Do feminized weed seeds male sterile plants

Creating Completely Both Male and Female Sterile Plants by Specifically Ablating Microspore and Megaspore Mother Cells

This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

Associated Data


Although genetically modified (GM) plants have improved commercially important traits, such as biomass and biofuel production, digestibility, bioremediation, ornamental value, and tolerance to biotic and abiotic stresses, there remain economic, political, or social concerns over potential ecological effects of transgene flow from GM plants. The current solution for preventing transgene flow from GM plants is genetically engineering sterility; however, approaches to generating both male and female sterility are limited. In addition, existing strategies for creating sterility lead to loss or modifications of entire flowers or floral organs. Here, we demonstrate that instead of the 1.5-kb promoter, the entire SOLO DANCERS (SDS) gene is required for its meiocyte-specific expression. We then developed an efficient method to specifically ablate microspore and megaspore mother cells using the SDS and BARNASE fusion gene, which resulted in complete sterility in both male and female reproductive organs in Arabidopsis (Arabidopsis thaliana) and tobacco (Nicotiana tabacum), but did not affect plant growth or development, including the formation of all flower organs. Therefore, our research provides a general and effective tool to prevent transgene flow in GM plants.

Keywords: completely both male and female sterile plants, flower structure, genetic ablation, gene flow, microspore and megaspore mother cells, SOLO DANCERS


Since genetically modified (GM) plants were produced in 1983 (Caplan et al., 1983; Murai et al., 1983), the number of GM plants has been rapidly increasing yearly (Husken et al., 2010). GM trees, turf grasses, biofuel and forage crops, and ornamentals have improved commercially important traits, including biomass and biofuel production, digestibility, bioremediation, ornamental value, and tolerance to biotic and abiotic stresses (Wang and Ge, 2006; Groover, 2007; Harfouche et al., 2011; Xu et al., 2011; Wang and Brummer, 2012; Wilkerson et al., 2014); however, the approval for commercialization of GM plants is subject to complicated and stringent government regulations due to economic, political, or social concerns over potential ecological effects of transgene flow and floral-modified plantations (Goeschl and Swanson, 2003; Hills et al., 2007; Van Acker et al., 2007; Strauss et al., 2009; Lombardo, 2014).

Transgene flow from GM plants to non-GM plants and wild populations is mainly mediated by dispersal of pollen and seeds. Early studies found that the pollen-mediated gene flew from GM Roundup Ready creeping bentgrass (Agrostis stolonifera) occurred within 2 to 21 km (Watrud et al., 2004). The non-GM rabbitfoot grass (Polypogon monspeliensis) could pollinate the GM creeping bentgrass to produce transgenic intergeneric hybrid offspring, suggesting that the transgene escape is also mediated by the female part of GM plants (Zapiola and Mallory-Smith, 2012). Long distance pollen-mediated gene flow occurred between weed beets (Beta vulgaris) as far as 9.6 km and the resulting interfield gene flow was unavoidable (Fenart et al., 2007). Pollen migration from poplars (Populus trichocarpa) often went beyond 10 km (Slavov et al., 2009; DiFazio et al., 2012), indicating that similar issues happened in GM trees. Moreover, gene flow from GM crops to native populations was detected in maize (Zea mays), soybean (Glycine max), wheat (Triticum aestivum), and canola (Brassica napus; Pineyro-Nelson et al., 2009; Liu et al., 2010b; Rieben et al., 2011; Wang and Li, 2012). To overcome regulatory hurdles to field research and, ultimately, commercial uses of GM plants, a practical solution is to create sterile plants by ablating floral organs/tissues using toxic genes under control of specific promoters or by altering flowering time and floral organs via manipulating genes critical for flower development.

Strategies on making male sterility have been extensively and successfully employed to prevent the pollen-mediated transgene flow. In the male reproduction organ anther, tapetum is a layer of nutritive cells, which is required for pollen development. Therefore, genetic ablation of tapetal cells by tapetum-specific promoters driven toxic genes, such as ribonuclease BARNASE and diphtheria toxin fragment A (DTA) genes, is commonly used to create male sterility in various plants. The widely used tobacco tapetum promoter TA29 was first employed to drive BARNASE to create male sterile tobacco and oilseed rape (Brassica napus) plants (Mariani et al., 1990). TA29::DTA tobacco transgenic plants are also male sterile (Koltunow et al., 1990). Since then, various male sterile plants were achieved using other tapetum or anther-specific promoters, including A9, A6, E1, T72, PS1, and PsEND1 in Arabidopsis, rapeseed (Brassica napus), rice (Oryza sativa), and pea (Pisum sativum) plants (Paul et al., 1992; Hird et al., 1993; Zhan et al., 1996; DeBlock et al., 1997; Roque et al., 2007). This strategy was also applied to perennial grasses and trees. The TAP::BARNASE creeping bentgrass is completely pollen sterile (Luo et al., 2005). PrMC2, a pine male cone-specific gene, was successfully used to generate male sterile pine (Pinus radiata) and Eucalyptus (sp.) plants by driving a modified BARNASE gene (Zhang et al., 2012). It was recently reported that the TA29::BARNASE transgenic poplar constantly showed robust male sterility during a 4-year field trial (Elorriaga et al., 2014). Attempts were also made to abolish male and female fertility together. In Arabidopsis, BARNASE driven by the second intron of AGAMOUS resulted in ablation of stamens and carpels (Liu and Liu, 2008). Male and female sterile tobacco plants were generated by expressing BARNASE under control of both the tapetum promoter p108 and the transmitting tract promoter sp41 (Gardner et al., 2009).

In addition, manipulating genes regulating flowering time, floral meristem identify, floral organ identity, and floral organ establishment is used to abolish plant fertility. Silencing the tobacco LEAFY genes NFL1 and NFL2 resulted in plants without flowers (An et al., 2011). Tomato (Lycopersicon lycopersicum) AGAMOUS RNAi lines showed “fruit-in-fruit” phenotype, but did not produce seeds (Pan et al., 2010). Down-regulation of APETALA3 genes OsMADS16 and MtNMH7 caused stamen to carpel transformation and male sterility in rice and Medicago truncatula, respectively (Xiao et al., 2003; Roque et al., 2013). Expression of TFL1, a strong floral repressing gene, led to the non-flowering phenotype in red fescue (Festuca rubra; Jensen et al., 2004). Moreover, overexpression of miR156 inhibited flowering in switchgrass (Panicum virgatum; Fu et al., 2012). Besides generating sterile plants, plastid transformation is also an excellent approach to prevent pollen-mediated transgene flow, since plastids, including chloroplasts, are maternally inherited in most plants (Ruf et al., 2007; Wani et al., 2010).

