Exceptional transmission of plastids and mitochondriafrom the transplastomic pollen parent and its impacton transgene containmentZora Svab and Pal Maliga*

Waksman Institute, Rutgers, The State University of New Jersey, 190 Frelinghuysen Road, Piscataway, NJ 08854-8020

Edited by Maarten Koornneef, Wageningen University and Research Centre, Wageningen, The Netherlands, and approved February 28, 2007(received for review January 4, 2007)

Plastids in Nicotiana tabacum are normally transmitted to theprogeny by the maternal parent only. However, low-frequencypaternal plastid transmission has been reported in crosses involv-ing parents with an alien cytoplasm. Our objective was to deter-mine whether paternal plastids are transmitted in crosses betweenparents with the normal cytoplasm. The transplastomic father linescarried a spectinomycin resistance (aadA) transgene incorporatedin the plastid genome. The mother lines in the crosses were either(i) alloplasmic, with the Nicotiana undulata cytoplasm that conferscytoplasmic male sterility (CMS92) or (ii) normal, with the fertile N.tabacum cytoplasm. Here we report that plastids from the trans-plastomic father were transmitted in both cases at low (10�4-10�5)frequencies; therefore, rare paternal pollen transmission is notsimply due to breakdown of normal controls caused by the aliencytoplasm. Furthermore, we have found that the entire plastidgenome was transmitted by pollen rather than small plastidgenome (ptDNA) fragments. Interestingly, the plants, which inher-ited paternal plastids, also carried paternal mitochondrial DNA,indicating cotransmission of plastids and mitochondria in the samepollen. The detection of rare paternal plastid transmission de-scribed here was facilitated by direct selection for the transplas-tomic spectinomycin resistance marker in tissue culture; therefore,recovery of rare paternal plastids in the germline is less likely tooccur under field conditions.

Nicotiana tabacum � organelle inheritance � plastid transformation �pollen transmission

DNA in a plant cell is found in three cellular compartments:the nucleus, plastids, and mitochondria. Genes encoded in

the nucleus are inherited biparentally, according to Mendel’srules. In contrast, plastids and mitochondria may be inheritedmaternally, paternally, or from both parents. In most crops, thematernal parent transmits plastids, because plastids are excludedfrom the sperm cell or, even if not excluded, are left behind insynergid cells during fertilization (1–3). In Chlamydomonasreinhardtii, a unicellular alga, in which maternal (mating-type�)and paternal (mating-type�) chloroplasts fuse, only the maternalplastid genome (ptDNA) is inherited, because the maternalptDNA is protected by methylation, whereas the nonmethylatedpaternal ptDNA is degraded (4). Once protocols for plastidtransformation were developed (5, 6), localization of transgenesin the plastid genome was advocated as a means of transgenecontainment (7). Utility of plastid localization for containmentbecame a hotly debated issue when the first herbicide-resistanttransplastomic plants were obtained (8, 9). The reason forskepticism was the reported relatively high-frequency paternalptDNA transmission in species in which plastids were assumedto be inherited strictly by the maternal parent. Paternal ptDNAtransmission was detected in crosses by using streptomycinresistance (in 0.07–2.5%) (10, 11) and tentoxin resistance (0.5–2.5%) (12) in tobacco, pigment deficiency in petunia (2%) (13,14), and atrazine resistance in foxtail or birdseed millet (0.03%)(15).

Common in these studies was utilization of alloplasmic sub-stitution lines, in which plastids and mitochondria of one specieswere combined with the nuclear genome of another species. Wetherefore decided to test whether rare paternal plastid trans-mission occurs in Nicotiana tabacum if no alloplasmic substitu-tion line is involved in the cross.

Here we report that paternal ptDNA is transmitted by pollenin crosses of plants with a normal cytoplasm in one of �10,000seedlings. We have also shown that the entire plastid genome istransmitted by pollen rather than small ptDNA fragments, andthat mitochondria (mitochondrial DNA) are always cotransmit-ted with the paternal ptDNA. Detection of rare paternal ptDNAtransmission described here is biased by tissue culture selection,and the transgenic ptDNA is less likely to get into the germ lineunder field conditions.

