List of Figures
Figure 1. List of world’s deadliest animals
Figure 2: Chart depicting the trend of malaria cases and deaths in India from the years 2001 to 2014
Figure 3. Life cycle of Plasmodium species
Figure 4. Algorithm for diagnosis and treatment of malaria
Figure 5. Schematic representation of the parasite numbers during various stages of Plasmodium developmental cycle.
Figure 6. Distribution map of the distribution of Anopheline species across Asia-Pacific regions.
Figure 7. Distribution of An. stephensi in India during the (a) pre-eradication era (1927–1958); (b) Post eradication period (1959–1999) and (c) Current scenario (2000–2015).
Figure 8. Computerized drawings showing the characteristics of Anopheline species
Figure 9. Pictorial keys to identify Anopheles at distinct stages.
Figure 10. Schematic of the application of various transcriptomic analyses of mosquitoes.
Figure 11: Workflow summary of RNA sequencing using TruSeq stranded mRNA sample preparation and sequencing kit.
Figure 12. RNA-seq data analysis pipeline.
Figure 13. Read quality representation of RNA-seq data – boxplot showing the average Q-score for each base in the read.
Figure 14. Heatmap representation of the correlation of data between the replicates within a tissue and between tissues.
Figure 15. Boxplot showing the FPKM distribution across the replicate datasets for the four tissues.
Figure 16. Representation of transcripts that belong to the three class-codes “=”, “j” and “u”.
Figure 17. A. Heatmap representation of differentially expressed transcripts in the four An. stephensi tissues – midgut, Malpighian tubules, ovary and fat body; B. Volcano plot matrix depicting the comparison of transcripts expressed in each of the four tissues (midgut, Malpighian tubules, ovary and fat body) with that of the others.
Figure 18. Gene ontology analysis of transcripts upregulated in female An. stephensi midgut. A. Cellular component B. Molecular function and C. Biological processes.
Figure 19. Distribution of upregulated transcripts in female An. stephensi Malpighian tubules. A. Cellular component B. Molecular function and C. Biological processes, based on gene ontology analysis.
Figure 20. Gene ontology analysis A. Molecular function and B. Biological processes, of transcripts enriched in ovaries.
Figure 21. Gene ontology analysis of fat body enriched transcripts. Transcript distribution in A. Cellular component B. Molecular function and C. Biological processes.
Figure 22. Composition of 4-plex iTRAQ labels.
Figure 23. Representative spectra of an iTRAQ-labeled peptide showing the reporter ions (inset).
Figure 24. Figure representing the iTRAQ-based proteomics workflow employed in the study.
Figure 25. SDS PAGE of total proteins isolated from 4 mosquito tissues (Malpighian tubules – MT; fat body – FB; ovary – Ova and midgut – MG) for input normalization
Figure 26: SCX profile plot of 4-plex iTRAQ labelled peptides separated over a 50-min run on a PolySULFOETHYL A column (PolyLC)
Figure 27: Summary of differentially expressed proteins in midgut, Malpighian tubules, ovary and fat body. Red color represents upregulated, while green color represents downregulated proteins in a given tissue compared to the other individual tissues and together (diagonal boxes with thick borders).
Figure 28. Comparison of biological processes in which differentially proteins expressed in female An. stephensi midgut are involved.
Figure 29. Biological processes of overexpressed (blue) and downregulated (yellow) proteins in Malpighian tubules of female An. stephensi mosquitoes.
Figure 30. Comparison of biological processes that are enriched among the proteins overexpressed (blue) and downregulated in ovaries of An. stephensi mosquitoes compared to midgut, Malpighian tubules and fat body.
Figure 31. Biological processes enriched among the proteins that were overexpressed in fat body (blue) and those that were downregulated in fat body (yellow) compared to midgut, Malpighian tubules and ovaries.
Figure 32. Revision of An. stephensi genome assembly by integration of transcriptomic and proteomic data.
Figure 33: Heatmap representation of RNA and protein expression across four mosquito tissues – midgut, Malpighian tubules, ovaries and fat body.
Figure 34. Heatmap representing the expression pattern of immune related genes.
Figure 35. Heatmap representation of vitellogenesis-related gene expression at transcript and protein levels
List of Tables
Table 1. Transcript distribution – number of transcripts in total, class code-based classification of transcripts in all four tissues and in individual tissues
Table 2: List of top 50 differentially expressed transcripts in midgut compared to transcripts expressed in Malpighian tubules, fat body and ovaries
Table 3. List of top transcripts found to be differentially expressed in female An. stephensi Malpighian tubules compared to those expressed in midgut, fat body and ovaries
Table 4. Top 50 differentially expressed transcripts in An. stephensi ovaries compared to midgut, Malpighian tubules and fat body
Table 5. Differentially expressed transcripts (top 50) in fat body compared to midgut, Malpighian tubules and ovaries
Table 6: Gradient used for SCX fractionation
Table 7: Partial list of proteins differentially expressed proteins in midgut of female An. stephensi
Table 8. Partial list of proteins found to be differentially expressed in female An. stephensi Malpighian tubules
Table 9. Partial list of differentially expressed proteins in ovaries of An. stephensi
Table 10. Partial list of fat body proteins that were found to be differentially expressed compared to that of midgut, Malpighian tubules and ovaries
Anopheles stephensi is one of the major malaria causing mosquito in India and other South Asian regions [Kumar et al., 2012, Sinka et al., 2012, Sinka et al., 2011]. There are three major variants – type form, which is the main competent vector; -mysorensis strain, which is mainly zoophilic but an important vector of Iran and -an intermediate form, whose vectorial capacities is relatively less studied. An. stephensi type form is a prominent urban vector of India and is reported to be the cause of about 12% of malaria transmission here [Gakhar et al., 2013, Hati, 1997]. Due to this, study and control of this vector forms an important basis for the control of malaria transmission in India. Most of the molecular level investigations on malaria vectors pertaining to vector behavior and competence, vector-pathogen interactions, transgenic mosquitoes and their applications, transmission-blocking vaccines have been carried out on Anopheles gambiae, because of the ready availability of whole genome sequences of different strains and several large-scale gene expression data derived using cDNA microarrays, NextGen sequencing technologies and mass spectrometry. However, An. gambiae is not of importance in the Indian scenario [Kumar et al., 2012, Sinka et al., 2011, Sharma, 1999]. Similar molecular level investigations of An. stephensi have been limited because of the lack of whole genome sequence data. The recent availability of its genome has provided an important avenue of research, which can be explored to efficiently understand the biology of this important Indian vector [Neafsey et al., 2015, Jiang et al., 2014]. Therefore, the Institute of Bioinformatics, Bangalore, in collaboration with National Institute of Malaria Research, field station, Goa are collaborating to investigate molecular profiling of the Indian vector mosquito Anopheles stephensi.
