Microsatellite primer development for the invasive perennial herb Gypsophila paniculata (Caryophyllaceae)

Abstract

Premise of the Study

Gypsophila paniculata (baby's breath; Caryophyllaceae) is a herbaceous perennial that has invaded much of northern and western United States and Canada, outcompeting and crowding out native and endemic species. Microsatellite primers were developed to analyze the genetic structure of invasive populations.

Methods and Results

We identified 16 polymorphic nuclear microsatellite loci for G. paniculata out of 73 loci that successfully amplified from a primer library created using Illumina sequencing technology. Microsatellite primers were developed to amplify di‐, tri‐, and tetranucleotide repeats and tested in three invasive populations in Michigan.

Conclusions

These markers will be useful in characterizing the genetic structure of invasive populations throughout North America to aid targeted management efforts, and in native Eurasian populations to better understand invasion history. Five of these developed primers also amplified in G. elegans.

The herbaceous perennial forb Gypsophila paniculata L. (Caryophyllaceae) was introduced to North America in the late 1800s (Darwent and Coupland, 1966). Invasive populations have since been documented throughout the northern and western United States and Canada, specifically in agriculture fields, rangeland, roadsides, and sandy coastlines along the Great Lakes (Darwent, 1975; Emery and Doran, 2013). Despite its wide range, little information exists on how populations throughout North America are related or spreading. Due to its aggressiveness, negative impacts on native biota (Emery and Doran, 2013), and a lack of data regarding its spread, it is important to develop molecular markers that can characterize the genetic structure of invasive populations of G. paniculata. These markers were developed to investigate invasions within the Lake Michigan coastal dune system where an 1800‐acre infestation has occurred, so that land managers can use population genetic approaches to inform their management efforts. These markers and optimized protocols can also be used to characterize populations of G. paniculata throughout its invasive and native ranges to further assess its invasion history and spread.

Calistri et al. (2014) examined the genetic relationship of five Gypsophila Boiss. spp. (including G. paniculata) within their native range and 13 commercial hybrid strains using a combination of amplified fragment length polymorphisms (AFLPs), inter‐simple sequence repeats (ISSRs), target region amplification polymorphisms (TRAP), and universal chloroplast simple sequence repeats (cpSSRs). However, the majority of these markers are dominant and thus do not fully distinguish between homozygotes and heterozygotes, a characteristic that would allow for fine‐scale population genetic analyses (Freeland et al., 2011). Thus, the development of microsatellite markers for G. paniculata is necessary to adequately characterize invasive populations throughout North America.

METHODS AND RESULTS

Microsatellite library development, assembly, and identification

Adventitious buds growing from the caudex of five G. paniculata plants were collected from Sleeping Bear Dunes National Lakeshore (hereafter Sleeping Bear Dunes or SBDNL) in 2015 to develop the microsatellite library. Tissue was stored in indicator silica until DNA extraction. Genomic DNA was extracted using DNeasy Plant Mini Kits (QIAGEN, Hilden, Germany), with modifications including extra wash steps with AW2 buffer (QIAGEN). Extracted DNA was run through OneStep PCR Inhibitor Removal Columns (Zymo Research, Irvine, California, USA) twice to remove any inhibitory compounds (e.g., phenolics, humic acid) that could interfere with downstream processes, and was checked using a NanoDrop 2000 (Thermo Fisher Scientific, Waltham, Massachusetts, USA).

For microsatellite library development, each sample was diluted to 50 ng/μL and submitted to the Cornell Life Science Core Laboratory Center Sequencing and Genotyping Facility where microsatellite libraries were created and sequenced using a 2 × 250 paired‐end format on an Illumina MiSeq (Illumina, San Diego, California, USA; Appendix S1). Raw sequence files for the microsatellite library have been deposited to NCBI's Sequence Read Archive (BioProject no. PRJNA431197). A total of 58,907 contigs containing microsatellite loci were obtained (Appendix S2), and details on filtering, assembly, and microsatellite primer identification are expanded upon in the Supporting Information (Appendix S1). MSATCOMMANDER (version 1.0.3; Faircloth, 2008) identified 3892 potentially unique primers that yielded products of 150–450 bp, had a GC content between 30–70%, and had a melting temperature (T _m) between 58–62°C, with an optimum of 60°C (Appendix S3).