ResultsPaternal ptDNA Transmission to Alloplasmic Cytoplasmic Male Sterile(CMS) 92 Mother. We first tested our protocols for paternalptDNA transmission using an alloplasmic N. tabacum line,CMS92, as maternal parent. Line CMS92 was obtained byrepeated backcrossing of Nicotiana undulata with N. tabacumthat resulted in the replacement of the N. tabacum cytoplasmwith the N. undulata cytoplasm. The ptDNA of paternal lines(Table 1) carried chimeric aadA genes in the trnV-3�rps12intergenic region that confer spectinomycin and streptomycinresistance. Thus, seedlings that acquired a few copies of paternalptDNA were expected to be spectinomycin- and streptomycin-resistant because of expression of aadA and male sterile, aCMS92 mitochondrial trait (12, 16) (Fig. 1A).

Seedlings were screened for paternal ptDNA by testing themfor the spectinomycin-resistance gene on a selective medium(500 mg/liter of spectinomycin). Four protocols were used. Theseeds were surface sterilized and (i) germinated on plant growth[revised medium for plant maintenance in sterile culture (RM)]medium; (ii) germinated on callus induction [revised medium fororganogenesis (shoot regeneration) of Nicotiana plumbaginifolia(RMOP)] medium; (iii) germinated on plant growth (RM)medium then transferred to callus induction (RMOP) medium(RM/RMOP protocol); (iv) germinated on callus induction

Author contributions: Z.S. and P.M. designed research; Z.S. and P.M. performed research;Z.S. and P.M. analyzed data; and P.M. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Freely available online through the PNAS open access option.

Abbreviations: CMS, cytoplasmic male sterile; mtDNA, mitochondrial genome; ptDNA,plastid genome; RM medium, revised medium for plant maintenance in sterile culture;RMOP medium, revised medium for organogenesis (shoot regeneration) of Nicotianaplumbaginifolia; RFLP, restriction fragment length polymorphism.

See Commentary on page 6879.

*To whom correspondence should be addressed. E-mail: maliga@waksman.rutgers.edu.

© 2007 by The National Academy of Sciences of the USA

www.pnas.org�cgi�doi�10.1073�pnas.0700063104 PNAS � April 24, 2007 � vol. 104 � no. 17 � 7003–7008

AG

RICU

LTU

RAL

SCIE

NCE

SSE

ECO

MM

ENTA

RY

Dow

nloa

ded

by g

uest

on

Mar

ch 1

2, 2

021

medium and transferred onto callus induction medium (RMOP/RMOP protocol). For protocols i and ii plates were inspected forresistance for up to 4 mo. For protocols iii and iv, seedlings weregerminated on the first selective medium (RM or RMOP) for2–3 wk (200–300 seedlings per 10-cm Petri dish), then trans-ferred to the RMOP medium (25 per 10-cm dish) and inspectedfor spectinomycin resistance for up to 4 mo. Results by paternalparent are listed in Table 1; cumulative data are in Table 2.

Spectinomycin-sensitive seedlings or calli were white, whereasresistant sectors or calli were green on the selective medium (Fig.1C). Resistance to spectinomycin could be due to expression ofthe Paternal aadA gene (lines designated PSpc and a number),or because of a new spontaneous point mutation in the 16SrRNA (lines designated Spc and a number). Seedlings carryingpaternal plastids (aadA gene) were also streptomycin-resistant(17). In contrast, spontaneous spectinomycin-resistant mutants(Spc93) were streptomycin-sensitive, because mutations in the

16S rRNA that prevent spectinomycin binding do not confercross resistance to streptomycin (Fig. 1D) (18, 19). PaternalptDNA transfer events were recovered following each of theprotocols (Table 1). The frequency of seedlings with paternalptDNA was in the range of 1–6 � 10�4. The overall frequency,10 events in 47,859 seedlings, corresponds to the frequency of2 � 10�4.

Paternal pollen transmission has been confirmed by DNAgel-blot analysis indicating the regenerated plants carried theplastid genome of the donor N. tabacum paternal parent and notthe maternal N. undulata ptDNA (see below). Plants regener-ated from 9 of the 10 (exception was PSpc1) paternal ptDNAtransmission lines have been purified to homoplasmy by oneadditional round of plant regeneration and transferred to thegreenhouse. The flowers in each of the nine lines had the CMS92petaloid phenotype of the maternal parent (Fig. 1B) and pro-duced seed after fertilization with wild-type pollen.

Table 1. Paternal ptDNA transmission to CMS maternal parent detected by selection for spectinomycin resistance (500 mg/liter)

Pollen donor Protocol No. of seedlings No. of ptDNA transfer events No. of mutants

Nt-pMHB10 RM 8,700 1 (PSpc1) 0Nt-pMSK56 RM 2,046 1 (PSpc4) 0Nt-pMSK56 6,590 0 0Nt-pMHB10 RMOP 1,000 1 (PSpc66) 2Nt-pMSK56 RMOP 2,610 2 (PSpc2, PSpc3) 1Nt-pMSK56 18,420 0 0Nt-pMSK56 RM/RMOP 128 0 0Nt-pMSK56 RM/RMOP 1,137 1 (PSpc7) 3Nt-pMSK56 RMOP/RMOP 4,050 0 4Nt-pJEK6 RMOP/RMOP 3,178 4 (PSpc23, PSpc24, PSpc25, PSpc32) 6

Fig. 1. Transmission of paternal ptDNA in alloplasmic crosses. (A) Flower morphology of the male sterile N. tabacum CMS92 mother line and fertile Nt-pMSK56transplastomic father line. Below are listed nuclear, plastid, and mitochondrial markers. (B) PSpc32 line has petaloid flowers and is male sterile, as the CMS92mother line. (C) The PSpc4 (RM protocol) and PSpc2, PSpc3, and PSpc66 (RMOP protocol) paternal pollen transmission events. Arrows point to green resistantsectors. (D) Classification of spectinomycin-resistant lines. The aadA gene in paternal ptDNA confers streptomycin resistance to PSpc4 cells. Spontaneousspectinomycin-resistant mutant Spc93 is sensitive to streptomycin. Mother line N. tabacum CMS92 (sensitive to both drugs) and father line Nt-pMSK56 (resistantto both drugs) were controls. Leaf sections were cultured on RMOP medium containing 500 mg/liter spectinomycin or 500 mg/liter of both drugs.

7004 � www.pnas.org�cgi�doi�10.1073�pnas.0700063104 Svab and Maliga

Dow

nloa

ded

by g

uest

on

Mar

ch 1

2, 2

021

In preliminary experiments, selection for streptomycin resis-tance (500 mg/liter) yielded two paternal ptDNA transmissionevents (PSt1 and PSt6). Because first selection on streptomycinmedium yielded seedlings that often turned out to be sensitivein the second test, experiments to recover seedlings with paternalptDNA by streptomycin selection were abandoned.

Paternal ptDNA Transmission in Cross with Normal Cytoplasm. Test-ing of paternal ptDNA transmission between parents with thenormal cytoplasm was carried out by using one transplastomicspectinomycin-resistant father, Nt-pMSK56, and mother lines,Nt-pHC18 or Nt-pHC19, that did not have any plastid geneticmarker but carried a (nonsegregating) nuclear gentamycin-resistance gene. Therefore, seedlings acquiring the aadA genefrom the paternal parent were expected to be resistant tospectinomycin, streptomycin, and gentamycin (Fig. 2A).

Three paternal ptDNA transfer events have been recovered in34,115 seedlings following the RMOP/RMOP protocol (Fig.2B), a frequency of 9 � 10�5 (Table 2). Transfer of paternalptDNA as the source of spectinomycin resistance has beenconfirmed by demonstrating resistance to streptomycin andgentamycin (Fig. 2C). Hybrid origin of the PSpc70 line was also

indicated by segregation of its seed progeny for gentamycinresistance (Fig. 2D).

Transmission of Entire ptDNA by Pollen. Transfer of ptDNA frag-ments to the nucleus has been shown during reproduction (20)and in somatic cells (21). Thus, transfer of paternal ptDNA mayoccur within an intact organelle or by a transformation-likeprocess. In case of the transfer of organelles, we expected to findthe entire paternal ptDNA, whereas transfer that involves trans-formation is likely to involve only ptDNA fragments. To distin-guish between the two possibilities, we tested species-specificptDNA restriction fragment length polymorphism (RFLP)markers in progeny to which paternal ptDNA has been trans-ferred. Because the ptDNA sequence is available only for N.tabacum (22), first we identified RFLP markers that are suitableto distinguish the N. tabacum and N. undulata ptDNA. This wasaccomplished by digesting total leaf DNA of the two parentallines with a battery of eight restriction endonucleases that yield100–400 ptDNA fragments each, then probing the blotted DNA(Fig. 3). We found polymorphic sites in five sets of genes in thelarge unique region (atpI-rpoC2, psbC-psaA, trnV-atpB, psbJ-psbE, and rps8-rps3), the insertion site of aadA in the invertedrepeat (3�-rps12-rrn16-trnV), and one set of genes (psaC-ndhA)

Table 2. Rare paternal ptDNA transmission detected by selection for spectinomycin resistance (500 mg/liter)

Maternal parentGerminate/

transferNo. of

seedlingsNo. of ptDNA

transfer events No. of mutantsFrequency of ptDNA

transferFrequency of

mutants

CMS RM 17,336 2 0 1 � 10�4 �6 � 10�5

RMOP 22,030 3 3 1 � 10�4 1 � 10�4

RM/RMOP 1,265 1 3 8 � 10�4 2 � 10�3

RMOP/RMOP 7,228 4 10 6 � 10�4 1 � 10-3

Fertile RMOP 31,897 0 1 �3 � 10�5 3 � 10�5

RM/RMOP 1,350 0 6 �6 � 10�4 4 � 10�3

RMOP/RMOP 34,115 3* 26 9 � 10�5 8 � 10�4

*PSpc64, PSpc68, and PSpc70.

Fig. 2. Transmission of paternal ptDNA between parents with normal cytoplasm. (A) Nuclear (N), plastid (Pt), and mitochondrial (Mt) genotypes of parentallines and exceptional hybrid with paternal ptDNA. (B) Identification of PSpc70 paternal ptDNA transmission event and Spc69 spontaneous spectinomycin-resistant mutant in RMOP/RMOP protocol. (C) Classification of spectinomycin-resistant lines. Paternal aadA gene confers streptomycin resistance to PSpc70. Spc69spontaneous spectinomycin-resistant mutant is sensitive to streptomycin. Both lines are gentamycin-resistant. Nt-pHC19 and Nt-pMSK56 are parental lines. (D)Germination of seedlings on spectinomycin medium indicates that the PSpc70 paternal ptDNA transmission line is homoplastomic for the aadA transgene andsegregates for the nuclear gentamycin resistance marker.

Svab and Maliga PNAS � April 24, 2007 � vol. 104 � no. 17 � 7005

AG

RICU

LTU

RAL

SCIE

NCE

SSE

ECO

MM

ENTA

RY

Dow

nloa

ded

by g

uest

on

Mar

ch 1

2, 2

021

in the small unique region. Probing of total cellular DNAisolated from greenhouse-grown homoplastomic plants revealedonly N. tabacum paternal ptDNA markers in the nine plantsselected by spectinomycin resistance (PSpc2-PSpc66) and in twoplants selected by streptomycin resistance (PSt1 and PSt6)(Table 3). We conclude, therefore, that the entire ptDNA istransmitted by pollen, most likely in intact organelles.

Cotransmission of Mitochondrial and Plastid DNAs. Paternal trans-mission of plastids naturally triggers the question whether thenonselected mitochondria (mitochondrial DNA) is cotransmit-ted with the ptDNA from the father line. To answer thisquestion, we first identified polymorphic restriction sites in themitochondrial genome (mtDNA) using a mtDNA fragmentcontaining the atp6 gene. Digestion of total cellular DNA withthe SspI enzyme and probing with the atp6 probe revealedunique fragments in both the N. tabacum and N. undulatamitochondria. Probing of each of the 11 paternal ptDNA transferlines revealed both N. tabacum- and N. undulata-specific frag-ments (Fig. 4), indicating that N. tabacum mitochondria weretransmitted with the plastids by pollen.