The goal of this Ph.D. study is to investigate the transcriptome and proteome profiles of four mosquito tissues – midgut, Malpighian tubules, fat body and ovary, using next generation sequencing technologies and high-resolution mass spectrometry. This investigation identifies the transcripts and proteins that are commonly and differentially expressed across these tissues. Functional annotation of the genes expressed in these tissues have been investigated using bioinformatics analysis to provide insights into the biology of this vector. Further, transcript and protein sequences derived from this study provide the data for validation and refining of predicted gene models in An. stephensi.
Mosquitoes are key representatives of the arthropod vectors, belonging to the family Culicidae. Both male and female mosquitoes feed on nectar and plant juices for energy. However, the females of most mosquito species require additional nutrition for the development of eggs, which they fulfill by feeding on the blood of various hosts, mostly vertebrates, through their specialized mouthparts known as proboscis. The female proboscis is a long needle-like structure that can pierce the host skin to feed on host blood. Owing to this feeding habit, the female mosquitoes present a major threat to humans and animals alike, due their ability to carry and transmit important disease-causing pathogens such as plasmodium, viruses and other parasites. Mosquitoes have earned the distinction of being the deadliest animal with an average of about 750,000 deaths attributed to mosquito bites every year, according to an article published in 2016 (Figure 1). About 3,500 species of mosquitoes have been reported to belong to the Anopheles sub-set from all parts of the world, of which about 40 are known to transmit malarial parasites. Efficient control of mosquitoes is a major task in the control and prevention of major debilitating diseases such as malaria, chikungunya, encephalitis, dengue and zika virus infections, among many others.
1.1. Malaria: current statistics
Malaria is a major killer with about 429,000 deaths worldwide and another 200 million cases of debilitation for several days according to the WHO World Malaria Report [Who, 2016]. India ranks third in the countries with the highest number of malaria related deaths, contributing to 6% of the total deaths from P. falciparum and P. vivax cases. Majority of the deaths in Africa are due to P. falciparum infections, while P. vivax infection has been reported to be a major cause of deaths in India with 51% of the total malaria deaths being from P. vivax infections [Who, 2016]. Claims of ~13 times higher malaria attributed deaths in India than the WHO estimates have been reported and intensifies the gravity of the situation in India [Kumar et al., 2011]. According to the report published by the Directorate General of Health Services as part of the National Vector Borne Disease Control Programme, Figure 2 represents the trend of malaria cases in India. It shows that although there has been a consistent decline in malaria cases from 2.08 million to 1.10 million during 2001 to 2014 in India, it is still significant. The peak in the number of deaths in 2006 is reported to be due to an epidemic reported in the North-Eastern States.
1.2. Life cycle of Plasmodium
Malaria is caused by the infection of the protozoan parasite belonging to the genus Plasmodium. There are over 100 species of Plasmodium that is known to infect various vertebrates. However, about 5 of them are known to infect humans under various conditions. The species causing malaria in humans are
1. Plasmodium falciparum: This is the deadliest species of Plasmodium distributed in the tropical and subtropical regions, especially in Africa. Due to its ability to rapidly multiply in the blood, it is known to cause severe anaemia. In addition, these species are reported to aggregate in the blood vessels, clogging the capillaries of the brain. This leads to the fatal complication called cerebral malaria.
2. Plasmodium vivax: This is the major species of Plasmodium found in Asia and Latin America. It has a dormant liver stage that can activate after several months or years of the actual infection and invade the blood. Therefore, the detection of an infection by P. vivax is harder compared to P. falciparum.
3. Plasmodium ovale: found mostly in the west African region, it is biologically and morphologically similar to that of P. vivax.
4. Plasmodium malariae: it is the only species of Plasmodium having a three-day/quartan cycle, while most other species have a two-day/tertian cycle. Distributed worldwide, this species can cause a chronic infection, which if untreated can last a lifetime leading to complications such as nephrotic syndrome.
5. Plasmodium knowlesi: usually found as a natural pathogen of long-tailed and pig-tailed macaques in Southeast Asia, it is recently reported to be a significant cause of zoonotic malaria. Fatal cases of infection by this species have been reported owing to the rapid progression of infection, due to its 24-hour replication cycle.
The life cycle of these Plasmodium species is fairly similar and comprises of a complex life cycle which can be broadly divided into two phases based on the host: (a) the mosquito phase and (b) the vertebrate phase. During a blood meal, if a mosquito feeds on an infected host, the Plasmodium species gains entry into the mosquito, in the form of gametocytes within the blood cells. These gametocytes enter the sporogonic cycle within the mosquito wherein they egress from the RBCs to differentiate into male and female gametes. These male gametes undergo ex-flagellation with the DNA replicating to 8N leading to the formation of eight haploid microgametes. The microgametes further search for the female gametes to fertilize, forming diploid zygotes within the mosquito midgut (Figure 3). The zygotes elongate and become motile ookinetes, which invades through the midgut wall to develop into oocysts within the wall of the mosquito midgut. These oocysts grow and rupture to release numerous sporozoites in to the hemolymph of the mosquito, through which it reaches the salivary glands [Smith et al., 2014]. Here, the sporozoites further mature and is released into the blood of another individual during the next blood feed cycle of the mosquito, thereby spreading the Plasmodium to another individual. This sets in the human/vertebrate phase of the Plasmodium life cycle. The plasmodium undergoes two stages within the human host – the exo-erythrocytic stages mainly in the liver and the erythrocytic stages. The sporozoites entering the humans after a bite from the infected mosquito enter the liver and infect the hepatocytes to replicate and develop into mature schizonts within the hepatocytes. These schizonts may remain dormant for months or years as in the case of infection with P. vivax or P. ovale or rupture to release the merozoites into the blood stream entering the erythrocytic schizogonic stages. The merozoites infect red blood cells to form the ring stage trophozoites, which mature into schizonts, rupturing and releasing merozoites to further the erythrocytic infection, leading to the clinical manifestations of the disease. Some of the parasites differentiate into gametocytes or the sexual erythrocytic stages which completes the life cycle within the mosquito vector upon ingestion.
1.3. Treatment strategies and combating malarial infections
Early diagnosis of malaria is critical to the successful treatment and the control of the spread of the disease. The diagnosis is based on the microscopic evaluation of blood smear or rapid diagnostic tests based on the circulating antigens in the blood. Though microscopic evaluation remains the gold standard and most sensitive, it is labor intensive. Rapid diagnostic tests are less specific to the Plasmodium species and care should be taken to choose the appropriate one for distinct species. It has been reported that these kits have to be correlated with the microscopic examination to confirm the infection. Infections with P. vivax is mostly treated with a combination of chloroquine and primaquine (to target the dormant hypnozoites in liver). However, P. falciparum confirmed cases are treated with artemisinin combination therapy (ACT) along with a single dose of primaquine. As artemisinin derivatives are the only active drugs, administration of monotherapy of these drugs in uncomplicated malaria is banned in India, although injectable artemisinin is used in malarial complications. The guidelines for the diagnosis and treatment of malaria by the National Vector borne disease control program (NVBDCP) and National Institute of Malaria Research (NIMR) is outlined in the Figure 4.
