Modern petunia varieties draw your attention with names such as ‘night sky’, ‘suncatcher pink lemonade’ or ‘kandy kane’, names with which their wild counterparts cannot compete. However, wild Petunia species do know how to be noticed as well – that is to say, by pollinators. For example, Petunia axillaris has large white flowers with petals that form a long and narrow corolla tube, which spread a strong fragrance during the night to attract nocturnally active hawkmoths (Ando et al., 2001). Another species, Petunia exserta, has bright red flowers that produce large amounts of nectar that attract hummingbirds (Stehmann et al., 2009). In fact, different species in the Petunia genus attract either hawkmoths, bees or hummingbirds as pollinators, depending on the features of the flower (Galliot et al., 2006). Animal pollinators are thought to exert selection pressure on floral traits, which drives the formation of new species (Fenster et al., 2004). To understand the mechanisms underlying speciation, evolutionary biologists are trying to identify the so‐called ‘speciation genes’. The variety in flowers that attract different types of pollinators makes petunias particularly suited for this search. Moreover, Petunia species can be easily crossed and yield normal, diploid offspring (Vandenbussche et al., 2016). Cris Kuhlemeier, in Bern, Switzerland, is one of the biologists investigating pollinator‐mediated speciation in petunia. Several years ago, his group tried to identify genes that cause differences in style length. The shape and size of the reproductive organs in petunia play an important role in efficient pollen transfer by its pollinator. One of the organs that varies in length between species that attract different pollinators is the style. Hence, style length is likely controlled by speciation genes. To discover these genes, the group applied quantitative trait loci (QTL) analysis but found that at least at one of the QTLs, recombination was severely suppressed, which made further fine‐mapping problematic. As it is becoming clear that suppression of recombination is a common feature of interspecific crosses, the authors decided to try a different approach. In this issue, they present a workflow that allowed them to identify a set of candidate genes involved in petunia style length (Yarahmadov et al., 2020).

The authors used three different Petunia species with different style lengths, i.e. two (P. axillaris and Petunia parodii) pollinated by hawkmoths and one (P. exserta) pollinated by hummingbirds (Figure). They first investigated how petunia style growth works on a mechanistic basis, and then they analysed growth rate, cell wall elasticity and cell division rate, and defined different developmental stages of style growth. Subsequently, they gathered gene expression data from each of the different stages and performed correlation analysis between each gene expression profile and style growth. Approaches that combine expression analysis with phenotypic data usually yield large sets of genes that are correlated with the phenotype, and finding the causative genes can be challenging. Therefore, the authors followed a more systematic approach by applying filters to narrow down the number of genes (Figure 1). Because transcription factors are overrepresented among genes underlying the evolution of plant development (Doebley and Lukens, 1998), they excluded all genes that did not encode transcription factors, and then they applied two different filters on the remaining set of genes. First, they selected genes that were differentially regulated due to cis factors. Differences in gene expression caused by trans factors cannot be explained by mutations in the genes themselves, but by differences in upstream factors. Hence, trans‐regulated genes are not candidate speciation genes. To identify which genes were regulated in cis or in trans, the authors generated interspecific F₁ hybrids to perform allele‐specific expression analysis. When the differential regulation of a gene between two species is caused by a cis factor, the two alleles will still show differential expression in the F₁ hybrid, whereas the two alleles will show equal expression when it is caused by a trans factor. Only 37 genes displayed strong allele‐specific expression, showing that this filtering step was important in reducing the number of genes. Second, the authors searched for genes that might encode proteins with diverged functions. The authors detected high‐impact single nucleotide polymorphisms, such as frameshifts and loss and gain of start and stop codons, in six of the genes encoding differentially expressed transcription factors, bringing the total list of candidate speciation genes to 43. To validate their approach, they performed virus‐induced gene silencing for six genes. For two of them, encoding the MYB transcription factors EOBI and EOBII, silencing caused a significant reduction in style length, which indicated that their method was effective.

The lab of Cris Kuhlemeier has two major research interests. Besides speciation in petunia, part of the group works on understanding phyllotaxis. They work with Arabidopsis and use computational, biomechanical and image tracking tools to uncover how plant organs are arranged in regular patterns. For their new approach to find speciation genes in petunia, two parts of the lab joined forces. Scientists that would normally work on how cell and organ size are determined were particularly instrumental in understanding how petunia styles obtain their final size. The researchers think that this alternative to QTL analysis should be widely applicable. And who knows, one day a petunia breeder might develop a new variety while using these new findings. There is no doubt this petunia will have a poetic name.

