Regulatory evolution acts by using available preexisting genetic elements to generate novelty. Likewise, the synthetic genotype created here is achieved by implementing known elements, however their selection and specific arrangement establishes a previously unknown synthetic species barrier ().
Design of a genetic circuit with selected components that form a synthetic species barrier.
The first element consist of null mutations in the glass (gl)
. The glass
product is a transcription factor of 604 amino acids with five zinc-fingers. Mutations in gl
specifically abolish photoreceptor cells resulting in blind, but viable flies 
. The gl60J
allele is a spontaneous mutant caused by the insertion of 30 kb of unknown DNA into the gl
locus and it is believed to be a null allele 
. Other alleles (gl3
) have also been used for this study.
The second element is formed by the Glass Multimer Reporter (GMR) 
, a heterologous promoter construct containing five tandem copies of a 27-bp glass
-binding site normally present in the regulatory region of ninaE
, the major rhodopsin
gene in Drosophila
. The GMR promoter can therefore drive glass
-dependent expression in the photoreceptor cells of Drosophila
The yeast protein GAL4 as a third building block can activate transcription in Drosophila
from promoters that bear GAL4 binding sites 
. In addition, the GMR sequence has been previously subcloned in front of gal4
, thus driving Gal4 expression under the control of Glass (GMR-gal4
Fourth, a tandem array of five GAL4 binding sites (5×UAS, for Upstream Activation Sequence) is employed where GAL4 binds with high affinity to induce the transcription of a downstream located gene.
The fifth element is a rasv12
allele, a mutant form of the Drosophila ras
. Conversion of the glycine residue at position 12 to valine constitutively activates the Ras protein. rasv12
has been previously subcloned behind GAL4 binding sites (UAS-rasv12
), which permits activation only within cells where GAL4 is expressed 
The inherent logic of the design relies on a “killing module” () and a regulator to switch it ON and OFF ():
The killing module is formed by GMR-gal4 and UAS-rasv12 whose activation is controlled by the presence or absence of the gene glass. When glass gene function is unperturbed, transcription of UAS-rasv12 driven by GMR-gal4 consistently kills 100% of the flies at any temperature from 17°C until 29°C (606/606 lethality at 17°C, 558/558 lethality at 23°C, 330/330 at 25°C, 110/110 at 29°C, ). Pupae arrest at mid pupation and due to abnormal tissue lysis the pupal case ends up almost empty.
Crosses between Drosophila synthetica and Drosophila melanogaster.
More than 20 other UAS
transgenes were tested, including UAS-caudal
, but UAS-rasv12
was the only one that resulted in 100% lethality when driven by the presence of glass
and the GMR-gal4
transgene at all temperatures (from 17°C to 29°C) ().
In the synthetic genotype GMR-gal4/GMR-gal4; UAS-rasv12, gl60J/UAS-rasv12, gl60J
mutant background, where no Glass protein is present) Rasv12
cannot be produced (OFF state, ). Surprisingly however, in addition to the small eye phenotype (), those flies showed a different wing morphology, with lateral extra veins (, compare with the wt wing pattern shown in 2c). Other alleles (gl3
) were also tested and yielded the same phenotype. Most likely the heat shock promoter (hsp70
) of the GMR-gal4
construct is leaky, leading to very low activation of UAS-rasv12
and consequently to the phenotype 
Morphological traits of Drosophila synthetica.
When hybrids between Drosophila melanogaster
and the synthetic genotype are produced, the “killing module” GMR-gal4; UAS-rasv12
is triggered by the presence of the glass
gene (, , ). This genetic network, while still allowing normal reproduction among flies with the synthetic genotype, completely isolates GMR-gal4/GMR-gal4; UAS-rasv12, gl60J/UAS-rasv12, gl60J
flies from normal D. melanogaster
due to hybrid early pupal lethality (, ). Unlike with the other known and naturally occurring speciation mutations 
, the sex of the parents did not influence the lethality of the hybrids in this case (, ). Experiments were performed at 17°C because flies of the synthetic genotype grew better and because due to the temperature sensitiveness of the Gal4 it is likely to be the temperature at which the killing module may be less effective. Despite this, the killing module was 100% effective even at 17°C (, ). In the initial population mutations in yellow (y1)
, which results in mild pigmentation, existed as a polymorphism in some individuals.
Creation of species boundaries by regulatory evolution.
It is often difficult to delineate “species boundaries” since they may carry identical mutations and are related to one another through common ancestors. However, most biologists agree on a very stringent definition for species, the Ernst Mayr’s Biological Species Concept, according to which species consist of populations of organisms that can reproduce with one another, but are reproductively isolated from other such groups 
(). This definition leads to a focus on the barriers to reproduction between species 
. Such barriers represented one of the main problems for Darwin who wrote: “How can we account for species, when crossed, being sterile (…), whereas, when varieties are crossed, their fertility is unimpaired?” 
. Because the postzygotically isolated population generated here conforms to the most stringent definition of species 
, it will subsequently be called Drosophila synthetica
To further prove that the synthetic genetic network allowed zero gene flow with D. melanogaster, co-cultures of both populations were performed for 13 generations (using D.melanogaster white (w) mutants with white eyes) and not a single hybrid was recovered (, ). Hybrids would have been easily recognizable by normally sized red eyes, because they would carry a normal copy of gl and two w+ copies from the transgenes (), but D. synthetica behaved like a stable species, did not interbreed and maintained its characteristic eyes (). Identical results were obtained when crossing D. synthetica with other melanogaster strains, including y,w flies and y,w,f flies (). In all cases synthetica and melanogaster did not interbreed ().
Assembling synthetic species boundaries can have practical applications. For example, the use of recombinant DNA technology to alter organisms for a specific purpose has raised controversy 
and is a growing problem due to the increasing number of transgenic organisms approved by regulatory agencies 
. A new framework where safety mechanisms are genetically designed along with desired modification could help to gain public support for a technology with the potential to satisfy future medical and nutritional needs 
. D. synthetica
is the first transgenic organism that cannot reproduce with the original wildtype
population. I therefore propose that synthetic species barriers may serve to compartmentalize dangers and protect natural species from interbreeding with emergent transgenic forms, therefore preserving natural biodiversity ().
Moreover, once a genetic network is identified, as is the case for the “ras-glass
” synthetic boundary described here, opening or closing of the barrier can be controlled at will. In case the interbreeding of populations appears beneficial, targeted strategies can be implemented to reverse hybridization barriers. To test this experimentally, the GAL4-inhibitor GAL80 was expressed from a tubulin
in D. melanogaster
in order to remove hybrid lethality and traverse the species barrier. Males of D. synthetica
hybridized successfully with tub-gal80 D. melanogaster
females and produced viable hybrids (), as predicted because Gal80 can block the “killing module”.