1.4. Malarial Vaccines
Considering the current spread of Plasmodium species and the lack of proper confirmative methods of species confirmation apart from microscopy, efforts towards the containing the spread of malarial infection gains utmost importance. Therefore, numerous efforts have been made towards the control of spread of infection including the development of vaccines. However, due to the complex biology of parasites than viruses or bacteria and the multiple stages of development in the human host and the expression of multiple unique antigens, it is much more difficult to develop vaccines against these parasites [Aggarwal and Garg, 2017]. Early efforts towards vaccine development had been focused mostly on the pre-erythrocytic stages, which include recombinant proteins from the pre-erythrocytic stages, DNA vaccines and live attenuated vaccines [Aggarwal and Garg, 2017, Ntege et al., 2017]. Recently, there has been other approaches that include targeting the erythrocytic stages as well, which are mainly aimed at reducing the number of parasites in blood and thereby reducing the severity of the disease [Ntege et al., 2017, Hoffman et al., 2015]. Several projects aimed at development of vaccines are underway with more than 30 vaccine candidates in the advanced preclinical or clinical stages of evaluation but none has been successful yet. RTS,S/AS01 is the only vaccine to have been in the phase 3 trials and is expected to complement other malaria treatment and control programs (WHO position paper on RTS,S/AS01).
Another strategy of vaccine development has been aimed at blocking the transmission of the parasites and is called as transmission blocking strategy. Three main approaches are used in blocking the transmission including drugs targeted at gametocytes, transmission blocking vaccines and changing mosquitoes to acquire refractory traits. During the life cycle of Plasmodium within mosquito, differentiation of the gametocytes in the ingested blood to micro and macrogametes results in the fusion of gametes in to a zygote. These diploid zygotes undergo meiotic division to form motile ookinetes that traverse the thick peritrophic membrane and the midgut wall to form oocysts, which leads to the formation of sporozoites through mitotic divisions. The sporozoites travel through the hemolymph and only about 20% of the total sporozoites in the hemolymph establish themselves in the salivary glands, while the rest are eliminated. Targeting malaria transmission in mosquitoes offers a natural advantage due to 2 major bottlenecks where the parasite numbers reduce drastically (Figure 3) i) during the ookinete-oocyst stage and ii) invasion of salivary glands by oocyst sporozoites and formation of salivary gland sporozoites [Goncalves and Hunziker, 2016, Sinden et al., 2007, Smith et al., 2014].
Studies in An. gambiae have shown that among the large number of gametocytes that is ingested by the mosquito in a blood meal, only a small fraction of them develop into oocytes. These stages may provide better targets of transmission control intervention. However, such intervention methods necessitate a sound knowledge of molecular mechanisms involving the interactions of various Plasmodium stages with the mosquito vector. Several such approaches are being employed in An. gambiae to control the mosquito population and malaria transmission [Bian et al., 2013, Wilke and Marrelli, 2012, Meredith et al., 2011, Hoffmann et al., 2011]. Although these studies have enormously increased the understanding of the vectorial capacities, they are not necessarily applicable in all the vector species due to the evolutionary changes in the proteins and genes among the different species of the same genera. A few studies have embarked on employing similar transmission blocking strategies in An. stephensi [Venkatrao et al., 2017, Kajla et al., 2016] but had been largely limited due to the lack of sufficient information regarding the gene and protein products in these species till recently.
1.5. Malarial vectors and their distribution
Among about 3500 species of mosquitoes, those that transmit human malarial parasites belong to the sub-set of Anopheles. Anopheles, as a genus of mosquitoes was first introduced by Johann Wilhelm Meigen in 1818, but was hardly studied thereafter, until late 1880s. The discovery of mosquitoes transmitting microfilariae and malarial protozoa in the late 19th century, initiated a drive to collect, name and classify these insects. Although, the earlier classification by Frederick V. Theobald was based on the distribution and shape of scales on the thorax and abdomen, the current classification of Anopheles sub-genera is based primarily on the number and positions of specialized setae on the gonocoxites of the male genitalia. This basis of classification, introduced by Sir (Samuel) Rickard Christophers in 1915, has been widely accepted [Harbach, 2013]. Owing to their impact on human health, Anopheline mosquitoes is the most studied and best known among the genus. The impact of this mosquito species on humans is underscored by the emergence of sickle cell anemia trait as a mode of resistance to malarial protozoa, transmitted by these species. However, studies to understand the evolution and phylogenetic relationships of these mosquitoes were performed only recently [Harbach and Kitching, 2005, Freitas et al., 2015].
Identification of the vector and non-vector mosquito species are largely based on their anatomical features such as the presence of setae in adults or their form and arrangement in larval stages [Harbach and Besansky, 2014]. Correct identification and speciation of these vector species become important for a judicious use of resources for vector control in the economically limited areas. Owing to the majority of malaria control interventions aimed at the human-vector contact, a detailed evaluation and understanding of the distribution, species composition and behavior of the vectors is extremely important. However, closely related species are hard to distinguish using only the anatomical features and hence molecular and morphometric methods have been devised to aid in the process of speciation [Garros and Dujardin, 2013]. Several attempts of morphometric methods were introduced, including that of differentiating the Anopheles species from others based on venation patterns on their wings [Dujardin, 2011]. Nevertheless, DNA-based barcoding techniques developed in 2003, proved to be more efficient when used in addition to these methods. DNA barcoding uses the variations in short, standardized gene regions, such as the Folmer region of Cytochrome oxidase I (COI) to identify the species [Hebert et al., 2003]. The introduction of these molecular methods of species identification have greatly improved the scenario and overcome the problems of speciation.