现代矮牵牛品种以“夜空”，“粉红色的柠檬蜜饯”或“康提凯恩”等名称引起您的注意，而野生品种则无法与之竞争。但是，野生矮牵牛物种确实也知道如何被传粉者注意到，也就是说。例如，矮牵牛具有白色的大花，花瓣形成长而狭窄的花冠管，在夜间散发出浓烈的香气，以吸引夜间活动的鹰蛾（Ando等，2001）。另一种是矮牵牛（Petunia exserta），具有鲜艳的红色花朵，会产生大量花蜜，吸引蜂鸟（Stehmann等，2009）。）。实际上，矮牵牛属中的不同物种会吸引鹰蛾，蜜蜂或蜂鸟作为授粉媒介，这取决于花的特征（Galliot等，2006）。动物传粉者被认为会对花卉性状施加选择压力，从而驱动新物种的形成（Fenster等，2004）。为了理解物种形成的机制，进化生物学家正在尝试鉴定所谓的“物种形成基因”。吸引不同类型传粉媒介的花朵种类繁多，使得矮牵牛特别适合这种搜索。此外，矮牵牛种很容易杂交并产生正常的二倍体后代（Vandenbussche等。，2016）。瑞士伯尔尼的克里斯·库勒迈耶（Cris Kuhlemeier）是调查矮牵牛授粉媒介介导物种的生物学家之一。几年前，他的小组试图鉴定出导致花型长度差异的基因。矮牵牛生殖器官的形状和大小在其传粉者有效的花粉转移中起重要作用。在吸引不同传粉媒介的物种之间长度变化的器官之一是花柱。因此，样式长度很可能受物种基因控制。为了发现这些基因，该研究小组应用了数量性状基因座（QTL）分析，但发现至少在其中一个QTL中，重组被严重抑制，这使进一步的精细映射成为问题。显而易见，抑制重组是种间杂交的共同特征，作者决定尝试另一种方法。在本期杂志中，他们提出了一种工作流程，使他们能够鉴定出一组与矮牵牛花型长度有关的候选基因（Yarahmadov等。，2020年）。

作者使用了3种不同样式长度的矮牵牛物种，即两种（P. axillaris和Petunia parodii）被鹰蛾授粉，另一种（P. exserta）。）被蜂鸟授粉（图）。他们首先研究了矮牵牛花式生长的机理，然后分析了生长速率，细胞壁弹性和细胞分裂速率，并定义了花式生长的不同发育阶段。随后，他们收集了每个不同阶段的基因表达数据，并进行了每个基因表达谱与样式生长之间的相关性分析。将表达分析与表型数据相结合的方法通常会产生大量与表型相关的基因，而寻找致病基因可能具有挑战性。因此，作者采用了一种更系统的方法，即通过应用过滤器来缩小基因数量（图1）。1998年），他们排除了所有不编码转录因子的基因，然后对其余的基因套用了两个不同的过滤器。首先，他们选择了由于顺式因子而受到不同调控的基因。反式因子引起的基因表达差异不能用基因本身的突变来解释，而不能用上游因子的差异来解释。因此，反式调控基因不是候选物种形成基因。为了确定顺式或反式调控的基因，作者生成了种间F ₁杂种执行等位基因特异性表达分析。当两个物种之间的基因差异调节是由顺式因子引起的时，两个等位基因在F ₁杂种中仍将显示差异表达，而当两个等位基因由反式引起时将显示相同表达。因子。只有37个基因显示出强大的等位基因特异性表达，表明这一过滤步骤对于减少基因数量很重要。其次，作者寻找可能编码具有不同功能的蛋白质的基因。作者在编码差异表达转录因子的六个基因中检测到了高影响力的单核苷酸多态性，例如移码以及起始密码子和终止密码子的丢失和获得，使候选物种形成基因的总数达到43个。为验证其方法，他们对六个基因进行了病毒诱导的基因沉默。对于其中的两个，编码MYB转录因子EOBI和EOBII，沉默导致样式长度显着减少，这表明它们的方法有效。

Cris Kuhlemeier的实验室有两个主要的研究兴趣。除了矮牵牛的物种形成外，该小组的一部分还致力于了解叶序。他们与拟南芥一起工作，并使用计算，生物力学和图像跟踪工具来发现植物器官如何以规则的方式排列。为了在矮牵牛中寻找物种形成基因的新方法，实验室的两个部分联合起来。通常会研究如何确定细胞和器官大小的科学家对理解矮牵牛花型如何获得最终大小特别有用。研究人员认为，这种替代QTL分析的方法应该广泛适用。谁知道，有一天，矮牵牛育种者可能会利用这些新发现开发出一种新品种。毫无疑问，这个矮牵牛将有一个诗意的名字。