1 Q UANTITATIVE DETERMINATION OF SELECTIVE ALTERNATIVE HOSTS OF CANDIDATUS LIBERIBACTER ASIATICUS AND POTENTIAL FOR TRANSMISSION TO CITRUS By HAO HU A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORI DA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 2012 Hao Hu
3 To my parents, for their unconditional love
4 ACKNOWLEDGMENTS I would like to express my deep gratitude to all those who made this dissertation possible. First of all, I am deeply indebted to my supervisory committee chair, Dr. Ronald H. Brlansky, for his guidance and support during my doctoral study. He is always willing to share his expertise and knowledge when I expe rience difficulties; forgiving when I make mistakes; and encouraging when I am down. I am greatly influenced and encouraged by his scientific spirit and attitude, and his rigorous scholarship will be the treasure of my life in my future academic career. I would like to acknowledge Drs. Jeffrey B. Jones, John S. Hartung, and Nian Wang for serving on my committee. Their academic expertise is an indispensable help for me to accomplish this study. I would like to give my special thanks to all members working in the Brlansky lab, Carmen Bierman, Rachana Dagli, Diann S. Acho r, Rebecca Clarke, Ananthakrish nan Govindarajulu, Nandlal Choudhary, and Avijit Roy, who kindly offer me help and support in all kinds of issues. I acknowledge Dr. Michael J. Davis and his l ab members, Huiqin Chen, and Maria E. Peacock for their help in the psyllid room and microscopic work. I am so lucky to have them not only as workmates, but also my friends in life. I personally thank all my friends in Gainesville, here in CREC, Lake Alf red, and in China, for their friendship accompanying me through these years. I wish all of them more success in their future study and career. Lastly and most importantly, I would like to thank my parents and sister in China for their unconditional love. W ithout their support, I would not have been able to come this far and pursue my study here.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ ........ 10 LIST OF ABBREVIATIONS ................................ ................................ ........................... 11 CHAPTER 1 LI TERATURE REVIEW ................................ ................................ .......................... 15 1.1 Citrus H uanglongbing ................................ ................................ ....................... 15 1.1.1 HLB t he D isease ................................ ................................ ................... 17 184.108.40.206 Disease t ype ................................ ................................ .................. 17 220.127.116.11 Geographic d istribution ................................ ................................ .. 17 18.104.22.168 Transmission ................................ ................................ .................. 18 22.214.171.124 Symptom development ................................ ................................ .. 20 126.96.36.199 Disease management ................................ ................................ .... 21 1.1.2 Candidatus Liberibact er spp. t he A ssociated P athogen s ...................... 23 188.8.131.52 Bacteria distribution in planta ................................ ......................... 24 184.108.40.206 Genetic research ................................ ................................ ............ 25 1.1.3 Citrus P syllid the V ector ................................ ................................ ....... 27 220.127.116.11 Life cycle of D. citri ................................ ................................ ......... 28 18.104.22.168 Clim atic requirements ................................ ................................ .... 28 22.214.171.124 Distribution ................................ ................................ ..................... 29 126.96.36.199 Ca Liberibacter spp. in psyllid s ................................ ..................... 30 1.2 H ost s in HLB ................................ ................................ ................................ ..... 31 1.2.1 Citrus H osts ................................ ................................ ............................. 32 1.2.2 Alternative H osts ................................ ................................ ..................... 34 1.3 HLB D etection ................................ ................................ ................................ ... 35 1.3.1 HLB D iagnosis ................................ ................................ ......................... 35 1.3.2 L ive vs. D ead B acteria ................................ ................................ ............. 38 2 PMA QPCR METHODOLOGY ................................ ................................ ............... 42 2.1 Introduction ................................ ................................ ................................ ....... 42 2.2 Materials and M ethods ................................ ................................ ...................... 46 2.2.1 Plant and P syllid M aterials ................................ ................................ ...... 46 2.2.2 PMA qPCR W orking P rotocol ................................ ................................ .. 47 2. 2.2.1 PMA pretreatment ................................ ................................ .......... 47 188.8.131.52 qPCR ................................ ................................ ............................. 49 2.2.3 Optimization of the PMA qPCR W orking P rotocol ................................ ... 51 184.108.40.206 With or without TissueLyser (TL) treatment ................................ ... 51
6 220.127.116.11 Centrifugation speed 16,100 g vs. 5,000 g ................................ ..... 52 18.104.22.168 Final concentration of PMA and EMA ................................ ............ 52 2.2.4 PMA qPCR vs. R ods/field ................................ ................................ ....... 53 2.2.5 Effect of TL T reatment on Live Bacterial Genome ................................ ... 54 2.2.6 Determination of Interference of Plant Tissue with PMA qPCR ............... 55 2.2. 7 Determination of the Live Bacteri al Genomes in S ymptomatic and A symptomatic Leaves from Las Positive Field Trees ................................ .... 57 2.3 Results ................................ ................................ ................................ .............. 58 2.3.1 Optimization of PMA qPCR W orking P rotocol ................................ ......... 58 22.214.171.124 With or without TL treatment ................................ .......................... 58 126.96.36.199 Centrifugation speed 16,100 g vs. 5,000 g ................................ ..... 58 188.8.131.52 Final concentration of PMA and EMA ................................ ............ 58 2.3.2 PMA qPCR versus D irect C ounting with M icroscope .............................. 59 2.3. 3 Effects of TissueLyser T reatment on L ive B acterial Genome .................. 59 2.3.4 Determination of Interference of Plant Tissue with PMA qPCR ............... 60 2.3.5 Live B acteria in S ymptomatic or A symptomatic HLB Leaf S amples ........ 61 2.3. 6 PMA qPCR on D ifferent H ost Plants ................................ ....................... 61 2.4 Discussion ................................ ................................ ................................ ........ 62 2.4.1 Optimization of PMA qPCR ................................ ................................ ..... 62 2.4.2 PMA qPCR vs. R ods/field ................................ ................................ ....... 6 5 2. 4 3 Effects of TissueLyser T reatment on L ive B acterial Genome .................. 66 2.4.4 Determination of Interference of Plant Tissue with PMA qPCR ............... 67 2. 4 .5 Live B acteria in S ymptomatic and A symptomatic Leaves ....................... 68 2.4. 6 Live Las in Different Hosts ................................ ................................ ....... 69 3 SEVERINIA BUXIFOLIA AS AN EXCELLENT ALTERNATIVE HOST OF LAS AND A SE ASONAL LIVE BACTERIAL DEVELOPMENT SHOWN BY PMA QPCR ................................ ................................ ................................ ..................... 79 3.1 Introduction ................................ ................................ ................................ ....... 79 3.2 Materials and Methods ................................ ................................ ...................... 82 3.2.1 Plants and P syllids ................................ ................................ .................. 82 3.2.2 Inoculation P rocedure ................................ ................................ .............. 83 3.2.3 Detection A ssays ................................ ................................ ..................... 85 3.3 Results ................................ ................................ ................................ .............. 88 3.3.1 Transmission E xperim ents ................................ ................................ ...... 88 184.108.40.206 From citrus to S. buxifolia transmission pathway 1 ..................... 89 220.127.116.11 From S. buxifolia to citrus transmission pathw ay 2 ..................... 89 18.104.22.168 From S. buxifolia to S. buxifolia transmission pathway 3 ............ 90 3.3.2 Las I dentity C onfirmation ................................ ................................ ......... 91 3.3.3 Symptom O bservation ................................ ................................ ............. 92 3.3.4 Dynamic C hange of L ive Las P opulation ................................ ................. 93 3.4 Discussion ................................ ................................ ................................ ........ 94 3.4.1 S. buxifolia as A lternative H ost of Las ................................ ..................... 95 3.4.2 Live Las P opulation in S. buxifolia and C itrus ................................ .......... 96 3.4.3. U ncommon Las Population F luctuation ................................ ................ 100
7 4 MULTIPLE CITRUS RELATIVES AS ALTERNATIVE HOSTS of LAS AND PSYLLID TRANSMISSION STUDIES ................................ ................................ .. 109 4.1 Introduction ................................ ................................ ................................ ..... 109 4.2 Materials and M ethods ................................ ................................ .................... 111 4 .2.1 Pla nts and P syllids ................................ ................................ ................ 111 4.2.2 Inoculation P rocedure ................................ ................................ ............ 111 4 .2.3 Detection A ssays. ................................ ................................ .................. 113 4.3 Results ................................ ................................ ................................ ............ 115 4.3.1 Transmission E xperiments ................................ ................................ .... 115 22.214.171.124 x Citrofortunella microcarpa ( Calamondin ) ................................ ... 115 126.96.36.199 Citropsis gilletiana ................................ ................................ ........ 117 188.8.131.52 Choisya spp. ................................ ................................ ................ 119 184.108.40.206 Zant hoxylum fagara ................................ ................................ ..... 121 220.127.116.11 Esenbeckia runyonii ................................ ................................ ..... 122 18.104.22.168 Amyris texana ................................ ................................ .............. 123 4.3.2 Las I dentification ................................ ................................ ................... 124 4.3.3 Symptom O bservation ................................ ................................ ........... 124 4.3.4 Las P opulation D ynamics in A lternative H osts ................................ ...... 125 4 .4 Discussion ................................ ................................ ................................ ...... 127 4.4.1 Transmission E xperiments and A lternative H ost S tatus ........................ 127 4.4.2 Live Las P opulation i n H ost P lants ................................ ........................ 129 5 SUMMARY AND FUTURE PERSPECTIVES ................................ ....................... 145 APPENDIX: SEQUENCES AND BLAS T RESULTS ................................ ................... 147 LIST OF REFERENCES ................................ ................................ ............................. 150 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 166
8 LIST OF TABLES Table page 3 1 qPCR and conventional PCR systems used in this study ................................ 102 3 2 Reaction of Severinia buxifolia to Las infection transmitted b y psyllids from infected citrus ................................ ................................ ................................ ... 103 3 3 Reaction of Valencia sweet orange to Las infection transmitted by psyllids from infected S. buxifolia ................................ ................................ .................. 104 3 4 Reaction of S. buxifolia to Las inoculation transmitted by psyllids from infected S. buxifolia ................................ ................................ .......................... 105 4 1 Results of psyllid transmission pathway 1 from citrus to Calamond in ............... 134 4 2 Results of psyllid transmission pathway 2 from Calamondin to citrus ............... 134 4 3 Results of psyllid transmission pathway 3 from Calamondin to Calamondin .... 134 4 4 Results of psyllid transmission pathway 1 from citrus to Citropsis gilletiana ..... 135 4 5 Res ults of psyllid transmission pathway 2 from C. gilletiana to citrus ............... 135 4 6 Results of psyllid transmission pathway 3 from C. gilletiana to C. gilletiana ..... 135 4 7 Results of psyllid transmission pathway 1 from citrus to Choisya spp ............. 136 4 8 Results of psyllid transmission pathway 2 from Choisya spp. to citrus ............. 136 4 9 Results of psyllid transmission pathway 3 from Choisya to Choisya ................ 136 4 10 Results of psyllid transmission pathway 1 from citrus t o Zanthoxylum fagara .. 137 4 11 Results of psyllid transmission pathway 2 from Z. fagara to citrus ................... 137 4 12 Results of psyllid tra nsmission pathway 3 from Z. fagara to Z. fagara .............. 138 4 13 Results of psyllid transmission pathway 1 from citrus to Esenbeckia runyonii .. 139 4 14 Results of psyllid transmission pathway 2 from E. runyonii to citrus ................. 139 4 15 Results of psyllid transmission pathway 3 from E. runyonii to E. runyonii ........ 139 4 16 Dynamic change of live Las population inside various graft transmitted host plants ................................ ................................ ................................ ................ 143
9 4 17 Dynamic change of cumulative Las genome inside vario us graft transmitted host plants ................................ ................................ ................................ ........ 143 4 18 Summary of HLB transmission experiments on all the tested citrus relatives ... 144 4 19 Su mmary of Las persistency and psyllid positive rate on all the tested citrus relatives ................................ ................................ ................................ ............ 144
10 LIST OF FIGURES Figure page 2 1 Comparison bet ween with TissueLyser treatment (TL) and without TL in PMA pretreatment. ................................ ................................ ................................ ..... 71 2 2 The effect of different concentrations of PMA or EMA on DNA yield s ............... 72 2 3 Correlation between common log value of bacterial concentration from microscopic counting and Cq values obtained by PMA qPCR. .......................... 73 2 4 Las observed under fluorescent microscope stained with both SYTO 13 and PMA. ................................ ................................ ................................ ................... 74 2 5 Effects of TissueLyser treatment on live and cumulative Las genome tested by PMA qPCR ................................ ................................ ................................ .. 75 2 6 PMA qPCR with defined ratios of live and isopropanol killed Xanthomonas citri ssp. citri ................................ ................................ ................................ ....... 76 2 7 Comparison of the l ive bacterial genome s of symptomatic and asymptomatic le aves of Hamlin sweet orange. ................................ ................................ .......... 77 2 8 Comparison of live Las genome in 24 different host plants. ............................... 78 3 1 Symptoms observed in HLB t ransmission experiments of S. buxifolia group. .. 106 3 2 Dynamics of live and cumulative genome of Ca L. asiaticus in S. buxifolia and citrus plants over a 20 month period.. ................................ ....................... 107 4 1 Psyllid activities observed on alternative hosts ................................ ................. 133 4 2 Leaf symptoms observed on Las infected plant hosts ................................ ...... 140 4 3 Gel images of PCR amplification of Las with different primers ......................... 141
11 LIST OF ABBREVIATION S AAP acqui sition access period ACP Asian citrus psyllid AH alternative host ANOVA analysis of variance bp base pair Cq quantitative cycle CREC Citrus Research and Education Center CTV citrus tristeza virus EMA ethidium monoazide f.c. final concentration HLB citrus huanglongbing IAP inoculation access period ICBR Interdisciplinary Center for Biotechnology Research k b kilo base pairs Laf Candidatus Liberibacter africanus Lam Candidatus Liberibacter americanus Las Candidatus Liberibacter asiaticus LBP live bacteri al percentage LB G live bacterial genome LSD least significant difference LT low temperature (treatment) Mb mega base pairs NCBI National Center for Biotechnology Information PCR polymerase chain reaction
12 PD physical disruption (treatment) PI 1 post inoculation PI 2 propidium iodide PMA propidium monoazide PMA qPCR real time fluorescent polymerase ch ain reaction with propidium monoazide pretreatment qPCR real time fluorescent polymerase chain reaction rDNA ribosomal deoxyribonucleic acid rRNA ribosomal ribonucleic acid SDW sterile distilled water TL TissueLyser (treatment)
13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Do ctor of Philosophy Q UANTITATIVE DETERMINATION OF SELECTIVE ALTERNATIVE HOSTS OF CANDIDATUS LIBERIBACTER ASIATICUS AND POTENTIAL FOR TRANSMISSION TO CITRUS By Hao Hu August 2012 Chair: Ronald H. Brlansky Major: Plant Pathology Citrus huanglongbing (H LB) is a devastating citrus disease. HLB in Florida is associated with bacterial pathogen Candidatus Liberibacter asiaticus (Las) and transmitted by Asian citrus psyllid ( Diaphorina citri ). Some citrus relatives have been listed as hosts of the psyllid and /or the associated bacterium based on field surveys or PCR tests on field samples, but their status has never been system at ically studied. In this work, 8 citrus relatives, i.e. Severinia buxifolia x Citrofortunella microcarpa Citropsis gilletiana Esen beckia runyonii Zanthoxylum fagara Choisya aztec Pearl Choisya ternata Sundance and Amyris texana were studied to investigate their alternative host status. P ossible transmission pathways for each plant were tested with repeat ed psyllid transmissi on experiments as well as grafting where compatible. After inoculation, plants were monitored for symptom development and tested by real time PCR (qPCR). The results showed that all plants studied except A. texana were infected by Las although their transm issibility varied a lot based on bacterial persistency and psyllid activities.
14 Although estimating live bacterial genome (LB G ) is critical for HLB research, PCR has limitations on differentiating live and dead cells. Propidium monoazide (PMA), a novel DNA binding dye, has already been successfully used on many bacterial plant pathogens to effectively remove DNA from dead cells, but no applications on uncultured bacteria like Las were reported. In this study, PMA qPCR protocols were first optimized to work with plant and psyllid materials, respectively. Then, they were used to determine LB G in various studies, such as establishing correlation between LB G and microscopic counting, checking the reactions of different citrus plants to Las infection, and checkin g the connection between LB G and leaf symptom expression. Lastly, the LB G dynamics inside HLB positive citrus and non citrus hosts was monitored monthly through a 20 month period, and a seasonal development pattern was observed in both hosts. T his study e xperimentally demonstrated the alternative host status of 7 pla nt species of which 6 plants were 1 st time reported in the world. T he optimized PMA qPCR provides an accurate way to determine LB G in hosts of Las which should benefit various HLB research an d serve as a crucial component in HLB management.
15 CHAPTER 1 LITERATURE REVIEW 1.1 Citrus H uanglongbing Citrus is the most important tree fruit crop in the world with production in over 100 countries of all the six continents (Saunt 2000). Among the many citrus diseases, huanglongbing (HLB) is probably the most destructive and devastating one worldwide (Halbert and Manjunath 2004; Bove 2006) The disease can reduce fruit quality and cause premature fruit drop, and eventually kill infected trees. Fruits fr om HLB affected trees are often small, misshaped with bitter taste, and they usually do not color properly, for which reason the disease is also called citrus greening in Africa and the Americas (like the United States) (Bove 2006; da Graca 1991 ; da Graca and Korsten 2004 ). For a long time, HLB only affect ed citrus producing countries in Africa (Pietersen et al. 2010) and Asia (Bove 2006; Ding et al. 2009; Gottwald et al. 1989; Manjunath 2009) the Arabian Peninsula, and the Reunion and Mauritius Islands, b ut in recent years, it has spread to the Americas. In 2004, it was first reported in the State of Sao Paulo, Brazil (Coletta Filho et al. 2004) which was also the first report of HLB in Western Hemisphere, then Florida, USA in 2005 (Halbert 2005) Cuba in 200 8 (Martinez et al. 2009; Luis et al. 2009) Dominican Republic in 2008 ( Matos et al. 2009 ) and Honduras in 2010 (Manjunath et al. 2010) This spread has caused great concern to the other states and neighboring countries with citrus production, especia lly those where the citrus psyllid vector is present. For all countries, HLB poses a great threat to the local citrus industry. For example, since the discovery of citrus HLB in south Florida in 2005, the disease has cost the State of Florida more than 3.6 3 billion dollars and more than 6611 (full time and part time) jobs.
16 T hree types of HLB have been reported ( Bove 2006; Zhang et al. 2011 b ) and they are associated with three species of phloem restricted G ram negative bacteria, Candidatus Liberibacter asiaticus (Las), Ca L. africanus (Laf) (da Graca, 1991; Planet et al. 1995) and Ca L. americanus (Lam) ( Teixeira et al., 2005 a ; Coletta Filho et al. 2005 ), which all belong to the alpha subdivision of Proteobacteria (Jagoueix et al. 1994) These fastid ious bacteria are not in pure culture been fulfilled for HLB, which poses a major obstacle in HLB research. Although HLB is graft transmissible, it is mainly transmitted by an insect vector, the citrus psyllid, in the field. Besides, for long distance transmission, citrus material transportation by human and wind driven insect vector are the major means reported (Halbert et al. 2010a; Halbert et al. 2010b; Ram a dugu et al. 2008). There are two types of citrus psyllids responsib le for HLB transmission, the Asian citrus psyllid (ACP) Diaphorina citri in Asia and the Americas; and the African citrus psyllid, Trioza erytreae in Africa. This disease was first demonstrated to be a graft transmissible (i.e. infectious) disease by C hinese phytopathologist Lin in 1954, therefore, the Chinese name of the disease that L in used in his work, or huanglongbing in English, which means conference of the International Organization of Citrus Virologists (IOCV) in Fuzhou, Fujian, China. Today the name of HLB is widely used for the three types of the disease. A ll known citrus species and cultivars are affected by HLB with little known resistance and no known cure. Nowadays, m ajor management strategies are insecticide applications to reduce psy llid populations removal of infected trees to eliminate sources of bacterial inoculum, and th e establishment of pathogen free nursery systems.