Anopheline species is one of the largest group of mosquito species with about 500 species, including 50 unnamed species complexes reported to be distributed across the world [Hay et al., 2010]. It is also one of the best studied mosquito species among those of medical importance, with seventy of them reported to be competent vectors of malarial parasites. However, only 41 of these Anopheline species have been reported as dominant vectors for transmission of human malaria [Sinka et al., 2012, Hay et al., 2010]. Among these, about nine species complexes have been identified to be the dominant vector species in America, with An. quadrimaculatus, An. albimanus and An. darlingi being found in majority of the regions [Sinka et al., 2012, Sinka et al., 2010a]. Among the six dominant vector species present across the Europe and the middle east, An. atroparvus was found in most of the locations, followed by An. messeae [Sinka et al., 2012, Sinka et al., 2010b]. A combination of the most competent, highly anthropophilic malarial vectors including An. gambiae, An. arabiensis and An. funestus were found in the majority of the locations in Africa among the seven dominant vectors [Sinka et al., 2012, Sinka et al., 2010b]. The distribution of the malarial vectors in Asia-pacific region have been found to the most diverse, with a total of nineteen Anopheline species reported to cause significant human malaria. Among them, at least ten of them are considered to be species complexes [Harbach, 2004]. An. farauti species complex was identified in most independent sites in the Asia-pacific region, partly due to the inclusion of two more comprehensive surveys in Papua New Guinea and northern Australia that provided the majority of these sites. The other dominant vector species complexes included An. barbirostris, An. sinensis, An. culicifacies, An. annularis, An. maculatus, An. minimus and An. subpictus [Sinka et al., 2012, Sinka et al., 2011].
Asia-pacific suffering with the second highest global malaria burden, after Africa, has a diverse distribution of malarial vectors. Ten out of the 19 vectors are considered to be species complexes, which consists of closely related, morphologically indistinguishable species that either occur in sympatry displaying different behavioural patterns. Six vector species are reported to be major vectors of human malaria in India, with An. culicifacies being the major rural malaria vector causing 65% of the cases reported per annum [Dev and Sharma, 2013]. An. annularis is a major vector of north-eastern Indian states of Orissa and Assam. Anopheles stephensi is the major malarial vector of the urban areas of western and north-western India extending across to Iran and Iraq [Kumar et al., 2012, Sinka et al., 2012, Sinka et al., 2011, Dev and Sharma, 2013].
1.6. Anopheles stephensi
An. stephensi is currently reported to be second most prevalent vector in India [Singh et al., 2017]. Its taxonomical classification is:
It is further classified into three forms of ecological variants – ‘type form’, ‘intermediate form’ and ‘mysorensis’ variety, based on the egg morphology [Subbarao et al., 1987]. The egg float consisted of 14-22, 12-17 and 9-15 ridges for the type, intermediate and var. mysorensis, respectively. Among these, the type form is considered to be an efficient malaria vector in urban areas, while the transmission efficiency of the intermediate form is not yet completely understood. The mysorensis variety is known to be mainly zoophilic and hence a poor vector for human malaria. A few studies have reported the refractory nature of the mysorensis type when infected with P. berghei and P. yoeli [Shinzawa et al., 2013].
The life cycle of the mosquito consists of four stages- egg, larva, pupa and adult. The initial three stages are aquatic and lasts for about 11-15 days. The eggs are boat-shaped with tapering edges, with lateral floats consisting of ridges that help them stay afloat. After the eggs hatch in 2-3 days, the larva undergoes three molts with four instars, differentiated based on the increase in size. The larval stages have distinct head, thorax and abdominal regions. The larvae usually lie parallelly, just below the water surface and actively feed with the help of feeding brushes. The four instar stages usually last for 7-10 days before the transformation to the pupal stage. The pupal stage usually lasts upto two days and is the least active stage. During this stage, it has a large cephalothorax with a pair of trumpets dorsally and the developing mouth parts, eyes, legs and wings, ventrally. The adult emerges from the pupae with the males emerging faster than the females. The adult mosquitoes have a life span of 7-25 days. The males are smaller with bushier mouth parts and a plumose type of antennae. The females have a sharper mouth parts with antennae and a long proboscis that helps in piercing the skin during blood feeding. The adult mosquitoes generally feed on plant nectar but the females require blood meal for the development and nourishment of eggs. Therefore, the females are active participants in the transmission of parasites.
An. stephensi actively reproduces during the rainy season with June – August being the peak season and hence causing increased transmission during these months. The type form is considered to be mainly endophilic and endophagic, although they bite outdoors in the warmer months of summer. They are reported to be breeding in artificial containers in homes and water collected at construction sites and other industrial areas in the urban regions. In rural areas, their breeding grounds are wells, canals and domestic water storage tanks [Dev and Sharma, 2013, Subbarao et al., 1987]. There have been indications of variation in the anthropologic behaviour of An. stephensi depending on the availability of alternate hosts [Sinka et al., 2011].
Anopheles vectors that can transmit malaria are not only found in the malaria-endemic areas but are also present in the malaria eliminated regions and therefore, present a constant risk of re-introduction of the disease in these areas. Efficient control of mosquito populations has been the foremost goal in the control of disease transmission. Vector control methods mainly involve distribution and use of insecticide treated nets, insecticide spraying, chemoprevention where the population (especially pregnant women and children) is provided with preventive medications and development and use of vaccines. The use of insecticide treated nets and indoor residual spraying of insecticides have so far been the most common method in preventing mosquito bites, however, the spread of insecticide resistance among the mosquito species has emerged as a new threat for this method of control. Although chemoprevention has reduced the rates of malaria in specific countries, they have not been widely employed.
1.7. OMICS era in vector biology
After the establishment of arthropod vectors as a cause of the spread of many human diseases in the late 19th century, a thorough understanding of the biology and the mechanisms involved is largely elusive even now. Medical entomologists have contributed to the understanding of the vector biology and thereby aiding the vector control strategies and limit their human contact. One of the earliest vector control strategies involved the control of breeding sites, followed by the use of chemical insecticides and intervene the vector-human contact by the use of repellents and nets. However, to overcome the variety of limitations by these initial methods, recent influence of rapidly expanding genetic information from the next generation sequencing and gene editing technologies have contributed a lot to the vector control initiatives. The decreasing costs of next generation sequencing technologies have facilitated the insect vector genome investigations and provided new insights into the physiology, behaviour and evolution of these vectors species. Omics technologies unravel the key to the deeper secrets of the biology of these vectors, which are hidden in the genetic code and the functional units.
