1 Long telomeres are associated with clonality in wild populations

2 of the fissiparous starfish Coscinasterias tenuispina

6 Alex Garcia-Cisneros*,a, Roco Prez-Portelab, Bethanie Carney Almrothc,

7 Sofie Degermand, Creu Palacna, and Helen Nilsson Sklde

a
8 Dept. Animal Biology (Invertebrates), University of Barcelona, and Biodiversity Research Institute (IRBIO),

9 643 Diagonal Avenue, 08028 Barcelona, Spain.

b
10 Center for Advanced Studies of Blanes (CEAB-CSIC), Acesso a la Cala Sant Francesc 14, Blanes, 17300

11 Girona, Spain.

c
12 Dept. Biological and Environmental Sciences, University of Gothenburg, Sweden.

d
13 Department of Medical Biosciensces, Ume University, SE-90185 Ume, Sweden.

e
14 Sven Lovn Centre for Marine Sciences - Kristineberg, University of Gothenburg, Kristineberg 566, SE 451

15 78 Fiskebckskil, Sweden.

16 *Corresponding author: [email protected].

17

1
18 Abstract

19 Telomeres usually shorten during an organisms lifespan and have thus been used as an aging and health marker.

20 When telomeres become sufficiently short, senescence is induced. The most common method of restoring

21 telomere length is via telomerase reverse transcriptase activity, highly expressed during embryogenesis.

22 However, although asexual reproduction from adult tissues plays an important role in the life cycles of certain

23 species, its effect on the aging and fitness of wild populations, as well as its implications for the long-term

24 survival of populations with limited genetic variation, is largely unknown. Here we compare relative telomere

25 length of fifty-eight individuals from four populations of the asexually-reproducing starfish Coscinasterias

26 tenuispina. Additionally, 12 individuals were used to compare telomere lengths in regenerating and non-

27 regenerating arms, in two different tissues (tube feet and pyloric caecum). The level of clonality was assessed by

28 genotyping the populations based on twelve specific microsatellite loci and relative telomere length was

29 measured via qPCR. The results revealed significantly longer telomeres in Mediterranean populations than

30 Atlantic ones as demonstrated by the Kruskal-Wallis test(K = 24.17, significant value: p-value < 0.001), with the

31 former also characterized by higher levels of clonality derived from asexual reproduction. Telomeres were

32 furthermore significantly longer in regenerating arms than in non-regenerating arms within individuals (pyloric

33 caecum tissue: Mann-Whitney test, V = 299, p-value < 10-6; and tube feet tissue Student's-t = 2.28, p-value =

34 0.029). Our study suggests that one of the mechanisms responsible for the long-term somatic maintenance and

35 persistence of clonal populations is telomere elongation.

36

37 Keywords: Aging, Asexual Reproduction, Clonal, Microsatellites, Population Genetics, Regeneration

2
38 Introduction

39 Although aging is observed in most organisms, there is a large degree of variation in the rate at which it occurs,

40 at both the species and individual level. Telomere length has frequently been used as an aging marker because

41 telomere caps normally become shorter during an organisms lifetime, primarily during DNA replication but also

42 in association with other factors such as stress (von Zglinicki, 2002; Epel et al., 2004; Kotrschal et al., 2007).

43 Critically short telomeres trigger a signal prompting the cell to permanently stop dividing, which leads to the

44 induction of cellular senescence (Herbig et al., 2006). Furthermore, long telomeres have been found to correlate

45 with good health and higher life expectancies in several species, thereby also serving as an indicator of somatic

46 fitness, which represents the boundary of aging diseases (Bize et al., 2009; Horn et al., 2010; Barrett et al.,

47 2013).

48 During fission or fragmentation in asexual organisms, two or more separate individuals are formed, resulting in

49 clonal offspring with genotypes identical to the parent and to each other. In wild asexual populations and after

50 recurrent fissions, it is difficult to determine the age of an individual by morphological means, not only for the

51 potential clone itself but also the original parental half. In these cases, genetic analyses can reveal the level of

52 clonality within a population as well as the extent of a clone, thus providing information regarding its age and

53 longevity (Ally et al., 2010). Indeed, extremely large and long-lived clonally propagating populations exist, such

54 as some sea grass species, with clones that are estimated to be 1,000 yr (Reusch et al., 1999) or more (Arnaud-

55 Haond et al., 2012). In the case of a terrestrial tree, the age of some clones have been estimated as old as 10,000

56 yr (Ally et al., 2010) and in cold waters, clonal individuals of the coral Lophelia pertusa are estimated to be

57 4,500-6,000 yr (Dahl et al., 2012). These estimations obtained for several plants and animal groups may in some

58 way indicate the existence of mechanism to largely delay, or even resist, aging in particular clones although the

59 evolutionary significance of these long-term resistance has not been clarify yet.

60 In sexually-reproducing species, telomere length is restored during embryogenesis by the reverse transcriptase

61 telomerase (Schaetzlein et al., 2004). However, little is known about aging in asexual organisms that propagate

62 via fission or budding; many questions remain unanswered as to whether they are able to fully maintain and/or

63 restore their telomeres to persist over time or whether they undergo somatic aging (Skld and Obst, 2011). In a

64 study using laboratory cultures of two invertebrate species, upregulation of telomerase has been shown to at least

65 partly restore telomeres in the clones of a flatworm (Tan et al., 2012). Telomerase is also upregulated during

3
66 budding in the colonial ascidian Botryllus schlosseri (Laird and Weissman, 2004). On the other hand,

67 experiments involving another colonial ascidian, Diplosoma listerianum, found that telomerase activity declined,

68 telomeres shortened and the growth rate slowed after prolonged asexual duplication, indicative of long-term

69 senescence in the studied clones (Skld and Obst et al., 2011). There is however very limited information

70 available regarding the effect of prolonged periods of asexual duplication on telomere length and aging in wild

71 populations. Whether molecular aging occurs and how it could potentially be delayed in wild clones is still

72 unknown.

