Limonene and a-terpinolene were the highest produced monoterpenes, which exhibited the strongest patterns consistently . To investigate the events leading to this accumulation of monoterpenes, we performed a Fisher’s exact test to identify enriched KEGG pathways in the modules. Terpenoid backbone biosynthesis was significantly enriched in the H-II-1 module . Figure 3.4 depicts the MEP terpene backbone biosynthesis producing GPP that leads to monoterpene biosynthesis. A flux of terpene biosynthesis occurred at the end of Stage II and the beginning of Stage III, which indicates precursors for monoterpene metabolism were being synthesized . From the WGCNA, we also identified the top 5% highest connected genes within the module network. In the H-II-1 module, the top highly connected genes included the gene encoding HDR and a limonene synthase in the terpene biosynthesis pathway.Color changes are a characteristic of fruit ripening. To further define ripening in the pistachio hull, we investigated the underlying biological cause of the change in fruit coloration from green-yellow to hues of red-pink observed in the hull during Stage IV . We found a significant correlation between red coloration increase in the hull with the H-IV-1 and H-IV-2 =0.73 modules during ripening. This was further supported by a Fisher’s exact test for enrichments of KEGG pathways in each module. The H-IV-2 module was significantly enriched for the carotenoid biosynthesis pathway. The B-carotene hydrolase was the highest expressed carotenoid gene in this module and is annotated to be involved in the production of lutelin and zeaxanthin. We examined the highest connectivity genes in this module and among them was a phytoene synthase gene with 887 connections, the rate-limiting step in the carotenoid pathway. Because pink coloration often comes from anthocyanins we also looked at anthocyanin biosynthesis in the hull. While expression was present in the phenylpropanoid and flavonoid pathways, grow bag expression was low in the steps exclusive toanthocyanin biosynthesis. We also found a high negative correlation between the hull redness and the H-III-1 module corresponding to a loss of green coloration.

A significant enrichment of photosynthesis genes in the same H-III-1 module , meaning gene expression of photosynthesis genes decreased after Stage III when fruit became less green.Pistachio kernels contain a high proportion of fatty acids and reach their maximum fat content as the kernel matures during ripening . To further understand the composition of the fat content, we measured unsaturated and saturated fatty acids across six time points during Stage III and IV of kernel development . Unsaturated fatty acids made up 87% of the total fatty acids present when the fruits were ready to be consumed . We confirmed that the unsaturated fatty acids were composed of a higher ratio of mono-unsaturated to poly-unsaturated . This ratio changed through time, such that by ripening MUFA were the predominant class of fatty acids present in the fruit. We determined alterations of metabolites within each class of fatty acid contributed to the changes in MUFA and PUFA ratios during maturation . To further understand what causes these alterations, we examined gene expression of kernel in gene modules associated with the increase in fat content. The module-trait relationships indicated that the increase in fat content was highly and significantly correlated with the K-III-1 module, along with K-IV-1, K-IV-2, and K-IV-3 . This same relationship was also evident for these same modules and the proportion of unsaturated fatty acids through time . We performed an enrichment of KEGG pathway annotations in kernel modules and found that fatty acid biosynthesis was significantly enriched in the K-III-1 module . The high expression of biosynthesis genes during Stage III indicates that fatty acids are produced early on at the start of kernel development, and taper off at the beginning of Stage IV . Within this module, 19 genes encoding fatty acid biosynthesis were found including key genes FAB2 and FAD2 which desaturates steric acid into oleic acid and oleic acid into linoleic acid, respectively . The FAB2 and FAD2 genes were the highest expressed genes in the pathway, and were among the top 5% of genes in the module. FAB2 peaked in expression with 12,500 normalized reads at 1508 GDD while FAD2 peaked with 14,700 normalized reads at 1749 GDD.

Consistent with the expression data, the metabolite data also showed that oleic and linoleic acid were the top two produced fatty acids, throughout development. Interestingly, the concentration of linoleic acid decreased over time while oleic acid increased, which was not evident in the expression data.Defining the biological events occurring during pistachio fruit development that lead to traits of interest can allow for breeding and management strategies to improve fruit quality. Further, a high-quality reference genome has been lacking, as previous genomes are incomplete and fragmented. Therefore, in order to facilitate molecular breeding and broaden the understanding of nut tree crop fruit developmental processes, we present for the first time an assembled 561 Mb reference-quality chromosome-scale genome of P. vera cv. Kerman. Based on k-mer distribution analysis with PacBio HiFi reads, the Kerman genome showed a moderate heterozygosity estimate in comparison with other outcrossing highly heterozygous crops, such as pear 1.6% and grape 1.6-1.7% . This is unexpected because the previously-reported heterozygosity levels of pistachio genome were higher, 1% and 1.72% , which was attributed to the nature of outcrossing by wind pollination and dioecy of pistachio trees . In addition, the genome size estimate of 521 Mb in our study was smaller than the first attempt for pistachio genome size estimate with 26.77 Gb whole genome sequencing data using 17-mers . However, later, the genome size estimated with the larger amount of data using 21-mers was rather similar in size to our assessment . In the final genome assembly, the Kerman genome size was larger than the estimated size but smaller than previously published genome assemblies of different pistachio cultivars, Batoury and Siirt and Bagyolu . Although the size variation of the estimated and assem-bled pistachio genome assemblies could be explained by possible genome size variation across different cultivars as documented in other plant genomes, it is likely that pseudoduplication in the assemblies, especially from the highly repetitive regions in chromosome arms in the case of pistachio, is the primary cause of assembly size variation . In 2015, Sola-Campoy and colleagues characterized massive enrichment of 180 bp repeat on one arm of 11 chromosomes in pistachio, which was also observed in the Kerman genome .

