Results and Discussion rep78 plant vector construction Initial strategies to express rep78 in A. thaliana involved cloning rep78 into a plant expression cassette in plasmid pRTL (Topfer et al., 1987). Under the regulation of the CaMV 35S promoter and the Nos terminator, the rep78 cassette was inserted into the Ti plasmid pBIB-HYG (Becker et al., 1992). This plant vector does not contain a reporter gene, and plants are selected by hygromycin resistance. However, no transformed plants were obtained, and an alternative expression strategy was employed. The pCAMBIA series of plant vectors is designed for Agrobacterium-mediated transformation (Hajdukiewicz et al., 1994). pCAMBIA-3302 contains GFP with 5 restriction sites, and a GFP fusion protein may be produced via in-frame cloning into one of these upstream sites. This allows for quick characterization of gene expression and/or protein location. Regulatory sequences within the plasmid include the cauliflower mosaic virus 35S promoter and the plant Nos terminator for constitutive expression of rep78-GFP. As disarmed vectors, chromosomally-located Agrobacterium genes orchestrate insertion of the plasmid T-DNA into the plant genome. pCAMBIA-3302 was chosen to express AAV-2 rep78 in A. thaliana (Fig. 4). This vector allows plant transformants to be selected by resistance to phosphinothricin -based herbicides (see Materials and Methods, rep78 Plant Vector Construction), while bacterial transformants are selected by kanamycin resistance. Samulski et al. (1987) created a recombinant plasmid carrying the infectious AAV-2 genome termed pSUB-201. No convenient restriction sites exist in pSUB-201 for rep excision, although a BglII site is present in the cloning region of pCAMBIA-3302. 30 Using rep-specific primers sRep3 and sRep4 that incorporate terminal BglII sites (Fig. 4), pSUB-201 was used as a template for rep78 amplification. Figure 6 shows the expected 1866 bp rep78 PCR product (Fig. 6, lane 2) with a 1 kb DNA standard (Fig. 6, lane 1). This amplification also incorporates three 5 nucleotides so as to provide in-frame cloning into the plant vector pCAMBIA-3302. Cloning of this rep78 product into BglIIdigested pCAMBIA-3302 yields pSdDan (Fig. 4). This strategy places rep78 in frame with GFP, ensuring expression of a functional Rep78-GFP fusion protein. A glycine residue is introduced between Rep78 and GFP in the process. 31 1 2 3 kb 1.6 kb 400 bp Figure 6. Amplification of rep78 from pSUB-201. Lane 1: 1 kb DNA standards. Lane 2: rep78. 32 Internal rep primers Rep5 and Rep6 were used to confirm insertion of rep78 into pCAMBIA-3302. Figure 7 shows the expected 330 bp rep78 PCR product with pSdDan as template (Fig. 7, lane 4). The figure also shows the results of the same PCR using empty plasmid (Fig. 7, lane 2) and pSUB-201 (Fig. 7, lane 3) as template. With positive and negative controls in place, this amplification demonstrates that rep78 has indeed been cloned into the vector. Lane 1 contains a 1 kb DNA standard (Fig. 7, lane 1). GFP and rep primers GFPL and Rep5 were used to determine the orientation of rep in pCAMBIA-3302. Figure 7 shows PCR products with empty plasmid (Fig. 7, lane 5) and pSdDan (Fig. 7, lane 6) as template. The 1900 bp product in lane 6 consists of a 1500 bp rep78 fragment and 400 bp GFP fragment amplified from the rep78-GFP fusion cassette. Were rep78 inserted in the wrong orientation (3 -5 ) no product would result, as with the negative control (Fig. 7, lane 5). This amplification using a rep78 primer and a GFP primer demonstrates that rep78 has been inserted in the correct orientation upstream of GFP in the plant vector. 