Page 4 - Mantovani1998
P. 4
satDNA Structure in Unisexual Stick Insects 1291
pNot13–15, pPdd16–18, pRag19–22, pSir23–25, and B. grandii clones are the same ones observed in a Downloaded from http://mbe.oxfordjournals.org/ by guest on April 9, 2016
pVen26–28, pVil29–31, and pVit32–34 (GenBank ac- previous paper (Mantovani et al. 1997).
cession numbers AF033617–AF033619, AF034422–
AF034433, and AF035230–AF035248). The nucleotide Mean p-distance values of Bacillus taxa range from
sequences of these clones are given in figure 3. The most 0.061 in grandii grandii–like sequences of B. lynceorum
frequent sequence length is 316 bp, but length variants to 0.148 in the sample comprising all intersubspecific B.
(ranging from 312 bp in pCas2 and pCat5 to 319 bp in grandii values of nucleotide substitutions (table 1, o and
pRag20) can be observed. Stop codons in all reading f, respectively). As is evident from table 2, in which the
frames are present in the consensus of both strands. most relevant comparisons between values reported in
table 1 are given, the sequences behave quite differently.
A wide range of sequence similarity (0.84–0.98) In particular, sequence diversity within B. grandii ben-
occurs between the AϩT-rich (75% on average) Bag320 azzii is significantly lower than sequence diversity with-
clones. Out of the 332 positions in the alignment, 241 in B.grandii grandii or B.grandii maretimi (comparisons
are variable, with transversions being more frequent 1–3). Moreover, sequence diversity within races of B.
than transitions in the great majority of comparisons. grandii is significantly lower than sequence diversity be-
The sites of the six endonucleases producing the Bag320 tween races (comparison 4). On the other hand, no dif-
restriction satellite pattern on ethidium-bromide- stained ferences can be observed in sequence diversity either
gels are present in the sequences, together with a DraI within or between races of B. atticus (comparisons 5–
site. 8).
The alignment of the Bag320 sequences of B. at- As far as interspecific comparisons are concerned
ticus atticus and B. grandii grandii—the taxa co-occur- (comparisons 9–12), sequence diversity in B. grandii
ring with B. lynceorum in southeastern Sicily—shows at grandii is about the same as it is in B. atticus atticus
least 11 diagnostic positions (positions 12, 93, 115, 133, (comparison 9); the same applies when the two species
176, 227, 229, 286, 312, 319, and 322). Out of the 34 are considered as a whole (comparison 12). On the other
B. lynceorum sequences, 22 clones show either the at- hand, sequence diversity within races of B. grandii is
ticus (Cat4, Mel10, Not15, Rag21, Ven27, Ven28, Vil30, lower than that within races of B.atticus (comparison
Vit33) or the grandii structure (Cas3, Cat5, Mas6, Mas7, 10), while sequence diversity between races of B. gran-
Mas8, Rag19, Rag20, Sir24, Sir25, Ven26, Vil29, Vil31, dii is higher than that between races of B. atticus (com-
Vit32, Vit34) at all diagnostic sites. The remaining parison 11). It must also be noted that when compari-
clones do not completely match either constitution ow- sons 10 and 11 are carried out with a B. grandii sample
ing to 1–5 substitutions. It is interesting to note that the devoid of B. grandii benazzii sequences, only the latter
clones Mel12, Not14, and Rag22 have an atticus com- is still significant (not shown).
position at most diagnostic loci but show a grandii nu-
cleotide at two consecutive diagnostic positions (229– With regard to B. lynceorum sequences, it must be
286, 115–133, and 229–286, respectively). first noted that sequence diversity among grandii-like
clones is significantly lower than that among atticus-like
In the neighbor-joining dendrogram (not shown) clones (comparison 13). The former is also significantly
obtained with Kimura two-parameter distances, B. lyn- lower than sequence diversity among B. grandii grandii
ceorum Bag320 sequences fall into two clades, the first clones (comparison 14), while the latter is about the
one consisting of Cas1, Cat4, Mel10, Mel11, Mel12, same as that found among B. atticus clones (comparison
Not14, Not15, Pdd18, Rag21, Rag22, Sir23, Ven27, 15).
Ven28, Vil30, and Vit33 clones, and the second one
comprising Cas2, Cas3, Cat5, Mas6, Mas7, Mas8, Mel9, Sequence diversity among grandii-like clones is
Not13, Pdd16, Pdd17, Rag19, Rag20, Sir24, Sir25, also significantly lower than that of B. grandii benazzii
Ven26, Vil29, Vil31, Vit32, and Vit34 sequences. No (P Ͻ 0.001; not shown).
apparent geographical trend emerges, because sequences
of the same origin may fall into one or the other cluster, Discussion
the only exception being Mascalucia clones. Both clus-
ters are well defined (bootstrap value ϭ 100). The present analyses first show the existence in the
triploid parthenogen B. lynceorum of AϩT-rich, tandem-
When B. atticus and B. grandii clones are consid- ly repeated, noncodifying units with a mean length of
ered in the dendrogram, a tricotomy with good bootstrap 316 bp; they belong to the Bag320 satDNA family of
values emerges (fig. 4): the first cluster comprises all B. the genus Bacillus and share the molecular characteris-
atticus clones intermigled with the B. lynceorum se- tics of satDNA families of other invertebrates.
quences of Cas1, Cat4, Mel10, Mel11, Mel12, Not14,
Not15, Pdd18, Rag21, Rag22, Sir23, Ven27, Ven28, This paper definitively verifies that B. lynceorum
Vil30, and Vit33; the second one shows the remaining embodies one haploid set of both B. grandii grandii and
B. lynceorum clones with the B. grandii grandii se- B. atticus. Their contributions must be added to that of
quences, three of which appear in a subcluster; B. gran- B. rossius, as already indicated on karyological and al-
dii benazzii and B. grandii maretimi clones build up the lozymic grounds (Mantovani, Scali, and Tinti 1992).
third cluster. UPGMA and maximum-parsimony dendro- Three different species were therefore involved in the
grams (not shown) have basically the same topology. It allotriploid constitution of B. lynceorum.
is to be noted that the groupings found for B. atticus
The contribution of B. grandii grandii and B. at-
ticus is evident on the basis of dendrogram topology.
Random cloning led to the detection of both atticus- and