73 Most asteroids have the ability to regenerate their body parts after autotomy or injuries, and about 26 species can

74 reproduce asexually via fission (Emson and Wilkie, 1980). All four known species of the cosmopolitan genus

75 Coscinasterias, commonly found in shallow waters, can reproduce both sexually and asexually via fission (Alves

76 et al., 2001; Lawrence 2013, among other references). When individuals of Coscinasterias reproduce sexually,

77 they release planktotrophic larvae that remain in the water column for several weeks, with a high potential to

78 colonize new habitats via dispersal (Karako et al., 2002). The species Coscinasterias tenuispina (Lamark, 1816),

79 widely distributed throughout the Atlantic Ocean and Mediterranean Sea, presents in some cases populations

80 either consisting of individuals of only one gender (usually males) or with an unbalanced proportion of males

81 and females (Alves et al., 2002, authors' unpublished data). The absence of one gender in some populations of

82 this species, as well dominance of a few genotypes - according to allozyme analyses (Ventura et al., 2004) -

83 suggests that maintenance of these populations takes place solely via asexual reproduction.

84 The aim of the present study was to assess the effect of asexual reproduction on relative telomere length in wild

85 populations of the starfish C. tenuispina, and its implication for the long-term survival of populations with

86 limited genetic variation. For this purpose, telomere length and its relationship with different levels of genetic

87 diversity related to asexuality were assessed using populations from two Mediterranean and two Atlantic sites.

88 Additionally, we explored the potential existence of mechanisms for telomere length control in somatic tissues

89 by comparing regenerating and non-regenerating arms within a set of individuals.

90

91 Material and Methods

92 Sampling

4
 93 Starfish of the species C. tenuispina (Supplementary material FS. 1) were collected from four different European

94 sites (Fig. 1A), two in the Atlantic basin and two in the Mediterranean basin, with between 13 and 17 individuals

95 collected per locality. The two Mediterranean sites, Llan (Costa Brava, Northwestern Mediterranean) and

96 Taormina (east of Sicily, Central Mediterranean), hereafter referred to as LLA and TAO, respectively, were

97 sampled in autumn 2011. Both Atlantic sites, Bocacangrejo and Abades (BOCA and ABA, separated by 33 km),

98 were located near Tenerife (Canary Islands) and were sampled in the spring (June) of 2012 (Table 1). These four

99 sampling locations were selected based on the abundance of the studied species and on the varying prevalence of

100 individuals undergoing fission. The starfish were sampled at between 0 and 20 m depth by snorkeling or

101 SCUBA-diving. Immediately after their removal from the sea, the animals were photographed on a millimeter-

102 scaled table in order to determine body size. Tube feet, used for the starfish locomotion and substrate attachment,

103 from the middle part of the longest arms of each individual were also collected (Fig. 1B) and preserved in either

104 RNAlater (Invitrogen, www.invitrogen.com) for telomere length analysis or absolute ethanol for microsatellite

105 genotyping. The animals were then released back into the sea. All tissue samples were stored at -20 C once in

106 the laboratory prior to analysis. Body size was assessed as the longest diameter across the starfish using the

107 ImageJ software program.

108 Additionally, both tube feet and pyloric caeca tissue (stomach extensions for processing and storage organ) were

109 collected from the middle part of the longest (non-regenerating) and shortest (regenerating arm; when the length

110 of this arm was less than a 50 % of the longest arm) arms of 12 asymmetric specimens sampled at site LLA (Fig.

111 1B, Supplementary material FS. 1). Pyloric caeca tissue was considered in the present analysis because it is a

112 distinct tissue that has further been shown to be involved in arm regeneration (Hernroth et al., 2010). Although

113 gonads may be also considered for the telomere length analysis, they cannot be used in this study because more

114 than 50 % of the individuals of the species lack gonads, even during the reproductive season (Crozier 1920).

115

116 Genetic analyses using microsatellites

117 Twelve microsatellite markers (m.ten1, m.ten6, m.ten13, m.ten14, m.ten19, m.ten24, m.ten25, m.ten27, m.ten30,

118 m.ten31, m.ten32, m.ten40) specifically designed for C. tenuispina (Garcia-Cisneros et al., 2013) were employed

119 for population genetic analyses (Table 1). Extractions and amplifications were performed using a REDExtract-

120 N-Amp Tissue PCR Kit (Sigma-Aldrich, www. sigmaaldrich.com). For PCR reactions we added 4 l of PCR

5
121 Ready Mix, 4 pmol of each primer, between 10 ng and 50 ng of DNA, and ultrapure water for molecular use to a

122 final reaction volume of 10 l. Forward primers for each locus were labeled with a fluorescent dye as described

123 in Garcia-Cisneros et al. (2013). Amplifications were performed on an S1000TM Thermal Cycler Dual 48/48

124 Reaction Module (BioRad Laboratories, HYPERLINK "http://www.bio-rad.com/"www.bio- HYPERLINK

125 "http://www.bio-rad.com/"rad.com), with an initial step of 1 min denaturation at 96 C followed by 35 cycles of

126 96 C for 1 min, 49 C for 30 seconds and 72 C for 20 seconds, and a final step at 72 C for 3 min.

127 Amplification products were analyzed on an ABI Prism 3730xl Genetic Analyzer (Applied Biosystems/Life

128 Technologies, www.lifetechnologies.com) by the Scientific-Technical Services of the University of Barcelona.

129 Allele size was estimated relative to an internal size standard 70-400 ROX (Bioventures Inc.,

130 www.bioventures.com) using the software program Peak-Scanner-96 (Applied Biosystems). The prevalence of

131 identical genotypes, considered as potential clones, within and between populations was tested using GenoDive

132 (Meirmans and Van Tienderen, 2004) and MLGsim (Stenberg et al., 2003). Both approaches were applied to

133 count the number of individuals with identical Multi Locus Genotypes (MLGs) and to calculate the likelihood of

134 observing identical MLGs in a population due to sexual events (Psex). MLGsim calculations of the Psex, and their

135 significance values, were based on 1,000 random simulations. MLGsim set a critical value for MLGs due to

136 sexual events at a minimum of 0.040. Values obtained for MLGs of C. tenuispina were significantly lower than

137 the critical value of 0.040 (see Table 1) pointing to clonal reproduction as the origin of identical MLGs. Genetic

138 diversity was measured by calculating allelic richness, including all individuals from the populations even if they

139 were considered the same clone, genotype diversity (synonym of clonal diversity) after rarefaction for each

140 population, heterozygosity expected and observed and the inbreeding coefficient (Fis) using the Hierfstat and

141 Vegan packages (Dixon, 2003; Goudet, 2005) in the software program R v. 3.0.0.

142 In order to assess genetic differentiation between populations, we calculated values of the D estimator (Jost,

143 2008), a statistic to measure differences in genetic structure between populations based on the allele frequencies,

144 using the R package DEMEtics (Gerlach et al., 2010). The significance of these D values was evaluated by

145 performing 10,000 permutations including all the individuals from the populations (Gerlach et al., 2010).