The largest region with dense distribution of 180 bp repeats reached about 9 Mb in chromosome 7, where no protein-coding gene was annotated . These extremely repetitive regions could have been the major issues of accurate pistachio genome assembly and chromosome construction. Although Omni-C reads are known to offer uni-form coverage across the genome without RE sites over represented, it was observed that the overall coverage of Omni-C reads was significantly lower in those regions in Omni-C analysis, likely due to the limitation of mapping capability . Therefore, more careful validation on those regions is needed to improve the pistachio genome. The annotation of intact and fragmented transposable elements in the Kerman assembly resulted in about 11% higher genome coverage than the estimated repetitiveness . As discussed in genome size and estimate, this is likely caused by the pseudo-duplication in the assembly or misestimation due to exceptionally repeat-dense regions in chromosome arms. Among 65% of repetitive regions, nearly 49% of the genome was composed of LTRs, which have been widely known as the dominant TE groups in plants and can play a major role in adaptation and evolution by introducing novel genetic material. The protein-coding gene annotation shows high completeness based on BUSCO assessment with almost 99% . However, minor improvements can still be made by filtering out false-positive gene models and recovering missing BUSCO genes. Macrosynteny patterns between Pistacia vera cv. Kerman, Mangifera indica , and Citrus sinensis provided evidence that P. vera has not experienced a lineage-specific whole genome duplication event , unlike the recent WGD which occurred in the mango genome as described in . The synteny between mango and pistachio genomes and the similarity between their fruit morphology and growth patterns provides an interesting evolutionary comparison within the Anacardiaceae family.During the growing season, pistachios undergo a unique asynchronous development of the kernel and maternal tissues. The hull and shell develop together in the first months, marking Stages I and II , grow bag gardening while embryo development takes place during Stages III and IV. In contrast to previous reports, we found that shell hardening continues to take place with kernel growth starting in late June at approximately 1000 GDD through late August at approximately 2000 GDD . The asynchronous developmental pattern between the fruit and embryo has not been well described in the literature for other tree crops. While peaches appear to exhibit a similar pattern in seed development, this trait does not seem to have been studied in a crop whose seed is consumed . Carbohydrate dynamics in the tree may offer some explanation of the asynchrony. Carbohydrates reserved from the prior year are utilized by the tree to produce buds and develop fruit in early spring, through Stage I . The lull in fruit growth identified as Stage II may serve as a transition between a net carbon loss and a net carbon gain in photosynthesis leading to the growth of the kernel. The RNAseq experiment assessed genetic changes through time and tissue type during fruit development. The shell and hull have the most similar gene expression patterns . This was obvious in the expression of hormone-related gene expression. The shell and hull tissues exhibited very similar expression patterns for each hormone biosynthesis pathway, while the kernel expression patterns were distinct . Interestingly, the hull and shell total gene expression became more similar over time . This contrasts with the morphology of the tissues, which early on in development are physically fused together and appear to become increasingly different through time as shells become woody and split, and the hull and shell tissues separate during ripening . The similarity in gene expression may be due to both tissues undergoing terminal developmental programs. This occurs earlier in the shell when the tissue reaches its peak firmness at the beginning of Stage IV, while in the hull this occurs at the end stages of Stage IV as ripening finishes. Shell lignification was previously reported to start as early as May-June, falling in Stage I-II . While the secondary cell walls become lignified, the shells are green and flexible at this point. However, as described above, the texture of the shell continues to change through Stage III leading to a woody tissue that then splits . The shell tissues appear to senesce and be fully lignified at around 2100 GDD, as RNA content became very low in shell tissues after this point. Our gene expres-sion analysis found a proportion of the genes involved in the phenylpropanoid pathway leading to monolignols to be expressed highest at Stage II followed by a sharp decline, marking the initial lignification . The genes exhibiting this pattern were among the highest expressed homologs; however, other copies of the genes displayed patterns with peak expression later on during Stage III or IV indicating lignin was still being produced, contributing to the increased firmness of the shell. This suggests that the lignification process does not complete until the shell reaches peak firmness, as has been described in walnuts . While continued lignification may be a factor leading to shell firmness changes, other factors such as cell wall modifications likely also contribute, but require further investigation. Overall, understanding the composition and alterations in the shell tissue will be important to ascertaining the underlying mechanisms leading to shell split for a higher quality nut.Although ripening has not previously been well explored in fruit tree crops, early reports suggest that pistachios are non-climacteric fruit . We confirmed ethylene is not produced in a climacteric pattern during ripening and remains at constant low levels, as shown through biosynthesis gene expression . In conjunction with this we found evidence that abscisic acid may be involved in regulating ripening in pistachio. NCED is the rate limiting enzyme in ABA biosynthesis . We found that a primary copy was expressed in the shell and hull tissues right before ripening changes began to occur, i.e., the transition between Stage III and Stage IV. This corresponded to an increase in ABA signaling genes such as, PYLs, PP2C, SNRK2, and ABFs, suggesting ABA is active at the onset of ripening .