Protocol measuring horizontal gene transfer from algae to non-photosynthetic organisms

Graphical abstract


Procedure
Protocol for measuring horizontal gene transfer 1 Prepare special plasmid containing designer transgene(s) of interest including an antibiotic selectable marker to create genetically engineered (GE) cyanobacteria to use as a transgene(s) donor for the horizontal gene transfer assay [1]. 2 Create special plasmid-based and/or genomic-based GE cyanobacteria (such as T. elongatus BP1) through genetic transformation using electroporation followed by transformant colony selection and verification with PCR. 3 Photoautotrophically grow GE T. elongatus BP1 in BG-11 liquid culture medium. 4 Grow wild-type E. coli cells in liquid LB medium. 5 Harvest photoautotrophically grown GE T. elongatus BP1 cells from actively growing liquid culture (at mid logarithm-growth phase) by centrifugation using Beckman Coulter Avanti J-26 Xp centrifuge. 6 Harvest wild-type E. coli cells from actively growing liquid culture (at mid logarithm-growth phase) by centrifugation using Beckman Coulter Avanti J-26 Xp centrifuge. 7 Make cell suspension of GE T. elongatus BP1 in BG-11 0 SA liquid medium at the cell population density of about 10 7 cells/mL. 8 Make cell resuspension of wild-type E. coli DH5α in LB liquid medium at the cell population density of about 10 7 cells/mL. 9 Mix 15 mL of GE T. elongatus BP1 (donor) cells resuspension (in BG-11 0 SA) with 15 mL of wild-type E. coli DH5α (recipient) cells resuspension (in LB). 10 Incubate the two (donor and recipient) organisms at 1:1 population ratio (the "15 mL + 15 mL" mixture) on a shaker at 37 C under continuous photosynthetic light intensity provided by daylight fluorescent lamps for a designated period of time (t Incubation ). 11 Take samples from liquid incubation co-culture and plate the liquid co-culture samples on LB agar medium in the presence and absence (control) of antibiotic; and spread (plate) a designated volume (V Incubation ), typically 100 ml, of the liquid co-culture (the donor-recipient co-incubation cell suspension liquid) onto the surface of a 45 mL antibiotic-containing LB agar medium per petri dish plate. 12 Incubate the LB plates at 37 C for E. coli colony development and observation to count for HGT events for a period of 1, 2, 4, 7 and up to 14 days in combination with control experiments. 13 Count the HGT events by counting the number of E. coli DH5α (recipient) colonies (N colony ) on the antibiotic LB agar plates. 14 Verify the HGT E. coli colonies (cells) by observation using Olympus DP672 Microscope in combination with control experiments. 15 Verify the HGT E. coli colonies by colony PCR in combination with control experiments. 16 Calculate the horizontal gene transfer (HGT) frequency from the number (N colony ) of HGT recipient (E. coli) colonies observed per antibiotic LB agar selective plate, the volume (V Incubation ) of the donorrecipient co-incubation cell suspension liquid that was used in spreading onto the surface of an antibiotic-containing LB agar medium plate, the concentration (C Recipient ) of the recipient (E. coli) cells in the co-incubation liquid, and the donor-recipient co-incubation time (t Incubation ) according to the following HGT Frequency equation: Method validation

Materials and strains preparation
To measure HGT in accordance of this protocol, it is a preferred practice to use genetically engineered cyanobacteria (blue-green algae) with a selectable marker such as an antibiotic-resistant transgene as a transgene donor. Currently, genetically engineered (GE) cyanobacteria can be created through two types of genetic transformation: 1) plasmid-based genetic transformation and 2) integrative genetic transformation through homologous recombination into the host genomic DNA. In a plasmid transformant, the transgenes such as an antibiotic-resistant transgene are in a self-replicable plasmid like pUC57 that can somehow replicate also in certain host cyanobacterial cell. The copy number of the plasmid per cell may vary depending on how well the plasmid could replicate in the cyanobacterial cell. During cyanobacterial cell division, the distribution of plasmid to the daughter cells may be random in nature. Consequently, plasmid transformants tend to be less stable than the genomic transformants whose transgenes are integrated into their genomic DNA. It is likely that the transgenes carried by a plasmid may be more mobile for HTG than those integrated into the cyanobacterial genomic DNA. The protocol may be used to test this feature by measuring HGT frequency for both the plasmid transgenes and/or genomic transgenes from GE cyanobacteria to E. coli. Examples of materials and strains preparation including plasmid preparation and genetic transformation of cyanobacteria such as Thermosynechococcus elongatus BP1 (BA000039.2) for creation of GE cyanobacteria including GE T. elongatus BP1 (pKB plasmid transformant) has been experimentally demonstrated in the Lee Laboratory at Old Dominion University and recently reported in Ref. [1].