146

147 Telomere measurements

6
148 The relative telomere lengths in either tube feet or pyloric caeca tissue were measured for each specimen via a

149 qPCR method (Farzaneh-far et al., 2008), and modified and optimized in our laboratory (Gothenburg, Sweden).

150 Telomere length assessment via qPCR has already been validated (Grabowski et al., 2005; OCallaghan et al.,

151 2008). However, although qPCR telomere measurements are normalized based on single copy genes in humans

152 and other model species, the lack of genomic data for non-model species such as C. tenuispina hinders the

153 identification of nuclear genes without paralogs for normalization. For this reason, telomeric DNA

154 measurements in the present study were performed relative to the total quantity of DNA in the samples.

155 Genomic DNA was extracted using the DNeasy Blood and Tissue kit (Qiagen, www.qiagen.com) according to

156 the manufacturer's instructions, and DNA concentrations assessed with a NanoDrop (Thermo Scientific,

157 www.thermoscientific.com) in triplicate to obtain an accurate value of DNA quantity. Absorbance ratios

158 measuring DNA purity at 260/280 averaged approximately 1.9 0.1, with the 260/230 absorbance ratio also

159 indicating acceptable sample purity (1.8 0.2). Small differences in absorbance ratios were not correlated with

160 final telomere length (260/280: R2 = 0.006, p = 0.96; 260/230: R2 = 0.001, p = 0.99).

161 The amount of DNA to be added in the qPCR reaction was estimated from a standard curve obtained via a

162 dilution series of a mixed sample DNA pool (10 0.0001 ng). The reaction efficiency calculated from the

163 standard curve was E = 104.1 %, R2 = 0.997. For the sample qPCR reaction, all DNA samples were adjusted to

164 the same concentration, and 0.5 ng of DNA was added for each PCR mix; this produced a concentration value

165 falling well within the linear range of the mixed sample standard curve.

166 Telomere analyses examining a broad range of species have indicated that the TTAGGG sequence is conserved

167 among deuterostomes (Gomes et al., 2010); the primers used for the qPCR reaction in C. tenuispina were

168 therefore those described for humans in Farzaneh-far et al. (2008):

169 (forward) 5 CGGTTTGTTTGGGTTTGGGTTTGGGTTTGGGTTTGGGTT 3, and

170 (reverse) 5 GGCTTGCCTTACCCTTACCCTTACCCTTACCCTTACCCT 3.

171 Concentrations used for the forward and reverse primers were 100 nM and 200 nM, respectively. A 10 l KAPA

172 SYBR FAST qPCR Kit (Kapa Biosystems, www.kapabiosystems.com) was employed as a Mastermix, with 3 l

173 of water added for a final PCR volume of 20 l following the protocol described by Carney Almroth et al.

174 (2012). Telomeres were amplified in triplicate for each sample to ensure accurate measurements using the

175 following qPCR protocol cycle: 3 min at 95 C followed by 25 cycles of 15 sec at 95 C and 1 min at 56 C. The

176 final step comprised 81 cycles of temperature increase from 55 C to 95 C in order to generate a melt curve,

7
177 indicating the presence of a single product. Relative telomere length is represented as a Cycle Threshold Value

178 (Ct Value), which is inversely proportional to telomere amount; longer telomeres thus produce an earlier

179 detectable signal than shorter telomeres. All measurements were analyzed in seven PCR plates, with the mixed

180 sample DNA dilution series included on each plate as an internal standard; Ct values for these differed by less

181 than 1.5 % between plates and were adjusted for.

182 Telomere length verification was performed in an independent laboratory (Ume, Sweden) using a similar

183 protocol for telomere length qPCR (Cawthon 2002). DNA from 11 samples were send to Ume and DNA

184 concentration were re-measured by the Nanodrop instrument (Thermo scientific). Each qPCR reaction contained:

185 17.5 ng DNA (diluted in TE/E.coli buffer), 0.1 M forward primer, 0.9 M reverse primer, 1X PCR Buffer 2,

186 1.7 mM MgCl2, 2.5 mM DTT, 0.2 mM dNTP, 150 nM ROX, 0.2X SYBR and 0.625 U AmpliTaq Gold (Applied

187 Biosystems).

188 Telomere primer sequence (53) Forward: CGGTTTGTTTGGGTTTGGGTTTGGGTTTGGGTTTGGGTT,

189 Reverse: GGCTTGCCTTACCCTTACCCTTACCCTTACCCTTACCCT. Each sample was analyzed two

190 separate times in triplicates on the ABI7900HT instrument (Applied Biosystems).

191

192 Statistical analyses

193 To compare body size between specimens from different populations, a non-parametric Mann-Whitney test was

194 performed since the data did not match our prior expectation of homoscedascity (Bartlett test = 35.97, and

195 signification value, p-value < 10-7). As body size data also did not match normality (Shapiro-Wilk test; W =

196 0.93, p-value = 0.0048), a Spearman correlation was used to test if telomere lengths were correlated to body size.

197 A mixed model analysis of variance, with telomere length (expressed as the Ct value) as a dependent variable,

198 was performed using basin (Atlantic and Mediterranean basins) as a fixed factor, and population and

199 MLGs as random factors. A non-significant effect was found for the MLGs (p-value > 0.98) and was therefore

200 not considered for further analyses (see Supplementary material FS. 3). Since our data did not match normality

201 (Shapiro-Wilk test; W = 0.96, p-value < 0.01), a Kruskal-Wallis test of the telomere length was performed

202 between the different sites, and a Mann-Whitney test between basins. Spearman correlation was used to test the

203 relationship between genetic diversity and mean telomere length.