The availability of the Anopheles gambiae genome in 2002 began a new era of comprehensive gene functional and evolutionary studies from the complete genomic sequences and represented a milestone in vector genomics [Holt et al., 2002]. This was followed by the sequencing of two other important vector species – Aedes aegypti [Nene et al., 2007] and Culex quinquefasciatus [Arensburger et al., 2010], aided in the understanding of the genomic evolution between these vector groups. These studies showed that while all the three organisms had 3 chromosomes, the gene repertoire and the genome sizes were quite different. The gene repertoire of An. gambiae consisted of 12,457 genes [Holt et al., 2002], while Ae. aegypti consisted of 15,419 genes [Nene et al., 2007] and C. quinquefasciatus [Arensburger et al., 2010] was the largest, consisting of 18,883 genes. These sequencing studies also showed that the larger genomes of the Ae. aegypti and C. quinquefasciatus consisted of large number of transposable elements. While the transposons made up to 42-47% of Ae. aegypti genome, their numbers were lesser in C. quinquefasciatus with about 29% of its genome. Further refinement of the genome annotations of An. gambiae based on the transcriptomic and proteomic studies have led to the latest genome annotation with 13,763 genes including 739 non-coding genes, according to VectorBase. The sequencing of the genomes 16 Anopheline species has been a valuable addition to the existing An. gambiae genome sequence revealed the rapid evolution of the Anopheles mosquitoes. It also revealed the tendencies of these species for X-chromosomal rearrangements [Neafsey et al., 2015]. This genomic information helped in deciphering the evolutionary picture of the Anopheline species revealing that An. gambiae and An. arabiensis were the first among the major vector species complex to diverge from other members of the species complex [Fontaine et al., 2015]. The genome sequence of the Asian tiger mosquito, Ae. albopictus, which is the highly invasive vector of Dengue and chikungunya, has also been sequenced recently, offering valuable insights to this highly competent and adaptable vector [Chen et al., 2015]. With the genome of 1,967 Mb, it is the largest genome of a mosquito till date and has an abundance of repetitive sequences with an expansion of the members of gene families involved in insecticide-resistance, sex determination, immunity, olfaction and others.
Numerous studies have been attempted to bridge the gap between the genomic data and the actual applications to demonstrate the role of genomic data in designing new vector management tools. As an example, a chemical compound was discovered to activate the highly conserved insect odorant receptor co-receptor (Orco), which could be used to intervene in the host-seeking behaviour of insect vectors [Jones et al., 2011]. In another instance, introduction of a small molecule inhibitor of a potassium channel (Kir) was found to disrupt the excretion of urine, rendering the mosquito flightless or dead within 24 hours [Raphemot et al., 2013]. Such mechanisms could be used for vector control in the growing scenario of insecticide resistance. In addition, several studies have been employed to develop genetically modified male mosquitoes that are either sterile or lead to incapacitated progeny, to actively replace the natural vector populations [Winskill et al., 2015, Carvalho et al., 2015, Harris et al., 2012, Fu et al., 2010]. Similar to these several other studies aimed at the discovery of novel vector control strategies have been facilitated by the availability of the genome sequences.
Similar to the genomic studies, transcriptomic profiling of tissues with specific functions such as chemoreception, reproduction and immunity led to a much deeper understanding of these physiological processes in the vector [Dixit et al., 2011, Jiang et al., 2017, Prasad et al., 2017, Thomas et al., 2016, Jiang et al., 2015]. Such applications of sequencing-based technologies in numerous studies aimed at understanding the biological processes of the An. gambiae vector have been exploited significantly to control malaria transmission. The vector species and the diseases transmitted by them are known to be the evolutionary by-products of complex host-parasite-vector interactions. The heterogeneity between the degree of vector competence even in the closely related Anopheline subspecies are proof to the need for comprehensive studies among the individual subspecies other than inference and incidental interpretations based on the available information. Duplication of advances achieved in An. gambiae post -omic studies, in An. stephensi would lead to a better understanding of the underlying molecular mechanisms within this species. In this aspect, the tissue-based transcriptomic and proteomic expression profiles will be a crucial link and basis for further studies in An. stephensi.
In this study, we report the transcriptomic and proteomic profiles of four tissues – midgut, Malpighian tubules, ovary and fat body. These mosquito tissues play an important role in the blood meal digestion and reproduction and are also some of the critical tissues involved in plasmodium life cycle. Mosquito midgut is involved in the initial storage and digestion of the ingested blood. Blood meal induces pathways such as TOR, which ultimately leads to synthesis of proteins required for egg development. Fat body and ovary are known to be involved in the utilization of the nutrients from blood to enable vitellogenesis. Malpighian tubules are known to play an important role in the mosquito xenobiotics. Fat body cells (trophoblasts) and recently, Malpighian tubules have also been shown to be involved in the immune responses [Verma and Tapadia, 2012, Martins et al., 2011, Dow, 2009]. However, there is no data on molecular profile of these tissues of An. stephensi obtained from NextGen sequencing technologies and high-resolution mass spectrometry. During the study, we surveyed the published literature for mosquito proteins to understand the function and relevance of various mosquito proteins. This led to a compilation of 94 mosquito proteins that were reported to have a role in either promoting or inhibiting the growth of Plasmodium within the mosquito vectors. We published this data in Malaria Journal [Sreenivasamurthy et al., 2013]. Most of these molecules have been described from An. gambiae, which is not an important vector in India. This study provides the transcript and protein expression pattern of many more proteins that could play a crucial role in vector competence or vector immunity or reproduction in An. stephensi. The transcriptomic and proteomic studies of midgut, Malpighian tubules, ovary and fat body bring to light the expression profiles of several genes that may have several biological implications in various physiological conditions.
Aims and Objectives
1. Transcriptomic analysis of midgut, Malpighian tubules, fat body and ovary of female An. stephensi
2. Proteomic analysis of midgut, Malpighian tubules, fat body and ovary of female An. stephensi
3. Integration of Transcriptome and proteome and functional annotation
The complexity of eukaryotes compared to that of prokaryotes are achieved at various levels. The complex network of assembly of cells into tissues and organs form the basis of evolution of organisms. In line with the central dogma of life, the basic framework of life maintained by the DNA or genome of a given organism is constant across the plethora of cell types within that organism. The functional components represented by the transcripts and proteins form the basis of diversity of various cell and tissue types. Knowledge of the various cells and tissues of a given organism provides critical insights that characterize the tissues and represent their functions.
Specific objective 1
Tissue specific transcript expression profiling of midgut, Malpighian tubules, ovaries and fat body of An. stephensi mosquito
The diversity in the function of the genes of a given organism is achieved in part by the transcript structures and expression in the component cells and tissues. The order of complexity increases at the transcript level owing to the variations in exon-splicing and expression of the genes. Tissue specific profiles of these transcript variations – at the level of spliced forms and expression helps in understanding the molecular differences in these tissues and probable functional implications.
Specific objective 2
Tissue specific protein expression profiling of four An. stephensi tissues – midgut, Malpighian tubules, fat body and ovaries
Proteins form the functional units of a cell and thus the understanding of the protein constituents of a tissue would be more functionally relevant in characterizing the tissue. Proteins represent another level of complexity of the genes introduced by the post-translational modifications and expression variations.
Specific Objective 3
Integration of tissue-specific transcriptome and proteomes
The integration of transcript and protein profiles across the tissues provides a more comprehensive understanding of the molecular aspects within these tissues.