17 1.1.1 HLB t he D isease 22.214.171.124 Disease t ype The Asian type of HLB associated with Las is mainly distributed in Asia and the Americas, and it is a heat tolerant type (Garnier and Bove 1993; Bove 2006). HLB in India does well in hot conditions above 25 C Asian HLB used to be kept out of the areas with cold winter climates, and its symptoms are less pronounced and di sappear above 1500 m, most probably due to the climate limit on vectors (Wang et al. 2006; Aubert 1987). The African type of HLB associated with Laf is mainly distributed in Africa, and it is heat sensitive. African HLB manifests symptoms primarily under c ool conditions (below 25C), and under greenhouse conditions, it does not show symptoms above 27 C In South Africa, HLB symptoms are more pronounced in winter than in summer, and African HLB is more serious in elevations above 700 m. S ymptoms of African H LB are moderate to severe at 22 to 24 C and disappear at 27 to 32 C while symptoms of Asian HLB from India and Philippines are severe at both temperature regimes The American type of HLB associated with Lam is only found in Brazil, and it is heat sensiti ve (Lopes et al. 2009b). Lam was the predominant species when HLB was first reported in Brazil in 2004 (Teixeira et al. 2005 a ), but there was a disproportionate increase in the occurrence of Las, and now Lam has bec o me more difficult to find (Lopes et al. 2009a) 126.96.36.199 Geographic d istribution For geographic distribution of HLB in the world, many review papers have compiled the information and listed all the countries and areas affected (Bove 2006; Halbert and Manjunath 2004), therefore, there is no need t o repeat their work here in this short introduction. T here are t wo general ideas about HLB distribution worth sharing :
18 one is that HLB is affecting all the major citrus producing countries and areas and the number of countries is still increasing (especial ly in the Americas where HLB is becoming endemic); the other one is that once a psyllid vector is present in certain area, it s only a matter of time that HLB also is discovered in the area So far, the citrus regions free of psyllids and HLB are the Medit erranean basin, most of Western Asia (Near and Middle East), Australia, and North and South Pacific islands, although they all are endangered by neighbor ing areas contaminated with the psyllid vectors and/or HLB. Along with the wide distribution in the wo rld, bacterial isolates from different HLB areas were found to have considerable genetic and physiological variations ( Bastianel et al. 2005; Tomimura et al. 2009 ; Ding et al. 2009; Furuya et al. 2010; Hu et al. 2008; Jagoueix et al. 1997; Katoh et al. 201 1; Ramudugu et al. 2011; Teixeira et al. 2008a; Tomimura et al. 2010), which was studied and used for different purposes, such as tracking the origin of the bacterium in a new HLB breakout to the source area. 188.8.131.52 Transmission Other than graft transmiss ion (mainly in the nursery) and psyllid transmission (natural way in the field), which are both well established transmission pathways of HLB, whether HLB is transmissible from infected trees to the next generation through seeds has not reached a definite answer so far. HLB infection has a major impact on fruit quality, and research show ed that seed quality including seed number per fruit, seed weight, seed germination, and seedling size, is also adversely affected (Albrecht and Bowman 2009). Regarding see d transmission, Tirtawidjaja (1981) reported that seeds taken from normal looking and symptomatic (small, misshaped, and discolored) fruits were planted and different results were observed: no symptoms were found on seedlings from seed taken from normal
19 fr uit (even when they are collected from infected trees) but some seedlings produced from seeds of fruits with HLB abnormality had stunted plants with chlorosis. Three of the seedlings had the same appearance as psyllid inoculated plants. However, no molecu lar diagnosis was included in this study, nor repeat s w ere adequate. In some of the most recent studies, a small percentage of the seedlings grown from seeds of HLB affected trees were positive for Las with quantitative PCR (Graham et al. 2008; Albrecht an d Bowman 2009; Hilf 2011 ), but the PCR results n either persist ed n or correlate d with disease symptoms (i.e. no HLB disease found on those seedlings later). These all suggested that the bacterial pathogen was not transmitted into the seedling s in a way per mitting disease development. It seems clear that all studies on HLB seed transmission which have relied solely on qPCR for pathogen detection can not derive a conclusive answer to this question (Hilf 2011). Further research, especially the ones with new app roaches and methodologies to address this question more thoroughly, is needed (Hilf 2011). Citrus fruit s are well vascularized organs and during seed development, the chalaza produces the integuments (inner and ou t er seedcoat) and the nucellus, thus prov iding a pathway for the bacterium to enter the seed (Albrecht and Bowman 2008). In an investigation of Las distribution throughout the plant, Tatineni et al. (2008) demonstrated the presence of Las in peduncles and seedcoat s of seeds collected from HLB af f ected citrus trees. But the bacterium was not detected in endosperm or embryos. A s imilar situation with other systemic plant disease systems occurs for example, the phloem limited mulberry dwarf phytoplasma was detected in the seedcoats (not in the embry o) of dwarf diseased mulberries (Jiang et al. 2004), but no
20 seed transmission of this disease has been reported. For citrus tristeza disease which is caused by phloem limited citrus tristeza virus (CTV), it s mainly transmitted by aphid vector, not by seed (Gonzalez Segnana et al. 1997). 184.108.40.206 Symptom d evelopment Some earlier observations and more recent work have shown the connection between disruption of sugar movement pathway in plant tissue and the appearance of chlorotic symptoms (i.e. typical blot chy mottle leaves) (Kim et al. 2009; Takushi et al. 2007). The current understanding of the HLB chlorotic leaf symptom is that chloroplast disruption is caused by severe internal starch granule accumulation, which is due to the lack of sugar export, i.e. phloem blockage. Sometimes, the chlorosis occurs in small sectors of leaves, which results in the blotchy mottle symptom; while sometimes, all upstream leaves become uniformly chlorotic, which suggests the phloem blockage occurs in the stem below. These un derstandings have supporting evidence from multiple anatomical and gen etic researches First of all, sucrose accumulation, plugged sieve pores, and phloem disruption were all observed with various microscopes (Folimonova and Achor 2010; Etxeberria et al. 2 009) Etxeberria et al. (2009) found extraordinarily high level of starch in all aerial tissues, photosynthetic cells, phloem elements, vascular parenchyma, xylem parenchyma, and phelloderm of HLB affected sweet orange trees; in contrast, roots of the HLB affected trees were depleted of starch whereas roots of healthy control plants contained substantial starch deposits. Second, a microarray research conducted by Kim et al. (2009) found that the e xpression of 624 genes in HLB af fected sweet orange was signi ficantly affected, and the up regulation of key starch biosynthetic genes ( ADP glucose pyrophosphorylase, starch synthase, granule bound starch synthase and starch debranching enzyme ) may contribute to the
21 starch accumulation process. R egarding phloem bloc kage, it is likely to be the result of plugged sieve pores rather than bacterial aggregates. Since the pp2 gene, which is related to callose deposition, is up regulated in infected host plant this could lead to sieve pores plugging Plus, Las does not for m aggregates in citrus tissue ( K im et al. 2009). Third, Folimonova et al. (2009) reported that continuous light conditions increased symptom expression in infected plants (especially in more tolerant citrus genotypes) and reduced the time before symptoms d eveloped. This could be more evidence for the understanding of HLB symptom development described above. T he amount of sugar production increase s due to the extended photosynthesis period and this aggr av ates the carbohydrates translocation problem. A direc t connect ion of Las bacterial activity and the leaf symptoms was not evident, because large numbers of bacteria could be found in phloem sieve tubes of pre symptomatic young flushes, but were not so abundant in highly symptomatic leaf samples. This may sug gest that a major proportion of the Las bacteri um is nonviable at more advanced stages of the disease (Folimonova and Achor 2010). 220.127.116.11 Disease management Considering the following fact s of HLB, (i) no resistant citrus genotypes available (ii) no curat ive methods to save infected trees (iii) high transmission efficiency in the field by psyllid vector and (iv) many potential alternative hosts in the field for the psyllid vector, the control of citrus HLB has to involve all aspects of an integrated pest m anagement program (Bove 2006; Halbert and Manjunath 2004). If HLB is not present, quarantine measures should be enforced to keep it out (Bove 2006). When HLB has entered a region previously HLB free, all possible measures should be considered. First of all a field survey should be conducted immediately to determine the extent of the
22 spread. Next prevent new trees from becoming infected with HLB because a successful management relies on the early and aggressive implementation of HLB control practices when the disease incidence is still very low (Belasque et al. 200 9; Manjunath et al. 2007a; Manjunath et al. 2008a ). There is no place in the world where citrus HLB occurs that it is under completely successful control (Halbert and Manjunath 2004). That having been said, various disease control practices have been applied to HLB (Bove 2006) Due to limited success or in consistent results, some practices, like physical and chemical therapie s targeting the systemic Las (Zhang et al. 2010 a ; Zhang et al. 2010b; Zhan g et al. 2011 a ; Zhang et al. 2012), biological control against the psyllid vector (Aubert 1987; Michaud 2002 a ; Michaud 2002b; Michaud 2004; Michaud 2001 ; Michaud and Olsen 2004; Pluke et al. 2005 ), host plant resistance (Koizumi et al. 1993; Albrecht et al 2012 ; Dutt et al. 2012; Bowman et al. 2009 ), and cultural control ( Bove 2006 ), were deemed ineffective or inadequate at least at current time (Westbrook et al. 2011). These measures are still under research or are used sometimes as supplements to other management practices For places where HLB occurs like Asian countries and American regions (Florida, USA, and Sao Paulo State, Brazil), three main measure s to manage the disease are commonly enforced: psyllid control, removal of potential inoculum source s and propagation of clean nursery stock (Hung et al. 2000; Brlansky and Rogers 2007; Belasque et al. 200 9; Belasque et al. 2010; Manjuanth et al. 2008b; Manjunath et al. 2007b; Ramudugu et al. 2007 ). However, regarding the removal of HLB positive trees, there is some controversy about to what extent this practice should be enforced, or whether it is practical for its designated purpose. G rowers and researchers are working on some alternatives to this drastic measure, for example,
23 mineral application or en hanced nutrition application on older trees to compensate the damage caused by HLB but mixed results are reported (unpublished results) 1.1. 2 Candidatus Liberibacter spp. t he A ssociated P athogen s Besides Laf, Las and Lam, Garnier et al. (2000) also de scribed a subspecies of Laf named Candidatus Liberibacter africanus subsp. capensis, from a native r utaceous plant Calodendrum capensis Thunb. in Western Cape Region of South Africa which can also infect citrus. The associated agents are highly fastidiou s, and so far, culturing the three bacteria has not been successful (Davis et al. 2008 ; Schuenzel et al. 2008 ), or widely accepted (Sechler et al. 2009). The identity of the HLB associated pathogen has gone through phases of virus, phytoplasma, bacteria like organism (BLO), and recently, bacteria (Bove 2006). Although the bacteria have not been cultured to complete Koch s postulates, all circumstantial evidence point s strongly to a bacterial agent behind this disease ( Bove and Ayres 2007) First, Ca L s pp. are detected in almost all the typical HLB samples. Some e xceptions include a few negative results with Ca L spp. in HLB samples in various reports (Chen et al. 2009; Chen et al. 200 8 ) A phytoplasma related to Ca Phytoplasma asteri was reported t o be present in citrus samples with HLB symptoms ( Chen et al. 2009; Marques et al. 2012; Teixeira et al. 2008c ). Second, various research es show that l iberibacters or some phytoplasma are transmitted to receptor plants with the disease ( Hung et al. 2000 ; M arques et al. 2012; Teixeira et al. 2008c ). Third, HLB disease symptoms abate temporarily when the plants are treated with antibiotics (Zhang et al. 2011; Zhang et al. 2012). Although some reports may suggest mix ed infection s in HLB for example, Chen repo rted that 69 out of 141 citrus samples
24 tested were PCR positive for both Las and Ca P. asteri (2009), the bacterial identity of the causal agent is still valid. 18.104.22.168 Bacteria distribution in planta It is critical to un derstand the distribution and m ovement of Ca L. spp inside an individual citrus tree, which can help in HLB management (Tatineni et al. 2008 ; Sagaram et al. 2008 ). Regarding bacteria l distribution, there is an assumption that liberibacters multipl y in the phloem tissue at the inoculati on site and move to other parts of the plant in both downward and upward directions This is based on electron microscopy studies and multiple observations of symptom development in naturally infected and experimentally inoculated trees. HLB symptoms usual ly start at the tip of one or more branches (i.e. yellow shoots), and they progress with time until the entire canopy is involved (Bove, 2006). This statement is supported by multiple researches Tatineni et al. (2008) reported that Las was detected in al l parts of graf t inoculated citrus trees including bark tissue, leaf midrib, roots, and different floral and fruit parts, but not in endosperm or embryo, which indicated systemic movement from infection site to different parts of the plant. With the help o f qPCR, they also confirmed that the bacterium was unevenly distributed in planta (ranging from 14 to 137,031 cells/ total DNA in different tissues ), and a relatively high bacteria l concentration was found in fruit peduncles. Ding et al. (2008) indica ted that Las could invade into the meristem, and 74.7% of plants obtained from conventional meristem tip culture were Las positive by nested PCR. Therefore, they used vitrification cryopreservation to eliminate Las from in vitro adult shoot tips and obtain ed up to 98.1% of Las free plants. C ryotherapy of shoot tips is a common technique for pathogen eradication to produce healthy planting materials (Wang et al. 2009), and it has the potential to help in certified clean nursery
25 stock programs for HLB managem ent. Folimonova et al. (2009) reported another type of uneven distribution of Las They found that high population of bacteria usually associated with symptomatic leaves, while for asymptomatic ones, either high or non detectable level of bacterial populat ion could occur This actually leaves a problem when selecting materials for diagnosis or graft inoculation work before any visual symptom is present. A pruning experiment carried out in Sao Paulo, Brazil (Lopes et al. 2007) also demonstrated downward move ment and overall systemic distribution for Lam in hundreds of field trees. Pruning was done by removing only the symptomatic shoots (cut them at the trunk) or by removing the entire canopy (cut the trunk 15 20 cm ab ove the graft line) M ottled leaves reapp eared on most of the lightly pruned trees (69.2%) and some of the heavily pruned ones (7.6%), regardless of the age, variety of the trees and pruning procedure. According to Folimonova et al. (2009), in graft inoculated plants, the movement of Las from don or tissue to receptor host is quite rapid after phloem connections are established because three out of four plants became positive even when inoculum was removed from receptor hosts after 4 weeks. 22.214.171.124 Gen etic r esearch Since no successful pure culture has been obtained for any Liberibacter species ( Bove 2006 ), to obtain genome DNA from these bacteria is extremely difficult (Zhang et al. 2011 b ). However, the Las psy62 genome (GenBank NC_012985.2) was successfully obtained by metagenomics using DNA extra cted from a single infected psyllid (Duan et al. 2009) after which molecular studies on Las became popular ( Akula et al. 2011a; Akula et al. 2011b; Zhang et al. 2011b) The genome size of Lam was found to be about 1.31 Mb, and the entire 23S and 5S rRNA g enes were already published (Wulff et al. 2009). In th e greatly reduced genome (1.23 Mb) of Las there is a high percentage of
26 genes involved in both cell motility (4.5%) and active transport (8.0%), while almost no offensive pathogenicity factors, such as plant colonizing extracellular degradative enzymes or type III and type IV secretion system components, are found, which is consistent with its intracellular lifestyle (Duan et al. 2009). A prophage like region of DNA was annotated in the Las psy62 genome and the gene variation in this region has been associated with the variations found in disease severities among different Las strains (Tsai et al. 2008; Tomimura et al. 2009). For lysogenic phages, lysogenic conversion genes (non lysis, nonstructural, an d nonvirion assembly genes expressed in prophages were considered lysogenic conversion genes when they alter the physiology or pathogenicity of the host) are thought to play an important role in the emergence of new pathogens and disease forms (Casjens 200 3; Boyd and Brussow 2002; Wagner and Waldor 2002; Boyd et al. 2001). I t was found that Las carries an excision plasmid prophage and a chromosomally integrated prophage that becomes lytic in plant infections (Zhang et al. 2011 b ). Zhang et al. (2011 b ) also r eported that phage particles associated with Las were found in the phloem of infected periwinkles by transmission electron microscopy, while in the psyllid, they were found only as prophage. With the help of microarray assays, the transcriptional response of citrus to Las infection w as studied. In one attempt to identify genes associated with tolerance to HLB, 326 genes were found to be significantly up regulated by at least 4 fold in the susceptible genotype compared with only 17 genes in tolerant one (Al brecht and Bowman 2011). While in another study, the microarray analysis indicated that Las infection significantly affected the expression of 624 genes whose encoded proteins were grouped into 18 categories, which included sugar metabolism, plant defense,
27 phytohormone, cell wall metabolism and 14 other categories ( K im et al. 2009). Albrecht and Bowman (2008) also reported change of the transcriptional profile of Valencia sweet orange in response to Las infection, and a notable pathogen induced accumulation of transcripts for a phloem specific lectin PP2 like protein was highlighted, which was later conf irmed by Kim et al. (2009) Based on a genome comparative study with four other members of the Rhizobiales, Hartung et al. (2011) reported that at least 50 c lusters of conserved genes were found on all the five metabolically diverse species ; and similar to Bartonella henselae an intracellular mammalian pathogen, the Las genome had all the hallmarks of a reduced genome of a pathogen adapted to an intracellular lifestyle. In order to understand the defense and stress response of citrus to Las infection, various investigations of the profile change at the transcriptional and post transcriptional level were conducted in recent years (Khalaf et al. 2010; Fan et al. 2010; Fan et al. 2011). 1.1.3 Citrus P syllid the V ector Citrus HLB probably is the most destructive disease of citrus caused by a vectored pathogen. The Asian citrus psyllid ( D citri Kuwayama, Hemiptera : Psyllidae) is therefore the most serious pest of citrus if any of the HLB associated pathogens are present. Unless the citrus trees are severely infested, direct psyllid feeding damage is minor to the plant (Halbert and Manjunath 2004). Populations of D. citri can reach extremely high levels, and more than 40,000 psyllids per tree were collected from a report of insecticide research. When such a heavy infestation happens, usually curled or notched new leaves will be killed (Halbert and Manjunath 2004), or sometimes eve n the entire terminals (Michaud 20 04). In total there are 13 different psyllid species reported to occur on citrus, and 7 of them are from the Diaphorina genus. T erytreae is the well known
28 African citrus psyllid, and it can be easily separated from D. citri based on its clear forewings t hat are pointed at the tips. Besides, nymphs of T. erytreae live in individual depressions on the undersides of citrus leaves, while nymphs of D. citri tend to colonize the stems of new shoots and never produce individ ual pits on the leaves (Halbert and Ma njunath 2004). 126.96.36.199 L ife cycle of D. citri 4 days and t he five nymphal instars are completed in 11 15 days (Aubert 1987 ; Halbert and Manjunath 2004 ). D. citri can finish a life cycle in 15 47 days dependi ng on the temperature, and adults may live up to several months and females may lay as many as 800 eggs in a lifetime (Mead 1977). Liu and Tsai reported the life cycle parameters of D. citri in Florida (2000). The time needed for a complete life cycle was found to be the same as what Mead reported (1977), and the best temperature range for D. citri development is 25 28C. T he female D. citri was found to lay most eggs (average number of 748.3 per insect) at 28C. 188.8.131.52 Climatic requirements It wa s reporte d that D. citri does not tolerate frost well, and the insect also does not do well when humidity is clos e to the saturation point perhaps because high humidity promotes fungal growth to which the nymp hs are very susceptible (Aubert 1987). However, things o bserved in Florida are somehow different. First of all, populations of D. citri are found in Gainesville, FL, where temperatures drop to at least 5C on several nights, and the high humidity in the Florida summer time seems not to prevent extremely high p opulations in loca l groves and backyards (Halbert and Manjunath 2004).
29 184.108.40.206 Distribution D. citri can be found in southeast Asia and the Indian subcontinent, the islands of Reunion and Mauritius, Saudi Arabia, and many citrus producing Latin Americ an countries, like Brazil and Argentina (da Graca 1991; Halbert and Manjunath 2004; Bove 2006 ; Halbert and Nunez 2004 ). In the United States, D. citri was first reported in Florida in June of 1998 and had quickly spread throughout the state to all the coun ties with citrus (Halbert et al. 2002 ; Tsai et al. 2000; Tsai et al. 2002 ). Reports o f psyllid discoveries in other states with citrus soon followed As of June, 2010, Texas (French et al. 2001), California (August, 2008), Arizona, Louisiana (June, 2008), Mississippi Alabama, Georgia, South Carolina, Hawaii, Puerto Rico, were quarantined for D. citri due to the concern of HLB. With the presence of D. citri it is usually just a matter of time before HLB is confirmed. For example, since the first discovery of HLB in south Florida in 2005, the disease has spread to all 34 counties with commercial citrus production (Halbert et al. 2008), and HLB was found in other states following the psyllid spread: Louisiana (June, 2008), Georgia, South Carolina (2009), Texa s (2012), California (2012). Two likely scenarios for the introduction of D. citri into Florida have been proposed. One scenario is that the insect could have spread naturally from South America, where it has established for many years, through Central Am erica and the Caribbean and finally found its way to Florida. Several interception records supported the hypothesis of D. citri gradual spread out in the Western Hemisphere (Halbert and Manjunath 2004) However, in th e first po ssible scenario, the psyllid population from Latin America should be free of HLB associated pathogens. The other possible scenario is that D. citri could have been introduced directly from Asian countries, and numerous
30 interception records of live D. citri found on imported citrus an d other rutaceous plants support this hypothesis. 220.127.116.11 Ca Liberibacter spp. in psyllid s It is widely accepted that the Ca L. spp can multiply in both of the psyllid vectors, but this had not been demonstrated with molecular evidence. One study found drastic increases of the number of the Laf inside T. erytreae over 9 days, and the researchers concluded that the bacteria were multiplying in the vector (Moll and Martin 1973) In th e current alternative host s work psyllid detection in single psyllids wa s conducted by quantitative PCR (i.e. the total number of bacteria in each psyllid) The data showed that there was a huge difference in the bacteria l population tested from single psyllids. The Cq values from single psyllids could range from 16 to 39, wh ich means a difference of almost 7 orders of magnitude (10 millions) (unpublished data in this study) This huge difference of bacterial population inside psyllids is not likely caused by different feeding time or merely by chance. Therefore, the data also supported that the Ca L. spp (or at least Las from this study) could multiply inside D. citri Considering the endosymbiont relation between Ca L. spp and the psyllid vectors suggested by research facts, like low bacteria vector specificity found in th is combination (Halbert and Manjunath 2004; Gottwald 2010), it is not hard to understand the bacteria should be able to multiply inside their long time host. It has been shown in laboratory studies that transmission of Las from parent to offspring (transov arial) occurs at a rate of 2 6%, and adult D. citri which acquires Las as nymphs are better vectors of the pathogen compared with adults that acquire the pathogen as adults (Inoue et al. 2009 ; Pelz Stelinski et al. 2010).
31 1.2 H ost s in HLB When discus sing hosts in HLB, two types of plant s are of concern, one is the plant to support the psyllid vectors, and the other is the plant in which the bacterial pathogen can multiply Research shows that the two types of plant s have different significance in HLB management C ompared to the wide physiological host range of the bacterial pathogens t he psyllid vectors have a relatively narrow host range i.e. the bacterial pathogen can live in various citrus and non citrus hosts while the psyllid vector seems to fe ed on only citrus and it s relatives. Considering the low vector pathogen specificity this may have potential implications for the disease epidemiology (Halbert and Manjunath 2004; Gottwald 2010). Halbert and Manjunath (2004) summarized the host range of D. citri and Ca Liberibacter spp., respectively. They pointed out that many hosts on the two lists were included based on field survey ( i.e. observations of plant symptoms or psyllid behavior) and/or some PCR tests, so their host status were not experime ntally established. For example, in the host list of D. citri there is only one host based on comparative laboratory study ( Tsai and Liu 2000). In their study, Tsai and Liu (2000) found that grapefruit was the best host of D. citri out of the four plants studied: Murraya paniculata (L.) Jack (orange jasmine), Citrus jambhiri Lushington ( R ough lemon), C aurantium L. (sour orange), and C x paradisi Macfad. (grapefruit), and there was no statistical difference among the other three hosts. Based on their own observations in the Citrus Arboretum in Winter Haven, Florida, Halbert and Manjunath (2004) suggested that the two Florida native Zanthoxylum plants, Z. clavahercules L. and Z. fagara (L.) Sarg., and Casimiroa edulis Llave & Lex. may be non hosts (or very poor hosts as in the case of Z. fagara ) of D. citri E ven though the se plants almost always have suitable flush, no D.
32 citri were ever (or rarely) found on th em In addition they also reported that no D. citri were found on Z. fagara plants growing next to an infested lime grove in South Florida (Halbert and Manjunath 2004). Due to the difficulty of detecting the HLB associated bacter ia with certainty, there is not much information on the host range of the liberibacters and m uch of the available infor mation is based on symptoms (Halbert and Manjunath 2004). For citrus cultivars, especially the commercial ones, most (if not all) of them are susceptible to some degree regardless of rootstocks (Halbert and Manjunath 2004; Bove 2006). With the help of mole cular assays, some studies have shown that several citrus relatives could harbor HLB associated bacteria for example, Severinia buxifolia (Poiret) Ten. (Hung et al. 2000; Hung et al. 2001 ; Ramadugu et al. 2010 ; Deng et al. 2008 ), Limonia acidissima L. (Ko izumi et al. 1996 ; Hung et al. 2000), Clausena lansium (De ng et al 2010; Ding et al. 2005 ), and Toddalia lanceolata Lam (Korsten et al. 1996). Some citrus species ( C indica Tan. and C macroptera Montr.) remained symptom free under heavy inoculum pressur e (Bhagabati 1993), which may indicate a certain degree of resistance. 1.2.1 Citrus H osts One reason HLB is so devastating is most (if not all) commercial citrus species and cultivars are sensitive to the disease regardless of rootstock (Bove et al. 2006; Halbert and Manjunath 2004). C itrus tristeza disease, another destructive citrus disease caused by CTV, can induce quick decline and death of citrus trees on sour orange rootstock C ontrol of the disease is achieved by replacing the sour orange rootstock with tolerant rootstocks, while this control measure is not available for HLB (Bove 2006). However, one characteristic of HLB is that different degrees of disease and symptoms are induced in different types of citrus, and for some citrus relatives, such as M
33 paniculata (L.) Jack (orange jasmine), only minim al symptoms will manifest on infected plants ( Damsteegt et al. 2010 ). Furthermore, different isolates of Las can cause different amount of disease in citrus cultivars as well (Tsai et al. 2008). There i s no real resistance to HLB in citrus species, but some species and cultivars have some tolerance. Several extensive field surveys showed that some cultivars were more susceptible to decl ine than others (Koizumi et al. 1993). For example, grapefruit was mo re tolerant than most of the sweet orange cultivars. Generally speaking, sweet oranges, mandarins and tangelos are most susceptible, lemon and grapefruit are more resistant, while limes, citranges and Poncirus trifoliata are the most tolerant species (Koiz umi et al. 1993; Bove 2006; Halbert and Manjunath 2004 ; Lopes and Frare 2008; Tsai et al. 2006 ). Most information on different citrus genotype reactions to HLB have been accumulated from observations of field trees made under different conditions, at diffe rent geographic locations, and at different times. And there are some disagreements between different works (Albrecht and Bowman 2011; Folimonova et al. 2009). In a study conducted under controlled greenhouse conditions, Albrecht et al. (2012) found that t olerance to HLB was higher in trees grafted on some rootstock selections even though different rootstocks did not affect disease incidence and trees on all rootstocks were considerably damaged by HLB. They reported that none of the 15 rootstocks tested in duced a high level of resistance in sweet orange scion in the early years following infection by Las but high vigor inducing rootstocks, particularly Volkamer lemon, might enable younger trees to outlast the damaging effects of the disease. Folimonova et al. (2009) tested the reaction of 30 different citrus genotypes to inoculation with 3 different Las isolates (no difference
34 observed among the isolates) collected from Florida and gave an evaluation to each of the genotypes. The genotypes were grouped into four categories based on symptoms and the ability of the plants to continue growth: sensitive (severe chlorosis on leaves, greatly reduced growth, and eventual death); moderately tolerant (some scattered symptoms but little or no growth reduction and no p lant death); tolerant (very minimal symptoms); and variable reaction (mixed reactions as described above showed on different shoots). Furthermore, they also observed similar bacterial population s in most citrus genotypes which showed substantially differen t reactions to Las infection. This demonstrated that there was no str ong correlation between bacterial population and host response. 1.2.2 Alternative H osts M paniculata is a preferred host of D. citri as confirmed by observations and laboratory research; however, its alternative host status of HLB associated bacteria was not clear (Hung et al. 2000; Kunta et al. 2010; Lopes et al. 2010; Walter et al. 2012 ). Hung et al. (2000) reported that Las would not multiply in M. paniculata or M. koenigii while anot her report found consistent symptoms on inoculated M. paniculata plants (Halbert and Manjunath 2004). More recently, from their controlled inoculation experiments with two isolates of Las using D. citri as vector, Damsteegt et al. (2010) concluded that M. paniculata is variable as a reservoir host of the HLB associated pathogen. Because the bacterial population in M paniculata become s extremely low after 5 months, M. paniculata (as well as another Murraya species, M. exotica ) could only serve as a bridging host if citrus are present during that period of time. The most critical epidemiological role played by M. paniculata growing near HLB af fected citrus would be to harbor large number of D. citri. The field survey conducted by Walter et al.
