Pijnackeria, Scali, 2009

Milani, Liliana, Scali, Valerio, Punzi, Elisabetta, Luchetti, Andrea & Ghiselli, Fabrizio, 2020, The puzzling taxonomic rank of Pijnackeria hispanica, a chimerical hybrid androgen (Insecta, Phasmida), Organisms Diversity & Evolution (New York, N. Y.) 20 (2), pp. 285-297 : 292-295

The Pijnackeria karyotype and the evolution of egg maturation modes

A great stability for chromosome number and structure within a cluster of related species is quite a rare finding among Phasmida : all species-rich genera investigated so far actually showed a marked karyotype plasticity for both number and structure of chromosomes, particularly when parthenogens are involved. Striking instances of this feature are the Australian Didymuria ( Craddock, 1972, 1975) and Sipyloidea ( John et al. 1987) , as well as the circum-Mediterranean Bacillus and Clonopsis ( Scali 2009b; Milani et al. 2010). Also, the Iberian genus Leptynia —defined on both morphology and genetic basis and encompassing only four species—revealed diploid chromosome sets ranging from 36 to 40 and a very significant structural variation, even entailing the sexchromosome formula with a shift from the usual XX/X0 to the rarer XX/XY one ( Scali 2009b; Scali et al. 2016). In addition, the karyotype stability of Pijnackeria species is mirrored by the number and localization of cytological NORbearing satellites ( Ghiselli et al. 2007; Scali et al. 2016): taking into account the tens of specimens analysed belonging to the seven Pijnackeria species, we were able to score 10 different satellite positions, but each species possesses a maximum of two different locations. This is a quite different scenario from that observed for NOR-bearing satellites of Bacillus , where a single species may encompass as much as 12 different locations for cytological satellites ( Manaresi et al. 1991, 1992a, b, 1993; Salvadori et al. 2018; Scali et al. 2020). Frequent changes in location and number of satellites are a common finding in stick insects, and it has been suggested that they could be an outcome of transposon activity ( Meyne et al. 1990; Zhdanova et al. 2007; Ruiz-Herrera et al. 2008; Ocalewicz 2013; Satović with dashed, grey lines; grey numbers in parentheses are the numbers of missing/ideal haplotypes along the connection et al. 2016). For example, R2 non-LTR transposons have been reported to be active in Bacillus species and even particularly prone to accumulate in parthenogenetic taxa ( Bonandin et al. 2014, 2017; Scavariello et al. 2017).

The cytological satellite features are not just a trait of inter- and intra-specific variability, since in all investigated phasmatodean species—13 all together up to now— chromosomal satellites have been always found to be sites of highly enriched and co-localized rDNA/telomeric sequences ( Scali et al. 2016; Liehr et al. 2017; Salvadori et al. 2018): therefore, this trait appears to have a biological and evolutionary bearing. Actually, it has been already possible to observe that in the di-hybrid Bacillus whitei and the threehybrid B. lynceorum , active NORs derive from all ancestors, although that of maternal B. rossius derivation appears as the most conserved ( Manaresi et al. 1991, 1992a, b, 1993). However, owing to the androgenetic structure of P. hispanica , such a feature cannot be verified.

A careful karyotype analysis of Pijnackeria polyploids gives us some indirect clues about their egg maturation mechanisms, since a direct investigation has not been possible. In stick insects, egg meiosis is blocked at pachytene during the first instar larva, to be resumed in adults at laying ( Taddei et al. 1993). Consequently, eggs can be collected at precisely scheduled times and investigated. Unfortunately, in Pijnackeria , such a direct analysis of oocyte maturation is not feasible, because the few laid eggs are firmly glued to the substratum and their chorionic capsule is too fragile to be handled and cut for fixation as it has been done in Bacillus and Clonopsis ( Marescalchi et al. 1991; Scali et al. 2010). Owing to such limitations, we can propose only conjectural hypotheses for P. hispanica , and considering its complex hybrid structure, an apomictic mechanism maintaining a steady karyotype structure appears the most likely. In this connection, we can with dashed, grey lines; grey numbers in parentheses are the number of missing/ideal haplotypes along the connection observe that P. masettii is a triploid hybrid with one chromosome set derived from P. lelongi and the other two from an unknown heterospecific paternal ancestor, as both the structure of several chromosome triplets and the cytological satellite features clearly support ( Fig. 6 , triplets 1–4, 6, 12, 17–19; see also Ghiselli et al. 2007; Scali et al. 2013). At the time of its formation, such a hybrid certainly suffered from a diffuse imbalance in pairing and segregation of the trivalents during

