Evolução Experimental da Forma da Asa de Drosophila ...€¦ · Drosophila melanogaster...
Transcript of Evolução Experimental da Forma da Asa de Drosophila ...€¦ · Drosophila melanogaster...
Universidade Federal do Rio de Janeiro Instituto de Biologia Pós-Graduação em Biodiversidade e Biologia Evolutiva
Evolução Experimental da Forma da Asa de Drosophila melanogaster
Integração morfológica, plasticidade fenotípica, bases celulares e expressão gênica
Daniel de Mattos Corrêa
Orientação: Blanche Christine Pires de Bitner-Mathé Leal
Rio de Janeiro 2015
I
Evolução Experimental da Forma da Asa de Drosophila melanogaster
Daniel de Mattos Corrêa
Tese de Doutorado apresentada ao Programa de Pós-graduação em Biodiversidade e Biologia Evolutiva, Instituto de Biologia, Universidade Federal do Rio de Janeiro, como parte dos requisitos necessários à obtenção do título de Doutor em Ciências Biológicas (Biodiversidade e Biologia Evolutiva).
Orientação: Blanche Christine Pires de Bitner-Mathé Leal
Rio de Janeiro
Abril de 2015
II
FICHA CATALOGRÁFICA:
Corrêa, Daniel de Mattos Evolução Experimental da Forma da Asa de Drosophila melanogaster/Daniel de Mattos Corrêa. Rio de Janeiro: UFRJ / IB, 2015. XV, 140 p. Orientadora: Blanche Christine Pires de Bitner-Mathé Leal Tese (Doutorado) – Universidade Federal do Rio de Janeiro, Insituto de Biologia, Programa de Pós-graduação em Biodiversidade e Biologia Evolutiva, 2015. Referências bibliográficas: f. 128-140 1. Seleção artificial. 2. Expressão Gênica 3. Drosophila melanogaster – Tese. I. Bitner-Mathé, B.C. II. Universidade Federal do Rio de Janeiro, Instituto de Biologia. III. Evolução Experimental da Forma da Asa de Drosophila melanogaster.
III
Evolução Experimental da Forma da Asa de Drosophila
Daniel de Mattos Corrêa
Orientação: Blanche Christine Pires de Bitner-Mathé Leal
Tese de Doutorado apresentada ao Programa de Pós-Graduação em Biodiversidade e Biologia Evolutiva, Instituto de Biologia, Universidade Federal do Rio de Janeiro, como parte dos requisitos necessários à obtenção do título de Doutor em Ciências Biológicas (Biodiversidade e Biologia Evolutiva).
Data: 28 de Abril de 2015 Aprovada por:
Prof. Dr. Antônio Solé-Cava (Departamento de Genética, IB - UFRJ)
Profª. Dra. Leila Maria Pessoa (Departamento de Zoologia, IB – UFRJ)
Profª. Dra. Helena Marcolla Araujo (Departamento de Histologia e Embriologia - UFRJ)
Prof. Dr. Paulo Cesar de Paiva (Departamento de Zoologia, IB – UFRJ)
Prof. Dr. Régis Lopes Corrêa (Departamento de Genética. IB - UFRJ)
Profª. Dra. Cássia Mônica Sakuragui (Instituto de Biologia – UFRJ. Suplente Interno)
IV
Profª. Dra. Katia Carneiro de Paula (Instituto de Ciências Biomédicas – UFRJ. Suplente Externo)
AGRADECIMENTOS
À Blanche, minha querida orientadora, por mais de uma década de ensinamentos, muito além
das técnicas em drosófila ou genética, mas de como ser professor, de como ter paixão por
ciência e de como viver uma vida laboratorial agradável!
Aos meus pais, Gerson e Izabel, pelo apoio irrestrito aos meus projetos pessoais e de trabalho.
Sem a semente e a rega de vocês, nada teria acontecido. Obrigado por acreditarem em mim
e por me darem uma rede de suporte sem a qual tudo seria muito mais difícil! À Lia e ao Celso
por expandirem minha definição de família e por trilharem ao meu lado essa jornada de todos
nós.
Ao André, pelo carinho, pelo companheirismo, pelas alegrias e por compreender e me ajudar
a atravessar de forma mais leve pelos estresses que vieram no último ano da elaboração dessa
tese. Obrigado!
Ao meu irmão Eduardo, pelas cervejas, pelos papos, pelas discussões políticas e pelo carinho.
À minha querida Raquel, pelas overdoses de cafeína e pelos litros de cerveja, compartilhando
medos, alegrias e as angústias dessa vida, sobretudo da vida acadêmica, mas também do resto
dela!
À Mariana, Renata e Clarice por aguentarem as incontáveis horas de prosa analítica,
escrutinando cada detalhe da vida. Por suportarem minhas paixões por discussões, minhas
relações intensas de prazer e angústia com os mínimos detalhes dessa empreitada de se saber
o que se puder saber.
À Fe Nunes por ser esquisita que nem eu e entender junto comigo o que parece ser uma
linguagem só nossa.
V
À Fe Braga, por todo o carinho, pelos papos e por ter dado vida ao Pedro e ao Luquinha; essas
duas crianças incríveis!
À Pat, Lucia, Dri, Aline e todos os amigos que têm feito esses 12 anos desde que entramos na
Biologia tão alegres e tão divertidos!
Aos amigos da vida, da praça e dos papos, Thaís, Tadeu, Mari Santana e todos os demais que
contribuem para dar sentido ao desenrolar dos dias.
À família Paiva por me ensinar que amigos são família também e que há poucos prazeres
maiores do que uma mesa com comida, vinho e amigos ao redor.
Ao meu fígado, fiel escudeiro, por aguentar firme, forte e jovial apesar de eu maltratá-lo
algumas vezes.
À todos os professores da banca, não somente pelas discussões que serão levantadas durante
a sabatina dessa tese, mas por terem contribuído com a minha formação durante esses 12
anos de Biologia e por terem sido inspiração para a construção do professor que eu gostaria
de um dia vir a ser.
Ao Programa de Pós-graduação em Biodiversidade e Biologia Evolutiva da UFRJ da qual tive o
prazer de fazer parte da primeira turma de ingressantes. Vida longa e próspera ao programa.
Às drosófilas, sem as quais esse trabalho jamais seria possível.
Este trabalho foi desenvolvido com recursos da Coordenação de Aperfeiçoamento Pessoal de
Ensino Superior (CAPES), incluindo a Bolsa de Doutorado pela qual sou grato.
VI
RESUMO
Fenótipos complexos são aqueles cuja variação não pode ser explicada por relações
mendelianas simples e suas bases genéticas apresentam alta pleiotropia com grande parte da
variação residindo nas interações entre os elementos ontogenéticos. Forma e tamanho de
estruturas biológicas, assim como o autismo e o câncer, são exemplos de traços complexos. O
ambiente também interfere no desenvolvimento desse tipo de caráter, processo conhecido
por plasticidade fenotípica. A identificação dos elementos que contribuem para a
determinação fenotípica torna-se igualmente complexa. A asa de Drosophila é um modelo
amplamente estudado que oferece uma grande cobertura de conhecimentos sobre sua
morfologia, desenvolvimento, genética e evolução, tornando-se, portanto, um alvo ideal para
um melhor entendimento da variação quantitativa de fenótipos complexos. Neste trabalho,
utilizamos linhagens de D. melanogaster selecionadas para formas extremas da asa a fim de
investigarmos diversos aspectos envolvidos na variação quantitativa de forma e tamanho.
Primeiramente, apresentamos um estudo descritivo da morfologia da asa, assim como
estimativas de integração genética e fenotípica de traços correlacionados com o intuito de
testarmos a previsibilidade de trajetórias evolutivas através de matrizes de correlação
genética da população inicial, antes do processo de seleção artificial ocorrer. No segundo
capítulo apresentamos um estudo sobre a plasticidade fenotípica relacionada à temperatura,
descrevendo as normas de reação em 10 temperaturas, cobrindo quase todo o espectro
possível para a espécie. Mostramos também que a média fenotípica influencia a norma de
reação, sugerindo que parte dos genes devem ser compartilhados entre as duas, cenário
chamado de sensitividade alélica. O terceiro capítulo aborda as bases celulares, mostrando
que o número de células da asa, mas não o tamanho delas, desempenham um importante
papel na resposta de forma à seleção, com um aumento no número de células em linhagens
com asa alongada. Por fim, investigamos os genes candidatos ao controle da forma da asa
através de um ensaio de microarranjo, real time qPCR e validações morfológicas pelo
silenciamento gênico tecido-específico mediado por RNAi. A natureza poligênica fica evidente,
sugerindo um modelo genético infinitesimal com múltiplos genes de efeitos pequenos. Ainda
assim, o desenho experimental permitiu a identificação de 11 genes fortemente candidatos a
terem uma grande contribuição relativa.
VII
ABSTRACT
Complex phenotypic traits are those whose variation cannot be explained by simple
mendelian inheritance and its genetic bases exhibit high pleiotropy, with a great amount of
variation residing on the interaction of ontogenetic elements. Shape and size of biological
structures, as well as autism and cancer, are classic examples of complex traits. Environment
also plays an important role in these traits through a process called phenotypic plasticity.
Identification of contributory elements to phenotypic determination is equally complex. The
Drosophila wing is a widely studied model, offering a broad coverture of published knowledge
on its morphology, development, genetics and evolution, thus making it an ideal target for a
better comprehension of the quantitative variation of complex traits. In the present work, we
use D. melanogaster strains artificially selected for extreme wing shapes in order to
investigate multiple layers involved in the quantitative variation of shape and size. First, we
present a descriptive study on wing morphology, as well as genetic and phenotypic integration
estimates of correlated traits. We tested the predictability of evolutionary trajectories
imposed by genetic correlation matrices of the initial population, prior to selection. On the
second chapter, we present a study on thermal-related plasticity that describes the reaction
norms on a gradient of 10 developmental temperatures, covering almost the entire spectrum
of livable temperatures for the species. We show that phenotypic mean influences its reaction
norm, suggesting that some genes might be shared, favoring a scenario called allelic
sensitivity. The third chapter addresses the cellular bases, showing that cell number, but not
their size, plays an important role on shape response to selection. Finally, we investigated
candidate genes for controlling wing shape through a microarray assay, real time qPCR and
morphological validation through tissue-specific gene silencing mediated by RNAi. The
polygenic nature becomes evident, suggesting an infinitesimal genetic model, with multiple
genes of small effects. Still, experimental design allowed the identification of 11 strong
candidate genes with a high relative contribution.
VIII
LISTA DE ILUSTRAÇÕES
INTRODUÇÃO
Figura 1: “INFINITAS FORMAS DE GRANDE BELEZA” ....................................................... 17
Figura 2: ASAS FOSSILIZADAS ........................................................................................... 21
Figura 3: DISCOS IMAGINAIS DE DROSOPHILA ................................................................. 23
Figura 4: DISCO IMAGINAL DE ASA ................................................................................... 24
Figura 5: EVERSÃO DO DISCO IMAGINAL .......................................................................... 24
Figura 6: EXPANSÃO DA LÂMINA DA ASA ......................................................................... 25
Figura 7: MODIFICAÇÕES NO FORMATO CELULAR ........................................................... 25
Figura 8: MODULARIZAÇÃO NO DESENVOLVIMENTO DO DISCO IMAGINAL ................... 27
Figura 9: ASAS DE SELEÇÃO COM FORMAS EXTREMAS .................................................... 30
Figura 10: MEDIDAS DE FORMA E TAMANHO DA ASA EM DROSOPHILA ......................... 30
CAPÍTULO I: MORFOLOGIA
Figura 1: THE ELLIPSE METHOD ......................................................................................... 37
Figura 2: HISTOGRAMS OF SIZE AND SHAPE ..................................................................... 41
Figura 3: ILLUSTRATIVE WINGS FROM L AND R STRAINS .................................................. 42
Figura 4: HISTOGRAMS OF WING LENGTH AND WIDTH ................................................... 43
Figura 5: PLOT OF WING LANDMARKS ............................................................................. 45
Figura 6: GRAPHICAL REPRESENTATION OF GENETIC AND PHENOTYPIC CORRELATIONS47
CAPÍTULO II: PLASTICIDADE FENOTÍPICA
Figura 1: HISTOGRAMS OF WING SHAPE AND SIZE .......................................................... 58
Figura 2: TEMPERURE REACTION NORMS FOR WING SHAPE AND SIZE ........................... 59
IX
Figura 3: RELATIONSHIP OF b_WSH AND PHENOTYPIC MEAN .......................................... 61
Figura 4: PROPORTIONAL VARIATION OF WW AND WL ..................................................... 62
CAPÍTULO III: BASES CELULARES
Figura 1: MORPHOMETRIC METHODS FOR INTERVEIN VARIATION ................................ 84
Figura 2: MEAN VALUES AND STANDARD ERRORS OF WING TRAITS ............................... 87
Figura 3: MEAN VALUES AND STANDARD ERROS PER INTERVEIN REGION ...................... 90
Figura 4: QUADRATIC CURVES FITTED TO AVERAGE CELL AREA ....................................... 91
Figura S1: ELLIPSE ADJUSTMENT ..................................................................................... 100
CAPÍTULO IV: GENES CANDIDATOS
Figura 1: EXPRESSION PATTERN OF GAL4 ....................................................................... 109
Figura 2: GAL4 / UAS SILENCING SYSTEM ....................................................................... 109
Figura 3: PIE CHART SUMMARIZING MICROARRAY ASSAY RESULTS .............................. 111
Figura 4: BOXPLOT OF WSH VARIATION ........................................................................... 112
Figura 5: HEATMAPS OF EXPRESSION RATIOS ................................................................ 113
Figura 6: VISUAL REPRESENTATION OF EXPRESSION LEVELS.......................................... 114
Figura 7: EFFECTS OF RNAi-MEDIATED SILENCING OF dp ............................................... 117
Figura 8: MALFORMATIONS ON THE WING OF D. melanogaster ................................... 117
CONCLUSÕES E PERSPECTIVAS FUTURAS
Figura 1: PADRÕES DE INTERFERÊNCIA NA ASA DE DROSOPHILA .................................. 126
X
LISTA DE TABELAS
CAPÍTULO I: MORFOLOGIA
Tabela 1: PHENOTYPIC AND GENETIC CORRELATION MATRICES FOR PRE-SELECTION .... 40
Tabela 2: ANCOVA FOR WSH AND WSI. ............................................................................... 42
Tabela 3: MANCOVA FOR ANGULAR MOVIMENTS OF LANDMARKS ................................ 44
Tabela 4: PHENOTYPIC CORRELATION FOR POST-SELECTION .......................................... 46
CAPÍTULO II: PLASTICIDADE FENOTÍPICA
Tabela 1: ANOVA FOR WSH AND WSI. ................................................................................. 58
Tabela 2: ANOVA FOR REACTION NORM PARAMETERS ................................................... 60
Tabela S1: MEAN AND STANDARD ERROR OF EXPERIMENTAL GROUPS. ......................... 71
Tabela S2: LINEAR AND QUADRATIC COEFFICIENTS ......................................................... 74
Tabela S3: PAIRWISE COMPARISON FOR THE SEL EFFECT OF ANOVA (TABLE2). ............. 76
CAPÍTULO III: BASES CELULARES
Tabela 1: ANOVA OF MORPHOLOGICAL AND CELLULAR TRAITS. ..................................... 89
Tabela S1: MEAN AND STANDARD ERROR FOR SELECTION STRAINS. ............................ 101
Tabela S2: MEAN AND STANDARD ERROR FOR BASELINE POPULATION... ..................... 102
Tabela S3: QUADRATIC REGRESSIONS OF AVERAGE CELL AREA. .................................... 103
CAPÍTULO IV: GENES CANDIDATOS
Tabela 1: F-TESTS FOR THE EFFECTS OS SEL . .................................................................. 112
Tabela 2: F-TESTS FOR RNAi STRAINS ............................................................................. 116
XI
LISTA DE ABREVIATURAS E SIGLAS
1C, 2C, 5C e 6C linhagens controle do processo de seleção
1L, 2L, 5L e 6L linhagens selecionadas para forma alongada da asa
1R, 2R, 5R e 6R linhagens selecionadas para forma alongada da asa
a raio maior da elipse
ANCOVA análise de covariância
ANOVA análise de variância
APF após formação da pupa
b raio menor da elipse
BIOREP réplicas biológicas do processo de seleção
b _WSH parâmetro da inclinação da reta que descreve a NR de WSH
BED biologia evolutiva do desenvolvimento
cDNA ácido desoxirribonucleico complementar
c _W_ parâmetro da curvatura da NR do traço analisado
CA área da célula da asa
CAaverage área média das células da asa
CD diâmetro celular
CN número de células
CNtotal número de células totais da asa
CNWW número de células na largura da asa
CNWL número de células no comprimento da asa
CNWW número de células no comprimento da asa
CNWW número de células no comprimento da asa
DE genes diferencialmente expressos
DNA ácido desoxirribonucleico
DV eixo do desenvolvimento dorsoventral
DT temperatura de desenvolvimento
EP erro padrão
G64 geração 64 das linhagens de seleção
G100 geração 100 das linhagens de seleção
XII
G1_TD16 1ª geração nascida em laboratório mantida a 16°C
h2 herdabilidade
IVR região interveia
IVRA-E região interveia especificada
L linhagens de seleção para asas longas
MANCOVA análise de variância multivariada
NER razão da expressão normalizada
NR norma de reação
PD eixo do desenvolvimento proximodistal
PF plasticidade fenotípica
qPCR reação em cadeia da polimerase quantificada em tempo real
QTL lócus de traço quantitativo
R linhagens de seleção para asa arredondada
rG correlação genética
RN norma de reação
RNAi ácido ribonucleico de interferência
RT transcrição reversa
SD dimorfismo sexual
SEL efeitos da seleção
TD temperatura de desenvolvimento
WA asa da área
WSH índice de forma da asa de Drosophila
WSHmean média de WSH
WSI índice de tamanho da asa de Drosophila
WL índice de comprimento da asa de Drosophila
WW índice de largura da asa de Drosophila
WWpredicted medida esperada de WW
ângulo que estima o posicionamento de veias
XIII
SUMÁRIO
INTRODUÇÃO ......................................................................................................................... 16
VARIAÇÃO FENOTÍPICA E EVOLUÇÃO BIOLÓGICA ............................................................ 16
A ASA DE DROSOPHILA ..................................................................................................... 19
1 ASPECTOS MACROEVOLUTIVOS .......................................................................... 19
2 A ASA ENQUANTO MODELO ............................................................................... 21
3 DESENVOLVIMENTO ............................................................................................ 22
4 MODULARIZAÇÃO E GENES CANDIDATOS AO CONTROLE DA VARIAÇÃO
QUANTITATIVA ....................................................................................................................... 25
5 PLASTICIDADE FENOTÍPICA ................................................................................. 28
6 SELEÇÃO ARTIFICIAL DA ASA DE D. MELANOGASTER ......................................... 29
OBJETIVOS E ORGANIZAÇÃO DA TESE .................................................................................. 31
CAPÍTULO I: MORFOLOGIA .................................................................................................... 32
Phenotypic and Genetic Integration in The Drosophila Wing
ABSTRACT .......................................................................................................................... 33
INTRODUCTION ................................................................................................................. 34
MATERIALS AND METHODS .............................................................................................. 35
ARTIFICIAL SELECTION PROGRAM .............................................................................. 35
PRE-SELECTION GENERATION HERITABILITY, PHENOTYPIC AND GENETIC
CORRELATION ESTIMATES .......................................................................................... 36
POST SELECTION GENERATION: MORPHOLOGICAL INTEGRATED RESPONSE AND
REACTION NORMS ...................................................................................................... 36
WING MORPHOLOGY MEASUREMENTS .................................................................... 36
STATISTICS .................................................................................................................. 37
RESULTS ............................................................................................................................. 38
DISCUSSION ....................................................................................................................... 47
REFERENCES ...................................................................................................................... 49
XIV
CAPÍTULO II: PLASTICIDADE FENOTÍPICA ............................................................................. 52
Thermal Plasticity Evolution in Strains of Drosophila melanogaster Selected for Divergent Wing Shape
ABSTRACT .......................................................................................................................... 53 INTRODUÇÃO ..................................................................................................................... 55
MATERIAL AND METHODS ................................................................................................ 55
STRAINS ...................................................................................................................... 55
REACTION NORMS ...................................................................................................... 55
WING SHAPE AND SIZE DESCRIPTORS ........................................................................ 56
RESULTS ............................................................................................................................. 57
DISCUSSION ....................................................................................................................... 63
ACKNOWLEDGEMENTS ..................................................................................................... 66
REFERENCES ...................................................................................................................... 66
SUPPORTING INFORMATION ............................................................................................. 71
CAPÍTULO III: BASES CELULARES ........................................................................................... 77
Cellular Basis of Morphological Variation and Temperature-related Plasticity in Drosophila melanogaster Strains with Divergent Wing Shapes
ABSTRACT .......................................................................................................................... 78
INTRODUÇÃO ..................................................................................................................... 81
MATERIALS AND METHODS .............................................................................................. 81
FLIES (STRAINS WITH ARTIFICIALLY SELECTED WING SHAPES) .................................. 81
EXPERIMENTAL DESIGN .............................................................................................. 82
WING MORPHOMETRICS ........................................................................................... 82
CELL SIZE (AREA) AND CELL NUMBER ........................................................................ 84
STATISTICAL ANALYSES AND DISTRIBUTION OF CELL AREA ACROSS INTERVEIN
REGIONS ................................................................................................................... 85
RESULTS ............................................................................................................................. 85
DISCUSSION ....................................................................................................................... 92
ACKNOWLEDGMENTS ....................................................................................................... 96
REFERENCES ...................................................................................................................... 96
SUPPORTING INFORMATION ........................................................................................... 100
XV
CAPÍTULO IV: GENES CANDIDATOS ..................................................................................... 104
Gene Expression Profile and Candidate Genes in Strains of Drosophila melanogaster Selected for Divergent Wing Shapes
ABSTRACT ........................................................................................................................ 105
INTRODUCTION ............................................................................................................... 106
MATERIAL AND METHODS .............................................................................................. 107
MICROARRAY ON 1L AND 1R STRAINS ..................................................................... 107
BIOLOGICAL REPLICATE STRAINS ............................................................................. 107
WING IMAGINAL DISC RNA EXTRACTION AND REAL TIME qPCR ............................. 107
MORPHOLOGICAL VALIDATION ............................................................................... 108
RESULTS ........................................................................................................................... 110
EXPRESSION PATTERNS ............................................................................................ 110
MORPHOLOGICAL VALIDATION ............................................................................... 115
DISCUSSION ..................................................................................................................... 118
ACKNOWLEDGMENT ....................................................................................................... 120
REFERENCES .................................................................................................................... 121
CONCLUSÕES E PERSPECTIVAS FUTURAS ........................................................................... 123
REFERÊNCIAS BIBLIOGRÁFICAS ........................................................................................... 125
I N T R O D U Ç Ã O
16
VARIAÇÃO FENOTÍPICA E EVOLUÇÃO BIOLÓGICA
Este planeta é uma vastidão de “infinitas formas de grande beleza” (figura 1) como
rapidamente percebeu Charles Darwin em sua longa jornada pelas florestas tropicais da
América do Sul. Sob a regência das leis básicas da física, fervilha aqui uma miríade de formas,
cores, sons, texturas e estratégias, “enquanto este planeta segue sua órbita, seguindo a lei
fixa da gravidade” (DARWIN, 1859). Essa vasta rede interconectada pela história evolutiva
compartilhada e pela ecologia se mostra imponente, compelindo muitos de nós a buscarmos
uma melhor compreensão sobre seus meandros e mecanismos.
Desde a formulação feita por Darwin e Wallace (WALLACE, 1858), a teoria da evolução
vem sofrendo transformações, corroborando os nortes básicos originais, mas agregando
novas informações e perspectivas para uma melhor compreensão do processo evolutivo. São
as “infinitas” variações entre indivíduos o substrato para a seleção natural, que irá atuar sobre
os componentes herdáveis que promovem as diferenças e não sobre aqueles que não podem
ser reproduzidos nas gerações subsequentes (RIDLEY, 2004). Por essa razão, os modelos
clássicos da biologia evolutiva confiaram, predominantemente, a variação fenotípica às
variações alélicas, em especial àquelas com efeitos aditivos (DAWKINS, 1976). Na realidade o
que se viu foi um cenário bem mais complexo do que o esperado com diversos fenótipos
observáveis não sendo facilmente explicados por diferenças em alelos. A dificuldade em se
explicar esses fenótipos ou associá-los à variantes genéticas, tem aberto novas frentes de
pesquisa. Essa dificuldade tem inclusive apontado para teorizações que recapitulam processos
antes descartados e apresentam perspectivas neo-lamarckistas, em especial para explicar a
evolução humana e seus aspectos culturais (JABLONKA; LAMB, 2005).
Esse tipo de variação é predominante na natureza e esses traços são considerados
fenótipos complexos, ou seja, aqueles cuja variação fenotípica não pode ser explicada por
relações mendelianas simples, demandando uma visão sistêmica para o seu estudo (SHAO et
al., 2008). Em geral são determinados por um grande número de genes com efeitos
pleiotrópicos e uma grande parte da variação reside nas interações entre genes, dificultando
o estabelecimento de relações causais entre variantes genéticas e as variações fenotípicas. O
tamanho e forma de estruturas, respostas comportamentais, doenças como o câncer, autismo
e esquizofrenia, são exemplos de caracteres complexos (AYROLES et al., 2009; PEARLSON;
FOLLEY, 2008; SHINJI IJICHI, NAOMI IJICHI, YUKINA IJICHI, 2011).
I N T R O D U Ç Ã O
17
Figura 1. “Infinitas formas de grande beleza”. Fotografias: Daniel Mattos.
I N T R O D U Ç Ã O
18
No que se refere aos elementos envolvidos nessa classe de determinação fenotípica,
variações nas sequências de nucleotídeos em regiões codificantes e em regiões promotoras
ou acentuadoras podem gerar fenótipos diferenciados. Por outro lado, fatores epigenéticos
como mudanças de estados da cromatina, padrões de metilação e acetilação de regiões
codificantes e reguladoras também promovem variação em fenótipos complexos (CARROLL,
2005; FALCONER; MACKAY, 1996; PIGLIUCCI, 2001). A metilação de regiões do DNA é um dos
mais estudados mecanismos para o controle dos níveis de expressão gênica e estudos têm
indicado que o padrão de metilação pode ser transmitido às gerações subsequentes e,
portanto, ter impactos nas trajetórias evolutivas dos caracteres. Por exemplo, Waterland e
Jirtle (2003) demonstraram que diferenças nas concentrações do radical metil na dieta de
camundongos alteram radicalmente a coloração dos pelos e a predisposição à obesidade, com
impactos fenotípicos nas gerações subsequentes, mesmo estas não tendo sido expostas ao
estímulo nutricional.
Finalmente, condições ambientais também são capazes de alterar trajetórias do
desenvolvimento, gerando fenótipos variados, processo chamado de plasticidade fenotípica
(PIGLIUCCI, 2001). Essa propriedade vem sendo estudada há mais de um século quando
Woltereck cunhou o termo norma de reação para descrever a função matemática das
variações morfológicas observadas em Dapnhia sp. em resposta à presença de predadores
(SARKAR, 1999; WOLTERECK, 1909). Ao longo do século XX, diversas pesquisas mostraram
que a plasticidade é um processo comum a praticamente todos os traços fenotípicos (WEST-
EBERHARD, 2003). A busca pelos elementos que promovem as “infinitas formas de grande
beleza” continua mais acirrada do que nunca.
Com esse intuito, a biologia evolutiva do desenvolvimento (BED) tem reunido os
conhecimentos de áreas que permaneceram por muito tempo separadas. Diversos estudos
têm relacionado a variação fenotípica aos reguladores da expressão gênica, em oposição à
visão clássica de que a variabilidade estaria contida majoritariamente em regiões codificantes
dos genes. A BED tem encontrado evidências de grandes mudanças evolutivas relacionadas às
sequências reguladoras que funcionariam como interruptores da expressão gênica em função
da localização e do tempo de desenvolvimento. Quanto ao plano básico e forma dos
organismos, há evidências de que poucos genes, com elevada conservação entre espécies
filogeneticamente distantes, seriam responsáveis pela determinação desses fenótipos, sendo
I N T R O D U Ç Ã O
19
reutilizados em diversas estruturas durante o desenvolvimento. Mudanças nas regiões
reguladoras desses genes seriam responsáveis por modificações fenotípicas de grande
importância evolutiva, possibilitando a enorme diversidade do grupo (CARROLL, 2006). A
expressão gênica pode ser regulada por diversos fatores como mudanças nas sequências
reguladoras, estados metilacionais, disponibilidade de fatores de transcrição, além de
reguladores pós-transcricionais como, por exemplo, o RNA de interferência (CARROLL, 2005;
CHENG et al., 2005; KING; WILSON, 1975). Entretanto, a hipótese de que modificações das
taxas de expressão gênica podem explicar parte das variações entre populações e espécies é
difícil de ser testada (LI; SAUNDERS, 2005) e há ainda grande demanda de estudos de
expressão gênica entre populações e espécies que possam justificar parte das divergências
morfológicas observadas, em especial em caracteres complexos.
Em sistemas como o desenvolvimento, a compreensão dos inúmeros fatores que
participam do processo é extremamente difícil. Contudo, é possível a identificação de
elementos cuja variação é capaz de alterar trajetórias do desenvolvimento, levando à
produção de fenótipos muito distintos. Nesse sentido, a asa de Drosophila oferece um modelo
único para o entendimento da evolução de caracteres complexos devido à enorme quantidade
de informação disponível e a possibilidade de se formular desenhos experimentais e analisá-
los sob uma gigantesca base de dados.
A ASA DE DROSOPHILA
1 ASPECTOS MACROEVOLUTIVOS
Drosófilas são insetos da ordem Diptera e possuem somente um par de asas. O
segundo par de asas foi perdido durante a evolução do grupo, tendo se reduzido a um par de
halteres, estruturas vestigiais que auxiliam o equilíbrio durante o voo. Por se tratarem de
estruturas muito finas e delicadas, as asas de dípteros não são facilmente encontradas no
registro fóssil, mas os mais antigos conhecidos datam do Permiano superior, há 250ma
(YEATES; WIEGMANN, 1999). Nesse período, a Terra presenciava também a diversificação dos
amniotas que, mais tarde, dariam origem aos mamíferos e répteis. Toda a massa continental
estava agrupada no supercontinente Pangeia e um clima seco e desértico predominava no
I N T R O D U Ç Ã O
20
ambiente terrestre. A radiação do gênero Drosophila é datada em 50ma, no início do Eoceno,
que viu o surgimento dos dois subgêneros: Drosophila e Sophophora na região tropical do
Velho Mundo. A espécie Drosophila (Sophophora) melanogaster é ainda mais recente,
surgindo há aproximadamente 2,3ma, no início do Pleistoceno, também no Velho Mundo
(RUSSO; TAKEZAKI; NEI, 1995). Acredita-se que a divisão entre grupos do Velho Mundo e
Neotropicais (há aproximadamente 40ma, no fim do Eoceno) tenha ocorrido pela invasão de
espécies pelo estreito de Bering e não pela separação entre América do Sul e África que já
estava ocorrendo no Cretáceo Superior (POWELL, 1997). Ainda não foi encontrado registro
fóssil de indivíduos pertences à espécie D. melanogaster, embora fósseis de grupos próximos
já tenham sido identificados. A figura 2 apresenta um díptero preservado em âmbar (resina
vegetal) datado em 42ma, no Eoceno, onde é possível verificar que a padronização de veias
das asas já era muito similar à encontrada em grupos atuais. Ainda hoje, há uma grande
conservação nos padrões de venação entre espécies do grupo melanogaster (CORRÊA, 2009)
e mesmo espécies que divergiram há milhões de anos apresentam variação quantitativa, e
não qualitativa, na forma da asa (OBBARD et al., 2012).