35 (2012) found th at Las incidence in ornamental M. paniculata and associated psyllids ( D. citri ) was extremely low (1.8% and 1% for plants and psyllids, respectively). Some hosts outside the Rutaceae family can be experimentally inoculat ed with Ca L. spp and they are used in various HLB research. For example, dodder ( Cuscuta sp p. in Cuscutaceae family) can be colonized by Las and Lam and the bacteria can multiply inside dodder to a high level T he bacteria are unevenly distributed in dodder as in citrus regardless the hi ghly reduced and simplified anatomy of the plant (Hartung et al. 2010 a ). Dodder can be used to transmit HLB associated pathogens to citrus (Zhou et al. 2007; Zhang et al. 2011 b ) and non Rutaceous pants such as periwinkle ( Catharanthus roseus L. G. Don, in Apocynaceae family) (Garnier and Bove 1983 ; Hartung et al. 2009 ) and several s olanaceous plants like tomato (Duan et al. 2008) and tobacco ( Nicotiana tobacum L. cv. Xanthii ) (Garnier and Bove 1993), which indicates that Las has a wide physiological host range. 1.3 HLB D etection 1.3.1 HLB D iagnosis Other than the phytoplasmas found in HLB samples, the l iberibacters are the bacterial pathogen s most commonly found associated with HLB. Because liberibacters have not been cultured HLB lab diagnosis has tota lly relied on DNA based molecular methods. F ield survey is difficult and inaccurate because HLB symptoms are often confused with those of disorders such as Zn 2+ deficiency ; besides, the latency period of HLB in diseased trees makes early d etection impossib le. In the laboratory, before molecular approaches such as PCR were u s ed, various lab techniques were u tilized but end ed up being of limited use (Wang et al. 2006; Bove 2006; Halbert and Manjunath 2004) E lectron microscopy wa s the first technology used to determine the bacterial
36 nature of the pathogen, but as a diagnostic assay, it is laborious, costly, and require s appropriate sample preparation. Biological indexing by grafting buds from the suspected plants onto susceptible citrus cultivars is direct and convincing, but it is time consuming and requires a large set of test plants. Serological methods have not been successfu l l y used in HLB detection due to the difficulties in raising specific antibodies to an uncultivated organism, and the variation of ant igen traits of the target pathogen. Molecular d iagnosis also has been a problem due to the uncultured HLB associated pathogen s Villechanoux et al. (1992) first isolated total DNA from periwinkle plants inoculated with Ca L. spp and then digested it wit h the restriction enzyme HindIII. After cloning and differential hybridization, they identified three clones with inserts of 2.6, 1.9, and 0.6 k b specific to the HLB associated bacteria (no reaction with healthy tissue). After sequencing, it was found that the 2.6 k b clone contained part of the gene cluster nusG rplKAJL rpoBC which confirmed the eubacterial nature of the HLB associated pathogen at the molecular level (Villechanoux et al. 1993). After the bacterial nature of HLB associated pathogen was esta blished, universal primers for general amplification of prokaryotic 16S rDNA were used with HLB samples, and specific primers were developed and an 1160 bp region of ribosomal DNA was obtained from both Asian and African type s of HLB samples (Jagoueix et a l. 1996). Further differentiation of the two types of HLB could be achieved by using restriction enzyme Xbal which yield s two fragments of 640 bp and 520 bp for Las and three fragments of 520 bp, 506 bp, and 130 bp for Laf B ased on sequences from the clo ned gene cluster of ribosomal protei n and 16S rDNA s everal pairs of PCR primers or hybridization probes have been developed ( Hocquellet et al. 1997; Hocquellet et a l. 1999; Hung et al.
37 1999; Hung et al. 2004; Teixeira et al. 2005 b ), which greatly helped r educe th e uncertainty in HLB diagnosis However, the resulting amplicon s were quite long and d id not always result in stable, sensitive PCR reactions (Wang et al. 2006) A further improvement was made when real time PCR (qPCR) wa s applied to HLB diagnosis The intrinsic superior properties of qPCR, such as high sensitivity, high specificity, time efficien cy and less contamination, make it a great tool in HLB diagnosis, and it has quickly become the most popular technique for HLB diagnosis. Multiple qPCR s ystems for HLB detection report that they have increased sensitivities at least 10 to 100 time higher than corresponding conventional PCR (Wang et a l 2006; Li et al. 200 6 ; Teixeira et al. 2008 b; Lin et al. 2010 ), which is especially important because earl y diagnosis (even before the symptoms manifest ) is critical for HLB management T he test could be completed within a couple of hours. Also, with the help of a standard curve and the knowledge of the known copy number of the target DNA, qPCR can be used to estimate the bacteria population in plant or insect samples (Wang et al. 2006; Li et al. 200 6 ; Teixeira et al 2008b; Lopes et al 2009 a; Lopes et al. 2009 b ), or even to monitor the live bacterial genome inside host plants (Trivedi et al. 2009). qPCR has a ls o been used in HLB research to study the distribution of the pathogen in citrus plant s the in fluence of temperature on l iberibacters multiplication in plant tissues, and the acquisition and transmission rate of the pathogen by the psyllid vector (Kim et al. 2009; Pelz Stelinski et al. 2010) With all the molecular methods available for lab diagnosis, an inexpensive, easy to handle and accurate field diagnostic test is still in great need, and an iodine test was developed for this purpose (Takushi et al. 2007). Multiple research ers have show n that
38 there is an abnormally high level of starch accumulation in the parenchyma cells of leaves from HLB affected trees, and the test is based on the fast reaction of iodine and starch in water which can produce a da rkish solution. In Florida, Chamberlain and Irey (2008) found that the iodine test and qPCR tests agreed for 76% of the samples, and it produced better results with blotchy mottle leaves and from samples collected in February and July. They found the iodin e test useful for field diagnosis but still not a substitute for PCR in HLB diagnosis. Based on the starch accumulation in HLB affected leaves, Sankaran et al. (2010) developed a technique of using mid infrared spectroscopy for HLB detection. 1.3.2 L ive vs. D ead B acteria T he knowledge about the populations of pathogens in hosts has critical value to help make decisions in disease management (Nocker et al. 2006) DNA based methods including qPCR have a major limitation in that they can not differentiate be tween live and dead cells After cell death, bacterial genomic DNA can persist from days to 3 weeks in host (Josephson et al. 1993; Masters et al. 1994), so DNA based diagnostics tend to overestimate the live bacterial population. In the case of HLB, sever al studies found evidence from qPCR that liberibacter is present in hosts at a relatively high population (Wa ng et al. 2006; Li et al. 2006) H owever, the bacteria l population observed with a microscope is usually far lower than the number estimat ed by qPC R (personal communication with Dr. Michael Davis in CREC, FL) Considering that direct counting us ing microscopes only focuses on bacteria with cell structure while DNA based assays detect all DNA in the sample, the discrepancy between data from the two me thods is therefore understandable In order to obtain an accurate estimate of the live bacterial
39 genome in host by DNA based molecular methods (including qPCR), an effective way to remove background noise (i.e. naked DNA and DNA from dead cells) is needed Membrane integrity is the most important criterion for distinguishing live cells from dead (or irreversibly damaged) ones ( Nocker et al. 2006 ). Live cells with intact membranes have the ability to exclude DNA binding dyes which can easily penetrate into dead or membrane compromised cells. This general rule is us ed by most practices that aim at discrimination of live dead cells Propidium iodide (PI) is a highly membrane impermeant DNA dye, and is gener ally excluded from live cells T herefore, PI has been extensively used in various research using microscop y and flow cytometry (Nebe von Caron et al. 2000; Nebe von Caron et al. 1998 ) to identify dead cells in mixed populations. After PI penetrates into dead cells, it binds to DNA by intercalating between th e bases with no sequence preference at the ratio of one dye molecule per 4 5 base pairs of DNA (Nocker et al. 2006). Ethidium monoazide (EMA) is another DNA binding dye with the azide group, which allows the dye to covalently bind to DNA upon exposure to b right visible light (maximum absorbance at 460 nm). In the sample treatment process, cells are incubated with EMA for 5 min which allows the penetration of EMA into cell s with compromised cell walls/membranes and the binding of EMA with DNA; then in the fo llowing photolysis step in which samples receive 2 min of bright light exposure, the dye inside the cells forms a covalent cross link with DNA, while free EMA (remaining in solution) is simultaneously inactivated by reacting with water molecules and is no longer capable of covalently binding to DNA. Photo induced cross linking was reported to inhibit PCR amplification of DNA from dead cells A nother study showed that the EMA cross linkage
40 with DNA also rendered the DNA insoluble and cross linked DNA was lo st together with cell debris during genomic DNA extraction (Nocker and Camper 2006). Th us in a bacterial culture with a live dead mixed population, the DNA from the dead cells will be selectively removed during DNA extraction after EMA treatment and the EMA cross linked DNA in the remaining DNA extraction will not be detected by PCR. The double effects of EMA on DNA were considered to be a total DNA loss (with or without subsequent PCR). This feature of EMA has been utilized in combination with various PC R technologies to evaluate live bacterial genome s of many bacterial species: Escherichia coli O 157:H7 (Nogva et al. 2003), Salmonella typhimurium (Nogva et al. 2003), Listeria monocytogenes (Nogva et al. 2003; Rudi et al. 2005a ; Rudi et al. 2005b), Campylo bacter jejuni (Rudi et al. 2005a), and Las (Trivedi et al. 2009). However, this promising technique was found to suffer from a major drawback. In the case of E. coli O157:H7 (Gram negative as liberibacters), while EMA successfully removes all genomic DNA from dead cells, the treatment also removes about 60% of the genomic DNA of the live cells harvested fr om log phase (Nocker and Camper 2006). EMA was also observed to readily penetrate into live cells of other bacterial species and cause partial DNA loss i n several other studies (Nocker et al. 2006; Cawthorn and Witthuhn 2008 ; Flekna et al. 2007; Kobayashi et al. 2009a). In order to overcome th is limitation, a novel chemical, propidium monoazide (PMA) was introduced into this area by Nocker et al. (2006). PMA is identical to PI except that the additional azide group allows covalent cross linkage to DNA upon light exposure (like EMA). Compared to EMA, PMA binds with DNA molecules in the same way and can effectively remove background DNA when in contact H owe ver, PMA is highly selective in penetrating only into dead
41 cells with compromised cell membrane s but not into live cells with intact cell membranes/cell walls, which is probably due to the higher charge of the molecule (two positive charges compared to onl y one in the case of EMA) (Nocker et al. 2006). PMA has been confirmed for use in a wide range of bacteria including Gram negative a nd Gram positive cells (Nocker et al. 2006), and thus PMA has quickly replace d EMA in studying live bacterial genome s (Nock er et al. 2007b; Nocker et al. 2010; Contreras et al. 2011; Nocker and Camper 2009 ; Nocker et al. 2009 ). Although PMA in combination with real time PCR (PMA qPCR) has been used to study live population of various pathogenic bacteria (Ya nez et al. 2011; No cker et al. 2007a; Liang et al. 2011; Kobayashi et al. 2009b ; Kobayashi et al. 2010; Kralik et al. 2010 ), there is so far no report of PMA qPCR application on uncultured bacteria like the HLB associated liberibacters.
42 CHAPTER 2 PMA QPCR METHODOLOGY 2.1 I ntroduction Citrus Huanglongbing (HLB), also called citrus greening disease is one of the most devastating citrus diseases worldwide which causes severe crop losses and thus threats the citrus industry (Bove 2006; da Graca 1991). The disease is associated with a phloem restricted, Gram negative bacteri um which belong s to the genus Candidatus Liberibac ter in the alpha subdivision of the Proteobacteria (J agoueix et al. 1994). T hree species of the associated bacteria have been reported: Ca ndidatus L iberibacte r asiaticus (Las) in As ia and America continents (Bove 2006; Halbert 2005; Teixeira et al. 2005 b ); Ca L. africanus (Laf) in Africa (da G raca 1991; Planet et al. 1995; Garnier et al. 2000); and Ca L. americanus (Lam) in Brazil only (Teixeira et al. 2005 a ) T he bacterium has not been cultured yet, nor have Koch s postulates been satisfied. Las is the most prevalent HLB associated liberibac ter species in the world (Bove 2006) and also is the only one found in the U.S. since the discovery in Florida in 2005 (Halbert 2005). Besides graft transmission, Las is naturally vectored and transmitted by the Asian citrus psyllid (ACP) ( Diaphorina citri ) in the field (Bove 2006; da Graca 199 1; Halbert and Manjunath 2004 ), and it is widely accepted that HLB associated li beribacters can multiply and accumulate to a relatively high population inside the ins ect host (Halbert and Manjunath 2004; Duan et al. 2009). Due to the uncultured bacteria and the confusion of HLB symptoms with those of disorders such as Zn 2+ deficiency HLB diagnosis usually relies on DNA based methods like various polymerase chain reaction assay s (PCR) (Bove 2006). The disease can be diagnosed with electr on microscopy (Garnier and Bove 1996; Aubert et al. 1988), but
43 such assays are time consuming and r equire appropriate sample preparation since the bacteria is known to be unevenly distrib uted in planta (Tatineni et al. 2008; da Graca 1991; Bov e 2006). Reliable serological assays have not been developed largely due to the difficulties in raising specific antibodies to an uncultured organism, and the variation of antigen trai ts of the target bacteria (Bove 2006). With the discovery of partial sequences from cloned HLB associated liberibacters gene cluster of ribosomal protein and 16S rDNA, several pairs of conventional PCR primers have been developed and quickly bec a me the routine diagnostic assays for HLB (Jagoueix et al. 1996; Hocquellet et al. 1999 b ; Teixeira et al. 2005a). However, the amplicons in these conventional PCR tests are quite long and do not lend themselves to stable, sensitive PCR reactions which is quite a disadvantage because early diagnosis is usually required for HLB control and in addition, a restricted fragment length polymorphism (RFLP) test need s to be performed to differentiate Ca L spp Therefore, real time fluorescent PCR (qPCR) wa s quickly adopted to provide better sensitivity, specificity to any of the three HLB associated liberibacter species and time efficiency for HLB diagnostic purpose s (Wang et al. 2006; Li et al. 2006; Li et al. 2008). With its quantit ative feature, qPCR is often used in various HLB research programs as it is for other pathogens, for example, determining the virulence factors of the pathogen, and the infection ab ility of insect vector For pathogenic orga nisms, knowledge of their population s in the ir hosts is of great importance in the investigat ion of virulence mechanisms or decision mak ing in disease management. However, DNA based methods including qPCR have the major limitation that they can not differen tiate between live and dead cells, but only live cell s are of
44 interest for most research purposes. Since bacterial genomic DNA can persist up to 3 weeks aft er cell death (Josephson et al. 1993; Masters et al. 1994), DNA based diagnostics tend to overestima te live population s In the case of HLB, it has been speculated by qPCR that liberibacter is present in hosts at a rel atively high population (Wang et al. 2006; Li et al. 2006 ; Trivedi et al. 2009 ) H owever, the bacteria population observed with a microsco pe is usually lower by orders of magnitude (Teixeira et al. 2008c) Considering that direct counting using microscopes only focuses on bacteria with cell structure while DNA based assays detect all DNA in the sample, the discrepancy between data from the t wo types of methods is therefore understandable In order to minimize the difference and focus on the live bacterial genome an effective way to remove background DNA (including naked DNA or DNA from dead cells) is needed. Ethidium monoazide (EMA), is a D NA intercalating dye with an azide group that allow s covalent binding to DNA upon bright visible l ight exposure and thus render s the DNA insoluble during DNA extraction and therefore undetectable by PCR EMA was used in combination with qPCR to remove the DNA from dead cell s and thus only detect the DNA extracted from live cells in the con sequent qPCR (Nocker and Camper 2006 a ; Rudi et al. 2005 a ; Wang and Levin 2006; Trivedi et al., 2009). But poor selectivity of EMA, i.e. EMA can also penetrate into live c ells during the pretreatment step and thus cause the DNA loss from live cells render ed the chemical problematic when applied to various bacteria (Nocker et al. 2006; Cawthorn and Witthuhn 2008 ; Flekna et al. 2007; Kobayashi et al. 2009a). For example, in a study of E. coli O157:H7 (Gram negative as liberibacters), while EMA successfully remove d all genomic DNA from dead cells, the
45 treatment also result ed in a loss of about 60% of the genomic DNA of live cells harvested fr om the log phase (Nocker and Campe r 2006 a ). In order to overcome th is limitation, a novel chemical, propidium monoazide (PMA) was tested. PMA is similar to propidium iodide (a commonly used membrane impermeant dye in commercial live/dead bacteria staining kits), was used by Nocker et al. ( 2006). PMA binds with DNA molecules in the same way as EMA and thus effectively remove s background DNA H owever, PMA is highly selective in penetrating only into dead cells with compromised cell membrane s but not live cells with i ntact cell membranes/cell walls This is probably due to its higher charge of the molecule (two positive charges compared to only one in the case of EMA) (Nocker et al. 2006). This feature of PMA wa s confirmed in a wide range of bacteria including Gram negative a nd positive ones (N ocker et al. 2006), and thus PMA has replace d EMA to work with qPCR (i.e. qPCR with PMA pretreatment, termed PMA qPCR ) in studying live bacterial genome s ( Nocker et al. 2007b; Nocker et al. 2010; Contreras et al. 2011; Nocker and Camper 2009). Although PMA qPCR has been used to study live population of various pa thogenic bacteria (Yanez et al. 2011; Nocker et al. 2007a; Liang et al. 2011; Kobayashi e t al. 2009b), there is so far no report of PMA qPCR application on uncultured bacteria like the HLB assoc iated liberibacters. In this study, we redesigned and optimized protocols of PMA qPCR which w ere originally designed to work with pure bacterial culture s and compared the difference between PMA and EMA treatments when working with HLB associated liberibac ters The effect of TissueLyser treatment was studied in a separate experiment to demonstrate the insignificant impact on live bacterial genome s. With this optimized methodology, a standard curve between qPCR quantification results and
46 microscopic counting was established In addition live and cumulative Las genome s were studied for different HLB samples, such as a collection of citrus species and cultivars, and leaves with or without symptoms Considering the severe threat that HLB has posed on Florida an d the world citrus industry, the rapid and quantitative PCR methodology developed in our study should provide a more accurate way to determine the live liberibacter population in plants and psyllids, which in turn should benefit various research such as di sease epidemiology, and thus serve as a crucial component in HLB disease management. 2.2 Materials and M ethods 2.2.1 Plant and P syllid M aterials Las infected citrus species and other alternative host plants were sampled from our HLB positive plant colle ction kept in the greenhouse with natural light and controlled temperature at 75 78 F. For the study of live and cumulative bacterial genome in different host plants of Las leaves from at least 3 plants of each plant species or cultivar were used. The pla nts used in this study included: Benton citrange ( Citrus insitorum Mabb. 'Benton' ) Carrizo citrange ( C insitorum Mabb. Carrizo ), C indica Tanaka Kikojii, Kohorski, Mexican lime ( C. aurantifolia (Christm.) Swingle) Rough lemon ( C x jambhiri Lush Rough lemon) Volkamer Lemon ( C. limonia Osbeck Volkameriana ), Sour orange ( C. aurantium L.) Cleopatra mandarin ( C. reticulata Blanco), Duncan grapefruit ( C. paradisi MacFadyen) Flame grapefruit ( C. paradisi Flame ), Hamlin sweet orange ( C sinensis Hamlin ), Iapar sweet orange ( C sinensis Iapar ), Madam Vinous sweet orange ( C sinensis Madam Vinous ), Nav e l sweet orange ( C sinensis Bahia ), Pineapple sweet orange ( C sinensis Pineapple ), Valencia sweet orange ( C sinensis Valencia ) Calam ondin ( x Citrofortunella
47 microcarpa ), Severinia buxifolia Zanthoxylum fagara Murraya paniculata periwinkle ( Catharanthus roseus ), and dodder ( Cuscuta indecora ) Valencia sweet orange was used in the optimization process of PMA qPCR working protocol Hea lthy citrus grown from seeds and kept in a separate greenhouse were confirmed to be HLB negative by qPCR first and then used as healthy control in all qPCR reactions. Healthy ACP ( D. citri ) were collected from M paniculata ornamentals located in Winter (BioQuip Products Inc., Rancho Dominguez, CA) with multiple flushing plants including sweet orange ( C sinensis ) and M koenigii Las p ositive psyllids were collected from HLB positive citrus groves in Winter Haven, Florida, and released on HLB positive citrus plants which were kept in a specialized psyllid room with controlled temperature of 75 to 78 F and controlled light of 12 hour art ificial light exposure each day Psyllid colon ies w ere established on both sites and constantly monitored by qPCR H ealthy control psyllids for all the qPCR reactions were collected from th e healthy cage while Las positive psyllids from psyllid room were used in microscopic observation and comparison work with PMA qPCR. 2.2.2 PMA qPCR W orking P rotocol 18.104.22.168 PMA pretreatment Plant material. For plant materials, midribs and petioles from randomly collected leaves were used for total DNA extraction. After being chopped into small pieces (about 0.5mm long) with sterile razor blade s about 120 mg of plant tissue (more than the required 100 mg tissue to compensate any material loss during the process) was put into a 2 mL Eppendorf tube and treated by TissueLy ser II ( Qiagen, Valencia, CA ) with liquid nitrogen; the pulverized tissue powder was then taken out of the tube and put onto
48 a piece of wax paper where it was mixed again with a toothpick to make a homogenized tissue pool; two portions of 50 mg of tissue p owder from the pool were weighed out on a Mettler Toledo GmbH TYPE BD202 digital balance (Greifensee, Switzerland) and each portion was put into a new 1.5 mL Eppendorf tube and 1 mL of sterile distilled water (SDW) was added into the tube and mixed briefly by vortexing. After this step, two tubes containing the same amount of starting plant tissue (50 mg each) would be obtained out of one test sample and read y for the following different treatment s A stock solution of 10 mg/mL PMA ( Biothium Inc. Hayward, CA) was made with SDW. The tube was wrapped in aluminum foil due to it s photosensitivity and stored at 4 C. 2.5 L PMA stock solution was added into one of the two tubes in the dark to make a final concentration (f.c.) of 25 g/mL (about 50 M of PMA) and then mixed briefly by inverting the tube (PMA treatment was for live bacteria data, therefore, for cumulative bacteria data, no PMA was added into the other tube in this step, but the physical treatments such as vortexing and incubation were the same f or both tubes); the tubes were then incubated at room temperature in the dark for 5 minu tes with occasional inverti on A fter incubation, the tubes were briefly vortexed for 5 seconds and pu t into crushed ice with their lids off and exposed to light from a halogen bulb (650W) for 2 minutes at a distance o f 20 cm from the light source T he tubes were then centrifuged in a n Eppendorf Centrifuge 5415D (Brinkmann Instruments Inc., Westbury, NY) at 13,200 rpm (= 16,100 g) for 5 minutes; and 800 L of the supernat ant was pipetted out and discarded, and the pellet with the remaining liquid was used to extract total DNA with DNeasy Plant Mini Kit (Qiagen) following the manufacturer s instructions with a minor
49 modification in the last step where only 50 L of elution buffer AE was use d to obtain the final DNA sample. Psyllid material. For psyllid material, a single psyllid was put into a 1.5 mL Eppendorf tube with 100 L SDW and ground with a pestle to homogenize the sample. Fifty microliters of the sample was transfer red into a new 1.5 mL Eppendorf tube with another 50 L SDW added to make a total volume of 100 L. As described for plant material, 2 tubes with the same amount of psyllid material in 100 L suspension were obtained for each psyllid sample tested. The two tubes were vortexed briefly and then PMA w as added into one tube in the dark to make a final concentration of 25 g/mL (as for live bacteria data, while for cumulative bacteria l genome data, no PMA was added in this step) T he tubes were incubated at room temperature in the dark for 5 minutes with occasional inverting; then after a brief vortex (5 seconds), the tubes were set into crushed ice with their lids off and exposed to light from a halogen bulb (650W) for 2 minutes at a distance of 20 cm from the l ight source T he tubes were briefly vortexed again and the whole suspension (about 100 L) was used for total DNA extraction with DNeasy Blood & Tissue Kit (Qiagen) following the manufacturer s instructions with the minor modification as described above After the total DNA was extracted, t he yield and purity of the DNA samples were determin ed with a NanoDrop Spectrophotometer ND 1000 (Wilmington, DE), and all DNA samples were stored at 20 C. 22.214.171.124 qPCR In this study, citrus material was used in the optimization process of the PMA qPCR working protocol, while in the part where microscopic work was involved, psyllid material was preferred considering its high bacterial population and low host tissue
50 background (Duan et al. 2009). Based on the different materials used in this study, two previously published qPCR systems for Las were employed For plant materia l, a duplex TaqMan qPCR system developed by Li et al. (2006) was utilized. In this duplex qPCR, two primer probe sets of HLBaspr (designed based on 16S rDNA of Las genome) and COXfpr (designed based on citrus mitochondrial cytochrome oxi dase gene and used as positive internal control to assess the quality of the DNA extracts and reaction mixtures) were us ed (Table 3 1) However, it was previously reported that multiple factors, such as host tissue and host species, may have an effec t on P CR amplification efficiencies due to the difference among remaining PCR inhibitors in t he DNA samples after extraction T herefore, an improved standard curve from Li et al. (2008) was used in data analysis in this study. For psyllid material, another duple x TaqMan qPCR system including the 16S rDNA primer/probe set for Las (Li et al. 2006) and a primer/probe set for internal control of psyllid material (based on wingless gene coding a glycoprotein of psyllid species) (Manjunath et al. 2008) was used (Table 3 1) Both qPCR reactions were performed on an Applied Biosystems 7500 Fast Real Time PCR System (Life Technologies, Carlsbad, CA), and shared the same protocol as follows: 2 min incubation at 50 C followed by 10 min incubation at 95 C, then 40 cycles of 15 s at 95 C and 1 min at 60 C, and fluorescent signals were collected at the 1 min at 60 C stage of each cycle The 20 L qPCR reaction mixture contained 10 L of 2x ABI TaqMan Universal PCR Master Mix ( Life Technologies ), 2 L of DNA template, appro priate amount of primer/probe stock (to reach the optimized final concentrations as reported ) and Nuclease Free W ater ( Qiagen ). All reactions were performed in tri plicate and each run contained one negative one positive and one
51 healthy control. qPCR data were analyzed with the Applied Biosystems software (Version 1.4.0.) Statistical analysis of the qPCR data in this entire dissertation was done with SAS 9.3 TS Level 1M0, and ANOVA with Least Significant Difference (LSD) was used at the significant level of P<0.05. 2.2.3 Optimization of the PMA qPCR W orking P rotocol The working protocol which was originally designed for bacterial culture s (Trivedi et al. 2009; Nocker et al. 2006) was used as a prototype for PMA qPCR work in this study, but later several m ajor modifications were made due to unsatisfying results obtained with the original protocol The differences (or improvements) before or after the modifications were demonstrated with parallel experiments, in which only the modifi ed step was different whi le the rest of the protocol remained the same. 126.96.36.199 With or without TissueLyser (TL) treatment Starting from a raw sample pool of 350 mg chopped citrus midrib/petiole tissue, 3 portions of 50 mg tissue were first weighed out and put into 3 individual 1.5 mL Eppendorf tubes T he 3 tubes then went directly to the DNA extraction steps with DNeasy Plant Mini Kit ( Qiagen) and the samples extracted in this way were marked as without TL Then, the rest of the chopped materials in the raw pool (about 200 mg tissue left) were processed by TissueLyser II (Qiagen) with liquid nitrogen, after which, 3 portions of 50 mg tissue were weighed out from the pulverized tissue pool and put into 3 tubes and used for DNA extraction; the second batch of 3 DNA extracts were marked as with TL The same parallel experiment was repeated 3 times with different tissue samples, and the Cq values from these DNA extracts were compared.