33 1 3 kb 1.6 kb 2 3 4 5 6 400 bp 300 bp Figure 7. Amplification of rep78 (lanes 2-4) and rep78-GFP (lanes 5-6) fragments. Lane 1: DNA standards. Lane 2: negative control; pCAMBIA-3302. Lane 3 positive control; pSUB-201. Lane 4: pSdDan. Lane 5: negative control; pCAMBIA3302. Lane 6: pSdDan. 34 rep78 in A. thaliana Wild type and AAVS1-engineered Arabidopsis plants were transformed with A. tumefaciens GV3101 carrying pSdDan. Approximately 0.2-1% of the seedlings produced from these transformed plants survived herbicide selection. This phenotype suggests genomic integration of pSdDan T-DNA carrying the herbicide-resistance gene. Plants that survived selection were tested for the presence of rep78 in chromosomal DNA by PCR. Genomic DNA from this first generation of plants (T1) was extracted and used as template with rep-specific primers (see Materials and Methods, PCR Analysis of Transformed Plants). Two out of 14 WT plants (#2 and #11) tested positive for rep amplification. The data presented here is from plant #2, though the results apply to both plants. No AAVS1-rep plants were obtained. The full 1866 bp rep gene was amplified as five ~400 bp fragments. Figure 8 shows these PCR products with wild type (Fig. 8, lanes 2, 4, 6, 8, and 10) and plant #2 (Fig. 8, lanes 3, 5, 7, 9, and 11) genomic DNA as template. All amplified fragments were the expected size based on the location of the primers within rep78. There were no products formed in the wild type reactions (Fig. 8, lanes 2, 4, 6, and 8), demonstrating that the products formed in the reactions using plant #2 DNA as template (Fig. 8, lanes 3, 5, 7, and 9) are in fact rep78 fragments and not Arabidopsis sequence. This PCR data suggests that this plant has indeed been stabily transformed with rep78. 35 1 2 3 4 5 6 7 8 9 10 11 1600 bp 500 bp 300 bp Figure 8. Amplification of rep78 fragments from transgenic plant #2. Lane 1: DNA standards. Lanes 2, 4, 6, 8, and 10: wild type genomic DNA with primers Rep3 and Rep7, Rep5 and Rep6, Rep8 and Rep9, Rep10 and Rep11, and Rep12 and Rep4, respectively. Lanes 3, 5, 7, 9, and 11: plant #2 genomic DNA with the same primer sets. 36 Both transformed plants were also tested for GFP by PCR. Using GFP primers GFPU and GFPL, plants testing positive for rep also tested positive for a ~400 bp GFP fragment. Figure 9 shows PCR products with wild type (Fig. 9, lane 2) and plant #2 (Fig. 9, lane 3) genomic DNA as template. The ~400 bp product in lane 3 is of the correct size based on the location of the GFP primers. No product was formed in the amplification reaction using untransformed plant DNA as template (Fig 9, lane 2), demonstrating that the product in lane 3 is indeed GFP, and not Arabidopsis, sequence. The presence of GFP in the plant genome further confirms successful transformation with pSdDan. It should be noted that the older leaves of plant #2 T1 generation exhibited curious mottling (Fig. 10). The T2 generation did not exhibit mottling. Normally associated with nutritional deficiency or infection, this lesion-like chlorosis was not observed with any other plants during the study. The possible pleiotropic effects of rep78 expression in Arabidopsis are not known. 37 1 2 3 1600 bp 1000 bp 500 bp 400 bp Figure 9. Amplification of GFP fragment from transformed plant #2. Lane 1: DNA standards. Lane 2: wild type genomic DNA with primers GFPL and GFPU. Lane 3: plant #2 genomic DNA with the same primers. 38 Figure 10. Chlorotic lesions as observed in the older leaves of the T1 generation of transformed plant #2. 39 rep78 sequence in A. thaliana The rep78 fragments amplified from plant #2 genomic DNA (Fig. 8) were purified and used as templates for nucleotide sequencing reactions. 