I N T R O D U Ç Ã O
21
2 A ASA ENQUANTO MODELO
A asa de Drosophila é um excelente modelo para estudos da evolução de traços
complexos já que é tradicionalmente utilizada em estudos de Biologia Evolutiva
(GIDASZEWSKI; BAYLAC; KLINGENBERG, 2009; KATAYAMA et al., 2014; POWELL, 1997;
TSUJINO; TAKAHASHI, 2014; YEH; TRUE, 2014) e Biologia do Desenvolvimento (AVANESOV et
al., 2012; BLAIR, 2007; BUCHMANN; ALBER; ZARTMAN, 2014; DE CELIS, 2003; LECUIT,
THOMAS; LE GOFF, 2007; NETO-SILVA; WELLS; JOHNSTON, 2009; STRIGINI; COHEN, 1999).
Soma-se a isso a vantagem dos genomas de diversas espécies desse gênero estarem
sequenciados (CONSORTIUM et al., 2007). Além disso, a similaridade na variação quantitativa
Figura 2. Fotomicrografia de Pareuthychaeta em âmbar do Eoceno. (A) P. electrica em âmbar Báltico e (B) P.
calpinei em âmbar Báltico. (C) Asa atual de D. melanogaster de linhagem controle do processo seletivo descrito
na seção Seleção Artificial da Asa de D. melanogaster desta tese. Figura modificada de GRIMALDI e SINGH (2012).
I N T R O D U Ç Ã O
22
da forma da asa entre populações de D. melanogaster e entre espécies filogeneticamente
distantes oferece uma oportunidade ímpar já que, possivelmente, os elementos envolvidos
na variação populacional podem ser extrapolados para a variação interespecífica (ALEXIS;
ISAAC; DAVID, 2015) .
3 DESENVOLVIMENTO
Por ser um modelo amplamente utilizado, o desenvolvimento da asa de Drosophila é
muito bem descrito. Os apêndices de Drosophila são formados a partir de estruturas
precursoras chamadas discos imaginais (figura 3; MORATA, 2001). O disco imaginal de asa se
forma a partir de um grupo de células precursoras que invaginam do ectoderma embrionário
(BATE; ARIAS, 1991). Os processos morfogenéticos que ocorrem no disco têm impacto direto
no órgão adulto (DOLEZAL et al., 2010). Portanto são nas mudanças das trajetórias de
desenvolvimento do disco imaginal e da fase pupal que residem grande parte das variações
quantitativas de tamanho e forma da asa.
O disco imaginal de larvas de 3° estadio é composto de uma monocamada com 20 a
50 células de tecido epitelial e sua parte mais central, de formato circular, irá, durante a
metamorfose, se diferenciar em uma coluna tridimensional de células que dará origem à asa
e à parte dorsal do tórax. Nesse momento, já há uma divisão clara de compartimentos
definidos pelos eixos dorsoventral (DV) e anteroposterior (AP), com a participação de
morfógenos clássicos como Wingless (Wg; especificação de DV) e Decapentaplegic (Dpp;
especificação de AP). Os genes scalloped e vestigial (vg) também participam da especificação
desses eixos (BRAY, 1999). Nessas etapas, genes como Engrailed, Hedehog, knot, EGF-R,
Notch, Hairless, entre outros, estão especificando os territórios de veias e diversos métodos
descritores de forma da asa são baseados na localização das veias, logo esses eventos estão
diretamente ligados às variações de forma (BIER, 2000; CROZATIER; GLISE; VINCENT, 2004;
CROZATIER et al., 2003; JOHANNES; PREISS, 2002). Em estágios mais tardios, genes da via de
Bmp interagem com integrinas e também alteram a posição de veias (ARAUJO, 2003). Ao final
do 3° estadio, o disco já é uma estrutura organizada que apresenta entre 30.000 e 50.000
células. Já na fase pupal, ocorre a eversão da parte mais central do disco, formando uma
bicamada celular, já com os compartimentos ventral e dorsal justapostos. A figura 4 ilustra e
I N T R O D U Ç Ã O
23
resume os processos envolvidos no desenvolvimento inicial do disco (ALEXIS; ISAAC; DAVID,
2015). Há então um alongamento pronunciado do tecido que envolve eventos mitóticos,
rearranjos celulares e modificações no formato celular (KANCA et al., 2014; TAYLOR, J.; ADLER,
2008). Por fim, a porção mais proximal da asa, próxima à articulação, se contrai, promovendo
a distensão do tecido ao longo do eixo proximodistal ao mesmo tempo em que o eixo
anteroposterior é comprimido (figura 5), com modificações na forma da asa (figura 6; AIGOUY
et al., 2010). Esse processo é mediado por divisões e rearranjos celulares que interferem na
forma da asa, enquanto as células se tornam mais hexagonais (figura 7; SUGIMURA; ISHIHARA,
2013).
Figura 3. Discos imaginais precursores dos apêndices do inseto adulto. Figura
adaptada de V. Hartenstein (http://flybase.bio.indiana.edu).
I N T R O D U Ç Ã O
24
Figura 4. Disco imaginal de asa. (a) disco em larvas de 3° estadio; compartimentos anterior e posterior
definidos pelo gradiente de Dpp e compartimentos dorsal e ventral definidos pelo gradiente de Wg. (b)
Disco em fase pupal, durante a metamorfose; eversão tridimensional do disco na região que irá formar a
asa e a porção dorsal do tórax. Note que em ambas as etapas, os territórios de veias já estão sendo
definidos. Figura modificada de ALEXIS; ISAAC; DAVID (2015).
Figura 5. Eversão do disco imaginal da asa. (a) Vista lateral do disco com os principais genes que determinam os
compartimentos dorsal e ventral entre parênteses. (b e c) Dobra do tecido do disco na região mais central de
aspecto circular, chamada de “bolsa do disco” (do inglês, disc pouch). (d) Pressão da hemolinfa contribui para a
eversão do disco. (e) Asa evertida. Nesse momento, ocorre grande extensão da área da asa. (f) No início do
processo, as células têm um formato colunar e ao fim adquirem formato cuboide, contribuindo para o aumento
de área. (g) Intercalação celular orientada contribui para o alongamento da lâmina da asa ao longo do eixo
proximodistal. Figura modificada de ALEXIS; ISAAC e DAVID (2015).
I N T R O D U Ç Ã O
25
4 MODULARIZAÇÃO E GENES CANDIDATOS AO CONTROLE DA VARIAÇÃO
QUANTITATIVA
O resumo dos principais eventos envolvidos na formação da asa evidencia uma
gigantesca complexidade no seu desenvolvimento com diversos processos capazes de
promover alterações quantitativas na estrutura adulta. O disco imaginal, assim como a asa
Figura 7. Modificações no formato celular em asas de pupa. As células são coloridas conforme o seu número de
lados (indicado em cima e à esquerda). Ao longo do desenvolvimento pupal, há um aumento na densidade de
células hexagonais (cinza). As modificações no formato celular contribuem para a determinação da morfologia
adulta. Figura modificada de (AIGOUY et al. (2010). APF refere-se a após formação da pupa.
Figura 6. (A-C) Expansão da lâmina da asa (vermelho) pela contração da articulação (azul). (D-F) Fluxo celular
durante a eversão. Migração celular ocorre a uma velocidade de 50m/h. Figura adaptada de AIGOUY et
al. (2010).
I N T R O D U Ç Ã O
26
adulta, apresenta indícios de modularização no seu desenvolvimento (BLAIR, 2003; GARCIA-
BELLIDO; RIPOLL; MORATA, 1973), apesar de a asa também responder de maneira bastante
integrada (KLINGENBERG, 2009). Diversos estudos mostraram ser possível a criação de
linhagens mutantes apresentando discos imaginais com taxas de crescimento heterogêneas
entre porções do disco, com grandes impactos na morfologia da asa adulta (ROGULJA;
RAUSKOLB; IRVINE, 2008; SCHWANK et al., 2011). Baena-lópez, Baonza e García-bellido (2005)
demonstraram que na parte mais central do disco, chamada de bolsa do disco (do inglês, disc
pouch), as divisões celulares são orientadas radialmente, com origem mais central, dividindo-
se em direção à todas as extremidades. Já nas regiões periféricas, as divisões são orientadas
tangencialmente à bolsa, evidenciando o caráter modular do desenvolvimento da estrutura
(figura 8). A orientação radial no centro parece ser coordenada pelo sistema Fat, Dachsous,
Four-jointed e Dachs que apresentam um gradiente também radial (AMBEGAONKAR et al.,
2012; HALE et al., 2015).
Diferenças não-sinônimas nas sequências dos genes envolvidos no desenvolvimento
podem gerar variações quantitativas, mas também variações nas sequências promotoras,
acentuadoras em cis ou trans, e em outros elementos capazes de alterar os padrões de
expressão gênica. Ao nível de tecidos, variações nos domínios de expressão, ainda que
pequenas, podem gerar variações morfológicas significativas. Variações ambientais também
interagem com o desenvolvimento e, por transdução, interferir nos padrões de expressão
(GILBERT; EPEL, 2008). Carreira e outros (2011) investigaram as bases genéticas da
morfogênese da asa, com ênfase em dimorfismo sexual e nos efeitos não-alométricos das
variações de forma e demonstraram que diversas mutações induzidas através da inserção de
elementos móveis causaram variações não-alométricas com efeitos compartimentalizados, ou
seja, restritos à uma região ou outra da asa.
Na seção anterior, diversos genes que participam da formação da asa foram descritos.
Fica evidente que há um grande número de genes participando do seu desenvolvimento. De
fato, 50% de todos os genes codificantes de proteínas apresentam variações nos níveis de
expressão durante a formação da asa, chegado a quase 80% o número de genes com
transcritos já detectados (O’KEEFE et al., 2012). Com relação às variações quantitativas da
forma, Matta e Bitner-Mathé (2010) identificaram diversos QTLs (loci de traço quantitativo,
do inglês quantitative trait loci) associados ao tamanho e forma da asa. Com o mesmo intuito,
I N T R O D U Ç Ã O
27
WEBER e outros (2008) identificaram centenas de genes candidatos ao controle da variação
de forma, evidenciando a grande dificuldade em se demonstrar quais seriam, de fato,
relevantes para a variação quantitativa. Fica evidente a necessidade de mais estudos sobre as
bases genéticas desse tipo de variação, com desenhos experimentais que permitam a
identificação de um número menor, porém mais relevante, de genes candidatos.
Faz-se necessária uma distinção. Os genes que participam diretamente no
desenvolvimento de estruturas não são obrigatoriamente os mesmos envolvidos na variação
quantitativa observada entre indivíduos ou populações. As técnicas para identificação de
genes participantes de vias do desenvolvimento (knockout, mutantes funcionais, etc.), em
geral, inferem na participação de um gene em particular pela malformação da estrutura
quando o gene alvo é perturbado. Porém, genes envolvidos com a variação quantitativa
podem atuar epistaticamente em relação às vias de desenvolvimento e, portanto, não serem
facilmente identificados pelas técnicas tradicionais.
Figura 8. Modularização no desenvolvimento do disco. (a) Esquema geral da
expressão periférica de Dachsous (roxo) e gradiente radial de Dachs (setas verdes).
Wg demarcando o eixo dorsoventral. (b) Proteína Dachsous corada em roxo na
periferia do disco. (c) Divisão celular orientada radialmente na porção central do
disco e tangencialmente à bolsa do disco na periferia. Figura adaptada de ALEXIS;
ISAAC; DAVID, 2015 e AMBEGAONKAR et al., 2012.
I N T R O D U Ç Ã O
28
5 PLASTICIDADE FENOTÍPICA
Plasticidade fenotípica é definida como a capacidade de um mesmo genótipo produzir
diferentes fenótipos em ambientes distintos e é usualmente descrita em termos da norma de
reação (NR) que é a representação gráfica das variações fenotípicas em função do gradiente
experimental (PIGLIUCCI; MURREN; SCHLICHTING, 2006). Mudanças nas condições ambientais
também são capazes de promover variações quantitativas na forma e tamanho da asa. A
temperatura é a principal condição ambiental testada em estudos de plasticidade da asa e,
impressionantemente, todos os estudos apontam para a mesma NR de tamanho, com moscas
desenvolvendo asas cada vez menores conforme a temperatura aumenta. Esse padrão é
observado, inclusive, em espécies filogeneticamente distantes e asas menores são
correlacionadas à diminuição corporal (BITNER-MATHÉ; KLACZKO, 1999b; DAVID et al., 1997,
2011; DE MOED; DE JONG; SCHARLOO, 1997; DEBAT; DEBELLE; DWORKIN, 2009; DEBAT et al.,
2003, 2008; KARAN et al., 2000; LOH et al., 2008; TROTTA et al., 2010; TROTTA et al., 2006).
Por outro lado, quanto às NR da forma da asa em relação à temperatura, também
analisadas em muitos dos estudos citados acima, elas apresentam variações mais erráticas e
de difícil comparação entre os diferentes trabalhos. Isso acontece por que os métodos
descritores de forma variam muito, dificultando associações. Muitos trabalhos utilizam a
morfometria geométrica como, por exemplo, o método Procrustes (KLINGENBERG, 2002) que
extrai informações de forma através de transformações matemáticas sobre coordenadas
tomadas em landmarks da asa com o intuito de se extrair influências alométricas. Essas
medidas, muitas vezes, tornam-se biologicamente abstratas, com pouca recapitulação de um
fenótipo compreensível.
Além da temperatura, a forma e tamanho da asa também respondem a outros
estímulos ambientais como latitudes (AZEVEDO et al., 1998; COLLINGE; HOFFMANN;
MCKECHNIE, 2006; GIBERT et al., 2004; LIEFTING; HOFFMANN; ELLERS, 2009), altitude
(BITNER-MATHÉ; PEIXOTO; KLACZKO, 1995; PITCHERS; POOL; DWORKIN, 2013),
endocruzamento (SCHOU; KRISTENSEN; LOESCHCKE, 2015; TROTTA, VINCENZO et al., 2011) e
nutrição (SOTO et al., 2008; VIJENDRAVARMA; NARASIMHA; KAWECKI, 2011).
Fica evidente que a asa de Drosophila tem servido como modelo para estudos de
plasticidade, abrindo possibilidades para diversos desenhos experimentais. Ainda assim,
muitas questões carecem de investigações. As bases genéticas da plasticidade, quer sejam
I N T R O D U Ç Ã O
29
elas da asa ou de outros traços fenotípicos, são pouco conhecidas. Sabe-se que a plasticidade
tem bases genéticas e evolui (SCHLICHTING, 1986), contudo, pouco sabe-se sobre quais genes
estariam envolvidos com a transdução de sinais ambientais para as vias de desenvolvimento,
com trabalhos apontando para a participação de chaperonas conhecidas como heat shock
proteins (Hsp), mas sem nenhuma indicação clara de que a plasticidade seria mediada por elas
(DEBAT et al., 2006). Novos estudos sobre as bases genéticas da plasticidade são necessários
para uma melhor compreensão das variações quantitativas causadas por modificações
ambientais.
6 SELEÇÃO ARTIFICIAL DA ASA DE D. MELANOGASTER
Em nosso laboratório foram estabelecidas, por um extensivo programa de evolução
experimental por seleção artificial, oito linhagens independentes de Drosophila melanogaster
divergentes para a forma das asas (MENEZES, 2007; TESSEROLI, 2005). 135 linhagens
isofêmeas forneceram um casal cada, dando início à uma linhagem massal. Todos as fêmeas
foram medidas e as 10 com asa mais arredondada foram cruzadas com dez machos
aleatoriamente determinados. As 10 com asa mais alongada também foram cruzadas com dez
machos. Dessa maneira, quatro linhagens foram formadas com as asas mais alongadas, sendo
chamadas de linhagens Longas e quatro com forma mais arredonda, sendo chamadas de
linhagens Redondas (figura 9). Cada linhagem estabelecida sofreu o processo de seleção
initerruptamente durante 21 gerações, quando o processo passou a ser intermitente. Os
fenótipos gerados neste processo extrapolam, inclusive, variações entre espécies (figura 10;
CORRÊA, 2009). Outros estudos também obtiveram uma grande resposta morfológica através
da seleção artificial na asa de Drosophila indicando que, apesar da conservação interespecífica
de forma, há variação genética aditiva (CARTER; HOULE, 2011; HOULE et al., 2003; LE ROUZIC;
HOULE; HANSEN, 2011; WEBER, 1990).
As linhagens de seleção com fenótipos extremos aliadas a todo o conhecimento já
publicado sobre morfologia, evolução, desenvolvimento e genética da asa de Drosophila
permitem a formulação de desenhos experimentais que possam responder a diversas
questões ainda a serem investigadas nessas áreas.
I N T R O D U Ç Ã O
30
Figura 10. Medidas de forma e tamanho da asa de espécies de Drosophila. Note que as linhagens selecionadas para formas extremas da asa (Redonda – círculo vermelho e Longa – círculo azul) extrapolam quase toda a variação de forma. Geração 67. Figura adaptada de CORRÊA (2009).
YAKUBAWILLISTONI
SIMULANSREDONDAMOJAVENSISMELANOGASTERLONGA
ERECTACONTROLEANANASSAE
ESPECIES$
0.3 0.4 0.5 0.6
SHAPE
200
300
400
500
600
SIZ
E
D. ananassae
Controle - D. melanogaster D. erecta
Longa - D. melanogaster D. melanogaster Selvagem
D. mojavensis Redonda - D. melanogaster D. simulans D. willistoni D. yakuba
WSI
– t
am
an
ho
da
asa
WSH
– forma da asa
Espécies
Figura 9. Asas de linhagens selecionadas para formas extremas. Linhagem Redonda (esquerda) e Longa (direita). Linhagens controle (não exibida) apresentam fenótipo intermediário. Geração 123.
O B J E T I V O S E O R G A N I Z A Ç Ã O D A T E S E
31
OBJETIVO GERAL
Investigar a complexidade de promotores de variação quantitativa da forma e
tamanho da asa de Drosophila melanogaster em diferentes níveis de variações biológica: as
correlações genéticas e fenotípicas, as respostas às variações ambientais, as bases celulares e
variações na expressão gênica.
OBJETIVOS ESPECÍFICOS E ORGANIZAÇÃO DA TESE
A presente tese é composta de 4 capítulos que abordam níveis diferentes da variação
biológica subjacentes à variação quantitativa de forma e tamanho da asa. No primeiro,
apresentamos uma descrição da morfologia da asa. No segundo, o ambiente enquanto
promotor de variação é estudado através de estimativas de plasticidade fenotípica. O terceiro
capítulo abarca as bases celulares e o último promove uma busca pelos genes candidatos ao
controle da variação de forma.
I. MORFOLOGIA
Analisar os padrões de integração morfológica na asa de Drosophila melanogaster e testar a
capacidade de previsibilidade da matriz de correlação genética sobre a trajetória evolutiva de
caracteres correlacionados.
II. PLASTICIDADE FENOTÍPICA (submetido ao Journal of Evolutionary Biology)
Avaliar os efeitos da seleção sobre a resposta plástica de tamanho e forma e averiguar a
influência da mudança na média fenotípica de forma sobre as normas de reação.
III. BASES CELULARES (publicado no periódico Genetica)
Avaliar as bases celulares (número e tamanho de células) subjacentes à variação de forma e
tamanho da asa de Drosophila melanogaster.
IV. GENES CANDIDATOS
Identificar genes candidatos ao controle da variação quantitativa de forma da asa através de um
ensaio de microarranjo, de análises dos níveis de expressão gênica em réplicas biológicas do
processo de seleção artificial através de PCR quantitativos em tempo real e por validações
morfológicas, através do silenciamento gênico tecido-específico e análise das consequências
morfológicas.
C A P Í T U L O I : M O R F O L O G I A
32
Phenotypic and Genetic Integration in the Drosophila wing
Daniel Mattos and Blanche Christine Bitner-Mathé*.
Manuscript under current refinement for publication.
Authors’ affiliations: Laboratório de Evolução de Caracteres Complexos – Drosophila,
Departamento de Genética, Instituto de Biologia, Universidade Federal do Rio de Janeiro –
Brasil
Running title: Drosophila wing correlated traits
C A P Í T U L O I : M O R F O L O G I A
33
ABSTRACT
Phenotypic integrations compromise the response of traits to selective pressures.
Although it is considered in most morphological evolutionary models, phenotypic integration
as a facilitator or a constraint to changes is hard to assess since most studies are hindsight,
departing from the established morphological divergence to infer the previous integration
patterns. Here we use artificially selected strains for divergent wing shapes in Drosophila
melanogaster to test predictions based on the genetic correlation matrix of the population
prior to selection. Based on Pre-Selection generation, we present heritability estimates,
phenotypic and genetic correlation matrices for wing shape, size and venation landmarks and
found that most of the correlated responses of wing traits followed predictions by genetic
parameters.
Keywords: Morphology ; wing shape; artificial selection.
C A P Í T U L O I : M O R F O L O G I A
34
INTRODUCTION
Organ development, as many complex phenotypic traits, is multifactorial, influenced
by many genes, many environmental conditions and their interaction. Organ development is
mediated by modularization of its parts, with a relative independence of each modular group
of traits. Modules are defined as “structural units that are internally integrated by
developmental interactions” (KLINGENBERG, 2002, 2014). Traits within modules are then
expected to be more phenotypically integrated than those in adjacent modules by sharing
genetic variation or by responding to common environmental cues. Evolutionary implications
of modules and the evolvability of the modules themselves are not straightforward and
require more studies (WAGNER; PAVLICEV; CHEVERUD, 2007). These phenotypic integration
(PI) by means of genetic correlation can be summarized by the genetic variance-covariance
matrix, which can change the evolutionary trajectory of individual traits (PHILLIPS; WHITLOCK;
FOWLER, 2001; STEPPAN; PHILLIPS; HOULE, 2002) e or constraining evolution. Some authors
defend that developmental modules evolve aligned with functional modules, thus facilitating
the organ evolution (CHEVERUD, 1984; WAGNER; ALTENBERG, 1996) while others emphasize
modularity as a constraint (ARTHUR, 2001). The two viewpoints are not, however, exclusive
and modularity will be a constraint or a facilitator depending on the selective pressures
imposed to the structure. Although it is acknowledged in morphological evolutionary models,
it is hard to target questions on the integration states and its relations to evolution since most
analyses have a hindsight perspective. Artificial selection programs are time and resource
consuming, but they offer a way to pose questions on phenotypic integration and modularity
on a forward perspective since one can assess the previous integration profile and the
consequences after a morphological divergence is established.
For such studies, the wing of Drosophila is an ideal target since both genetic and
phenotypic aspects have been largely described (BITNER-MATHÉ; KLACZKO, 1999a, b; DAVID
et al., 2003; MATTA; BITNER-MATHÉ, 2004, 2010; MATTA; BITNER-MATHÉ; ALVES-FERREIRA,
2011; TORQUATO et al., 2014). Five longitudinal veins run along the proximodistal axis and
two transversal veins run along the anteroposterior axis. This venation pattern is widely
conserved across Drosophila species. While the expression domains of some genes encompass
the entire wing imaginal disc, others are only expressed in one compartment or along the
developing veins (BLAIR, 2007; KOLZER, 2003; LECUIT; LENNE, 2007). This mosaic of different
C A P Í T U L O I : M O R F O L O G I A
35
genes being expressed locally or widespread and at a particular time may account for part of
the phenotypic integrations of wing features.
On the morphological aspect, no evidence of modularization of wing compartments
was found (KLINGENBERG, 2009). However, there is evidence of genetic correlation among
wing traits (MATTA; BITNER-MATHÉ, 2004), but wether genetic correlation determines the
trajectory of correlated traits is still unknown. Lastly, the reaction norms of each of these traits
can be more or less coupled with the genetic architecture underlying wing shape. Here we
analyze Drosophila melanogaster strains that were submitted to an extensive artificial
selection program targeting wing overall shape at the generation prior to the program (Pre-
Selection) and the 64th generation after selection (Post-Selection). We describe morphological
variation of wing features, phenotypic integration patterns, genetic parameters such as
heritability estimates and the genetic correlation matrix for Pre-Selection population and
conclude that most of the traits trajectories were determined by the genetic correlation
matrix, although unpredicted trajectories were also observed.
MATERIAL AND METHODS
ARTIFICIAL SELECTION PROGRAM
The strains used in this work were previously established by B. C. Bitner-Mathé, D.
Tesseroli & B. F. Menezes. The artificial selection program is explained in details in Menezes
et al (2013). Briefly, 135 isofemale lines founded the initial baseline population from which
strains were established by decreasing or increasing a shape index based on width-to-length
ratio. Selection was applied for 21 consecutive generations and intermittently after that. Here
we use the most divergent strains selected for elongated wings (named L strain) and for
rounded shaped wings (R strain) at the 64th generation after selection started (hereafter
referred to as Post-Selection). Strains are assumed fairly homozygous.
C A P Í T U L O I : M O R F O L O G I A
36
PRE-SELECTION GENERATION HERITABILITY, PHENOTYPIC AND GENETIC CORRELATION
ESTIMATES
From the offspring of each isofemale line, two virgin females were taken and allowed
to mate with one male collected in the field. Females were then separated and put to oviposit
in separate vials at 16° for rearing. Wings were measured for two female offspring and the
mother. A regression of the offspring on the mother was performed to assess heritability
estimates, which equal two times the regression slope (FALCONER; MACKAY, 1996).
Heritability, genetic and phenotypic correlations from Pre-Selection flies are used to predict
the morphological consequences on the correlated traits and compared to the actual
trajectory of the Post-Selection flies.
POST SELECTION GENERATION: MORPHOLOGICAL INTEGRATED RESPONSE AND
REACTION NORMS
At the 64th generation, 10 couples of flies with virgin females from each strain were
transferred to two different temperatures (16°C or 25°C) for mating and oviposition in a
standard Drosophila medium. Every two days, adults were transferred to a different bottle for
oviposition, hence creating replicates. Measurements were taken for approximately 15 left
wings (up to 5 in each replicate) of females per strain submitted to each selection program
and reared in one of the two developmental temperatures.
WING MORPHOLOGY MEASUREMENTS
Since selection was applied to females only (with males being randomly chosen at each
generation), all analyses in this paper refer to female individuals. All left wings were mounted
on slides and photographed with a digital camera attached to a stereoscope microscope. The
program ImageJ (http://rsbweb.nih.gov/ij/) was used to take the coordinates of 20 semi-
landmarks around the wing border and the position of 11 landmarks at the intersection of
veins or at their extremities, ensuring homology (Fig. 1). Shape and size estimates were
assessed by the Ellipse Method (Klaczko and Bitner-Mathé, 1990; Klaczko, 2006)) which
estimates the best-fitted ellipse to the given coordinates taken around wing contour. Wing
shape (WSH) is defined as the ratio between the ellipse smallest and largest radius (WW/WL)
C A P Í T U L O I : M O R F O L O G I A
37
and size (WSI) is the square root of the product of those measures (√𝑊𝐿 ∗ 𝑊𝑊). The Ellipse
method also provides a positional description of venation pattern by Polar Coordinates, with
each landmark being characterized by the angle formed between WL and the radius
connecting the landmark to the center of the ellipse. Since radius lengths are highly correlated
to size, we analyzed landmark variation by the angular component only which will be here
forth addressed as , followed by the letter corresponding to the respective landmark.
Statistical analyses were carried out using SYSTAT© v.13.0 (SPSS Inc.).
STATISTICS
Homogeneity of distributions was tested by an analysis of covariance (ANCOVA) for
univariate traits and a multivariate analysis of covariance (MANCOVA) for the group of
landmarks. Both analyses were carried out using the directions of selection (SEL: L x R strains)
as a fixed effect while developmental temperature (DT: 16ºC and 25ºC) was not considered a
fixed effect. Replicates were nested within the interaction term between those factors.
Figure 1. (a) Wing semi-landmarks (in red) to estimate wing contour best-fitted ellipse (red ellipse). Landmarks describing wing venation in purple. Landmark H was not included in statistical analyses due to a method’s limitation. (b) Estimated ellipse and determination of wing largest radius (WL), smallest radius (WW) and
illustrative description of landmark B by B.
C A P Í T U L O I : M O R F O L O G I A
38
Phenotypic correlation matrices were estimated by the Pearson’s product-moment
correlation. The total genetic correlation (rG) was calculated as the arithmetic mean of two
reciprocal between-trait daughter–mother covariances divided by the geometric mean of the
within-trait daughter–mother covariances.
rG=(COVX1Z2+COVX2Z1)/(2√(COVX1Z1COVX2Z2)), where COVX1Z2 is the covariance between
trait 1 of the parents and trait 2 of the offspring, COVX2Z1 is the covariance between trait 2 of
the parents and trait 1 of the offspring, COVX1Z1 and COVX2Z2 are the offspring–parent
covariances of traits 1 and 2, respectively (Becker, 1992; Falconer; Mackay, 1996; as explained
in Matta; Bitner-Mathé 2004).
RESULTS
Heritability estimates, phenotypic and genetic correlations for wing traits of the Pre-
Selection baseline population are exhibited on Table 1. WSH has high heritability (h2=0.59),
suggesting high additive genetic variance to respond to the selective pressure. For WSI,
heritability is low and not significant, indicating a high environmental regulation with little
genetic variance contributing to total phenotypic variance. Most landmark angles also
presented high heritability.
The Pre-Selection population was then submitted to a bidirectional artificial selection
that stretched the variation of WSH by selecting flies with elongated wings (low WSH values, L
strains) or rounder wings ((high WSH values, R strains). The effects of this program on the 64th
generation can be appreciated in Fig. 2 that presents the histograms of Pre-Selection
population reared at 16ºC and the selected strains reared at 16ºC and 25ºC. Both L and R
strains responded to the selection applied, with a slightly stronger response in the R Strain. It
is noteworthy that despite the great divergence in WSH, distribution of WSI remains overlapped
showing that selecting for wing shape did not impose a change in wing size. Differences
between the distribution of WSI from Post-Selection flies reared at 16ºC and 25ºC shows a high
temperature-related plasticity of this trait. Fig. 3 shows two illustrative wings from L and R
strains with the mean estimated ellipses superposed. To test homogeneity of these
distributions, we performed an Analysis of Covariance (ANCOVA) on the Post-Selection flies,
presented on Table 2. As expected by the high heritability of WSH, we detected a strong
response effect of direction of selection (SEL: L x R strains) holding 88% of WSH variance. We
C A P Í T U L O I : M O R F O L O G I A
39
detected no change in WSH due to developmental temperature (DT). Difference is detected in
WSI between L and R strains. WSI was highly influenced by DT and this result is in accordance
with the low heritability found for this trait. Moreover, we detected a significant interaction
SEL*DT for WSH, indicating a change in the reaction norm of L and R strains due to selection.
Predictions of evolutionary trajectories after selection are based on the genetic
correlations of the Pre-Selection population. According to the genetic correlations with WSH,
an increase in this value (i.e. selecting for a rounder wing), would impose a reduction in WSI
(rG = -0.31), and increase in A (rG=0.45) and so forth. OnlyD and I do not exhibit significant
rG with WSH. Since angular movements are harder to visualize, these correlations can be best
viewed in Fig. 5, which graphically summarizes these variations.
C A P Í T U L O I : M O R F O L O G I A
40
WSH WSI A B C D E F G I J K O
WSH 0.59 -0.309 0.446 -0.454 -0.186 0.137 0.348 0.414 0.320 -0.145 0.705 0.463 0.495
WSI -0.269 0.27 -0.139 0.392 -0.066 -0.078 -0.261 0.125 -0.035 -0.353 -0.167 -0.121 -0.317
A 0.092 0.250 0.42 -0.024 -0.253 -0.527 -0.750 0.678 0.593 -0.639 0.771 0.368 0.541
B -0.526 0.391 0.291 0.37 0.467 -0.191 -0.404 0.039 -0.033 -0.250 -0.497 -0.316 -0.440
C 0.002 -0.102 0.053 0.143 0.57 0.560 -0.041 -0.089 -0.350 0.104 -0.523 -0.794 -0.685
D 0.284 -0.230 -0.120 -0.271 0.556 0.78 0.584 -0.221 -0.419 0.127 -0.478 -0.674 -0.588
E 0.339 -0.279 -0.528 -0.514 0.154 0.623 0.71 -0.427 -0.521 0.465 -0.379 -0.384 -0.414
F 0.020 0.247 0.796 0.384 -0.057 -0.184 -0.501 0.57 0.872 -0.504 0.771 0.590 0.785
G -0.095 0.147 0.696 0.381 -0.131 -0.249 -0.524 0.900 0.42 -0.408 0.819 0.780 0.959
I 0.131 -0.258 -0.455 -0.409 0.103 0.361 0.691 -0.461 -0.466 0.62 -0.556 -0.457 -0.581
J 0.027 0.253 0.863 0.293 -0.048 -0.147 -0.463 0.900 0.843 -0.426 0.29 0.759 1.013
K 0.005 0.075 0.688 0.172 -0.166 -0.204 -0.344 0.774 0.793 -0.317 0.885 0.39 0.980
O -0.023 0.179 0.786 0.237 -0.144 -0.203 -0.424 0.855 0.834 -0.365 0.943 0.942 0.49
Table 1. Phenotypic (lower diagonal) and genetic (upper) correlation matrices in the Pre-Selection generation. Heritability estimates in blue.