52 188.8.131.52 Centrifugation speed 16,100 g vs. 5,000 g For plant materials, in the last centrifugatio n step of PMA pretreatment (right before DNA extraction ) the difference between centrifugation at 5,000 g and 16,100 g was compared by doing the following: starting from a pulverized sample pool, one portion of 50 mg tissue was weighed out and used direct ly fo r DNA extraction and a nother portion of 50 mg went through the PMA pretreatment process but only without PMA added (as for cumulative bacteria genome data) and centrifuged at 5,000 g for 5 minutes before DNA extraction T he same process as above was applied to another pulverized sample pool with t he last centrifugation step at 16,100 g The same process was repeated 3 times with different tissue samples. 184.108.40.206 Final concentration of PMA and EMA EMA powder ( Molecular Probes, Inc. Eugene, OR) was us ed to prepare a stock solution of 10 mg/mL in SDW in the dark The tube was wrapped in aluminum foil due to photosensitivity and stored at 4 C. Different final concentrations of PMA and EMA used in the pretreatment were compared in this study. Starting f rom a sample pool of pulverized citrus midrib/petiole tissue, portions of 50 mg of tissue were weighed out and put into 1.5 mL Eppendorf tubes with 1 mL SDW, which then received same processing steps as in the optimized workin g pro tocol to get DNA extract s which differed only in the amount of PMA (or EMA) added. One tube had no PMA (or EMA) added and DNA extract was marked as Cumulative T hree tubes had different amount s of PMA added (0.5 2.5 and 10 L of PMA 10 mg/mL stock solution to reach final concentration s of 5, 25, 100 g/mL, respectively ) and these DNA extracts were marked as PMA 5 g/mL PMA 25 g/mL and PMA 100 g/mL respectively Another three tubes had similar amounts of EMA and tho se DNA extracts were marked as EMA 5 g/mL EMA 25
53 g/mL and EMA 100 g/mL respectively. Each sample which received different amount s of PMA (or EMA) had three replicates from the same sample pool In the data analysis, a comparison was made between different DNA yield in percent which w as calculated by dividing bacterial genome tested from each DNA extract with PMA (or EMA) added by that tested from the Cumulative 2.2.4 PMA qPCR vs. R ods/field Due to the high bacterial concentrations present in some psyllid s and the low host tissue background (Duan et al. 2009), psyllids were often used in the microscopic work for HLB associated liberibacters observation T herefore, they were chosen to be the study material in this work to help correlate direc t counting numbers (by microscope) with quantitative PCR results (by PMA qPCR). Psyllids were collected from the psyllid room and immobilized by putting them in 20 C freezer for 15 minutes. For surface cleaning psyllids were first wash ed with 70% ethan ol once, and rinsed with SDW 15 to 20 times. The psyllids were dissected under a dissecting microscope and the midguts (i.e. alimentary canal) were removed and transferred into a 0.5 mL PCR tube with 30 L SDW (one midgut in each tube). After vortex mixing for 1 minute and a brief centrifugation the midguts were pipetted out along with 1 L of the liquid. PMA was added to the tube (f. c. = 25 g/mL) and incubated for 5 minutes in the dark. Four microlit ers of the buffer from the tube was pipetted onto a slide and stained with 1 L 25 M SYTO 13 Green Fluorescent Nucleic Acid Stain (Molecular Probes, Invitrogen, Eugene, OR ), and then covered with a cover slip and mounted for direct counting with an Axio Imager A1 Microscope (Carl Zeiss) using a green fluorescence F ilter Set 38 (excitation: 470/40 nm; emission: 525/50 nm) Photomicrographs were also taken with an AxioVision LE camera attached to the
54 microscope using the green fluorescence Filter Set 38 (fo r SYTO 13 excitation/emission maximum: 488/506 nm ) and the red fluorescence Filter Set 20 (for PMA excitation/emission maximum: 464/610 nm; parameters for Filter Set 20, excitation: 546/12, emission: 608/32 ), respectively The software used for visualiza tion was AxioVision LE, Release 4.6 (Carl Zeiss) The remaining liquid in the tube (about 25 L) followed the PMA qPCR protocol for psyllid samples as described above and was used for DNA extraction with DNeasy Blood & Tissue Kit (Qiagen) (25 L elution bu ffer AE was used in the last step to obtain the same volume of the starting liquid). For direct counting, a 10X/23 eyepiece and a 63X/1.25 Oil Iris objective lens were chosen as the working combination. Twenty fields were randomly screened and only typica l rod shaped liberibacter bacteria were counted towards a total number, based on which, a rods per field number was first calculated (total number/20) and then converted into rods per L liquid with the microscopic parameters. The logarithmic values of rods per L liquid data were then plotted against Cq values of each corresponding psyllid DNA extracts in qPCR, and linear regression was used to help find a possible correlation (i.e. standard curve). 2.2.5 Effect of TL T reatment on Live Bacterial Genome In order to test the effect of the pulverization process on the live bacterial genome obtained by PMA qPCR methodology developed in this study, the TissueLyser treatment was analyzed in a separate experiment. Considering the low temperature (caused by liquid nitrogen, LT ) and physical disruption (caused by intense pounding with metal beads, PD ) might have different effects on the bacterial genome tested the TissueLyser treatment w as dissected into two separ ate treatments, LT and PD. T he experiment design was as follows: about 16 mature leaves with typical blotchy mottle
55 symptoms were collected from a HLB positive citrus plant and a sample pool (about 1.6 gram in total) was made by chopping mid ribs and petioles into small pieces (about 0.5mm long) T he sample pool was divided into four equal portions (as four groups, 0.4 gram per group), and each group was then further divided into three equal portions (as three samples, 0.13 gram p er sample) and each sample was put into individual Eppendorf tubes E ach group ( three samples) received different treatments (1) No LT+PD ( untreated sample as negative control) (2) LT ( the sample tubes were placed in liquid nitrogen, no physical disruption with metal beads pounding) (3) PD ( samples in tubes received only physical disruption in TissueLyser with metal beads, no LT treatment prior to this) (4) With LT+PD (sample tube first soaked in liquid nitrogen then disrupted in TissueLyser with metal bead s, the regular TissueLyser treatment as in PMA pretreatment) A fter the different treatments, two portions of 0. 0 5 gram tissue were weighed out from each sample and p laced into two individual Eppendorf tubes (one tube add PMA, no PMA for the other one), wh ich then would receive the rest of the PMA pretreatment process as described in the PMA qPCR working procedure above. The whole process was repeated three times, and all the leaves were sampled from the same HLB positive citrus plant. 2.2.6 Determinatio n of Interference of Plant Tissue with PMA qPCR Xanthomonas citri ssp. citri strain 306 was used in this study. The bacterium was grown in nutrient broth on a shaker at 200 rpm and 28 C overnight to the log phase. Bacterial cells were collected by centrif ugation at 5,000 g for 5 min and re suspended in sterile distilled water (SDW). The bacterial cell suspension was adjusted to an optical density at 600 nm of 0.0 3 by dilution with SDW. The bacterial suspension was kill ed by exposure to isopropanol (final co ncentration, 70%) for 10 min and mixed with a fresh
56 culture suspension at the defined ratios of viable versus isopropanol killed as follows : 1000 to 0 900 to 100 500 to 50 0, 100 to 900 and 0 to 100 0 (Nocker and Camper 2006) ( See table in Figure 2 6 ) T he isopropanol killed bacterium was checked for viability by plating on nutrient agar, followed by overnight incubation at 28 C. Leaf samples collected from healthy citrus tree s were pulverized by TissueLyser with liquid nitrogen. Aliquots of 50 mg of tis sue powder from the same sample pool were weighed out and put into individual 1.5 ml Eppendorf tubes. One milliliter of the bacterial suspension ( of each mixed ratio ) was added to two tubes of plant tissue. One tube was subject to PMA treatment (as for liv e bacterial genome), and the other one followed the treatment only without PMA added (as for cumulative bacterial genome) PMA qPCR pretreatment was follow ed by D NA extraction and qPCR. Three replicates were included for each bacterial suspension. After qP CR, the numbers of live and dead cells were first calculated from the Cq values with the standard curve method (Cubero and Graham 2005) and the ratios were compared with the original mixed ratios The TaqMan qPCR system with a standard curve developed by Cubero and Graham (2005) was used. In this qPCR system, a set of primers (J RTpth3: 5 ACCGTCCCCTACTTCAACTCAA 3 and J RTpth4: 5 CGCACCTCGAACGATTGC 3 ) and a TaqMan probe (J Taqpth2: 5 FAM ATGCGCCCAGCCCAACGC TAMRA 3 ) were designed based on sequences of the pth gene, a major virulence gene used specifically to detect strains of citrus bacterial canker (Cubero and Graham 2002; Mavrodieva et al. 2004). Another set of TaqMan primer/probe, COXfpr (Li et al. 2006), was also included in this qPCR as internal co ntrol for plant material The qPCR reaction w as performed on an Applied Biosystems 7500 Fast Real Time PCR System (Life
57 Technologies, Carlsbad, CA), and the PCR protocol was as follows: 2 min incubation at 50 C followed by 10 min incubation at 95 C, then 40 cycles of 15 s at 95 C and 1 min at 60 C, and fluorescent signals were collected at the 1 min at 60 C stage of each cycle The 2 5 L qPCR reaction mixture contained 1 2.5 L of 2x ABI TaqMan Universal PCR Master Mix ( Life Technologies ), 2 L of DNA template, 1 L of 10 M forward/reverse primers (J RTpth3/ J RTpth4), 0.5 L of 10 M TaqMan probe (J Taqpth2), and Nuclease Free W ater ( Qiagen ). All reactions were performed in tri plicate and each run contained one negative one positive and one healthy control. qPCR values were analyzed with the Applied Biosystems software (Version 1.4.0.) Statistical analysis of the qPCR data was done with SAS 9.3 TS Level 1M0, and ANOVA with Least Significant Difference (LSD) was used at the significant level of P < 0 .05. 2.2. 7 Determination of the Live Bacterial Genome s in S ymptomatic and A symptomatic Leaves from Las Positive Field Trees New mature leaves with or without symptoms (100 each) were collected from Las positive Hamlin sweet orange trees ( C sinensis Ham lin ) located at CREC, Lake Alfred (Figure 2 7A). The 200 leaves from symptomatic and asymptomatic group were tested individually with PMA qPCR to determine the live bacterial genome inside each leaf, and to assess if the live bacterial genome was differen t between the two groups. The data distribution of both leaf groups was presented in the Box and whisker plot, and the statistical analysis was done with ANOVA at the significant level of P < 0.05.
58 2.3 Results 2.3.1 Optimization of PMA qPCR W orking P rotoco l 220.127.116.11 With or without TL treatment The difference with or without th e TL pulverization step was tested by parallel experiments and the results are shown in Fig ure 2 1. Basically, starting f rom the same material, a higher bacterial genome in the sample could be detected with TL treatment (shown as lower Cq values compared to the ones without TL), which could be caused by more bacteria released due to smaller tissue particles. M ore importantly, a much smaller variation occurred among the three tubes that received TL treatment (shown as smaller standard deviation) which demonstrated greater homogeneity in the sample pool with TL treatment 18.104.22.168 Centrifugation speed 16,100 g vs. 5,000 g In all three repeats, lower centrifugation speed (i.e. 5,000 g) us ually caused a loss of DNA (as much as 68%) with the supernatant discarded after this step ( data not shown ). However, a much higher centrifugation speed (i.e. 16,100 g) greatly reduce d the loss and therefore worked better in the PMA pretreatment protocol where we wanted to obtain as much DNA as possible 22.214.171.124 Final concentration of PMA and EMA A series of different final concentrations of PMA and EMA 5, 25 and 100 g/mL were tested and compared in order to find the best working concentration for th is research. The evaluation was based on DNA yield in percent (See Fig ure 2 2) The was calculated from Cq values of EMA (or PMA ) treated samples divided by the untreated sample which gave a general DNA removal efficiency after EMA (or PMA) treatment. As shown in Fig ure 2 2, starting with the same tissue sample pool, the
59 DNA removed by EMA treatment was usually more than that removed by PMA at the same concentration More DNA was remo ved as EMA concentrations increased H owever, DNA removal b y PMA did not increase when dye concentration exceeded 25 g/mL Therefore, the f.c. of 25 g/mL PMA was chosen in the working protocol. 2.3.2 PMA qPCR versus D irect C ounting with M icroscope After an optimized PMA qPCR working protocol was established, t he correlation between PCR based molecular quantification and microscopic direct counting was estimated as described above For direct counting ( s ee Fig ure 2 4 A for live bacteria observed with green fluorescence filter set 38 ) rods/field data were first c onverted to rods/ L using parameters from the microscope, and then the common logarithm ic values of the rods/ L data were plotted a gainst Cq values from qPCR (Fig ure 2 3). A standard curve by linear regression showed a good fit of the data, and the R 2 value of 0.9476 indi cated a good accuracy over a relatively wide range of Cq values (from 26 to 34). With the standard curve in Fig ure 2 3, an estimation of the live Las genome or even how many live liberibacter rods could be observed under the microscope (a linear conversio n process needed) can be made based on a Cq va lue obtained by PMA qPCR, or vi c e versa. 2.3. 3 Effects of TissueLyser T reatment on L ive B acterial Genome The four different treatments g a ve significantly different estimates of Las genome when used on samples from the same plant. C ompared to untreated samples, low temperature treatment (using liquid nitrogen as in With LT and With LT+PD ) did not cause any significant change (or loss) in the resulting genome but physical disruption alone (using TissueLyser machine without liquid nitrogen With PD ) caused a significant rise in Cq value, which meant a significant loss in the live bacterial genome
60 detected. At the same time, samples that received the two treatments separately ( With LT or With PD ) res ulted in significantly lower live bacterial genome than the ones with two treatments together (i.e. the regular TissueLyser treatment as in With LT+PD ) F or the cumulative bacterial genome compared to samples with no treatment, each treatment alone ( W ith LT or With PD ) did not have significant impact on the cumulative genome detected, while combined treatments resulted in a significantly higher cumulative genome In Figure 2 5, most of the groups (four different treatments in live and cumulative ) ha d considerable variations within the data collected from 9 samples (3 samples from each of the three repeats), while only the live bacterial genome obtained with regular TL treatment had the smallest variations (0.56 compared with 2.8 in live bacterial gen ome from untreated samples). 2.3.4 Determination of Interference of Plant Tissue with PMA qPCR After exposure to isopropanol for 10 min, the viability check by plating on nutrient agar showed no colonies after 3 three days incubation The isopropanol kill ed cells were mixed with the untreated original culture in defined ratios (see table in Figure 2 6), and then 1 mL of the mixed bacterial suspension was added to 50 mg of pulverized plant tissue and the mixture was treated following the PMA qPCR protocol, which resulted in two Cq values representing live and cumulative bacterial genomes, respectively. The DNA yield in percent was calculated from the two Cq values, and it was compared with the defined ratios. With an increase of dead cells in the bacterial suspension, the Cq values of the live bacterial genome also increased (representing a decrease in detectable live bacterial genomes), while the cumulative bacterial genome ( from samples without PMA ) remained unchanged. As a result with three replicates, the DNA yield in percent data agreed well with the defined ratios (Figure 2 6). However,
61 the internal control of COX showed significant decrease of that detectable genome after PMA treatment (data not shown), which was not observed in the case of Las inf ected plant samples. 2.3.5 Live B acteria in S ymptomatic or A symptomatic HLB L eaf S amples S ymptomatic and asymptomatic sweet orange leaves collected from field all showed a wide range of live bacterial genomes usually ranging over more than 4 orders of m agnitudes (Figure 2 7B ) However, the live bacterial genome detected in individual symptomatic leaves clustered at a significantly higher level (Group mean: 2.22 x 10 6 Las bacteria per gram tissue; and 95% confidence interval: 1.49 3.30 x 10 6 Las bacteri a per gram tissue ) than the cluster found in asympt o matic leaves (Group mean: 5.44 x 10 4 Las bacteria per gram tissue; and 95% confidence interval: 3.41 8.69 x 10 4 Las bacteria per gram tissue) ( P <0.0001) (Figure 2 7B). In addition, Las was not detected i n some asymptomatic leaves (6 samples undetected out of 100 leaves tested), which never happened in symptomatic leaves. 2.3. 6 PMA qPCR on D ifferent H ost Plants Whether or not different host plants which showed different levels of resistan ce to HLB disease also had different levels of live and cumulative bacterial genome was checked with 24 citrus and non citrus plants infected with Las A s shown in Figure 2 8 a wide range of live bacterial genomes were obtained from most of the hosts tested, which made th e difference among different citrus varieties insignificant, and this was also observed in cumulative bacterial genomes (data not shown). However, some significant differences still existed, for example, dodder ( Cuscuta indecora ) had a significantly higher amount of live Las genome than the rest, while M. paniculata and Z.