8
204 Potential differences in telomere length between regenerating and non-regenerating arms within individuals were

205 tested using a mixed model involving the logarithmic transformation of Ct values, tissue type as a factor, and

206 interaction with regeneration. Moreover, we separately tested the effects of regeneration for both tissue types

207 via a paired-data students t-test for tube feet, and a paired Mann-Whitney test for pyloric caeca telomere length

208 since the data did not match normality (non-normal distribution; Shapiro-Wilk test; W = 0.8482, p-value =

209 0.002).

210 All statistical analyses of telomere length and box plots were performed in R v. 3.0.0.

211

212 Results

213 Genetic diversity and genetic distance between localities

214 The low and significant values (p-values) of the Psex (Table 2) indicated that the identical genotypes observed

215 among individuals are a consequence of clonal propagation in this species. The studied populations of C.

216 tenuispina presented different levels of allelic richness and clonality as defined by identical genotypes (Table 1).

217 Populations from the Atlantic Ocean were genetically more diverse than those from the Mediterranean Sea,

218 presenting twelve out of the thirteen single multilocus genotypes (MLGs) found in the whole study area.

219 Furthermore, excess of heterozygotes were found in all populations except Taormina. The D estimator, used to

220 assess differences in genetic structure between populations, revealed significant genetic differences among all

221 four populations here analyzed (Table 3). Large and significant differences were recorded between the two

222 Mediterranean populations, with the latter characterized by a higher prevalence of genetic clones compared to

223 the Atlantic populations. Indeed, only one multilocus genotype was detected at Llan (Table 1).

224

225 Differences in telomere length and body size between and within populations

226 Telomeres were significantly longer in the Mediterranean than in the Atlantic starfish populations as shown by

227 the Kruskal-Wallis test (K = 24.17, significant value: p-value < 0.001) (Figs. 2A, 2B). When the four

228 populations were analyzed separately, significantly longer telomeres were again observed in the two

9
229 Mediterranean populations (Kruskal-Wallis test; K = 37.03, p-value < 0.001), with individuals from Llan

230 presenting the longest telomeres (Fig. 2B). Telomere length measurements were double-checked and verified by

231 an independent laboratory (Ume, Sweden), and the result showed a strong correlation between the

232 measurements between both independent sets of analyses (value of the correlation for the regression: R2=0.88),

233 (Supplementary material FS 2).

234 Genotype diversity (also expressed as clonal diversity) in this species depends on the relative ratio of fission

235 (asexual reproduction) versus sexual reproduction. Here, populations with lower genotype diversity, and

236 therefore higher fission rates had longer telomeres at the population level, as demonstrated by a significant

237 correlation (Pearson correlation: R = 0.99, p-value < 0.007). In Figure 3, it is presented the high correlation

238 between mean of genotype diversity per population and the Ct vale (which is inversely proportional to telomere

239 length).

240 Populations analyzed showed differences in the mean body size of the starfish, demonstrated by a significant

241 value of the Mann-Whitney test (W = 740, p-value < 10-7), with considerably larger specimens observed in the

242 Atlantic populations (mean body size = 6.36 SD 3.8 cm) than in those from the Mediterranean Sea (mean body

243 size = 2.81 SD 1.1 cm). However, telomere length did not depend on the starfish body size, as demonstrated by

244 the absence of correlation between these two variables (Correlation value: rho = 0.065, p-value = 0.63, non-

245 significant) (Supplementary material FS. 3).

246

247 Telomere length in relation to arm regeneration

248 Our results showed that telomere length was significantly longer in regenerating (short) arms than in non-

249 regenerating (long) arms (Fig.1B, SF. 1), as demonstrate by the significance of the different tests applied (F =

250 52.26, p-value < 0.001), for both tissue types analyzed, tube feet (Students-t = 2.28, p-value = 0.029) and

251 pyloric caecum tissue (Mann-Whitney, V = 299, p-value < 10-6) (Fig. 4). Pyloric caecum telomeres were always

252 longer in regenerating arms in all individuals, while eight out of twelve specimens displayed longer telomere

253 lengths in regenerating arm tube feet. .

254

10
255 Discussion

256 Life expectancy in clonal lineages remains unclear due the lack of understanding of different phenomena that

257 would influence on its time survival. Firstly, we do not know the real consequences of the accumulation of

258 somatic mutations and their deleterious effects on the individual and their clonal offspring. Secondly, we ignore

259 the real effect in wild populations of the lack of genetic diversity for adaptive potential, and finally the

260 mechanisms to avoid senescence. The first two difficulties are usually overcome when the species are able to

261 maintain sex, even at low rates or sporadic events, by combining genomes and eliminating phenotypic

262 expression of deleterious mutations when recessive. Here, in this study, we shed some light on the third problem,

263 evidencing telomere elongation during asexual reproduction of the starfish Coscinasterias tenuispina.

264 Our work represents the first study to explore the potential implications of asexual reproduction on relative

265 telomere length in wild populations of a clonal starfish. To date, most ecological studies examining telomere

266 length and/or telomerase activity have focused on obligate sexually-reproducing species or clonal organisms

267 maintained in laboratory cultures (see examples in Klapper et al., 1998; Ojimi and Hidaka, 2010; Horn et al.,

268 2010; Skld and Obst, 2011; Carney Almroth et al., 2012; Tan et al., 2012), and no previous study based on

269 telomeres and aging has been conducted on wild clonal populations of any species.

270 Longer telomeres were recorded in the Mediterranean populations of the starfish C. tenuispina, while these

271 specimens also significantly smaller in size. Although shorter telomeres were observed in the Atlantic

272 populations characterized by a larger mean body size, our results did not detect any correlation between the two

273 variables. This finding is consistent with the lack of age-related telomere shortening demonstrated for other

274 marine species, including sea urchins and lobsters, and may be attributed to high phenotypic plasticity in body

275 size and/or to continuous telomerase activity throughout their life-span (Klapper et al., 1998; Ebert et al., 2008).