Methods for measuring horizontal gene transfer
Horizontal gene transfer assay GE T. elongatus BP1 (pKB plasmid transformant) cells were photoautotrophically grown at 45 C in a Percival growth chamber in BG-11 0 SA medium containing kanamycin (40 mg/ml), and wild-type E. coli DH5α cells were grown in LB broth at 37 C in a shaking incubator. The photoautotrophically grown cells of GE T. elongatus BP1 were collected by centrifugation. The supernatant was discarded and the cells were resuspended in fresh BG-11 0 SA medium without antibiotic. Both GE T. elongatus BP1 cells and wild-type E. coli cells were resuspended to population density of about 10 7 cells/mL (Fig. 1). Using the cells resuspension at a concentration of 10 7 cells/mL, 15 mL cells resuspension of GE T. elongatus BP1 in BG-11 0 SA were mixed with 15 mL cells resuspension of wild-type E. coli DH5α in LB as shown in Fig. 2. The "15 mL + 15 mL" GE T. elongatus and wild-type E. coli liquid co-culture was incubated in a shaker at 37 C under continuous photosynthetic light intensity of about 8 mE m -2 s -1 provided by daylight fluorescent lamps. In addition to the liquid incubation co-cultures, two controls were also set up using the T. elongatus BP1 transformant and wild-type E. coli DH5α. Control 1 consisted of 15 mL of T. elongatus BP1 transformant (10 7 cells/mL) and 15 mL of LB broth without wildtype E. coli DH5α cells. Control 2 consisted of 15 mL of wild-type E. coli DH5α (10 7 cells/mL) and 15 mL of BG-11 0 SA medium without T. elongatus BP1 transformant. These controls were also incubated in a shaking incubator at 37 C under the same lighting conditions. Samples from each liquid co-culture and control were collected after 1, 2, and 3 days and were then sampled and spread on control LB agar plates and selection LB plates containing kanamycin. Typically, about 100 ml of liquid incubation coculture sample was used to spread onto the surface of a 45 mL antibiotic-containing LB agar medium per petri dish (100 mm diameter, 15 mm deep) plate (Fig. 3). The LB plates were incubated at 37 C under the same continuous photosynthetic lighting condition of 8 mE m -2 s -1 . PCR was used to verify the presence of the transgene cassette DNA within these colonies [1].
As an example, to study whether GE T. elongatus BP1 (pKB plasmid transformant) has ability to transfer its pKB plasmid into wild-type E. coli DH5α, the two (donor and recipient) organisms were coincubated at 1:1 population ratio for a designated period of time [1]. Then, the co-incubated liquid cell culture was sampled and plated on LB agar medium. The results for the control 1 (Fig. 3A), which only contained T. elongatus BP1 transformant, showed no colony growth on LB plates with and without kanamycin after 48 h of plate incubation at 37 C under lighting. This indicated that T. elongatus BP1 could not grow on LB plates. Therefore, any colony formed on LB agar plates from this experiment would be E. coli. Bacterial growth was seen on LB plates without antibiotic for control 2 which contained wild-type E. coli only. However, control 2 did not have any colony growth on LB plates with kanamycin (Fig. 3B). The only colony growth observed on LB agar plates containing kanamycin was from co-cultured samples (the "15 mL + 15 mL" GE T. elongatus and wild-type E. coli liquid incubation co-cultures). Typically, the observed number of HGT E. coli colonies ranged from 20 to 75 per plate [1].
Because control 1 (Fig. 3A) eliminated the possibility of T. elongatus BP1 growing on LB agar plates, the colonies on LB plates with kanamycin were indeed E. coli (Fig. 3C). It is worth noting that these HGT E. coli colonies grew much slower on LB agar plates containing kanamycin compared to the control wildtype E. coli on plates without antibiotic which only took 24 h to form observable colonies instead of  48 h. Randomly selected HGT E. coli colonies on kanamycin plates were further analyzed by PCR to confirm the presence of the target plasmid within the HGT E. coli cells [1].