62 fagara had significantly lower amount of live Las genome compared to most citrus genotypes ( P <0.0001). 2.4 Discussion 2.4.1 Optimization of PMA qPCR Unlike PMA EMA was reported to be pro blematic when used to selectively remove DNA of dead cells (Nocker et al. 2006; Cawthorn and Witthuhn 2008 ; Flekna et al. 2007; Kobayashi et al. 2009a ). EMA could pass through intact cell membrane s of live cells of many different bacteria species tested, a nd cause DNA loss in DNA extraction and subsequent PCR assays, while no staining effect on live bacteria was observed with PMA (Nocker et al. 2006). Therefore, PMA was chosen to work with qPCR in this study to gain a better understanding of live bacterial genome s in HLB af fected hosts. However, when published EMA qPCR or PMA qPCR working protocols (Trivedi et al. 2009; Nocker et al. 2006) were directly utiliz ed in this study, some major problems occurred which led to inconsistent results or even total fail ure of the experiments. For example, a large variation of Cq values was observed among repeated ly sampled portions from the same sample pool which rendered results unreli able A fter PMA pretreatment Las was sometimes not detected in some previously positi ve plant samples which made the data tot ally useless. T he previous working protocol was originally designed for cultur able bacteria and therefore had fewer variables to control as compared with uncultured Las. N o protocol was available for working with p syllid material Therefore, optimization of the previous working protocol s was necessary. In order to get a more accurate estimat e of live Las genome in the hosts, a homogeneous starting material is quite important. For plant materials, finely c hopped midrib and petiole tissue is good enough for a positive or negative result however,
63 whether the roughly treated material could meet the needs of qPCR or PMA qPCR was in question. D ue to the fact s that Ca L. spp are inside the plant phloem tissue ( da Graca 1991; Bove 2006), and that they are unevenly distributed in planta (Tatineni et al. 2008; Bove 2006), a severe disruption protocol was needed to minimize these effects. Therefore, a pulverization step using TissueLyser and liquid nitrogen was adde d to reduce the tissue particles size and make a more uniform sample pool. The results not only showed that pulverized tissue had a smaller deviation among portions from the same pool, but also gave a higher DNA extraction efficiency demonstrated by a gene ral ly lower Cq value from pulverized tissue compared to the same sample without pulverization (Figure 2 1) The increased DNA extraction efficiency could help minimize experimental error thus is an additional value of the TissueLyser step. After the Tissu e L yser treatment was routinely used in PMA work, Cq values increased, indicating fewer live bacteria l genome s as a portion of the corresponding cumulative bacteria genome Live bacteria l genome were sometimes not detectable. Therefore, the live bacterial p ercentages (LBP) were usually less than 10% which were much lower than th e 17 31% reported in citrus with EMA ( Trivedi et al. 2009). Besides, when the cumulative bacteria l genome data from PMA treatment was compared with data from the no PMA pretreatment s ample there was a big difference as well. These findings ruled out the possibility that the DNA removal effect of PMA was the cause of the bacteria genome change, but rather a total DNA loss by the PMA pretreatment process The discarded supernatant in th e centrifugation step was the only possible step that could account for the loss, and later, a considerable amount of DNA material including Las DNA was detected in the discarded supernatant (data not shown). The
64 centrifugation speed of 16, 1 00 g was trie d in an effort to reduce the DNA loss. The data showed that the higher speed reduce d the loss to a minimum level, and in turn, LBP results started to become reasonable (ranging from less than 1% to near 100%). In EMA or PMA work published previously (Trive di et al. 2009; Nocker et al. 2006; Nocker and Camper 2006a) different final concentrations were used for different target bacteria or specific application s which made it uncertain which concentration would work for Las For example, EMA was used at the f.c. of 240 M (about 100 g/mL ) in studies by Rudi et al. ( 2004 ), Trivedi et al. ( 2009 ), and Nocker and Camper ( 2006 a ) and at 50 M (about 25 g/mL ) in another study by Nocker et al. ( 2006 ) PMA was mostly used at 50 M and 5 M (Nocker et al., 2006; Liang et al., 2 011) A ll of the published EMA or PMA work dealt with bacterial pure culture s rather than uncultured ones like Las in this study Since these concentrations might not work the same with Las they were compared in this study. As shown in Fig ure 2 2, with in crease d chemical concentration s in all three replicates PMA reach ed a s aturated stage in which added chemical did not cause further DNA reduction, while EMA had no such phenome non. Another fact worth pointing out was that sample s treated with same amount of PMA or EMA, the former one usually had a higher DNA yield than the latter (Figure 2 2) Th e results confirmed the fact that EMA may have substantial DNA removal effect on live bacteria in suspension i.e. low selectivity of EMA compared with PMA (Nocker et al. 2006), and thus is not suitable for this work Therefore, the final concentration of 25 g/mL was chosen for PMA qPCR detection of Las This concentration was considered sufficient to effectively remove genomic DNA from dead cells while avoiding potential DNA loss due to excessive dye
65 2.4.2 PMA qPCR vs. R ods/field With an effective and relia ble PMA qPCR protocol and proper experimental material (midguts dissected from psyllid s ), a linear relationship between microscopically observable liberibacter rods and quantitative PCR results was established in this work With this standard curve, the li ve Las population inside hosts (plants or psyllid) c an be estimat ed with data obtained from either of the two methodologies. Because microscopic observation of Las only count s cells with rod shaped cell structure, while regular qPCR (without PMA pretreatme nt) detect s the whole bacterial genome no matter from live or dead cells there i s always a considerable discrepancy between the genome estimat e s made by these two methodologies ( Li et al. 2006 ; Trivedi et al., 2009). PMA can effectively remove background DNA (DNA from dead cells and naked DNA molecules) from the cumulative genome and make only DNA of live cells detec table in subsequent qPCR assays T herefore, this additional treatment made qPCR results theoretically comparable to what you could actually se e under microscopes and thus a possible correlation standard curve. When converting qPCR results to actual bacterial genome s, standard curves developed with the PCR target on a recombinant plasmid are a common option (Li et al. 2008; Trivedi et al. 2009; Wang et al. 2006) However, such standard curves were problematic when used for field samples in HLB research, and optimizations for them were constantly being made, for example, a major modification was made for standard curves by Li et al. (2008) so that the factors of host tissue, host species and even geographic locations could be taken into account In another study, the host tissue was proved not to be a significantly factor in their standard curve (Trivedi et al., 2009) T he differen ce in performance of recombinant plasmid and bacterial genomic DNA in qPCR
66 assays remained unstudied and thus left the application of such a standard curve in question. When counting Las from psyllids with a microscope, there was a general impression that live bacteria wer e much more prevalent than dead ones in psyllid s (see Fig ure 2 4A vs. 2 4B), however, the LBP in psyllids obtained by PMA qPCR were usually just around 20% (data not shown). This discrepancy could possibly be due to the naked DNA in the background which c ould not be seen with either dye under the microscope but still detectable in qPCR (Nocker and Camper 2009). 2. 4 3 Effects of TissueLyser T reatment on L ive B acterial Genome PMA qPCR was developed to estimate the live bacterial genome inside hosts. H owe ver, certain process in the PMA protocol might have adverse effects on the integrity of the bacterial cell structures, which could render the data inaccurate and produc e an underestimate of the actual live population For example, the extreme ly low tempera ture in liquid nitrogen treatment could cause the bacterial cells to freeze and thaw. Also no research data support s the assumption that the intense pounding by metal beads would not harm the bacterium inside the tissues. Although in long term storage usin g liquid nitrogen, bacteria experience extreme temperature change but still remain viable. H owever, that practice usually is done with pure bacterial culture s and has to follow certain procedures (like using glycerol to protect the membrane from freeze dam age) Therefore, investigation into this matter is necessary. T he TissueLyser treatment in the PMA pretreatment not only had no significant impact on the live bacterial genome inside plant tissue, but also tended to increase the estimate of the cumulative bacterial genome In addition, the liquid nitrogen treatment alone had no significant impact on either live or cumulative bacterial genome However
67 the pounding with steel beads seemed to have an adverse effect on live bacterial genome ( fewer live bacteria detected), but not on the cumulative The possible adverse effect of liquid nitrogen treatment on live population was ruled out. The smallest variations of live bacterial genome detected in the samples With LT+PD demonstrated the reliability of PMA qPC R in fulfilling its purpose. Besides, the big variations of cumulative genome in all the 4 groups including With LT+PD implied that cumulative bacterial genome is not a suitable target to be used for monitoring the dynamic change of the bacterial populat ion in HLB research. 2.4.4 Determination of Interference of Plant Tissue with PMA qPCR PMA qPCR has been repeatedly reported to work well with culturable bacteria including various plant pathoge n s ( Yanez et al. 2011; Nocker et al. 2007a; Liang et al. 20 11; Kobayashi et al. 2009b ; Kobayashi et al. 2010; Kralik et al. 2010), but there is no report of PMA qPCR application on uncultured bacteria like the HLB associated liberibacters In order to demonstrate the effectiveness of the optimized PMA qPCR protoco l with plant samples, the experiment with the mixture of plant tissue and bacterial suspensions of defined live/dead cell ratios was used to mimic the situation in Las infected plant samples. Half of the bacterial suspension from fresh culture of log phase was subject to exposure to 70% isopropanol for 10 min, resulting in a decrease in culturable bacterial cell s to zero. With the increase of dead cells in the mixed bacterial suspension, the live bacterial genome s dete c t ed by qPCR after PMA treatment decrea sed as expected, and the DNA yield in percent data agreed well with the defined live/dead bacterial ratios. This showed that PMA qPCR was not interfered with by plant tissue Regard ing the significantly decrease of COX detected after PMA treat ment it co uld be caused by certain interaction between plant tissue and high
68 concentration of bacteria used. Another possibility is that there is certain interaction between the two primer/probe sets, which has never been studied for these two specific sets before. 2. 4 .5 Live B acteria in S ymptomatic and A symptomatic Leaves In order to determine the correlation between symptom expression and the bacterial population, an experiment was conducted using each single leaf as a sample, rather than at least three leaves as in regular Las detection. With a sample size of 100 leaves in each group (symptomatic and asymptomatic) individual leaves with or without symptoms both showed a wide range of live bacterial populations, and their distribution in the bacterial population o verlapped with each other (Figure 2 7B). T he symptomatic leaves had a significantly higher level of live bacterial population than the asymptomatic ones ( P <0.0001), which indicated a substantial difference of bacterial populations regarding symptom express ion. Based on the 95% confidence intervals of the means of both groups (i.e., 95% confidence interval of the mean in symptomatic group: 1.49 3.30 x 10 6 Las bacteria per gram tissue; 95% confidence interval of the mean in asymptomatic group: 3.41 8.69 x 10 4 Las bacteria per gram tissue), the intermediate region of the bacterial population, from 8.69 x 10 4 to 1.49 x 10 6 Las bacteria per gram tissue, could be considered as the bacterial population range with which Las infected Hamlin sweet orange leaves sta rted showing symptoms, i.e., the threshold region of internal bacterial population for HLB symptom expression in Hamlin sweet orange. Another way to interpret this threshold region is that Hamlin leaves with live bacterial population below 8.69 x 10 4 Las b acteria per gram tissue are more likely to be asymptomatic, while leaves with more live bacterial population than 1.49 x 10 6 Las
69 bacteria per gram tissue are more likely to show HLB symptoms already (95% confidence level). Regarding the 6 asymptomatic lea ves with undetectable Las, it could be due to the new growth of the plant had not been colonized by the bacteria yet, and this also confirmed that symptomatic leaves should be the first choice for Las detection. 2.4. 6 Live Las in Different Hosts According to the results shown in Figure 2 8 different citrus species and cultivars were all able to accumulate high amount of live Las genome regardless of their susceptibility to HLB disease. In general, no significant difference in live bacterial genome was obse rved between a s usceptible plant species (or cultivar) and a more resistant one. This suggests that there is no strong correlation between citrus susceptibility to HLB and the internal bacterial populations. Folimonova et al. (2009) reached a similar concl usion based on qPCR tests of their plant collection, although it s cumulative Las genome results. The data of live Las genome obtained in this study was not significant ly differen t from the cumulative genome (data not shown) However, if considering all ci trus varieties as a group the citrus relatives or HLB related plants (like dodder, periwinkle) included in this study showed significant difference s with this group S. buxifolia and Calamondin showed no difference with Citrus spp. as hosts of Las which was not unexpected due to their excellent alternative host status found in this research (discussed in the following chapters) In contrast, M. paniculata and Z. fagara had significantly lower Las genome (live and cumulative ) compared to citrus. This means that Las could not accumulate to a high level in these two plants, and may further indicate an unsuitable living environment inside these two plants for the bacterium Las genome was significantly higher in dodder than in citrus, which helps to
70 explain wh y dodder serve s as an excellent host and transmission vector for Las ( Hartung et al. 2010 a; Zhou et al. 2007; Zhang et al. 2011 b; Duan et al. 2008 )
71 Figure 2 1 Comparison between with TissueLyser treatment (TL) and without TL in PMA pretreatment Aliq uots from the same plant sample were treated with or without TL, and tested with qPCR after DNA extraction. No PMA treatment was included in this experiment. Error bars represented standard deviations from 3 replicates and the experiment was repeated thre e times
72 Figure 2 2. The effect of different concentrations of PMA or EMA on DNA yield s Sample aliquots were made from the same sample pool, and treated with P MA (or E MA) at different final concentrations. Live bacterial genome data from P MA (or E MA ) treated samples were divided by the cumulative genome data from untreated samples, and the resulting DNA yields (in percent) were compared. Error bars represented standard deviations from 3 replicates.
73 Figure 2 3. Correlation between common log va lue of bacterial concentration from microscopic counting and Cq values obtained by PMA qPCR Psyllid samples were first dissected, and the suspension of midguts were observed with microscope as well as tested by qPCR after treated with both SYTO 13 and PMA The common log values of bacterial concentrations from microscopic observation were plotted against the corresponding Cq values obtained from qPCR, and the linear regression resulted in a standard curve as above. Error bars represented standard deviation s from at least 3 replicates T he standard curve could be further converted to show the common log value of bacterial cells per gram plant tissue versus Cq values, which is y = 0.2819x + 13.576 (y = common log value of bacterial cells per gram tissue, x = Cq), if this specific data format was needed for certain comparative purpose.
74 Figure 2 4. Las observed under fluorescent microscope stained with both SYTO 13 and PMA SYTO 13 could stain all bacterial cells including live and dead ones, while PM A could only stain dead ones. Different filter sets had different emission/excitation wavelength ranges thus with specific filter set choice, live and dead cells could be observed separately. A) Green Filter Set 38 for SYTO 13 view of live cells and B) R ed Filter Set 20 for PMA view of dead cells A B 5 m 5 m
75 Figure 2 5. Effects of TissueLyser treatment on live and cumulative Las genome tested by PMA qPCR Aliquots from the same sample pool received different treatments, i.e., with either low temperature (LT, tr eated with liquid nitrogen) or physical disruption (PD, treated with metal beads pounding), with both LT+PD, or without LT+PD. After the 4 different treatments, all aliquots of samples were tested following the same PMA qPCR protocol, and the resulting Cq values of both live and cumulative genome were compared. Error bars represented standard deviations from 9 replicates of each sample (3 replicates from 3 repeated experiments). Statistical analysis was done with SAS 9.3 TS Level 1M0 at the significant leve l of P <0.05. bc ab a c A A A B No LT+PD
76 Figure 2 6 PMA qPCR with defined ratios of live and isopropanol killed Xanthomonas citri ssp. citri The table show s the ratios of live and isopropanol killed bacteria The n umbers represent volumes in microliters. One mill iliter of mixed bacterial suspension was mixed with 50 mg of plant tissue to mimic the situation of Las infected plant samples. After PMA qPCR procedure, the DNA yield (in percent) was calculated from the Cq value s obtained from PMA treated samples divided by those from no PMA added samples Error bars represented the standard deviations from 3 replicates.
77 Figure 2 7 Comparison of the l ive bacterial genome s of symptomatic and asymptomatic leaves of Hamlin sweet orange A) HLB s ymptomatic and a sympto matic leaves Leaves were tested individually by PMA qPCR to check the liv e bacterial genome in each leaf. B) Box and Whisker plot showing the data distribution of the 100 leaves ( 100 w ith and 100 without symptoms) tested. The upper and lower bar of each g roup show s the maximum and minimum values, excluding outliers (dots) The box contain s the middle 50% of the data (inter quartile range), which was from the 25 th percentile (lower edge) to the 75 th percentile (upper edge). The statistical analysis was done with SAS 9.3 TS Level 1M0 at the significant level of P <0.05. Sym ptomatic As ymptomatic A B a b
78 Figure 2 8 Comparison of live Las genome in 24 different host plants Plants of 1 to 18 were different citrus varieties and species. Plants of 19 to 24 were citrus relatives or HLB related plants. Error bars represented standard deviations among 3 to 8 replicates of each plant tested. 1. Navel sweet orange ( C itrus x sinensis 2. Benton citrange ( C insitorum Mabb. 'Benton') 3. Carrizo citrange ( C. insitorum Mabb. 'Carr izo') 4. Kohorski 5. Sour orange ( C. aurantium L.) 6. Mexican lime ( C. aurantifolia (Christm.) Swingle), 7. Cleopatra mandarin ( C. reticulata Blanco) 8. Rough lemon ( C x jambhiri Lush. Rough lemon) 9. C. indica Tanaka 10. Volkamer Lemon ( C. limonia 11. Duncan grapefruit ( C. paradisi MacFadyen), 12. Valencia sweet orange ( C x sinensis 13. Kikojii 14. Flame grapefruit ( C. paradisi 15. Iapar sweet orange ( C x sinensis 16. Pineapple sweet orange ( C x sinensis 17. Madam Vinous sweet orange ( C x sinensis 18. Hamlin sweet orange ( C x sinensis 19. Calamondin ( x Citrofortunella microcarpa ), 20. Severinia buxifolia 21. Zanthoxylum fagara 22. Murraya panicu lata 23. periwinkle ( Catharanthus roseus ), 24. dodder ( Cuscuta indecora )
79 CHAPTER 3 S EVERINIA BUXIFOLIA AS AN EXCE LLENT ALTERNATIVE HO ST OF LAS AND A SEASONAL LIVE BACTERIAL DEVEL OPMENT SHOWN BY PMA QPCR 3.1 Introduction Citrus huanglongbing (HLB) is a devastating citrus disease worldwide. It can cause premature fruit drop and eventually kill the infected trees. It affects all the major citrus producing areas in the world, such as the U.S., China, and Brazil, and causes substantial economic loss. Most importantly, there is no effective cure for this disease yet. There are three types of HLB ( Bove 2006; Zhang et al. 2011 ), and they are associated with three phloem restricted bacterial agents, Candidatus Liberibacter asiaticus (Las), Ca L. africanus (La f), and Ca L. americanus (Lam). These fastidious bacteria have not been obtained in pure culture yet, nor has Koch s postulates been fulfilled for HLB, which poses a major obstacle in HLB research. Although HLB is graft transmissible, it is natural ly tran smitted in the field by the insect vector, citrus psyllid There are two types of citrus psyllids responsible for HLB transmission, one is Asian citrus psyllid (ACP) Diaphorina citri which is mainly distributed in Asia and the Americas ; the other one is African citrus psyllid, Trioza erytreae which is in Africa. Even though ACP was discovered in 1998 in Florida, it was not until 2005 that citrus HLB was first reported in south Florida ( Halbert 2005 ), which was also the first report of HLB in the United States. After that, HLB quickly spread to all the citrus producing counties in Florida. The only bacterial agent found in Florida is Las It was reported that, since 2005, HLB has cost the S tate of Florida more than $3.6 billion and more than 6,600 part ti me and full time jobs. Considering that the major citrus cultivars in Florida are sweet oranges and grapefruit, which are all susceptible to the disease
80 (Folimonova et al. 2009; Halbert and Manjunath 2004) HLB has posed great threat to the citrus industry in Florida. With the rapid spread of HLB in Florida, alternative hosts of the associated bacterium need more study due to their potential to serve as se c ondary sources of the bacterium Alternative hosts are concealed havens for pathogen survival, but th ey are often neglected in epidemiological researches especially when they are not recognized (Hung et al. 2001) Therefore, t he knowledge about alternative hosts is important for the integrated pest management of HLB (Hung et al. 2000; Halbert and Manjunat h 2004). There are many citrus relatives in the Rutaceae family present in Florida that are grown mainly as ornamentals Some of the citrus relatives are listed as hosts of the HLB associated bacterial agents and/or the psyllid vectors based on field surve ys and PCR results However, their status as alternative hosts of HLB has never been systematically studied except for two studies on Severinia buxifolia (Hung et al. 2001) and Murraya paniculata (Damsteegt et al. 2010), respectively Hung et al. (2001) re ported that S. buxifolia is an alternative host of Las because it can be infected by Las through grafting and psyllid transmission. However, the study on transmission was not complete (i.e., the psyllid transmission pathway from Las positive S. buxifolia t o healthy ones was not included), nor the repeats were sufficient. Besides, the transmissibility of S. buxifolia to citrus was not studied. It was reported that M. paniculata supports good psyllid populations, but Las bacterium does not to persist long in it, usually becoming undetectable after 5 months, which makes them minimally important as hosts of Las (Damsteegt et al. 2010).