Transcriptomic analysis of An. stephensi
Next generation sequencing has opened new avenues of comprehensive whole transcriptome profiling studies by providing information on the quantitative transcript abundance, sequence, expression and isoform-based information for the majority of genes encoded in the genome of various vector species. In the initial stages, transcriptome studies preceded the genome sequencing studies due to the option of de novo assembly of transcripts to obtain transcript and isoform information without the genome sequence. Nevertheless, the validity and interpretation of these data in the absence of genome sequence was a challenge in itself. However, reduction in the next generation sequencing costs have bridged this gap considerably.
Transcriptomic studies in An. gambiae, like that of the genome sequencing evolved as a turning point in the vector biology research. Analysis of the transcriptome of an organism is a reflection of its physiological state at different conditions and help us understand important information regarding its vector competence. A summary of the transcriptomic analyses in An. gambiae and their applications have been well summarized in the review by Domingos et al. [Domingos et al., 2017].
An. stephensi is an important vector in the Indian malarial context and studies similar to those carried out in the case of An. gambiae is extremely important to understand the vector competence and the behavior of this important vector species. The recent sequencing of the An. stephensi genome independently by two groups has sparked a tremendous interest and activity towards understanding of the behavioral traits, physiological conditions and vector-pathogen interactions in this vector species. As is the case with most of the genome sequence drafts, the genome assembly of this vector also was far from complete. Transcriptome data has been of immense help in improving the assembly and annotation of An. stephensi genome [Neafsey et al., 2015, Jiang et al., 2014, Prasad et al., 2017]. The two genomes published by the two individual groups are of two different strains of An. stephensi type form, which is the Indian strain [Jiang et al., 2014] and the SDA strain that was originally isolated from Pakistan [Neafsey et al., 2015]. The transcriptome analysis in these studies has been at the level of whole adult mosquitoes or developmental stages [Neafsey et al., 2015, Jiang et al., 2014], the major purpose of which was to aid in the annotation of the newly sequenced genome of both the strains. However, this has been largely supplemented by numerous studies under different experimental conditions [Jiang et al., 2014, Jiang et al., 2015], at different tissue levels such as, salivary glands [Dixit et al., 2011], hemocytes [Thomas et al., 2016], midgut [Patil et al., 2009, Xu et al., 2005] and ovaries. Although not all of these have been next generation sequencing studies, they still provide significant insights to the physiology of this vector. Thus, there have been a few expression studies at the tissue level but have not been adequate covering all the tissues. Here we present the tissue-based RNA-seq expression profile of four mosquito tissues from sugar-fed female adult mosquitoes, that have not been studied earlier thereby, complementing the existing tissue-based expression splicing information profiles [Sreenivasamurthy et al., 2017] and helping in the improvement of the existing predicted gene models [Prasad et al., 2017]. The information provided by such studies is unparalleled in enhancing the knowledge of these tissues.
3.2. Materials and methods
3.2.1. Tissue dissection and RNA isolation
Midgut, Malpighian tubules, ovaries and fat body was dissected from 2 to 3-day old adult female An. stephensi mosquitoes maintained in the insectary of National Institute of Malaria Research, field station, Goa. The sugar fed mosquitoes were dissected under a stereomicroscope and the tissues were washed in PBS and stored in RNAlater. The tissues were homogenized using Precellys Lysis Kit (PEQLAB Biotechnologie GmbH) in the MINILYS benchtop homogenizer. Total RNA was isolated from the homogenized tissue using the Qiagen miRNeasy kit according to the manufacturer’s protocol. The quality of the isolated RNA was estimated using RNA nano 6000 chip in an Agilent BioAnalyzer 2100. RNA Integrity Number (RIN) for the isolated RNA was found to be in the range of 9 and 10. This was further used for the library preparation. The RNA-seq libraries were constructed for each tissue using the Illumina TruSeq RNA Sample Preparation Kit (Illumina, San Diego, CA).
3.2.2. Library preparation, cluster generation and sequencing
For the stranded RNA library preparation, 500 ng of total RNA from each of the four tissues was combined with 1 μl of ERCC RNA spike-in control mix 1 or mix 2 (cat. #s – 4456740, 4456739, Life Technologies), which was diluted 100 times. Poly adenylated mRNA was captured using oligo-dT magnetic beads provided in the library preparation kit in two rounds. During the second round of purification, the captured mRNA is fragmented using heat and primed using random hexamers for the cDNA synthesis. Post-first and second strand synthesis, the template RNA was removed and purified ds cDNA was taken through end repair, adenylation of 3’ ends, and adapter ligation steps. Enrichment of the successful adapter ligated library was carried out by 15 cycles of
PCR and the size distribution was validated on the Agilent BioAnalyzer using a DNA 1000 kit. The final library was prepared after size selection from a band between 210-500bp with a peak at approximately 286 bp. All libraries were quantitated with Qubit 2.0 Fluorometer (Life Technologies, Grand Island, NY) and each library was loaded onto two lanes with one sample per lane. The RNA-Seq libraries were clustered using Illumina TruSeq SBS kit v3. flow cell and were then further sequenced on an Illumina HiScanSQ system (Illumina, San Diego, CA).
3.2.3. RNA-seq data analysis and generation of transcript models
The quality of the raw reads was assessed using FastQC (Version 0.10.1) tool. Figure 5 represents an example of the read quality. Raw reads from both lanes of the flow cell were quality filtered to remove ambiguous bases due to sequencing errors at the 3’ end of the reads using fqtrim v0.9.4. A Phred-based score threshold of Q≥20 was applied for the base quality and only reads longer than 60 bp after trimming were considered for further analysis. Initial alignment and assembly of the transcripts was carried out using the TopHat and Cufflinks pipeline as described in Genome Research paper (Prasad, TSK., et al., 2017). The analysis was repeated using the latest HISAT-StingTie pipeline outlined in the Nature Protocols paper (Pertea, M. et al., 2015). Anopheles stephensi (Indian strain) genome (build ASTEI2) was downloaded from “VectorBase” (http://www.vectorbase.org/) and indexed in HISAT (Version 2.1.0) (Kim D. et al., 2015) aligner using default parameters. Known annotations and Gene Transfer File (GTF) – AsteI2.2 from VectorBase was supplied to HISAT2 aligner and reads from each lane for individual tissue was aligned against the indexed genome to obtain eight ‘Binary Alignment Map’ (BAM) files. Merged BAM file for each tissue was obtained by combining the replicate BAM files. Transcript assembly was carried out using StringTie (version 1.2.1) assembler (Pertea, M. et al., 2015) against AsteI2.2 gene annotations, as reference. Annotation of assembled transcripts into known and novel categories was performed using ‘gffcompare’ in StringTie package as described earlier (Pertea, M. et al., 2016). StringTie assemblies built for four tissues were merged using StringTie-merge option. Classification of assembled transcripts as annotated, alternate isoforms and intergenic transcripts and their expression was obtained by comparing the merged StringTie assembly to VectorBase annotated transcripts using gffcompare. Figure 12 outlines the data analysis workflow followed for the transcriptomics data from the four tissues.
