Computational Identification of Candidate Gene Families for Volatile Sulfur Compound Biosynthesis in Cannabis sativa Using Profile Hidden Markov Models

Abstract

Cannabis is well known for its pungent, skunk-like aroma. Recent chemical studies have identified prenylated and C6 volatile sulfur compounds as contributors to its skunky and citrus-like aromas, but the pathways that produce these compounds remain unknown. This gap limits efforts to explain variation in sulfur-aroma traits and to selectively enhance or reduce those traits. To address this gap, we used the known chemistry of sulfur-containing volatiles in Cannabis and characterized sulfur and volatile biosynthetic pathways in other plant species to select candidate enzyme groups. Because the GMO cultivar is anecdotally associated with a pronounced sulfurous aroma, reference protein sequences and profile hidden Markov models were used to search its version 1 (v1) primary high-confidence protein set of 55,790 sequences. These searches recovered 975 unique proteins. Sequence screening retained 941 candidates across 20 reporting categories; 939 contained all expected domains, while the two candidates assigned to the methionine gamma-lyase (MGL)-nearest category had no category-specific expected-domain rule. The largest reporting category comprised 359 proteins containing a cytochrome P450 domain, recovered through a search motivated by cytochrome P450 family 74 (CYP74) enzymes involved in oxylipin and plant volatile formation. Thirteen of these proteins were also recovered by at least one full-length CYP74 reference search. Other large reporting categories included 218 sugar-transferase, 83 glutathione-transferase, and 61 alcohol dehydrogenase candidates. Comparison with the Cannabis Expression Atlas linked 168 candidates to 128 annotated genes through 100%-identity amino-acid matches spanning at least 80% of each GMO v1 candidate protein. Twenty-nine genes were tissue-specific, including 13 root-specific and 6 trichome-specific genes. These results define candidates for biochemical testing and direct searches for additional enzymes acting upstream and downstream in Cannabis sulfur-volatile pathways.