The techniques of microscopic imaging can certainly be employed to visualize the donor (T. elongatus BP1 transformants) and recipient (E. coli) cells in the co-incubation liquid and to see the cells from selected HGT E. coli colonies. Fig. 4 presents a microscopic image of donor (T. elongatus BP1 transformants) and recipient (E. coli DH5α) cells in the co-incubation liquid that was examined under a light microscope (Olympus DP672 Microscope) after 2 days of co-incubation. Fig. 5 presents a microscopic image of the HGT E. coli DH5α cells from a colony obtained from LB kanamycin plate of coculture sample containing T. elongatus BP1 transformants and E. coli DH5α. These microscopic examination results indicate that the protocol is working since the cells from the selected HGT recipient colonies were indeed the HGT E. coli DH5α cells as expected [1].

Calculating the horizontal gene transfer frequency from experimental data
Based on the protocol, the HGT frequency may be calculated from the number (N colony ) of HGT recipient (E. coli) colonies observed per antibiotic LB agar selective plate, the volume (V Incubation ) of the donor-recipient co-incubation cell suspension liquid that was used in spreading onto the surface of an  antibiotic-containing LB agar medium plate, the concentration (C Recipient ) of the recipient (E. coli) cells in the co-incubation liquid, and the donor-recipient co-incubation time (t Incubation ) according to the HGT Frequency Eq. (1) described in the procedure above.
The HGT frequency expressed by this equation may be considered also as the probability for an HGT event to occur per recipient cell in a period of time under a given experimental condition. Table 1 presents an example of using the Eq. (1) to calculate HGT frequency from the experimentally observed numbers of HGT recipient (E. coli) colonies per antibiotic LB agar plate under the given experimental conditions including the plated co-incubation liquid cell suspension volume of 100 ml, recipient cell concentration of 1/2 Â 10 7 E. coli cells /1000 ml in the co-incubation liquid, and the donor-recipient co-incubation time of 2 days. The HGT frequency (probability) from the plasmid transformants of cyanobacterium Thermosynechococcus elongatus BP1 (donor) to wild-type E. coli (recipient) was calculated to be in a range from 2.0 Â 10 À5 to 7.5 10 À5 per cell day. That is, there could be about 20 to 75 HGT events per million recipient (E. coli) cells daily from the plasmid-based GE bluegreen alga Thermosynechococcus under the given experimental conditions. As a conclusion, the HGT frequency can be calculated from the number of HGT recipient colonies observed, the number of recipient cells plated, and the donor-recipient co-incubation time. This approach can also help test for the possible HGT routes to assess whether a HGT is through a direct cell-to-cell interaction or by an indirect cell-to-liquid environment-to-cell process. The protocol may be applied in full and/or in part with adjustments to measure HGT for a wide range of donor and recipient organisms of interest, for example, from algae to non-photosynthetic organisms.
"Tricks" and notes 1 The inability for certain cyanobacteria such as Thermosynechococcus elongatus BP1 to grow on LB agar plate makes the counting of HGT E coli colonies quite convenient. However, the protocol may be used  for all cyanobacteria including those that can grow on LB agar plate as well, since cyanobacterial colonies typically have a blue-green color readily distinguishable from E coli colonies. 2 It is important to properly select the HGT co-incubation liquid medium, which should be able to accommodate both the donor (cyanobacteria) and recipient (E. coli) cells. Sometimes, it may be challenging to find a liquid culture medium that could accommodate both the donor (cyanobacteria) and recipient (E. coli) cells. A logical solution may be a 50-50 mixture of the donor liquid culture medium (such as BG-11 0 SA) and the recipient liquid culture medium (such as LB) as used in the present experiment. 3 The liquid co-incubation of the donor (cyanobacteria) and recipient (E. coli) cells should be performed under proper experimental conditions including not only the temperature but also the lighting conditions to accommodate both the donor and recipient cells in consideration of a natural environment such as in case of GE cyanobacteria accidently released into a natural pond environment, where they may meet with other bacteria like E. coli. The light intensity typically needs to be at least about 5 mE m -2 s -1 to ensure the blue-green algae being able to function properly with photosynthetic energy support. 4 After plating of HGT co-incubation liquid sample onto the surface of antibiotic LB agar selective medium, it is a preferred practice to incubate the antibiotic LB selective plates not only at a proper temperature (such as 37 C) for HGT E. coli cells to grow, but also under a light intensity of at least about 5 mE m -2 s -1 to ensure the chlorophyll synthesis pathway in GE cyanobacteria is activated so that any potential cyanobacterial colony must have its green color that can be easily distinguished from an E. coli colony. 