276 Nevertheless, telomere length is regarded as an indicator of health and somatic fitness, and its variation observed

277 in our populations may be influenced by both inherited and/or environmental components (Epel et al., 2004).

278 Thus, populations of C. tenuispina comprising only one clone may be healthily maintained, including those in

279 Llan, which were also characterized by the longest telomeres.

280 The longer telomeres found in regenerating compared to longer non-regenerating arms are indicative of telomere

281 elongation and the preservation of chromosome ends in somatic tissue over the asexual cycle. These results may

11
282 explain the positive correlation between telomere length and level of clonality. Asexual reproduction via fission

283 has been proposed as more prevalent in small specimens of Coscinasterias (Emson and Wilkie, 1980), which is

284 also consistent with our results and observations. It is therefore possible that the key to retain long telomeres in

285 these starfish is to frequently undergo fission. On the other hand, fission may be influenced by the environment,

286 either directly, e.g. by different temperature regimes between the Mediterranean and Atlantic basins here

287 analyzed, or indirectly, e.g. by growth limitation and therefore it could be more prevalent in such situations

288 (Haramoto et al. 2007). Therefore, in C. tenuispina, longer telomeres and greater somatic fitness may be

289 triggered by unstable environmental conditions in wild populations.

290 Elongation of telomeres in populations of C. tenuispina may be one of the mechanisms related to the absence of

291 senescence and genetic defects associated to prolonged periods of asexual propagation. Studies investigating

292 terrestrial species have demonstrated that both telomere length and telomere erosion are predictors of survival

293 and somatic fitness (Bize et al., 2009; Horn et al., 2010), with the combination of long telomeres and telomere

294 elongation in regenerating tissues potentially providing these clonal starfish a high probability of survival.

295 Although the results of a previous study examining a colonial ascidian suggest that passing through a sexual

296 reproductive phase is required in order to avoid senescence after prolonged periods of asexual budding (Skld et

297 al., 2011), it is not known how generally this finding can be applied to other species. Despite the fact that

298 asexual reproduction facilitates clonal dispersion and renewal, it depends on mitotic divisions, which may

299 increase the accumulation of somatic mutations, the Muller ratchet phenomena. Different studies with asexual

300 species already revealed accumulation of somatic mutations on non-synonymus position in clonal lineages

301 compared to their non-clonal sibling species (Paland and Lynch 2006; Barraclough et al., 2007). Unfortunately,

302 the negative effects of deleterious mutations in wild populations of clonal species have never been fully

303 investigated or proved, and these effects have been only supposed, but asexual lineages persist over short

304 evolutionary periods (Schwander and Crespi 2009). However, older clones of aspen species has been found to

305 exhibit a significant reduction in reproductive performance associated with male sexual fitness decline,

306 suggesting that at least some long-lived clonal organisms may be vulnerable to senescence over long periods of

307 time (Ally et al., 2008). A hypothesis to explain the excess of heterozygosity found in the studied populations of

308 C. tenuispina may be a positive selection of heterozygotes to keep genetic diversity, besides having greater

309 individual adaptability by high phenotype plasticity (Hrandl, 2009; Goudie et al., 2012). Nevertheless, our

310 current results cannot actually test this hypothesis, and the heterozygotes excess found in all populations may be

12
311 results from other stochastic processes that have not been controlled in this study. In other organisms as plants

312 that maintain asexual lineages has been observed that polyploidy and hybridization commonly generate

313 heterozygotes. In other species as fur seals, inbreeding is avoided by an active selection from females for non-

314 relative males or heterozygotes (Hoffman et al., 2007). However, we do not have evidences of any of these

315 processes, and positive selection of heterozygotes is the only hypothesis that can be presented here, but further

316 studies on this particular point may shed light on it.

317 Although the molecular mechanisms responsible for telomere elongation and its preservation in C. tenuispina

318 remain unknown, studies examining other asexual and sexual organisms indicate a pivotal role for telomerase in

319 the telomere length regulation of somatic cell lineages. In Asterias rubens, a sexually reproducing starfish,

320 telomerase activity is high throughout the animal irrespectively of mitotic activity and there was no difference in

321 telomerase activity nor telomere length between regenerating and non-regenerating arms (Hernroth et al., 2010),

322 but clonal starfishes may have higher telomerase activities after fission as other asexual species. In flatworms,

323 maintenance of somatic telomere length seems to be an adaptation of asexual but not sexual strains, and is based

324 on different levels of telomerase activity (Tan et al., 2012). Furthermore, in the colonial ascidian Botryllus

325 schlosseri, telomerase activity is up-regulated in early bud rudiments, and declines during zooid development

326 (Laird and Weissman, 2004). Differences in the relative abundance of somatic versus stem cells might also

327 determine variation in telomere length (Ojimi and Hidaka, 2010), but we cannot either discard recombination

328 between homologous telomeres as a means to elongate telomeres (Liu et al., 2007). Even in plants, telomere

329 restoration is only present in meristomatic and reproductive tissues, with exceptions in long-lived species

330 (Flanary and Kletetschka, 2005; Watson and Riha, 2010; among other references), while in some algae it has

331 been described telomerase activity during all life cycles, but large differences in telomerase activity have been

332 found across algae groups (Fulneckov et al., 2013; evkov et al., 2013). Further comparison of telomerase

333 and stem cells in relation to regeneration in the analyzed populations of C. tenuispina would thus be of future

334 interest.

335 The results presented here reveal the need for further research exploring the ecological and evolutionary

336 significance of asexual reproduction and telomere elongation in clonal lineages. Future empirical studies

337 measuring the success of Atlantic and Mediterranean populations should consider additional variables such as

338 sexual reproductive success, as well as an evaluation of whether genetic diversity is fundamental to the

13
339 maintenance of clonal populations, concepts that have been only theoretical explored for few authors as

340 Weissman et al. (2009) and Marriage and Orive (2012). Those theoretical models for asexual species have

341 evaluated and proposed that clones with large population sizes exhibit high successful levels (Weissman et al.,

342 2009; Marriage and Orive, 2012). Nevertheless, further analysis regarding possible somatic deleterious

343 mutations, as well as health and population size monitoring, is essential in order to understand the life

344 expectancy of clonal populations.