StringTie output files from individual tissues were merged using the StringTie-merge option and the differential expression of the transcripts in each of the four tissues were identified using the Cuffdiff software. The data from the two lanes were treated as technical replicated and normalization of the reads were performed by calculating the Fragments per Kilobase of exon per Million Fragments Mapped (FPKM). R-package (version 2.18.0) was used for the computation of the differential expression using Cuffdiff and cummeRbund was used for the visualization of the RNA-seq data including the heatmaps and the clustering analysis [L. Goff, 2013]. The data was filtered based on their FPKM values for the differential expression analysis. Only those transcripts with FPKM values of ≥ 0.1 in at least one of the four tissues were considered for tissue-specific expression analysis. The identification of significant tissue-specific transcripts were based on a right tailed t-test.
To identify the long non-coding RNAs, the coding potential of identified transcripts was estimated using Coding Potential Assessment Tool (CPATv1.2.2). Transcripts longer than 200 bp with a CPAT score of <0.39 (set for the prebuilt fly gene models) were predicted to be long non-coding RNAs.
RNA sequencing of four adult female An. stephensi mosquito tissues resulted in about 500 million paired-end reads with about 55 million reads per tissue from technical replicates. Of the reads obtained >90% had a Q-score of 20 and above. Figure 13 provides a representative read quality scores from the FastQ analysis. During the filtering process, reads with the Phred score of <20 and <60bp were trimmed using fqtrim. About 6% of the reads from each tissue were excluded based on these stringent criteria. The alignment and assembly of these reads against An. stephensi genome from VectorBase resulted in the identification of 25,795 transcripts in total corresponding to more than 13,000 gene loci. Transcripts that had an FPKM value of ≥0.1 in at least one tissue was considered to be expressed. Upon FPKM-based filtering only 23,009 transcripts remained. The distribution of these transcripts according to their classification is provided in Table 1 along with the transcripts identified in each of the tissues.
Table 1. Transcript distribution – number of transcripts in total, class code-based classification of transcripts in all four tissues and in individual tissues
All 4 tissues Midgut Malpighian tubule Ovary Fat body
Total number of transcripts identified 23,009 17,461 18,812 18,616 18,685
Corresponding gene location identified 13,099 10,357 11,107 10,973 11,371
Total number of known/annotated transcripts – “=” 9,722 7,508 7,883 8,001 8,015
Number of alternate isoforms/transcripts – “j” 8,820 7,603 8,232 7,992 8,037
Number of novel transcripts (intergenic) – “u” 2,694 2,136 2,458 2,396 2,398
Possible pre-mRNA fragment – “e” 93 77 86 86 87
Intronic transfrag – “i” 80 64 71 65 74
Exonic overlap with reference transcript – “o” 655 592 616 613 618
Possible polymerase run-on fragment – “p” 324 280 307 289 313
Repeat – “r” 356 261 326 307 304
Intronic on opposite strand – “s” 1 0 1 1 1
Exonic overlap with reference on the opposite strand – “x” 264 214 239 227 235
The correlation between the replicates are represented in Figure 14 based on a distance matrix. It was found that there was good agreement between the replicates of a given tissue with the maximum difference noted in the midgut replicates. Figure 15 representing the log of FPKM values across the replicate datasets for the four tissue shows that the median FPKM is comparable across the tissues with the median FPKM in the range of 8-10. The transcripts identified in the four tissues varied in length with majority of them in the range of 1000 to 3000 bp. The transcripts were categorized into various classes based on the class codes using the VectorBase gene annotation as the reference by gffcompare. Transcripts belonging to the three major class codes, i.e. “=”, “j” and “u” were used for the differential expression analysis of the identified transcripts across the four tissues. Figure 16 provides the representation of the three classes of transcripts that were used for further analysis. The differential expression of transcripts was computed against individual tissues and in comparison, with all the tissues.
3.3.1. Transcripts expressed in midgut of female An. stephensi
About 17,400 transcripts in total were identified in the adult midgut with an FPKM of ≥0.1. Among the four tissues, we identified the lowest number of transcripts in midgut, and these corresponded to 10,357 gene loci. We identified 7,508 of the known/annotated transcripts with an additional 7,603 transcripts assembled as alternative transcripts in these annotated gene loci. Most of the transcripts had multiple alternatively spliced forms that were not annotated for a given gene loci in the reference gene build. Therefore, with this data we provide evidence for at least 7,603 transcript isoforms that are expressed in midgut. In addition, we identified reads in the intergenic regions of An. stephensi that were not annotated for transcription. Our analysis showed about 2,136 transcripts were transcribed from these unannotated genome regions, which could be potentially novel gene coding regions. Efforts of proteogenomics analysis of annotating and improving the annotation of the genome using transcriptomic and proteomic approaches forms the basis of our paper published in Genome Research (Prasad et al., 2016). Among the transcripts that were identified, 859 transcripts were found to be significantly differentially expressed in midgut compared to other tissues. Among these, 472 were found to be significantly upregulated ≥2-fold in midgut compared to that of other tissues, while 387 of them were found to be significantly downregulated (≥2-fold) in midgut compared to other tissues. The list of top 50 transcripts that were differentially expressed in midgut is provided in the Table 2. Of the transcripts that were found to be upregulated, 209 of them were previously annotated while 217 were alternatively spliced and 46 were intergenic transcripts. Similarly, among those found to be downregulated, 104 were previously annotated, 232 were alternatively spliced forms and 51 intergenic transcripts. Gene ontology analysis of the transcripts upregulated in midgut showed that these transcripts are mainly membrane components and are involved largely in the process of proteolysis (Figure 18).