Primer optimization

Prior to PCR optimization, contigs containing potential primers were aligned using Clustal Omega (Sievers et al., 2011) to ensure they were targeting unique microsatellite regions. We tested 107 primer pairs that consisted of either tetrameric, trimeric, or dimeric motifs, and yielded products between 150–300 bp. Of these, 73 successfully amplified (Appendix S4), and 16 were polymorphic and easily scorable (Appendix S5). DNA from leaf tissue collected in 2016 from three populations (Zetterberg Preserve, SBDNL, Petoskey State Park) along eastern Lake Michigan was used for primer optimization (Fig. 1). We collected tissue samples from a minimum of 30 individual plants from each population, and herbarium voucher specimens were collected for each location (Appendix 1). Tissue storage and DNA extraction methods are the same as previously stated.

Figure 1.

Map of Gypsophila paniculata sampling locations in the northwest corner of Michigan's Lower Peninsula.

PCR reactions consisted of 1× Taq + KCl buffer (Thermo Fisher Scientific), 2.0–2.5 mM MgCl₂ (Table 1) (Thermo Fisher Scientific), 300 μM dNTP (New England BioLabs, Ipswich, Massachusetts, USA), 0.08 mg/mL bovine serum albumin (BSA; Thermo Fisher Scientific), 0.4 μM forward primer fluorescently labeled with either FAM, VIC, NED, or PET (Applied Biosystems, Foster City, California, USA), 0.4 μM reverse primer (Integrated DNA Technologies, Coralville, Iowa, USA), 0.25 units of recombinant Taq DNA polymerase 5 U/μL (Thermo Fisher Scientific), and a minimum of 50 ng of DNA template, all in a 10.0‐μL reaction volume. The thermal cycle profile consisted of 94°C for 5 min; 35 cycles of 94°C for 1 min, 62°C annealing temperature for 1 min, 72°C for 1 min; and a final elongation step of 72°C for 10 min. Successful amplification was determined by visualizing the amplicons on a 2% agarose gel stained with ethidium bromide. Fragment analysis of the amplicons was performed on an ABI 3130xl Genetic Analyzer (Applied Biosystems).

Table 1.

Characteristics of 16 polymorphic nuclear microsatellite loci developed for Gypsophila paniculata