345

346 Acknowledgments

347 We thank Dr. Owen Wangensteen for his support during sampling and for his fruitful ideas. We are also

348 indebted to Dr. Jose Carlos Hernandez, Dr. Sabrina Clemente, and the staff of the University of La Laguna for

349 their help and logistic support during sampling in Tenerife. We thank Susann Haraldsson, Ume, for laboratory

350 assistance with qPCR. This research was financially supported by a PhD fellowship FPI-MICINN (BES-2011-

351 044154) (ACG), the European ASSEMBLY project (227799), the Swedish Royal Academy of Sciences (ACG),

352 and the Spanish Government project CTM2010-22218-C02-. The research was also supported by a Juan de la

353 Cierva contract from the Spanish Government (RPP) and by the Adlerbertska Research Foundation (HNS).

354

355 References

356 Ally D, Ritland K, Otto SP (2008). Can clone size serve as a proxy for clone age? An exploration using
357 microsatellite divergence in Populus tremuloides. Mol. Ecol. 17: 4897911.
358 Ally D, Ritland K, Otto SP (2010). Aging in a long-lived clonal tree. PLoS Biol. 8: e1000454.
359 Alves LS, Pereira A, Ventura C (2002). Sexual and asexual reproduction of Coscinasterias tenuispina
360 (Echinodermata: Asteroidea) from Rio de Janeiro, Brazil. Mar. Biol. 140: 95101.
361 Arnaud-Haond S, Duarte CM, Diaz-Almela E, Marb N, Sintes T, Serro EA (2012). Implications of extreme
362 life span in clonal organisms: millenary clones in meadows of the threatened seagrass Posidonia oceanica. PLoS
363 One 7: e30454.
364 Barraclough TG, Fontaneto D, Ricci C, Herniou EA (2007). Evidence for inefficient selection against deleterious
365 mutations in cytochrome oxidase I of asexual bdelloid rotifers. Mol. Biol. Evol. 24: 195262.
366 Barrett ELB, Burke T a, Hammers M, Komdeur J, Richardson DS (2013). Telomere length and dynamics predict
367 mortality in a wild longitudinal study. Mol. Ecol. 22: 24959.
368 Bize P, Criscuolo F, Metcalfe NB, Nasir L, Monaghan P (2009). Telomere dynamics rather than age predict life
369 expectancy in the wild. Proc. Biol. Sci. 276: 167983.
370 Carney Almroth B, Skld M, Skld HN (2012). Gender differences in health and aging of Atlantic cod subject to
371 size selective fishery. Biol. Open 1: 9228.

14
372 Cawthon RM (2002). Telomere measurement by quantitative PCR. Nucleic Acids Res. 30: e47
373 Dahl MP, Pereyra RT, Lundlv T, Andr C (2012). Fine-scale spatial genetic structure and clonal distribution of
374 the cold-water coral Lophelia pertusa. Coral Reefs 31: 11351148.
375 Dixon P (2003). VEGAN, a package of R functions for community ecology. J. Veg. Sci. 14: 927930.
376 Ebert TA, Russell MP, Gamba G, Bodnar A, Russel MP (2008). Growth, survival, and longevity estimates for
377 the rock-boring sea urchin Echinometra lucunter lucunter (Echinodermata, Echinoidea) in Bermuda. Bull. Mar.
378 Sci. 82: 381403.
379 Emson RH, Wilkie IC (1980). Fission and Autotomy in Echinoderms. (eds) Oceanogr. Mar. Bio. Ann. Rev.
380 Aberdeen University Press 18: 155:250
381 Epel ES, Blackburn EH, Lin J, Dhabhar FS, Adler NE, Morrow JD, et al. (2004). Accelerated telomere
382 shortening in response to life stress. Proc. Natl. Acad. Sci. U. S. A. 101: 173125.
383 Farzaneh-far R, Cawthon RM, Na B, Warren BS, Schiller NB, Whooley MA (2008). Prognostic Value of
384 Leukocyte Telomere Length in Patients with Stable Coronary Artery Disease: Data from the Heart and Soul
385 Study (R1). Arter. Thromb Vasc. Biol. 28: 13791384.
386 Flanary BE, Kletetschka G (2005). Analysis of telomere length and telomerase activity in tree species of various
387 life-spans, and with age in the bristlecone pine Pinus longaeva. Biogerontology 6: 10111.
388 Fulneckov J, Sevckov T, Fajkus J, Lukesov A, Lukes M, Vlcek C, et al. (2013). A broad phylogenetic
389 survey unveils the diversity and evolution of telomeres in eukaryotes. Genome Biol. Evol. 5: 46883.
390 Garcia-Cisneros A, Valero-Jimnez C, Palacn C, Prez-Portela R (2013). Characterization of thirty two
391 microsatellite loci for three Atlanto-Mediterranean echinoderm species. Conserv. Genet. Resour. 5: 749753.
392 Gerlach G, Jueterbock A, Kraemer P, Deppermann J, Harmand P (2010). Calculations of population
393 differentiation based on GST and D: forget GST but not all of statistics! Mol. Ecol. 19: 384552.
394 Gomes NM V, Shay JW, Wright WE (2010). Telomere biology in Metazoa. FEBS Lett. 584: 374151.
395 Goudet J (2005). Hierfstat, a package for r to compute and test hierarchical F-statistics. Mol. Ecol. Notes 5: 184
396 186.
397 Goudie F, Allsopp MH, Beekman M, Oxley PR, Lim J, Oldroyd BP (2012). Maintenance and loss of
398 heterozygosity in a thelytokous lineage of honey bees (Apis mellifera capensis). Evolution 66: 1897906.
399 Grabowski P, Hultdin M, Karlsson K, Tobin G, Aleskog A, Thunberg U, et al. (2005). Telomere length as a
400 prognostic parameter in chronic lymphocytic leukemia with special reference to VH gene mutation status. Blood
401 105: 480712.
402 Haramoto S, Komatsu M, Yamazaki Y (2007). Patterns of asexual reproduction in the fissiparous seastar
403 Coscinasterias acutispina (Asteroidea: Echinodermata) in Japan. Zoolog. Sci. 24: 107581.
404 Herbig U, Ferreira M, Condel L, Carey D, Sedivy JM (2006). Cellular senescence in aging primates. Science.
405 311: 1257.
406 Hernroth B, Farahani F, Brunborg G, Dupont S, Dejmek A, Skld HN (2010). Possibility of mixed progenitor
407 cells in sea star arm regeneration. J. Exp. Zool. 314: 45768.
408 Hoffman JI, Forcada J, Trathan PN, Amos W (2007). Female fur seals show active choice for males that are
409 heterozygous and unrelated. Nature 445: 9124.
410 Hrandl E (2009). A combinational theory for maintenance of sex. Heredity. 103: 44557.
411 Horn T, Robertson BC, Gemmell NJ (2010). The use of telomere length in ecology and evolutionary biology.
412 Heredity. 105: 497506.
413 Jost L (2008). G(ST) and its relatives do not measure differentiation. Mol. Ecol. 17: 40154026.
414 Karako S, Achituv Y, Perl-Treves R, Katcoff D (2002). Asterina burtoni (Asteroidea; Echinodermata) in the
415 Mediterranean and the Red Sea: does asexual reproduction enhance colonization? Mar. Ecol. Prog. Ser. 234:
416 139145.