Table 2: List of top 50 differentially expressed transcripts in midgut of female An. stephensi compared to that of Malpighian tubules, fat body and ovaries
Transcripts upregulated in midgut compared to other tissues
Transcript_ID Gene_ID FPKM Log2(fold change) Average p-value
MG MT Ov FB MG/MT MG/Ov MG/FB Average
ASTEI04993-RA ASTEI04993 51497.3 171.6 39.5 493.5 8.2 10.3 6.7 8.4 0.000
ANSTF.8643.1 ASTEI09365 722.3 2.4 0.2 66.7 8.2 11.8 3.4 7.8 0.001
ANSTF.9505.3 ASTEI10274 4227.3 26.6 1.0 20.6 7.3 12.1 7.7 9.0 0.000
ASTEI00941-RA ASTEI00941 1306.4 6.5 3.2 13.0 7.7 8.7 6.6 7.7 0.000
ASTEI10276-RA ASTEI10276 4502.1 37.8 2.3 23.8 6.9 10.9 7.6 8.5 0.000
ANSTF.2802.3 ASTEI02990 16096.1 196.8 41.9 90.1 6.4 8.6 7.5 7.5 0.000
ASTEI07506-RA ASTEI07506 1311.6 12.0 1.8 8.0 6.8 9.5 7.3 7.9 0.000
ASTEI02990-RA ASTEI02990 2612.3 27.2 14.5 17.1 6.6 7.5 7.3 7.1 0.000
ANSTF.9505.1 ASTEI10274 243.6 2.4 1.8 1.9 6.7 7.1 7.0 6.9 0.000
ANSTF.3931.2 ASTEI04209 190.4 1.8 0.4 1.7 6.8 9.0 6.8 7.5 0.000
ANSTF.5961.1 ASTEI06382 76.9 1.2 0.3 0.7 6.0 8.0 6.8 6.9 0.000
ANSTF.3930.2 ASTEI04208 586.9 7.1 1.4 5.4 6.4 8.7 6.8 7.3 0.000
ASTEI09414-RA ASTEI09414 2178.9 73.5 5.3 21.2 4.9 8.7 6.7 6.8 0.000
ASTEI11436-RA ASTEI11436 14448.6 146.5 39.5 212.8 6.6 8.5 6.1 7.1 0.000
ASTEI10564-RA ASTEI10564 6816.2 70.7 8.6 126.7 6.6 9.6 5.7 7.3 0.000
ASTEI05935-RA ASTEI05935 3161.9 34.6 11.3 45.8 6.5 8.1 6.1 6.9 0.000
ASTEI03729-RA ASTEI03729 775.8 8.9 1.7 10.5 6.5 8.8 6.2 7.2 0.000
ASTEI06461-RA ASTEI06461 192.0 2.9 0.3 2.3 6.1 9.3 6.4 7.2 0.000
ASTEI10652-RA ASTEI10652 912.0 11.0 1.6 16.0 6.4 9.2 5.8 7.1 0.000
ASTEI05824-RA ASTEI05824 21.6 1.3 0.0 0.3 4.1 10.4 6.3 6.9 0.003
ASTEI06639-RA ASTEI06639 172.8 2.2 0.2 4.1 6.3 9.9 5.4 7.2 0.001
ASTEI06641-RA ASTEI06641 849.7 11.4 1.5 11.8 6.2 9.2 6.2 7.2 0.000
ANSTF.9506.2 ASTEI10275 2778.6 41.0 7.2 55.1 6.1 8.6 5.7 6.8 0.000
ASTEI10275-RA ASTEI10275 3186.5 47.9 7.3 66.7 6.1 8.8 5.6 6.8 0.000
ASTEI10033-RA ASTEI10033 1429.0 31.3 0.8 28.7 5.5 10.8 5.6 7.3 0.000
Transcripts downregulated in midgut compared to other tissues
Transcript_ID Gene_ID FPKM Log2(fold change) Average p-value
MG MT Ov FB MG/MT MG/Ov MG/FB Average
ASTEI07530-RA ASTEI07530 2.6 23.5 57.3 550.5 -3.2 -4.5 -7.7 -5.1 0.000
ANSTF.6181.3 ASTEI06644 0.1 3.4 2.5 485.4 -4.9 -4.5 -12.1 -7.1 0.000
ANSTF.2700.1 ANSTF.2700 0.3 2.9 2.3 232.0 -3.2 -2.9 -9.5 -5.2 0.001
ANSTF.1341.1 ASTEI01412 1.8 40.3 29.7 166.1 -4.5 -4.1 -6.6 -5.1 0.000
ANSTF.6181.4 ASTEI06644 0.2 1.6 2.3 147.6 -3.3 -3.8 -9.8 -5.7 0.000
ANSTF.894.1 ASTEI00959 0.2 1.7 10.1 56.3 -2.9 -5.4 -7.9 -5.4 0.005
ANSTF.5191.1 ASTEI05569 0.3 4.8 51.1 5.5 -4.1 -7.5 -4.3 -5.3 0.000
ASTEI11306-RA ASTEI11306 0.8 21.0 20.3 50.0 -4.7 -4.7 -6.0 -5.1 0.000
ASTEI05326-RA ASTEI05326 0.3 3.7 10.3 46.1 -3.6 -5.0 -7.2 -5.3 0.000
ANSTF.187.1 ASTEI00214 0.1 2.8 7.0 25.8 -4.3 -5.6 -7.5 -5.8 0.000
ANSTF.5741.1 ANSTF.5741 0.1 1.3 4.2 24.1 -3.2 -4.9 -7.4 -5.2 0.002
ASTEI03946-RA ASTEI03946 0.2 1.8 21.3 3.2 -3.6 -7.1 -4.4 -5.0 0.000
ANSTF.2756.1 ASTEI02935 0.3 16.7 15.9 19.9 -5.9 -5.8 -6.2 -6.0 0.000
ANSTF.7908.4 ASTEI08579 0.1 1.5 19.7 1.4 -4.4 -8.1 -4.3 -5.6 0.022
ASTEI06406-RA ASTEI06406 0.2 15.0 19.0 2.4 -6.1 -6.4 -3.5 -5.3 0.002
ANSTF.9861.1 ANSTF.9861 0.2 2.9 18.7 5.2 -4.0 -6.7 -4.8 -5.2 0.000
ASTEI03548-RA ASTEI03548 0.1 0.9 15.2 2.2 -3.3 -7.3 -4.5 -5.0 0.009
ANSTF.4078.1 ASTEI04372 0.1 2.5 14.7 2.4 -4.6 -7.1 -4.5 -5.4 0.001
ASTEI07346-RA ASTEI07346 0.1 2.3 0.6 12.5 -5.0 -3.1 -7.5 -5.2 0.002
ANSTF.6500.1 ASTEI07015 0.1 9.0 12.3 0.8 -7.1 -7.5 -3.6 -6.1 0.006
ANSTF.6181.7 ASTEI06644 0.0 0.8 1.4 12.0 -4.0 -4.9 -7.9 -5.6 0.011
ASTEI07935-RA ASTEI07935 0.1 1.6 11.6 1.5 -4.4 -7.2 -4.3 -5.3 0.016
ANSTF.5782.1 ASTEI06196 0.1 2.5 11.2 2.1 -5.3 -7.5 -5.1 -6.0 0.025
ANSTF.2977.1 ASTEI03182 0.1 6.4 1.9 4.5 -6.2 -4.4 -5.7 -5.4 0.000
ANSTF.5442.1 ASTEI05836 0.0 0.9 1.7 1.2 -5.6 -6.6 -6.0 -6.1 0.001
Cite This Work
To export a reference to this article please select a referencing stye below:
Related ServicesView all
DMCA / Removal Request
If you are the original writer of this dissertation and no longer wish to have your work published on the UKDiss.com website then please: