In all Arthropods, the steroid hormone ecdysone (E) acts to control molting and metamorphosis (Qu et al. 2015). During the past two decades, many of the enzymatic steps that convert dietary cholesterol to ecdysone in the insect Drosophila melanogaster have been identified and their activities ordered within a biosynthetic pathway. (Figure 1A, Chávez et al. 2000, Warren et al. 2002, Warren et al. 2004, Niwa et al. 2004, Yoshiyama-Yanagawa et al. 2011). Collectively, these genes are referred to as the Halloween group (Warren et al. 2002) and include phantom (phm), disembodied (dib), shadow (sad), shroud (sro), and spook (spo) which were first identified in the embryonic lethal screens of Nüsslein-Volhard, Wieschaus, and Jürgens as mutants that failed to produce differentiated cuticle (Nüsslein-Volhard et al. 1984, Wieschaus et al. 1984, Jürgens et al. 1984). A closer examination of the embryo proper revealed that null mutants in these genes all exhibit a common phenotype characterized by arrest at embryonic stages 15-17 with failures of head involution, dorsal closure, and cuticle production (Chávez et al. 2000, Warren et al. 2002, Petryk et al. 2003, Niwa et al. 2004, Warren et al. 2004). Subsequent genetic and biochemical analysis demonstrated that each of these gene products acts at a specific step in the biosynthetic pathway and their loss results in failure to synthesize ecdysone, and hence they all produce an identical loss-of-function phenotype.

Despite this progress, there remains uncertainty in the enzymes and reactions involved in an early step of the pathway termed the “Black Box”. The name arises because the chemical intermediates between the upstream 7-dehydrocholesterol (7dC) and the downstream 5beta-ketodiol have eluded isolation (Gilbert et al. 2002). Nevertheless, the activity of several gene products including Shroud (Sro), Spo/Spok, Cyp6t3 and Cyp6u1 have been suggested to act within the Black Box based on rescue experiments that involve feeding mutant larvae intermediate compounds that are thought to be upstream and downstream of the Black Box enzymes. In these experiments, feeding the downstream intermediate 5beta-ketodiol rescues loss-of-function larvae and enables them to progress through additional molts while feeding the upstream intermediate 7dC does not (Figure 1A). (Niwa et al. 2010; Ono et al. 2006 ; Ou et al. 2011; Christesen et al. 2017).

Since all Halloween null mutants are embryonic lethal, the feeding experiments which helped position various gene products in the biosynthetic pathway sometimes involved bypassing the embryonic lethality by using RNAi to specifically knockdown the function of these genes in the prothoracic gland (PG), the source of larval ecdysone. The embryonic tissue source of E is still unknown. Strong RNAi knockdown in the PG of any of the Halloween gene products leads to first instar arrest. However, feeding the larvae various intermediates often enables them to progress into the second and third instar stage, provided the intermediate is downstream of the enzymatic block. These types of experiments suggested that the Sro and Spo/Spok enzymes are components of the Black Box.

The identification of Cyp6t3 (FBgn0033697) and Cyp6u1 (FBgn0033121) genes as encoding possible factors of the E biosynthetic pathway came about differently from classical genetic studies. Cyp6t3 was identified as a gene that was specifically up-regulated in the PG in response to loss of the transcription factor DHR4. (Ou et al. 2011). DHR4 mutants pupariate early likely because of precocious ecdysone production, and simultaneous knockdown of Cyp6t3 in a DHR4 mutant background suppressed the early pupation phenotype. In addition, RNAi knockdown of Cyp6t3 in the PGs of wild type animals resulted in slow growth and large pupae, both of which were suppressed by feeding knockdown larvae ecdysone or 5beta-ketodiol but not 7-dC suggesting that, like Spo/Spok and Sro, this enzyme works at the Black Box step. Because of the low transcript abundance of Cyp6t3 , it was also hypothesized that Cyp6t3 might be a “bottleneck” in ecdysone biosynthesis, or even rate limiting. Cyp6u1 was identified in a genome wide transcriptome analysis of the PG as a P450 that showed high level expression specifically in the PG, and RNAi knockdown in the PG led to a developmental delay/arrest phenotype which was rescued by feeding ecdysone or 5beta-ketodiol but not 7-dC, again positioning this enzyme at the Black Box step.

Since it has been reported that tissue-specific knockdown of genes can have more severe phenotypes than genetic null mutations (Gibbens et al. 2011, Ohhara et al. 2015), we made Crispr-mediated null mutations in Cyp6t3 as an additional means to assess its function during Drosophila development . Guide RNAs from both ends of the Cyp6t3 cytochrome P450 domain were simultaneously injected into a Cas9 stock (Bestgene Inc.) for the purpose of generating mutations at each individual site as well as potential deletions between the sites. After screening the progeny of the G0 flies, we obtained 3 mutations Cyp6t3[Cr1] , a 1 bp deletion at Guide 1, Cyp6t3[Cr2] , a 2 bp deletion at Guide 2, and Cyp6t3[Cr3] , a 1 kbp deletion between Guides 1 and 2 (Figure 1B). Translation of these mutated genes yields proteins with different premature stops or a large deletion within the Cytochrome P450 domain of Cyp6t3 (Figure 1C) . All three mutations are homozygous viable. For this reason alone, we deemed Cyp6t3 a non-essential gene for ecdysone biosynthesis.

Although loss of Cyp6t3 is not essential for animal survival, we reasoned that it could affect the rate of ecdysone biosynthesis. Previous work has shown that either raising or lowering the basal E biosynthetic rate leads to advanced or delayed timing of pupariation, respectively (Ou et al. 2016, Koyama et al. 2014, and Shimell et al. 2018). To address this issue, we assessed the time to pupariation of various Cyp6t3 mutant combinations (Figure 1D). In sum, we find no delay in developmental timing compared to the wild type control suggesting that the basal rate of ecdysone synthesis in the Cyp6t3 mutants is on par with wild type.

The fact that Cyp6t3 is not an essential component of the E biosynthetic pathway under lab growth conditions should not be surprising since, unlike the other Halloween genes that are highly conserved across all arthropods examined to date, Cyp6t3 has been lost in some clades (Ou et al. 2011 and Good RT et al. 2014). In addition, as previously noted, the transcript abundance of Cypt6t3 is very low by qRT-PCR (Ou et al. 2011) and did not make the cutoff in 2 different ring gland-specific transcriptional profiles, the only putative E biosynthetic P450 gene failing to do so (Ou et al. 2016 and Christesen et al. 2017). Despite these observations, the regulation of Cyp6t3 transcription by DHR4 and its ability to suppress the precocious DHR4 mutant phenotype when knocked down in the PG remains intriguing. It would be interesting to determine whether a Cyp6t3, DHR[1] double mutant shows the same suppression of precocious pupariation as found when RNAi is used to knockdown Cyp6t3 in the PG of a DHR[1] mutant. It is also possible that Cyp6t3 is important under specialized growth conditions that modulate DHR4 activity. Another possibility is that Cyp6t3 function is redundant in D. melanogaster , and perhaps other species, with another P450 enzyme. Precedence for the existence of partial redundance in E biosynthetic enzymes in insects is exemplified by the Spo/Spok relationship (Ono et al. 2006). Both are thought to function within the Black Box but at different developmental stages, Spo acting embryonically and in female follicle cells while Spok acts in the PG during larval stages. One intriguing candidate for a redundant activity could be Cyp6u1 . This gene product is also a member of the P450 clade 6 subgroup but is highly conserved across Arthropoda (Good RT et al. 2014). Examining the phenotype of Cyp6u1 null mutants and Cyp6t3;Cyp6u1 double mutants might provide insight on the redundancy issue.

In summary, our studies do not resolve the enigma of the Black Box reaction mechanism in the E biosynthetic pathway. However, our elimination of Cyp6t3 involvement under standard growth conditions helps reduce the complexity of the necessary components. This should aid in simplifying the design of future studies aimed at understanding the details of the biochemical reactions catalyzed by the Blackbox enzymes and whether their regulation serves as a rate limiting step in E biosynthesis as previously suggested (Gilbert et al. 2002).

Crispr mutant generation

The CRISPR Optimal Target Finder (Gratz et al. 2014) was employed to identify two target sites within Cyp6t3 with no off-target sites. The targets with the PAM are: 5’ GGATCCCG¯AACCATAAGCGC CGG 3’ and 5’ GGATCAAA ¯ GACGCAGGGCTC CGG 3’. Sense and antisense oligonucleotides (Oligo Table) for the two guides RNAs were cloned into pU6-BbsI-chiRNA (Addgene #45946; Gratz et al. 2013). Both guide RNAs were injected into vas-Cas9 embryos (BDSC51323, BestGene, Chino Hills CA). Single G0 flies were screened by PCR using primers cyp6t3 01F and cyp6t3 02R (Oligo Table) for the presence of deletions, making them candidates for further analysis. Ten to fifteen individual F1 progeny of G0 flies that displayed multiple deletion bands by PCR (possible Cr mutant/+) were then crossed to a CyO* balancer (BDSC 3198), homozygosed, the DNA region amplified by PCR and then sequenced for the presence of a mutation and its precise location. Both Cyp6t3[Cr1] and Cyp6t3[Cr2] mutations were obtained from a single G0 fly.

Clustal omega analysis used SnapGene Version 6.0.7 software.

Developmental timing

Crosses were set up in cages. Four-hour collections of embryos on yeast-apple juice plates (yeast paste spread on 25% apple juice, 25 mg/mL dextrose, 1.5 mg/mL tegosept, 2.2% agar) were aged about 16 hours, until 1 ^st instar larvae started to emerge. 30-35 larvae were transferred to vial food (Bloomington standard cornmeal food) sprinkled with dry yeast and allowed to grow at 25° with a 12-hour light/dark cycle. Pupariation was scored every 2 hours with the data from 4 vials combined due to the speed of pupariation and lack of data points necessary for non-linear regression analysis in Prism (asymmetric Sigmoidal, 5PL).

Oligo Table

Fly lines