Pearson’s product-moment correlations between wing traits, bold values are significant (P < 0.05).
C A P Í T U L O I : M O R F O L O G I A
41
Figure 2. Histograms of wing size (WSI; left) and shape (WSH; right) from D. melanogaster unselected females
(Pre-Selection) that later originated the artificially selected strains (Post-Selection from the 64th generation
after selection initiated reared at 16°C and 25°C). Note that WSI distributions overlap while WSH exhibit great
divergence between L and R, with R strain presenting a slightly stronger response to selection when
compared to Pre-Selection distribution.
C A P Í T U L O I : M O R F O L O G I A
42
Effects D.F WSH WSI
Direction of Selection (SEL) 2 0.876 *** 0.010 Developmental Temperature (DT) 1 0.004 0.971 *** SEL x DT 2 0.104 *** 0.008 Replicate (SEL x DT) 9 0.009 0.007 Residuals 63 *** p < 0.001
The estimates of WSH and WSI are composed by the measures of the wing length (WL)
and width (WW) as described in the Material and Methods section. The selection program
focused on the ratio WW/WL to isolate the targeted phenotype. Fig. 4 shows the histograms of
these measures. Since WL and WW present temperature-related phenotypic plasticity, the
effects of selection compared to Pre-Selection flies should be restrained to flies reared at 16ºC
only. Regarding WL, we found a more intense response in the R strain while for WW, both R
and L strains equally responded. This scenario suggests that reducing the biological axes might
Figure 3. Illustrative wings from L Strains (left) and R Strains (right). Superimposition of ellipses drawn by the
parameters estimated by the Ellipse method from mean value of 1L and 1R strains.
Table 2: Results of the ANCOVA for wing shape (WSH) and size (WSI) testing the homogeneity
of distribution between directions of selection (L x R strains) and developmental
temperature (16ºC and 25ºC). Reaction norms changes are assessed by the interaction term
(SEL*DT). Percentage of variance explained by each effect.
C A P Í T U L O I : M O R F O L O G I A
43
be easier than stretching. Similar variation haS been observed in other experiments with these
strains (unpublished data).
Figure 4: Histograms of wing length (WL; upper four graphs) and width (WW; lowest four
graphs) from D. melanogaster unselected females (Pre-Selection) that later originated the
artificially selected strains (Post-Selection from the 64th generation after selection initiated
reared at 16°C and 25°C). Since wing length and width are size-sensitive to developmental
temperature, artificial selection effects should be noticed on the left graphs since all flies were
reared at 16°C. Note that WL exhibits a more pronounced divergence in the R Strains while
for WW, L and R Strains diverged from the Pre-Selection distribution.
C A P Í T U L O I : M O R F O L O G I A
44
We then analyzed the integration pattern in Post-Selection flies. Polar coordinates
provided by the Ellipse Method were transformed into Cartesian coordinates following the
equation x= R*cos and y= R*sin, where R is the radius and the angle of each landmark
(Fig. 5). Landmarks A, C, D, E, all along the wing border, are highly divergent between L and R
strains. Proximal landmarks F, G, J, K, O and distal landmarks C, D and E are very divergent,
suggesting that selection affected the whole wing blade and was not constrained to some of
its compartments. To test homogeneity of variation of landmarks, we performed a
multivariate analysis of variance (MANCOVA) presented on Table 3. We detect a significant
effect on landmarks of SEL (L x R) and of developmental temperature (16ºC x 25ºC) and ad
interaction between those factors indicating a change in the reaction norms. These results
indicate a high degree of integration between wing traits, with a change in wing overall shape
imposing a change in most of its components. Venation pattern also exhibits temperature-
related phenotypic plasticity. Phenotypic correlation matrices are exhibited on Table 4 and
summarized in Fig. 6. The increase in WSH (R Strains) is correlated with an increase in A, C,
D, F, G, JG, G, O and a decrease in I. Comparing the predicted movements from Pre-
Selection genetic correlation with the actual results in Post-Selection, we observe that the first
successfully predicted the direction of movements for A, D, F, G, I, J, K, O, but not
for WSI, B, C and E.
Ellipse
Effects Wilk’s F-ratio D.F. p
Direction of Selection (SEL) 0.289 30.621 11, 137 0.000
Developmental Temperature (DT) 0.420 17.196 11, 137 0.000
SEL x DT 0.841 2.357 11, 137 0.011
Strain (SEL) 0.161 7.311 44, 526 0.000
N = 155
Table 3: Summary of the results of MANCOVA for variation in angular movements of
landmarks (Ellipse ).
C A P Í T U L O I : M O R F O L O G I A
45
- 320 - 240 - 160 - 80 0 80 160 240 320 400
- 300
- 250
- 200
- 150
- 100
- 50
0
50
100
150
200
A B
C
D
E
F
G H
I
O
JK
16°C
- 320 - 240 - 160 - 80 0 80 160 240 320 400
A x
- 300
- 250
- 200
- 150
- 100
- 50
0
50
100
150
200
Ay
AB
C
D
E
F
GH
I
O
JK
A
B
C
D
E
F
G H
I
O
JK
16°C
25°C
Pre
-Sele
cti
on
Po
st-
Se
lecti
on
- 320 - 240 - 160 - 80 0 80 160 240 320 400
- 300
- 250
- 200
- 150
- 100
- 50
0
50
100
150
200
Figure 5. Plot of landmarks from Pre-Selection (black dots – upper graph) and from L Strain
(blue dots) and R Strain (red dots) reared in two developmental temperatures (16°C and
25°C). Polar coordinates extracted from the Ellipse method and transformed to Cartesian
coordinates. Upper wings on the right indicate orientation of the wing with landmarks.
Letters A-O indicate the landmark represented by the nearest 95% confidence ellipse.
Landmark H excluded from statistical analyses.
C A P Í T U L O I : M O R F O L O G I A
46
WSH WSI A B C D E F G I J K O WSH 1
WSI 0.347 1
A 0.926 0.225 1
a B 0.168 0.068 0.183 1
C 0.783 0.138 0.712 0.415 1
D 0.850 0.271 0.775 0.307 0.851 1
E -0.377 -0.140 -0.403 -0.204 -0.136 -0.046 1
F 0.868 0.224 0.953 0.255 0.669 0.725 -0.489 1
G 0.768 0.054 0.884 0.351 0.641 0.648 -0.552 0.955 1
I -0.812 -0.505 -0.716 -0.203 -0.528 -0.645 0.498 -0.706 -0.627 1
J 0.878 0.153 0.957 0.256 0.696 0.732 -0.473 0.981 0.947 -0.702 1
K 0.731 0.022 0.880 0.275 0.561 0.558 -0.523 0.933 0.947 -0.573 0.954 1
O 0.811 0.112 0.939 0.248 0.652 0.675 -0.452 0.965 0.929 -0.615 0.976 0.968 1
WSH WSI A B C D E F G I J K O WSH 1
WSI -0.045 1
A 0.935 -0.006 1 b
B 0.562 0.228 0.658 1
C 0.811 0.064 0.815 0.690 1
D 0.872 -0.004 0.867 0.636 0.919 1
E -0.434 -0.100 -0.564 -0.586 -0.505 -0.349 1
F 0.918 -0.027 0.945 0.626 0.834 0.861 -0.576 1
G 0.822 0.050 0.889 0.608 0.819 0.810 -0.599 0.928 1
I -0.711 -0.157 -0.769 -0.601 -0.734 -0.674 0.759 -0.787 -0.767 1
J 0.933 -0.058 0.955 0.607 0.816 0.851 -0.583 0.981 0.930 -0.786 1
K 0.880 -0.060 0.928 0.606 0.780 0.842 -0.554 0.969 0.927 -0.726 0.967 1
O 0.921 -0.053 0.955 0.629 0.816 0.866 -0.566 0.976 0.935 -0.761 0.979 0.974 1
Table 4. Phenotypic correlation matrices for Post-Selection generation reared at 16ºC (a) and 25ºC (b). N=33 (a) and 38 (b). p<0.05 in bold.
C A P Í T U L O I – M O R F O L O G I A
47
DISCUSSION
Heritability found for wing shape was high and significant (h2=0.59; p<0.05) while for
size was non-significant, a pattern similar to what has been reported for D. simulans, D.
serrata, D. gouveai, D. mercatorum, D. paranaensis, Z. indianus and other D. melanogaster
populations, indicating the conservation in the amount of additive genetic variance for these
traits (GILCHRIST; HUEY; SERRA, 2001; HOFFMANN; SHIRRIFFS, 2002; LEIBOWITZ; SANTOS;
FONTDEVILA, 1995; LOH; BITNER-MATHÉ, 2005; MATTA; BITNER-MATHÉ, 2004; MORAES et
al., 2004; MORAES; SENE, 2004; SZTEPANACZ; BLOWS, 2015). The high value found for wing
shape is also in accordance with the large morphological divergence achieved by the selection
program in Post-Selection strains. The low and non-significant h2 observed for wing size may
also explain how Post-Selection strains kept an unchanged size, retaining temperature-related
plastic response with little additive genetic variance relative to the total phenotypic variance.
The contrasting results for wing shape and size suggest the first is under a tighter genetic
control while size is more responsive to environmental variations such as temperature. The
high heritability for wing shape and the fact that shapes achieved by artificial selection are
hardly seen in nature suggest that wing shape is under selective pressure, although the
Figure 6. Graphical representation of Genetic and Phenotypic Correlations (see tables 1 and 3) from Pre (left) and
Post-Selection (right) reared at 16ºC. Arrow size reflects intensity of correlation with WSH. The angular variation
of any given trait tends to be followed by the variation shown for all others. Orientation of arrows shows only
one direction of the correlations and all can be inverted once the sign of WSH is also inverted.
.
C A P Í T U L O I : M O R F O L O G I A
48
adaptive character of wing shape has not yet been satisfactorily demonstrated (BIRDSALL et
al., 2000; WEBER et al., 1999; ZIMMERMAN; PALSSON; GIBSON, 2000).
Regarding the integration patterns of the wing in Pre-Selection flies, the only traits with
a significant phenotypic correlation to wing shape are wing size, D, B, D and E, but
almost all exhibit high genetic correlations, i.e., covariation in these traits in the offspring are
partially explained by covariation in those of the mother. As wings depart from the
intermediate shape by means of the selection program, trajectories of other wing traits are
expected to be influenced by the genetic correlation matrix. As seen on figure 5, for most of
the wing traits, genetic estimates were successfully able to predict the new locations of
landmarks in R and L strains. In fact, differences between predicted and observed are roughly
in intensity of the displacement of landmarks, rather than direction, with few exceptions as
B, C, E and I. All these landmarks are located on the distal portion of the wing with
most being along the wing border.
Two morphological compartments with an apparent high genetic modularization can
be observed. On the most proximal portion of the wings, encompassing landmarks A, F,
G, J, K and O, there appears to be a consistent developmental module with all
landmarks responding to the same direction and with equal intensities. The most distal
portion of the wing also appears as a module, with landmarks C and D responding similarly.
These results are in contradiction with those obtained by Klingenberg (2009) and by
Klingenberg and Aklan (2000) through a different approach for module recognition. They did
not find evidence of morphological modularization on the wing. Also interestingly, wing size
responded in the opposite direction predicted by the genetic correlation estimates, slightly
increasing in R strains, although this difference was not significant, as seen on Table 2.
The effects of selection on shape variation can be appreciated on figure 1. The
response to selection of R strains was more pronounced than in L strains. Furthermore,
selection affected both axes (length and width) in R strains, whereas in L strains only the width
exhibited a pronounced divergence (figure 3, considering only flies reared at 16°C, since
comparisons with 25°C cannot depict allometric variations). However, small random
fluctuations in shape and size are expected when different generations are analyzed, and this
might account for these effects.
C A P Í T U L O I : M O R F O L O G I A
49
Despite the great divergence in shape, strains did not present any significant size
difference, although all strains had their size reduced compared to the Pre-Selection
population. This is surprising considering that both size and shape variables are calculated by
the same ellipse axes provided by the adjustment of the fittest ellipse to the set of
coordinates. The selection program also did not alter wing size thermal plasticity and the
reaction norm found was similar to those reported in the literature, with wings becoming
smaller at higher temperatures (DAVID et al., 2011; DEBAT; DEBELLE; DWORKIN, 2009). Clinal
studies have repeatedly found smaller wings at higher latitudes, a pattern usually explained
by temperature-related selection (COYNE; BEECHAM, 1987; GOCKEL et al., 2001; HOFFMANN;
SHIRRIFFS, 2002). Other studies have reported a certain freedom between the variation of
these two traits (DEBAT; DEBELLE; DWORKIN, 2009; DEBAT et al., 2003; GILCHRIST; HUEY;
SERRA, 2001). Nonetheless, these results do not imply a total independency of the traits and
it is more likely that the variance in one trait will interfere, at least partially, with the other,
especially when considering its evolution, since the wing is perceived by selection as a whole
structure and, in nature, a trade-off between these two features of the wing might occur to
regulate the outcome phenotype.
REFERENCES
ARTHUR, W. Developmental drive: An important determinant of the direction of phenotypic evolution. Evolution and Development, v.3 , p. 271-278, 2001.
BECKER, W. A. Manual of quantitative genetic. 5th. ed. Pullman, WA, U.S.A: Academic Enterprises, 1992.
BIRDSALL K et al. Genetic variation for the positioning of wing veins in Drosophila melanogaster. Evolution and Development, v. 2, p. 16–24, 2000.
BITNER-MATHÉ, B. C.; KLACZKO, L. B. Heritability , phenotypic and genetic correlations of size and shape of Drosophila mediopunctata wings. Heredity, v. 83, n. 6, p. 688–696, 1999a.
BITNER-MATHÉ, B. C.; KLACZKO, L. B. Plasticity of Drosophila melanogaster wing morphology: effects of sex, temperature and density. Genetica, v. 105, n. 2, p. 203–210, 1999b.
BLAIR, S. S. Wing vein patterning in Drosophila and the analysis of intercellular signaling. Annual review of cell and developmental biology, v. 23, p. 293–319, 2007.
C A P Í T U L O I : M O R F O L O G I A
50
CHEVERUD, J. M. Quantitative genetics and developmental constraints on evolution by selection. Journal of theoretical biology, v. 110, p. 155–171, 1984.
COYNE, J. A.; BEECHAM, E. Heritability of two morphological characters within and among natural populations of Drosophila melanogaster. Genetics, v. 117, n. 4, p. 727–737, 1987.
DAVID, J. R. et al. Genetic variability of sexual size dimorphism in a natural population of Drosophila melanogaster : an isofemale-line approach. Journa of Genetics, v. 82, n. 3, p. 79–88, 2003.
DAVID, J. R. et al. Thermal phenotypic plasticity of body size in Drosophila melanogaster: sexual dimorphism and genetic correlations. Journal of Genetics, v. 90, n. 2, p. 295–302, 2011.
DEBAT, V. et al. Allometric and nonallometric components of Drosophila wing shape respond differently to developmental temperature. Evolution; international journal of organic evolution, v. 57, n. 12, p. 2773–2784, 2003.
DEBAT, V.; DEBELLE, A.; DWORKIN, I. Plasticity, canalization, and developmental stability of the Drosophila wing: joint effects of mutations and developmental temperature. Evolution, v. 63, n. 11, p. 2864–2876, 2009.
FALCONER, D. S.; MACKAY, T. F. C. Introduction to quantitative genetics. Longmans Green. Harlow, Essex, Uk, 1996.
GILCHRIST, G. W.; HUEY, R. B.; SERRA, L. Rapid evolution of wing size clines in Drosophila subobscura. Genetica, v. 112-113, p. 273–286, 2001.
GOCKEL, J. et al. Nonclinality of molecular variation implicates selection in maintaining a morphological cline of Drosophila melanogaster. Genetics, v. 158, n. 1, p. 319–323, 2001.
HOFFMANN, A. A.; JENNIFER SHIRRIFFS. Geographic variation for wing shape in Drosophila serrata. Evolution & development, v. 56, n. 5, p. 1068–1073, 2002.
KLINGENBERG, C. P. Morphometric integration and modularity in configurations of landmarks: tools for evaluating a priori hypotheses. Evolution & Development, v. 11, n. 4, p. 405–421, 2009.
KLINGENBERG, C. P. Morphometrics and the role of the phenotype in studies of the evolution of developmental mechanisms. Gene, v. 287, p. 3–10, 2002.
KLINGENBERG, C. P. Studying morphological integration and modularity at multiple levels: concepts and analysis. Philosophical transactions of the Royal Society of London., v. 369, p. 201-249, 2014.
KLINGENBERG, C. P.; AKLAN, S. T. D. Z. Morphological integration between developmental compartments in the Drosophila wing. Evolution, v. 54, n. 4, p. 1273–1285, 2000.
KOLZER, S. Defective proventriculus is required for pattern formation along the proximodistal axis, cell proliferation and formation of veins in the Drosophila wing. Development, v. 130, n. 17, p. 4135–4147, 1 set. 2003.
LECUIT, T.; LENNE, P. F. Cell surface mechanics and the control of cell shape, tissue patterns and morphogenesis. Nature Reviews Molecular Cell Biology, v. 8, n. 8, p. 633–644, ago. 2007.
C A P Í T U L O I : M O R F O L O G I A
51
LEIBOWITZ, A.; SANTOS, M.; FONTDEVILA, A. Heritability and selection on body size in a natural population of Drosophila buzzatii. Genetics, v. 141, n. 1, p. 181–189, 1995.
LOH, R.; BITNER-MATHÉ, B. C. Variability of wing size and shape in three populations of a recent Brazilian invader, Zaprionus indianus (Diptera: Drosophilidae), from different habitats. Genetica, v. 125, n. 2-3, p. 271–281, 2005.
MATTA, B. P.; BITNER-MATHÉ, B. C. An interspecific QTL study of Drosophila wing size and shape variation to investigate the genetic basis of morphological differences. Genetics and molecular research, v. 9, n. 4, p. 2032–2049, 2010.
MATTA, B. P.; BITNER-MATHÉ, B. C. Genetic architecture of wing morphology in Drosophila simulans and an analysis of temperature effects on genetic parameter estimates. Heredity, v. 93, n. 4, p. 330–41, out. 2004.
MATTA, B. P.; BITNER-MATHÉ, B. C.; ALVES-FERREIRA, M. Getting real with real-time qPCR: a case study of reference gene selection for morphological variation in Drosophila melanogaster wings. Development genes and evolution, v. 221, n. 1, p. 49–57, 2011.
MENEZES, B. F. et al. The influence of male wing shape on mating success in Drosophila melanogaster. Animal Behaviour, v. 85, n. 6, p. 1217–1223, 2013.
MORAES, E. M. et al. Wing shape heritability and morphological divergence of the sibling species Drosophila mercatorum and Drosophila paranaensis. Heredity, v. 92, n. 5, p. 466–73, 2004.
MORAES, E. M.; SENE, F. M. Heritability of wing morphology in a natural population of Drosophila gouveai. Genetica. v. 121, p. 119–123, 2004.
PHILLIPS, P. C.; WHITLOCK, M. C.; FOWLER, K. Inbreeding Changes the Shape of the Genetic Covariance Matrix in Drosophila melanogaster. Genetics, v. 3, p. 1137-1145, 2001.
STEPPAN, S. J.; PHILLIPS, P. C.; HOULE, D. Comparative quantitative genetics: evolution of the G matrix. Trends in Ecology & Evolution, v. 17, n. 7, p. 320–327, 2002.
TORQUATO, L. et al. Cellular basis of morphological variation and temperature-related plasticity in Drosophila melanogaster strains with divergent wing shapes. Genetica, p. 1–11, 2014.
WAGNER, G. P.; ALTENBERG, L. Perspective: Complex adaptations and the evolution of evolvability. Evolution, v. 50, p. 967–976, 1996.
WAGNER, G. P.; PAVLICEV, M.; CHEVERUD, J. M. The road to modularity. Nature Reviews Genetics, v. 8, n. 12, p. 921–31, 2007.
WEBER, K. et al. An Analysis of Polygenes Affecting Wing Shape on Chromosome 3 in Drosophila melanogaster. Genetics, v. 786, p. 773–786, 1999.
ZIMMERMAN, E.; PALSSON, A.; GIBSON, G. Quantitative Trait Loci Affecting Components of Wing Shape in Drosophila melanogaster. Genetics, v. 155, n. 2, p. 671–683, 2000.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
52
Thermal Plasticity Evolution in Strains of Drosophila melanogaster Selected for Divergent
Wing Shape
Daniel Mattos1, Felipe Rocha2, Louis Bernard Klaczko2 and Blanche Christine Bitner-Mathé1*.
Submitted to the Journal of Evolutionary Biology on 03/12/2014 and resubmitted (the
current revised version) on 08/03/2015.
Authors’ affiliations:
1. Universidade Federal do Rio de Janeiro, Departamento de Genética, Instituto de
Biologia, Brasil
2. Universidade Estadual de Campinas, Departamento de Genética e Evolução, Instituto
de Biologia, Brasil
Running title: influence of wing shape mean on reaction norm
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
53
ABSTRACT
Phenotypic plasticity is the property of genotypes that allows an increase in phenotypic
variance when organisms develop in different environmental conditions by adjustment of
developmental pathways, residing on the frontier of eco-evo-devo studies. Few empirical
studies have addressed the long debate on the relationship between the mean phenotypic
value of a trait and its plastic. Here we used strains of Drosophila melanogaster artificially
selected for divergence of wing shape that retain remarkably similar wing sizes to investigate
the independency between mean value and reaction norm. Flies from elongated, rounded and
unselected control strains, with three distinct phenotypic mean values, developed at
a thermal gradient ranging from 14°C to 30°C. Wing size was highly responsive to
developmental temperature and sex, but similar amongst all strains. On the other hand,
reaction norms of wing shape were altered by the selection program and the phenotypic mean
value of wing shape explained 36% of the variation amongst reaction norms. The contrasting
responses of wing shape and size is intriguing, especially since both are estimated by the same
measures of wing length and width. Wing length and width exhibited a proportional response,
allowing size to vary without imposing great disturbance on wing shape variance. Results
indicate that wing shape and its reaction norm cannot be assumed independent while wing
size and shape can have independent responses.
keywords: artificial selection – ecological genetics - insects - morphometrics – reaction
norm – wing size
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
54
INTRODUCTION
Phenotypic plasticity is a central issue in eco-evo-devo studies since it allows variation
by changes in developmental programs when organisms are exposed to different
environmental conditions producing different phenotypes (WEST-EBERHARD, 2003). Plasticity
has been demonstrated to evolve and thus has a genetic basis (DEBAT; DEBELLE; DWORKIN,
2009; PIGLIUCCI, 2005). Plastic variation is often represented by the reaction norm (RN) that
plots the phenotypic values across an environmental gradient. Two possibly overlapping
genetic mechanisms might underlie RN shapes. RN might be influenced by allelic sensitivity,
where some alleles responsible for trait establishment have varying effects across
environments, or by regulatory sensitive genes regulating the expression of downstream
genes (VIA et al., 1995). Allelic sensitivity implies a higher dependency between the RN and
the trait mean, since evolution of the trait mean would have a direct impact on the genetic
basis of its plasticity. The relative contribution of allelic sensitivity or regulatory genes and
hence the independency between the trait mean and plasticity has long been debated (DAVID
et al., 2005; SCHEINER, 1993; VIA, 1993). Rocha et al. (2009) addressed this issue and found a
high correlation between the trait mean and the RN showing that, for some traits,
independency should not be assumed. Trait mean and RN are likely to show variable degrees
of independence for different traits, and more empirical data is necessary to provide a better
framework for understanding of the genetic basis and evolution of phenotypic plasticity. Since
plasticity is a property hard to manipulate, one can address this issue by imposing variation
on the population mean in order to investigate the side effects on the plastic response.
Phenotypic plasticity has long been studied on the wing of Drosophila, mostly because
the wing a virtually two-dimensional organ and much of its genetics and development is
largely described, thus making it an interesting model for such studies (BITNER-MATHÉ;
KLACZKO, 1999a, b; BLAIR, 2007; DAVID, JEAN et al., 2011; GIRALDEZ; COHEN, 2003; KARAN;
MORETEAU; DAVID, 1999; NEUFELD et al., 1998). Wing size is sensitive to rearing temperature,
with flies exhibiting bigger wings at lower temperatures (DAVID et al., 1994). The extent to
which wing plasticity is independent from the population mean has never been addressed and
whether the wing shape RN depends on the phenotypic value remains unclear. Furthermore,
most studies focusing wing plasticity use a narrow range temperature gradient with only two
or three temperatures, thus leaving those RN still to be more fully described.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
55
Here we used strains of Drosophila melanogaster artificially selected for extreme
values of wing shape and control unselected strains with intermediate values to investigate
the selection effects on the wing plasticity and to test the independency between wing shape
mean value and its reaction norm. The strains were founded from the same baseline
population and the artificial selection program focused on the wing shape described by width-
to-length ratio. Our results indicate that wing shape temperature-related RN is influenced by
the phenotypic mean value, with rounded and unselected control strains exhibiting opposite
slope signs while L strains developed an upward parabolic RN. Divergently selected strains
exhibit very similar size, showing independency between wing size and shape variation.
Furthermore, no change in size RN was detected. Lastly we show that a proportional variation
between the wing length and width, which compose both our measures of shape and size,
explain how these two features can respond differently to the analyzed factors.
MATERIAL AND METHODS
STRAINS
We used flies from strains previously established by B. C. Bitner-Mathé, D. Tesseroli
and B. F. Menezes (MENEZES et al., 2013). Briefly, a baseline population founded from 135
isofemale lines was used to found independent strains artificially selected for decreasing or
increasing a shape index based on a width-to-length ratio. Selection was applied for 21
consecutive generations and intermittently applied after that with no generation overlap.
Here we use 3 biological replicate strains, independently selected for elongated wings (named
L strains), 3 for rounded shaped wings (R strains) and 2 unselected controls (C strains) from
the 123rd generation. Due to the prolonged and strong selection program, strains are assumed
to be fairly homozygous.
REACTION NORMS
Strains were reared in a thermal gradient (modifed from Fogleman 1978) ranging from
14°C to 30°C with two degrees interval. Five randomly chosen couples were kept in each vial
at 23°C for two days for mating and oviposition in a standard Drosophila medium, after which
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
56
adults were removed and vials placed in the thermal gradient. Emerged adults aging 13-19
days were stored in 80% ethanol and their wings were mounted on a microscope slide and
photographed with a digital camera (Optronics DEI-750, 1.3 megapixels, software Axion Vision
3.0) attached to a stereoscope microscope (Zeiss Stemi SV11). Left wing measurements were
taken for up to 43 females and 40 males per strain and temperature, for a total of 1136 wings.
Due to viability issues of these strains, balanced analyses were not possible. All emerged flies
with undamaged wing were included.
Reaction norms were described both by linear (f(T)= a + bT) and quadratic (f(T) a + bT
+ cT2) models, where f is the trait of interest (wing shape or size) and T is the temperature.
Quadratic was considered the best-fit model when most of the reaction norms curvatures (c)
were significantly different from zero (ROCHA, F.; MEDEIROS; KLACZKO, 2009; ZAR 2010;
ROCHA; KLACZKO 2012 for a better description of the procedure). For traits with c not
significantly different from zero, a linear regression on developmental temperature was
performed and the regression slope (b) used as the reaction norm parameter. b can be
interpreted as a plasticity index, where a value approaching zero means canalization while
either negative and positive values indicate an increase in plasticity with opposite phenotypic
response to the environmental condition. Traits for which the curvature (c) was significantly
different from zero were then described by a quadratic regression with c used as the reaction
norm descriptor. A c approaching zero means that RN is essentially linear, whereas a positive
c means an upward parabola and a negative a downward one.
WING SHAPE AND SIZE DESCRIPTORS
Wing length (WL) was measured as the distance between the alula opening and the
distal edge of the 3rd longitudinal vein (L3) and wing width (WW) as the line from the 5th
longitudinal vein (L5) to the top of the wing, approximately perpendicular to WL. Shape and
size estimates followed the Ellipse method where wing shape (WSH) wass calculated as the
ratio WW/WL and wing size (WSI) estimated as √(Ww/2) (WL/2 ) (Klaczko & Bitner-Mathé,
1990; Klaczko, 2006). See supporting material Fig. S1 for the fitting of the estimated ellipse
drawn by these parameters onto illustrative wings. Measurements of WW and WL were taken
on the program ImageJ 1.47v (http://rsbweb.nih.gov/ij/). The artificial selection program was
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
57
based on the WW/WL ratio and in these strains more variation is expected to be found along
these axes.
All statistical analyses were conducted on SYSTAT© v.13.0 (SPSS Inc.).
RESULTS
Fig. 1 shows wing shape (WSH) and size (WSI) variation between flies from artificially
selected strains (L and R) in the 123rd generation. The high divergence for shape and the
similarity for size between the selected strains are evidenced by the histograms, which
includes all data, regardless of sex or developmental temperature. Mean values, standard
errors and group sample sizes by strain, sex and temperature can be found at supporting
information Table S1. The effects of direction of selection (L × C × R), sex, temperature, their
interactions and biological replicates nested within direction of selection were tested by an
analysis of variance (ANOVA) on Table 1. Direction of selection (SEL) is the major source of
variation of WSH, encompassing 78% of the total variation of WSH and 8% among biological
replicates within each SEL. For wing size (WSI), the main factor of variation was Temperature
(59%), followed by Sex (24%). No significant difference of WSI was detected among SEL,
showing that despite the large divergence in WSH, wing size was similar among directions of
selections.
Linear and quadratic models were used to describe the reaction norms along the
temperature gradient of the artificially selected strains. Linear (b) and quadratic coefficients
(c) with standard errors are exhibited on Table S2. For WSH, c is not significantly different from
zero in 10 out of the 16 groups (8 strains with 2 sexes), indicating that WSH reaction norm is
predominantly linear; except for L strains, for which the quadratic model best describes WSH
variation. For WSI, c is significant for most of the RNs; hence, the quadratic model was used.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
58
%SS
Source df WSH WSI
SEL 2 77.51 ** 1.36 Temperature (T) 8 0.45 59.08 * Sex 1 2.43 23.79 ** SEL x T 16 2.03 1.35 SEL x Sex 2 0.13 0.08 Sex x T 8 0.13 0.07 SEL x Sex x T 16 0.25 0.24 Replicates (SEL) 5 7.96 *** 3.78 *** Error 1077 9.11 10.25
Level of significance: *P<0.05; **P<0.0l; ***P<0.00l. df: degrees of freedom.
Figure 1. Histograms of wing shape (WSH; left) and size (WSI; right), including males and females from the artificially selected strains at all temperatures, showing great shape variation while size distributions overlap. Illustrative wings of Elongated strains (L) and Round Strains (R) exhibited. Control unselected strains not included for clarity.
Table 1: ANOVA for wing shape (WSH) and size (WSI) testing the effect of directions of selections (SEL: L x C x R), sex, temperatures, their interactions and the effect of biological replicates nested within direction of selection (Replicates (SEL)). F value is estimated using the Replicates (SEL) effect as the error term. The table shows the percentage of the total variance explained by each effect.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
59
Fig. 2 shows the adjustment of the best-fit model for the wing traits by direction of selection and sex along the temperature gradient; for WSH in L strain, both adjustments are presented. Wings from R strains become more elongated as temperature increases; conversely, C strains become rounder and, for L strains, the linear regression suggests a WSH homogeneity (with a b not significantly different from zero), but the quadratic adjustment shows a slight increase for this trait at extreme temperatures. Concerning WSI, the influence of temperature variation is more evident, with values decreasing as the temperature increases forming a downward parabola.
Therefore, we used the linear regression coefficient to describe the reaction norm of
WSH (b_WSH) and the quadratic coefficient to describe the nonlinear reaction norms of WSI,
(c_WSI). The extent of the effect of direction of selection (SEL) on the reaction norms and was
assessed by the ANOVA on Table 2. The effects of SEL, Sex, their interactions and biological
replicates nested within SEL were tested. SEL is responsible for 81% of the variation of b_WSH,
showing that divergence in wing shape among SEL was accompanied by modification in the
plastic response of WSH. Differences among biological replicates was assessed by the nested
Figure 2. Temperature reaction norms of wing shape (WSH) and wing size (WSI) for artificially selected strains (Round – R; Elongated – L; unselected control – C). Symbols of sex and letters representing the strains highlight important sources of variation within the data.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
60
effect of Replicates(SEL), which was nonsignificant for b_WSH, indicating the homogeneity of
reaction norms among strains selected for the same direction (nonetheless, F values were still
calculated using the Replicates(SEL) as the error term for a more conservative analysis). A
post-hoc pairwise comparison for b_WSH (Tukey’s Honestly-Significant-Difference Test) was
then performed for the SEL effect. Directions of selection were significantly different from
each other (L × R, L × C and R × C: p<0.000002 – Table S3). No effect for Sex or for SEL × Sex
interaction was found.
The relationship between the phenotypic value of wing shape and its reaction norm
(b_WSH) is shown in Fig. 3. A quadratic regression was performed since data is nonlinear, with
intermediate phenotypes (C Strains) exhibiting extreme values of b_WSH. WSHmean explained
36% of the variation of b_WSH, thus corroborating the above results, in that selecting for
different wing shapes systematically changed the wing shape reaction norm.
%SS
Source df b_WSH c_WSI
SEL 2 80.6 ** 23.2
Sex 1 1.3 17.8 SEL x Sex 2 1.0 0.7 Replicates (SEL) 5 13.0 55.6 ** Error 5 4.0 2.7 Level of significance: *P<0.05; **P<0.0l; ***P<0.00l. df: degrees of freedom.
Table 2: ANOVA for wing traits temperature-reaction norms parameters, testing the effect of directions of selections (SEL), sex, their interactions and biological replicates nested within direction of selection (Replicates (SEL)). Reaction norms are described by the angular coefficient of wing shape (b_WSH) and by the curvature of the quadratic polynomials of wing size (c_WSI). F value is estimated using the Replicates (SEL) as the error term. The table shows the percentage of the total variance explained by each effect.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
61
Differences of wing size reaction norms were nonsignificant (effect of SEL on c_WSI).
Effects of Sex and SEL × Sex interaction were also not significant. Biological replicates held
most of the variation of c_WSI (56%). Overall, our results reveal a high independence pattern
of variation between WSI and WSH (and their reaction norm parameters). This contrasting
response to the experimental factors is surprising because both variables were extracted from
the same biological axes (WW and WL) and one might expect more interference between them.
To analyze how these axes responded to size-imposing sources of variation once the shape
values had been established by selection. In order to allow variance in size while attempting
to maintain a somewhat stable shape, a compensatory pattern between WL and WW is
expected and values of the latter were predicted following the equation (Wwpredicted=
(MeanWSH)(WL)). Wwpredicted represents the expected values in case the mean shape of each
strain were to be kept unchanged despite variation of size in WL. Fig. 4 shows the line
representing the relations between values of Wwpredicted and real WL. WW is scattered around
Figure 3. Relationship of b_WSH (reaction norm parameter for wing shape) on the total phenotypic mean of each strain and sex (total mean calculated on the pool of individuals from all temperatures by strains and sex). The adjusted quadratic regression explains 36% of reaction norms variation.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
62
the prediction line, showing that real values tend to follow the prediction equation. These
results indicate that once wing shape is established for a particular strain, variation in size is
proportional and affects both biological axes with the same intensity. This proportional
variation explains how wings exhibit great variance in size across temperatures while not
imposing variation of the same order of magnitude to wing shape.
Figure 4. Proportional variation of WW and WL. Thick line represents the predicted values of wing width (WWpredicted) in a function of the real values of wing length (WL) that allows variation of size while preserving the mean WSH of each direction of selection. WWpredicted = (MeanWSH)*WL. Note that real values of WW are distributed around the prediction line (thick line) showing a proportional growth between axes that allows variance in wing size while maintaining a stable WSH. Mean WSH for L=0.431; C=0.474 and R=0.517. Females on the left and males on the right. Thin long line represents the linear regression of WW on WL.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
63
DISCUSSION
The artificially selected strains used in this work exhibit great divergence in wing shape,
while their size is remarkably similar. When these strains were submitted to a large
temperature gradient, each direction of selection exhibited a different wing shape reaction
norm and a similar one for size. Control unselected strains exhibit a shape RN where wings
become rounder as temperature increased, a pattern also found in closely related species
(LOH et al., 2008; MATTA; BITNER-MATHÉ, 2004). However, the intense selection program led
rounded winged flies (R strains) to develop an opposite trend, with wings getting more
elongated as temperature increases. Elongated winged flies (L strains), despite their apparent
strong canalization when analyzed by the linear model (i.e. they exhibit a b=0), evolved a
quadratic RN, with flies reared at the extreme temperatures developing rounder wings. This
change in RN is consistent within the independently selected strains within each direction of
selection. On the other hand, for wing size reaction norms, curvature of the RNs was not
affected by the selection program and are similar to size-related RNs described for many
different populations of Drosophila species (DAVID et al., 1994; MORIN et al., 1999; PÉTAVY
et al., 1997).
The association between the mean value of a trait and its plastic response has long
been debated, with a tendency to assume their independence (see Rocha et al. 2009 for a
brief review on this issue). The discussion gained new insights from an empirical study
demonstrating that reaction norms of abdominal pigmentation in D. mediopunctata are highly
dependent on the phenotypic mean (ROCHA, F.; MEDEIROS; KLACZKO, 2009) adding up to a
few plant studies pointing in the same direction (ELBERSE et al., 2004; KLIEBENSTEIN; FIGUTH;
MITCHELL-OLDS, 2002; STINCHCOMBE; DORN; SCHMITT, 2004). Our results also empirically
show evidence supporting the dependency point-of-view, presenting evidence that wing
shape mean value and its plasticity are, at least partially, dependent, with 36% of the variation
of shape RN explained by the phenotypic mean of wing shape. Part of the genetic basis
associated with wing shape seems to be tightly coupled with the trait capability to present
variation at different developmental temperatures, favoring an allelic sensitivity scenario in
opposition to exclusive plasticity regulatory genes. Genetic linkage is unlikely to explain these
results since wing shape is a complex trait influenced by several genes and one would not
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
64
expect the consistency found here among reaction norms of independently selected strains
within each phenotypic group.
The slope (b) of linear reaction norms is an index of plasticity, with b=0 meaning
canalization and values departing from zero, an increase in plasticity (ROCHA; KLACZKO,
2012,2014). Small changes in reaction norms have profound impacts on phenotypes and on
trait evolution. The average function provided by the linear regressions (from which slopes b
were extracted and used in the ANOVA) for R strains was f(T)=-0.0008(T)+0.54, hence, the
predicted phenotype for R strains at 30ºC is WSH=0.516 (which is close to the real mean of WSH
at 30ºC). For C strains was f(T)=0.0016(T)+0.44. We then computed what would be the
phenotype of R strains if the reaction norm slope had not deviated from the C strains. Hence,
we computed f(30)=0.0016(30)+0.54=0.588 (C strain slope and R strain constant) and found a
difference of almost 14% in wing shape had the slopes of R strains been the same as C strains.
Since b is not commonly presented in papers, comparison with other reaction norms is not
straightforward. Therefore, we visually estimated b from published shape reaction norms. Loh
et al. (2008) analyzed natural populations of Z. indianus, using a similar wing shape descriptor.
The rough estimate was b=0.0003, close to the order of magnitude observed here (Table S2).
Even when we analyzed wing shape RN that used different shape descriptors, we found b with
similar magnitudes. For (AZEVEDO et al., 1998), we estimated a b=0.002 and in (TROTTA et al.,
2010), using the Procrustes generalized least square procedure on a set of landmarks taken
on the wing, we estimated b=-0.001 for the most plastic strain. The order of magnitude of
wing shape RN seems to be similar and small, despite the wing descriptor. On the other hand,
although wing size RN is similar amongst all strains regardless of the direction of selection
applied, it displays more plasticity than shape. For size, the average b=-5.7, an order of
magnitude 4 times higher than for shape RN. This is compelling evidence that wing shape
variation is less plastic than size, presumably because an optimum shape must be more
regulated than size.
Less plasticity of wing shape in contrast to size was also reported in D. mediopunctata
(BITNER‐MATHÉ; KLACZKO 1999c) and by Torquato et al.(2014) using the strains produced by
the same artificial selection program in the 64th generation. (DEBAT; DEBELLE; DWORKIN,
2009) proposed that conservation of wing shape in Drosophila genus is likely to be associated
with stabilizing selection since many studies (HOULE et al., 2003; PÉLABON et al., 2006;
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
65
WEBER, 1992,1990), including ours, have shown an intense response to shape selection;
therefore it is not the lack of genetic variation nor any constraint in development that prevents
more divergent shapes in nature. The association between shape RN and the phenotypic mean
might contribute to the stabilization of wing shape at intermediate optimum temperatures.
Because of the intense selection program, flies still exhibit great divergence in wing shape, but
smaller genetic divergences might be optimized by different RN associated to them, stabilizing
the phenotype at optimum temperatures.
The contrasting response of wing size and shape to temperature is especially
impressive because both WSH and WSI are composed by the same biological axes (WL, proxy of
proximodistal, and WW, proxy of anteroposterior, developmental axes) and one might expect
more interferences between them. In any biological structure, variation of size and shape
must be orchestrated by developmental axes. The predominance of anteroposterior,
dorsoventral and proximodistal axes specifying tissue organization during development
(NIEHRS, 2010) is an indication that morphological variation should be, whenever possible,
described in terms of those axes because those variation can then be comprehended in terms
of its ontogeny. In the wing of Drosophila, the developmental genetics underlying the
establishment of biological axes have been well described (CIFUENTES; GARCÍA-BELLIDO,
1997; STRIGINI; COHEN, 1999). But it remains unclear how these axes cope with shape and
size variation to produce a phenotype with a size that usually correlates with total body size
and yet preserving a functional shape.
The strains in this work were obtained by an artificial selection program focused on the
ratio WW/WL to generate changes in the shape of the wing, unintentionally preserving overall
wing size. Once the divergent shape was established among C, L and R strains, size-imposing
sources of variation, such as developmental temperatures, affected both axes with similar
intensity and direction, promoting a proportional response that allowed massive size variation
across the temperatures while not imposing great disturbance on wing shape. This
coordinated pattern suggests that genes involved in variation of size are more likely to be
those whose expression domain is on the entire pouch of the wing imaginal disc, rather than
those with restricted territories. A broad expression might explain the proportional response
of the biological axes, both in the composition of size and shape.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
66
Evolution of plasticity is a property hard to manipulate and study. The large
temperature gradient used here provided a thorough description of wing shape and size
reaction norms. Two main properties previously observed remained unaltered despite great
shape divergence: low plasticity of wing shape and a high plasticity of wing size. On the other
hand, divergent wing shape led to different reaction norms, revealing that genetic variation
for wing shape plasticity is influenced by the mean value.
ACKNOWLEDGEMENTS
We thank Danielle Tesseroli and Bianca Menezes for the establishment and maintenance of
the artificial selection strains, and Dulcinea da Rocha for technical assistance. We also thank
Professor Jean David for helpful comments on an early draft of the manuscript and valuable
suggestions from the two anonymous reviewers. This paper is part of the D. Sc. requirements
of Daniel Mattos at the Biodiversity and Evolutionary Biology Graduate Program of the Federal
University of Rio de Janeiro and was supported by Coordenação de Aperfeiçoamento de
Pessoal de Nível Superior (CAPES; graduate scholarship of D. Mattos), Conselho Nacional de
Desenvolvimento Científico e Tecnológico (CNPq; B.C.B-M.: #485332/2007-8; L.B.K:
#312292/2009-0 and #312066/2014-7), Fundação de Amparo à Pesquisa do Rio de Janeiro
(FAPERJ; B.C.B-M.: #E26/171.314/2008) and Fundação de Amparo à Pesquisa de São Paulo
(FAPESP; L.B.K.: #2012/03144-0 and #2013/04980-0).
REFERENCES
AZEVEDO, R. B. R. et al. Latitudinal variation of wing: thorax size ratio and wing-aspect ratio in
Drosophila. Evolution, v. 52, p. 1353–1362, 1998.
BITNER-MATHÉ, B. C.; KLACZKO, L. B. Heritability , phenotypic and genetic correlations of size
and shape of Drosophila mediopunctata wings. Heredity, v. 83, p. 688–696, 1999a.
BITNER-MATHÉ, B. C.; KLACZKO, L. B. Plasticity of Drosophila melanogaster wing morphology:
effects of sex, temperature and density. Genetica, v. 105, n. 2, p. 203–210, 1999b.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
67
BITNER-MATHÉ, B. C.; KLACZKO, L. B. Size and shape heritability in natural populations of
Drosophila mediopunctata: temporal and microgeographical variation. Genetica, v. 105, n. 1,
p. 35–42, 1999c.
BLAIR, S. S. Wing vein patterning in Drosophila and the analysis of intercellular signaling.
Annual review of cell and developmental biology, v. 23, p. 293–319, 2007.
CIFUENTES, F. J.; GARCÍA-BELLIDO, A. Proximo-distal specification in the wing disc of
Drosophila by the nubbin gene. Proceedings of the National Academy of Sciences of the
United States of America, v. 94, n. 21, p. 11405–11410, 1997.
DAVID, J. R. et al. Isofemale lines in Drosophila: an empirical approach to quantitative trait
analysis in natural populations. Heredity, v. 94, n. 1, p. 3–12, 2005.
DAVID, J. R. et al. Reaction norms of size characters in relation to growth temperature in
Drosophila melanogaster: an isofemale lines analysis. Genetics, selection, evolution, v. 26, n.
3, p. 229–251, 1994.
DAVID, J. R. et al. Thermal phenotypic plasticity of body size in Drosophila melanogaster:
sexual dimorphism and genetic correlations. Journal of Genetics, v. 90, n. 2, p. 295–302, 2011.
DEBAT, V.; DEBELLE, A.; DWORKIN, I. Plasticity, canalization, and developmental stability of
the Drosophila wing: joint effects of mutations and developmental temperature. Evolution, v.
63, n. 11, p. 2864–2876, 2009.
ELBERSE, I. A. M. et al. Quantitative trait loci affecting growth-related traits in wild barley
(Hordeum spontaneum) grown under different levels of nutrient supply. Heredity, v. 93, p.
22–33, 2004.
FOGLEMAN, J. C. A thermal gradient bar for the study of Drosophila. Drosophila Information
Service, v. 53, p. 212–213, 1978.
GIRALDEZ, A J.; COHEN, S. M. Wingless and Notch signaling provide cell survival cues and
control cell proliferation during wing development. Development, v. 130, n. 26, p. 6533–6543,
2003.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
68
HOULE, D. et al. Automated measurement of Drosophila wings. Bmc Evolutionary Biology, v.
3, p. 1–13, 2003.
KARAN, D.; MORETEAU, B.; DAVID, J. R. Growth temperature and reaction norms of
morphometrical traits in a tropical drosophilid: Zaprionus indianus. Heredity, v. 83 ( Pt 4), p.
398–407, 1999.
KLACZKO, L. B. Evolutionary genetics of Drosophila mediopunctata. Genetica, v. 126, n. 1-2, p.
43–55, 2006.
KLACZKO, L. B.; BITNER-MATHÉ, B. C. On the edge of a wing. Nature, v. 346, p. 231, 1990.
KLIEBENSTEIN, D. J.; FIGUTH, A.; MITCHELL-OLDS, T. Genetic architecture of plastic methyl
jasmonate responses in Arabidopsis thaliana. Genetics, v. 161, p. 1685–1696, 2002.
LOH, R. et al. Adaptation to different climates results in divergent phenotypic plasticity of wing
size and shape in an invasive drosophilid. Journal of genetics, v. 87, n. 3, p. 209–17, 2008.
MATTA, B. P.; BITNER-MATHÉ, B. C. Genetic architecture of wing morphology in Drosophila
simulans and an analysis of temperature effects on genetic parameter estimates. Heredity, v.
93, n. 4, p. 330–41, 2004.
MENEZES, B. F. et al. The influence of male wing shape on mating success in Drosophila
melanogaster. Animal Behaviour, v. 85, n. 6, p. 1217–1223, 2013.
MORIN, J. P. et al. Divergence of reaction norms of size characters between tropical and
temperate populations of Drosophila melanogaster and D. simulans. Journal of Evolutionary
Biology, v. 12, p. 329–339, 1999.
NEUFELD, T. P. et al. Coordination of growth and cell division in the Drosophila wing. Cell, v.
93, n. 7, p. 1183–1193, 1998.
NIEHRS, C. On growth and form: a Cartesian coordinate system of Wnt and BMP signaling
specifies bilaterian body axes. Development, v. 137, p. 845–857, 2010.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
69
PÉLABON, C. et al. Response of fluctuating and directional asymmetry to selection on wing
shape in Drosophila melanogaster. Journal of Evolutionary Biology, v. 19, p. 764–776, 2006.
PÉTAVY, G. et al. Growth temperature and phenotypic plasticity in two Drosophila sibling
species : probable adaptive changes in flight capacities. Journal of Evolutionary Biology v. 10,
p. 875–887, 1997.
PIGLIUCCI, M. Evolution of phenotypic plasticity: where are we going now? Trends in ecology
& evolution, v. 20, n. 9, p. 481–486, 2005.
ROCHA, F. B.; KLACZKO, L. B. Connecting the dots of nonlinear reaction norms unravels the
threads of genotype-environment interaction in Drosophila. Evolution, v. 66, n. 11, p. 3404–
3416, 2012.
ROCHA, F.; MEDEIROS, H. F.; KLACZKO, L. B. The reaction norm for abdominal pigmentation
and its curve in Drosophila mediopunctata depend on the mean phenotypic value. Evolution,
v. 63, n. 1, p. 280–287, 2009.
ROCHA, Felipe Bastos; KLACZKO, Louis Bernard. Undesirable consequences of neglecting
nonlinearity: Response to comments by liefting et al. (2013) on Rocha & Klaczko (2012).
Evolution, 2014.
SCHEINER, S. M. Plasticity as a selectable trait: reply to Via. American Naturalist, v. 142, n. 2,
p. 371–373, 1993.
STINCHCOMBE, J. R.; DORN, L. A.; SCHMITT, J. Flowering time plasticity in Arabidopsis thaliana:
A reanalysis of Westerman & Lawrence (1970). Journal of Evolutionary Biology, v. 17, p. 197–
207, 2004.
STRIGINI, M.; COHEN, S. M. Formation of morphogen gradients in the Drosophila wing.
Seminars in cell developmental biology, v. 10, n. 3, p. 335–344, 1999.
TORQUATO, L. et al. Cellular basis of morphological variation and temperature-related
plasticity in Drosophila melanogaster strains with divergent wing shapes. Genetica, p. 1–11,
2014.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
70
TROTTA, V. et al. Thermal plasticity of wing size and shape in Drosophila melanogaster, D.
simulans and their hybrids. Climate Research, v. 43, p. 71–79, 2010.
VIA, S. et al. Adaptive phenotypic plasticity: consensus and controversy. Trends in ecology &
evolution, v. 10, n. 5, p. 212–217, 1995.
VIA, S. Adaptive Phenotypic Plasticity: Target or By-Product of Selection in a Variable
Environment? The American Naturalist, v. 142, n. 2, p. 352–365, 1993.
WEBER, K. E. How small are the smallest selectable domains of form? Genetics, v. 130, p. 345–
353, 1992.
WEBER, K. E. Selection on wing allometry in Drosophila melanogaster. Genetics, v. 126, n. 4,
p. 975–989, 1990.
WEST-EBERHARD, M. J. Developmental plasticity and evolution. New York: Oxford University
Press, 2003. v. 424.
ZAR, J. H. Biostatistical Analysis. 4th. ed. New Jersey: Prentice Hall, 2010.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
71
SUPPORTING INFORMATION
All Temperatures 14°C 16°C 18°C
Fem
ales
R
WSH 0.511 ± 0.001 ( 201 ) 0.509 ± 0.006 ( 14 ) 0.526 ± 0.003 ( 15 ) 0.520 ± 0.004 ( 12 )
WSI 404.5 ± 2.256 ( 201 ) 438.5 ± 4.351 ( 14 ) 450.1 ± 2.585 ( 15 ) 444.2 ± 5.538 ( 12 )
WL 1132 ± 6.234 ( 201 ) 1231 ± 17.33 ( 14 ) 1241 ± 8.322 ( 15 ) 1232 ± 18.33 ( 12 )
WW 579 ± 3.456 ( 201 ) 625 ± 4.610 ( 14 ) 653 ± 3.923 ( 15 ) 641 ± 7.172 ( 12 )
L
WSH 0.420 ± 0.001 ( 201 ) 0.428 ± 0.003 ( 14 ) 0.419 ± 0.003 ( 30 ) 0.420 ± 0.002 ( 22 )
WSI 402.5 ± 2.007 ( 201 ) 420.9 ± 2.394 ( 14 ) 428.1 ± 2.180 ( 30 ) 433.8 ± 1.996 ( 22 )
WL 1242 ± 6.438 ( 201 ) 1287 ± 8.390 ( 14 ) 1323 ± 6.643 ( 30 ) 1339 ± 6.714 ( 22 )
WW 522 ± 2.646 ( 201 ) 551 ± 3.661 ( 14 ) 554 ± 3.781 ( 30 ) 562 ± 3.164 ( 22 )
C
WSH 0.467 ± 0.001 ( 162 ) 0.457 ± 0.002 ( 21 ) 0.455 ± 0.003 ( 19 ) 0.466 ± 0.004 ( 15 )
WSI 410.6 ± 2.978 ( 162 ) 445.9 ± 2.763 ( 21 ) 456.2 ± 2.524 ( 19 ) 432.9 ± 8.075 ( 15 )
WL 1203 ± 9.371 ( 162 ) 1320 ± 9.683 ( 21 ) 1352 ± 6.975 ( 19 ) 1268 ± 21.93 ( 15 )
WW 561 ± 3.886 ( 162 ) 603 ± 3.555 ( 21 ) 616 ± 4.645 ( 19 ) 591 ± 12.33 ( 15 )
Mal
es
R
WSH 0.523 ± 0.001 ( 201 ) 0.527 ± 0.006 ( 12 ) 0.532 ± 0.004 ( 11 ) 0.531 ± 0.006 ( 13 )
WSI 360.1 ± 2.085 ( 201 ) 395.4 ± 3.072 ( 12 ) 403.4 ± 2.974 ( 11 ) 396.8 ± 3.199 ( 13 )
WL 996 ± 5.631 ( 201 ) 1090 ± 12.40 ( 12 ) 1106 ± 8.305 ( 11 ) 1089 ± 11.53 ( 13 )
WW 521 ± 3.237 ( 201 ) 574 ± 4.289 ( 12 ) 588 ± 5.373 ( 11 ) 578 ± 5.024 ( 13 )
L
WSH 0.440 ± 0.001 ( 205 ) 0.444 ± 0.002 ( 24 ) 0.440 ± 0.003 ( 29 ) 0.436 ± 0.004 ( 24 )
WSI 362.7 ± 2.145 ( 205 ) 390.1 ± 2.730 ( 24 ) 391.1 ± 1.902 ( 29 ) 389.0 ± 1.918 ( 24 )
WL 1094 ± 6.631 ( 205 ) 1171 ± 7.061 ( 24 ) 1179 ± 5.836 ( 29 ) 1179 ± 6.940 ( 24 )
WW 481 ± 2.883 ( 205 ) 520 ± 4.512 ( 24 ) 519 ± 3.250 ( 29 ) 514 ± 3.551 ( 24 )
C
WSH 0.480 ± 0.001 ( 166 ) 0.464 ± 0.003 ( 9 ) 0.479 ± 0.003 ( 22 ) 0.483 ± 0.005 ( 10 )
WSI 367.4 ± 2.494 ( 166 ) 398.1 ± 1.520 ( 9 ) 409.4 ± 3.263 ( 22 ) 395.5 ± 5.683 ( 10 )
WL 1062 ± 7.573 ( 166 ) 1169 ± 7.097 ( 9 ) 1183 ± 9.236 ( 22 ) 1139 ± 20.37 ( 10 )
WW 509 ± 3.399 ( 166 ) 542 ± 2.318 ( 9 ) 567 ± 5.236 ( 22 ) 549 ± 6.603 ( 10 )
Table S1: Mean ± Standard error estimated using the pool of all individuals by direction of selection, sex and temperature (sample size in parenthesis). Round strains (R), Elongated
Strains (L) and Control unselected Strains (C). Wing shape (WSH), wing size (WSI), wing length (WL) and wing width (WW). Units of WSI, WL and WW are pixels (px). 1px=0,0017mm. WSH is a
dimensionless ratio.
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
72
20°C 22°C 24°C 26°C
Fem
ales
R
WSH 0.506 ± 0.004 ( 24 ) 0.515 ± 0.004 ( 27 ) 0.509 ± 0.004 ( 34 ) 0.509 ± 0.004 ( 26 )
WSI 427.4 ± 2.438 ( 24 ) 408.9 ± 2.437 ( 27 ) 406.7 ± 2.902 ( 34 ) 389.7 ± 1.896 ( 26 )
WL 1202 ± 7.603 ( 24 ) 1140 ± 7.505 ( 27 ) 1141 ± 8.042 ( 34 ) 1093 ± 6.111 ( 26 )
WW 608 ± 4.417 ( 24 ) 587 ± 4.247 ( 27 ) 580 ± 5.211 ( 34 ) 556 ± 3.711 ( 26 )
L
WSH 0.420 ± 0.002 ( 26 ) 0.415 ± 0.002 ( 23 ) 0.415 ± 0.003 ( 43 ) 0.419 ± 0.005 ( 13 )
WSI 417.4 ± 3.146 ( 26 ) 409.6 ± 1.362 ( 23 ) 388.3 ± 1.844 ( 43 ) 381.8 ± 3.421 ( 13 )
WL 1288 ± 7.817 ( 26 ) 1272 ± 5.119 ( 23 ) 1206 ± 7.642 ( 43 ) 1179 ± 8.538 ( 13 )
WW 541 ± 5.110 ( 26 ) 528 ± 2.471 ( 23 ) 500 ± 2.453 ( 43 ) 495 ± 6.416 ( 13 )
C
WSH 0.465 ± 0.002 ( 28 ) 0.467 ± 0.003 ( 9 ) 0.484 ± 0.004 ( 18 ) 0.471 ± 0.002 ( 18 )
WSI 427.4 ± 4.961 ( 28 ) 420.8 ± 5.919 ( 9 ) 376.1 ± 3.483 ( 18 ) 395.0 ± 1.746 ( 18 )
WL 1253 ± 14.76 ( 28 ) 1232 ± 19.52 ( 9 ) 1082 ± 12.77 ( 18 ) 1151 ± 5.200 ( 18 )
WW 583 ± 6.870 ( 28 ) 575 ± 7.374 ( 9 ) 523 ± 4.426 ( 18 ) 542 ± 2.654 ( 18 )
Mal
es
R
WSH 0.517 ± 0.003 ( 24 ) 0.523 ± 0.003 ( 33 ) 0.528 ± 0.003 ( 35 ) 0.516 ± 0.004 ( 29 )
WSI 381.5 ± 2.446 ( 24 ) 364.9 ± 2.172 ( 33 ) 364.1 ± 2.409 ( 35 ) 345.2 ± 2.526 ( 29 )
WL 1062 ± 6.265 ( 24 ) 1010 ± 6.094 ( 33 ) 1003 ± 7.070 ( 35 ) 961 ± 6.139 ( 29 )
WW 549 ± 4.487 ( 24 ) 528 ± 3.944 ( 33 ) 529 ± 3.996 ( 35 ) 496 ± 4.815 ( 29 )
L
WSH 0.436 ± 0.002 ( 33 ) 0.438 ± 0.003 ( 14 ) 0.437 ± 0.003 ( 37 ) 0.449 ± 0.004 ( 12 )
WSI 372.8 ± 1.863 ( 33 ) 368.6 ± 3.382 ( 14 ) 346.6 ± 1.894 ( 37 ) 341.6 ± 2.309 ( 12 )
WL 1129 ± 5.237 ( 33 ) 1114 ± 11.58 ( 14 ) 1050 ± 6.993 ( 37 ) 1020 ± 7.445 ( 12 )
WW 492 ± 3.103 ( 33 ) 488 ± 4.350 ( 14 ) 458 ± 2.820 ( 37 ) 458 ± 3.987 ( 12 )
C
WSH 0.477 ± 0.002 ( 40 ) 0.504 ± 0.008 ( 2 ) 0.481 ± 0.003 ( 36 ) 0.486 ± 0.004 ( 19 )
WSI 381.9 ± 2.623 ( 40 ) 379.6 ± 4.843 ( 2 ) 351.3 ± 3.269 ( 36 ) 355.8 ± 1.754 ( 19 )
WL 1106 ± 8.397 ( 40 ) 1069 ± 5.245 ( 2 ) 1014 ± 11.24 ( 36 ) 1021 ± 4.652 ( 19 )
WW 527 ± 3.615 ( 40 ) 539 ± 11.11 ( 2 ) 487 ± 4.082 ( 36 ) 496 ± 3.738 ( 19 )
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
73
28°C 30°C
Fem
ales
R
WSH 0.512 ± 0.004 ( 27 ) 0.504 ± 0.005 ( 22 )
WSI 367.1 ± 2.043 ( 27 ) 360.1 ± 3.858 ( 22 )
WL 1026 ± 7.37 ( 27 ) 1014 ± 8.85 ( 22 )
WW 525 ± 3.60 ( 27 ) 512 ± 7.14 ( 22 )
L
WSH 0.430 ± 0.003 ( 21 ) 0.429 ± 0.006 ( 9 )
WSI 363.1 ± 2.950 ( 21 ) 339.7 ± 2.728 ( 9 )
WL 1108 ± 10.65 ( 21 ) 1038 ± 8.92 ( 9 )
WW 476 ± 3.87 ( 21 ) 445 ± 5.41 ( 9 )
C
WSH 0.455 ± 0.005 ( 16 ) 0.483 ± 0.004 ( 18 )
WSI 363.3 ± 2.348 ( 16 ) 363.2 ± 1.752 ( 18 )
WL 1077 ± 7.08 ( 16 ) 1045 ± 5.74 ( 18 )
WW 490 ± 4.75 ( 16 ) 505 ± 3.49 ( 18 )
Mal
es
R
WSH 0.526 ± 0.004 ( 29 ) 0.512 ± 0.005 ( 15 )
WSI 323.6 ± 2.235 ( 29 ) 313.6 ± 4.124 ( 15 )
WL 893 ± 5.90 ( 29 ) 877 ± 10.46 ( 15 )
WW 470 ± 4.28 ( 29 ) 449 ± 7.08 ( 15 )
L
WSH 0.447 ± 0.002 ( 23 ) 0.441 ± 0.005 ( 9 )
WSI 313.9 ± 3.290 ( 23 ) 301.4 ± 4.095 ( 9 )
WL 939 ± 10.17 ( 23 ) 908 ± 10.31 ( 9 )
WW 420 ± 4.53 ( 23 ) 400 ± 6.93 ( 9 )
C
WSH 0.471 ± 0.004 ( 14 ) 0.492 ± 0.004 ( 14 )
WSI 323.6 ± 1.782 ( 14 ) 319.9 ± 3.588 ( 14 )
WL 943 ± 4.73 ( 14 ) 913 ± 10.30 ( 14 )
WW 444 ± 3.58 ( 14 ) 449 ± 5.56 ( 14 )
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
74
Trait Direction
of Selection
Strain Sex Linear Regression Quadratic Regression
b SE c SE
Win
g Sh
ape
(WSH
)
R Strains
1R F -0.0009 0.0003 ** -0.0001 0.0001
Negative Linear RN
1R M -0.0007 0.0003 * -0.0002 0.0001 *
5R F -0.0006 0.0003 * 0.0000 0.0001
5R M -0.0004 0.0004 0.0001 0.0001
6R F -0.0009 0.0004 * 0.0000 0.0001
6R M -0.0012 0.0003 *** 0.0000 0.0001
C Strains
6C F 0.0011 0.0002 *** 0.0000 0.0001
Positive Linear RN
6C M 0.0009 0.0003 ** 0.0001 0.0001
1C F 0.0026 0.0007 *** 0.0006 0.0002 **
1C M 0.0017 0.0006 ** 0.0003 0.0002
L Strains
1L F -0.0001 0.0003 0.0002 0.0001 *
Quadratic RN
1L M -0.0004 0.0003 0.0001 0.0001
2L F -0.0003 0.0003 0.0002 0.0001 **
2L M 0.0001 0.0004 0.0004 0.0001 ***
5L F 0.0011 0.0002 *** 0.0001 0.0001 *
5L M 0.0001 0.0002 0.0001 0.0001
Table S2: Linear (b) and quadratic (c) coefficients with standard error (SE) from the respective regressions for each
artificially selected strains used as the reaction norm parameter in all analyses. Note that RN of both R and C
Strains are predominantly linear with inverse signs for the linear coefficient while L Strains exhibit quadratic RN.
Females (F) and males (M). 10/16 reaction norms of WSH are significantly linear
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
75
Table S2 continuation.
Trait Direction
of Selection
Strain Sex Linear Regression Quadratic Regression
b SE
SE
Win
g Si
ze (
WSI
)
R Strains
1R F -5.1903 0.4209 *** -0.3704 0.0797 ***
1R M -5.9272 0.3268 *** -0.3151 0.0584 ***
5R F -4.5799 0.4170 *** -0.1674 0.0777 *
5R M -3.3777 0.4996 -0.1053 0.0935
6R F -6.8319 0.2932 *** -0.2159 0.0560 ***
6R M -6.7955 0.2427 *** -0.1327 0.0513 *
C Strains
6C F -6.3435 0.2393 *** -0.4030 0.0473 ***
6C M -6.2545 0.2687 *** -0.3456 0.0528 ***
1C F -6.0621 0.8873 -0.3062 0.2691
1C M -7.2668 0.6247 -0.2335 0.2019
L Strains
1L F -5.0341 0.4237 *** -0.3158 0.0768 ***
1L M -6.4387 0.3979 * -0.1857 0.0778 *
2L F -4.5394 0.3637 *** -0.2823 0.0775 ***
2L M -3.7554 0.3139 *** -0.2507 0.0662 ***
5L F -6.5762 0.3451 *** -0.4120 0.0798 ***
5L M -6.8940 0.2659 *** -0.2834 0.0639 ***
C A P Í T U L O I I : P L A S T I C I D A D E F E N O T Í P I C A
76
Selection Selection Difference p-Value 95% Confidence Interval
Lower Upper
C L 0.0457 < 10 -5 0.04338 0.04802
C R -0.0452 < 10 -5 -0.04756 -0.0429
L R -0.0909 < 10 -5 -0.09313 -0.0887
Table S3: Results of the Pairwise comparison of the Selection effect of ANOVA
presented on Table 2. Tukey’s Honestly-Significance-Difference Test.
C A P Í T U L O I I I : B A S E S C E L U L A R E S
77
Cellular basis of morphological variation and temperature-related plasticity in Drosophila
melanogaster strains with divergent wing shapes
Libéria Souza Torquato; Daniel Mattos; Bruna Palma Matta; Blanche Christine Bitner-Mathé
Published in Genetica (2014) 142:495-505 DOI 10.1007/s10709-014-9795-0
Authors’ affiliations:
Laboratório de Evolução de Caracteres Complexos – Drosophila, Departamento de Genética,
Instituto de Biologia, Universidade Federal do Rio de Janeiro – Brasil
Running title: cellular basis of wing shape and size variation
C A P Í T U L O I I I : B A S E S C E L U L A R E S
78
ABSTRACT
Organ shape evolves through cross-generational changes in developmental patterns at
cellular and/or tissue levels that ultimately alter tissue dimensions and final adult proportions.
Here, we investigated the cellular basis of an artificially selected divergence in the outline
shape of Drosophila melanogaster wings, by comparing flies with elongated or rounded wing
shapes but with remarkably similar wing sizes. We also tested whether cellular plasticity in
response to developmental temperature was altered by such selection. Results show that
variation in cellular traits is associated with wing shape differences, and that cell number may
play an important role in wing shape response to selection. Regarding the effects of
developmental temperature, a size-related plastic response was observed, in that flies reared
at 16ºC developed larger wings with larger and more numerous cells across all intervein
regions relative to flies reared at 25ºC. Nevertheless, no conclusive indication of altered
phenotypic plasticity was found between selection strains for any wing or cellular trait. We
also described how cell area is distributed across different intervein regions. It follows that
cell area tends to decrease along the anterior wing compartment and increase along the
posterior one. Remarkably, such pattern was observed not only in the selected strains but also
in the natural baseline population, suggesting that it might be canalized during development
and was not altered by the intense program of artificial selection for divergent wing shapes.
Keywords: artificial selection; morphological evolution; phenotypic plasticity; size; wing
development
C A P Í T U L O I I I : B A S E S C E L U L A R E S
79
INTRODUCTION
For organ shape to evolve, phenotypic variation in tissue dimensions and final adult
proportions must be altered across generations. Natural selection can promote such changes,
by acting on the heritable counterpart of phenotypic variation for the target trait, which may
reflect the existence of heritable variation in subordinate traits at lower levels of biological
organization (GARLAND; KELLY, 2006). For instance, when variation in organ shape is selected
for, effective response to selection should be expected to occur through evolutionary changes
in cellular parameters during tissue development, such as cell size, shape, number and/or
organization (LECUIT; LE GOFF, 2007). Another predicted outcome of natural or artificial
selection, not often discussed in the literature (GARLAND; KELLY, 2006), is that average
phenotypic plasticity for the trait under selection (or for subordinate traits) could also be
affected, in a correlated response to selection on the target trait. Therefore, investigating the
cellular basis of evolution and plasticity in organ morphology is crucial to the understanding
of which cellular traits and developmental processes are able to respond to selective and
environmental factors that affect organ development.
Describing and isolating size and shape components of organ morphology is one
essential step in studies of morphological evolution. The Drosophila wing has been widely
used as a model for such studies, given some experimental advantages to other organs like:
adult wing is essentially bidimensional, which facilitates the characterization and
interpretation of its morphological variation (KLACZKO; BITNER-MATHÉ, 1990; KLINGENBERG,
2002), and wing development is relatively well known (BLAIR, 2007; GARCIA-BELLIDO; DE
CELIS, 1992; NETO-SILVA; WELLS; JOHNSTON, 2009). And since each wing cell produces a
single trichome (DOBZHANSKY, 1929), estimation of cell number and cell area can be
performed through trichome density in regions of adult wing. In natural populations, genetic
basis for latitudinal and altitudinal variation in wing shape traits has been reported; the
geographical variation being generally non-linear (BITNER-MATHÉ; PEIXOTO; KLACZKO, 1995;
GILCHRIST et al., 2000; HOFFMANN; SHIRRIFFS, 2002; IMASHEVA et al., 1995; LOH; BITNER-
MATHÉ, 2005; PITCHERS; POOL; DWORKIN, 2013). But the adaptive nature of wing shape is
difficult to be evidenced. One successful demonstration was made by Menezes et al. (2013),
using replicate strains of D. melanogaster that were artificially selected for increasing the
quantitative divergence in the outline shape of the wings. These authors observed higher
C A P Í T U L O I I I : B A S E S C E L U L A R E S
80
mating success for males from strains with elongated wings and found no consistent evidence
that general size (wing or thorax size) is related to such success. In fact, a large amount of
additive genetic variation for the outline shape of the wing has been observed, even in species
from different Drosophila subgenera (BITNER-MATHÉ; KLACZKO, 1999a; MATTA; BITNER-
MATHÉ, 2004), over which natural and artificial might act. So, overall, wing shape should be
considered as a meaningful component of reproductive fitness in Drosophila and might be a
direct target for selection in nature.
Despite the different methods used to estimate morphological variation, the adaptive
nature of wing size has been extensively reported, at least as a proxy for body size (ARENDT,
2007; KLEPSATEL et al., 2014; NIJHOUT, 2003). Ectotherms generally follow Bergmann’s rule,
in that animals from higher latitudes tend to have larger bodies and body parts than those
from lower latitudes. This clinal variation has been commonly considered as an adaptive
response to the latitudinal variation in temperature, but contradicting results indicate that
other selective pressures should also be acting. In a thorough review, Arendt (2007) discusses
the patterns of cellular alterations that underlie wing size variation, especially in D.
melanogaster. As reported by the author, plastic change of wing size in response to rearing
temperature is consistently due to changes in cell size: flies that develop at lower
temperatures have larger wings as a result of increased cell size, with little or no contribution
of cell number. A similar pattern is observed in thermal selection experiments. When reared
at different temperatures for extended periods of time, cold-adapted flies show an evolved
(genetic) response to temperature that is mostly due to larger cells in their (larger) wings. On
the other hand, the evolved response to natural latitudinal variation generally follows an
opposite trend: when flies from different populations are reared under similar conditions,
high-latitude (larger) wings show a consistent increase in cell number, with little or no
contribution of cell size, relative to low-latitude (smaller) wings (ARENDT, 2007; KLEPSATEL et
al., 2014). In this context, therefore, the means to be larger depends on which environmental
and/or evolutionary factors are acting on the correlated subordinate mechanisms that affect
cellular alterations during development.
However, it remains unclear which cellular processes are involved with evolution of
wing shape. One way to find associations between variation in cellular traits and differences
in organ shape would be through comparison of populations or strains that significantly
C A P Í T U L O I I I : B A S E S C E L U L A R E S
81
diverge in organ shape but have similar organ sizes, so that confounding effects of size
components could be largely minimized without the need of any scaling transformation. The
D. melanogaster strains used by Menezes et al. (2013) present such features. These strains
were previously established in our laboratory, by B. C. Bitner-Mathé, D. Tesseroli and B. F.
Menezes (unpublished data), through artificial selection for increasing or decreasing the
values of a wing shape index, which resulted in flies with elongated (L strains) or rounded (R
strains) wings. It is noteworthy that variances in wing outline shape of L and R strains do not
show any overlap after selection (MENEZES et al., 2013). Moreover, despite the shape
divergence, these strains have evolved similar wing sizes (MATTA, BRUNA P; BITNER-MATHÉ;
ALVES-FERREIRA, 2011; MENEZES et al., 2013). So, here we investigated the cellular basis
involved in wing morphological variation in these strains, as well as a possible plastic response
to developmental temperature of wing features and cellular traits. Moreover, we tested for
any experimentally evolved change in temperature-related plasticity, which could have arisen
as a byproduct of selection on wing outline shape.
MATERIALS AND METHODS
FLIES (STRAINS WITH ARTIFICIALLY SELECTED WING SHAPES)
Detailed information on the program of bidirectional artificial selection that generated
the fly strains can be found in Menezes et al. (2013). Briefly, wild-caught D. melanogaster
females were collected (Rio de Janeiro: 22º95’52”S / 43º19’68”W) and were individually set
in vials with cornmeal sucrose medium at 16ºC, resulting in 135 isofemale lines. After eclosion,
males and virgin-females were used as the baseline laboratory population for the selection
program, which was henceforth performed at 22ºC. Using the width-to-length ratio as an
index of wing shape (WSH – see next section), bidirectional selection was simultaneously
applied to four independent biological replicates. As a result, four strains with elongated (L)
wings were established through selection for decreasing WSH (L strains), while four strains with
rounded (R) wings were set by selection for increasing WSH (R strains). Selection was
performed at every generation until the 21st. Subsequent selection was applied intermittently,
but extra care was always taken to not overlap any generation.
C A P Í T U L O I I I : B A S E S C E L U L A R E S
82
EXPERIMENTAL DESIGN
This work was performed at generation 64 (G64), for which the last preceding selection
was made at generation 62. For comparison, we also used the wild-caught females (Wild) and
their laboratory daughters (G1) reared at 16ºC. All flies were fixed in ethanol 70%.
Given the labor intensive requirements of trichome counting, only females from two
biological replicates were analyzed in each direction of selection: 1L and 5L (L strains), 1R and
5R (R strains); a sample of each was transferred to 25ºC two generations prior the following
experiment. For each selection strain, ten random pairs of G64 flies were placed into bottles
containing 40mL of cornmeal sucrose medium for oviposition. Egg density was controlled by
transferring parental flies to new bottles, daily, during six days. To allow oviposition at
different temperatures of development (TD), bottles from days 1, 3 and 5 were maintained at
25°C while bottles from days 2, 4 and 6 were maintained at 16°C. Hence, three batches of
siblings raised in each TD were obtained for each selection strain (1L, 5L, 1R and 5R). Up to
five females from each batch had their left wings mounted on microscope slides.
WING MORPHOMETRICS
Wing length (WL) was estimated as the distance between the alula opening (AO) and
the distal edge of the 3rd longitudinal vein (L3), while wing width (WW) was estimated as a
straight line approximately perpendicular to WL, using the posterior edge of the 5th
longitudinal vein (L5) as a landmark (Fig. 1a).
Klaczko and Bitner-Mathé (1990) demonstrated that the outline shape of Drosophila
wing can be geometrically described by the fitting of an ellipse to semi-landmarks taken on
the wing contour, thereby allowing an estimation of its outline shape (named SH) through the
ratio between minor and major ellipse radii. Wing size (named SI) is then estimated by the
geometric mean between both ellipse radii, which corresponds to the radius of a circle with
the same ellipse area. Because no scaling transformation is applied (see also Klaczko 2006),
any information obtained with the ellipse adjustment can be traced back to the original
landmarks taken from the wing, making biological inferences straightforward. Here, a proxy
of such ellipse indexes were estimated: wing shape (named WSH) is given by the width-to-
length ratio (WW/WL); while wing size (named WSI) is given by the geometric mean between
WW/2 and WL/2, and has the same measurement unit of WW and WL. By superposing the
C A P Í T U L O I I I : B A S E S C E L U L A R E S
83
estimated ellipse over the original wing image (see Fig. S1), we show that the ellipse proxy can
be considered a good approximation to the ellipse adjustment, and that both of these
methods capture most of the outline shape and size of each original wing. So wings with bigger
SH or WSH ratios are rounder because they have larger width and/or smaller length; the
contrary is true for more elongated wings.
We acknowledged that methods of geometric morphometrics are broadly used and
have remarkable statistical power for estimating size and shape components of wing
morphology, especially non-allometric components in different aspects of wing shape (see
KLINGENBERG, CHRISTIAN PETER, 2002; PITCHERS; POOL; DWORKIN, 2013). In this work,
however, we have used the WSH proxy mainly because: (1) it produces a straightforward
description of wing outline shape without the need of any scaling transformation (as shown
in Fig. S1); (2) the fly strains used here were generated through selection on WSH ratios, and
so should vary the most along WW and WL measurements.
Regarding internal aspects of the wing, five intervein regions (IVR) named A-E were
analyzed by estimating their sizes through polygonal surface areas, with no scaling applied.
More specifically, each IVR area was estimated as the area of a polygon delimited by
landmarks at intersections of wing veins, as well as by 5 to 10 semi-landmarks on curved
regions; see the respective traced areas in Fig. 1a.
Whole wing images were taken at 36X magnification using digital camera (Optronics
DEI-750, 1.3 megapixels, software Axion Vision 3.0) attached to a stereoscope microscope
(Zeiss Stemi SV11). WW and WL were measured using ImageJ v.1.46r
(http://rsbweb.nih.gov/ij/) and were scaled to millimeters. The polygonal surface area of each
IVR (in mm²) was estimated using ImageJ’s polygon tool.
Raw data will be available from Dryad Digital Repository (http://datadryad.org/).
C A P Í T U L O I I I : B A S E S C E L U L A R E S
84
CELL SIZE (AREA) AND CELL NUMBER
Images of fixed-size rectangles within each intervein region (IVR) at dorsal wing surface
(Fig. 1b) were captured with above mentioned digital camera (and software) on a optic
microscope (Zeiss Stemi Axioskop 2) and used for trichome counting (1000X magnification;
fixed rectangle area of 0.00906 mm²). Trichome counting was performed using the cell
counter plugin v.2010 on ImageJ v.1.46r; all visible trichomes were counted. Given that cell
density might not be homogeneous throughout the wing, we aimed at taking each rectangle
image in equivalent positions for all analyzed wings. A proxy for average cell area (CA) in each
IVR was estimated through the ratio between the fixed surface area of rectangle image
(0.00906 mm2) and the respective number of counted trichomes (Fig. 1b). We also
investigated how variation in cell area is distributed across the wing surface and whether this
distribution was affected by response to artificial selection and/or temperature variation
during the development (details in next section). In turn, cell number (CN) in each IVR was
given by the ratio between the estimated IVR area (traced areas in Fig. 1a) and the respective
CA. We note that a thorough estimation of CA and CN across the wing blade was performed
by analyzing five different intervein regions, instead of using only one or two regions as
Figure 1. Morphometric methods presented in a wild-type female wing of Drosophila melanogaster. (a) Wing showing wing length (WL), wing width (WW) and the area of each intervein region (IVR) A-E (traced lines). Within each IVR, rectangles indicate approximate locations where all trichomes were counted. (b) Magnification used for image taking and trichome counting at each rectangle region shown in (a); each wing cell has a single characteristic trichome (Colour figure online)
C A P Í T U L O I I I : B A S E S C E L U L A R E S
85
commonly performed in the literature. Total cell number at dorsal wing surface (CNTotal)
was calculated through the ratio between the estimates of whole wing area (π * WL/2
* WW/2; a proxy to the ellipse area) and average cell area across all intervein regions ( ).
STATISTICAL ANALYSES AND DISTRIBUTION OF CELL AREA ACROSS INTERVEIN REGIONS
Two-way nested ANOVAs were performed to test the homogeneity between the
following fixed factors: direction of selection (SEL: R×L strains), temperature of development
(TD: 16ºC×25oC), interactions between these factors (SEL×TD), plus the nested effects of
biological replicates within such interaction (SEL×TD{REP}), in order to enhance statistical
power. Whenever the SEL×TD{REP} effect was significant, it was used as the error term for the
remaining effects (for details on nested ANOVA, see SOKAL; ROHLF, 1981). Almost all traits
(including WSH ratios) were normally distributed (Kolmogorov-Smirnov test: all P > 0.00625;
alfa from Bonferroni correction), except for CAD in 1R strain TD 16ºC (P = 0.00260); so no data
transformation was applied.
The distribution of cell area across the intervein regions of each wing was estimated
through the following polynomial regression: Y = g0 + g1X + g2X²; where Y is CA (dependent
variable), X is the rank order of intervein region (independent variable), and coefficients are
g0 (intercept), g1 (slope) and g2 (quadratic). To do so, a rank order was attributed to each
intervein region according to their proximity from the most anterior region of the wing: IVRA
was ranked as 1, IVRB as 2 and so forth. The following polynomial parameter and characteristic
values were then used as variables that describe cell area distribution: the shape of the curve
(g2), minimum value (MV) and intervein region of minimum value (IVR-MV); for details see
David et al. (1997).
All statistical analyses were performed using SYSTAT© v.13.0 (SPSS Inc.).
RESULTS
Mean values of wing traits (WSH, WSI, WW, WL), plus total cell number (CNTotal) and
average cell area (CAAverage) at dorsal wing surface are shown in Fig. 2, for the artificially
selected strains with rounded (R) or elongated (L) wings (values per biological replicates are
presented in Table S1). ANOVA results for each trait are reported in Table 1. Regarding overall
CA
C A P Í T U L O I I I : B A S E S C E L U L A R E S
86
shape divergence, significant effects of direction of selection (SEL: R×L strains) were found for
WSH, WW and WL. It follows that wings from R flies are rounder (bigger WSH ratios) than wings
from L flies, and this difference in outline shape is accounted by significant differences in both
width and length: R flies have wider (bigger WW) and shorter (smaller WL) wings than L flies.
No difference in wing size associated to direction of selection was found (non-significant SEL
effect for WSI). This result suggests that SEL effects can be considered a shape-related source
of variation, with no detectable influence on wing size. As for the cellular traits in whole dorsal
surface, we found that wings from R flies have significantly more cells (increased CNTotal) than
wings from L flies (Fig. 2 and Table 1), but no significant SEL effect was detected for average
cell area (CAAverage).
The effect of temperature of development (TD: 16ºC×25oC) was significant for all size-
related traits, but not for WSH (Table 1). Wings are on average bigger and present larger width
and length when flies are raised at 16ºC (Fig. 2 and Table S1), a pattern repeatedly described
in the literature. Both cell number and cell area account for such temperature-related
variation in size: bigger wings result from bigger and more numerous cells when flies are
reared at 16ºC. Moreover, we did not detect any significant SEL×TD effect. Therefore, we have
no indication of variation in the plasticity of any trait between the strains selected for
divergent wing shapes.
C A P Í T U L O I I I : B A S E S C E L U L A R E S
87
For comparison purposes, mean values of wild-caught females (Wild) and their
daughters (G1) are also presented in Fig. 2 (and detailed in Table S2). We note that, due to
logistics in our laboratory at the time of the experiment, G1 flies had to be maintained at 16oC,
but all subsequent generations of the selection program were maintained at 22oC. It is clear
that average WSH values have diverged from Wild and G1 flies, increasing in R strains and
decreasing in L strains. Differences in size-related (WSI, WW and WL) and cellular traits (CNTotal
and CAAverage) can also be observed between baseline population (Wild and G1) and selection
strains from G64. Nevertheless, it is interesting to note that, when reared at similar conditions
(16oC), mean values of size-related and cellular traits in R and L strains tend to approach those
of G1 16oC-reared flies; a pattern not observed for WSH.
Figure 2. Mean values and standard-errors of wing traits according to temperature of development (16oC or 25oC) in females from L strains (dashed line) and R strains (dotted line), sampled at the 64th generation of artificial selection. Mean values and standard-errors for wild-caught females (Wild) and their laboratory daughters (G1), which founded the selection strains, are also presented. Abbreviations: wing outline shape (WSH), wing size (WSI, in mm), wing width (WW, in mm) and length (WL, in mm), plus total cell number (CNTotal) and average cell area (CAAverage, in mm2 × 10-4) at dorsal wing blade
C A P Í T U L O I I I : B A S E S C E L U L A R E S
88
To further explore the cellular basis of temperature-related plasticity subjacent to the
morphological variation between R and L strains, intervein regions (A-E) were individually
analyzed. Fig. 3 presents mean and standard error for each intervein area (IVR), cell area (CA)
and cell number (CN) (detailed in Table S1 and S2). ANOVA in Table 1 shows that IVRA was
significantly different between selection strains. It follows that flies from R strains have bigger
IVRA than flies from L strains. This morphological variation seem to result from increased cell
number: CNA is significantly bigger in R strains relative to L strains, but no significant change
between R and L strains was detected for cell area at intervein region A (CAA). In addition,
significant but opposite SEL differences were found for CAE and CNE, which might explain the
non-significant morphological difference in the area of IVRE. That is, cells at region E of R wings
are on average smaller but more numerous relative to the same region in L wings, so that the
area on intervein region E does not significantly differ between R and L wings. No other
significant SEL difference was found for any IVR, CA or CN estimates. Regarding the effects of
temperature, all intervein regions were significantly bigger at 16ºC with larger and more
numerous cells, relative to flies reared at 25oC (except for CNB; see TD effects in Table 1).
Table 1: ANOVA of each morphological and cellular trait in D. melanogaster strains with divergent wing shapes.
C A P Í T U L O I I I : B A S E S C E L U L A R E S
89
ANOVA – Mean Squares
Trait SEL TD SEL×TD SEL×TD{REP} error
WSH 17.511** 0.014 0.121 0.324*** 0.008
WSI 0.403 12.423*** 0.078 0.117* 0.041
WW 26.487** 23.214** 0.965 0.400** 0.090
WL 47.423* 108.701** 0.173 2.862*** 0.368
IVRA 1.096*** 2.333*** 0.033 0.043 0.027
IVRB 0.796 6.316** 0.112 0.139** 0.036
IVRC 0.121 10.968*** 0.028 0.041 0.048
IVRD 0.016 8.067*** 0.051 0.077* 0.023
IVRE 0.071 18.215*** 0.003 0.033 0.075
CAA 0.021 86.694*** 10.629 1.642 3.143
CAB 0.659 190.978** 6.487 8.291** 2.295
CAC 0.984 37.632** 0.085 7.466 3.488
CAD 1.610 98.665*** 4.652 1.655 1.936
CAE 39.184** 149.136*** 8.767 1.799 5.091
CAAverage 3.593 105.720*** 4.522 1.959 1.149
CNA 253,995.430*** 158,029.631*** 2,716.003 3,756.846 9,969.001
CNB 298,286.299 92,097.881 4,536.593 78,961.420** 17,081.196
CNC 2,111.319 1,610,532.739*** 20,533.579 39,795.842 33,675.012
CND 29,304.266 765,368.141*** 3,711.323 15,841.698 10,978.311
CNE 816,289.880** 850,056.972** 131,497.232 60,167.240 82,163.968
CNTotal 5,629,535.408*** 14,228,061.620*** 110,600.228 551,419.905 278,865.760
Wing traits: outline shape (WSH), size (WSI), width (WW), length (WL), and the area of intervein regions A-E (IVRA-
E). Cellular traits: average cell area (CA) or cell number (CN) at intervein regions A-E (subscripts), plus the average cell area across all intervein regions (CAAverage) and total cell number (CNTotal) at dorsal wing blade. Units: WSH is a ratio; WSI, WW and WL are in millimeters; and all CA estimates are in mm2 ×10-4. ANOVA model: direction of selection (SEL: L strains × R strains; df=1), temperature of development (TD: 16oC × 25oC; df=1), SEL×TD interaction (df=1), nested effects of biological replicates (SEL×TD{REP}; df=4), and model error (df=72); whenever SEL×TD{REP} effect was significant, it was used as error term for the remaining effects. N=10 for each group (1L, 5L, 1R and 5R strains reared at each TD). Mean squares (MS) are presented ×10-2, except for cell number estimates. ***P ≤ 0.001; **P ≤ 0.01; *P ≤ 0.05.
C A P Í T U L O I I I : B A S E S C E L U L A R E S
90
Figure 3 Mean values and standard-errors of wing traits per intervein region (A-E) according to temperature of development (16oC or 25oC), in females from L strains (dashed line) and R strains (dotted line) that were sampled at the 64th generation of artificial selection. Mean values and standard errors for wild-caught females (Wild) and their laboratory daughters (G1), which founded the selection strains, are also presented. Abbreviations: area of each intervein region (IRVA-E, in mm2), cell number (CNA-E) and cell area (CAA-E, in mm2 × 10-4)
C A P Í T U L O I I I : B A S E S C E L U L A R E S
91
To investigate how variations in cell area are distributed across the wing surface, each
CA mean value was plotted against the respective IVR rank order (Fig. 4). It is clear that average
cell area tend to decrease from intervein regions A-C and to increase from regions D-E, despite
the direction of selection or temperature at which flies were raised. Interestingly, a similar
pattern was observed in wings of Wild and G1 flies, which were never subjected to artificial
selection. This result demonstrates that such distribution was already present in wings from
the natural baseline population and was not altered by the intense program of artificial
selection. Adjustment of the quadratic regression was significant for all experimental groups
and explains from 18% to 38% of cell area variation across the wing blade. No differences
between L and R strains, or between developmental temperatures, were found for the
parameters that describe each polynomial regression (Table S3), except for minimum value of
16oC × 25oC-reared flies. But this is an expected result, since wings of 16ºC-reared flies tend
to have larger cells across the whole wing surface, regardless the selection strain.
C A P Í T U L O I I I : B A S E S C E L U L A R E S
92
DISCUSSION
In this study, we investigated the cellular basis of an artificially selected divergence on
the outline shape of D. melanogaster wings, as well as of temperature-related plasticity in
these selected strains. Confounding effects of wing size on wing shape estimates are
commonly observed in natural populations and do not allow a direct association between
cellular variation and wing shape differences. One advantage of the selection strains used in
our work is that, despite their wide divergence in shape, wing size is remarkably similar. So, in
this specific case, most of the phenotypic variation in the wing contour is directly related to
variation in shape, even without scaling transformations. This is particularly interesting since,
when no scaling is applied, cellular variation can be directly related to the morphological
variation that is being captured. We were thus able to isolate cellular variation involved with
shape-related effects of artificial selection on wing outline shape from cellular variation
associated with size-related effects of plastic response to developmental temperature; with
minimum or no interaction between such effects.
Figure 4. Quadratic curves fitted to average cell area (CAAverage, in mm2 × 10-4) across the ranked intervein regions for the complete data set of each experimental group; rank order was attributed respecting their proximity from the most anterior region of the wing: IVRA was ranked as 1, IVRB as 2 and so forth. Legend of experimental groups also presents the percentage of variation explained (R2) by the respective polynomial adjustment (Colour figure online)
C A P Í T U L O I I I : B A S E S C E L U L A R E S
93
In respect to the cellular basis of artificially selected divergence in outline shape, our
results show a significant increase in total cell number of rounded wings relative to elongated
wings. It seems that cell number may play an important role in wing shape response to
selection. Regarding the relative contribution of intervein regions, a significant shape-related
effect of direction of selection was detected only for IRVA and CNA. Although not statistically
significant, intervein region B also seems to exhibit a small divergence between R and L strains,
regarding both IVR area (IRVB) and cell number CNB (see Fig. 3). Hence, we found that response
to selection on wing shape was primarily associated to changes in the anterior region of the
wing (A and possibly B). Such result is not in agreement with previous studies that identified
a more prominent response of posterior and distal regions in relation to different aspects of
wing shape variation in Drosophila (Gilchrist et al. 2000, Pezzoli et al. 1997). This discrepancy
might be related to the fact that, here, wing outline shape was the direct target of selection;
while in natural populations it is not straightforward to isolate the amount of wing shape
variation that actually results from direct selection on wing shape (especially from optimizing
selection; see Gilchrist et al. 2000). But we note that the variation in the outline shape of the
wing might not be completely independent from variation in the positioning of some wing
veins. For instance, genetic correlation between outline shape and the positioning of second,
fourth and fifth longitudinal veins has already been found, even in different Drosophila species
(BITNER-MATHÉ; KLACZKO, 1999; MATTA; BITNER-MATHÉ, 2004). Given that IRVA and IRVB are
both associated to the second longitudinal vein, it is possible that the positioning of such wing
vein have also changed as a correlated response to selection on outline shape. But a thorough
description of correlated responses from different size and shape traits to the selection on
outline shape still needs to be performed.
Regardless the wing compartments, studies on the cellular basis of natural adaptation
and artificial selection associated with Drosophila wing morphology have also found a primary
role of cell number, for both shape and size-related traits. In experiments of artificial selection
for body size (reviewed in ARENDT, 2007), flies selected to be larger tend to have larger wings
with increased number of cells than control flies (NOACH; DE JONG; SCHARLOO, 1997;
PARTRIDGE et al., 1999). The contrary was found for flies selected to be smaller, although a
large effect of (reduced) cell size was also observed (GUERRA et al., 1997; PARTRIDGE et al.,
1999). When comparing natural populations along latitudinal clines, several studies have
C A P Í T U L O I I I : B A S E S C E L U L A R E S
94
reported that flies from higher latitudes tend to have larger wings than those from lower
latitudes, and this wing size difference is mainly explained by an increase in cell number, with
small or no contribution of cell size (ARENDT, 2007; KLEPSATEL et al. 2014). Although it is
commonly assumed that adaptation to temperature is the most important factor for the
establishment of latitudinal clines, thermal selection experiments have shown a contradicting
result: the wing size increase observed in laboratory cold-adapted flies largely result from an
increase in cell size rather than cell number (ARENDT, 2007). Therefore, it is reasonable to
assume that other selective pressures (apart from developmental temperature) should be
acting in latitudinal clines of wing size, such as balancing selection for optimal wing shape
(GILCHRIST et al., 2000). Regarding the cellular basis of wing shape adaptation in nature,
Pezzoli et al. (1997) examined temperate and subtropical populations of D. melanogaster and
found that significant differences in components of wing shape were mainly explained by
variation in cell number. But given the differences in wing size, the exact contribution of cell
number and cell area to such differences in wing shape components could not be estimated.