1856 out of the 1866 amplified bp (~99%) were sequenced and shown to have 100% identity to AAV-2 rep78 (Fig. 11). This information confirms the presence of the complete rep78 gene in the plant genome. AAV-2 rep78 Plant 2 sequence 1 Plant 2 sequence 2 5 -ACGATCCCCAATATGCGGCTTCCGTCTTTCTGGGATGGGCCACG-3 5 -ACGATCCCCAATATGCGGCTTCCGTCTTTCTGGGATGGGCCACG-3 3 -ACGATCCCCAATATGCGGCTTCCGTCTTTCTGGGATGGGCCACG-5 Figure 11. Alignment of rep78 (930 to 973 bp shown) and plant #2 sequence data. Rep78 expression in A. thaliana rep transcription Our transformation strategy relies on a functional Rep enzyme in the plant system. Rep expression was therefore characterized in Rep-transformed plants. Qualitative mRNA assays such as RT-PCR indicate transcription of specific genes. RNA was isolated from plant #11 (T2 generation) and used as template with rep-specific primers. If rep is being transcribed in the plant, the message should reverse transcribe and amplify via RT-PCR; wild type and non-rep transcribing plants should not yield RTPCR product. Figure 12 shows these RT-PCR products with wild type (Fig. 12, lanes 2, 4, and 6) and plant #11 (Fig. 12, lanes 3, 5, and 7) RNA as template. The sizes of the products in lanes 5 and 7, and the lowest molecular weight product in lane 3, correspond to those 40 of the expected amplified rep fragments (~400 bp, see Fig. 8) based on the location of the primers within rep78. There were no products formed in the wild type reactions (Fig. 12, lanes 2, 4, and 6), demonstrating that the products formed in the reactions using plant #11 RNA as template (Fig. 12, lanes 3, 5, 7) are in fact based on rep transcript and not Arabidopsis messages. This RT-PCR data suggests that the rep-GFP fusion is being transcribed in plant #11. This hypothesis is further confirmed as plant #11 also tested positive for GFP transcription by RT-PCR. Figure 13 shows RT-PCR products with GFP-expressing plant (Haseloff et al., 1997; a gift of Brenda Winkel; Fig. 13, lane 2) and transformed plant #11 (Fig13, lane 3) RNA as template. The size of both products corresponds to the expected size of the amplified GFP fragment (400 bp, see Fig. 9). The product of the reaction with GFP-expressing plant RNA (Fig. 13, lane 2) is of the same size as the transformant RNA product (Fig. 13, lane 3), demonstrating that the product in lane 3 is in fact based on GFP transcript and not Arabidopsis transcript. This RT-PCR data further supports rep-GFP transcription in transformed plant #11. RT-PCR is sensitive to genomic contamination. In order to show that the products were amplified from reverse transcribed RNA, and not DNA, standard PCR reactions with Taq polymerase were also performed with the isolated RNA. Without reverse transcriptase, no products should form if the RNA sample is devoid of genomic DNA. No PCR products formed in reactions containing plant #11 RNA template and Rep or GFP primer pairs (Fig. 14 A, lanes 3 and 5; B, lane 3). 41 1 2 3 4 5 6 7 1000 bp 400 bp 200 bp Figure 12. rep RT-PCR from transgenic plant #11. Lane 1: DNA standards. Lanes 2, 4, and 6: wild type RNA with primers Rep5 and Rep6, Rep10 and Rep11, and Rep12 and Rep4, respectively. Lanes 3, 5, and 7: plant #11 RNA with the same respective primer sets. 42 1 2 3 3000 bp 1000 bp 400 bp Figure 13. GFP RT-PCR from transgenic plant #11. Lane 1: DNA standards. Lane 2: GFP-expressing plant RNA with primers GFPU and GFPL. Lane 3: plant #11 RNA with the same primers. 43 A 3000 bp 1 2 3 4 5 B 1 2 3 1000 bp 600 bp 400 bp 500 bp Figure 14. rep and GFP PCR from transgenic plant RNA and positive templates. A. Lane 1: DNA standards. Lane 2: E. coli containing pSdDan with primers Rep5 and Rep6. Lane 3: transgenic plant #11 RNA with the same primers. Lane 4: E. coli containing pSdDan with primers Rep10 and Rep11. Lane 5: transgenic plant #11 RNA with the same primers. B. Lane 1: DNA standards. Lane 2: E. coli containing pSdDan with primers GFPU and GFPL. Lane 3: transgenic plant #11 RNA with the same primers. 