DiscussionPollen Transmission of ptDNA. We report here that pollen trans-mission of ptDNA occurs at a low frequency in crosses of N.tabacum lines carrying the normal cytoplasm. Thus, it appearsthat exclusion of paternal ptDNA fails at the frequency of 1 of10,000 seedlings. Indeed, during male reproductive cell matu-ration, occasional plastids have been observed in tobacco spermcells that could form the cytological bases of paternal pollentransmission (23). Alternatively, organellar inclusions in thesperm nucleus could provide a potential mechanism for trans-mitting organellar DNA into the next generation (24).

We normally evaluate seedling phenotype 2 wk after spreading

200–300 seeds on a selective (500 mg/liter) spectinomycin me-dium. Formation of resistant sectors due to expression ofpaternal aadA genes may take longer, 6–16 wk, and the sectorsare relatively small. Thus, they are likely to remain undetectedin a standard seed assay unless a large number of seedlings orseedling calli are cultivated for extended periods. Because of thelack of suitable genetic markers, transmission of paternal ptDNAby pollen thus far had been shown only in alloplasmic crosses,leaving open the possibility that paternal ptDNA transmission isdue to the breakdown of normal control processes in thealloplasmic cross. Indeed, the frequency of paternal pollentransmission reported in alloplasmic crosses, in general, wassignificantly higher, 0.5–2.5% of seed progeny (10–14) than thevalue we found, 9 � 10�5, in the cross with the normal cytoplasm.However, the frequency of paternal pollen transmission in ouralloplasmic cross was also relatively low, 2 � 10�4, and only�2-fold higher than the frequency in the cross between thenormal parental lines. We believe this is due to our choice oftobacco line, N. tabacum cv. Petit Havana. Indeed, line-specificdifferences within a species are important; in Petunia, low-frequency paternal ptDNA transmission was detectable in only5 of the 22 inbreds tested (25).

Protocols for Testing Paternal ptDNA Transmission. The number ofcell divisions in a developing plant is determined by the plant’sdevelopmental program. In contrast, tobacco cells in a culturecan be grown indefinitely, so long as the cells are transferred tofresh medium at regular intervals. We were interested to find outwhether the limitation of cell division in a seedling wouldinterfere with the recovery of paternal ptDNA transfer events.Based on data in Table 2, it appears it is not necessary to callusthe seedlings to identify paternal pollen transmission eventswhen screening for a transgenic spectinomycin resistancemarker. Although the frequency of paternal ptDNA transmis-

Table 3. RFLP markers in paternal ptDNA transmission lines

Clones

atpI-rpoC1 psbC-psaA trnV-atpB psbJ-psbE rps8-rps3 3�-rps12-rrn16 psaC-ndhA

AseI AluI HpaII AvaII EcoRI CIaI EcoRI HpaII StyI BamHI HaeIII HpaII

PSpc2 t t t t t t t t t t tPSpc3 t t t t t t t t t t tPSpc4 t t t t t t t t t t tPSpc7 t t t t t t t t t t tPSpc23 t t t t t t t t t tPSpc24 t t t t t t t t t tPSpc25 t t t t t t t t t tPSpc32 t t t t t t t tPSpc66 t t t t t t tPSt1 t t t t t t t t t t tPSt6 t t t t t t t t t t

Fig. 3. RFLP mapping of paternal ptDNA. DNA gel blot to identify RFLP markers in the CMS92 ptDNA. Lanes contain total cellular DNA of N. tabacum cv. PetitHavana (t), the CMS92 line (u), and transplastomic Nt-pMSK56 line (TP). The genes covered by the probe and restriction enzymes used for DNA digestion are listedon top. Fragment sizes are shown for N. tabacum.

7006 � www.pnas.org�cgi�doi�10.1073�pnas.0700063104 Svab and Maliga

Dow

nloa

ded

by g

uest

on

Mar

ch 1

2, 2

021

sion events was independent of the protocol, the number ofspontaneous spectinomycin-resistant mutants was not; the two-step RMOP/RMOP protocol with repeated subculture on callusinduction medium yielded 10 times more mutants than proto-cols, including only one subculture on callus induction medium(Table 2). Thus, longer culture and transfer onto fresh medium(more cell divisions) enabled recovery of more spontaneousmutants.

Cotransmission of Paternal Plastid and Mitochondrial DNA. Takingadvantage of species-specific plastid and mitochondrial RFLPmarkers in plants derived from the alloplasmic cross, we couldaddress two important issues that are untraceable in the normalcross. First, we found by testing seven regions in the ptDNA, allRFLP markers derive from the paternal parent (Table 3). Thus,the entire ptDNA is transmitted, rather than small ptDNAfragments from the paternal parent, most likely in an intactplastid. There is no evidence for a transformation-like processthat was reported for ptDNA transferred from the plastid to thenucleus (20, 21) or ptDNA recombination detected after chlo-roplast fusion (26). This conclusion is in agreement with earlierreports that failed to show any deviation from paternal ptDNA(10–13).

We also found that each of the alloplasmic lines that acquiredpaternal ptDNA also has the maternal CMS92 phenotype.Maintenance of maternal CMS phenotype and mtDNA restric-tion patterns led earlier investigators to the conclusion that themtDNA is not transmitted together with the selected ptDNA(10–12). Interestingly, we found that each of the paternal ptDNAtransfer lines has mixed mitochondrial DNA RFLP markers (Fig.4). Thus, the mitochondria in these plants carry some of the N.tabacum mtDNA sequence in addition to the mitochondrialsequence causing the petaloid CMS92 phenotype (16, 27).Finding mixed parental mtDNA is not surprising. Mitochondriain plant cells may participate in a massive fusion cycle (28), andrecombination of mtDNA following mitochondrial fusion is welldocumented (see, for example, ref. 29). Cotransfer of mtDNAand ptDNA in earlier studies may have been missed because oflimited probing, or because it does not occur in all speciescombinations. Mitochondria are more frequent in sperm cellsthan plastids, and inclusion of both plastids and mitochondria in

the cytoplasm or nucleus of exceptional sperm cells couldprovide a shared mechanism for transmitting both organellesinto the next generation (23, 24).

Transgene Containment. One of our objectives was to determinethe absolute number of paternal ptDNA transmission events ina normal cross. When using a one-step protocol, we may havemissed some paternal ptDNA transmission events because ofnutritional limitation that prevented formation of a visible sectorfrom each of the carrier cells. With the two-step protocol, werecovered significantly more spontaneous mutants but not sig-nificantly more transmission events; thus, the one event per�10,000 seedlings is the absolute number of ptDNA transmis-sion events in N. tabacum cv. Petit Havana, the tobacco line westudied. The absolute number of ptDNA transmission eventsseen here is probably much higher than the number of eventslikely to appear under field conditions. The plastid genome ishighly polyploid (30) and, when the cells divide, there is no exactduplication of the cytoplasm. Presence of ptDNA copies in thevegetative plant tissue depends on the paternal ptDNA copiesbeing present in the shoot apex. Paternal ptDNA copies wereclearly present in the original shoot apex of only two of theseedlings, PSpc3 and PSpc4 (Fig. 1C) of the 12 events thatyielded plants. (Note that we disregard newly differentiatedshoot apexes, such as PSpc66 in Fig. 1C.) Formation of raresectors with a few copies of paternal ptDNA are biased by tissueculture selection; therefore, they are less likely to get into thegerm line under field condition.

Transgene flow within crops and between crops and wildrelatives is a genuine concern (31, 32). Experiments reportedhere are the first steps to design genetic screens that will lead toidentification of nuclear genes controlling plastid pollen trans-mission. Understanding the genetic control of plastid inheri-tance will facilitate rational design of new transgenic crops inwhich escape of plastid transgenes by pollen can be minimized.

Materials and MethodsPlant Lines. Mother lines with the normal cytoplasm were Nt-pHC18 or Nt-pHC19, N. tabacum cv. Petit Havana plantstransformed with an aacC1 gentamycin resistance gene ex-pressed in a caulif lower mosaic virus 35S cassette (33). In theNt-pHC19 plants, an XbaI site is blunted downstream of theaacC1 coding region. CMS92 seed was kindly provided by EzraGalun in the N. tabacum cv. Samsun background (16). TheCMS92 mother line was an F1 hybrid of cv. Samsun and cv. PetitHavana. Father lines N. tabacum cv. Petit Havana Nt-pMHB10(34), Nt-pMSK56 (35), and Nt-pJEK6 plants (36) carry trans-plastomic spectinomycin resistance (aadA) genes. The fertilemother line was manually emasculated before pollination.

Tissue Culture and Selection. The RM plant growth medium wasagar-solidified containing Murashige and Skoog salts (37) and3% sucrose (pH 5.8). The RMOP medium is an RM mediumcontaining benzyladenine (1 mg/liter), naphthaleneacetic acid,

Table 4. DNA probes to test N. tabacum and N. undulata (CMS92) ptDNA polymorphisms

Region Probe Fragment, ptDNA DNA source Ref.

atpI-rpoC1 SacI; 7.2 kb SacI; 7.2 kb; 15662–22868 pGS98 42psbC-psaA PstI-SacII; 5.1 kb PstI-SacII; 5.1 kb; 36830–42000 pRB10 43trnV (trnA-Val(UAC)-atpB EcoRI; 3.5 kb Sal6; 49841–65310 pBR322::Sal6 26psbJ-psbE SalI-SpeI; 2.5 kb SalI-SpeI; 65310–67693 pRB1 44rps8-rps3 SalI; 2.9 kb SalI; 2.88 kb; 82754–85634 pBR322::Sal10 263�-rps12-rrn16 BamHI-ApaI; 2.1 kb BglII-EcoRI; 101145–104081 pPRV1 45psaC-ndhA PCR fragment; 3.2 kb Primers in ptDNA at: 119134–122304* Tobacco leaf DNA This paper

*Primers: 5�-CTCGAACGTATCAATAAGCTAGAC-3�; 5�-CGTATGAGATGAAAATCTCACGTAC-3�.

Fig. 4. Plants with paternal ptDNA have mtDNA from both parents. Note N.tabacum- (t) and N. undulata- (u) specific mtDNA fragments in parental linesand in plants obtained by spectinomycin (PSpc2–66) and streptomycin (PSt1,PSt6) selection. Total cellular DNA was probed with the 6.9-kb atp6 probe.

Svab and Maliga PNAS � April 24, 2007 � vol. 104 � no. 17 � 7007

AG

RICU

LTU

RAL

SCIE

NCE

SSE

ECO

MM

ENTA

RY

Dow

nloa

ded

by g

uest

on

Mar

ch 1

2, 2

021

(0.1 mg/liter), and thiamine (1 mg/liter) (38). Antibiotics spec-tinomycin dihydrochloride and streptomycin sulfate were filter-sterilized and added after autoclaving. The seeds were sterilizedby exposure to the vapor of 100 ml of 6% sodium hypochlorite(Clorox bleach) and 3 ml of concentrated HCl for 8 h. Seeds werecounted with AlphaImager 2000 (Alpha Innotech, San Leandro,CA). General protocols for plant tissue culture have beendescribed elsewhere (17, 39).

DNA Gel Blot Analysis. The protocols for DNA isolation, diges-tion, and probing have been described (17, 39). The plastid

DNA probes are listed in Table 4. Mitochondrial DNA wasprobed with a 6.9-kb PstI fragment containing the 3�-half ofatp6 gene (40) generously provided by C. S. Levings, NorthCarolina State University, Raleigh, NC. For the complete N.tabacum mtDNA sequence, see GenBank accession no.BA00004 (41).

We thank Arun K. Azhagiri and Kerry Ann Lutz for discussions. Thisresearch was supported by Award 2004-39454-15192 of the U.S. Depart-ment of Agriculture Biotechnology Risk Assessment Research GrantProgram.

1. Reboud X, Zeyl C (1994) Heredity 72:132–140.2. Mogensen HL (1996) Am J Bot 83:383–404.3. Hagemann R (2002) Prog Bot 63:5–51.4. Nishiyama R, Ito M, Yamaguchi Y, Koizumi N, Sano H (2002) Proc Natl Acad

Sci USA 99:5925–5930.5. Boynton JE, Gillham NW, Harris EH, Hosler JP, Johnson AM, Jones AR,

Randolph-Anderson BL, Robertson D, Klein TM, Shark KB, Sanford JC(1988) Science 240:1534–1538.

6. Svab Z, Hajdukiewicz P, Maliga P (1990) Proc Natl Acad Sci USA 87:8526–8530.

7. Maliga P (1993) Trends Biotech 11:101–107.8. Daniell H, Datta R, Varma S, Gray S, Lee SB (1998) Nat Biotechnol

16:345–348.9. Bilang R, Potrykus I (1998) Nat Biotechnol 16:333–334.

10. Medgyesy P, Pay A, Marton L (1986) Mol Gen Genet 204:195–198.11. Horlow C, Goujaud J, Lepingle A, Missonier C, Bourgin JP (1990) Plant Cell

Rep 9:249–252.12. Avni A, Edelman M (1991) Mol Gen Genet 225:273–277.13. Cornu A, Dulieu H (1988) J Hered 79:40–44.14. Drepas A, Dulieu H (1992) J Hered 83:6–10.15. Wang T, Li Y, Shi Y, Reboud X, Darmency H, Gressel J (2004) Theor Appl

Genet 108:315–320.16. Galun E, Arze-Gonen P, Fluhr R, Edelman M, Aviv D (1982) Mol Gen Genet

186:50–56.17. Svab Z, Maliga P (1993) Proc Natl Acad Sci USA 90:913–917.18. Fromm H, Edelman M, Aviv D, Galun E (1987) EMBO J 6:3233–3237.19. Svab Z, Maliga P (1991) Mol Gen Genet 228:316–319.20. Huang CY, Ayliffe MA, Timmis JN (2003) Nature 422:72–76.21. Stegemann S, Hartmann S, Ruf S, Bock R (2003) Proc Natl Acad Sci USA

100:8828–8833.

22. Shinozaki K, Ohme M, Tanaka M, Wakasugi T, Hayashida N, MatsubayashiT, Zaita N, Chunwongse J, Obokata J, Yamaguchi-Shinozaki K, et al. (1986)EMBO J 5:2043–2049.

23. Yu HS, Russel SD (1994) Planta 193:115–122.24. Yu HS, Russell SD (1994) Plant Cell 6:1477–1484.25. Derepas A, Dulieu H (1992) J Hered 83:6–10.26. Medgyesy P, Fejes E, Maliga P (1985) Proc Natl Acad Sci USA 82:6960–6964.27. Hanson MR, Bentolila S (2004) Plant Cell 16 (Suppl):S154–SI69.28. Sheahan MB, McCurdy DW, Rose RJ (2005) Plant J 44:744–755.29. Boeshore ML, LIfshitz I, Hanson MR, Izhar S (1983) Mol Gen Genet

190:459–467.30. Bendich AJ (1987) BioEssays 6:279–282.31. Ellstrand NC, Prentice HC, Hancock JF (1999) Annu Rev Ecol Syst 30:539–563.32. Riegel MA, Lamond M, Preston C, Powles SB, Roush RT (2002) Science

296:2386–2388.33. Carrer H, Staub JM, Maliga P (1990) Plant Mol Biol 17:301–303.34. Lutz KA, Bosacchi MH, Maliga P (2006) Plant J 45:447–456.35. Khan MS, Maliga P (1999) Nat Biotechnol 17:910–915.36. Lutz KA, Knapp JE, Maliga P (2001) Plant Physiol 125:1585–1590.37. Murashige T, Skoog F (1962) Physiol Plant 15:473–497.38. Sidorov V, Menczel L, Maliga P (1981) Nature 294:87–88.39. Lutz KA, Svab Z, Maliga P (2006) Nat Protoc 1:900–910.40. Bland MM, Levings CS, III, Matzinger DF (1987) Curr Genet 12:475–481.41. Sugiyama Y, Watase Y, Nagase M, Makita N, Yagura S, Hirai A, Sugiura M

(2005) Mol Genet Genomics 272:603–615.42. Serino G, Maliga P (1998) Plant Physiol 117:1165–1170.43. Bock R, Maliga P (1995) Mol Gen Genet 247:439–443.44. Bock R, Kossel H, Maliga P (1994) EMBO J 13:4623–4628.45. Zoubenko OV, Allison LA, Svab Z, Maliga P (1994) Nucleic Acids Res

22:3819–3824.

7008 � www.pnas.org�cgi�doi�10.1073�pnas.0700063104 Svab and Maliga

Dow

nloa

ded

by g

uest

on

Mar

ch 1

2, 2

021