Our results add to this discussion, since we showed that morphological responses to a direct
selection on wing outline shape were achieved mainly through changes in cell number. So it
appears that, when caused by direct selection on wing morphology (for shape or size-related
traits) or by selective pressures in latitudinal clines other than developmental temperature,
evolutionary changes in wing size and shape might be essentially achieved through
subordinate changes in cell number.
As for the size-related plastic response to developmental temperature, we observed
that flies reared at the lowest temperature developed larger wings with bigger and more
numerous cells across the whole wing surface. Moreover, WSI was high and similarly
correlated with both cell size and number (Pearson’s correlation using whole dataset: r =
0.753, P < 0.0001 for WSI×CNTotal; and r = 0.724, P < 0.0001 for WSI×CAAverage). Remarkably, we
did not detect a correlation between CNtotal and CAaverage (r = 0.093; P = 1.000). This suggest
that developmental processes involved in establishment of final organ size might be capable
of regulating both cell number and size but with some independency between them. In fact,
it has been repeatedly demonstrated that cell size has a major role in the plasticity of
Drosophila wing size to developmental temperature (reviewed in ARENDT, 2007). But the
importance of variation in cell number to such temperature-related plasticity is not clear, and
C A P Í T U L O I I I : B A S E S C E L U L A R E S
95
has only been reported as a conditional sex-effect (ARENDT, 2007; NOACH et al. 1997). So
further investigation is needed in order to improve the validity of such findings; ideally using
a greater number of biological replicates, different selection populations and gradients of
developmental temperature.
Regarding the distribution of cell area across the different intervein regions, we also
observed an interesting counterclockwise-like pattern: cell area tends to decrease along the
anterior compartment of the wing (from IVRA to IVRC) and increase along the posterior wing
compartment (from IVRD to IVRE), despite the direction of selection or temperature in which
flies were raised. Remarkably, this pattern of distribution was observed not only for L and R
strains, but also for the wild-caught baseline population and its G1 daughters reared under
laboratory conditions, which were never subjected to artificial selection. This result indicates
that the quadratic pattern of cell area distribution might be canalized during development and
did not change in face of developmental temperature or artificial selection for rounded or
elongated wings (Fig. 4). We note that this is the first time in which such pattern is thoroughly
analyzed, but evidence for its existence in natural populations can be obtained by plotting
results from Pezzoli et al. (1997). So what developmental processes could generate such cell
area distribution across the wing surface? In wing imaginal discs, the location of presumptive
vein and intervein regions along the anteroposterior axis are determined during late third
instar wing disc development (BIER, 2000; BLAIR, 2007; RESTREPO; ZARTMAN; BASLER, 2014).
From this stage on, presumptive intervein regions are recognized as developmental subunits
that largely restrict cell proliferation within their boundaries (GARCIA-BELLIDO; MERRIAM,
1971; GONZÁLEZ-GAITÁN; CAPDEVILA; GARCIA-BELLIDO, 1994). So it is tempting to argue that
genes activated or repressed downstream of the signaling pathways that pattern presumptive
vein and intervein regions (such as Decapentaplegic signaling pathway) could also be
responsible for the final definition of cell area at each intervein region. It is also possible that
such pattern might be achieved or refined during hinge contraction along the proximodistal
axis, in the development of Drosophila pupal wing (OLGUIN; MLODZIK, 2010).
One major goal of experiments that perform artificial selection is gaining insight into
evolutionary patterns that might also occur in nature, not only for the focal trait, but also for
subordinate or correlated traits (FULLER; BAER; TRAVIS, 2005). Expected outcomes of natural
selection, like possible changes in average phenotypic plasticity due to directional selection
C A P Í T U L O I I I : B A S E S C E L U L A R E S
96
(GARLAND; KELLY, 2006), can also be tested in such experiments. In our study, no conclusive
indication of altered phenotypic plasticity due to wing shape selection was found, given that
no significant SEL×TD interaction was observed for any wing or cellular trait. Nevertheless,
reaction norms can be quite complex and linearity cannot be assumed (DAVID et al., 1997;
ROCHA; KLACZKO, 2012). So, possible changes in phenotypic plasticity due to the intense
artificial selection are yet to be thoroughly tested, particularly on a wider and more refined
gradient of developmental temperature.
Investigating the cellular basis of wing shape divergence and associated changes in
plasticity might be a difficult task, especially because of confounding wing size effects. But this
is a worthwhile quest, since the evolution of organ shape is a fascinating subject in
evolutionary developmental biology. In this study, we were able to directly relate cellular
variation to wing shape differences and show that cell number may have an important role in
response to selection on wing shape.
ACKNOWLEDGMENTS
We thank Danielle Tesseroli and Bianca Menezes for the establishment and
maintenance of the artificial selection strains, and Dulcinea da Rocha for technical assistance.
We also thank Louis Bernard Klaczko and reviewers for helpful comments. This study is part
of the thesis research of L.M.S.T. in pursuit of her Ph.D. in Genetics at the Genetics Department
of Universidade Federal do Rio de Janeiro (UFRJ), and was supported by Conselho Nacional de
Desenvolvimento Científico e Tecnológico – CNPq (B.C.B.-M. research project: 485332/2007-
8; L.S.T. graduate scholarship: 140993/2010-0) and by Coordenação de Aperfeiçoamento de
Pessoal de Nível Superior – CAPES (B.P.M. research project and postgraduate scholarship:
23038.0080/2010-97 AUXPE-PNPD 2799/2010; D.M. graduate scholarship).
REFERENCES
ARENDT, J. Ecological correlates of body size in relation to cell size and cell number: patterns
in flies, fish, fruits and foliage. Biological reviews of the Cambridge Philosophical Society, v.
82, n. 2, p. 241–56, 2007.
C A P Í T U L O I I I : B A S E S C E L U L A R E S
97
BIER, E. Drawing lines in the Drosophila wing: Initiation of wing vein development. Current
Opinion in Genetics and Development. v.10, p.393-398 , 2000.
BITNER-MATHÉ, B. C.; KLACZKO, L. B. Heritability , phenotypic and genetic correlations of size
and shape of Drosophila mediopunctata wings. Heredity, v. 83, p. 688–696, 1999.
BITNER-MATHÉ, B. C.; PEIXOTO, A. A.; KLACZKO, L. B. Morphological variation in a natural
population of Drosophila mediopunctata: altitudinal cline, temporal changes and influence of
chromosome inversions. Heredity, v. 75, p. 54–61, 1995.
BLAIR, S. S. Wing vein patterning in Drosophila and the analysis of intercellular signaling.
Annual review of cell and developmental biology, v. 23, p. 293–319, 2007.
DAVID, J. R. et al. Phenotypic plasticity and developmental temperature in Drosophila: analysis
and significance of reaction norms of morphometrical traits. Journal of Thermal Biology, v.
22, n. 6, p. 441–451, 1997.
DOBZHANSKY, T. The influence of the quantity and quality of chromosomal material on the
size of the cells in Drosophila melanogaster. Wilhelm Roux’ Archiv für Entwicklungsmechanik
der Organismen, v. 115, n. 3, p. 363–379, 1929.
FULLER, R. C.; BAER, C. F.; TRAVIS, J. How and When Selection Experiments Might Actually be
Useful. Integrative and Comparative Biology, v. 45, n. 3, p. 391–404, 2005.
GARCIA-BELLIDO, A.; DE CELIS, J. F. Developmental genetics of the venation pattern of
Drosophila. Annual review of genetics, v. 26, p. 277–304, 1992.
GARCIA-BELLIDO, A.; MERRIAM, J. R. Parameters of the wing imaginal disc development of
Drosophila melanogaster. Developmental Biology, v. 24, n. 1, p. 61–87, 1971.
GARLAND, T.; KELLY, S. A. Phenotypic plasticity and experimental evolution. The Journal of
experimental biology, v. 209, n. Pt 12, p. 2344–61, 2006.
GILCHRIST, A. S. et al. Adaptation and constraint in the evolution of Drosophila melanogaster
wing shape. Evolution development, v. 2, n. 2, p. 114–124, 2000.
C A P Í T U L O I I I : B A S E S C E L U L A R E S
98
GONZÁLEZ-GAITÁN, M.; CAPDEVILA, M. P.; GARCIA-BELLIDO, A. Cell proliferation patterns in
the wing imaginal disc of Drosophila. Mechanisms of development, v. 46, n. 3, p. 183–200,
1994.
GUERRA, D. et al. Developmental constraints in the Drosophila wing. Heredity, v. 79, p. 564–
571, 1997.
HOFFMANN, A. A.; JENNIFER SHIRRIFFS. Geographic variation for wing shape in Drosophila serrata. Evolution; international journal of organic evolution, v. 56, n. 5, p. 1068–1073, 2002.
IMASHEVA, A. G. et al. Geographic differentiation in wing shape in Drosophila melanogaster.
Genetica, v. 96, n. 3, p. 303–306, 1995.
KLACZKO, L. B.; BITNER-MATHÉ, B. C. On the edge of a wing. Nature, v. 346, p. 231, 1990.
KLEPSATEL, P. et al. Similarities and differences in altitudinal versus latitudinal variation for
morphological traits in Drosophila melanogaster. Evolution; international journal of organic
evolution, v. 68, n. 5, p. 1385–98, 2014.
KLINGENBERG, C. P. Morphometrics and the role of the phenotype in studies of the evolution
of developmental mechanisms. Gene, v. 287, p. 3–10, 2002.
LECUIT, T.; LE GOFF, L. Orchestrating size and shape during morphogenesis. Nature, v. 450, n.
7167, p. 189–92, 8 2007.
LOH, R.; BITNER-MATHÉ, B. C. Variability of wing size and shape in three populations of a
recent Brazilian invader, Zaprionus indianus (Diptera: Drosophilidae), from different habitats.
Genetica, v. 125, n. 2-3, p. 271–81, 2005.
MATTA, B. P.; BITNER-MATHÉ, B. C. Genetic architecture of wing morphology in Drosophila
simulans and an analysis of temperature effects on genetic parameter estimates. Heredity, v.
93, n. 4, p. 330–41, 2004.
MATTA, B. P.; BITNER-MATHÉ, B. C.; ALVES-FERREIRA, M. Getting real with real-time qPCR: a
case study of reference gene selection for morphological variation in Drosophila melanogaster
wings. Development genes and evolution, v. 221, n. 1, p. 49–57, 2011.
C A P Í T U L O I I I : B A S E S C E L U L A R E S
99
MENEZES, B. F. et al. The influence of male wing shape on mating success in Drosophila
melanogaster. Animal Behaviour, v. 85, n. 6, p. 1217–1223, 2013.
NETO-SILVA, R. M.; WELLS, B. S.; JOHNSTON, L. A. Mechanisms of growth and homeostasis in
the Drosophila wing. Annual Review of Cell and Developmental Biology, v. 25, p. 197–220,
2009.
NIJHOUT, H. F. The control of body size in insects. Developmental Biology, v. 261, n. 1, p. 1–
9, 2003.
NOACH, E. J. K.; DE JONG, G.; SCHARLOO, W. Phenotypic plasticity of wings in selection lines
of Drosophila melanogaster. Heredity, v. 79, p. 1–9, 1997.
OLGUIN, P.; MLODZIK, M. A new spin on planar cell polarity. Cell, v. 142, n. 5, p. 674–6, set.
2010.
PARTRIDGE, L. et al. Correlated responses to selection on body size in Drosophila
melanogaster. Genetical Research, v. 74, n. 1, p. 43–54, 1999.
PEZZOLI, M. C. et al. Developmental constraints and wing shape variation in natural
populations of Drosophila melanogaster. Heredity, v. 79, n. 1996, p. 572–577, 1997.
PITCHERS, W.; POOL, J. E.; DWORKIN, I. Altitudinal clinal variation in wing size and shape in
African Drosophila melanogaster: one cline or many? Evolution; international journal of
organic evolution, v. 67, n. 2, p. 438–52, 2013.
RESTREPO, S.; ZARTMAN, J. J.; BASLER, K. Coordination of Patterning and Growth by the
Morphogen DPP. Current biology, v. 24, n. 6, p. R245–R255, 2014.
ROCHA, F. B.; KLACZKO, L. B. Connecting the dots of nonlinear reaction norms unravels the
threads of genotype-environment interaction in Drosophila. Evolution; international journal
of organic evolution, v. 66, n. 11, p. 3404–16, 2012.
SOKAL, R. R.; ROHLF, F. J. Biometry. 2nd. ed. New York, NY: Freeman, 1981. p. 859
C A P Í T U L O I I I : B A S E S C E L U L A R E S
100
SUPPORTING INFORMATION
Figure S1. Original wings images superposed with respective estimated ellipses, using actual data from ellipse adjustment to semi-landmarks (traced blue lines) and its proxy via wing width (WW) and length (WL) measurements (red solid lines). Presented wings were randomly chosen: first specimen in alphabetical order for each D. melanogaster selection strain (1L, 1R, 5L, 5R). Ellipse adjustment was performed using a least square procedure to solve the ellipse equation for Cartesian coordinates of 20 semi-landmarks taken on the wing contour (details in Klaczko and Bitner-Mathé 1990; and Klaczko 2006; see main text). In all four cases, correlation between x-observed and their y-expected ellipse values was greater than 99%. Given that no scaling transformation is applied, the estimated ellipses were reconstructed from values of minor (b) and major (a) ellipse radii (in mm), each multiplied by 2. In turn, ellipses for the proxy of such adjustment were reconstructed using WW and WL values (in mm). On average, WW estimates 94.5% of the respective 2b value, while WL estimates 98.7% of the respective 2a value; similar values were found in data from other studies (not shown).
C A P Í T U L O I I I : B A S E S C E L U L A R E S
101
Table S1: Mean ± SE for wing morphology and cellular variations at intervein regions (Fig. 1a) by individual strain and temperature of development (TD).
TD 25°C (mean ± SE) TD 16°C (mean ± SE) TD 25°C (mean ± SE) TD 16°C (mean ± SE)
Trait 1L (n=10) 1R (n=10) 1L (n=10) 1R (n=10) 5L (n=10) 5R (n=10) 5L (n=10) 5R (n=10)
WSH 0.445 ± 0.002 0.542 ± 0.001 0.438 ± 0.003 0.564 ± 0.003 0.453 ± 0.002 0.528 ± 0.004 0.439 ± 0.002 0.516 ± 0.004
WSI 0.678 ± 0.003 0.666 ± 0.007 0.745 ± 0.006 0.760 ± 0.005 0.656 ± 0.003 0.684 ± 0.006 0.734 ± 0.011 0.760 ± 0.006
WW 0.904 ± 0.005 0.980 ± 0.010 0.986 ± 0.011 1.141 ± 0.009 0.883 ± 0.005 0.994 ± 0.008 0.973 ± 0.015 1.092 ± 0.009
WL 2.033 ± 0.007 1.809 ± 0.020 2.251 ± 0.017 2.023 ± 0.011 1.949 ± 0.011 1.885 ± 0.021 2.217 ± 0.035 2.118 ± 0.017
IVRA 0.129 ± 0.004 0.140 ± 0.003 0.162 ± 0.008 0.184 ± 0.004 0.128 ± 0.002 0.156 ± 0.003 0.154 ± 0.007 0.188 ± 0.007
IVRB 0.250 ± 0.004 0.240 ± 0.005 0.289 ± 0.008 0.317 ± 0.004 0.227 ± 0.004 0.262 ± 0.004 0.285 ± 0.009 0.312 ± 0.007
IVRC 0.280 ± 0.004 0.259 ± 0.006 0.348 ± 0.008 0.348 ± 0.005 0.274 ± 0.004 0.273 ± 0.006 0.347 ± 0.012 0.338 ± 0.005
IVRD 0.225 ± 0.003 0.210 ± 0.005 0.280 ± 0.006 0.280 ± 0.003 0.216 ± 0.004 0.227 ± 0.005 0.280 ± 0.009 0.296 ± 0.004
IVRE 0.371 ± 0.004 0.373 ± 0.009 0.468 ± 0.011 0.473 ± 0.008 0.359 ± 0.003 0.371 ± 0.007 0.461 ± 0.018 0.465 ± 0.007
CAA 2.084 ± 0.049 1.939 ± 0.073 2.207 ± 0.044 2.234 ± 0.059 2.010 ± 0.036 2.002 ± 0.049 2.158 ± 0.063 2.270 ± 0.067
CAB 1.796 ± 0.035 1.687 ± 0.048 1.901 ± 0.053 2.085 ± 0.079 1.751 ± 0.031 1.710 ± 0.027 2.150 ± 0.046 2.044 ± 0.043
CAC 1.735 ± 0.043 1.630 ± 0.071 1.851 ± 0.036 1.846 ± 0.058 1.773 ± 0.036 1.848 ± 0.058 1.945 ± 0.071 1.892 ± 0.081
CAD 1.772 ± 0.042 1.618 ± 0.046 1.901 ± 0.067 1.941 ± 0.045 1.690 ± 0.025 1.691 ± 0.038 1.909 ± 0.035 1.908 ± 0.043
CAE 1.869 ± 0.080 1.609 ± 0.054 2.020 ± 0.081 1.969 ± 0.092 1.838 ± 0.049 1.685 ± 0.054 2.101 ± 0.073 2.004 ± 0.075
CAAverage 1.851 ± 0.025 1.697 ± 0.046 1.976 ± 0.037 2.015 ± 0.045 1.813 ± 0.021 1.787 ± 0.032 2.053 ± 0.025 2.024 ± 0.030
CNA 623.5 ± 28.1 728.7 ± 30.7 736.3 ± 33.1 827.8 ± 28.7 636.7 ± 15.0 780.2 ± 16.4 725.0 ± 48.1 835.5 ± 39.0
CNB 1,393.9 ± 16.6 1,427.1 ± 40.3 1,529.5 ± 47.3 1,541.3 ± 61.1 1,296.1 ± 23.8 1,537.3 ± 42.5 1,326.3 ± 42.8 1,528.6 ± 40.0
CNC 1,621.1 ± 45.4 1,603.8 ± 54.1 1,882.2 ± 51.1 1,899.5 ± 47.0 1,553.5 ± 38.7 1,486.2 ± 49.3 1,795.9 ± 67.0 1,822.1 ± 93.3
CND 1,258.6 ± 25.5 1,300.9 ± 29.7 1,488.4 ± 48.0 1,456.5 ± 26.8 1,280.8 ± 24.1 1,344.4 ± 33.2 1,478.1 ± 34.6 1,560.3 ± 38.6
CNE 2,024.1 ± 105.9 2,352.6 ± 97.7 2,341.5 ± 94.3 2,481.4 ± 115.4 1,963.2 ± 54.8 2,217.3 ± 57.8 2,237.0 ± 124.3 2,346.0 ± 87.9
CNTotal 7,811.0 ± 102.7 8,243.3 ± 194.4 8,841.1 ± 166.9 9,029.2 ± 163.9 7,467.8 ± 89.8 8,245.4 ± 127.3 8,273.4 ± 238.9 8,997.6 ± 196.6
Wing traits: outline shape (WSH), size (WSI), width (WW), length (WL), and the area of intervein regions A-E (IVRA-E). Cellular traits: average cell area (CA) or cell number (CN) at intervein regions A-E (subscripts), plus the average cell area across all intervein regions (CAAverage) and total cell number in dorsal wing blade (CNTotal). Units: WSH is a ratio (WW/WL); WSI, WW and WL are in millimeters, IVR areas are in mm2; and all CA estimates are in mm2 ×10-4.
C A P Í T U L O I I I : B A S E S C E L U L A R E S
102
Table S2: Mean ± SE for wing morphology and cellular variations at intervein regions in baseline population.
Baseline population
Trait Wild (n=23) G1 (n=20)
WSH 0.485 ± 0.002 0.471 ± 0.003
WSI 0.632 ± 0.008 0.782 ± 0.006
WW 0.881 ± 0.012 1.073 ± 0.010
WL 1.815 ± 0.025 2.281 ± 0.018
IVRA 0.125 ± 0.004 0.184 ± 0.004
IVRB 0.226 ± 0.006 0.329 ± 0.006
IVRC 0.251 ± 0.006 0.385 ± 0.006
IVRD 0.197 ± 0.005 0.309 ± 0.005
IVRE 0.322 ± 0.010 0.505 ± 0.010
CAA 1.989 ± 0.027 2.349 ± 0.050
CAB 1.797 ± 0.035 2.303 ± 0.055
CAC 1.685 ± 0.044 2.068 ± 0.035
CAD 1.700 ± 0.032 2.087 ± 0.031
CAE 1.816 ± 0.039 2.358 ± 0.062
CAAverage 1.798 ± 0.025 2.233 ± 0.024
CNA 631.4 ± 19.8 790.3 ± 22.0
CNB 1,260.6 ± 32.0 1,436.0 ± 33.3
CNC 1,506.8 ± 43.8 1,873.5 ± 46.4
CND 1,166.3 ± 31.8 1,489.7 ± 34.1
CNE 1,777.9 ± 46.6 2,173.2 ± 72.8
CNTotal 7,008.0 ± 148.1 8,638.8 ± 164.9
Abbreviations are the same presented in Table S1. G1 flies were reared at
16°C.
C A P Í T U L O I I I : B A S E S C E L U L A R E S
103
TABLE S3: Quadratic regressions of average cell area across rank order of all five intervein regions fitted to complete data set of each experimental group, and F-tests of group differences from quadratic regressions adjusted to each individual wing.
Abbreviations: coefficient that describes the shape of the curve (g2) and characteristic estimates of minimum value (MV) and intervein region that holds the minimum value (IVR-MV). A rank order was attributed to each intervein region according to their proximity from the most anterior region of the wing: IVRA was ranked as 1, IVRB as 2 and so forth. See Fig. 4 for visual representation of the quadratic curves fitted to complete data set of each experimental group. ANOVA (F-ratio) of group differences was performed in data derived from quadratic regressions adjusted to each individual wing: group differences (df=5), model error: df=112. aTukeys’s a posteriori pairwise comparisons for the significant group differences in MV; six outlier cases were excluded: (Wild = R_TD25 = L_TD25) ≠ (G1 = L_TD16 = R_TD16); α = 0.05. G1 flies were reared at 16°C. ***P ≤ 0.001.
Experimental
quadratic regression fitted to complete data set of each group
Mean ± SE from quadratic regressions adjusted to each individual wing
group n g2 MV IRV-MV F-ratio R2 g2 MV IRV-MV
WILD 23 0.053 1.682 3.417 23.167*** 28.0% 0.053 ± 0.009 1.669 ± 0.035 3.588 ± 0.257
G1 20 0.063 2.105 3.156 11.992*** 18.2% 0.063 ± 0.013 2.092 ± 0.040 2.944 ± 0.222
L_TD16 20 0.054 1.900 3.337 14.876*** 21.9% 0.054 ± 0.011 1.934 ± 0.059 2.556 ± 0.713
R_TD16 20 0.053 1.891 3.627 20.260*** 28.0% 0.053 ± 0.015 1.942 ± 0.053 2.708 ± 0.354
L_TD25 20 0.056 1.711 3.382 30.752*** 37.5% 0.056 ± 0.010 1.742 ± 0.063 2.261 ± 1.420
R_TD25 20 0.029 1.643 4.192 18.451*** 26.1% 0.029 ± 0.009 1.680 ± 0.042 3.018 ± 0.455
F-ratio of group
differencesa
1.067 12.289*** 0.459
C A P Í T U L O I V : G E N E S C A N D I D A T O S
104
Gene Expression Profile and Candidate Genes in Strains of Drosophila melanogaster
Selected for Divergent Wing Shape
Daniel Mattos, Bruna Palma da Matta, Márcio Alves and Blanche Christine Bitner-Mathé.
Manuscript under current refinement for publication.
Running title: gene expression of Drosophila wing shape
C A P Í T U L O I V : G E N E S C A N D I D A T O S
105
ABSTRACT
Mapping the elements involved in the phenotypic determination of complex traits has
been a challenge. The Drosophila wing is a great model for such studies because much of its
developmental biology, genetics, and environmental responses are well known. However,
models still fail to satisfactorily map genetic determinants controlling quantitative wing shape
variation. Here we investigated candidate genes through an extensive search in the expression
profile; both by a microarray assay and by real time quantitative PCR on biological replicates.
Results indicate that genetics underlying wing shape follows an infinitesimal genetic model,
with a great number of genes with small effects, aligned with results obtained for other shape
indexes in the literature. We also identified strong candidate genes (lft, Trl, Idgf4, sgg,
CG17919, ems, GstD3, dp, CG10208, rho and deltaTry) that should be tested on interspecific
variation due to their consistent response in all biological replicates. Furthermore, we discuss
that a large amount of shape variation seems to lie within the interaction of nuclear elements
rather than deterministic elements with major effects alone.
Keywords: microarray – real time qPCR – evolution - complex traits
C A P Í T U L O I V : G E N E S C A N D I D A T O S
106
INTRODUCTION
Since the rediscovery of Mendel’s work in the beginning of the 20th century, geneticists
worldwide have tried to map the underlying genetics of perceivable phenotypes. More than
a century later, astounding breakthroughs have been accomplished, including the discovery
of the DNA and part of the intricate code within it. Many phenotypes have been successfully
mapped, with huge impact on many scientific fields. However, many others refuse to fit into
simplistic models and fall into a class of complex traits. In common, these traits share a
multidimensional phenotypic space (e.g. shape variation or medical syndromes with multiple
symptoms) and a great number of epistatic and pleiotropic genes. The infinitesimal genetic
model is usually assumed with many genes of small effect underlying phenotypic variation.
Moreover, these traits are usually influenced by environmental conditions that blur the effects
of single genetic variants. Expression patterns, regulation and epigenetics then became the
focus of recent studies in an attempt to break the genetic codes of such phenotypes. Still,
there is a large amount of variation not accounted for by such processes. Baranzini et al. (2010)
made an extensive research on genetic and epigenetic elements underlying multiple sclerosis
and found no single variant underlying the disease. However, the study points to many
elements that together contribute to the syndrome. Similar inferences about genetic
establishment and evolutionary persistence were made for the causes of schizophrenia
(NETTLE; CLEGG, 2006; PEARLSON; FOLLEY, 2008; WILKINS, 2011).
Drosophila wing offers great opportunities for the understanding of the evolution of
complex traits. Developmental pathways are well described (ALEXIS; ISAAC; DAVID, 2015;
BIER, 2000; LEGOFF; ROUAULT; LECUIT, 2013) and many studies on shape variation are
available (CARREIRA et al., 2011; DEBAT; DEBELLE; DWORKIN, 2009; DEBAT et al., 2003;
MENEZES et al., 2013; TORQUATO et al., 2014). In addition, the wing is formed from a
modularized structure in the larva, the wing imaginal disc, isolating the genetic framework
from the development of other structures.
Here we use strains submitted to an intense selection program on wing shape that
generated strains with divergent phenotypes to (1) identify wing shape candidate genes and
(2) test the replicability of expression profiles in biological replicates. Finally, we tested
whether strong candidate genes could alone account for shape variation through tissue-
specific RNAi-mediated silencing. We found support for an infinitesimal genetic model, with
C A P Í T U L O I V : G E N E S C A N D I D A T O S
107
many genes with small effects behind wing shape. Furthermore, results suggest that biological
replicates achieved similar shapes through alternate genetic pathways.
MATERIAL AND METHODS
MICROARRAY ON 1L AND 1R STRAINS
Experimental conditions for Microarray assays were previously described by Matta
(2010). Briefly, 1L (elongated winged strain) and 1R (rounded winged strain) from the 67th
generation were analyzed, with four biological replicates. 100 imaginal disc were collected
from each strain and replicate. The microarray chip produced by Drosophila Genomics
Resource Center (DGRC-2 Ologonucleotide Array; https://dgrc.cgb.indiana.edu/), was read on
VersArray ChipReaderTM 3m form BioRad. Details on the microarray essay can be found in
Matta (2010).
BIOLOGICAL REPLICATE STRAINS
Three round strains (1R, 2R and 5R), three elongated strains (1L, 2L and 5L) and
unselected controls (1C, 2C and 5C) were analyzed. For each of the nine strains, two biological
replicate crosses (named A and B) were done. Eight couples from generations 120 and 121
were randomly put in vials containing standard Drosophila medium with blue bromophenol
for third instar larvae identification (ASHBURNER, 1989). Adults were allowed to mate and
oviposit for three days after which they were transferred to a new vial, hence generating six
technical replicates. On the 4th day after initial contact, third instar larvae were sexed and had
wing imaginal discs collected. Remaining larvae were allowed to fully develop and adults were
stored in 95% alcohol.
WING IMAGINAL DISC RNA EXTRACTION AND REAL TIME qPCR
For each biological replicate (A and B) of L and R strains, 60 imaginal discs were
collected, half of each sex and stored in 500µL of RNA Later at -20°C. For unselected controls,
40 discs were collected. In total, 1280 imaginal discs were extracted. RNA was extracted by
RNeasy mini kit (QIAGEN), following manufacturer’s instructions. RNA was quantified on
C A P Í T U L O I V : G E N E S C A N D I D A T O S
108
Nanodrop (Thermo Fisher) and Bioanalyzer (Agilent Technologies); all samples presented good
quality and quantity. cDNA was then synthetized and used on RT-qPCR for expression level
quantification, through the incorporation of SybrGreen during qPCR. 22 genes were tested.
Reference genes were established by Matta (2010).
MORPHOLOGICAL VALIDATION
Candidate genes had their expression silenced by enhancer_GAL4 / UAS_RNAi
silencing system and morphological consequences evaluated. Briefly, a Drosophila strain
produced with an element consisting of a specific genomic enhancer (here we used vestigial
and nubbin enhancers due to their known expression on the wing disc pouch; CIFUENTES;
GARCÍA-BELLIDO, 1997) attached to the yeast transcription factor GAL4. When genomic
enhancer is activated (Fig. 1), GAL4 is produced and binds to the UAS promoter associated to
the silencing target gene (Fig. 2). Nubbin_Gal4 strain also expresses the protein Dicer (Dcr)
which intensifies gene silencing by degrading the silencing complex formed. Each tested gene
was crossed with vg_Gal4 and with nubbinDcr_GAL4. Control strains (GD-60.000 and KK-
60.100) with no element inserted were also crossed to both GAL4 strains. Reciprocal crosses
and (A and B) were performed with five couples at each cross. Strains were acquired at
Bloomington stock center (http://flystocks.bio.indiana.edu/) and Vienna stock center
(http://stockcenter.vdrc.at/control/main).
Cross 1A: 5♀ UAS-RNAi X 5♂ GAL4-nub-Dcr
Cross 1B: 5♂ UAS-RNAi X 5♀ GAL4-nub-Dcr
Cross 2A: 5♀ UAS-RNAi X 5♂ GAL4-vg
Cross 2B: 5♂ UAS-RNAi X 5♀ GAL4-vg
Controls
Cross 1A: 5♀ Control 60.000 X 5♂ GAL4-nub-Dcr
Cross 1B: 5♂ Control 60.000 X 5♀ GAL4-nub-Dcr
Cross 2A: 5♀ Control 60.000 X 5♂ GAL4-vg
Cross 2B: 5♂ Control 60.000 X 5♀ GAL4-vg
C A P Í T U L O I V : G E N E S C A N D I D A T O S
109
Figure 1. Expression pattern of protein GAL4 on the wing imaginal disc. nubbin_GAL4 element (a and b) showing
GAL4 in the nubbin expression domain and vestigial_GAL4 (c and d). Images taken on a fluorescence microscope.
For both elements, GAL4 is expressed in the imaginal disc pouch, which originates adult wing blade. GAL4 will
then drive the expression of the silencing element, hence silencing the target gene.
Figure 2. GAL4 / UAS silencing system. When genomic enhancer is activated, GAL4 is expressed and binds to the UAS promoter, driving the expression of silencing gene X. Figure
modified from ST Johnston (2002).
C A P Í T U L O I V : G E N E S C A N D I D A T O S
110
RESULTS
EXPRESSION PATTERNS
The microarray performed on the 1L/1R67thGen strains indicated 150 differentially
expressed genes (DE). Fig. 3 groups DE genes by biological function; enzymes are the most
represented group. Ten transcription factors and 11 signaling transduction constitute good
candidates genes due to their ability to regulate downstream genes and change
developmental pathways.
