Organelle genomes. Chloroplasts and mitochondria of Lophophora

Three locations for DNA storage

A cactus cell stores DNA in its nucleus and in two types of organelles, chloroplasts and mitochondria. Chloroplasts participate in photosynthesis, while mitochondria participate in cellular respiration and ATP production. Their small gene sets are part of a system also dependent on the nuclear genome. Comparing organelle genomes helps investigate sequence history, changes in DNA organisation and plant relationships. [Zhang et al. 2026, Zhang et al., p. 675. Liu et al., p. 2][Liu et al. 2026, Zhang et al., p. 675. Liu et al., p. 2]

In both 2026 studies, DNA was isolated from outer stem tissues of plants maintained in a Beijing laboratory. Specimens came from shipments intercepted by Chinese customs authorities. Voucher collections were designated WYY20240510-01 for L. diffusa and WYY20240510-02 for L. williamsii. These are sequences from particular specimens. The publications did not establish their natural populations of origin. [Zhang et al. 2026, Zhang et al., p. 676. Liu et al., p. 2][Liu et al. 2026, Zhang et al., p. 676. Liu et al., p. 2]

Long reads and sequence assembly

Both papers used PacBio HiFi technology. Repeated reading of the same DNA fragment increases accuracy, while fragment length helps connect regions separated by repeats. Software assembles these reads into longer continuous sequences called contigs. The result is checked through methods including mapping reads back to the assembled genome. A high number of covering reads supports sequence accuracy but does not alone establish all molecular forms present in a living cell. [Zhang et al. 2026, Zhang et al., p. 676. Liu et al., pp. 3–4, 12][Liu et al. 2026, Zhang et al., p. 676. Liu et al., pp. 3–4, 12]

A short chloroplast genome

Zhang and colleagues assembled a L. diffusa chloroplast genome of 115 689 base pairs. A base pair is a unit of length of a double-stranded DNA sequence. The large single-copy region had 79 403 base pairs, the small region 32 910, and the two inverted repeats 1688 each. The repeats contain corresponding sequences arranged in opposite directions. Their very short length distinguishes the genome described. A total of 100 unique genes were annotated, including 67 protein-coding genes, 29 tRNA genes and four rRNA genes. [Zhang et al. 2026, p. 677, Fig. 2]

The new sequence was compared with a previously deposited L. diffusa sequence. The difference in total length was 93 base pairs, with identical inverted-repeat length. The authors consider intraspecific variation and differences in technology and assembly methods. The comparison does not distinguish these causes. Shortened repeats are a feature of DNA structure, and their effects on drought tolerance or photosynthetic efficiency were not measured in this experiment. [Zhang et al. 2026, pp. 677–678]

An extensive mitochondrial genome

Liu and colleagues presented a L. williamsii mitochondrial genome with a total length of 2 422 778 base pairs and GC content of 42,84%. GC means the proportion of guanine and cytosine in the sequence. The assembly initially formed a branched graph comprising 57 nodes. For subsequent analyses, the authors recorded the sequence as five linear contigs. The longest had 2 165 823 base pairs and the others 131 437, 80 851, 31 979 and 12 688. Their sum corresponds to the reported total length. [Liu et al. 2026, p. 4, Fig. 1]

This representation organises the continuity of the DNA read. The physical presence of exactly five distinct, stable chromosomes in living mitochondria requires further verification. The authors identify alternative sequence forms and structural dynamics in vivo as a further research stage. Plant mitochondrial genomes contain repeats between which recombination can occur. A single assembly representation therefore need not encompass all molecular configurations. [Liu et al. 2026, pp. 10–12]

Repeats and traces of transfer between organelles

The mitochondrial genome contained 587 simple sequence repeats, 124 tandem repeats and 5519 pairs of dispersed repeats at least 30 base pairs long. Individual values depend on the search criteria used. Simple repeats consist of a short motif repeated consecutively. Dispersed repeats connect distant, similar sequence fragments. In the paper, their distribution describes the genome's complex architecture. [Liu et al. 2026, pp. 3, 5–6, Fig. 2]

Sequence comparison identified 18 mitochondrial DNA fragments similar to the chloroplast genome. Together, they comprised 31 614 base pairs, with matches longer than 1000 base pairs selected for comparison. The authors interpret them as traces of intracellular DNA transfer from chloroplasts. This homology documents sequence similarity and position. It does not establish the time of transfer or activity of all transferred genes in the mitochondrion. [Liu et al. 2026, pp. 3, 7–8, Fig. 4]

RNA editing as a prediction

A DNA sequence and the RNA information read from it can differ through editing. In plant mitochondria, a frequent change is replacement of cytidine with uridine in RNA. This can change a codon and an amino acid in the resulting protein. The software used by Liu and colleagues predicted 431 such sites in 32 protein-coding genes. The most were identified in nad2, and none in cox1. [Liu et al. 2026, pp. 3, 7–8]

The result came from bioinformatic prediction. Confirmation requires comparison of DNA with transcripts, the actual RNA molecules. The publication considered whether editing might help maintain mitochondrial function in a dry, hot environment, but did not measure this relationship during stress. The current result identifies sites to be tested, retaining the distinction between a predicted change, an active process and physiological function. [Liu et al. 2026, pp. 11–12]

Gene trees and the choice of plants compared

The chloroplast analysis united L. diffusa, L. fricii and available L. williamsii sequences in one group, sister to the three included Mammillaria species. The mitochondrial analysis used 30 comparable genes from 28 plant species. Among its five Cactaceae representatives, Mammillaria huitzilopochtli was closest to L. williamsii. The results establish relationships in two different sets of taxa and genomes. [Zhang et al. 2026, Zhang et al., pp. 676–678, Fig. 3. Liu et al., pp. 3, 8–9, Fig. 5][Liu et al. 2026, Zhang et al., pp. 676–678, Fig. 3. Liu et al., pp. 3, 8–9, Fig. 5]

Other Lophophora species were absent from the mitochondrial study, while L. alberto-vojtechii was not included in the chloroplast study. The trees therefore do not compare all four accepted species. Increasing sample numbers and comparing nuclear markers will allow similarity in an individual organelle genome to be distinguished from the history of population and species divergence. The new sequences provide reference material for these analyses. [Zhang et al. 2026, Zhang et al., pp. 678–679. Liu et al., p. 12][Liu et al. 2026, Zhang et al., pp. 678–679. Liu et al., p. 12][2-WWW]

DNA sequencing and further biological studies

Most of the mitochondrial genes compared showed a signal of strong constraints on sequence changes, termed purifying selection. For atp8 and matR, the authors described an elevated ratio of nonsynonymous to synonymous changes, interpreted as a possible signal of positive selection. The former change the protein sequence, while the latter retain the same amino acid. The relationship of these genes to adaptation to drought and temperature remains a hypothesis requiring functional validation. [Liu et al. 2026, pp. 8–9, 11]

Genomic data allow regions to be selected for studies of relatedness and identification and variation among individuals to be examined. Marker validation requires samples reflecting the scope of the species studied and similar taxa. For desert-cactus biology, the next link is measurement of gene expression, RNA changes and physiology under defined light, temperature and water availability. Connecting these levels reveals how the sequence read participates in plant function. [Zhang et al. 2026, Zhang et al., p. 679. Liu et al., pp. 11–12][Liu et al. 2026, Zhang et al., p. 679. Liu et al., pp. 11–12]