44 Rep translation Western blots using Rep and GFP-specific antibodies were performed in an effort to detect the Rep-GFP fusion protein in the T2 generation of transformed plants. The encouraging RT-PCR data suggest rep is being transcribed, therefore Western blots might reveal detectable levels of Rep protein being translated. However, neither anti-Rep nor anti-GFP antibody could detect the fusion protein in transformed plant #2 or #11. Figure 15 shows the GFP-specific antibody reaction with purified GFP (Fig. 15, lane 2), wild type Arabidopsis lysate (Fig. 15, lane 3), transformed plant #2 lysate (Fig. 15, lane 4), and transformed plant #11 lysate (Fig. 15, lane 5). Shown also is the Rep78/68-specific antibody reaction with AAV-infected HeLa cell lysate (Fig. 15, lane 6), wild type Arabidopsis lysate (Fig. 15, lane 7), transformed plant #2 lysate (Fig. 15, lane 8) and transformed plant #11 lysate (Fig. 15, lane 9). The appropriately sized GFP and Rep bands (Fig. 15, lanes 2 and 6, respectively) indicate that the antibody system was able to detect target protein (the Rep protein is deceivingly large, as the lower numbered lanes electrophoresed ahead of the higher lanes). However, no protein bands were evident in the transformed plant lysate lanes (Fig. 15, lane 4, 5, 8, and 9). GFP fluorescence was not observed by fluorescent microscopy in the transformed plants through the T2 generation (data not shown). Consistent with the Western blot data, this suggests that the Rep-GFP fusion protein is not produced at detectable levels. suggests that, while the rep-GFP fusion gene may be transcribed, it is not being translated to a detectable level in the transformed plants (see Conclusions for discussion). This 45 anti-GFP 1 200 kDa 116 97 66 2 3 4 5 6 anti-Rep 7 8 9 Rep 45 GFP Figure 15. Western blot of rep/GFP transformed plants. Lanes 2-5 use primary antibody against GFP. Lanes 6-9 use primary antibody against Rep78/68. Lane 1: protein standards. Lane 2: purified GFP. Lane 3: wild type Arabidopsis lysate. Lane 4: transformed plant #2. Lane 5: transformed plant #11. Lane 6: AAV-2+Ad-5 infected HeLa cell lysate. Lane 7: wild type plant lysate. Lane 8: plant #2. Lane 9: plant #11. 46 Rep78 expression in vitro Without access to purified Rep protein, a positive control was needed for use with transformed plant Western blot data (Fig. 15, lane 6). Therefore, AAV-infected <a href="/keyword/hela-cells/" >hela cells</a> were used to ensure the efficacy of Rep antibodies. A productive AAV-2 HeLa cell infection can be achieved by co-infection with adenovirus type 5. Doubly infected cells were harvested at 8 hours (Fig. 16, lane 3), 12 hours (Fig. 16, lane 4), 24 hours (Fig. 16, lane 5), and 36 hours (Fig. 16, lane 6) post infection. Cells infected with Ad-5 only were also harvested at 36 hours and were used as a negative control for AAV-2 Rep78 synthesis (Fig. 16, lane 2). Cells infected with Ad-5 alone show diffuse reaction to the Rep78/68-raised monoclonal antibody and one band at ~130 kDa. Most importantly, the 78 and 68 kDa Rep bands were not visible in this Ad-5 only sample, while they were clearly visible in the Ad-5 and AAV-2 infected sample (Fig. 16, lane 6). While nonstructural Rep proteins have been detected as early as 4 hours post infection (Redemann et al., 1989), figure 16 shows the detection of Rep78/68 by monoclonal antibody 36 hours post infection (Fig. 16, lane 6). The antibody s ability to detect Rep78 is demonstrated here, and it will be used to characterize Rep78 expression in transformed plants. 