From these candidate genes, 22 were tested through qPCR on three biological
replicates from 120th and 121st generations, forming pairs with the highest phenotypic
divergence. Fig. 4 shows the morphological variation between pairs and Table 1 shows the F-
tests. Genes were chosen according the following criteria: (1) the 5 genes with highest ranks,
taking into account both FC and significance; (2) highly ranked genes with known participation
on wing imaginal disc developmental pathways and (3) highly ranked transcription factors due
to their regulatory function. Fig. 5 exhibits the expression profile by heatmap. Differential
expression has low consistency among biological replicates for most genes. In fact, only three
genes (lft, sgg and Idgf4) have consistent and significant variation in all three replicates and
eight genes are consistent in at least two (CG10208, lft, Trl, sgg, ems, Idgf4, GstD3 and
CG17919). These genes represent the strongest candidates for wing shape variation that
emerged from this analysis. The large number of DE genes in the microarray and the low
consistency among biological replicates are evidence supporting the infinitesimal genetic
model for shape variation.
Comparing L/R Strains expressions provides information on DE genes, but is
uninformative regarding the direction of the response related to the original state prior to
selection. Therefore, for the eight strong candidates, we quantified the expression for an
unselected control strains (1C Strain) in order to find if divergence in expression was
bidirectional (i.e., both L and R strains shifted expression levels) or unidirectional. Fig. 6 shows
the plot of the normalized expression ratios (NER) for L, C and R Strains. C Strain is exhibited
in the center and each is connected by lines to the others representing the expression shift.
To our knowledge, this form of presentation is original in such studies and can provide an easy-
to-visualize comparison with the treatments of interest when the control state is available.
C A P Í T U L O I V : G E N E S C A N D I D A T O S
111
Figure 3. Pie chart summarizing results of the Microarray, grouped by biological function of significantly regulated genes. Note the prevalence of genes with unknown biological functions and an enrichment of transcription factors.
C A P Í T U L O I V : G E N E S C A N D I D A T O S
112
1L / 1R 2L / 5R 5L / 2R
SEL Error SEL Error SEL Error
DF 1 48 1 48 1 48
MS 0.1713 0.0001 0.1715 0.0001 0.051 0.000 p 0.0000 0.0000 0.000
Table 1. F-Tests for the effects of direction of selection ((SEL: L x R Strains) for each pair of biological replicates in the 123rd generation analyzed by qPCR.
Figure 4. Boxplot of WSH variation in the 123rd generation with the pairs of strains analyzed by the RT-qPCR. L Strains in blue and R Strains in red. Pairs were formed aiming highest morphological divergence.
C A P Í T U L O I V : G E N E S C A N D I D A T O S
113
Figure 5. Heatmaps of expression ratios of candidate genes by the microarray of the 67th generation (array) and by real time qPCR testing replicability of three biological replicates of the 121st and 122nd generations. Green refers to genes upregulated on the L Strains and red refers to those downregulated. Significantly DE genes marked by *.
C A P Í T U L O I V : G E N E S C A N D I D A T O S
114
Figure 6. Visual representation of the variations of expression in L (Log L) and R (Log R) Strains compared to unselected control (Log C). Log C represents expression levels of intermediate phenotype. Normalized expression ratio (NER) used on the Y axe of the graphs as the expression quantifier. Only 1L/1R (contiguous line) and 2L/5R (traced line) are shown because morphological divergence is greater. This graph allows a rapid visualization of which direction of selection had its expression shifted from the control dosage. E.g., for sgg, L strains diverged from the control while R strains remained unaltered. Blue line indicates significant expression differences.
C A P Í T U L O I V : G E N E S C A N D I D A T O S
115
MORPHOLOGICAL VALIDATION
RNAi-mediated gene silencing targeting transcripts present on the imaginal disc pouch
of third instar larvae were performed for all 8 candidate genes and to 12 others due to their
rank in the microarray, consistency among biological replicates or to known biological
functions during wing development. Table 2 shows mean and standard error for wing shape,
size, length and width of control and expression-silenced strains with the respective F-tests,
exhibiting only results containing the enhancer element (nubbinDcr_GAL4 or vg-GAL4) that
provided information on quantitative variation; i.e. crosses that generated no progeny or
malformed wing are not exhibited. Only 4 silenced strains had their wing shape significantly
altered, two of them from the list of 8 candidates (Idgf4, sgg, dp and deltaTry). The effect of
silencing dp was rather impressive, with silenced strains exhibiting a very round wing, mainly
explained by a shortening in wing length (figure 7). However, others were able to significantly
alter at least one of the biological axes (length or width) that compose our index of wing shape
(lft, CG17919, rho and Trl) and are also good candidates for controlling wing shape
quantitative variation. Regarding experimental designs with RNAi-mediated silencing strains,
the element nubbinDcr_GAL4 had a higher frequency of malformed wings or unviable
progeny, probably due to its intense silencing or to unspecific effects of the higher dosage of
the Dicer protein during wing development. Figure 8 shows four examples of malformed wings
with nubbinDcr_GAL4 (for which quantitative variation was analyzed by the vg_GAL4 cross).
C A P Í T U L O I V : G E N E S C A N D I D A T O S
116
Table 2. RNAi strains wing shape phenotypic mean and standard error. F-test between silenced strain (gene) and its
respective control (60.000 or 60.100). Control strains were tested against an unspecific control strain with Y
chromosome genes known to be expressed in the testis (Ppr-y and KL-5) silencing element. Wing shape (WSH), size
(WSI), length (WL) and width (WW).
WSH WSI WL WW
Gene Gal4 Control Mean SE Mean SE Mean SE Mean SE
60.000 KL-5 0.460 0.002 807.313 12.625 *** 1191.104 19.815 * 547.216 8.098
60.100 Ppr-y 0.457 0.003 797.106 11.704 1180.391 20.310 538.397 6.787
Idgf4 nubDcr 60.100 0.484 0.003 * 721.462 11.411 1037.643 17.027 501.691 7.881
lft nubDcr 60.100 0.482 0.002 738.585 13.415 *** 1064.032 20.019 *** 512.734 9.118
CG15353 nubDcr 60.100 0.462 0.002 797.402 13.201 1174.301 20.678 541.535 8.567
CG17919 nubDcr 60.000 0.449 0.003 762.009 12.600 1137.431 19.819 510.601 8.327 *
chit nubDcr 60.100 0.458 0.002 771.613 13.299 1140.803 20.920 521.963 8.576
elB nubDcr 60.100 0.463 0.003 797.713 13.870 1172.702 20.399 542.707 9.642
rho nubDcr 60.100 0.479 0.002 690.550 11.284 998.544 17.328 * 477.598 7.435
CG10208 nubDcr 60.000 0.459 0.002 760.961 12.251 1124.050 18.965 515.204 8.035
CG32373 nubDcr 60.100 0.481 0.002 784.600 11.148 1131.316 17.347 544.203 7.348
deltaTry nubDcr 60.000 0.465 0.002 *** 788.495 12.014 1157.202 18.769 537.331 7.874
Jon65Aiii nubDcr 60.100 0.459 0.003 801.682 12.585 1183.793 21.138 543.000 7.505
Vang nubDcr 60.100 0.486 0.003 770.979 11.778 1106.388 19.415 537.335 7.101
Dr / Drop nubDcr 60.100 0.474 0.003 797.599 14.480 1159.625 24.141 548.694 8.570
60.000 KL-5 0.477 0.003 795.497 12.735 1151.763 19.623 549.513 8.460
60.100 Ppr-y 0.468 0.003 810.065 12.203 1185.707 21.029 553.536 6.983
ems Vg 60.000 0.479 0.004 805.076 13.730 * 1163.889 22.637 557.013 8.540
bun Vg 60.000 0.470 0.003 680.738 16.119 994.106 25.725 466.224 10.151
sgg Vg 60.100 0.454 0.005 * 713.893 15.268 *** 1060.719 25.872 480.682 9.381 ***
dp Vg 60.000 0.754 0.008 *** 532.707 9.548 *** 614.414 12.462 *** 462.097 7.842 *
Trl Vg 60.100 0.471 0.003 732.421 14.910 *** 1067.171 21.779 * 502.756 10.449 ***
C A P Í T U L O I V : G E N E S C A N D I D A T O S
117
WSH WSI
WL
WW
Gene Gal4 Control Mean SE Mean SE Mean SE Mean SE
GstD3 Vg 60.100 0.468 0.003 786.074 13.141 1150.010 22.098 537.427 7.840
vito Vg 60.100 0.475 0.005 778.101 10.834 1130.682 18.813 535.764 7.017
* P < 0.05 ; *** P < 0.01
Figure 7. Effects of RNAi-mediated silencing of dp on the wing imaginal disc of 3rd instar larvae of D. melanogaster. Control wing (left) and dp silenced (right). Note that most of the shape variation is due to a shortening on the wing length.
Figure 8. Malformations on the wing of D. melanogaster due to RNAi-mediated gene silencing by the nubbinDcr_GAL4 element.
C A P Í T U L O I V : G E N E S C A N D I D A T O S
118
DISCUSSION
Microarray indicated 150 DE genes underlying wing shape quantitative variation.
Taking under consideration all analyses, we identified as strong candidates the following: lft,
Trl, Idgf4, sgg, CG17919, ems, GstD3, dp, CG10208, rho and deltaTry. Testing those genes in
interspecific variation by real time qPCR is the next step towards identifying major quantitative
wing shape genes. However, consistency of expression ratios across biological replicates is
usually low and even when expression analysis was repeated for the 1L/1R pair in a later
generation through real time qPCR, some genes were no longer significantly regulated. Nine
out of the twenty tested genes by RNAi-mediated silencing were able to alter the phenotype
apart from the control; four with detectable shape differences and the others interfering with
wing length or width. All tested silenced genes, with the exception of dp, had a very small
phenotypic impact, suggesting that it is not the variation of expression in one or a few genes
that promotes quantitative variation but, rather, the cumulative effects of multiple genes
interacting, supporting the infinitesimal genetic model for Drosophila wing shape. Similar
conclusions were reached for other wing shape index (WEBER et al., 2008). However, when
lower levels of phenotypic variance are analyzed, identification of genes with major effects
seems more efficient (or viable). Umetsu; Dunst; Dahmann (2014) identified only the action
of Eph receptor in an RNAi screening targeting almost 3000 genes required for shaping the
anteroposterior compartment boundary of the wing imaginal disc.
Weber et al. (2008) used a microarray analysis to identify wing shape candidate genes.
Their shape index is different from ours and captures variations mostly along the
anteroposterior axis. Still, they also found a very large number of genes (198-534 depending
on the lines compared on the microarrays) and interpreted results as confirming “the highly
polygenic and degenerate basis of wing shape”, although accounting some genes to “fixation
of noncontributory alleles” in lines selected from larger population sizes. LaayounI et al.
(2007), analyzing expression profile of thermal tolerance in D. suboscura, also found a large
number of genes and recognized the same issue. The number of DE genes indicated by the
microarray on our strains indicated a smaller number of genes (150), however still large. Our
selection strains were founded from 135 isofemale lines and the same effect might explain
the still large number of genes we found here and the low replicability among replicates.
C A P Í T U L O I V : G E N E S C A N D I D A T O S
119
Besides the identification of noncontributory alleles, another issue common to studies
with expression ratios is that genes with higher FC are considered the best candidates,
although analyses try to correct this assumption by ranking genes based on the replicability as
well as the FC. Technical replicates might reduce experimental errors due to differential RNA
extraction or treatment conditions but they still do not account for the replicability of the
effects of such a dosage ratio. Only biological replicates might reduce (but not extinguish) the
biological relevance of FC intensity. A gene 100x more expressed might have little or no impact
on any phenotype depending on the genetic and epigenetic background of the particular cell
while an increase of 0.5x in the expression of another might have profound impacts by
regulating downstream genes. Wilkins (2011) discusses the nonlinearity of the effects of gene
expressions shifts, but argues in favor of qualitative conclusions that can be addressed under
the linearity assumption. Comparison of the expression levels of the selected strains (L and R)
and the unselected control (C) indicates a nonlinear effect of those genes on wing shape.
Taking these issues in consideration, morphological validation is extremely important, but
challenging when complex traits are being analyzed since, as seen in this work, in which most
of the individual gene silencing did not recapitulate the expected phenotypes.
From the list of genes assessed by the Microarray, many do not participate in known
wing developmental pathways. In fact, many genes are involved with general metabolic
processes that are unlikely to be directly related to organ morphogenesis. This suggests that
wing development and wing quantitative shape variation might be regulated by different
genes, with those involved in the quantitative variation epistatically interacting with the
known developmental pathways.
Finding robust candidate genes, epigenetics or any nuclear causal element for
quantitative traits has proven to be challenging. The interactions of all these elements seem
to underlie the quantitative phenotypes observed rather than the allelic or even the dosage
variation of expression of few genes. Epistatic interactions between genes, genetic and
epigenetic markers might have such a huge effect on phenotypes that isolating each variant
will account for a small part of the observed phenotypic variance. Assessing the variance
associated with the interactions is harder but may be crucial for a better comprehension of
phenotypes, including for areas like medicine. For instance, when heart diseases are
considered, isolated genetic variants and individual habits (such as smoking or obesity) are
C A P Í T U L O I V : G E N E S C A N D I D A T O S
120
associated with higher risks, but neither alone is able to account for a major amount of the
associated risk. Hence, it is likely that the increase in risk is correlated with the interaction of
those and other factors. This seems to be true for most complex traits (BARANZINI et al.,
2010).
Considering the effects of the artificial selection program on the strains, it remains
unclear which elements actually responded. Differences in expressions levels can be explained
by changes in cis or trans regulatory genetic regions and by epigenetic shifts. From the list of
strong candidate genes, analyses of the cis-regulatory regions as well as epigenetic markers
might elucidate the participation of both types of nuclear variation contributing to short-term
phenotypic evolution. Wittkopp; Haerum; Clark (2008), analyzing gene expression divergence
and the contribution of cis and trans regulatory regions between D. melanogaster and D.
simulans, found that “cis-regulatory changes seem to accumulate preferentially over time.”
The same might not be true for short-term evolution and comparing both would bring
important insights.
Li and Saunders (2005) argue that testing the hypothesis that gene expression shifts
might account for a significant amount of the observed phenotypic variance among
populations and species is hard, but crucial. The genetics of complex traits keeps challenging
the most current techniques of biological exploration. Understanding the relative contribution
and the interaction of the components underlying variation in such traits is imperative both
for our understanding of evolution and for medical research.
ACKNOWLEDGMENT
We thank Danielle Tesseroli and Bianca Menezes for the establishment and
maintenance of the artificial selection strains and Dulcinea da Rocha for technical assistance.
This paper is part of the D. Sc. requirements of Daniel Mattos at the Biodiversity and
Evolutionary Biology Graduate Program of the Federal University of Rio de Janeiro and was
supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – CAPES.
C A P Í T U L O I V : G E N E S C A N D I D A T O S
121
REFERENCES
ALEXIS, M.-V.; ISAAC, S.-C.; DAVID, H. Making quantitative morphological variation from basic developmental processes: Where are we? The case of the Drosophila wing. Developmental dynamics, 23. 2015.
ASHBURNER, M. Drosophila: A Laboratory Handbook. 1989.
BARANZINI, S. E. et al. Genome, epigenome and RNA sequences of monozygotic twins discordant for multiple sclerosis. Nature, v. 464, p. 1351–1356, 2010.
BIER, E. Drawing lines in the Drosophila wing: Initiation of wing vein development. Current Opinion in Genetics and Development, 2000
CARREIRA, V. P. et al. Genetic basis of wing morphogenesis in Drosophila: sexual dimorphism and non-allometric effects of shape variation. BMC developmental biology, v. 11, p. 32, 2011.
CIFUENTES, F. J.; GARCÍA-BELLIDO, A. Proximo-distal specification in the wing disc of Drosophila by the nubbin gene. Proceedings of the National Academy of Sciences of the United States of America, v. 94, n. 21, p. 11405–11410, 1997.
DEBAT, V. et al. Allometric and nonallometric components of Drosophila wing shape respond differently to developmental temperature. Evolution, v. 57, n. 12, p. 2773–2784, 2003.
DEBAT, V.; DEBELLE, A.; DWORKIN, I. Plasticity, canalization, and developmental stability of the Drosophila wing: joint effects of mutations and developmental temperature. Evolution, v. 63, n. 11, p. 2864–2876, 2009.
LAAYOUNI, H. et al. Thermal evolution of gene expression profiles in Drosophila subobscura. BMC Evolutionary Biology, v. 7, p. 42, 2007.
LEGOFF, L.; ROUAULT, H.; LECUIT, T. A global pattern of mechanical stress polarizes cell divisions and cell shape in the growing Drosophila wing disc. Development, v. 140, n. 19, p. 4051–9, 2013.
LI, W.-H.; SAUNDERS, M. A. The chimpanzee and us. Nature, v. 437, p. 50–51, 2005.
MATTA, B. P. Identificaçao de lócus candidatos ao controle da variação de tamanho e forma nas asas de Drosophila, em especies do grupo melanogaster. 2010. Tese (Doutorado em Ciências Biológicas) – PGGEN. 2010.
MENEZES, B. F. et al. The influence of male wing shape on mating success in Drosophila melanogaster. Animal Behaviour, v. 85, n. 6, p. 1217–1223, 2013.
NETTLE, D.; CLEGG, H. Schizotypy, creativity and mating success in humans. Proceedings. Biological sciences / The Royal Society, v. 273, n. 1586, p. 611–615, 2006.
C A P Í T U L O I V : G E N E S C A N D I D A T O S
122
PEARLSON, G. D.; FOLLEY, B. S. Schizophrenia, psychiatric genetics, and Darwinian psychiatry: An evolutionary framework. Schizophrenia Bulletin, 2008
ST JOHNSTON, D. The art and design of genetic screens: Drosophila melanogaster. Nature reviews. Genetics, v. 3, n. 3, p. 176–88, 2002.
TORQUATO, L. et al. Cellular basis of morphological variation and temperature-related plasticity in Drosophila melanogaster strains with divergent wing shapes. Genetica, p. 1–11, 2014.
UMETSU, D.; DUNST, S.; DAHMANN, C. An RNA interference screen for genes required to shape the anteroposterior compartment boundary in Drosophila identifies the Eph receptor. PloS one, v. 9, n. 12, p. e114340, 2014.
WEBER, K. E. et al. Microarray analysis of replicate populations selected against a wing-shape correlation in Drosophila melanogaster. Genetics, v. 178, n. 2, p. 1093–1108, 2008.
WILKINS, J. F. Genomic imprinting and conflict-induced decanalization. Evolution, v. 65, n. 2, p. 537–53, 2011.
WITTKOPP, P. J.; HAERUM, B. K.; CLARK, A. G. Regulatory changes underlying expression differences within and between Drosophila species. Nature genetics, v. 40, n. 3, p. 346–50, 2008.
C O N C L U S Õ E S E P E R S P E C T I V A S F U T U R A S
123
CONCLUSÕES
Os estudos apresentados nesta tese evidenciam o caráter complexo da variação
quantitativa da forma e tamanho da asa. O programa de evolução experimental da asa
forneceu linhagens com fenótipos extremos que permitem a investigação de diversos níveis
de variação biológica.
O programa de seleção artificial gerou linhagens com asas alongadas e linhagens com
asas arredondadas. Asas redondas foram alcançadas por um aumento da largura de mesma
intensidade da diminuição do comprimento. Já asas alongadas parecem ser fruto
primariamente da redução da largura. Dados não apresentados do início da seleção indicam
que linhagens redondas apresentaram uma resposta mais acentuada diminuindo o
comprimento e só após algumas gerações de seleção houve resposta da largura.
Aparentemente, a redução de medidas é uma resposta morfológica mais facilmente
recrutada. Além disso, a resposta à seleção ocorreu através de modificações na quantidade
de células da asa, sem mudanças de tamanho celular. Dois compartimentos foram
identificados, em especial, pela uniformidade de intensidade e direção de landmarks
localizados na porção mais proximal da asa. Além disso, o padrão de correlação genética da
população prévia ao processo seletivo foi capaz de prever parte das respostas de traços
correlacionados indicando que as matrizes de correlação genética têm grande impacto nas
trajetórias evolutivas.
O estudo da plasticidade fenotípica (PF) trouxe importantes insights nos anos recentes.
Em especial, a área tem despertado interesse devido à possibilidade de se incorporar essas
informações aos modelos de previsões de impacto das mudanças climáticas e suas
consequências sobre a diversidade no planeta. Por exemplo, um estudo recente mostrou que,
na Europa, insetos de coloração mais escura são favorecidos em climas mais amenos
enquanto os de coloração mais clara são favorecidos em ambientes mais quentes (ZEUSS et
al., 2014). Outro trabalho avaliou a influência da pigmentação abdominal em Drosophila sobre
o dessecamento e resistência a raios ultravioleta (MATUTE; HARRIS, 2013), cujas normas de
reação são bem conhecidas (ROCHA; KLACZKO, 2009). A reunião dessas informações pode
facilitar previsões locais de perdas de biodiversidade de insetos. A importância da PF para a
sobrevivência de espécies torna-se cada vez mais evidente. Ramp e outros (2015)
C O N C L U S Õ E S E P E R S P E C T I V A S F U T U R A S
124
argumentam que a PF pode explicar, parcialmente, como baleias de barbatana puderam
sobreviver a mudanças nas temperaturas oceânicas nos últimos milhões de anos. Outro
estudo recente mostrou que a PF do mimetismo de peixes da espécie Pseudochromis fuscus é
capaz de conferir diversos ganhos em fitness. A PF retornou ao centro das atenções com tanta
força que novas formas de se analisar as normas de reação estão sendo constantemente
propostas (PERTOLDI et al., 2014; ROCHA; KLACZKO, 2012, 2014), fornecendo novos olhares
sobre problemas tão antigos.
No presente estudo, mostramos a influência da média fenotípica da forma da asa
sobre a inclinação da norma de reação, invertendo a orientação da variação termal da forma
da asa em linhagens com asas arredondadas. A dependência entre média fenotípica e norma
de reação para a forma da asa tem consequências importantes para o inseto, uma vez que
possíveis aumentos de temperatura que selecionem populações com diferentes formatos
terão impacto sobre a capacidade dessas populações em responderem à própria variação
termal.
Entretanto, apesar do alto número de publicações descrevendo as normas de reação
de tamanho e forma da asa de Drosophila, pouco se sabe sobre as bases genéticas que
permitem a plasticidade. A dependência entre média fenotípica e plasticidade evidenciada
aqui sugere que, para a forma da asa, a PF é em parte explicada por sensitividade alélica, onde
alguns dos genes envolvidos na determinação do fenótipo responderiam de maneiras
diferentes às mudanças ambientais. No capítulo de expressão gênica, apontamos também
alguns dos genes envolvidos na determinação fenotípica e uma busca aprofundada dentre os
candidatos poderia apontar novas frentes de estudo para identificação de genes também
envolvidos na resposta plástica. Um estudo recente em insetos da infraordem Fulgoromorpha
(Ordem: Hemiptera) identificou dois receptores de insulina envolvidos na determinação
plástica de dois morfotipos de asa, mostrando ser possível a identificação de genes capazes
de apresentar respostas às variações ambientais. Nesse modelo de estudo, trata-se de um
polifenismo com apenas dois morfotipos e alternativas binárias nas vias de desenvolvimento,
facilitando a identificação de genes reguladores das vias.
Por fim, o estudo apontou 11 genes candidatos ao controle da variação quantitativa
da forma da asa (lft, Trl, Idgf4, sgg, CG17919, ems, GstD3, dp, CG10208, rho and deltaTry).
C O N C L U S Õ E S E P E R S P E C T I V A S F U T U R A S
125
Alguns desses genes têm participação conhecida na morfogênese da asa (lft, Trl, Idgf4, rho;
flybase.org/). Os genes Trl e ems se associam ao promotor da RNA polimerase II, regulando a
expressão de diversos genes (CHOPRA et al., 2008; TAYLOR, H. S., 1998) e, portanto, têm
grande capacidade de promover modificações globais nos padrões de expressão das células
da asa. dp está associado com a aposição das lâminas dorsal e ventral do disco imaginal
(PROUT et al., 1997). Há pouca ou nenhuma informação conhecida sobre os genes CG17919,
CG10208, GstD3 (glutationa) e deltaTry (peptidase) e nenhuma informação fenotípica; em
especial, nada foi reportado sobre o envolvimento na formação da asa. Já o gene rho é
conhecido por especificar a diferenciação celular e, portanto, também consiste em um
importante candidato (MARENDA et al., 2006).
Um cenário bastante interessante está sendo construído para um melhor
entendimento do desenvolvimento de asa em insetos. Aproveitando o largo conhecimento já
publicado sobre genes envolvidos na formação da asa de Drosophila, Linz e Tomoyasu (2015)
promoveram uma extensiva busca pelos genes candidatos em Tribolium castaneum através
de silenciamento por RNAi. Seus resultados mostram uma alta conservação nos genes de
desenvolvimento, sem ter encontrado um único gene exclusivo de T. castaneum. Avançar os
conhecimentos sobre a asa de Drosophila permitirá, cada vez mais, um melhor entendimento
sobre a genética, o desenvolvimento e a evolução de asas de insetos em geral.
PERSPECTIVAS FUTURAS
Uma resposta, em geral, abre novas perguntas. Essa tese arranha a superfície de um
problema cujo núcleo ainda está distante. Um melhor entendimento das variações
quantitativas em traços ditos complexos é premente, seja para o avanço da nossa
compreensão dos processos evolutivos, seja para novas soluções em medicina. A asa de
Drosophila tem se mostrado um promissor modelo para estudo desse tipo de variação e novos
insights são obtidos a todo tempo. Ano passado, um novo olhar foi lançando sobre as asas,
revelando que há seleção sexual em padrões de coloração iridescentes por interferência
luminosa; um efeito ignorado durante um século de estudos com Drosophila, mostrando que
apesar de todo o conhecimento sobre elas, as asas ainda guardam surpresas (figura 1,
KATAYAMA et al., 2014). As linhagens de seleção geradas por Danielle Tesseroli e Bianca
C O N C L U S Õ E S E P E R S P E C T I V A S F U T U R A S
126
Menezes ainda prometem ser generosas no estudo das variações quantitativas de forma e
tamanho. Experimentos de seleção artificial permitem uma ampla gama de desenhos
experimentais, porém há uma grande demanda de tempo, espaço e recursos humanos para o
processo seletivo. Trabalhos em genética quantitativa também exigem tamanhos amostrais
que demandam altos recursos. Entretanto, um novo sistema automatiza a determinação
sexual, medição de peso, tamanho corporal e caracteres morfométricos, incluindo forma e
tamanho da asa em moscas vivas para experimentos de seleção artificial em larga escala
(MEDICI et al., 2015). Tal ferramenta promete abrir caminhos para perguntas ainda mais
instigantes, com tamanhos amostrais antes limitantes.
Dos elementos envolvidos na variação de forma da asa de Drosophila, algumas
perguntas se impõem. Qual a participação de outros eventos celulares, como migração e
mudanças no formato celular, no estabelecimento da forma da asa? Qual a contribuição de
diferenças alélicas em sequências codificantes de genes envolvidos na determinação
fenotípica? Quais elementos contribuem para a variação observada nos níveis de expressão
dos genes candidatos; seriam predominantemente diferenças em sequências reguladoras ou
elementos epigenéticos capazes de alterar os níveis de expressão? Alguns dos genes
Figura 1. Padrões de interferência da asa de Drosophila melanogaster. Machos com colorações mais vívidas são mais sexualmente atraentes para fêmeas. Figura adaptada de Katayama et al. (2014)
C O N C L U S Õ E S E P E R S P E C T I V A S F U T U R A S
127
candidatos ao controle da variação de forma também controlam a variação por plasticidade
fenotípica? Os genes candidatos também estão envolvidos na diferença morfológica
interespecífica? Mudanças evolutivas de curto prazo são mediadas por elementos
qualitativamente diferentes daqueles envolvidos em variações a longo prazo; i.e. há uma
maior participação de mudanças na regulação gênica por fatores epigenéticos na primeira e
por mudanças nas sequências reguladoras entre espécies?
Com sorte e trabalho, novos estudos, novas colaborações, novos alunos devem tentar
dar resposta a essas e tantas outras questões.
R E F E R Ê N C I A S B I B L I O G R Á F I C A S
128
AIGOUY, B. et al. Cell flow teorients the axis of planar polarity in the wing epithelium of Drosophila. Cell, v. 142, n. 5, p. 773–786, 2010.
ALEXIS, M.-V.; ISAAC, S.-C.; DAVID, H. Making quantitative morphological variation from basic developmental processes: Where are we? The case of the Drosophila wing. Developmental dynamics, 2015.
AMBEGAONKAR, A. A. et al. Propagation of dachsous-fat planar cell polarity. Current Biology, v. 22, n. 14, p. 1302–1308, 2012.
ARAUJO, H. Integrins modulate Sog activity in the Drosophila wing. Development, v. 130, n. 16, p. 3851–3864, 2003.
ARENDT, J. Ecological correlates of body size in relation to cell size and cell number: patterns in flies, fish, fruits and foliage. Biological reviews of the Cambridge Philosophical Society, v. 82, n. 2, p. 241–56, 2007.
ARTHUR, W. Developmental drive: An important determinant of the direction of phenotypic evolution. Evolution and Development, v.3,n.4, 2001.
ASHBURNER, M. Drosophila: A Laboratory Handbook. Plainview, NY: Cold Spring Harbor Laboratory Press, 1989.
AVANESOV, A. et al. The role of glypicans in Wnt inhibitory factor-1 activity and the structural basis of Wif1’s effects on Wnt and Hedgehog signaling. PLoS genetics, v. 8, n. 2, p. e1002503, jan. 2012.
AYROLES, J. F. et al. Systems genetics of complex traits in Drosophila melanogaster. Nature genetics, v. 41, n. 3, p. 299–307, 2009.
AZEVEDO, R. B. R. et al. Latitudinal variation of wing: thorax size ratio and wing-aspect ratio in Drosophila. Evolution, v. 52, p. 1353–1362, 1998.
BAENA-LÓPEZ, L. A.; BAONZA, A.; GARCÍA-BELLIDO, A. The orientation of cell divisions determines the shape of Drosophila organs. Current biology, v. 15, n. 18, p. 1640–1644, 2005.
BARANZINI, S. E. et al. Genome, epigenome and RNA sequences of monozygotic twins discordant for multiple sclerosis. Nature, v. 464, p. 1351–1356, 2010.
BATE, M.; ARIAS, A. M. The embryonic origin of imaginal discs in Drosophila. Development , v. 112, p. 755–761, 1991.
BECKER, W. A. Manual of quantitative genetic. 5th. ed. Pullman, WA, U.S.A: Academic Enterprises, 1992.
BIER, E. Drawing lines in the Drosophila wing: Initiation of wing vein development. Current Opinion in Genetics and Development, v.10, n.4, 2000.
R E F E R Ê N C I A S B I B L I O G R Á F I C A S
129
BIRDSALL K et al. Genetic variation for the positioning of wing veins in Drosophila melanogaster. Evolution and Development Development, v. 2, n. 1, p. 16–24, 2000.
BITNER-MATHÉ, B. C.; KLACZKO, L. B. Heritability , phenotypic and genetic correlations of size and shape of Drosophila mediopunctata wings. Heredity, v. 83, n. 6, p. 688–696, 1999a.
BITNER-MATHÉ, B. C.; KLACZKO, L. B. Plasticity of Drosophila melanogaster wing morphology: effects of sex, temperature and density. Genetica, v. 105, n. 2, p. 203–10, 1999b.
BITNER-MATHÉ, B. C.; KLACZKO, L. B. Size and shape heritability in natural populations of Drosophila mediopunctata: temporal and microgeographical variation. Genetica, v. 105, n. 1, p. 35–42, 1999c.
BITNER-MATHÉ, B. C.; PEIXOTO, A. A.; KLACZKO, L. B. Morphological variation in a natural population of Drosophila mediopunctata: altitudinal cline, temporal changes and influence of chromosome inversions. Heredity, v. 75 , p. 54–61, 1995.
BLAIR, S. S. Lineage compartments in Drosophila. Current biology, v. 13, n. 14, p. R548–R551, 2003.