Locusa	Primer sequences (5′–3′)	Repeat motif	Allele size range (bp)	mM MgCl2	Fluorescent label	GenBank accession no.
BB_3335	F: TCCACCAAACTCTTAAACTGCC	(AGG)5	215–244	2.5	NED	MH704701
	R: CACAGACACAAAGGATCCAACC
BB_3913	F: GGCTGTCGGGTAATAAACACAG	(ACAG)5	159–171	2.0	PET	MH704702
	R: TCCCAACTCAAGTCATAGCCTAG
BB_5567	F: GGCTAGGGAAAGTAGGAAGACC	(ATT)5	198–222	2.0	VIC	MH704703
	R: CGTGTCCTGTTTCTCCATGATC
BB_4443	F: TAGGGTGGGTGCTTGTACTAAC	(AAG)16	171–211	2.0	NED	MH704704
	R: AAAGTGGTGCTGCAGAAGAATC
BB_21680	F: ACTACACACAGACTCGATCCTC	(AAAG)5	199–218	2.0	PET	MH704705
	R: CTTTGATTGTTTGGTGTAAGTTGC
BB_3968	F: CATGGAGGACAATGAGAAGACG	(AGG)6	207–219	2.0	FAM	MH704706
	R: ACGGTGGTAATGAAGTTTGGTG
BB_1355	F: GCTGATCTTTGTCGTCAGGAAG	(AAAC)5	220–224	2.0	NED	MH704711
	R: ACTCTAGGTGTTAGGAAGGCAC
BB_5151	F: TCCACCTTATAACTCACCACCC	(ACC)5	205–210	2.0	PET	MH704712
	R: TGAGGAAGGATAACAGCTCTCG
BB_14751	F: CCTCAAACCCTAACAATGCTCC	(AAG)12	195–248	2.5	FAM	MH704713
	R: TCAGCCGATCCTCTAACACG
BB_4258	F: TCACAAGAGGCCCAATTTCTTC	(AAT)5	178–195	2.5	VIC	MH704714
	R: ACTTGAACCCGAACCTATACCC
BB_6627	F: CAAACTCAACCAACCAGACACC	(AAAC)5	151–155	2.0	FAM	MH704715
	R: CACCTCAGCAACAACAGAGTG
BB_5021	F: ATTGTCGGTGGTCATTGGTTTC	(AC)8	162–207	2.0	VIC	MH704707
	R: CTTAGTCCGCAGTGTAAACAAAG
BB_7213	F: TTGCATTCCCACCATTTCATCC	(AC)7	161–248	2.5	PET	MH704708
	R: AGCCAACCTCGTATTA21ATTGCC
BB_2888	F: CTTCATTCATGTACAAGAGCGC	(AC)16	219–232	2.5	FAM	MH704709
	R: AGAACTGGCTATGGATCGAAATG
BB_8681	F: ATCTCCAGTTTCCGTGATTTGC	(ACC)8	204–222	2.5	NED	MH704710
	R: TACGTCACAAGAGCTTTCAACC
BB_31555	F: TGTATAACTGAGATAACCCAGACG	(AC)7	150–156	2.0	VIC	MH704716
	R: TTGTTACCTTGTTCCGGCAAAG

^aOptimal annealing temperature was 62°C for all loci except for BB_2888, where it was 63°C.

Microsatellite marker data analysis

Alleles were scored using GeneMapper version 5 (Applied Bio‐systems), and MICRO‐CHECKER version 2.2.3 (van Oosterhout et al., 2004, 2006) was used to identify null alleles and potential scoring errors from stuttering or large allele dropout. There was no significant evidence of null alleles (P > 0.05) in the Zetterberg Preserve and SBDNL populations. However, null alleles were suggested for loci BB_3968, BB_5021, and BB_8681 in the Petoskey State Park population (Table 2). Homozygote excess for the Petoskey State Park population is not surprising, given this population's reduced number of alleles at each locus and small comparative population size. We characterized genetic diversity by examining the number of alleles and expected and observed heterozygosity for each locus averaged over each population (Table 2) using the package STRATAG in the R statistical program (Archer et al., 2016). The number of observed alleles ranged from one to 10. Some loci were monomorphic for one population but polymorphic when analysis included all populations (e.g., BB_4258).

Table 2.

Genetic properties of the 16 developed polymorphic nuclear microsatellite loci for three populations of Gypsophila paniculata.a

Locus	Zetterberg Preserve (n = 30)					Sleeping Bear Dunes (n = 33)					Petoskey State Park (n = 30)
	A	H o b	H e	F IS	r	A	H o b	H e	F IS	r	A	H o b	H e	F IS	r
BB_3335	8	0.70	0.82	0.09	0.07	7	0.80	0.80	−0.01	0.00	3	0.50	0.56	0.02	0.05
BB_3913	4	0.74	0.68	−0.04	−0.06	4	0.59	0.61	0.05	0.00	2	0.13	0.13	−0.06	−0.06
BB_5567	4	0.63	0.60	−0.04	−0.04	4	0.63	0.63	0.01	−0.01	3	0.53	0.62	0.14	0.05
BB_4443	10	0.60	0.64	0.03	0.03	8	0.76	0.77	−0.02	0.00	5	0.80	0.71	−0.06	−0.09
BB_21680	4	0.53	0.56	0.01	0.02	3	0.52	0.56	0.02	0.05	4	0.67*	0.58	0.23	−0.10
BB_3968	3	0.50	0.42	−0.14	−0.11	4	0.35*	0.49	0.15	0.13	2	0.04	0.04	—	0.28
BB_1355	2	0.17	0.26	0.37	0.13	2	0.14	0.13	−0.07	−0.08	2	0.37	0.51	0.29	0.13
BB_5151	2	0.20	0.18	−0.10	−0.11	2	0.49	0.51	0.11	0.00	2	0.03	0.03	—	−0.02
BB_14751	6	0.79	0.71	−0.06	−0.07	6	0.76	0.76	−0.02	−0.01	3	0.20	0.29	0.21	0.12
BB_4258	3	0.40	0.33	−0.11	−0.22	1	0.00	0.00	—	0.00	1	—	—	—	0.00
BB_6627	2	0.30	0.26	−0.16	−0.16	2	0.46	0.49	−0.03	0.04	1	—	—	—	0.00
BB_5021	5	0.77	0.73	0.00	−0.05	5	0.54	0.60	0.15	0.02	1	—	—	—	0.30
BB_7213	5	0.63	0.61	−0.05	−0.02	3	0.59	0.58	−0.13	0.04	2	0.37	0.38	0.04	0.01
BB_2888	6	0.90	0.80	−0.11	−0.07	6	0.71	0.82	0.20	0.06	2	0.57	0.50	−0.13	−0.08
BB_8681	4	0.47	0.51	0.23	0.01	4	0.38	0.36	−0.07	−0.05	3	0.41	0.53	0.28	0.17
BB_31555	4	0.60	0.67	0.07	0.05	4	0.46	0.59	0.11	0.12	2	0.23	0.26	0.10	0.04

— = data not available because only one allele present at locus or because A > 1 is due to rare allele; A = number of alleles; F _IS = fixation index; H _e = expected heterozygosity; H _o = observed heterozygosity; n = number of individuals sampled; r = null allele frequency.

^aVoucher and locality information for the populations are provided in Appendix 1.

^bSignificant deviation from Hardy–Weinberg equilibrium (*P < 0.05).

The Zetterberg Preserve population displayed slightly higher heterozygosity values (0.17–0.90) than Sleeping Bear Dunes (0.14–0.80), but the Petoskey State Park population (0.03–0.80) had much lower heterozygosity values in comparison. A probability test for Hardy–Weinberg equilibrium (HWE), calculation of the fixation index (F _IS), and linkage disequilibrium were performed in GENEPOP 4.2 (Raymond and Rousset, 1995; Rousset, 2008). The default parameters for Markov chain Monte Carlo (MCMC) iterations were used to calculate HWE. All loci were in HWE except locus BB_3968 for SBDNL and locus BB_21680 for Petoskey State Park (Table 2). The F _IS estimates were calculated using the probability model following Robertson and Hill (1984). Statistical tests for genetic linkage disequilibrium were performed using the log likelihood ratio statistic (G‐test) and MCMC algorithm by Raymond and Rousset (1995). Two pairs of loci were significantly out of linkage disequilibrium (P < 0.05) for both the Zetterberg Preserve and Sleeping Bear Dunes populations (BB_5021 and BB_2888, BB_3913 and BB_1355). Out of 16 loci, five successfully amplified in the related G. elegans M. Bieb. (BB_4443, BB_4258, BB_7213, BB_5151, BB_1355) (Table 3).

Table 3.

Results of cross‐amplification of microsatellite loci isolated from Gypsophila paniculata and tested in 12 G. elegans individuals.a

Locus	Allele size range (bp)	GYEL 01	GYEL02	GYEL03	GYEL04	GYEL07	GYEL08	GYEL09	GYEL12	GYEL13	GYEL15	GYEL18	GYEL19
BB_3335
BB_3913
BB_5567
BB_4443	152	X	X	X	X	X	X	X	X	X	X	X	X
BB_21680
BB_3968
BB_1355	209–222	X	X	X	X	X	X	X	X	X	X	X	X
BB_5151	202–205	X	X	X	X	X	X	X	X	X	X	X	X
BB_14751
BB_4258	160–169	X	X	X	X	X	X	X	X	X	X	X	X
BB_6627
BB_5021
BB_7213	161–171	X	X	X	X	X	X	X	X	X	X	X	X
BB_2888
BB_8681
BB_31555

X = successful amplification in each G. elegans individual (n = 12).

^a Gypsophila elegans tissue was sourced from individuals grown from seed purchased from Everwilde Farms (#FGYPELE‐05, lots 36863 and 39619) in a greenhouse at the AWRI‐GVSU facilities.

CONCLUSIONS

The 16 microsatellite primers developed for G. paniculata provide a tool for estimating genetic diversity and structure of invasive populations, which will aid in understanding its invasion history, identifying source populations, and examining dispersal patterns. Although we developed these markers to specifically study the Lake Michigan dune system populations, G. paniculata is invasive throughout North America. With these markers, we can begin to understand the invasion of G. paniculata in North America in order to improve management efforts and prevent the further spread of this species.

DATA ACCESSIBILITY

A summary of the microsatellite library development and sequence analysis protocols provided to us by the Department of Ecology and Evolutionary Biology at Cornell University are in Appendix S1. FASTA sequences for the 16 microsatellite primers developed here are in Appendix S5. The FASTA file listing all identified contigs containing microsatellite regions is in Appendix S2. Potential primer pairs for the identified microsatellite‐containing contigs are in Appendix S3. The 107 G. paniculata–specific primer pairs tested during primer optimization are in Appendix S4. Raw sequence files for the microsatellite library have been deposited to the National Center for Biotechnology Information (NCBI) Sequence Read Archive (BioProject no. PRJNA431197), and microsatellite sequences have been deposited to GenBank (Table 1). Voucher specimens for each population have been deposited into the Grand Valley State University Herbarium (GVSC), Grand Valley State University Department of Biology, Allendale, Michigan, USA (Appendix 1).

Supporting information

APPENDIX S1. Microsatellite library development and sequencing methods for Gypsophila paniculata.

APPENDIX S2. A total of 58,907 contigs potentially containing microsatellite regions from Gypsophila paniculata.

APPENDIX S3. Information about 7904 unique microsatellite primer pairs identified by MSATCOMMANDER version 1.0.3 for Gypsophila paniculata.

APPENDIX S4. One hundred and seven Gypsophila paniculata–specific microsatellite primer pairs tested during primer optimization.

APPENDIX S5. Sequences for the 16 Gypsophila paniculata microsatellite loci.

APPENDIX 1. Voucher and location information for Gypsophila paniculata populations used in this study.

Voucher specimen accession no.a	Collection locality	Geographic coordinates	N
VPSMI551, VPSMI554, VPSMI569–VPSMI571	Petoskey State Park, Emmet County, MI, USA	45.40288418°N, 84.91271857°W	30
VSBMI001–VSBMI016	Sleeping Bear Dunes National Lakeshore, Leelanau County, MI, USA	44.87372°N, 86.06170°W	33
VZPMI001–VZPMI005	Zetterberg Preserve at Point Betsie, Benzie County, MI, USA	44.68231°N, 86.25322°W	30

N = number of individuals sampled.

^aVouchers deposited at Grand Valley State University Herbarium (GVSC), Grand Valley State University Department of Biology, Allendale, Michigan, USA.

Leimbach‐Maus, H. B. , Parks S. R., and Partridge C. G.. 2018. Microsatellite primer development for the invasive perennial herb Gypsophila paniculata (Caryophyllaceae). Applications in Plant Sciences 6(12): e1203.