15
417 Klapper W, Khne K, Singh KK, Heidorn K, Parwaresch R, Krupp G (1998). Longevity of lobsters is linked to
418 ubiquitous telomerase expression. FEBS Lett. 439: 1436.
419 Kotrschal A, Ilmonen P, Penn DJ (2007). Stress impacts telomere dynamics. Biol. Lett. 3: 12830.
420 Laird DJ, Weissman IL (2004). Telomerase maintained in self-renewing tissues during serial regeneration of the
421 urochordate Botryllus schlosseri. Dev. Biol. 273: 18594.
422 Lawrence J (2013). Starfish : biology and ecology of the Asteroidea. Johns Hopkins University Press: Baltimore.
423 Liu L, Bailey SM, Okuka M, Muoz P, Li C, Zhou L, et al. (2007). Telomere lengthening early in development.
424 Nat. Cell Biol. 9: 143641.
425 Marriage TN, Orive ME (2012). Mutation-selection balance and mixed mating with asexual reproduction. J.
426 Theor. Biol. 308: 2535.
427 Meirmans PG, Van Tienderen PH (2004). Genotype and Genodive: Two Programs for the Analysis of Genetic
428 Diversity of Asexual Organisms. Mol. Ecol. Notes 4: 792794.
429 OCallaghan N, Dhillon V, Thomas P, Fenech M (2008). A quantitative real-time PCR method for absolute
430 telomere length. Biotechniques 44: 8079.
431 Ojimi MC, Hidaka M (2010). Comparison of telomere length among different life cycle stages of the jellyfish
432 Cassiopea andromeda. Mar. Biol. 157: 22792287.
433 Paland S, Lynch M (2006). Transitions to asexuality result in excess amino acid substitutions. Science. 311:
434 9902.
435 Reusch TBH, Bostrm C, Stam WT, Olsen JL (1999). An ancient eelgrass clone in the Baltic. Mar. Ecol. Prog.
436 Ser. 183: 301304.
437 Schaetzlein S, Lucas-Hahn A, Lemme E, Kues WA, Dorsch M, Manns MP, et al. (2004). Telomere length is
438 reset during early mammalian embryogenesis. Proc. Natl. Acad. Sci. U. S. A. 101: 80348.
439 Schwander T, Crespi BJ (2009). Twigs on the tree of life? Neutral and selective models for integrating
440 macroevolutionary patterns with microevolutionary processes in the analysis of asexuality. Mol. Ecol. 18: 2842.
441 evkov T, Biov K, Fojtov M, Lukeov A, Hrkov K, Skorov E (2013). Completion of cell division is
442 associated with maximum telomerase activity in naturally synchronized cultures of the green alga Desmodesmus
443 quadricauda. FEBS Lett. 587: 7438.
444 Skld HN, Asplund ME, Wood C a, Bishop JDD (2011). Telomerase deficiency in a colonial ascidian after
445 prolonged asexual propagation. J. Exp. Zool. B. Mol. Dev. Evol. 316: 27683.
446 Skld HN, Obst M (2011). Potential for clonal animals in longevity and ageing studies. Biogerontology 12: 387
447 96.
448 Stenberg P, Lundmark M, Saura A (2003). mlgsim: a program for detecting clones using a simulation approach.
449 Mol. Ecol. Notes 3: 329331.
450 Tan TCJ, Rahman R, Jaber-Hijazi F, Felix DA, Chen C, Louis EJ, et al. (2012). Telomere maintenance and
451 telomerase activity are differentially regulated in asexual and sexual worms. Proc. Natl. Acad. Sci. U. S. A. 109:
452 420914.
453 Ventura CRR, Alves SLS, Maurcio CEP, Silva EP (2004). Reproduction and population genetics of
454 Coscinasterias tenuispina ( Asteroidea: Asteriidae ) on the Brazilian coast . In: Nebelsick J. H., Heinzeller T.
455 (eds) Echinoderms: Munchen Proceedings of the 11th International Echinoderm Conference, 6-10 October
456 2003, Munich, Germany, Taylor & Francis 2004: Munich, pp 7377.
457 Watson JM, Riha K (2011). Telomeres, aging, and plants: from weeds to Methuselah - a mini-review.
458 Gerontology 57: 12936.
459 Weissman DB, Desai MM, Fisher DS, Feldman MW (2009). The rate at which asexual populations cross fitness
460 valleys. Theor. Popul. Biol. 75: 286300.
461 Von Zglinicki T (2002). Oxidative stress shortens telomeres. Trends Biochem. Sci. 27: 339344.
462

16
463

17
 464 Tables

465 Table 1. Geographical coordinates and genetic information of the sampled Coscinasterias tenuispina

466 populations. Data presented include: Sample size (n), Number of different Multi Locus Genotypes (MLGs),

467 Clonal MLGs (found in more than one individual) and Single MLGs (found in only one individual), as well as

468 Allelic richness after rarefaction to 13, Clonal richness (R), Genotype diversity measured after rarefaction,

469 observed and expected heterozigosity (Ho/He) and inbreeding coefficient (Fis).