Although male sterility has been successfully achieved via different approaches in various plant species, it cannot completely prevent transgene flow. Seed development in male sterile GM plants can be rescued by the long-distance transfer of pollen from non-GM plants. The same is also true for female sterile GM plants which disperse pollen to non-GM or male sterile GM plants. Thus, completely abolishing both male and female fertility is the only fail-safe way to prevent transgene flow (Stewart, 2007). So far, approaches to generating complete both male and female sterility are limited. Moreover, existing strategies for creating male and/or female sterility lead to loss or modifications of entire flowers or floral organs (Xiao et al., 2003; Roque et al., 2007; Pan et al., 2010; An et al., 2011), which may cause potential ecological effects on biodiversity of species associated with flowers, such as insects. In economically interesting species, for example ornamentals, altered flowers may also be undesirable. Furthermore, since the remaining toxicity of BARNASE or DTA in non-target organs due to the non-specific basal activities of employed promoters often inhibits plant survival and growth (Lannanpaa et al., 2005; Wei et al., 2007), it is difficult to obtain usable sterile plants that have normal biomass and yield. Therefore, it is imperative to generate sterility in both male and female reproductive organs without affecting plant growth or modifying flower structure.

In Arabidopsis, the SOLO DANCERS (SDS) gene, which encodes a meiosis-specific cyclin, is required for homolog interaction during meiotic prophase I in Arabidopsis (Azumi et al., 2002). With normal growth and development, the sds mutant is both male and female sterile. RNA in situ hybridization analysis showed that SDS transcripts were specifically present in microspore mother cells (male meiocytes) in anthers and megaspore mother cells (female meiocytes) in ovules (Azumi et al., 2002). Here, we report our new approach to create complete both male and female sterility in Arabidopsis and tobacco by specifically ablating microspore and megaspore mother cells using the SDS and BARNASE fusion gene. Our research provides a general and effective tool to prevent transgene flow in GM plants.

Materials and Methods

Plant Materials and Growth Condition

Arabidopsis thaliana Landsberg erecta (Ler) and tobacco (Nicotiana tabacum Petit Havana SR1) were used in this study. Plants were grown in Metro-Mix 360 soil (Sun-Gro Horticulture, Agawam, MA, USA) in a growth chamber under a 16-h light/8-h dark photoperiod regime at 22°C and 50% of humidity.

Generation of Constructs and Transgenic Plants

PCR reactions (see all primers in Supplementary Table S1) were performed using Phusion High-Fidelity DNA Polymerase (New England Biolabs, Ipswich, MA, USA). The 1.5-kb promoter of the SDS gene (upstream of the SDS coding region and the 3′ non-coding region of the SDS adjacent gene) was amplified and cloned into the pENTR/D-TOPO vector (Invitrogen, Grand Island, NY, USA) to generate pENTR-SDS. The SDS genomic fragment from the beginning of the 1.5-kb promoter region to the last exon was introduced into the pENTR/D-TOPO vector to generate pENTR-SDS::SDS. The BARSTAR gene amplified from the pABGCZ vector (Zhang et al., 2012) was introduced to the pEarleyGate303 vector at the Nsi site to generate pEarleyGate303-BARSTAR. An XhoI site was introduced between BglII and XbaI sites right after attR2 to generate pEarleyGate303-BARSTAR(XhoI). The BARNASE fragment amplified from pABGCZ was cloned into pEarleyGate303-BARSTAR(XhoI) using the XhoI and XbaI sites to generate pEarleyGate303-BARSTAR-BARNASE. Using the Gateway LR recombinase II enzyme mix (Invitrogen, Grand Island, NY, USA), SDS::GUS, SDS::BARNASE, SDS::SDS-GFP, and SDS::SDS-BARNASE binary vectors were generated between pENTR-SDS and pGBW3-GUS, pENTR-SDS and pEarleyGate303-BARSTAR-BARNASE, pENTR-SDS::SDS and pGBW4-GFP, as well as pENTR-SDS::SDS and pEarleyGate303-BARSTAR-BARNASE, respectively.

The floral dip method was used to generate Arabidopsis transgenic plants (Clough and Bent, 1998). Transformants of SDS::GUS and SDS::SDS-GFP were screened on 50 μg/mL of kanamycin and 25 μg/mL of hygromycin. Transformants of SDS::BARNASE and SDS::SDS-BARNASE were screened on 1% of Basta (PlantMedia, Lubbock, TX, USA).

Tobacco transformation was performed as described previously (Curtis et al., 1995). Briefly, leaf disks were inoculated with the Agrobacterium strain GV3101 containing the SDS::SDS-BARNASE binary vector and cultured for 1 day in the dark, followed by 2 days under light. Then, leaf disks were screened on shoot and root selection medium containing 4% of Basta. The regenerated seedlings were transferred into soil and sprayed with 4% of Basta solution one week later. The surviving plants were used for further analyses.

Pollen Staining and Anther Semi-Thin Sections

To examine pollen viability in Arabidopsis plants, Alexander pollen staining was carried out as described previously (Zhao et al., 2002). Briefly, main inflorescences were collected when 1–4 flower(s) were opened. Inflorescences were fixed for 24 h in the fixative containing methanol, 60 mL; chloroform, 30 mL; distilled water, 20 mL; picric acid, 1 g; and HgCl2, 1 g. After transferring through 70, 50, and 30% ethanol series (30 min in each change), inflorescences were finally incubated with water. Inflorescences were them transferred into the staining solution (ethanol 95%, 10 ml; malachite green, 10 mg; acid fuchsin, 50 mg; orange G, 5 mg; phenol, 5 g; glacial acetic acid, 2 ml; glycerol, 25 ml; and distilled water 50 ml) and kept at 50°C for 48 h. Individual anthers were dissected out from flowers and then mounted on the glass slides together with the staining solution for observation. Mature anthers from tobacco plants were collected and analyzed using the same method. Pollen grains were released from anthers before imaging.

Semi-thin sectioning was performed as described in our previous studies (Zhao et al., 2002; Jia et al., 2008). Briefly, dissected floral buds were fixed in 2.5% (vol/vol) glutaraldehyde in 0.1 M HEPES (N-2-Hydroxyethyl piperazine-N_-2-ethanesulfonic acid) buffer (pH 7.2) and 0.02% TritonX-100 overnight at room temperature. Samples were washed three times for 15 min each in 0.1 M HEPES buffer with 0.02% Triton X-100 and then fixed in 1% OsO4 overnight at room temperature. Samples were then dehydrated in a graded acetone series (10% increments) for 60 min each. Infiltration started with 20% Spurr’s resin and then 40, 60, and 80% Spurr’s resin every 3 h. Samples were transferred to 100% Spurr’s resin three times for 24 hours each. Samples were finally embedded in 100% Spurr’s resin and polymerized at 60°C overnight. Semi-thin (0.5 μm) sections were made using an Ultracut E ultramicrotome (Reichert–Jung) and were stained with 0.25% Toluidine Blue O.

GUS Staining Assay

Histochemical GUS staining assay was performed as previously described (Liu et al., 2010a).

Briefly, tissues were collected and fixed for 1 h in 90% acetone at –20°C. After washing tissues in washing buffer [0.1 M phosphate (pH 7.0), 10 mM EDTA, and 2 mM K3Fe(CN)6] twice for 5 min under the vacuum, the drained tissues were transferred into the GUS staining buffer [0.1 M phosphate (pH 7.0), 10 mM EDTA, 1 mM K3Fe(CN)6, 1 mM K4Fe(CN)6⋅3H2O, and 1 mg/ml X-GLUC)] and incubated overnight at 37°C. GUS-stained tissues were then fixed in a 3:1 mixture of ethanol and acetic acid. Tissues were mounted onto the glass slides for observation.

Real-Time qRT-PCR

Inflorescences of wild-type (WT) and SDS::SDS-BARNASE independent Arabidopsis transgenic plants were collected for RNA isolation using the RNeasy Plant Mini Kit (Qiagen, Valencia, CA, USA). After determining the RNA quantification by the NanoDrop 2000c (Thermo Scientific, Bannockburn, IL, USA), RNA reverse transcription was conducted using the QuantiTect Reverse Transcription Kit (Qiagen, Valencia, CA, USA). Real-time PCR (DNA Engine Opticon 2 system, Hercules, CA, USA) and data analysis were performed as previously described (Liu et al., 2010a) to evaluate expressions of A9, ATA7, DMC1, and SWI1 (Supplementary Table S1). The ACTIN2 gene was used as an internal control. Three independent biological repeats were carried out.


Pollen staining samples, GUS staining and semi-thin sections were observed and imaged with Olympus SZX7 and BX51 microscopes (Olympus, Center Valley, PA, USA), respectively. Images were obtained with an Olympus DP 70 digital camera (Olympus, Center Valley, PA, USA). For confocal microscopy analysis, anthers and ovules were dissected and mounted in water. The GFP signal was observed with a Leica TCS SP2 laser scanning confocal microscope (Leica, Buffalo Grove, IL, USA) using a 63×/1.4 water immersion objective lens. The 488-nm laser line was used to excite GFP and it also induced chlorophyll autofluorescence. The PMT gain settings was held at 650. GFP and chlorophyll autofluorescence were detected at 505–530 nm and 644–719 nm, respectively.


BARNASE Driven by the 1.5-kb Promoter of the SDS Gene Caused Various Defects in Growth and Reproduction

To create completely both male and female sterile plants without altering flower structure, we first generated the SDS::BARNASE construct using the 1.5-kb promoter of the SDS gene and a modified BARNASE (Zhang et al., 2012) to genetically ablate microspore and megaspore mother cells in Arabidopsis (Figure ​ Figure1A 1A ). Among 66 examined SDS::BARNASE transgenic plants, none of them showed the specific phenotype in sterility. Instead, compared with the wild type (Figure ​ Figure2A 2A ), SDS::BARNASE young plants were defective in vegetative growth, indicated by abnormal shape and numbers of rosette leaves (Figures 2B,C ). Different from the WT adult plant (Figure ​ Figure2D 2D ), SDS::BARNASE adult plants also exhibited various abnormal phenotypes, such as dwarf and fertile (Figure ​ Figure2E 2E ), dwarf and sterile (Figure ​ Figure2F 2F ), and even no inflorescence (Figure ​ Figure2G 2G ). The height of mature SDS::BARNASE plants was significantly reduced (Figure ​ Figure2H 2H ). Moreover, SDS::BARNASE plants produced significantly fewer rosette leaves than that of wild type (Figure ​ Figure2I 2I ). Various defects of SDS::BARNASE plants in growth and development suggest that the 1.5-kb promoter of the SDS gene failed to drive the specific expression of BARNASE in microspore and megaspore mother cells.

Schematic diagrams of constructs generated in this study. (A) SDS::BARNASE. (B) SDS::GUS. (C) SDS::SDS-GFP. (D) SDS::SDS-BARNASE. LB and RB, the T-DNA left and right border, respectively; BAR, the gene conferring resistance to the herbicide Basta; SDS::, the 1.5-kb promoter of the SDS gene; BARNASE, the bacterial ribonuclease; KAN, the kanamycin resistance gene; GUS, the gene encoding β-glucuronidase; GFP, the gene encoding green fluorescent protein; HPT, the hygromycin phosphotransferase gene; and SDS::SDS, the SDS genomic fragment containing a 1.5-kb promoter followed by a DNA fragment consisting of seven exons and six introns.

SDS::BARNASE Arabidopsis plants were abnormal in growth and development. (A–C) Compared to wild type (A), three-week old SDS::BARNASE (B,C) plants produced less rosette leaves with irregular shape. Bars = 0.5 cm. (D–G) Six-week old wild-type (WT, D) and SDS::BARNASE plants showing fertile but dwarf (E), dwarf and sterile (F), and no inflorescence (G) phenotypes. Bars = 1 cm. (H) Six-week old SDS::BARNASE plants were significantly shorter than the wild type. (I) The rosette leaf number of SDS::BARNASE adult plants was significantly reduced. “n” indicates the number of examined plants. Stars indicate significant difference (P < 0.01).

The 1.5 kb Upstream Region of the SDS Gene did not Confer its Meiocyte-Specific Expression

Genetic ablation relies on the specificity of employed promoters. To examine why BARNASE under the control of the 1.5-kb SDS promoter did not achieve specific ablation effects on microspore and megaspore mother cells, we generated SDS::GUS plants to test the transcriptional activity of the 1.5-kb promoter (Figure ​ Figure1B 1B ). Among 25 examined SDS::GUS transgenic plants, GUS signals were detected in cotyledons, true leaves, and shoot apical meristem of young seedlings (Figure ​ Figure3A 3A ), as well as in carpels and stigmas of young buds (Figures 3B–D ). Thus, our results suggest that the 1.5-kb promoter of the SDS gene was not sufficient for conferring its meiocyte-specific expression, which resulted in abnormal plant growth and development when it drove the expression of BARNASE.

The entire SDS gene but not the SDS 1.5-kb promoter confers the SDS meiocyte-specific expression. (A–D) GUS staining of SDS::GUS plants showing GUS signals in cotyledons, true leaves, and shoot apical meristem of a young seedling (A), as well as in carpels and stigmas of young buds (B–D). (E) A confocal image from an SDS::SDS-GFP stage-5 anther showing the GFP signal (green color) only in microspore mother cells (arrows). Red and yellow colors showing merged autofluorescences. (F) A confocal image from an SDS::SDS-GFP stage 2-IV ovule showing the GFP signal only in the megaspore mother cell (arrow). Bars = 0.1 cm (A,B), 0.5 mm (C,D), 50 μm (E), and 10 μm (F).

The Entire SDS Gene Led to the Meiocyte-Specific Expression of the SDS Protein

The possible existence of regulatory elements in SDS introns may contribute to the SDS meiocyte-specific expression. To test, how to achieve the specific expression of SDS in microspore and megaspore mother cells, we generated SDS::SDS-GFP constructs by fusing the SDS genomic fragment, containing the 1.5-kb promoter, seven exons and six introns, with the GFP gene (Figure ​ Figure1C 1C ). In examined 18 SDS::SDS-GFP transgenic plants, we did not detect the GFP signal during the seedling stage and later in the vegetative growth stage. We, however, observed GFP signals only in microspore mother cells in anthers (Figure ​ Figure3E 3E ) and megaspore mother cell in ovule during the reproductive stage (Figure ​ Figure3F 3F ). Therefore, our results indicate that the entire SDS gene led to the meiocyte-specific expression of the SDS protein.

SDS::SDS-BARNASE Caused Complete Both Male and Female Sterility but did not Affect Growth or Development in Arabidopsis

To generate complete both male and female sterility by specifically ablating microspore and megaspore mother cells, we made the SDS::SDS-BARNASE construct by fusing the SDS entire gene with the BARNASE gene (Figure ​ Figure1D 1D ). We performed three transformations, resulting in 97, 80, and 126 SDS::SDS-BARNASE transgenic plants, respectively. All independent transgenic plants were sterile. We first evaluated the effects of SDS::SDS-BARNASE on growth and development. SDS::SDS-BARNASE transgenic plants produced rosette leaves with the same number, size, and shape as that of WT plants (Figures 4A,B ). No morphological changes were observed in SDS::SDS-BARNASE inflorescences and flowers (Figures 4C,D ). Moreover, mature SDS::SDS-BARNASE plants had a similar height to the wild type (Figures 4E–G ). The flowering time of SDS::SDS-BARNASE plants was not affected either, because the same rosette leaf numbers as the wild type were produced when flowering (Figure ​ Figure4H 4H ).

SDS::SDS-BARNASE Arabidopsis plants showed normal growth and development. (A,B) Three-week old WT (A) and SDS::SDS-BARNASE (B) plants. Bars = 0.5 cm. (C,D) Five-week old WT (C) and SDS::SDS-BARNASE (D) inflorescences. Bars = 0.5 cm. (E,F) Six-week old WT (E) and SDS::SDS-BARNASE (F) plants. Bars = 1 cm. (G) No difference in average height between six-week old WT and SDS::SDS-BARNASE plants. (H) Similar rosette leaf numbers indicating no difference in flowering time between WT and SDS::SDS-BARNASE plants. “n” in (G,H) indicates the number of examined plants.

To further investigate sterility of SDS::SDS-BARNASE transgenic plants, we analyzed both male and female fertilities. Compared with the wild type (Figures 5A,H ), SDS::SDS-BARNASE plants produced short siliques (Figures 5B,I ). Except short filaments, SDS::SDS-BARNASE plants formed flowers that were the same as the wild type, indicated by four sepals, four petals, six stamens, and two carpels (Figures 5D,E ). In the WT flower, pollen grains were released from anthers that reached the stigma (Figure ​ Figure5D 5D ), whereas in the SDS::SDS-BARNASE flower, no pollen grains were observed on the anther surface and anthers did not reach the stigma (Figure ​ Figure5E 5E ). Furthermore, different from the WT anther (Figure ​ Figure5F 5F ), the SDS::SDS-BARNASE anther did not produce pollen grains (Figure ​ Figure5G 5G ), indicating that SDS::SDS-BARNASE plants were male sterile. Because pollination using the WT pollen did not rescue the fertility (Figures 5C,J ), SDS::SDS-BARNASE plants were female sterile too. Thus, using SDS::SDS-BARNASE, we efficiently created completely both male and female sterile Arabidopsis plants that had normal vegetative and reproductive growth and development, including the formation of all flower organs.

SDS::SDS-BARNASE Arabidopsis plants were completely both male and female sterile. (A–C) Primary branches showing normal siliques in wild type (A) and short siliques indicating no developing seeds in SDS::SDS-BARNASE plants without (B) and with (C) pollination. Bars = 1 cm. (D,E) Side view of mature flowers (One sepal was removed, respectively) showing the SDS::SDS-BARNASE flower (E) is similar to the wild type (D) except short filaments. Pollen grains released from WT anthers (D, inset), while no pollen grains from SDS::SDS-BARNASE anthers (E, inset). Bars = 0.5 mm. (F,G) Pollen staining showing the WT anther full of viable pollen grains (F), but no pollen grains from the SDS::SDS-BARNASE anther (G). Bars = 30 μm. (H–J) Dissected individual siliques from primary inflorescences (positions 1–9) were long in wild type (H), but short in SDS::SDS-BARNASE plants (I, without pollination; J, pollinated with WT pollen). Bars = 1 cm.

SDS::SDS-BARNASE Inhibited Both Male and Female Gamete Formation

To further understand ablation effects on microspore and megaspore mother cells, we did semi-thin sectioning of anthers and whole-mount squashes of ovules. At stage 5 (Sanders et al., 1999; Zhao et al., 2002), when compared with the WT anther (Figure ​ Figure6A 6A ), the SDS::SDS-BARNASE anther showed vacuolated microsporocytes (microspore mother cells) and tapetal cells (Figure ​ Figure6D 6D ), indicating the degeneration of both cells. At stage 7 in the WT anther, successful male meiosis resulted in the formation of tetrads (Figure ​ Figure6B 6B ), whereas in the SDS::SDS-BARNASE anther, tetrads, and tapetal cells were collapsed (Figure ​ Figure6E 6E ). At stage 9, the WT anther contains developing pollen grains (Figure ​ Figure6C 6C ), but the SDS::SDS-BARNASE anther lacked developing microspores (Figure ​ Figure6F 6F ).

Formation of male gametes was arrested in SDS::SDS-BARNASE Arabidopsis plants. (A–C) WT anthers showing microsporocytes (microspore mother cells) and surrounding tapetal cells at stage 5 (A), tetrads and tapetal cells at stage 7 (B), and developing pollen grains at stage 9 (C). (D–F) SDS::SDS-BARNASE anthers showing degenerating microsporocytes and precociously vacuolated tapetal cells at stage 5 (D), dead microsporocytes and tapetal cells at stage 7 (E), and a nearly empty anther lobe at stage 9 (only one dead pollen, F). M, microsporocytes (microspore mother cells); DP, developing pollen; T, tapetal cell; and Tds, tetrads.

In embryo sacs of WT ovules, two nuclei at stage FG3 (Pagnussat et al., 2009) (Figure ​ Figure7A 7A ) and four nuclei at stage FG4 (Figure ​ Figure7B 7B ) were observed; however, in SDS::SDS-BARNASE embryo sacs, only a single nucleus was produced (Figures 7D,E ). At stage FG6, the WT embryo sac showed the central cell, the egg cell, and synergid cells (Figure ​ Figure7C 7C ), but the SDS::SDS-BARNASE embryo sac is empty (Figure ​ Figure7F 7F ).

Formation of female gamete was arrested in SDS::SDS-BARNASE Arabidopsis plants. (A–C) WT ovules showing two separated nuclei (arrows) at the FG3 stage (A), four nuclei (arrows) at the FG4 stage (B), and the central cell, the egg cell, and synergid cells in a mature embryo sac (white dots outlined) at the FG6 stage (C). (D–F) SDS::SDS-BARNASE ovules showing one small nucleus (arrow) at both FG3 (D) and FG4 (E) stages and a small empty embryo sac (white dots outlined) at the FG6 stage (F). Bars = 10 μm. cc, central cell; ec, egg cell; and syn, synergid cells.

Furthermore, our results showed that expressions of tapetal cell marker genes A9 and ATA7 as well as microspore and megaspore mother cell marker genes DMC1 and SWI1 were significantly decreased in SDS::SDS-BARNASE buds in comparison to the wild type (Figure ​ Figure8 8 ). In summary, the specific expression of the SDS-BARNASE toxic fusion protein in microspore and megaspore mother cells efficiently impaired the production of both male and female gametes, which led to absolute both male and female sterility, but did not affect flower organ formation or plant growth and development.

Expressions of tapetal cell as well as microspore and megaspore mother cell marker genes. Real-time qRT-PCR showing decreased expressions of tapetal cell marker genes A9 and ATA7 as well as microspore and megaspore mother cell marker genes DMC1 and SWI1. Stars indicate significant difference (P < 0.01).

SDS::SDS-BARNASE Caused Complete Both Male and Female Sterility but did not Affect Growth or Development in Tobacco

To test whether SDS::SDS-BARNASE can provide a general tool to create both male and female sterile plants, we transformed it into tobacco and generated SDS::SDS-BARNASE tobacco transgenic plants by tissue culture. Among 14 examined SDS::SDS-BARNASE tobacco transgenic lines, leaf shape and size (Figures 9A–C ), as well as the plant height (Figures 9B–D ) were the same as that of WT plants. In addition, the SDS::SDS-BARNASE tobacco flower had the same size, color, and structure as that of wild type (Figures 9E,F ). Therefore, SDS::SDS-BARNASE did not affect growth or development in tobacco plants.

SDS::SDS-BARNASE tobacco plants showed normal growth and development. (A) Forty-day old tobacco WT and SDS::SDS-BARNASE plants. Bar = 5 cm. (B,C) Sixty-day old WT (B) and SDS::SDS-BARNASE (C) plants. Bars = 10 cm. (D) No difference in average height between WT and SDS::SDS-BARNASE adult plants. (E,F) Flower size, color, and structure remained the same in WT and SDS::SDS-BARNASE plants. Bars = 1 cm.

Ten examined SDS::SDS-BARNASE tobacco transgenic lines were completely sterile. WT tobacco plants produced large fruits and per fruit averagely contained 0.11 g of seeds (Figures 10A,D ). Conversely, SDS::SDS-BARNASE plants produced small fruits and no seeds were found when self-pollenated (Figures 10B,D , e.g., plants #1, 3, 5, and 7). Further pollen viability analysis showed that WT tobacco anthers produced viable pollen, indicated by red color (Figure 10E ), whereas anthers from sterile tobacco plants either lacked pollen grains (Figure 10F ) or formed dead pollen grains (Figure 10G ). The four non-absolutely sterile lines produced a few seeds (Figure 10D , e.g., plants #2, and 14) and only some functional pollen grains were found in anthers of those lines (Figure 10H , e.g., plant #2). Our results suggest that SDS::SDS-BARNASE impaired male fertility in tobacco.

SDS::SDS-BARNASE tobacco plants were completely both male and female sterile. (A–C) Large fruits from the WT plant (A) and small fruits from SDS::SDS-BARNASE plants without (B) and with (C) manual pollination with WT pollen grains. Bars = 1 cm. (D) The weight of seeds per self-pollinated and manually pollinated fruit (n = 5), respectively. Numbers indicate examined independent transgenic lines. (E) WT viable pollen grains in red color. (F–H) no (F), all dead (G) and a few viable (H) pollen grains in SDS::SDS-BARNASE plants. Numbers indicate examined independent transgenic lines. Bars = 100 μm.

We then examined the female fertility in sterile tobacco transgenic plants. The fertility of manually male-sterilized WT flowers could be rescued by cross-pollination with WT pollen (Figure 10D ), but following cross-pollination with WT pollen, the fruit size of SDS::SDS-BARNASE sterile tobacco plants did not change (Figure 10C ) and no seeds were produced (Figure 10D , e.g., plants #1, 3, and 5). Thus, SDS::SDS-BARNASE tobacco transgenic plants were also female sterile. Manual pollination partially rescued the fertility of line #7, indicating that the line #7 is a completely male but partially female sterile plant, while lines #2 and 14 were nearly male and female sterile plants (Figure 10D ). Collectively, a majority of SDS::SDS-BARNASE tobacco transgenic plants were completely male and female sterile, suggesting that SDS::SDS-BARNASE is functionally conserved, which can be used to create both male and female sterility in general.


SOLO DANCERS, a unique type of (SDS-type) meiosis-specific cyclin, is conserved in flowering plants (Wang et al., 2004; Zhang et al., 2014; Wu et al., 2015). In our studies, the 1.5-kb SDS promoter did not achieve the specific expression of SDS in either microspore or megaspore mother cells. Conversely, the entire SDS gene genomic fragment did. The intron-dependent spatial expression has been revealed in different genes from various species, including SUS3 and SUS4 in potato (Fu et al., 1995a,b), OsTubA1 in rice (Jeon et al., 2000; Giani et al., 2009), PhADF1 in petunia (Jeong et al., 2007), as well as AGAMOUS, ACT1 and PRF in Arabidopsis (An et al., 1996; Sieburth and Meyerowitz, 1997; Jeong et al., 2006). Therefore, it is possible that regulatory motifs in SDS introns contribute to its specific spatial and temporal expression. Future studies should be focused on dissecting the functions of unknown regulatory motifs and then making a synthetic promoter that confers the strong and specific expression of SDS in microspore and megaspore mother cells. We found a few non-completely sterile tobacco transgenic plants, suggesting that the Arabidopsis SDS gene did not work efficiently in tobacco. In order to achieve accurate and efficient ablation effects, it would be more practical to use SDS orthologous genes of target species or the synthetic promoter to drive BARNASE.

The existing methods for creating male sterile only GM plants are not able to completely prevent transgene flow, because pollen from non-GM plants can rescue seed development. In addition, current sterilization technologies either suppress the production of entire flower or some floral organs, which may cause potential ecological effects besides transgene flow. Furthermore, BARNASE and DTA are very toxic. Many “specific” promoters still have basal activities in other organs; therefore, the low expression of BARNASE/DTA in non-target organs often reduces plant survival rate and inhibits plant growth. Microspore and megaspore mother cells are two small groups of male and female reproductive cells, which are differentiated after all floral organs are established. Ablating microspore and megaspore mother cells only leads to elimination of male and female gametes, but it does not affect other somatic cell differentiation and flower development. In this study, we specifically ablated microspore and megaspore mother cells using the SDS and BARNASE fusion protein. Thus, our research developed an efficient strategy to successfully create completely both male and female sterile plants; however, the plant growth and development, including the formation of all flower organs, were not affected.

Genetically modified crops have been widely cultivated in many countries due to their improved agronomic traits; however, the adoption of GM trees (e.g., poplar, eucalypts, and pines) and perennial grasses (e.g., miscanthus and switchgrass) is limited, because those plants are long-lived, weakly domesticated, and important to ecosystems. Various studies has been done to increase cold tolerance and biomass, modify lignin and cellulose biosynthesis, or alter growth and flowering of Eucalyptus (Girijashankar, 2011; Hinchee et al., 2011; Klocko et al., 2015), aspen (Etchells et al., 2015), poplar (Wilkerson et al., 2014), and switchgrass (Fu et al., 2011; Shen et al., 2013; Baxter et al., 2014; Poovaiah et al., 2015). Our research developed a general and effective approach to create completely both male and female sterile plants by specifically ablating microspore and megaspore mother cells, which provides a solution for overcoming regulatory hurdles to field research and ultimately commercial uses of GM plants.

Besides the transgene containment, our method can be applied for modifying invasive and ornamental plants. Male and female sterilized invasive plants generated by our method can be planted for multiple purposes, such as forestry and horticulture. The main valuable trait for many ornamental trees, such as cherries and plums, is the beauty of flowers. Fruits often make the garden messy. Moreover, nutrient competition from fruit setting affects flower organ differentiation, and consequently reduces flower numbers in the coming year. Our new method serves as an excellent tool to engineer ornamental trees that still produce attractive intact flowers without fruit setting, which, therefore, maintains their ornamental value.

Author Contributions

DZ conceived and designed the experiments. JH, AS, and TZ performed experiments. All authors analyzed data. DZ and JH wrote the article.

Conflict of Interest Statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Dazhong Zhao has filed a patent application for the sterilization technique reported in this article.


We thank H.A. Owen for critical reading of this manuscript, C. Zhang for the pABGCZ vector, T. Nakagawa for Gateway binary vectors, C. Pikaard for the pEARLEY303 vector, and J. Gonnering and Tom Schuck for plant care.


Funding. This work was supported in part by the National Science Foundation (NSF IOS-0721192 and IOS-1322796), the Research Growth Initiative (RGI) at the University of Wisconsin-Milwaukee, and the UW-Madison/UW-Milwaukee Intercampus Research Incentive Grants Program to DZ. DZ also gratefully acknowledges the support of the Shaw Scientist Award from the Greater Milwaukee Foundation.

How to Make Feminized Seeds at Home

First of all, what are “feminized” marijuana seeds? Although male and female plants look the same when young, only female cannabis plants make buds. Male cannabis plants grow non-potent pollen sacs instead. Male plants also lower yields and cause seedy buds if they’re left among your female plants for too long. Because male plants don’t make buds and their presence reduces the quality of buds, most growers toss male plants the moment they see pollen sacs forming. However, since about half of regular cannabis seeds end up being male, that means you end up tossing half your plants once they start flowering.

Feminized seeds come from two female plants being bred together, causing all offspring to be female (which means every plant makes buds)

While regular seeds make female plants about half the time, feminized seeds only create female plants. That means you won’t need to toss half the plants once they start flowering. But how are feminized seeds made and can you make them at home?

Feminized seeds are created by breeding two female plants together. Because there are no male parents, all the resulting seeds end up being bud-bearing female plants. With feminized seeds, you can count on every plant to produce buds. Learn more about male vs female plants and feminized cannabis seeds.

Feminized seeds are available from seed banks for nearly every popular or famous strain. Breeders understand that a lot of people just want to grow plants for buds, and don’t care about making a robust breeding program (which is one of the main reasons growers want male plants).

You can’t see the difference between male and female plants until they start flowering (unless you do a genetic test). Feminized seeds ensure all plants are female so you don’t need to worry about it. You know that every plant is female from germination.

So how do seed banks feminize their seeds? How do you breed two female plants together?

The main idea is to force a female plant to produce pollen sacs like a male plant. These flowers (growing on a female plant) create pollen, which can be harvested and used to pollinate another female cannabis plant. The resulting seeds will all end up being female. Can feminizing seeds cause hermaphrodite plants?

Growers can force a female plant to make pollen sacs, and the “feminized” pollen produced can be used to fertilize another female plant. The resulting seeds will only produce female plants.

How do you force a female plant to make pollen?

There are two main ways to make feminized pollen:

  1. Induce feminized pollen chemically (Recommended) – This is the professional way to feminize seeds and is how reputable seed banks and breeders create feminized seeds to sell to the public. Substances that interact with plant processes such as colloidal silver or gibberellic acid are applied to bud sites of a female plant when they start flowering. Bud sites are drenched daily for the first 3-4 weeks after the switch to 12/12. This causes a female plant to produce pollen sacs which release feminized pollen when they open up. This pollen is used on another female plant to produce feminized seeds. This article will give you step-by-step instructions on how to feminize cannabis seeds using this method.
  2. Rhodelization (Not Recommended!) – In some cases, a female cannabis plant may naturally start making male pollen sacs or bananas, which can self-pollinate the plant. This happens if the plant is stressed, or if the plant is not harvested in time and buds start to die of old age. The plant is basically doing everything it can to make seeds and save the next generation. This method is “natural” and these seeds end up being mostly female. The problem with this method is you’re selecting for plants that naturally turn into hermies (grow both male and female sex organs) without any chemical induction. This means the resulting seeds are much more likely to turn hermie in natural conditions, too. That’s a problem if you don’t want seedy buds every time you harvest. For that reason, it’s highly recommended you don’t feminize seeds this way. It’s also a good idea to toss any and all seeds that are the result of natural herming (for example seeds you find in your buds even though you didn’t grow any male plants).

Read this article for more in-depth discussion about the pros and cons of each method, and how to avoid hermaphrodite plants when producing your own feminized seeds.

Overview: How to Make Feminized Seeds

This is a quick overview of the process, and then I’ll give the full details and steps below.

1.) Buy or Make Colloidal Silver – The article below will teach you how to make colloidal silver at home, as well as show you where to buy it if you don’t want to make it (it’s actually pretty cheap). It’s basically a solution of silver suspended in water and is available online and in health stores as a dietary supplement.

What about gibberellic acid? From what I understand it can be used exactly the same way as colloidal silver to induce female plants to produce pollen, but I don’t know the recipe for an effective gibberellic acid solution. On the other hand, I know that this exact colloidal silver method works for making feminized seeds. That’s why I’ve only included instructions for colloidal silver. If you’ve used gibberellic acid to make feminized seeds; we’d love to hear from you.

2.) Spray the bud sites of your known female plant daily during the first 3-4 weeks of the flowering stage (until pollen sacs form and start splitting open) – After switching to a 12/12 light schedule to initiate the flowering stage, choose bud sites on your known female plant, and spray/drench them daily with colloidal silver (or gibberellic acid). As the treated flowers develop, they will form into male pollen sacs instead of regular buds. Untreated bud sites on the plant will form into female buds as usual; however, these buds are unsafe to smoke unless you’ve been very careful to make sure they didn’t come into contact with colloidal silver or gibberellic acid during the feminization process as these are unsafe for smoking.

3.) Harvest “Feminized” Pollen – When pollen sacs are ready to be harvested they swell like a balloon and start to open up. Don’t harvest early! Keep spraying the bud sites daily until pollen sacs open or you might end up with empty pollen sacs. When the pollen sacs are ready, the leaf section protecting the pollen will start to crack. At this point, it’s time to collect the feminized pollen. One of the easiest ways to do this is to collect the pollen sacs directly and let them dry for a week. At that point, they can be placed in a bag and shaken to collect all the pollen.

4.) Pollinate Another Female Plant – At this point, take the feminized pollen you’ve collected and use it to pollinate a female plant that has been flowering for about 2-3 weeks (full detailed instructions with a video on how to do this below). Although it’s possible to pollinate the same plant as the original, it’s not recommended in part because the timing doesn’t match up (pollinating buds late in the flowering stage doesn’t produce many seeds). It’s best to pollinate a different female plant that you started budding a few weeks after the original. This increases the number of seeds produced as well as gives the new female plant enough time to develop them to maturity. It also increases genetic diversity compared to self-pollination.

5.) Wait ~6 Weeks After Pollination to Harvest Seeds – After about 6 weeks from pollination, the calyxes on the buds of your female plant will be swollen and fat. You know it’s time to harvest your seeds when they start bursting out. At this point, it’s time to congratulate yourself because you’ve got feminized seeds!

Now that you’ve gotten the overview, here’s the feminization process with detailed step-by-step instructions…

Step-By-Step Instructions (with pics!)

1.) Buy or Make Colloidal Silver (or Gibberellic Acid)

Where to get Colloidal Silver (your options):

    Buy Ready-To-Use Colloidal Silver: Colloidal silver is sometimes used as a dietary supplement, so it’s relatively easy to find (never take it without talking to a doctor first though!). If you’re purchasing colloidal silver, try to find a solution that has at least 30 PPM (parts per million) of silver or higher.
  1. Buy A Colloidal Silver Generator Kit: If you plan on feminizing a lot of seeds, you may want to invest in a generator kit so you can easily make your own endless supply of colloidal silver. This is cheaper in the long run compared to buying it ready-to-use.
  2. Make Your Own: You can make your own colloidal silver generator at home. The following diagram illustrates what you need to do.

Note: You can purchase gibberellic acid online (a gibberellic acid solution can be used the same way as colloidal silver for feminizing seeds). However, I do not have experience with the gibberellic acid method and don’t know the best way to prepare the solution.

2.) Spray the bud sites of your known female plant daily during the first 3-4 weeks of the flowering stage (until pollen sacs form and start splitting open)

Wait until your plant is 5-6 weeks old before initiating the flowering stage. Some young plants seem to have trouble (and take much longer) to go through the feminization process, and their pollen may not be as fertile, so start with a more mature plant.

When plants are ready, change to a 12/12 light schedule to initiate flower formation and put cannabis plants in the flowering stage. Note: If you’re feminizing an auto-flowering plant, start spraying daily when the plant is about 20 days old from seed. This is when most auto-flowering cannabis strains start making flowers.

As soon as you change the light schedule (and maybe even a day or two before) start spraying your plants thoroughly with colloidal silver at every bud site you want to form into pollen sacs.

Spray bud sites thoroughly, drenching them with colloidal silver every single day. Bud sites are located wherever leaves meet stems.

The above pic shows you where pollen sacs form on the plant (the same places female buds form).

Important! Keep spraying daily until pollen sacs open up. Don’t stop spraying early, even if pollen sacs appear to be already formed, otherwise they may not produce much pollen

A spray bottle / mister is really helpful for spraying bud sites evenly and thoroughly

You can choose to treat a single bud site or all the bud sites on the plant. Any untreated bud sites will develop into female buds as usual. If you want to smoke these buds, it’s incredibly important to avoid letting them come into contact with colloidal silver because silver is not safe to smoke. (Don’t worry, feminized seeds don’t contain any silver). I highly recommend letting the whole plant be your test subject so you don’t have to worry about that

3.) Harvest Your Feminized Pollen

When pollen sacs are starting to crack and look like they’re about to open up (or if you can see one has already opened) then your pollen is ready for harvest!

When pollen sacs are cracking and opening up, you’re ready to harvest your pollen!

Pollen spilling onto a nearby leaf

One way to harvest your pollen is to gently and carefully remove all the pollen sacs. Let them dry in open air for a week, and then put them in a resealable bag. If you shake the bag the pollen should easily spill out. You may need to cut a few open yourself.

How to Store Feminized Pollen: Moisture is your main enemy when storing pollen. It can help to double the mass of the pollen collected by adding regular cooking flour. This absorbs moisture during storage and as an added bonus, it increases the volume to make application easier when you get to pollinating. If pollen is totally dry and you triple-bag the pollen-flour mixture and stick it in the freezer (with a good nametag so you know where the pollen came from), your pollen can be stored for a year or longer. You can add a few silica packs (which suck out any remaining moisture) in the bag to make extra sure that the pollen stays totally dry.

4.) Pollinate Another Female Plant

When your chosen mother is 2-3 weeks into the flowering stage, take a small paintbrush or powder brush and ‘paint’ your feminized pollen on the developing bud sites you want to pollinate. Bud sites (for both male and female plants) are located wherever you can see leaves meet a stem.

Buds are ready to get pollinated when they look like little bunches of white hairs

Only the buds that come in contact with pollen will grow seeds. You can choose to pollinate all of your buds or just a few on the plant.

Here’s a video by Ed Rosenthal on Youtube showing you how to pollinate buds with pollen. Make sure to touch all the female pistils/hairs with your pollen.

5.) Wait About 6 Weeks Then Harvest Seeds

It usually takes about 6 weeks for your feminized seeds to fully develop. Some plants are literally dying right as the seeds become ready, so to get the most viable seeds, you need to try to keep it alive until the seeds actually start dropping. The seeds can be used right away, or stored in a cool, dry place for a few years. Don’t forget to label them with the date.

This seed is about to burst out of its calyx

This is what it looks like when the seed is exposed

Picture Journal of Making Feminized Pollen with Colloidal Silver

This grower initiated the feminization process on a seedling that was only a few weeks old. As a result, the plant wasn’t able to get big enough to produce many pollen sacs. You will get even better results if you start with a plant that is at least 5 weeks old

October 18 – Plant right before the switch to 12/12

October 27 – After being drenched with colloidal silver daily for a little over a week

October 30 – Pollen sacs are forming

November 15 – Pollen sacs appear to be almost fully formed and are swelling in size, but haven’t opened up yet. Don’t stop spraying colloidal silver or you may end up with empty sacs!

November 27 – Pollen sacs are opening up! Collect the pollen before they’re all open!

FAQs – Frequently Asked Questions

How can I identify plant gender before the plant actually starts flowering?

There are a few ways to identify plant gender before the plant actually starts flowering, and each is helpful in different situations.

  1. Use feminized seeds – All your plants will be female if you start with pre-made feminized seeds.
  2. Start with a clone – A clone is an exact copy of another plant. If the “mother” of the clone is a female plant, it means the clone is also female
  3. Look at preflowers (identify plants when they’re 3-6 weeks from seed) – If you know where to look, cannabis plants will actually reveal their gender in the vegetative stage when they’re just 3-6 weeks from seeds. Male plants usually show their gender by 3-6 weeks and female plants usually show their gender around week 4-8 from seed. Learn how to determine the sex in the veg stage by looking at preflowers.
  4. Test the leaves of your seedling in a lab – It’s possible to send in a leaf from a young cannabis plant to a specialized testing company, and they will be able to determine the gender as soon as 3 weeks from seed! Although I haven’t used any of these companies and can’t recommend any in particular, here’s a link to one example just so you can see what I’m talking about. From talking to other growers who use this method, it appears to be accurate.
  5. Take a clone and force it to start flowering – if you take a clone from a vegetative plant, you can force that clone to start flowering and reveal its gender. You’ll know the sex of the “parent” plant by the sex expressed by the clone. This is what I do to determine the sex. I cut off a piece of the plant, stick it in a glass of water (don’t forget to label it with the strain), and give it a 12/12 light schedule until the little piece starts forming either pollen sacs or buds.

The easiest way to identify sex with unknown seedlings? Cut off a small branch of the plant, stick it in a cup of water with a label, and keep it in a sunny window on a 12/12 light schedule until buds or pollen sacs start forming. As long as the plant is getting bright light in the day and long dark nights, it will reveal its sex in just a few weeks (you don’t even need to wait for roots to form).

Can I Make a Breeding Program Using Just Female Plants and Feminized Seeds?

Yes, it’s possible to use just female plants and feminized seeds for further breeding, with one major caveat.

Without careful and thorough testing, it may be possible to accidentally select cannabis plants that tend to herm (make male flowers or pollen) and cause seedy buds when you don’t want them to.

For each possible “mother,” clones should be grown in several different environments and tested thoroughly to make sure that the mother plant does not have any tendency to make pollen naturally in normal or stressful conditions. It’s okay if plants grow pollen sacs if induced chemically since that is very unlikely to happen in someone’s garden on accident, but you don’t want plants that will start growing male flowers on their own without chemical induction. Thorough testing of plant hardiness is always important when breeding, but it may be especially important when breeding feminized seeds together.

Are there other reasons I should avoid breeding seeds without males?

The most common reason growers say you shouldn’t do this is because it’s “unnatural” or doesn’t “seem right.” Some growers say you need male plants for genetic diversity. I’ve also heard growers say that the resulting plants will be weaker, sterile, and less potent. Someone once even told me that resulting plants “will be worse in every way.”

As of yet, I haven’t seen any of these claims backed by actual personal experience, or any real-life examples showing why using feminized seeds is not a viable way to breed new strains.

To those who say this type of reproduction just doesn’t seem right, the evolutionary strategy of plants using only female and hermaphrodite plants to breed is actually pretty common and is known as gynodioecy. One example of a plant that only reproduces this way is a flower found in Canada and the US called Lobelia siphilitica, also known as the Great Lobelia. Obviously, this reproduction method isn’t exactly the same as artificial feminization since the pollen production is caused naturally instead of induced chemically, but examples of gynodioecy show that a female flower-based breeding population can exist in the wild even when no plants are purely male.

The Great Lobelia naturally reproduces using only female and hermaphrodite plants. This is similar to the cannabis feminization process because it results in a population of plants that all primarily grow female flowers, with no pure male plants

When it comes to genetic diversity, the ability to cross out to thousands of different cannabis strains allows you to dramatically increase the gene pool without using male plants. So those are my answers to the common objections of a feminized seed-based breeding program, however I am just a theory-crafter when it comes to this topic. It certainly seems possible that a feminized-only breeding program could run into unforeseen problems down the road, but as far as I know there isn’t any evidence of that so far.

Although I have a few anecdotes from growers who have used only feminized seeds for a few generations, it would be much better to share information from someone who has conducted plenty of testing over several generations. We’d love to hear from you if you have bred more than a few generations using only feminized seeds and want to share your experience.

What are the positive aspects of breeding two feminized seeds together?

Besides not having to worry about male plants in the next generation, the main advantage of doing this is you have a much better idea of what you’re working with when it comes to producing the type of buds you’re looking for. When you’re growing a male plant, it has several genes it will pass to its offspring that has to do with how buds develop, but since it’s a male plant those genes aren’t expressed and it’s hard to figure out what they are.

Historically, the way to learn more about the “hidden genes” contained in a male plant is to breed it to several well-known female plants and see how the offspring compare to each other. The genes that don’t come from the known mother plant are assumed to come from the male. Another way of going about this is to take several clones of the same well-known female plant and breed them with many different male plants to see which ones produce the best offspring.

After testing with several pairings, you start to get an idea of the hidden genes a male plant has to offer to its female offspring. This time-consuming process of documenting and identifying good male plants is why proven stud male plants are one of the most valuable and closely guarded types of clones available today.

But the process of finding the right “father” is a little different when you start with two female plants. In this case, you already know quite a bit about the genes of both parents because you can just look at and test the buds of both plants directly. This allows you to pinpoint desirable genes with less guessing and much less time spent growing out and cataloging plants.

Breeding two female plants together offer hints about what kind of buds their genes will produce

Why even have male plants then?

In nature, male plants are effective at increasing genetic diversity by ensuring cross-pollination. With only purely male and purely female plants, every resulting seed will have two different parents.

Another big advantage in the wild of having separate female and male plants is sexual specialization. In other words, plants are able to evolve male and female traits separately, so each type of flower can become more specialized at its unique “job.”

However, this isn’t the only successful breeding strategy for plants. In fact, only 6-7% of plants have completely separate male and female plants like cannabis plants do (known as dioecious plants). Most plants grow some mix of male and female flowers on each plant, with different combinations offering different evolutionary benefits.

You might enjoy this scientific article if you want to learn more about the evolution of sex determination in plants and animals: Sex Determination: Why So Many Ways of Doing It?

And although most cannabis strains (at least the good ones) display either purely male or purely female flowers, there are some wild populations (and some strains of hemp) that regularly produce plants with male and female parts on the same plant.

When it comes to artificial selection for breeding new strains, the grower is in charge of cross-pollination, so there’s no need for the plant to specialize in male parts. Pretty much the only thing most growers care about is how female flowers develop. So (unlike in nature) growers have the freedom to choose plants that improve female buds without even having to consider how it might affect male plants.

Only a small percentage of plant species produce male plants like cannabis.

Can feminizing seeds result in hermaphrodite plants?

The answer is yes. If you do it the wrong way then feminization can lead to plants with an increased chance of herming. However, with a well-tested and well-bred feminization program, one of the main goals is to breed out any plants with hermaphroditic tendencies that show up under normal conditions. When you buy feminized seeds from trustworthy breeders, you can count on the fact that every plant will end up growing only female flowers and that’s it.

This is a relatively big topic with a lot of opposing opinions so I wrote a whole article about it.

Feminizing seeds the wrong way can result in hermaphrodite plants.

Can I pollinate the same plant I collected the pollen from?

Yes, it’s possible. However, it’s not really recommended because, for one, the timing doesn’t match up. By the time your pollen is ready to use, your original plant will already be several weeks past the optimum pollination point. It’s best to pollinate a female plant that has only been flowering about 2-3 weeks, but pollen sacs need more time before pollen is ready to use. It’s also possible to run into unwanted side effects from self-pollination/in-breeding.

One thing to keep in mind is even if you pollinate a plant to itself, the resulting seeds are likely not going to be exact copies of the original (unless the original plant is extremely inbred). The resulting seeds include not just the mother’s expressed genes but also her hidden ones.