BLAIR, S. S. Wing vein patterning in Drosophila and the analysis of intercellular signaling. Annual review of cell and developmental biology, v. 23, p. 293–319, 2007.
BRAY, S. Drosophila development: Scalloped and vestigial take wing. Current Biology, v. 9, 1999.
BUCHMANN, A.; ALBER, M.; ZARTMAN, J. J. Sizing it up: the mechanical feedback hypothesis of organ growth regulation. Seminars in cell & developmental biology, v. 35, p. 73–81, 2014.
CARREIRA, V. P. et al. Genetic basis of wing morphogenesis in Drosophila: sexual dimorphism and non-allometric effects of shape variation. BMC developmental biology, v. 11, p. 32, 2011.
CARROL, S. B. Infinitas formas de grande beleza: como a evolução forjou a grande quantidade de criaturas que habitam o nosso planeta. Rio de Janeiro: Ed. Jorge Zahar, 2006.
CARROLL, S. B. Evolution at two levels: on genes and form. PLoS biology, v. 3, n. 7, p. e245, 2005.
CARTER, A. J. R.; HOULE, D. Artificial selection reveals heritable variation for developmental instability. Evolution, v. 65, p. 3558–3564, 2011.
CHENG, Z. et al. A genome-wide comparison of recent chimpanzee and human segmental duplications. Nature, v. 437, n. 7055, p. 88–93, 2005.
CHEVERUD, J. M. Quantitative genetics and developmental constraints on evolution by selection. Journal of theoretical biology, v. 110, p. 155–171, 1984.
R E F E R Ê N C I A S B I B L I O G R Á F I C A S
130
CHOPRA, V. S. et al. Transcriptional activation by GAGA factor is through its direct interaction with dmTAF3. Developmental biology, v. 317, n. 2, p. 660–70, 2008.
CIFUENTES, F. J.; GARCÍA-BELLIDO, A. Proximo-distal specification in the wing disc of Drosophila by the nubbin gene. Proceedings of the National Academy of Sciences of the United States of America, v. 94, n. 21, p. 11405–11410, 1997
COLLINGE, J. E.; HOFFMANN, A A; MCKECHNIE, S. W. Altitudinal patterns for latitudinally varying traits and polymorphic markers in Drosophila melanogaster from eastern Australia. Journal of Evolutionary Biology, v. 19, n. 2, p. 473–482, 2006.
CONSORTIUM, D. G. et al. Evolution of genes and genomes on the Drosophila phylogeny. Nature, v. 450, n. 7167, p. 203–218, 2007.
CORRÊA, D. DE M. Evolução nos Padrões de Covariação entre Traços da Asa em Linhagens de Drosophila melanogaster submetidas à Seleção. 2009. Dissertação de Mestrado, Departamento de Genética, Universidade Federal do Rio de Janeiro. 2009.
COYNE, J. A.; BEECHAM, E. Heritability of two morphological characters within and among natural populations of Drosophila melanogaster. Genetics, v. 117, n. 4, p. 727–737, 1987.
CROZATIER, M. et al. Vein-positioning in the Drosophila wing in response to Hh; new roles of Notch signaling. Mechanisms of Development, v. 120, p. 529–535, 2003.
CROZATIER, M.; GLISE, B.; VINCENT, A. Patterns in evolution: Veins of the Drosophila wing. Trends in Genetics, v.20,n.10, 2004.
DARWIN, C. On the Origin of Species by Means of Natural Selection , or the Preservation of Favoured Races in the Struggle for Life. London: John Murray, Albermale Street, 1859.
DAVID, J. R. et al. Genetic variability of sexual size dimorphism in a natural population of Drosophila melanogaster : an isofemale-line approach. Journal of Genetics, v. 82, n. 3, p. 79–88, 2003.
DAVID, J. R. et al. Isofemale lines in Drosophila: an empirical approach to quantitative trait analysis in natural populations. Heredity, v. 94, n. 1, p. 3–12, 2005.
DAVID, J. R. et al. Phenotypic plasticity and developmental temperature in Drosophila: Analysis and significance of reaction norms of morphometrical traits. Journal of Thermal Biology, v. 22, n. 6, p. 441–451, 1997.
DAVID, J. R. et al. Reaction norms of size characters in relation to growth temperature in Drosophila melanogaster: an isofemale lines analysis. Genetics, selection, evolution, v. 26, n. 3, p. 229–251, 1994.
DAVID, J. R. et al. Thermal phenotypic plasticity of body size in Drosophila melanogaster: sexual dimorphism and genetic correlations. Journal of Genetics, v. 90, n. 2, p. 295–302, 2011.
R E F E R Ê N C I A S B I B L I O G R Á F I C A S
131
DAWKINS, R. The Selfish Gene. Oxford: Oxford University Press, 1976.
DE CELIS, J. F. Pattern formation in the Drosophila wing: The development of the veins. BioEssays, v. 25, n. 5, p. 443–51, 2003.
DE MOED, G. H.; DE JONG, G.; SCHARLOO, W. The phenotypic plasticity of wing size in Drosophila melanogaster: the cellular basis of its genetic variation. Heredity, v. 79 ( Pt 3), n. August 1996, p. 260–267, 1997.
DEBAT, V. et al. Allometric and nonallometric components of Drosophila wing shape respond differently to developmental temperature. Evolution, v. 57, n. 12, p. 2773–2784, 2003.
DEBAT, V. et al. Hsp90 and the quantitative variation of wing shape in Drosophila melanogaster. Evolution, v. 60, p. 2529–2538, 2006.
DEBAT, V. et al. Multidimensional analysis of Drosophila wing variation in Evolution Canyon. Journal of Genetics, v. 87, n. 4, p. 407–419, 2008.
DEBAT, V.; DEBELLE, A.; DWORKIN, I. Plasticity, canalization, and developmental stability of the Drosophila wing: joint effects of mutations and developmental temperature. Evolution, v. 63, n. 11, p. 2864–2876, 2009.
DOBZHANSKY, T. The influence of the quantity and quality of chromosomal material on the size of the cells in Drosophila melanogaster. Wilhelm Roux’ Archiv für Entwicklungsmechanik der Organismen, v. 115, n. 3, p. 363–379, 1929.
DOLEZAL, T. et al. Casein kinase i epsilon somatic mutations found in breast cancer cause overgrowth in Drosophila. International Journal of Developmental Biology, v. 54, p. 1419–1424, 2010.
ELBERSE, I. A. M. et al. Quantitative trait loci affecting growth-related traits in wild barley (Hordeum spontaneum) grown under different levels of nutrient supply. Heredity, v. 93, p. 22–33, 2004.
FALCONER, D. S.; MACKAY, T. F. C. Introduction to quantitative genetics. Harlow, Essex, UK: Logmans Green, 1996.
FOGLEMAN, J. C. A thermal gradient bar for the study of Drosophila. Drosophila Information Service, v. 53, p. 212–213, 1978.
FULLER, R. C.; BAER, C. F.; TRAVIS, J. How and when selection experiments might actually be useful. Integrative and Comparative Biology, v. 45, n. 3, p. 391–404, 2005.
GARCIA-BELLIDO, A.; DE CELIS, J. F. Developmental genetics of the venation pattern of Drosophila. Annual review of genetics, v. 26, p. 277–304, 1992.
R E F E R Ê N C I A S B I B L I O G R Á F I C A S
132
GARCIA-BELLIDO, A.; MERRIAM, J. R. Parameters of the wing imaginal disc development of Drosophila melanogaster. Developmental Biology, v. 24, n. 1, p. 61–87, 1971.
GARCIA-BELLIDO, A.; RIPOLL, P.; MORATA, G. Developmental compartmentalisation of the wing disk of Drosophila. Nature, v. 245, n. 147, p. 251–253, 1973.
GARLAND, T.; KELLY, S. A. Phenotypic plasticity and experimental evolution. The Journal of experimental biology, v. 209, n. Pt 12, p. 2344–61, 2006.
GIBERT, P. et al. Comparative analysis of morphological traits among Drosophila melanogaster and D. simulans: genetic variability, clines and phenotypic plasticity. Genetica, v. 120, n. 1-3, p. 165–79, 2004.
GIDASZEWSKI, N. A.; BAYLAC, M.; KLINGENBERG, C. P. Evolution of sexual dimorphism of wing shape in the Drosophila melanogaster subgroup. BMC Evol Biol, v. 9, p. 110, 2009.
GILBERT, S. F.; EPEL, D. Ecological Developmental Biology: Integrating Epigenetics, Medicine, and Evolution. Sunderland, MA: Sinauer Associates Inc, 2008.
GILCHRIST, A. S. et al. Adaptation and constraint in the evolution of Drosophila melanogaster wing shape. Evolution and development, v. 2, n. 2, p. 114–124, 2000.
GILCHRIST, G. W.; HUEY, R. B.; SERRA, L. Rapid evolution of wing size clines in Drosophila subobscura. Genetica, v. 112-113, p. 273–286, 2001.
GIRALDEZ, A J.; COHEN, S. M. Wingless and Notch signaling provide cell survival cues and control cell proliferation during wing development. Development, v. 130, n. 26, p. 6533–6543, 2003.
GOCKEL, J. et al. Nonclinality of molecular variation implicates selection in maintaining a morphological cline of Drosophila melanogaster. Genetics, v. 158, n. 1, p. 319–323, 2001.
GONZÁLEZ-GAITÁN, M.; CAPDEVILA, M. P.; GARCIA-BELLIDO, A. Cell proliferation patterns in the wing imaginal disc of Drosophila. Mechanisms of development, v. 46, n. 3, p. 183–200, 1994.
GRIMALDI, D.; SINGH, H. The extinct genus Pareuthychaeta in Eocene ambers (Diptera: Schizophora: Ephydroidea). The Canadian Entomologist, v. 144, n. 01, p. 17–28, 2012.
GUERRA, D. et al. Developmental constraints in the Drosophila wing. Heredity, v. 79 ( Pt 6), p. 564–571, 1997.
HALE, R. et al. Cellular interpretation of the long-range gradient of Four-jointed activity in the Drosophila wing. eLife, v. 4, 2015.
HOFFMANN, A. A.; JENNIFER SHIRRIFFS. Geographic variation for wing shape in Drosophila serrata. Evolution and development, v. 56, n. 5, p. 1068–1073, 2002.
R E F E R Ê N C I A S B I B L I O G R Á F I C A S
133
HOULE, D. et al. Automated measurement of Drosophila wings. Bmc Evolutionary Biology, v. 3, p. 1–13, 2003.
IMASHEVA, A. G. et al. Geographic differentiation in wing shape in Drosophila melanogaster. Genetica, v. 96, n. 3, p. 303–306, 1995.
JABLONKA, E.; LAMB, M. J. Evolution in four dimensions: Genetic, epigenetic, behavioral, and symbolic variation in the history of life. 1. ed. Cambridge, Massachusetts, USA: The MIT Press, 2005.
JOHANNES, B.; PREISS, A. Wing vein formation in Drosophila melanogaster: Hairless is involved in the cross-talk between Notch and EGF signaling pathways. Mechanisms of Development, v. 115, p. 3–14, 2002.
KANCA, O. et al. Raeppli: a whole-tissue labeling tool for live imaging of Drosophila development. Development, v. 141, n. 2, p. 472–80, 2014.
KARAN, D. et al. The genetics of phenotypic plasticity. IX. Genetic architecture, temperature, and sex differences in Drosophila melanogaster. Evolution, v. 54, n. 3, p. 1035–40, 2000.
KARAN, D.; MORETEAU, B.; DAVID, J. R. Growth temperature and reaction norms of morphometrical traits in a tropical drosophilid: Zaprionus indianus. Heredity, v. 83 ( Pt 4), n. June, p. 398–407, 1999.
KATAYAMA, N. et al. Sexual selection on wing interference patterns in Drosophila melanogaster. Proceedings of the National Academy of Sciences of the United States of America, v. 111, n. 42, p. 15144–8, 2014.
KING, M. C.; WILSON, A. C. Evolution at two levels in humans and chimpanzees. Science, v. 188, n. 4184, p. 107–116, 1975.
KLACZKO, L. B. Evolutionary genetics of Drosophila mediopunctata. Genetica, v. 126, n. 1-2, p. 43–55, 2006.
KLACZKO, L. B.; BITNER-MATHÉ, B. C. On the edge of a wing. Nature, v. 346, p. 231, 1990.
KLEPSATEL, P. et al. Similarities and differences in altitudinal versus latitudinal variation for morphological traits in Drosophila melanogaster. Evolution, v. 68, n. 5, p. 1385–98, 2014.
KLIEBENSTEIN, D. J.; FIGUTH, A.; MITCHELL-OLDS, T. Genetic architecture of plastic methyl jasmonate responses in Arabidopsis thaliana. Genetics, v. 161, p. 1685–1696, 2002.
KLINGENBERG, C. P. Morphometric integration and modularity in configurations of landmarks: tools for evaluating a priori hypotheses. Evolution & Devevelopment, v. 11, n. 4, p. 405–421, 2009.
R E F E R Ê N C I A S B I B L I O G R Á F I C A S
134
KLINGENBERG, C. P. Morphometrics and the role of the phenotype in studies of the evolution of developmental mechanisms. Gene, v. 287, p. 3–10, 2002.
KLINGENBERG, C. P. Studying morphological integration and modularity at multiple levels: concepts and analysis. Philosophical transactions of the Royal Society of London, v. 369, p. 2010249, 2014.
KLINGENBERG, C. P.; AKLAN, S. T. D. Z. Morphological integration between developmental compartments in the Drosophila wing. Evolution, v. 54, n. 4, p. 1273–1285, 2000.
KOLZER, S. Defective proventriculus is required for pattern formation along the proximodistal axis, cell proliferation and formation of veins in the Drosophila wing. Development, v. 130, n. 17, p. 4135–4147, 2003.
LAAYOUNI, H. et al. Thermal evolution of gene expression profiles in Drosophila subobscura. BMC Evolutionary Biology, v. 7, p. 42, 2007.
LE ROUZIC, A.; HOULE, D.; HANSEN, T. F. A modelling framework for the analysis of artificial-selection time series. Genetics Research, v.93, n.2, p155-173, 2011.
LECUIT, T.; LE GOFF, L. Orchestrating size and shape during morphogenesis. Nature, v. 450, n. 7167, p. 189–92, 2007.
LECUIT, T.; LENNE, P. F. Cell surface mechanics and the control of cell shape, tissue patterns and morphogenesis. Nature Reviews Molecular Cell Biology, v. 8, n. 8, p. 633–644, ago. 2007.
LEGOFF, L.; ROUAULT, H.; LECUIT, T. A global pattern of mechanical stress polarizes cell divisions and cell shape in the growing Drosophila wing disc. Development, v. 140, n. 19, p. 4051–9, 2013.
LEIBOWITZ, A.; SANTOS, M.; FONTDEVILA, A. Heritability and selection on body size in a natural population of Drosophila buzzatii. Genetics, v. 141, n. 1, p. 181–189, 1995.
LI, W.-H.; SAUNDERS, M. A. The chimpanzee and us. Nature, v. 437, p. 50–51, 2005.
LIEFTING, M.; HOFFMANN, A.; ELLERS, J. Plasticity versus environmental canalization: population differences in thermal responses along a latitudinal gradient in Drosophila serrata. Evolution, v. 63, n. 8, p. 1954–1963, 2009.
LINZ, D. M.; TOMOYASU, Y. RNAi screening of developmental toolkit genes: a search for novel wing genes in the red flour beetle, Tribolium castaneum. Development genes and evolution, v. 225, n. 1, p. 11–22, 2015.
LOH, R. et al. Adaptation to different climates results in divergent phenotypic plasticity of wing size and shape in an invasive drosophilid. Journal of Genetics, v. 87, n. 3, p. 209–17,. 2008.
R E F E R Ê N C I A S B I B L I O G R Á F I C A S
135
LOH, R.; BITNER-MATHÉ, B. C. Variability of wing size and shape in three populations of a recent Brazilian invader, Zaprionus indianus (Diptera: Drosophilidae), from different habitats. Genetica, v. 125, n. 2-3, p. 271–81, 2005.
MARENDA, D. R. et al. MAP kinase subcellular localization controls both pattern and proliferation in the developing Drosophila wing. Development, v. 133, n. 1, p. 43–51, 2006.
MATTA, B. P. Identificaçao de lócus candidatos ao controle da variação de tamanho e forma nas asas de Drosophila, em especies do grupo melanogaster. 2010. Tese de Doutorado, Departamento de Genética, UFRJ, 2010.
MATTA, B. P.; BITNER-MATHÉ, B. C. An interspecific QTL study of Drosophila wing size and shape variation to investigate the genetic basis of morphological differences. Genetics and molecular research, v. 9, n. 4, p. 2032–49, 2010.
MATTA, B. P.; BITNER-MATHÉ, B. C. Genetic architecture of wing morphology in Drosophila simulans and an analysis of temperature effects on genetic parameter estimates. Heredity, v. 93, n. 4, p. 330–41, 2004.
MATTA, B. P.; BITNER-MATHÉ, B. C.; ALVES-FERREIRA, M. Getting real with real-time qPCR: a case study of reference gene selection for morphological variation in Drosophila melanogaster wings. Development genes and evolution, v. 221, n. 1, p. 49–57, 2011.
MATUTE, D. R.; HARRIS, A. The influence of abdominal pigmentation on desiccation and ultraviolet resistance in two species of Drosophila. Evolution, v. 67, n. 8, p. 2451–2460, 2013.
MEDICI, V. et al. The FlyCatwalk: A High-Throughput Feature-Based Sorting System for Artificial Selection in Drosophila. G3, v. 5, n. 3, p. 317–27, 2015.
MENEZES, B. Influência da Forma das Asas para o Sucesso no Acasalamento em Drosophila melanogaster. 2007. Monografia, Departamento de Genética, UFRJ, 2007.
MENEZES, B. F. et al. The influence of male wing shape on mating success in Drosophila melanogaster. Animal Behaviour, v. 85, n. 6, p. 1217–1223, 2013.
MORAES, E. M. et al. Wing shape heritability and morphological divergence of the sibling species Drosophila mercatorum and Drosophila paranaensis. Heredity, v. 92, n. 5, p. 466–73, 2004.
MORAES, E. M.; SENE, F. M. Heritability of wing morphology in a natural population of Drosophila gouveai. Genetica, v. 121, p. 119–123, 2004.
MORATA, G. How Drosophila appendages develop. Nature reviews: Molecular cell biology, v. 2, n. 2, p. 89–97, 2001.
R E F E R Ê N C I A S B I B L I O G R Á F I C A S
136
MORIN, J. P. et al. Divergence of reaction norms of size characters between tropical and temperate populations of Drosophila melanogaster and D. simulans. Journal of Evolutionary Biology, v. 12, n. 2, p. 329–339, 1999.
NETO-SILVA, R. M.; WELLS, B. S.; JOHNSTON, L. A. Mechanisms of growth and homeostasis in the Drosophila wing. Annual review of cell and developmental biology, v. 25, p. 197–220, 2009.
NETTLE, D.; CLEGG, H. Schizotypy, creativity and mating success in humans. Proceedings. Biological sciences / The Royal Society, v. 273, n. 1586, p. 611–615, 2006.
NEUFELD, T. P. et al. Coordination of growth and cell division in the Drosophila wing. Cell, v. 93, n. 7, p. 1183–1193, 1998.
NIEHRS, C. On growth and form: a Cartesian coordinate system of Wnt and BMP signaling specifies bilaterian body axes. Development, v. 137, p. 845–857, 2010.
NIJHOUT, H. F. The control of body size in insects. Developmental Biology, v.261,n.1,p1-9, 2003.
NOACH, E. J. K.; DE JONG, G.; SCHARLOO, W. Phenotypic plasticity of wings in selection lines of Drosophila melanogaster. Heredity, v. 79, p. 1–9, 1997.
O’KEEFE, D. D. et al. Combinatorial control of temporal gene expression in the Drosophila wing by enhancers and core promoters. BMC genomics, v. 13, n. 1, p. 498, 2012.
OBBARD, D. J. et al. Estimating divergence dates and substitution rates in the Drosophila phylogeny. Molecular Biology and Evolution, v. 29, p. 3459–3473, 2012.
OLGUIN, P.; MLODZIK, M. A new spin on planar cell polarity. Cell, v. 142, n. 5, p. 674–676, set. 2010.
PARTRIDGE, L. et al. Correlated responses to selection on body size in Drosophila melanogaster. Genetical Research, v. 74, n. 1, p. 43–54, 1999.
PEARLSON, G. D.; FOLLEY, B. S. Schizophrenia, psychiatric genetics, and Darwinian psychiatry: An evolutionary framework. Schizophrenia Bulletin , v.34, n.4, p. 722-733, 2008.
PÉLABON, C. et al. Response of fluctuating and directional asymmetry to selection on wing shape in Drosophila melanogaster. Journal of Evolutionary Biology, v. 19, p. 764–776, 2006.
PERTOLDI, C. et al. The phenotypic variance gradient - a novel concept. Ecology and evolution, v. 4, n. 22, p. 4230–4236, 2014.
PÉTAVY, G. et al. Growth temperature and phenotypic plasticity in two Drosophila sibling species : probable adaptive changes in flight capacities. Journal of Evolutionary Biology, v. 10, p. 875–887, 1997.
R E F E R Ê N C I A S B I B L I O G R Á F I C A S
137
PEZZOLI, M. C. et al. Developmental constraints and wing shape variation in natural populations of Drosophila melanogaster. Heredity, v. 79, n. 1996, p. 572–577, 1997.
PHILLIPS, P. C.; WHITLOCK, M. C.; FOWLER, K. Inbreeding Changes the Shape of the Genetic Covariance Matrix in Drosophila melanogaster. Genetics, 2001.
PIGLIUCCI, M. Evolution of phenotypic plasticity: where are we going now? Trends in ecology & evolution, v. 20, n. 9, p. 481–486, 2005.
PIGLIUCCI, M. Phenotipic plasticity: beyond nature and nurture. Baltimore: The John Hopkins University Press, 2001.
PIGLIUCCI, M.; MURREN, C. J.; SCHLICHTING, C. D. Phenotypic plasticity and evolution by genetic assimilation. The Journal of experimental biology, v. 209, n. Pt 12, p. 2362–7, 2006.
PITCHERS, W.; POOL, J. E.; DWORKIN, I. Altitudinal clinal variation in wing size and shape in African Drosophila melanogaster: one cline or many? Evolution, v. 67, n. 2, p. 438–452, 2013.
POWELL, J. R. Progress and Prospects in Evolutionary Biology: The Drosophila Model. New York: Oxford University Press, 1997.
PROUT, M. et al. Autosomal mutations affecting adhesion between wing surfaces in Drosophila melanogaster. Genetics, v. 146, n. 1, p. 275–285, 1997.
RAMP, C. et al. Adapting to a warmer ocean-seasonal shift of baleen whale movements over three decades. PloS one, v. 10, n. 3, 2015.
RESTREPO, S.; ZARTMAN, J. J.; BASLER, K. Coordination of patterning and growth by the morphogen DPP. Current biology : CB, v. 24, n. 6, p. R245–R255, 2014.
RIDLEY, M. Evolution. Oxford: 3. ed, Blackwell Publishing, 2004.
ROCHA, F. B.; KLACZKO, L. B. Connecting the dots of nonlinear reaction norms unravels the threads of genotype-environment interaction in Drosophila. Evolution, v. 66, n. 11, p. 3404–16, 2012.
ROCHA, F.; MEDEIROS, H. F.; KLACZKO, L. B. The reaction norm for abdominal pigmentation and its curve in Drosophila mediopunctata depend on the mean phenotypic value. Evolution, v. 63, n. 1, p. 280–287, 2009.
ROCHA, Felipe Bastos; KLACZKO, Louis Bernard. Undesirable consequences of neglecting nonlinearity: Response to comments by liefting et al. (2013) on Rocha & Klaczko (2012). Evolution, 2014.
ROGULJA, D.; RAUSKOLB, C.; IRVINE, K. D. Morphogen control of wing growth through the Fat signaling pathway. Developmental Cell, v. 15, n. 2, p. 309–321, 2008.
R E F E R Ê N C I A S B I B L I O G R Á F I C A S
138
RUSSO, C. A M.; TAKEZAKI, N.; NEI, M. Molecular phylogeny and divergence times of drosophilid species. Molecular biology and evolution, v. 12, n. 3, p. 391–404, 1995.
SARKAR, S. From the reaktionsnorm to the adaptive norm: the norm of reaction, 1909–1960. Biology and Philosophy, v. 14, n. 2, p. 235–252, 1999.
SCHEINER, S. M. Plasticity as a selectable trait: reply to Via. American Naturalist, v. 142, n. 2, p. 371–373, 1993.
SCHLICHTING, C. The Evolution of Phenotypic Plasticity in Plants. Annual Review of Ecology and Systematics. , 1986
SCHOU, M. F.; KRISTENSEN, T. N.; LOESCHCKE, V. Trait-specific consequences of inbreeding on adaptive phenotypic plasticity. Ecology and evolution, v. 5, n. 1, p. 1–6, 2015.
SCHWANK, G. et al. Antagonistic Growth Regulation by Dpp and Fat Drives Uniform Cell Proliferation. Developmental Cell, v. 20, n. 1, p. 123–130, 2011.
SHAO, H. et al. Genetic architecture of complex traits: large phenotypic effects and pervasive epistasis. Proceedings of the National Academy of Sciences of the United States of America, v. 105, n. 50, p. 19910–19914, 2008.
SHINJI IJICHI, NAOMI IJICHI, YUKINA IJICHI, H. S. AND H. M. Autism Spectrum Disorders: The Role of Genetics in Diagnosis and Treatment. InTech, 2011.
SOKAL, R. R.; ROHLF, F. J. Biometry. 2nd. ed. New York, NY: Freeman, 1981.
SOTO, I. M. et al. Wing morphology and fluctuating asymmetry depend on the host plant in cactophilic Drosophila. Journal of evolutionary biology, v. 21, n. 2, p. 598–609, 2008..
ST JOHNSTON, D. The art and design of genetic screens: Drosophila melanogaster. Nature reviews. Genetics, v. 3, n. 3, p. 176–188, 2002.
STEPPAN, S. J.; PHILLIPS, P. C.; HOULE, D. Comparative quantitative genetics: evolution of the G matrix. Trends in Ecology & Evolution, v. 17, n. 7, p. 320–327, 2002.
STINCHCOMBE, J. R.; DORN, L. A.; SCHMITT, J. Flowering time plasticity in Arabidopsis thaliana: A reanalysis of Westerman & Lawrence (1970). Journal of Evolutionary Biology, v. 17, p. 197–207, 2004.
STRIGINI, M.; COHEN, S. M. Formation of morphogen gradients in the Drosophila wing. Seminars in cell developmental biology, v. 10, n. 3, p. 335–344, 1999.
SUGIMURA, K.; ISHIHARA, S. The mechanical anisotropy in a tissue promotes ordering in hexagonal cell packing. Development, v. 140, n. 19, p. 4091–101, 2013.
R E F E R Ê N C I A S B I B L I O G R Á F I C A S
139
TAYLOR, H. S. A regulatory element of the empty spiracles homeobox gene is composed of three distinct conserved regions that bind regulatory proteins. Molecular reproduction and development, v. 49, n. 3, p. 246–253, 1998.
TAYLOR, J.; ADLER, P. N. Cell rearrangement and cell division during the tissue level morphogenesis of evaginating Drosophila imaginal discs. Developmental Biology, v. 313, n. 2, p. 739–751, 2008.
TESSEROLI, D. Seleção direcional sobre as formas das asas de Drosophila melanogaster. 2005. Dissertação de Mestrado, Departamento de Genética, UFRJ, 2005.
TORQUATO, L. et al. Cellular basis of morphological variation and temperature-related plasticity in Drosophila melanogaster strains with divergent wing shapes. Genetica, p. 1–11, 2014.
TROTTA, V. et al. Allometric and non-allometric consequences of inbreeding on Drosophila melanogaster wings. Biological Journal of the Linnean Society, v. 102, p. 626–634, 2011.
TROTTA, V. et al. Thermal plasticity in Drosophila melanogaster: a comparison of geographic populations. BMC evolutionary biology, v. 6, p. 67, 2006.
TROTTA, V. et al. Thermal plasticity of wing size and shape in Drosophila melanogaster, D. simulans and their hybrids. Climate Research, v. 43, p. 71–79, 2010.
TSUJINO, M.; TAKAHASHI, K. H. Lack of response to artificial selection on developmental stability of partial wing shape components in Drosophila melanogaster. Genetica, v. 142, n. 2, p. 177–184, 2014.
UMETSU, D.; DUNST, S.; DAHMANN, C. An RNA interference screen for genes required to shape the anteroposterior compartment boundary in Drosophila identifies the Eph receptor. PloS one, v. 9, n. 12, p. e114340, 2014.
VIA, S. et al. Adaptive phenotypic plasticity: consensus and controversy. Trends in ecology & evolution, v. 10, n. 5, p. 212–217, 1995.
VIA, S. Adaptive phenotypic plasticity: target or by-product of selection in a variable environment? The American Naturalist, v. 142, n. 2, p. 352–365, 1993.
VIJENDRAVARMA, R. K.; NARASIMHA, S.; KAWECKI, T. J. Plastic and evolutionary responses of cell size and number to larval malnutrition in Drosophila melanogaster. Journal of evolutionary biology, v. 24, n. 4, p. 897–903, 2011.
WAGNER, G. P.; ALTENBERG, L. Perspective: complex adaptations and the evolution of evolvability. Evolution, v. 50, p. 967–976, 1996.
WAGNER, G. P.; PAVLICEV, M.; CHEVERUD, J. M. The road to modularity. Nature reviews. Genetics, v. 8, n. 12, p. 921–31, 2007.
R E F E R Ê N C I A S B I B L I O G R Á F I C A S
140
WALLACE, A. R. On the tendency of varieties to depart indefinitely from the original type ( 1858 ). Proceedings of the Linnean Society Of London, 1858.
WATERLAND, R. A.; JIRTLE, R. L. Transposable elements: targets for early nutritional effects on epigenetic gene regulation. Molecular and Cellular Biology, v. 23, n. 15, p. 5293–5300, 2003.
WEBER, K. et al. An Analysis of Polygenes Affecting Wing Shape on Chromosome 3 in Drosophila melanogaster. Genetics, v. 786, p. 773–786, 1999.
WEBER, K. E. How small are the smallest selectable domains of form? Genetics, v. 130, p. 345–353, 1992.
WEBER, K. E. et al. Microarray analysis of replicate populations selected against a wing-shape correlation in Drosophila melanogaster. Genetics, v. 178, n. 2, p. 1093–1108, 2008.
WEBER, K. E. Selection on wing allometry in Drosophila melanogaster. Genetics, v. 126, n. 4, p. 975–989, 1990.
WEST-EBERHARD, M. J. Developmental plasticity and evolution. New York: Oxford University Press, 2003.
WILKINS, J. F. Genomic imprinting and conflict-induced decanalization. Evolution, v. 65, n. 2, p. 537–53, 2011.
WITTKOPP, P. J.; HAERUM, B. K.; CLARK, A. G. Regulatory changes underlying expression differences within and between Drosophila species. Nature genetics, v. 40, n. 3, p. 346–50, 2008.
WOLTERECK, R. Weitere experimentelle Untersuchungen über Artveränderung, speziel über das Wesen quantitativer Artunterschiede bei Daphnien. Verhandlungen der deutschen zoologischen Gesellschaf, v. 19, p. 110–173, 1909.
YEATES, D. K.; WIEGMANN, B. M. Congruence and controversy: toward a higher-level phylogeny of Diptera. Annual review of entomology, v. 44, p. 397–428, 1999.
YEH, S.-D.; TRUE, J. R. The genetic architecture of coordinately evolving male wing pigmentation and courtship behavior in Drosophila elegans and Drosophila gunungcola. G3 , v. 4, n. 11, p. 2079–93, 2014.
ZAR, J. H. Biostatistical Analysis. 4th. ed. New Jersey: Prentice Hall, 2010.
ZEUSS, D. et al. Global warming favours light-coloured insects in Europe. Nature communications, v. 5, n. May 2013, p. 3874, 2014.
ZIMMERMAN, E.; PALSSON, A.; GIBSON, G. Quantitative trait loci affecting components of wing shape in Drosophila melanogaster. Genetics, v. 155, n. 2, p. 671–683, 2000.