81 Research was initiated to study the potential of several rutaceous plant species as hosts for Las. E ight different plant speci es studied were Severinia buxifolia Calamondin ( x Citrofortunella microcarpa ), Citropsis gilletiana Choisya ternata Sundance Choisya aztec Pearl Zanthoxylum fagara Esenbeckia runyonii and Amyris texana Transmission trials were set up in greenhouse s (most trials were done by psyllid transmission; graft transmissions were also set up when plants are compatible with citrus material), and the plants were later tested with real time PCR (qPCR) T he living situation of the bacteri um inside alternative ho sts (even all the hosts including psyllids), and whether the passage through alternative hosts will a ffect th e pathogenicity of the bacteria are both important aspects for evaluation of their host status. T herefore, the new methodology to detect the live b acterial genome inside hosts developed in this study (see Chapter 2) was used to monitor the dynamic changes of live bacterial genome inside all the hosts found in this work. Besides, since it s the first time report of their host status of Las for most of the plants in this work, the identities of the bacterial agents were double checked with conventional PCR followed by sequencing for further confirmation. This study showed that seven plant species can be infected by Las out of which six are first time r eported to be alternative host s of Las A mong those alternative hosts identified in this work, the bacteria persistence and psyllid activities associated with the plants were quite different, which ma de their transmissibility back to citrus differ. Based o n the transmissibility back to citrus, the plants are grouped into four categories: high, medium high, medium, and no transmissibility S buxifolia from the high transmissibility group is a n excellent alternative host of Las and can also support psyllid vector well
82 Besides, the work of monitoring live bacterial genome dynamics in this plant is the most completed one therefore, this part of work will be discussed in a separate chapter, while the other plants will be discussed all together in the next cha pter. 3.2 Materials and Methods 3.2.1 Plants and P syllids Seedlings of S. buxifolia were grown from seeds collected at the University of Florida, Citrus Research and Education Center (CREC), Lake Alfred, Florida. The source plants were assumed and later q PCR confirmed by the 16S system ( Li et al. 2006) ( see Table 3 1 ) to be Las free Valencia sweet orange ( C itrus sinensis Valencia ) seedlings were grown from seeds purchased from California Citrus and Severinia seedlings were grown in an insect free green house w ith natural temperature and daylight. Healthy seedlings of S. buxifolia and citrus were usually kept in this greenhouse for 2 to 3 months after transplanting, and moved to a second greenhouse ( referred to as the ) with controlled tempe rature of 75 to 78 F and controlled light of 12 hour artificial light exposure each day. A ll the psyllid transmission experiments were set up in the psyllid room and kept for the first 2 months post inoculation (PI). After all inoculated plants were qPCR t ested for the first time and the remaining psyllids were remov ed the plants were moved to the third greenhouse with controlled temperature of 75 to 78 F and natural light for long term monitor and observation due to space limit in the psyllid room. Health y ACP were collected from M paniculata ornamentals located in Winter (BioQuip Products Inc., Rancho Dominguez, CA) with multiple flushing plants including sweet orange ( C sinen sis ) and M koenigii A
83 p syllid colony was established in the healthy cage with supplementations from the same location in Winter Haven. Psyllid and plants kept in this cage were constantly monitored by qPCR to en sure a Las free condition. Besides, all plants and psyllids used in this study were tested for latent Las infection by qPCR prior to use in any experiment. 3.2.2 Inoculation P rocedure Because S. buxifolia is graft compatible with citru s graft transmission was also done The budding method was used, and t wo S. buxifolia plants were grafted with multiple bud s (usually mature buds with wood attached) collected from HLB positive citrus plant (with strong HLB symptoms like blotchy mottle leaves) Four months af ter grafting leaves from the S. buxifolia plants were collected and tested with qPCR to confirm the infection. Since p syllid transmission is the primary means of HLB spread in the field (Halbert and Manjunath 2004; Bove, 2006; da Graca, 1991), it was t he major inoculation method u sed in this study. In all the three psyllid transmission pathways investigated in this study, (1) from citrus to S. buxifolia (2) from S. buxifolia to citrus, (3) from S. buxifolia to new S. buxifolia a standard inoculation p rocedure was followed Las free psyllids ( D. citri ) were collected from healthy cage and colonized on source plants (for example, for pathway 1, from citrus to S. buxifolia Las free psyllids colonized Las infected sweet orange plants, while for pathway 2 and 3, Las infected S. buxifolia plants as source plants were colonized by psyllids ) for a 3 week acquisition access period (AAP) (Damsteegt et al. 2010), during which psyllids were mating and reproducing, but only adults were used for inoculations for t heir better transmission efficiency ( Pelz Stelinski et al. 2010) W hen new flushes were available on receptor plants for psyllid feeding, check plants for any pests and treat with appropriate methods
84 to remove them if any presented A fter visible pests rem oved, wash plants thoroughly with tap water and let the plants sit till dry; c ollect psyllids (10 psyllids as a set) from for inoculation and wait at least 14 days o f inoculation access period (IAP) (Damsteegt et al. 2010) at the same time, 10 psyllids (1 set) were qPCR assayed individually to obtain Las positive percentage data for each transmission experiment H ealthy receptor plants were mock inoculated with healt hy psyllids as negative controls, while healthy sweet orange plants inoculated with psyllids from source plants served as positive controls with each experimental inoculation. For each transmission pathway, using psyllids collected from the same cage of so urce plants, 3 batches of receptor plants (2 to 12 plants in each batch depending on the plant and psyllid availability ) were used to set up 3 separate transmission experiments which s erv e d as three repeats of each pathway to validate the results For tra nsmission pathway 1, a total of 12 S. buxifolia plants were inoculated over a period of 9 months, for pathway 2, 28 Valencia sweet orange ( C sinensis Valencia ) plants were inoculated over a 5 month period, and for pathway 3, 16 S. buxifolia plants used over a 22 month period. Positive and negative controls were set up for each of the 3 transmission experiment repeats, respectively. As mentioned above, following inoculation, plants were allowed to grow in the psyllid room for the first 2 months PI afte r which inoculated plants were cleaned with appropriate treatments and then moved to the third greenhouse for long term test ing and observation At the same time, psyllids left on the receptor plants were all collected and 1 set of 10 psyllids was tested f or Las positive percentage. At 2 months PI leaves from the receptor plants were first collected for qPCR test, and then tested monthly until
85 positive results were revealed. Negative plants w ere monitored for one year, after which the plants were discarded Because the information about the live bacteria l genome inside hosts is important for evaluation of the host status and one of our research objectives wa s to know if passage through alternative host s will affect the pathogenicity of the bacterium, the n ew ly developed PMA qPCR method was used to monitor the live and the cumulative bacterial genome in selected plants from transmission pathways 2 and 3. This monitor ing lasted more than 20 months for most of the plants in the Severinia work 3.2.3 Detection A ssays All plants and psyllid materials used in this study were assayed by qPCR for Las presence and/or development. For plant material DNA samples were extracted from newly matured leaves of each inoculated plant using the DNeasy Plant Mini Kit (Qiagen, ally 3 leaves for citrus plant s and 8 leaves for S. buxifolia ) were randomly selected and aseptically removed from each plant; leaf lamina were discarded and midrib/petiole materia ls were finely chopped with sterile razor blades; 100 mg of chopped tissue were weighed out and put into a 2 mL microcentrifuge tube and pulverized using TissueLyser II (Qiagen) with liquid nitrogen, after which the tissue w as ready for DNA extraction with the Plant Kit (Qiagen). For psyllid s the DNeasy Blood & Tissue Kit (Qiagen) was used for DNA extraction. Briefly, a single psyllid (or a group of psyllids depending on the purpose of the assay) was placed into a 1.5 mL microcentrifuge tube with 180 L AT L buffer (tissue lysis buffer from the DNeasy Blood & Tissue Kit) and gr oun d with a small plastic pestle, and then the
86 and purity of the DNA samples w as estimated with a NanoDrop Spec trophotometer ND 1000 (Wilmington, DE), and all DNA samples were stored at 20 C. Due to the uncultured nature of the associated agent, although the 16S qPCR system previously published by Li et al. (2006) was used as the primary method to detect Las pres ence and to monitor Las development throughout this study, several other PCR detection systems, real time or conventional, were also includ ed to further confirm the identity of the transmitted agent (Table 3 1). I nternal controls (IC) were included to asse ss the quality of the DNA extracts and the reaction mixtures F or plant samples, the IC was a plant cytochrome oxidase (COX) ba sed primer/probe set (Li et al. 2006), and for psyllid samples, a psyllid glycoprotein gene ( WG ) was used (Manjunath et al. 2008a ) All primers and probes were synthesized by Integrated DNA Technologies Inc. (Coralville, IA). All qPCR reactions were performed on an Applied Biosystems 7500 Fast Real Time PCR System (Life Technologies, Carlsbad, CA), and all the qPCR reactions shared the same protocol as follows: 2 min incubation at 50 C foll owed by 10 min incubation at 95 C, then 40 cycles of 15 s at 95 C and 1 min at 60C, and fluorescent signals were collected at the 1 min at 60C stage of each cycle. Data were analyzed with the App lied Biosystems software Version 1.4.0. A 20 L qPCR reaction mixture contained 10 L of 2x ABI TaqMan Universal PCR Master Mix (Life Technologies), 2 L of DNA template, appropriate amount of primer/probe stock (to reach the optimized final concentration s as listed in Table 3 1) and Nuclease Free Water (Qiagen). In order to confirm the results from qPCR reactions, qPCR reactions were run again in the conventional PCR system (only without probes while PCR reaction mixture
87 and the PCR protocol remained the same) to check for actual bands under the UV light (Table 3 1). For conventional PCR, all reactions were run on a n S1000 Thermal Cyc ler (Bio Rad, Hercules, CA) The PCR protocol for T7 1/Sp6 1 was different due to it s bigger PCR product size ( 2588 bp ), wh ich was as follows: first 3 min incubation at 95 C, then 40 cycles of 45 s at 94 C, 90 s at 58 C and 2 min at 72 C, and a final extension of 10 min at 72 C. A 50 L PCR reaction mixture contained 25 L of 2x GoTaq Green Master Mix (Promega, Madison, WI), 6 L of DNA template, and appropriate amount of primer stock and Nuclease Free Water (Qiagen). Amplified PCR products were run using 1% agarose gel stained with GelRed (Biotium, Hay ward, CA), and visualized with Alpha Imager EP (AlphaInnotech, San Lea ndro, CA) After visualization of the bands of expected size, all PCR products ( about 100 bp) were sent for sequencing at the Interdisciplinary Center for Biotechnology Research (ICBR), University of Florida, Gainesville, FL. The 2588 bp PCR product w as fi rst cloned and then sequenced at ICBR due to its size All sequence data were blasted online at National Center for Biotechnology Information (NCBI) website (http://blast.ncbi.nlm.nih.gov/Blast.cgi). After receptor plants were confirmed positive for Las infection, some of them were chosen to monitor the live population dynamics of Las inside them over time. In order to check if the passage through S. buxifolia will affect the pathogenicity of the bacteria, the long term monitor ing was mainly conducted in receptor plants from transmission pathway 2 and 3. For transmission pathway 2, three citrus seedlings were randomly picked from each repeat, and tested by PMA qPCR every other month (later increased to every month for more data). For transmission pathway 3 due to early plant death the
88 20 month monitor ing was only accomplished with 3 positive S. buxifolia seedlings from the 2nd repeat PMA qPCR work followed the 2.2.2 PMA qPCR working protocol in Chapter 2. 3.3 Results 3.3.1 Transmission E xperiments Tw o S. buxifolia plants were initially graft transmitted with buds obtained from HLB af fected citrus and then kept in psyllid room for more than two years, and the infect ed plants w ere confirmed with qPCR. When t he two Severinia plants (referred as Sb1 and S b2 ) had plenty of new shoots available, healthy psyllids were released in to the cage to feed on them. Psyllids adapted to the new plants without any obvious adapt ive phase, and colonies built up on Severinia almost as quickly as on citrus or Murraya panicu lata plants. Honeydew, eggs and nymphs, all typical si gns of psyllid activities (Bove 2006), could be easily found on S. buxifolia plants usually one or two weeks later After several months of feeding and multiplying (longer than the required 3 week AAP), 10 psyllids were randomly collected and tested individually with qPCR for evaluation every couple of weeks. When the majority of the psyllid population (usually >70%) all acquired high levels of Las, with quantitative cycle (Cq) v alues ranging from 16.9 t o 30.0 they were used in transmission experiments w ith Severinia plants as source plants (i.e. pathway 2 and 3). However, in the 3rd repeat of pathway 3, a couple of S. buxifolia plants that were inoculat ed via psyllid transmission in this study were use d as source plants to raise new batches of infected psyllids, so that this type of transmission, from psyllid transmitted Severinia plants to healthy Severinia via psyllids, was also done.
89 126.96.36.199 From citrus to S. buxifolia transmission pathway 1 The natural transmission via psyllid vector s which naturally occur s in the field was test ed by three repeated experiments in this study. With psyllids raised on infected R ough lemon plants, 3, 5, and 4 Severinia seedlings were inoculated in three separate tri als. In the 1st qPCR test 2 month PI, only 33%, 20 % and 25 % of inoculated plants from each of the repeat, respectively, were positive, while in the 2nd test of 5 month PI, all the seedlings were positive for Las (Cq ranging from 2 6 to 39, see Table 3 2). T his demonstrated a high transmissibility of Las from citrus to S. buxifolia via the psyllid vector. D ue to some fungal infection in these trials, only a couple psyllids were collected before plants were cleaned and moved to another greenhouse and only one of them were qPCR positive with a Cq value close to 36 (data not shown). Valencia sweet orange positive controls which were inoculated with psyllids collected from the same Rough Lemon plants had Cq values of around 30s when first tested 2 month PI, and l ater reach ed the low 20s; mock inoculated Severinia seedlings with Las free psyllids showed high Cq values (close to 40) sometimes wi th 16S primer/probe (Li et al. 2006) but totally undetermined Cq values by rpoB primer/probe (data not shown). 188.8.131.52 Fro m S. buxifolia to citrus transmission pathway 2 In this pathway psyllids raised on positive Severinia plants were used to inoculate 3 groups of Valencia sweet orange plants separately (28 plants in total, see Table 3 3 ). Psyllids collected from the posi tive Severinia plants usually had a high percentage of positive results (70% 100%) with Cq values ranging from 16 to 30. However, p syllids collected from receptor citrus plants after inoculation had a considerable variability in Cq value levels, ranging fr om low 20s to high 30s and about 80 % with undetermined values (data not shown). Unlike Severinia 9 out the 13 positive citrus plants revealed their
90 infections in the 1st qPCR test (2 month PI) and all 13 plants (out of 28 total inoculated plants) show ed i nfection in the 2nd test (5 month PI) and infected citrus plants quickly reach a stable stage of high levels of Las population represented by constantly low Cq values. For citrus plants inoculated in this pathway, the stable stage usually started from the 5 month PI ( some late ones started from 12 month PI) and stayed on the Cq level of 18~2 7 for at least 1 2 months in this study (Table 3 3 ). Valencia sweet orange plants which were mock inoculated with healthy psyllids as negative controls were all qPCR neg ative throughout the whole monitoring period, and showed healthy looking growth especially with much bigger plants, which was the same as unsuccessfully inoculated ( i.e., escape d ) Valencia plants in these trials. 184.108.40.206 From S. buxifolia to S. buxifo lia transmission pathway 3 Three groups of Severinia plants (2, 9, and 5 plants in each group, respectively, 16 in total) were used to demonstrate the transmissibility from infected S. buxifolia to healthy ones via psyllids. The first two groups were ino culated with psyllids collected from graft transmitted Sb1 and Sb2, while the 3rd group of 5 plants was inoculated with psyllids from psyllid transmitted Severinia plants (plant 3 1 to 3 5 in Table 3 4). Like the Severinia plants in pathway 1, only a small part of the successfully inoculated plants were qPCR positive in the 1st qPCR test of 2 month PI, and all showed positive in the 2nd test of 5 month PI with Cq values ranging from the low 20s to 30s. Variable qPCR results (ranging from 16 to high 30s and about 70 % with undetermined values) of the psyllids collected from the inoculated Severinia plants after 2 month period were also foun d (data not shown). Among all the positive plants obtained in this pathway, most r each ed a stable stage of high bacteria p opulation with minor fluctuation s which was similar to what citrus plants showed after inoculation, however, a couple of plants (plant
91 1 1 and 1 2 in Table 3 4) showed an uncommon fluctuation in Cq values (from low 20s to undetectable Las population ) For Severinia plants inoculated in this pathway, the stable stage usually started from 5 month PI ( some late ones started from 10 month PI) and lasted for at least 14 months ( Cq values ranging from 2 1 to 28 ) in this study (Table 3 4). Valencia plants inoculat ed with psyllids collected from the same Severinia plants had Cq values of high 30s when first tested 2 month PI, and later w ould reach the low 20s; mock inoculated Severinia seedlings with Las free psyllids remained qPCR negative throughout the entire mon itoring period. For the two types of transmissions conduct ed in this p athway of which graft transmitted Severinia plants (i.e. Sb1 and Sb2 in the 1st and 2nd repeats) and psyllid transmitted ones (i.e. 2 3 and 2 6 from 2nd repeat used for the 3rd repeat ) were used as source plants, respectively, no obvious difference w as observed based on the data collected. Besides, the 3rd repeat also demonstrated that Las transmission through S. buxifolia could be successful for at least 3 generations. 3.3.2 Las I dent ity C onfirmation Due to the uncultured fact of the associated bacterial agent of HLB, a lot of efforts were put into this study trying to confirm the identity of the agent involved. First, several qPCR systems which targeted different regions of the bacte rial genome, for example, CQULA04F/R CQULA10P set based on 50S ribosomal subunit protein L10 ( rp lJ ) of Las (Wang, et al. 2006) and rpoBf/r/p set based on beta subunit of RNA polymerase ( rpoB ) of l iberibacters were us ed to retest the positive plant sample s from the psyllid transmission trials P ositive PCR results were obtained with CQULA04F/R CQULA10P, rpoBf/r/p, and Las spef/r/p systems while negative results with Lam spef/r/p and Laf spef/r/ p systems, as expected. Second qPCR systems plus
92 one conventi onal PCR system (Table 3 1) were run on the regular thermocycler and actual bands with expected amplicon sizes, for example, 75 bp for 16S and 111 bp for rpoB primer sets were visualized in gel (Figure 4 3B showed this results with all other alternative h osts) Third PCR products were purified and sent for sequenc ing while the 2588 bp product of the conventional PCR were first cloned and then sent for sequencing due to the big size of the PCR product (see Figure 4 3C). S equence data were compared online and confirmed to be Las sequence with more than 99% similarity ( See BLAST results in Appendix ). 3.3.3 Symptom O bservation As a part of the transmission pathway 1, Sb1 and Sb2 were kept under observation and were tested periodically for Las presence by q PCR The two plants both grew slowly with small leaves and corky veins, and some branches even showed blotchy mottle leaves (Figure 3 1A) the typical symptoms described for H LB af fected citrus plants (Bove 2006; da Graca 1991). However, symptoms observed on psyllid inoculated Severinia seedlings did not always agree with PCR results. For example, small leaves and leaf yellowing could be found on both PCR positive and negative plants, besides, small plants rarely had raised veins on their leaves like the tw o older Severinia plants and leaf mottling was hard to define on Severinia plants comparing to citrus plants. The only symptom observed in this study that was consistent with the infection status was the stunted plants, i.e. all infected Severinia plants showed significantly smaller plants than un infected ones including negative controls ( Figure 3 1B ). Citrus seedlings successfully inoculated in this work showed typical HLB symptoms like blotchy mottle, small leaves, and yellowing. Positive citrus seedl ings all
93 grew really slow and some die back occur red Stunted plants were observed in citrus plants as well (Figure 3 1C), which were 100% consistent with PCR results. 3.3.4 Dynamic C hange of L ive Las P opulation Since August 2010, leaf samples from positi ve S. buxifolia and citrus seedlings inoculated in this study were collected and tested with PMA qPCR in a monthly manner, and the mon itoring lasted for 20 month s The dynamic change of the live and cumulative Las genome w as monitored. For S buxifolia live Las began at a lower population level (represented by a higher level of Cq values ranging from 27 to 32) and this lower level lasted about 8 months until April, 2011, when the live bacterial genome had a sudden rise of more than 30 time s and stay ed on a significant higher level ( P <0.05 ) (represented by a lower level of Cq ranging from 23 to 26). However, in contrast to this seasonal development in the live bacterial genome, cumulative Las genome fluctuated from 20 to 26 throughout the entire 20 month o f monitored period and no obvious pattern w as observed (Figure 3 2A). There was only one month exception for this general pattern found in live bacterial genome development, January 2012, in which the live Las population (Cq = 28) dropped out of the range o f 23 to 26. Another situation noticeable in Figure 3 2A is that the standard deviations were generally much smaller for live bacterial data set than the cumulative ones, which also confirmed the results from the previous PMA qPCR methodology chapter. Fo r citrus a similar seasonal development pattern was also observed in the plants from the 3rd repeat (3 plants from each of the 3 transmission repeats were chosen for live bacterial genome monitor work) which was lower population level from August, 2010 t ill February, 2011 (Cq from 27 to 28) and a sudden rise of almost 60 time
94 occurr ed in April, 2011 and the months after that (Cq from 20 to 22). No obvious staged pattern was observed for cumulative Las in citrus either, only fluctuation from Cq values of 2 0 to 25 through the entire period. The exceptional month of January, 2012 also gave a Cq value rise for live Las in citrus (Figure 3 2B), which indicated the live bacterial genome drop might be a common inci dent for that particular month. Citrus plants fro m t he first two repeats for this work showed moderate fluctuation (within Cq values of 22 to 25) throughout the entire period from August, 2010 to June, 2011. 3.4 Discussion S buxifolia is a spiny, low branching, compact evergreen shrub related to Citr us Fortunella and Citropsis in Rutaceae family. It s common name is Box Orange or Boxthorn, which is native to southern China, and widely distributed in Asian countries such as India, Malaysia, Vietnam, the Philippines and Japan In the United States, it is planted in the entire state of Florida and coastal regions of states like California, Texas, and Louisiana. Boxthorn often appears in the vicinity of citrus orchards, and is attractive to citrus psyllids (Hung et al. 2001). With the fast spread of this devastating citrus disease in Florida, citrus relatives gained much attention largely due to the fact that they could potentially serve as alternative hosts of Las or an inoculum reservoir, which would pos e much complexity and uncertainty to the HLB mana gement program in Florida. Las was detected in S. buxifolia by PCR in a previous research ( Hung et al. 2000 ), but the status of S. buxifolia as an alternative host of Las has never been system at ically studied and a lot of questions concerning S. buxifolia and HLB remained unsolved, for example, how well is the bacterium survi ving inside S. buxifolia and will the passage through S. buxifolia
95 affect the pathogenicity of the bacterium or not? T his study was designed to answer those questions. 3.4.1 S. buxifol ia as A lternative H ost of Las In this study it s been demonstrated in repeated transmission experiments of multiple plants that Las could be easily transmitted by psyllid in all the transmission pathways involving S. buxifolia (i.e. from citrus to S. bux ifolia from S. buxifolia back to citrus, and from S. buxifolia to S. buxifolia ), and by graft ing from citrus to S. buxifolia The p syllid vector easi ly infest ed this host without any obvious adapt ive phase, and psyllid activities such as feeding and repro ducing were just as common as on citrus plants. Plus, psyllid raised on positive Severinia plants could easily acquire Las bacterium (70 100% of population with Las Cq values of 20s). In order to serve as an inoculum reservoir, a plant species should be a ble to retain the pathogen in an infectious state over an ext ended period of time (Damsteegt et al. 2010). The data showed that Las persist ed in S. buxifolia plants for more than 30 months, and the psyllid acquired bacteria from positive Severinia plants f rom as early as 5 month PI. Therefore, S. buxifolia i s an excellent alternative host and inoculum reservoir for citrus HLB. The risk that this plant pose s to HLB management is that it can support the psyllid vector well, which is harmful even when it is La s free, and is an inoculum reservoir of the pathogen Compared to orange jasmine ( M. paniculata ), S. buxifolia is definitely more of a problem in HLB management. Recently, Ramadugu et al. (2010) reported naturally infected S. buxifolia by Las in the field, which shows that S. buxifolia may be a threat if it is present in the vicinity of citrus groves. Orange jasmine supports good psyllid populations, but Las bacterium d oes not to persist long in it, usually becoming undetectable after 5 months (Damsteegt et al. 2010). The psyllid may still acquire
96 inoculum from Las infected orange jasmine during the short period s of time, but these short periods of infectivity probably makes them minimally important as hosts of Las (Damsteegt et al. 2010). E ven though it ha s been shown that S. buxifolia i s an excellent alternative host for both the bacterial agent and psyllid vector, some relatively l ow er transmission rate s were found in Severinia group, i.e. 13 out of 28 (46.4%) receptor citrus plants in the transmission p athway 2 trials were positive, which is the lowest one in all the back inoculation experiments ( see Chapter 4 for transmission rates of all other alternative hosts). Other than some random errors, such as the psyllids happened to be at low positive percent age with Las or they were inactive when used in inoculation experiments it was most probably due to the inoculation procedure used for the Severinia group. A set of 10 psyllids was used to inoculate individual plant s while at least 40 to 50 psyllids wer e used in Dr. Damsteegt s inoculation (2010). Another disagreement worth mentioning is that Folimonova et al. (2009) reported that S. buxifolia was highly tolerant to HLB because it showed little or no visual symptoms and plants continued to grow vigorous ly which was like uninoculated control plants. In this study, no leaf symptoms on S. buxifolia were consistent with PCR results (which made them all suspicious symptoms), however, the plant growth was inhibited seriously after Las infection, and stunted plants was the major symptom found in HLB affected S. buxifolia 3.4.2 Live Las P opulation in S. buxifolia and C itrus T he cumulative bacterial genome detected by regular qPCR is the sum of live bacteria, dead bacteria (with integrity compromised cell str ucture) and naked DNA of the target organism in the sample. The information of the cumulative bacterial genome
97 may fulfill some diagnostic purpose; however, the live bacterial genome is what is needed. For example, information on the live population of pat hogenic microorganisms in host s is widely used in various investigations of virulence mechanisms and to help make decisions in disease management practice, such as the timing of pesticide sprays Multiple studies have been published on bacterial population s in HLB af fected citrus hosts. Some focus ed on bacterial distribution in different tissue or parts of the host plant ( Tatineni et al. 2008 ), and some have tr ied to determine the dynamic change of the bacterial population over a certain period of time ( Hu et al. 2006 ) H owever, the value of the data is questionable since regular qPCR was used After PMA qPCR was developed in this study f or live bacteria detection, two major questions of interest, the amount of Las inside S. buxifolia ( or all the hosts) and the effect of the passage through S. buxifolia on Las pathogenicity, can be studied With the help of PMA qPCR, a different yet quite obvious seasonal pattern of bacterial development was discovered in both S. buxifolia and citrus host. The first stage wi th lower bacterial population (higher Cq values around 28) could be a phase of early adaptation and slo w development after inoculation ; however, the sudden rise of bacterial population for both plant species raised a question: what event or factor drove th is change? When Las genome sequence was published, it was revealed that there are 12 phage related genes inside the 1.23 Mb genome, indicating that a phage or proph age may be present (Duan et al. 2009). Later, it was reported that Las also carrie d an excis ion plasmid prophage and a chromosomally integrated prophage that bec a me lytic in plant infections (Zhang et al. 2011). These prophages were associated with the pathogenicity of Las. Therefore, one possible explanation for this sudden change is that
98 someth ing changed for these phage elements inside Las in April, 2011 For example, the lytic phage entered the lysogenic cycle, which led to a sudden change (i.e. increase) of the bacterial population Some molecular tests targeting the phage (or prophage) sequ ence in the rel e vant DNA samples may help answer this question However, before identify ing the real reason driving the change, the information revealed in this development pattern (or growth curve) co uld be useful in HLB management F or example, if a pest icide application is need ed twice a year for HLB treatment, knowing that a bacterial population outbreak would occur in April could indicate that psyllid control is needed b efore the bacterial population increase s. Regarding the first two repeats of citru s plants monitored for live bacterial genome dynamics, the moderate fluctuation throughout the period seems to be in consistent with the seasonal development pattern. However, the first two trial s were set up in summer 2009, while the 3 rd repeat of citrus a nd the only repeat of S. buxifolia ( from which the seasonal pattern was observed) were both set up in Spring 2010. Therefore, the live bacterial genome in the first two repeats of citrus may already reach the high level when the monitoring work started in August, 2010, which actually confirm ed the seasonal pattern. As discussed above, the cumulative bacterial genome may be less informative due to the mixed nature A s imilar situation found in this study when dealing with cumulative bacterial genome : no obvi ous pattern was found in either S. buxifolia or citrus host. Hu et al. (2006) reported that Las population should reach the highest level from October to December in citrus samples collected from field, while the specific period showed no such feature in t his study (Figure 3 2B) This disagreement among different research on
99 the cumulative bacterial genome raised a major problem when interpreting the data and/or supporting decision making for HLB management. PMA qPCR for monitoring the live bacterial genome s may produce more useful data. There are quite a lot of factors could affect Las growth, and thus contribut e to the Las population inside hosts such as the initial inoculum, plant age, bacteria growth phase and environmental conditions (including light, temperatur e, water, nutrition, etc.) However, in this greenhouse study with controlled conditions, many factors could be eliminated when analyzing Las growt h. In January 2012, a slight drop in live Las population occurr ed in both plant hosts (Figure 3 2A & B). The pro phage hypothesis discussed above is not likely behind this event due to too much change in one month (up and down within one month) Some other factors are also ruled out due to the controlled conditions, like temperature, plant age and bacte rial growth phase Therefore, one possible factor causing this change is daylight time Since January is one of the months with shortest daylight time in Florida and daylight exposure is directly related to the ability of plant photosynthesis, which affec ts plant growth and in turn the growth of intracellular Las After successfully inoculat ed by psyllid, citrus seedlings in this study revealed their infection status as early as 2 month PI, while most S. buxifolia seedlings were qPCR positive at 5 month P I. After 5 month PI, new positive citrus and Severinia plants both can quickly reach a stable stage of bacterial growth with no significant difference in bacterial population ( P <0.05). Both plants inoculated at seedling stage s develop stunted symptoms. Bes ides, Las can be transmitted through S. buxifolia multiple times (at least 3 times demonstrated in this study) without any obvious change in pathogenicity
100 behavior. Furthermore, c onsidering the new discovery that Las living situations inside both plant spe cies were a lso quite comparable ( seasonal pattern of live Las population development), it can be assume d that S. buxifolia and citrus provide a quite similar living environment to Las and the psyllid and thus findings regarding HLB on either plant species could be interchangeable or predictable for the other 3.4.3. U ncommon Las Population F luctuation T hroughout the four years of observation, the two Severinia plants Sb1 and Sb2 had a lot of fluctuation in Las population compar ed to Las infected citrus ( see Table 4 6 ). A fter confirmed qPCR positive with Las (low Cq value could reach 20s), the two plants c ould produce undetermined Cq values (i.e. negative results) for a couple of months and then gradually decreased to 30s later. Furthermore, a similar fluc tuation of qPCR results also was observed in several other Severinia seedlings from the psyllid transmission experiments ( see Severinia plant 1 1 in Table 3 4 ). In order to investigate this uncommon fluctuation of Cq values, which could represent the drama tically changing situation of Las surviving inside the Severinia plants, the qPCR system with generic primer/probe set ( based on rpoB gene ) developed in our lab was us ed as a routine method to monitor those plants that showed the extreme fluctuation s. Usin g the two systems there were s ome difference in Cq values, i.e. usually lower Cq values with the rpoB gene primer/probe set than with the 16S primer set This was not expected since the 16S system usually has lower Cq value s than the rpoB system due to t he 3 copies of the 16S gene in Las genome compared to 1 copy of rpoB (Las psy62 genome in GenBank NC_012985.2 ) This may indicate that a certain unknown liberibacter was present in these S. buxifolia plants, which is not detectable by the Las specific 16S primer/probe set The other possib ility but less likely, is that the 16S system may have
101 affinity problem with these alternati ve host samples. More work is need ed to test these hypotheses.