(million years ago) and 95% high posterior density

(HPD) calculated for main

Pijnackeria and

Leptynia cladogenetic events the first meiotic division, as actually observed in triploid specimens of Bacillus atticus ( Marescalchi and Scali 1997, 2003); therefore, a modified meiosis escaping such constraints is likely to have been evolved by selecting an apomictic egg maturation mechanism. A similar meiotic pathway could have been evolved in the tetraploid P. hispanica androgenetic parthenogen with a high (76) chromosome number.

Gathering all relevant observations on the issue, the most parsimonious series of gains/losses of whole chromosome sets leading to the extant structure of P. hispanica can be envisaged as follows: a seminal parthenogenetic P. originis / P. recondita hybrid with an apomictic reproduction owing to the marked genetic differentiation of the parental taxa was produced. Pre-mating isolating mechanisms were easily overcome, since in phasmids they are rather ineffective even between utterly differentiated species ( Scali et al. 1995). Backcrosses to P. recondita males were still possible (see Tinti and Scali 1996) and, thanks to the physiological eggpolyspermy ( Scali 1972), an all-paternal progeny was originated when syngamy with the hybrid egg nucleus failed and two spermatozoa fused to originate a 2 n androgen, which only kept the mitochondrial DNA of the mother but continued an apomictic reproduction ( Mantovani and Scali 1992). The 4 n ploidy of P. hispanica could then be reached through a twostep acquisition of additional Pijnackeria genomes by the androgen. After the original hybridization of P. recondita with P. originis leading to an early diploid androgen, an additional fathering taxon, providing the third haploset, should have been different from P. recondita and likely similar to the unknown paternal ancestor of P. masettii . The last contribution of a fourth genome could have been provided by a P. barbarae -like paternal ancestor: the heterozygous structure of several quartets of P. hispanica ( Fig. 6 ) and the high variability of its ef1-α sequences ( Fig. 9 ) are consistent with the above-outlined assumptions. These being correct, P. hispanica could be then envisaged as a polyphyletic hybrid complex endowed with a high colonizing potential. Moreover, it is tempting to speculate that such kind of multi-hybrid origin could be the reason behind the long evolutionary persistence (1.96–3.31 Myr) of this parthenogenetic taxon. The high variability produced by multiple hybridization events might have compensated for the absence of sexual recombination ( Ghiselli et al. 2007 and references therein).

On the issue, we would like to mention that androgenetic stick-insect strains of Bacillus and Clonopsis contributed to the formation of a complex network among parental and derived taxa, so that their reproductive and micro-evolutionary features were defined as “reticulate evolution” (reviewed in Scali 2009a), and, within it, androgenesis has been proposed as a short-cut pathway for speciation ( Ghiselli et al. 2007; Milani et al. 2010, 2014). To better envisage the cladogenetic potential of androgenesis, a simple model of hybrid egg maturation and genome transmission has been worked out for Clonopsis hybrids, which would even explain the ascertained diploid structure of polyploid karyotypes ( Milani et al. 2009, 2010): the Clonopsis model also accommodates quite easily the otherwise inexplicable chromosomal findings reported in the Australian Sipyloidea nelida species complex by John et al. (1987).

On the whole, the targeted cytogenetic insight and transmission analysis of genomes, although rather limited, appears to be an effective tool to reveal the exploitation of a wide array of reproductive modes and evolutionary pathways in stick insects: these insights seem to really add to the routinely accepted ideas about reproductive features, evolutionary modes, and phylogenetic relationships in animals.

Figures

Fig. 6 Karyotype of the triploid Pijnackeria masettii (on the top) and of the tetraploid Pijnackeria hispanica (on the bottom) modified from Scali et al. (2013). P. masettii triplets 1–4, 6, 12, 17– 19 clearly support a 2 + 1 structure; 1st, 2nd, 13th, 15th, and 19th quartets of P. hispanica seem to suggest either a 2 + 2 structure or, better, a 2 + 1 + 1 structure

Fig. 9 Templeton network of the ef1-α gene sequences. Circle size is proportional to haplotype frequency; black dots represent missing/ideal haplotypes. Connections obtained with relaxed parameters are indicated