Population Code Coordinates n MLG Clonal Single Allelic R (Clonal Genotype Ho/He Fis
MLG MLG richness richness) diversity r[13]
r[13]
Llana, Costa LLA 4223 N, 17 1 1 0 1.33 0.00 1.00 (0.0) 0.33/0.17 -1
Brava 309 E
Taormina, Sicily TAO 3751 N, 15 4 3 1 1.75 0.21 3.80 (0.4) 0.24/0.27 0.11
1518 E
Bocacangrejo, BOCA 2824 N, 13 8 2 6 2.06 0.58 7.54 (0.5) 0.53/0.27 -0.29
Tenerife 1619 W
Abades, Tenerife ABA 2808 N, 13 7 1 6 2.14 0.50 6.54 (0.5) 0.43/0.25 -0.22
1626 W
TOTAL 58 20 7 13 0.33

18
470 Table 2. Different Multi Locus Genotypes (MLGs) found in more than one individual in the four localities.

471 Llan (LLA), Taormina (TAO), Boca Cangrejo (BOCA), and Abades (ABA). Data presented include the

472 number of individuals sharing the same MLG (n), the probability of obtaining the same MLG from different

473 sexual events (Psex), and the associated p-value of Psex. * Significant when p-values < 0.01.

Clonal MLG Locality n Psex p-value

-14
MLG 1 LLA 17 < 10 0.000*
MLG 2 TAO 5 0.000 0.000*
MLG 4 TAO 3 < 10 -8 0.007*
-15
MLG 8 TAO 6 < 10 0.000*
-10
MLG 9 BOCA 2 < 10 0.000*
MLG 11 BOCA 5 0.000 0.000*
-14
MLG 12 ABA 7 < 10 0.000*
474

19
475 Table 3. Values of the D genetic differentiation estimator between populations of Coscinasterias tenuispina.

476 Llan (LLA), Taormina (TAO), Boca Cangrejo (BOCA), and Abades (ABA). * Indicates significant p-values <

477 0.01.

TAO LLA BOCA

LLA 0.167*
BOCA 0.159* 0.233*
ABA 0.143* 0.148* 0.138*

20
478 Figure legends

479 Figure 1. Map of sampling locations, including those in Mediterranean Sea and Northeastern Atlantic Ocean.

480 Circles highlight the three sampling areas. Two different populations were sampled in Tenerife.

481 Figure 2. Box plots of telomere qPCR Ct values obtained from tube feet for the Coscinasterias tenuispina

482 populations from different seas and localities: A) Ct values of grouped Atlantic and Mediterranean populations,

483 and B) Ct values of each separate locality. Llan (LLA), Taormina (TAO), Abades (ABA), and Bocacangrejo

484 (BOCA). Lower Ct values indicate longer telomeres. Boxes are represented by the first and third quartile, the

485 dark line is the median, and dots are outliers.

486 Figure 3. Correlation between relative telomere length in tube feet (telomere qPCR Ct Value) and genotype

487 diversity for the four Coscinasterias tenuispina populations analyzed. Horizontal and vertical bars represent the

488 standard error for genotype diversity and relative telomere length, respectively. The relative telomere length was

489 represented as the mean value per locality.

490 Figure 4. Box plot of Coscinasterias tenuispina telomere qPCR Ct values in regenerating and non-regenerating

491 arms for two different tissues from individuals of Llana: pyloric caecum (Pyl) and tube feet (TF). Pyloric

492 caecum from non-regenerating arms (Pyl NR); Pyloric caecum from regenerating arms (Pyl R); Tube feet from

493 non-regenerating arms (TF NR); Tube feet from regenerating arms (TF R). Boxes are represented by the first and

494 third quartile, the dark line is the median, and dots are outliers.

495 FS 1. Picture of the aboral side of a Cosciansterias tenuispina individual (black and white image). The

496 individual had three longer non-regenerating arms, and five shorter regenerating arms.

497 FS 2. Telomere length PCR verification. Eleven samples were sent to an independent laboratory in Ume,

498 Sweden, and reanalyzed for telomere length. The samples telomere Ct-values were compared between

499 Gothenburg and Ume laboratories.

500 FS3. Dot plot representing the mean of Ct. value at the y-axis, and the longest arm length at the x-axis. Full

501 filled dots represent Mediterranean individuals while white dots represent individuals from the Atlantic side.

502 Vertical bars represent the standard deviation from the three replicate analyses per each individual.

21
A B
Normal (long)
Costa Brava, arms
Llan (LLA)

Sicily,
Taormina
(TAO)

Regenerating (short)
Tenerife, Bocacangrejo Tube feet arms
(BOCA)
Tenerife, Abades Pyloric caeca
(ABA)
 A) B)

14.0

14.0

Ct. Value

Ct. Value
13.0

13.0

12.0

12.0

11.0

11.0
Mediterranean Atlantic LLA TAO BOCA ABA

Localities
 R = 0.99
12.6

LLA
TAO
ABA
BOCA
12.4

Ct. Value

12.2
12.0
11.8

1 2 3 4 5 6 7 8

Genotype diverstity

18
16

Ct. Value

14

12

NR.Pyl R.Pyl NR.TF R.TF

FS1. Picture of the aboral side of a Cosciansterias tenuispina individual (black and white image). The individual

had three longer non-regenerating arms, and five shorter regenerating arms.
1 FS2. Telomere length PCR verification. Eleven samples were sent to an independent laboratory in Ume,

2 Sweden, and reanalyzed for telomere length. The samples telomere Ct-values were compared between

3 Gothenburg and Ume laboratories.

 15 A tlantic
M editerranean

14

C t. Value

13

12

11

0 5 10 15 20

A rm size (cm )
SF 3. Dot plot representing the mean Ct. value, at the y-axis, and the longest arm length. at the x-axis. Full filled dots
represent Mediterranean individuals while white dots individuals from the Atlantic. Vertical bars represent the standard
deviation from the three replicas analysis per each individual.