102 Table 3 1. qPCR and conventional PCR systems used in this study System type Primer/probe Product size Region Final conc Source Duplex TaqMan qPCR 1F: HLBas 1R: HLBr 1P: HLBp 2F: COXf 2R: COXr 2P: COXp 75 bp 68 bp 16S rDNA cytochrome oxidase (COX) gene 250 nM 250 nM 150 nM 300 nM 300 nM 150 nM Li et al. 2006 Dup lex TaqMan qPCR 1F: HLBas 1R: HLBr 1P: HLBp 2F: WGf 2R: WGr 2P: WGp 75 bp 74 bp 16S rDNA g lycoprotein gene wingless ( wg ) 250 nM 250 nM 150 nM 300 nM 300 nM 150 nM Li et al. 2006 Manjunath et al. 2008a TaqMan qPCR F: CQULA04F R: CQULA04R P: CQULAP1 0 87 bp 50S ribosomal subunit protein L10 ( rplJ ) 800 nM 800 nM 400 nM Wang et al. 2006 TaqMan qPCR F: rpoBf R: rpoBr P: rpoBp 111 bp subunit of RNA polymerase ( rpoB ) 400 nM 400 nM 200 nM Developed in TaqMan qPCR F: Las spef R: Las sper P : Las spep 182 bp 50S ribosomal subunit protein L10 (rplJ ) 250 nM 250 nM 150 nM Developed in TaqMan qPCR F: Lam spef R: Lam sper P: Lam spep 174 bp 50S ribosomal subunit protein L10 (rplJ ) 250 nM 250 nM 150 nM Developed in T aqMan qPCR F: Laf spef R: Laf spcr P: Laf spep 157 bp 50S ribosomal subunit protein L11 ( rplK ) 250 nM 250 nM 150 nM Developed in Conventional PCR F: T7 1 R: Sp6 1 2588 bp 16S/23S rRNA 200 nM 200 nM Developed in
103 Table 3 2. R eaction of Severinia buxifolia to Las infection transmitted by psyllids from infected citrus S. buxifolia plant no. Assay 1 2 month PI (Cq) Assay 2 5 month PI (Cq) Assay 3 8 month PI (Cq) Assay 4 12 month PI (Cq) 1 1 34.9 38.2 0.0 0.0 1 2 0.0 35.0 38.6 27.9 1 3 0.0 32.8 2 1 0.0 39.0 0.0 0.0 2 2 0.0 38.4 0.0 0.0 2 3 35.3 33.4 38.7 29.8 2 4 0.0 27.0 33.0 35.7 2 5 0.0 39.5 0.0 0.0 3 1 0.0 37.3 39.6 32.6 3 2 33.3 26.7 30.2 34.6 3 3 0.0 38.2 0.0 0.0 3 4 0.0 39.0 0.0 0.0 Note: 0.0 means undeterm ined Cq value, i.e. negative results; empty box means plant died or was discarded; PI means post inoculation; Cq means quantitative cy cle. This note applies to Table 3 2, 3 3 3 4 and Table 4 1 through Table 4 17
104 Table 3 3. Reaction of Valencia sweet orange to Las infection transmitted by psyllids from infected S. buxifolia Assay/Months PI (Cq) P lant 1/2 2/5 3/8 4/10 5/12 6/14 7/16 8/18 9/20 10/24 1 1 0.0 24.7 25.8 24.1 22.5 21.6 23.4 21.0 20.6 18.4 1 2 0.0 0.0 0.0 0.0 0.0 1 3 0.0 0.0 0.0 0.0 0.0 1 4 0.0 27.3 23.9 22.4 20.9 21.2 23.1 21.9 20.6 19.3 1 5 35.2 27.7 24.5 22.9 21.4 23.4 23.6 20.9 19.4 1 6 0.0 0.0 0.0 0.0 0.0 1 7 0.0 26.8 24.0 22.4 21.0 22.2 23.6 20.8 20.2 18.3 1 8 0.0 0.0 0. 0 0.0 0.0 1 9 0.0 0.0 0.0 0.0 0.0 1 10 37.3 0.0 0.0 0.0 0.0 2 1 0.0 0.0 0.0 0.0 0.0 2 2 0.0 35.6 26.3 24.6 23.0 21.5 22.2 20.0 19.9 2 3 0.0 0.0 25.9 24.2 22.6 22.3 23.3 20.4 21.4 19.5 2 4 0.0 0.0 0.0 0.0 0.0 2 5 0.0 0.0 0.0 0.0 0.0 2 6 26.6 25.3 24.4 22.8 21.3 22.7 21.6 19.1 19.0 19.1 3 1 0.0 0.0 0.0 0.0 0.0 3 2 36. 29.4 31.1 29.1 27.2 22.6 26.5 19.4 20.1 19.0 3 3 0.0 0.0 0.0 0.0 0.0 3 4 0.0 0.0 0.0 0.0 0.0 3 5 0.0 0.0 0.0 0.0 0.0 3 6 31.5 30. 9 24.0 22.4 20.9 20.6 24.8 21.0 20.7 21.2 3 7 31.2 28.9 23.1 21.5 20.1 24.6 25.0 20 20.1 18.7 3 8 0.0 0.0 0.0 0.0 0.0 3 9 0.0 0.0 0.0 0.0 0.0 3 10 25.9 25.1 22.9 21.4 20.0 24.3 24.3 20.8 19.1 20.7 3 11 27.0 25.9 28.5 26.7 24.9 21.6 24.8 20.9 19.8 20.7 3 12 28.2 31.3 24.3 22.7 21.2 22.1 24.0 20.7 21.1 19.6 PI means post inoculation; Cq means quantitative cycle.
105 Table 3 4. Reactio n of S. buxifolia to Las inoculation transmitted by psyllid s from infected S. buxifolia Assay/Month s PI (Cq Plant 1/2 2/5 3/8 4/9 5/50 6/11 7/12 8/13 9/14 10/15 11/16 12/17 13/18 14/19 15/20 16/24 1 1 0.0 33.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 33.5 32.5 23. 3 0.0 0.0 1 2 35.7 0.0 26.7 24.1 26 26.1 26.1 23.8 22.5 25.4 25.5 22 24 23.1 23 26.2 2 1 0.0 0.0 0.0 2 2 0.0 0.0 0.0 2 3 0.0 26 27.9 27.4 23.2 27.7 0.0 26.8 25.3 23.8 24 22.4 26.8 24.3 24.6 23.9 2 4 34.5 26.8 22.6 22.6 25.6 25.4 27.9 23.3 25.3 21.5 24.1 22.5 24.8 25.1 23.8 24.9 2 5 0.0 25.8 24 27.4 22.1 24.5 21.3 23.7 23 21.2 22.6 20 25.5 27.1 26 24.8 2 6 0.0 25.8 25.8 27.8 25.5 28.7 28.1 23.8 23.4 21.8 23.7 23.4 23.9 28.3 23.9 26.6 2 7 0.0 0.0 0.0 2 8 0.0 2 3.9 23.3 24.2 21.5 22 22.3 25.3 24.4 22.6 25.5 22.5 22.2 26.9 28.4 24 2 9 0.0 0.0 24 36.3 23.5 27.4 26.9 27.4 24.3 22.2 25.2 20.4 24.3 28.2 25.2 24.9 3 1 0.0 38.3 35.7 ... ... ... ... ... ... ... ... ... ... ... ... ... 3 2 0.0 0.0 0.0 ... ... ... ... .. ... ... ... ... ... ... ... ... 3 3 0.0 38.9 36.3 ... ... ... ... ... ... ... ... ... ... ... ... ... 3 4 37.8 29.2 27.2 ... ... ... ... ... ... ... ... ... ... ... ... ... 3 5 0.0 37.6 35.1 ... ... ... ... ... ... ... ... ... ... ... ... ... Note: point; PI means post inoculation; Cq means quantitative cycle.
106 Fi gure 3 1. Sympto ms observed in HLB transmission experiments of S. buxifolia group. A ) Suspicious blotchy mottle on S. buxifolia After infected with Las, S. buxifolia plants showed some ch lorotic leaves, but this symptom was not consistent with qPCR results. I t was consid ered suspicious symptom. The only symptom always consistent with qPCR results was stunted plants as shown in B) and C). Tall plants in B) and C) were negative controls. B ) Stunted S. buxifolia and C ) Stunted Valencia citrus A B C A B
107 Figure 3 2. Dynami cs of live and cumulative genome of Ca L. asiaticus in S. buxifolia and citrus plants over a 20 month period. After inoculated with Las, 3 positive seedlings were chosen from citrus and S. buxifolia group, respectively, and monitored monthly by PMA qPCR t o check the bacterial dynamics inside the two plant species The monitor started from August 2010 and lasted for 20 months through April 2012. Error bars represented standard deviations from 3 replicates. A) Las dynamics in S. buxifolia B ) Las dynamics in citrus A
108 Figure 3 2. Continued. B
109 CHAPTER 4 MULTIPLE CITRUS RELA TIVES AS ALTERNATIVE HOSTS OF LAS AND PSYLLID TRANSMISSION STUDIES 4. 1 Introduction In this chapter, alternative hosts were studied as possible alternative hosts for huanglongbing (HLB ). These hosts represented 7 different plant species, x Citrofortunella microcarpa (Calamondin), Citropsis gilletiana Choisya ternata Sundance Choisya aztec Pearl Zanthoxylum fagara Esenbeckia runyonii and Amyris texana whi ch belong to 6 differen t genera A ll the plants discussed in this chapter are citrus relatives in the Rutaceae family. A brief introduction on each plant is as follows: Calamondin is a shrub or small tree growing to 3 6 meters. It is a cross between Citrus reticulata (Mandarin orange group) and Fortunella japonica (Kumquat group), and it was treated as an intergeneric hybrid in the nothogenus Citrofortunella as Citrofortunella microcarpa It is believed to originate from China and has spread throughout Southeast Asia, India, Hawaii, central and North America. In North America, Calamondin is primarily grown as an ornamental plant in gardens and pots, and the plant is especially attractive when the fruits are onset. The plant is frost sensitive; therefore, it is limited to fros t free climates such as Florida, coastal California, South Texas and Hawaii in the United States. Other than ornamental function, the plant is also used for fruit consumption, culinary, medicine and many other purposes. Choisya is a genus of aromatic ever green shrubs in the Rutaceae family. Members of this genus are commonly known as Mexican Orange or Mock Orange because of the similarity with orange, both in shape and the scent. They are native to southern North America, and from the southwestern states l ike Arizona, New Mexico
110 and Texas in the United States and south through most of Mexico. Choisya spp. are popular ornamentals in areas with mild winters for their abundant and fragrant flowers, and the most commonly used cultivars are the golden leaved C. ternata Sundance and C. aztec Pearl Z fagara is a spreading shrub or small tree in the Rutaceae family, which is native to southern Florida and Texas in the United States, and Mexico, Central America, the Caribbean, and South America as far south a s Paraguay. As indicated in its common name Lime Prickly ash, the plant is quite thorny and smells similar to citrus. As a significant food source and cover for native wildlife, Z. fagara has a high drought tolerance and grows best in full s un, but it can also survive as an understory shrub. E runyonii is a species of flowering tree in the Rutaceae family, which is native to northeastern Mexico with a small, disjunctive population in the Rio Grande Valley of Texas, USA (possibly arose from seeds dispersed by flooding). The showy foliage and blooms make the plant an attractive ornamental ; besides, they are also used as living fence in Mexico by planting their branch in the ground during dry season. Amyris is a genus of flowering plant in the Rutaceae family The generic name is derived from Greek meaning intensely scented which refers to the strong odor of the resin produced by the plant. And because of it s high resin content, the wood from plants of this genus is often used for torches and firewood, and thus their common name Torchwoods.
111 4.2 Materials and M ethods 4 .2.1 Plants and P syllids Citrus seedlings were grown from seeds purchased from California. Calamondin plants used in this study were usually grafted o n Cleopatra mandarin root stock Seedlings o f Z. fagara E. runyonii A. texana C. gilletiana C. ternata Sundance and C. aztec Pearl were first purchased from nurseries in Florida and Texas, however, some seedlings of C. gilletiana and cuttings of Choisya spp. were also grown in our greenhou se and used as backup plant source. All the plants were kept in an insect free greenhouse w ith natural temperature and daylight until ready for psyllid inoculation purpose. Healthy ACP s were collected from Murraya paniculata ornamentals located in Winter H aven, Florida, and raised in the healthy cage A ll the psyllid transmission experiments were set up in the psyllid room and kept for the first 2 months PI. After all inoculated plants were qPCR tested for the first time and remaining psyllids were remove d, the plants were moved to the third greenhouse with controlled temperature of 75 to 78 F and natural light for long term monitor and observation due to space limit in the psyllid room. A ll plants and psyllids used in this study were tested for latent Las infection by qPCR prior to use in any transmission experiment. 4.2.2 Inoculation P rocedure Because Calamondin is graft compati ble with citrus material, graft transmission was also conducted in this study. The budding method was used, and 9 Calamondin plan ts were grafted in summer 2009 with multiple bud s (usually mature buds with wood attached) collected from HLB positive Rough lemon plant ( C x jambhiri Lush. Rough
112 lemon ) with strong HLB symptoms Four months after grafting leaves were collected and teste d with qPCR to confirm the infection status. The other 6 plant species were not grafted with positive citrus scions in this study due to incompatibility with citrus materials Psyllid transmission is the major means used in the transmission experiments o f all the 7 plant species. For each of the 7 plants on the list, the three transmission pathways: (1) from citrus to alternative host (2) from alternative host to citrus, (3) from alternative host to new alternative host were tested with 3 separate trial s to validate the results. Regarding inoculation procedure, most trials just followed the standard steps, basically, put healthy psyllids on source plants for a 3 week of AAP (Dams teegt et al. 2010) ; then transfer the psyllids onto individual receptor plan t (better with new flush ready) and let them feed for at least 14 days of IAP (Damsteegt et al. 2010) ; at the same time set up positive and negative controls At 2 month PI, leaf samples for qPCR detection were collected. Then the plants were cleaned of al l psyllids and sprayed with pesticide and moved to the secure greenhouse for long term observation. Psyllid s from the source and receptor plants were all tested individually for Las using qPCR as described previously and positive percentage data were colle cted The transmission rate was defined as the number of plants which were tested Las positive by qPCR out of the total number of plants inoculated (by grafting or psyllid transmission). S ometimes psyllids in this study did not like th e pla nts as indicate d in literatures For example, psyllids put on Esenbeckia and Amyris plants w ould all disappear overnight, or sometimes, psyllids collected from some of the source plants showed 0% Las positive which made them unusable for transmission experiments. Therefo re, some
113 modifications were made to the standard inoculation procedure as foll ows Mesh bags were placed over the leaves to force psyllids to feed on alternative host plants tested and to keep track of the number of psyllids used ( See Figure 4 1F ) If psyl lid died too fast on the source plants, we only let them feed on source plants for 24 hours and then transfer r ed them onto receptor plants after that The psyllids were left on the receptor plants unt il they could not be found ( or dead ) so the 14 day of I AP did not apply in these cases In addition the psyllids fed on source plant s for at least 24 hours and then receptor plants were placed in to the same cage with source plants to allow transmission to occur natur ally. Finally the number of psyllids used f or transmission was increased from 10 to 30 40 psyllids per plant. 4 .2.3 Detection A ssays. All plants and psyllid materials were assayed by 16S qPCR system (Li et al. 2006) for Las presence and PMA qPCR was used to determine the live bacterial genome (See 2.2.2 PMA qPCR working protocol for details) Since this is the first report of the alternative host status for many of the plants in this chapter, several PCR detection systems, real time or conventional, were utiliz ed to further confirm the identity o f the transmitted Las (Table 3 1). For plant s DNA samples were extracted from newly matured leaves of each inoculated plant using DNeasy Plant Mini Kit (Qiagen, Valencia, CA) according to the a t least 3 leaves for plants with big leaves like citrus Calamondin and Citropsis while 8 leaves for smaller plants like Amyris Choisya ; for plant with tiny leaves like Zanthoxylum the branches with multiple opposite leaves were collected and used for D NA extraction ) were randomly selected and aseptically removed from each plant; leaf lamina were discarded and
114 midrib/petiole materials were finely chopped with sterile razor blades; 100 mg of chopped tissue were weighed out and put into a 2 mL microcentrif uge tube and pulverized using TissueLyser II (Qiagen) with liquid nitrogen, after which the tissue were ready for DNA extraction with the Plant Kit (Qiagen). For psyllid s the DNeasy Blood & Tissue Kit (Qiagen) was used for DNA extraction. Briefly, a singl e psyllid (or a group of psyllids depending on the purpose of the assay) was placed into a 1.5 mL microcentrifuge tube with 180 L ATL buffer (tissue lysis buffer from the DNeasy Blood & Tissue Kit) and gr oun d with a small plastic pestle until homogenized and then total DNA samples were estimated with a NanoDrop Spectrophotometer ND 1000 (Wilmington, DE), and all DNA samples were stored at 20 C. After valid Cq values wer e obtained, primers from the same system (without probe) were used in conventional PCR to visualize actual bands in gel. PCR products ( around 100bp) were sent for sequencing directly, while the 2588bp products from the conventional PCR (Table 3 1) were clo ned and sequenced as described in Chapter 3. All sequence data were blasted online at NCBI website (http://blast.ncbi.nlm.nih.gov/Blast.cgi). After receptor plants were confirmed positive with Las infection some were chosen and assayed for the live bact erial genome i.e. to monitor the live population dynamics of Las inside them over time by PMA qPCR However, due to the time limits, only some Calamondin plants were monitored for more than one year.
115 4.3 Results 4.3.1 Transmission E xperiments In the work discussed in this chapter, 6 plants were found to be able to serve as alternative hosts of Las and all were the first time reports. The transmission rates from all the transmission experiments are summarized in Table 4 18 The persistency of Las and psyl lid activities were the two major factors to evaluate the transmissibility and risk level of the hosts (see results in Table 4 19 ). The detailed results of each plant are individually presented. 220.127.116.11 x Citrofortunella microcarpa ( Calamondin ) For graft t ransmission, two out of nine plants ( 22.2 %) grafted with positive citrus bud s became Las positive after 4 months, and the bacterium persisted in Calamondin for greater than 30 months The positive Calamondin plants were used as source plants to raise psyll ids for inoculation of healthy citrus and healthy Calamondin plants in the transmission pathway 2 and 3. The following are those psyllid transmission results. From citrus to Calamondin transmission pathway 1. With psyllids raised on infected rough lemon plants, 2 3 and 4 Calamondin plants were inoculated in 3 separate trials which were set up in different months The two plants from the 1st trial were infected with an unknown fu gal disease and died before they were tested. For the 2nd and 3rd repeats, p lants showed Las positive qPCR results as early as 2 month PI ( 33% and 75 % for 2nd and 3rd trial respectively ) with moderate Cq values, while almost all the plants revealed their infection status at 4 month PI (Table 4 1) This demonstrated a high transmi ssibility of Las from citrus to Calamondin via the psyllid vector. Like S. buxifolia Calamondin can also be infested by psyllids without any obvious adapt ive phase and various psyllid activities, i.e. feeding and breeding (eggs
116 and nymphs), were easily observed on the Calamondin plants (Figure 4 1A). Valencia sweet orange positive controls which were inoculated with psyllids collected from the same Rough Lemon plants showed positive qPCR values with a Cq= 30 when first tested at 2 month PI, and later the Cq reach ed low 20s Calamondin plants mock inoculated with Las free psyllids remain ed Las qPCR negative throughout the whole testing period. From Calamondin to citrus transmission pathway 2. P syllids raised on graft transmitted positive Calamondin plants were used to inoculate 3 groups of Pineapple sweet orange plants separately ( 8 plants in total, see Table 4 2 ). After the 3 week of AAP, the majority of the psyllids collected from the positive Calamondin plants were Las positive (80% 100%) with Cq values ranging from 19 to 34. However, no citrus plants were infected in the 1st qPCR test (2 month PI) but were qPCR positive at 3 month PI B acterial growth was hindered in citrus since the Cq values remained in the 30s even at 5 month PI. From Calamondin to Calamondin transmission pathway 3 Three groups of Calamondin plants ( 3 plants in each group Table 4 3 ) were used to demonstrate the transmissibility from infected Calamondin to healthy ones via psyllids. The 3 plants of t he 1st trial were put in the sa me cage wit h the source Calamondin plants ( with a high percentage of psyllids carrying Las ) for 2 months, while the 2nd and 3rd trial followed the standard inoculation procedure described in Materials and methods Plants from the 1st trial were not posit ive until 4 month PI and the Las population remained at low level for the rest of the 2 months, while the other two trials had some different results The plants were qPCR positive at 2 month PI and the Las population was already quite
117 high (Cq values in t he middle 20s), which may reflect the effect of different inoculation procedures used. But the high transmission rate (88.9% in total) demonstrated the high transmissibility between Calamondin plants. 18.104.22.168 Citropsis gilletiana No graft transmission was attempted on C. gilletiana due to its graft incompatibility with citrus P syllid transmission was successful in all the three transmission pathways tested in this study. From citrus to C. gilletiana transmission pathway 1. Positive psyllids were raised and collected from HLB af fected rough lemon plants, with which 3 separate trials (6 3 and 3 C. gilletiana plants in each trial) were inoculated in the same months (weeks apart for each trial). At first, psyllids did not seem to feed on the Citropsis rec eptor plants and they usually disappeared after days. Therefore, the Mesh Bag method (See Figure 4 1F ) was used and new round of inoculation s w ere done Five out of the six positive plants were Las positive in the 1st qPCR test 2 month PI (83.3%, Table 4 4 ). The b acterial population s never accumulated to a high level in Citropsis (Cq values from 30 to 39), and some plants even lost the bacterium after a period of time (see undetectable results of plants 1 2, 1 6, 2 2 and 3 1 in Table 4 4 ) H owever, the bact erium survived in the host for more than 14 mont hs Psyllid seemed to not like Citropsis plants in these three trials, be cause no infestation was observed. The situation changed when the positive Citropsis plants from this pathway were used for the other t wo transmission pathways, and some psyllid activities (but limited compared to S. buxifolia and Calamondin) were observed. From C. gilletiana to citrus transmission pathway 2. P ositive Citropsis plants from pathway 1 were used to raise healthy psyllids a nd to inoculate 3 groups of
118 Pineapple sweet orange plants separately ( 7 plants in total, see Table 4 5 ). As mentioned above, the psyll id somehow started to infest the positive Citropsis (Figure 4 1B), only no breeding acti vities were found, i.e. no eggs o r nymphs were found After the 3 week of AAP, the majority of psyllids collected from the positive Citropsis plants were Las positive (80% 100%) with Cq values ranging from 16 to 34. Due to limited psyllids available (no breeding), citrus seedlings (recept or plants) with new flush were put in the same cage with the source plants until the 1st qPCR test (2 month PI), and the natural movement from Citropsis to citrus was observed. Citrus did get inoculated in this way and qPCR positive plants were detected in the 1st qPCR test with high Cq values (100% of the plants ) The Las bacterium in these citrus plants from this pathway seemed to grow slowly with constantly high Cq values above 37. From C. gilletiana to C. gilletiana transmission pathway 3 Three group s of C. gilletiana seedling s ( 14 in total, Table 4 6 ) were used to demonstrate the transmissibility from infected C. gilletiana to healthy ones via psyllids. The 1st trial followed the standard inoculation procedure (the artificial psyllid transfer process ), while in the other two trials, small C. gilletiana seedlings were put in the same cage with the source C gilletiana plants (with high percentage of psyllids carrying Las ) for 2 months due to limited psyllids availability (no breeding) Plants from the 1st trial were both successfully inoculated and the Las population remained at a low level (Cq from 34 to 38) to date For the other two trials, some psyllids did move from source plants and feed on the small seedlings, however, there was only one positive result (1 out of 12 plants total) with a Cq of 34. T he t ransmission rate varied a lot in this pathway most probably due to the different inoculation procedure used.