47 Ad 12 200 kDa 116 97 66 AAV+Ad 456 3 Rep78 Rep68 45 36 hrs 8 12 24 36 Figure 16. Western blot of AAV-2 and Ad-5 infected HeLa cell lysate. Lane 1: protein standards. Lane 2: Ad-5 alone, 36 hours post infection. Lane 3: AAV-2+Ad-5, 8 hours post-infection. Lane 4: AAV-2+Ad-5, 12 hours post infection. Lane 5: AAV-2+Ad-5, 24 hours post infection. Lane 6: AAV-2+Ad-5, 36 hours post infection. 48 Conclusions In the absence of helper virus and during subsequent low levels of replication, adeno-associated virus type 2 selectively integrates into a region termed AAVS1 on human chromosome 19 (Berns and Linden, 1995). Rep78/68 is a viral enzyme that catalyzes this reaction (Linden et al., 1996). Since AAVS1 shares homology with AAV2 termini, Rep binds and nicks each strand (Im and Muzyczka, 1990). Integration is believed to occur as the cellular replication machinery switches templates between the proximal AAV-2 and AAVS1 (Linden et al., 1996). Consistent with this function are Rep s biochemical activities, including site-specific endonuclease activity, DNA binding, and helicase activity (Berns and Linden, 1995). This mechanism has been employed for non-homologous, targeted integration in mammalian cell culture and for gene therapy studies (Ponnazhagan et al., 2001; Larson et al., 2001). We hope to develop a novel plant transformation strategy using this recombination technology. To this end, the goal of this project is to express AAV-2 Rep78/68 in plants. rep78 was introduced into plant vector pCAMBIA-3302, producing a rep78-GFP fusion cassette (pSdDan, Fig. 4). A. tumefaciens harboring pSdDan was used to transform wild type A. thaliana. The presence of rep78 in the plant genome of two plant lines was demonstrated by PCR (Fig. 8). Sequence analysis further confirmed the introduction of the complete rep78 gene into Arabidopsis (Fig. 11). RT-PCR demonstrated rep transcription in one transformed plant line (Fig. 12), but Western analysis failed to detect Rep-GFP protein in the T2 generation of either transformed plant line. 49 There are a number of potential explanations for the lack of detectable protein in the plants. The rep-GFP gene may simply be expressed at levels too low to detect. This may be due to the location of the gene(s) within the plant genome, yielding the sequence susceptible to position effects and inefficient expression (Jin et al., 2002). The location(s), orientation, and copy number of rep within the plant genome have not been investigated. The codon bias of A. thaliana, may be a factor in the low rep expression levels. Transgenes are often reconstructed with preferred codons to accommodate the host s translational machinery (Pan et al., 1994). CUG (Leu) and CCC (Pro) are used four times less frequently in Arabidopsis than in humans, where AAV-2 Rep is normally synthesized. GCC (Ala) is used nearly three times less frequently. All other codons are used at comparable frequencies. Each of these three codons comprises about 6% total of those in Rep78. Therefore, it seems unlikely that a lack of available tRNAs in the plant would account for undetectable levels of Rep, though each individual codon bias could theoretically compound and make for inefficient translation. Post-transcriptional gene silencing would also explain the lack of Rep protein in the transformed plants. Believed to be a defense mechanism against pathogens such as RNA viruses, post-transcriptional gene silencing is often a problem with transgene expression in plants (Matzke et al., 2002). Small plant RNAs mediate homologydependent dsRNA formation and degradation, and methylation of the foreign DNA. This commonly observed but poorly understood silencing mechanism offers an attractive explanation (Chicas et al., 2001), as transcript was present but no Rep protein was detected. 50 This problem of Rep expression in Arabidopsis should be approached several ways...