5 Note, the C Recipient used in the HGT frequency Eq. (1) described above is the cell population density (concentration) of the recipient (E. coli) in the co-incubation liquid, but not that of the donor (GE cyanobacteria). Therefore, in a "15 mL + 15 mL" GE T. elongatus and wild-type E. coli co-incubation liquid, the value for C Recipient was a half of the E. coli cell stock concentration (10 7 E. coli cells/1000 ml) used in making the 50-50 co-culture of GE T. elongatus and wild-type E. coli. 6 Based on the protocol, the donor of genetic material for HGT assays does not have to be GE cyanobacteria. It can be a plasmid material such as the pKB plasmid tested in the experiment [1]. It can also be other sources of genetic materials such as DNA materials from certain dead organisms of interest. Therefore, it is practically appropriate to express the HGT frequency (probability) on the recipient basis as shown in Eq. (1) using N colony over C Recipient , V Incubation and t Incubation . 7 The protocol may be used to also test for certain possible routes of HGT such as to answer the questions of: 1) Does a HGT require a direct cell-to-cell interaction? or 2) Could a recipient cell pick up a piece of DNA in a liquid environment such as a plasmid that may be released from a donor cell (an indirect cell-to-liquid environment-to-cell process for HGT)? This type of questions can be answered by comparative HGT assays in presence and absence of GE cells and/or free plasmid DNA in incubation liquid as demonstrated in the experiments of Fig. 3 using the protocol. 8 Based on the present experimental result [1], the HGT frequency (probability) from the plasmid transformants of cyanobacterium T. elongatus BP1 to wild-type E. coli was estimated here to be in a range from 2.0 Â 10 À5 to 7.5 10 À5 per cell day. The HGT probability from the genomic transformants of cyanobacteria to wild-type E. coli is likely to be rather low, which is yet to be measured. Special efforts such as increasing the incubation time and the donor and recipient cell population density for HGT liquid co-incubation with multiple replications may be needed to measure for such a potentially low probability genomic HGT event. This type of potentially challenging experiments with multiple replications may take months and even years of efforts to accomplish. It is a preferred practice to run a positive control experiment measuring HGT from a well-characterized plasmid transformant such as the pKB plasmid T. elongatus BP1 transformant to E. coli as a control in parallel with the long-duration main assay experiments to measure the genomic HGT event. This may help to ensure the long-duration experimental setups and conditions are all properly working so that the ultimate assay result whether positive or negative will represent the true outcome from the genomic HGT process of interest. 9 This protocol may be adjusted and applied in full and/or in part in combination with adjustments to measure HGT for a wide range of donor and recipient organisms from algae to non-photosynthetic organisms. For example, the recipient cell does not necessarily have to be E. coli. Other recipient organisms of interest such as yeast and human health-related bacteria including (but not limited to) Shigella, Campylobacter, and Salmonella may also be used for HGT assays in accordance of the protocol. Therefore, both the co-incubation liquid media and selective agar media plates may be adjusted in accordance of the protocol for the HGT assays based on the specific donor and recipient organisms of interest.

Additional information
In contrast to the "vertical" transmission of DNA from the parent to its offspring of the same species, horizontal gene transfer (HGT) is a natural process for an organism to transfer genetic material to another organism that is not of the same species, for example, from a blue-green alga to a totally distinct non-photosynthetic bacterium. It is generally believed that HGT is used as a means of evolution [2]. The phenomenon of HGT sometimes is noticeable in certain genomic analysis. For example, there may be a piece of very different genomic DNA such as a distinct GC-rich DNA "island" comprising certain genes in otherwise a largely AT-rich genome of an organism. This type of distinct genetic material likely came from a very different (organism) species through a HGT event in the natural evolutionary process. Does such a HGT event occur at a time scale of million years, hundreds of years, or a few days? Currently, it is not entirely clear how often such a HGT event could happen between two different species. This paper outline a protocol to measure HGT from genetically engineered (GE) cyanobacteria which are also known as blue-green algae to non-photosynthetic organisms such as wild-type E. coli.
