farm Bureau Scholarship Aug. 2013-May 2016 Roger Mitchell Scholarship Aug. 2014-Dec. 2014 Campus Dining Services Student Employee Scholarship JAn. 2015-May 2015 Luverne Walton Scholarship Aug. 2015-May 2016 MO Student Unions/US Bank Social Justice Scholarship May 2018-Jul. 2018 Global Health Established Field Placement Award Aug. 2018-May 2019 Maryland Senatorial Scholarship Aug. 2018-Aug. 2019 AAAS/Science Program for Excellence in Science

Professional Memberships 2018-Present American Society of Tropical Medicine and Hygiene, Member 2018-Present American Association for the Advancement of Science, Member 2018-Present Tropical Medicine Dinner Club, Member and Student Representative 2017-Present Maryland Entomological Society, Member

Poster Presentations Hoffman, J.E., Simubali, L., Mudenda, T., Moss, W.J., Norris, D.E., Stevenson, J.C. Exploring the bionomics of a secondary vector of malaria in southern Zambia. Presented at the 4th annual JHMRI Future of Malaria Research Symposium, Rockville, Maryland, 2018.

Hoffman, J.E., Simubali, L., Mudenda, T., Moss, W.J., Norris, D.E., Stevenson, J.C. Investigating the biology & behavior of Anopheles squamosus and its role in malaria transmission in southern Zambia. Presented at the 67th annual meeting of the American Society of Tropical Medicine and Hygiene, New Orleans, Louisiana, 2018.

Kobayashi, T., Bobanga, T., Umesumbu, S., Schatz, S., Gebhardt, M., Ciubotariu, I.I., Hoffman, J.E., Jones, C.M., Stevenson, J.C., Norris, D.E., Moss, W.J. The impact of mass bed net distribution on vector species and malaria prevalence in Kilwa and Kashobwe, Haut-Katanga Province, Democratic Republic of the Congo. Presented at the 67th annual meeting of the American Society of Tropical Medicine and Hygiene, New Orleans, Louisiana, 2018.

Hoffman, J.E., Simubali, L., Mudenda, T., Moss, W.J., Norris, D.E., Stevenson, J.C. Investigating the biology & behavior of Anopheles squamosus and its role in malaria transmission in southern Zambia. Presented at the Johns Hopkins Bloomberg School of Public Health’s Annual GIS Day, Baltimore, Maryland, 2018.

Hoffman, J.E., Simubali, L., Mudenda, T., Moss, W.J., Norris, D.E., Stevenson, J.C. Evaluating the role of a secondary vector in residual malaria transmission in southern Zambia. Presented at the Johns Hopkins Bloomberg School of Public Health’s Global Health Day, Baltimore, Maryland, 2019.