*3.1. Mitochondrial Processing in P. duplex Resembles That Seen in C. reinhardtii*

We mapped the 5 and 3 termini of presumably mature *P. duplex* mitochondrial mR-NAs and found the 5- UTRs to be very short or non-existent, the 3- UTRs of varying lengths, the polycytidylation of 3 termini, and circularized full-length mRNAs. In the only other chlorophyte where mitochondrial RNA processing has been documented, *C. reinhardtii*, individual mRNAs are endonucleolytically cleaved from primary transcripts directly adjacent to the start codons, leaving no 5- UTR. The remaining intergenic region becomes the 3- UTR of an adjacent gene and is of varying lengths, presumably due to exonucleolytic processing [56]. It has also been demonstrated that non-template polycytidylation occurs on the 3 termini of *C. reinhardtii*, and that this phenomenon may be limited to the Phylum Chlorophyta since it has been found in representative species from Peridinophyceae, Prasinophyceae, Trebouxiophyceae, and Chlorophyceae, but not in a red alga (*Chondrus crispus* Stackhouse), a glaucophyte (*Cyanophora paradoxa* Korshikov), or embryophytes (*Physcomitrella patens* (Hedw.) Bruch & Schimp., *Arabidopsis thaliana* (L.) Heynh., and *Solanum tuberosum* L.) [55]. *Chlamydomonas* mitochondrial mRNAs are also 3- polyuridylylated and polyadenylated [54–57], but this was not observed in *P. duplex*.

Circular mRNAs are a common phenomenon across the biological spectrum, but their purpose has been difficult to determine [78]. In algal mitochondria mRNA circularization has been demonstrated in *Chlamydomonas*, where it appears to create translatable mRNAs and was hypothesized to be linked to polycytidylation [56]. In *P. duplex*, full-length coding regions were found to be circularized for seven of the twelve transcripts that were analyzed. The remaining five yielded no data so we are uncertain if they form circular transcripts. Circularization for five of the transcripts coincided with tandemly repeated templatederived motifs, AU repeats, GAACGAA, and GCGUCUU, which was not reported in *C. reinhardtii*. RNA circularization among nuclear genes occurs by way of an intron-exon back-splicing mechanism involving protein factors and conserved cis-elements [79,80], and perhaps a similar mechanism involving repeated elements occurs in mitochondria. The co-incidence of poly-cytidylation and circularization in *C. reinhardtii* led to a hypothesis that the poly(C) acted as a cis-element for circularization. In *P. duplex*, only one transcript, *atp9*, had circular mRNAs with full-length coding regions and a poly(C) addition. The other two transcripts with poly(C) additions, *atp6* and *cob*, were fragments of the coding region, suggesting that the poly(C) addition is not universally linked to the creation of translatable circular mRNAs and may not act as a signal for circularization.

These data demonstrate that mitochondrial mRNA processing is conserved in the Chlorophyceae algal clade. The polycytidylation of mitochondrial mRNAs has already been shown to be conserved across the chlorophytic algae [55]. This study extends the similarities of mitochondrial mRNA processing to include the absence of a 5- UTR, despite *P. duplex* having larger intergenic regions than *C. reinhardtii*. Our data suggest that the cleavage of tRNAs plays a pivotal role in the maturation of two mRNAs (*cob* and *atp9*) in *P. duplex*, which is consistent with mitochondrial transcripts in all systems studied to date, from *C. reinhardtii* to humans. The difference is the size of the chondriome and the lengths of the intergenic regions. We found that the removal of tRNAs in *P. duplex* creates mature 3 ends that, at least in two cases, require no further processing other than polycytidylation. The 5 ends, however, are further processed until no 5- UTR remains (Figure 4). The lack of a 5- UTR and the accompanying translation mechanism required is not only due to the compressed chondriome of *C. reinhardtii*, but also its purposeful removal prior to translation. *Pediastrum* is in the order Sphaeropleales and is a sister clade to Volvocales, which contains *Chlamydomonas* [81], so it is possible that these processing events could be limited to those two Orders; however, we hypothesize that these processes are conserved across the Chlorophyceae.