The phenomenon of HGT is not only of an interest to the modern science of molecular genetics and biology, but also to the biosafety issue of genetic engineering in relation to biofuel energy and environmental sustainability as well as public health and wellbeing. Synthetic biology using genetically engineered (GE) cyanobacteria has the potential to produce valuable products such as biofuels. In recent years, GE cyanobacteria have become a promising new agriculture avenue for photosynthetic renewable energy. This approach addresses concerns of fossil fuels including their contribution to the increase in greenhouse gases (such as CO 2 ) in the Earth's atmosphere and their inevitable depletion as the world's major fuel source [3][4][5][6][7][8][9][10][11]. Photosynthetic production of various biofuels in GE cyanobacteria is accomplished by inserting a series of transgenes that will enable the direct conversion of photosynthetic metabolic intermediates to products such as ethanol and butanol [12][13][14]. Cyanobacteria are considered to be advantageous for several reasons which include the use of abundant raw materials (e.g. sunlight, water and CO 2 ), decrease in competition for land and crops, and potential in alleviating the increase of CO 2 in the atmosphere by being net carbon neutral [15][16][17][18].
While the synthetic biology approach to photosynthetic biofuel production is promising, there are several key unresolved bio-safety questions that should be addressed before such methods are considered for wide commercial use. One of the main concerns is the ability of GE cyanobacteria to transfer their modified genes to other bacteria through HGT if they were to escape containment [19,20]. Through this process, it is possible that wild-type bacteria will be able to acquire designer transgenes such as those intended for biofuel production and/or antibiotic resistance. Because antibiotic resistance genes are widely used as a selective method for the genetic transformation of photosynthetic organisms and bacteria, HGT could increase the risk of proliferation which can have potential ramifications to human and animal health among other areas [21].
Better understanding the phenomenon of HGT between engineered and wild-type organisms (whether from the environment or common lab strains) will advance basic science knowledge in synthetic biology with microorganisms and could assist government agencies in the development of important policies and regulations for genetic engineering and use of GE cyanobacteria for biofuel production as well as other areas.
Therefore, this protocol is developed to determine whether a blue-green alga such as GE Thermosynechococcus elongatus BP1 has the potential to transfer its transgenes to wild-type E. coli and assess their possible HGT frequency. Earlier studies of HGT showed that E. coli can transfer genes to cyanobacteria [22][23][24][25][26][27]. In the present work, the protocol of measuring horizontal gene transfer from GE blue-green algae to wild-type E. coli was experimentally demonstrated through the use of GE T. elongatus BP1 carrying a designer DNA construct with a plasmid to produce alcohol that was coincubated with wild-type E. coli DH5α in a 1:1 ratio. HGT was then monitored on solid Luria Broth (LB) media with selective antibiotic for E. coli growth. The experimental results indicate that plasmids can be transferred from GE T. elongatus BP1 (transgene donor) to wild-type E.coli (recipient) after two days of liquid co-incubation [1]. The frequency (probability) of HGT for a plasmid transgene from the donor to the recipient organism was calculated, for the first time, from the experimental data based on the protocol.
In summary, this protocol could be extremely useful for testing of HGT between diverse microorganisms employing recombinant transgene DNA constructs. It utilizes the intrinsic differences in appearance and metabolism, with selectable antibiotic markers, to specifically screen for the transfer of gene(s) between transgenic blue-green algae and non-photosynthetic bacteria. Presence of the transgene(s) in the putative HGT recipients is then confirmed by PCR in combination of colony formation and re-streaking on culture plates with antibiotics. This type of assay may provide quantifiable information on possible unintentional genetic transfer from recombinant microalgae developed for commercial production to other microorganisms present in the environment. This approach could also be used to estimate horizontal gene transfer between Escherichia coli or other bacteria containing recombinant DNA transgenes and cyanobacteria (HGT in the opposite direction). Another possibility is to use an E. coli strain carrying a plasmid with a selectable marker different from that present in the cyanobacterial donor in this experimental scheme, then to subsequently screen for both selectable markers following the co-incubation step. Therefore, this novel protocol may be adjusted and applied in full and/or in part in combination with adjustments to measure HGT for a wide range of donor and recipient organisms from algae to non-photosynthetic organisms, which could be highly useful for a number of applications in various fields from the science of molecular genetics and biology to the biosafety of genetic engineering.

Author's contribution
The author JWL conceived the original idea of measuring horizontal gene transfer from algae to non-photosynthetic organisms and wrote this article for the novel protocol here with his original equation on HGT Frequency calculation. JWL first proposed the idea and wrote the method for measuring horizontal gene transfer in a research grant application in 2015. The proposed project was subsequently funded by the Biotechnology Risk Assessment Grant Program competitive grant award no. 2016-33522-25624 from the U.S. Department of Agriculture. This method has now been successfully practiced experimentally through the funded project work activity in measuring horizontal gene transfer. The graphic abstract was adapted and modified from Nguyen et al. 2019 Gene, 704:49-58 [1].