119 22.214.171.124 Choisya spp. Due to incompatibility no graft transmission was attempted on the tw o Choisya species, C ternata Sundance and C aztec Pearl Psyllid transmission was successful in all the three transmission pathways tested. No obvious difference was observed between the two Choisya species in transmission experiments. Psyllid activi ties were similar to what was observed on C. gilletiana From citrus to Ch o isya spp. transmission pathway 1. Positive psyllids were raised and collected from HLB af fected R ough lemon plants, and were transferred to Choisya plants in 3 separate trials (4 4 and 3 Choisya plants in each trial, Table 4 7 ). Plants of the two Choisya species, C ternata Sundance and C aztec Pearl were randomly mixed in each inoculation trial, but no different reactions, i.e. psyllid preference, transmission rate, Las p ersistence, etc., were observed in this study. At first, psyllids seemed to have problem feeding on the Choisya receptor plants, and they usually died after days. Therefore, the same Mesh Bag method was utilized in this group, and this helped achieve tran smission to almost 100% success. All plants were Las positive in the first two qPCR tests, but the bacterial population never accumulated to a high level (Cq values from 32 to 39), and some plants even lost the bacterium over time (see undetectable results of plants 1 4, 2 1, 2 3, 3 2, and 3 3 in Table 4 7 ). La s survived in the host for more than 13 months (2 1 and 2 4 in Table 4 7 ). Psyllid seemed to have a phase to adjust to the Choisya plants in these three trials, and no psyllid colonization was observe d (Figure 4 1C) From Choisya spp. to citrus transmission pathway 2. P ositive Choisya plants ( both C ternata Sundance and C aztec Pearl ) from pathway 1 were used to raise healthy psyllids and to inoculate 3 groups of Pineapple sweet orange ( Citrus sinensis
120 Pineapple ) plants separately ( 15 plants in total, see Table 4 8 ). Although no psyllid infestation occurred on Choisya a good number of psyllids collected from positive Choisya plants were Las positive (30% 100%) with Cq values ranging from 20 t o 38 after the 3 week AAP. Due to the limited psyllids available (no breeding), citrus seedlings (receptor plants) with new flush were put in the same cage with the source plants until the 1st qPCR test (2 month PI), and the natural movement from Choisya t o citrus was observed. All of the citrus plants were infected in this way and positive plants showed up in the 1st qPCR test with high Cq values, however, the Las bacterium in citrus plants of this pathway seemed to grow slower when compared to the recepto r citrus plants in Severinia research which may be due to less inoculum obtained from the plants From Choisya to Choisya transmission pathway 3 Three groups of Choisya ( 17 in total, Table 4 9 ) were used to demonstrate the transmissibility from infecte d Choisya plants to healthy ones via psyllids. Due to plant availability, small Choisya cuttings from the backup stock (all cuttings were kept in insect free greenhouse for at least 6 months) were used in these transmission trials, and they were put in the same cage with the source Choisya plants (with high percentage of psyllids carrying Las ) for 2 months until the 1st qPCR test. During the two month IAP, the movement of psyllids from source plants to receptor cuttings was observed. Most of the cuttings we re Las positive in the 1st qPCR (10 out of 17), and more were infected in the 2nd test (13 out of 17), although Las growth was still limited in Choisya F or the two Choisya species, no different reactions were observed in the trials of this pathway.
121 4.3.1 .4 Zanthoxylum fagara No graft transmission was attempted on Z. fagara due to incompatibility Psyllid transmission was successful in all the three transmission pathways tested. From citrus to Z. fagara transmission pathway 1. With positive psyllids rai sed and collected from HLB af fected R ough lemon plants, 4 separate trials (2 2 3, and 3 Z. fagara plants in each trial, see Table 4 10 ) were set up. P syllids did not feed well on this thorny plant ; however, the transmission was accomplished (100% positiv e). Some psyllid activities were observed, but quite limited compared to S. buxifolia and Calamondin (even less activities than Citropsis and Choisya most of time Figure 4 1D ). Mesh Bag method (See Figure 4 1F ) was also used for certain trials when psylli d feeding seemed to be a problem. After infection Las did not accumulated to a high population level (Cq values stayed in the upper 30s, rarely below 35) S ome of the Zanthoxylum plants started to lose the bacterium from as early as 3 month PI, and the lo ngest time the bacterium persisted inside this host was about 12 months. From Z. fagara to citrus transmission pathway 2. T here was no psyllid reproduction on Zanthoxylum plants, and the number of psyllid s decreas ed over time since they were released on to the source plants. However, surviving psyllids fed and acquired the bacterium from the plant (20% 40% positive psyllids ) Cq values from 17 to 32 were obtained probably due to prolonged feeding time. Due to the limited number of psyllids available, citr us plants of the se 3 trials were put in the same cage with the source Zanthoxylum plants until the 1st qPCR (2 month period), and natural movement from Zanthoxylum to citrus occurred Most citrus plants were inoculated with Las in this way (12 out 13 plant s in total, see Table 4 11 ). Las seemed to develop slowly in these plants probably due to less bacterial inoculum transmitted at the beginning
122 From Z. fagara to Z. fagara transmission pathway 3 In the first two trials in this transmission pathway, psyl lids raised on positive Zanthoxylum plants were collected and transmissions were done using the Mesh bag method and psyllid activity was monitored because of the limited number of psyllids This method worked quite well, and all three plants were inoculate d with Las As observed in the transmission pathway 1, the qPCR values of Las in creased slowly and the Zanthoxylum plants began to lose the bacterium from as early as 5 months PI in this pathway (plant 1 1, see Table 4 12 ). 126.96.36.199 Esenbeckia runyonii No g raft transmission was attempted with E. runyonii due to graft incompatibility Psyllid transmission was successful in all the three transmission pathways tested even though no psyllids were collected from E. runyonii source plants From citrus to E. runyo nii transmission pathway 1. P ositive psyllids were raised and collected from HLB af fected R ough lemon plants and 3 separate trials (6 6 and 1 E. runyonii plants in each trial, see Table 4 13 ) were done Psyllids were not actively feeding on this plant ; therefore, mesh bags were used to force them to feed. P lants were in fected with Las in this way (5 out of 13 plants in total). Las concentration was low represe nted by Cq values above 35 all the time and the bacterium was only detected for 8 month s Psyl lid activities on Esenbeckia plants were extremely limited in this study. Little psyllid feeding was observed and eggs or nymphs were never found From E. runyonii to citrus transmission pathway 2. When positive Esenbeckia plants from trials of transmi ssion pathway 1 were used to r ear psyllids, the psyllids were extremely hard to find after AAP (Figure 4 1E) No psyllid data for any E. runyonii source plants were collected in this study because psyllids died overnight. However, when the back inoculation was attempted anyway by putting citrus seedlings
123 (with plenty of new flushes) in the same cage with positive E. runyonii source plants ( healthy psyllids were released onto E. runyonii plants at least 24 hour prior to this), transmission did occur Limited psyllid movement from E. runyonii plants to citrus plants was observed in any of the three trials. T he citrus plants usually were Las positive in the 1st qPCR test (2 month PI). Las seemed to develop slow ly since qPCR values decreased slowly over time (Ta ble 4 14) From E. runyonii to E. runyonii transmission pathway 3 N o psyllids were collectable from E. runyonii source plants and therefore all receptor E. runyonii plants of the 3 separate trials were put in the same cage with the source plants until t he 1st qPCR 2 months later. N o movement of psyllids was observed in these trials. However, the transmission still occurred, but at a low rate in this pathway (only 20%). Las did not accumulate to high levels in the positive E. runyonii plants (Cq values ab ove 35, Table 4 15 ). 188.8.131.52 Amyris texana No graft transmission was attempted on A. texana due to graft incompatibility with citrus Psyllid transmission failed for all the three transmission pathways tested with multiple trials. All modifications of the inoculation procedure were applied to this host plant, for example, mesh bag, increased number of psyllids for inoculation and prolonged IAP, but none of these means led to successful Las transmission. P syllids did not appear to feed on this host (Figure 4 1F), and usually died days after released onto the plants. No psyllid breeding activities were observed. C itrus plant s of positive control were infected by Las, while mock inoculated A. texana plants remained qPCR negative throughout the entire monitor p eriod.
124 4.3.2 Las I dentification The identity of the bacterial agent, Las, in these hosts was determined and confirmed with different methods. After positive qPCR results (i.e. valid Cq values) were obtained with the routine assay of 16S qPCR, several othe r diagnostic assays were also utilized to verify results (Table 3 1). First, several other qPCR systems target ing different regions of the bacterial genome were u s ed to retest the positive plant samples from the psyllid transmission trials for example, CQ ULA04F/R CQULA10P set based on the gene encoding the 50S ri bosomal subunit protein L10 ( rplJ ) (Wang et al. 2006) of Las and rpoBf/r/p set based on the gene encoding the beta subunit of RNA polymerase ( rpoB ) of l iberibacters P ositive results were obtained with CQULA04F/R CQULA10P, rpoBf/r/p, and Las spef/r/p systems while negative results with Lam spef/r/p and Laf spef/r/p systems. Some of the qPCR systems plus one conventional PCR system (Table 3 1) were run on the regular thermocycler and actual bands wi th expected amp licon sizes were visualized For example, 75 bp and 111 bp for 16S and rpoB primer sets, respectively, were visualized in gels (See Figure 4 3B) In addition, the PCR products were pu rified and sequenced T he 2588 bp PCR product was first cl oned (See Figure 4 3C) and then sent for sequencing S equence data were blasted online at NCBI and confirmed to be the appropriate Las sequence with high similarity ( See BLAST results in Appendix ). 4.3.3 Symptom O bservation After the 1st qPCR test 2 mont hs PI, all of the inoculated plants were cleaned of psyllids and moved to the greenhouse with controlled temperature and natural daylight conditions for incubation and symptom observation. In the large collection of plant materials kept in the greenhouse ( about 300 plants) various plant symptoms were
125 observed. First of all all citrus plants (mostly sweet oranges seedlings, such as Pineapple, Valencia and Hamlin) which were successfully inoculated with Las in this study showed certain typical HLB symptoms, for example, leaves with blotchy mottle yellowing leaves, small leaves, leaves with corky vein, leaves with green island (Figure 4 2A) and stunt ed plants (F igure 3 1C ). L as positive Calamondin plants also showed typical HLB symptoms such as blotchy mo ttle leaves (Figure 2 6 B). Second, citrus relatives used in this study also developed some symptoms, for example, leaf chlorosis (Figure 4 2B to F ) H owever, these leaf symptoms were not always consistent with qPCR test results Third, some symptoms observ ed in this study were always consistent with qPCR results, i.e. the Las infection status, so they could be considered as typical HLB symptoms on certain alternative host s For example, the stunting symptom found in all positive S buxifolia plants tested was discussed in Chapter 3, and could be used as a good indicator of Las infection in S. buxifolia Stunting was also observed in Las positive Calamon din plants. However, the se stunting symptoms were not observed in other alternative hosts. This is probabl y due to the less development of Las bacteria inside these hosts (represented by constant Cq values above 30). 4.3.4 Las P opulation D ynamics in A lternative H osts Along with the plants in Severinia group, two graft transmitted Calamondin plants were tested regularly every other month beginning August 2010 till April 2012 by PMA qPCR to determine the amount of living Las in the plants Two citrus plants (Pineapple sweet orange) were used as positive control s for the grafting experiments of the Calamondin grou p, and they were also included in this monitoring process to check the difference of Las development. N o general pattern could be demonstrat ed with statistical significance. However, as shown in Table 4 16 a basic trend was still
126 noticeable from the data obtained: the graft inoculated Calamondin plants tended to have lower live bacterial genome s ( represented by higher Cq values around 30s ) in winter time (from December to Februarys ); while in summer time (from June to October), the live Las population w oul d accumulate to a higher level (Cq values around 25). This basic trend also was found in the data from the two graft inoculat ed Severinia plants ( Sb 1 and Sb 2 in Table 4 16 and 4 17 ). In the graft inoculat ed Calamondin and Severinia plants, the cumulative bacterial genome also showed this basic trend (Table 4 16 and 4 17 ). In contrast, graft inoculat ed citrus plants (Citrus 1 and 2 in Table 4 16 and 4 17 ) showed no such trend for either live or cumulative bacterial genome s throughout the time period monit ored. The bacterial genome (live and cumulative ) fluctuated on a rather high level (Cq values ranging from 21 to 25 and from 19 to 24 for live and cumulative bacterial genome respectively) for a period of more than 20 months. For the other 4 plant groups i.e. Citropsis Choisya Zanthoxylum and Esenbeckia selected plants were also tested with PMA qPCR after their Las infection status was confirmed H owever, due to the low bacterial population in these hosts (as shown in Table 4 4 to Table 4 15 ), the l ive bacterial genome s obtained by PMA qPCR method were often low either high Cq values around 39, or undete ctable (undetermined results data not shown). Besides, the time period for these host plants to be tested by PMA qPCR was quite short, mostly becau se of the short persistenc e of Las bacterium in some of these hosts (i.e. the bacteria died out with in couple of months). Therefore, the data of live/ cumulative bacterial genome change collected from these 4 plant hosts w as limited.
127 4 .4 Discussion 4.4.1 Transmission E xperiments and A lternative H ost S tatus For most of the alternative hosts in this study, the psyllid transmission experiments were implemented in a manner of no choice inoculation, which means psyllids were manually transferred to the source /receptor plants and forced to feed. This may not be common in the field when there are preferred hosts i.e., citrus, available for the psyllid vector. However, it may happen in the case that all citrus plants are sprayed with insecticide, and other rutac eous plants in the neighborhood become the only choice of the psyllids, which is highly possible under the current HLB management (Halbert and Manjunath 2004). The results of transmission experiments as well as qPCR tests of Las from all the tested citrus relatives (including S. buxifolia discussed in Chapter 3) and psyllids were summarized in Table 4 18 and 4 19 7 plants were demo nstrated to be alternative hosts of Las of which 6 were reported for the first time. The 6 new discovered alternative hosts o f Las are x C microcarpa (Calamondin), C gilletiana C ternata Sundance C aztec Pearl Z fagara E runyonii A texana was shown to not be a host of Las Furthermore, based on the persistence of Las (Table 4 19 ) and psyllid activities on these c itrus relatives, they were grouped into 4 different categories : x C microcarpa (Calamondin) in the high transmissibility group with S. buxifolia ( Chapter 3 ) ; Choisya spp. (two species), C gilletiana and Z fagara form a medium high transmissibility grou p; E runyonii in a medium transmissibility group; A texana in the no transmissibility group. The alternative host information including transmissibility evaluation obtained in this study should provide help in HLB management, for example, alternative hos ts with high or medium high transmissibility should be monitored by PCR if they present in the
128 vicinity of citrus groves, and the sale and transportation of these plant materials should be restricted especially between HLB affected and HLB free areas. Lik e S. buxifolia and most citrus cultivars, Calamondin was a good host for both the psyllid vector and Las bacterium in this study. After being transferred to Calamondin, psyllid s could rapidly infest the plant, and lay eggs, hatch and develop into nymphs (F igure 4 1A). After infection with Las, Calamondin expressed typical HLB symptoms such as leaves with blotchy mottle and stunted growth. Las transmission from citrus to Calamondin was not successful in the first trial, probably because the trial was comple ted in the winter time when the bacterial population happened to be really low in source plants. All the other trials were successful with a good transmission rates. After the 3 week A AP, psyllids collected from positive Calamondin plants were mostly posit ive with high Las population s (Table 4 19 ), which demonstrated that Calamondin was a good inoculum reservoir. The population of Las in Calamondin was quite high comparable to that of citrus plant s throughout the entire monitored period (>30 months). Given that Calamondin is a common ornamental plant in Florida, it should be considered as a high risk alternative host. As alternative hosts, C. gilletiana Choisya spp., and Z. fagara should be considered in the group of medium high risk plants for Las transmi ssibility. Successful transmissions were obtained in all the transmission pathways tested, but the Las bacterium never accumulated to a high population inside these hosts, and even died out in Z. fagara within one year. This may indicate these plants have some resistance to the bacterium. In addition, they are not preferred hosts for the psyllid. Psyllids were forced to feed on them in no choice experiments, or even by the mesh bag method. Trials
129 without certain enforcement all failed. No eggs or nymphs w ere ever found on these plants, which showed limited psyllid activities associated with these plants (especially on Z. fagara ). However, the psyllid did acquire Las from these plants (Table 4 19 ) and accomplish transmission back to citrus. All plants in th is group showed minimal symptoms from Las infection, which indicated that HLB may not be a disease for them. Nevertheless, their risk as bridging alternative hosts in the field should be considered medium to high. E. runyonii was not a good host for eithe r the psyllid vector or Las bacterium in this study. Very limited psyllid activities were observed (no eggs, no nymphs, and even no psyllid adults available for qPCR test after AAP and IAP), and the inoculated Las bacterium died out within 5 months. No sym ptoms were observed. However, transmission was successful for all the pathways tested. All facts considered, E. runyonii is an alternative host with medium risk in HLB management. A texana was not infected by Las after multiple trials in this study, and p syllids did not readily feed on this plant. Therefore, A. texana was not considered as an alternative host of Las, and thus has no transmissibility for Las 4.4.2 Live Las P opulation i n H ost P lants Da ta from the graft transmitted plants were collected ove r a 20 month period, but no general pattern was found Because fewer plants were included in this part (2 plants from each group, see Table 4 16 and 4 17 ) no statistically valid pattern could be generalized. However, the basic trend (i.e. high bacterial population in summer time and low in winter) observed in these plant samples did not agree with the seasonal pattern in some particular months, and these differences could be caused by the different plant ages and inoculation methods. C ompared to the seedl ings used in the
130 psyllid transmission experiments, plants used in graft transmission experiments were at least several years old The different growth stages of plants could have considerable impact on how the plants react to infection This could also be associated with the development of defense system s in plants which could be responsible for the differen ces in development of the Las bacterium inside host plants. Second, the group s of plants shown in Table 4 1 6 and 4 17 were all inoculated by bud grafti ng while the other group s discussed in Chapter 3 w ere all inoculated by psyllid s Considering that different inoculation methods were used, different results could be expected to occur A recent study reported that with a side grafting method using Las po sitive lemon trees, the scion survival rate and transmission rate could reach as high as 81.2% and 88.8%, r espectively (Zhang et al. 2012) H owever, this was not found in our graft transmission tests (for example, much lower budding survival rates in the C alamondin group). Further comparative research work to demonstrate the difference in Las development in plant hosts following different inoculation methods needs study With all the factors affecting Las development considered, the basic trend of liv e/ cumulative bacterial genome dynamic s found in the graft inoculat ed plants was still quite informative. First of all, this trend confirmed the pattern found in the Severinia work As discussed in Chapter 3, the reason why the seasonal pattern could be obs erved in those two Severinia and citrus groups (3 plants in each group) was because they were synchronized plant seedlings and the pattern probably revealed the early adaptation and development stage of Las in the hosts. Therefore, although the other two g roups of citrus plants (from Severinia transmission pathway 2) did not show the seasonal pattern (only one stage of high bacterial population was observed), they
131 still complied with the seasonal pattern. Because the se plants could be interpreted as in the second stage of development since they were monitored at least one year after the inoculation. The same interpretation could be applied here. The Calamondin and citrus plants were graft inoculated in summer of 2009 ( Sb 1 and Sb 2 were grafted even earlier) and they were monitored by PMA qPCR at least one year after inoculation. T he refore, the lower bacterial population in winter time was part of the seasonal pattern. Second, this basic trend indicated a general time table for the ups and downs in Las devel opment inside host s which could be used in various applications. For example, in order to ensure a successful transmission, knowing that Las bacterial population is relatively low inside Calamondin and Severinia hosts (or maybe all alternative hosts) duri ng winter time, all the transmission work using these positive hosts as source plants should avoid winter time, and choose the hot season ( i.e. high bacterial population in plants ) to implement. This also explained why some of the psyllid transmission e xperiments in this study ended up with extraordinarily low transmission rates (sometimes even a total failure), such as the 1st trial in transmission pathway 2 of Severinia group (40% transmission rate) and the 1 st trial in transmission pathway 1 of Calamo ndin group (total failure), most probably because they were set up in winter time, and the psyllids used for transmission could not acquire enough bacteria from the source plants. I f there is any pesti cide application program for HLB management, this infor mation suggests a better timing for application, which should be at the end of winter, right before the Las bacteri al population starts to increase This timing coincides with the conclusion derived from the seasonal pattern.
132 Regarding the data of live La s population s in the other 4 plant hosts ( Citropsis Choisya Zanthoxylum and Esenbeckia ) the erratic results might reveal a limit of the PMA qPCR methodology when applied to samples with extremely low bacterial population s. W hen the cumulative bacterial genome was still detectable (Cq values around 37), the live bacterial genome turned out to be undetectable. This could possibly be due to some detection limit issue with the new methodology i.e. the extra pre treatment step somehow reduced the sensitivi ty of the following qPCR H owever, another possibility which should not be ruled out was that the live bacteria in those samples were indeed much fewer in number than the dead ones, and therefore, the failure to detect live Las might actually reveal this f act. Another issue found in the live bacterial genome data was that it could show higher population than the corresponding ones of cumulative bacterial genome for example, several sets of data from Citropsis group showed that cumulative bacterial genome s were above Cq values of 38, while the corresponding live ones were all below 37. This was logically impossible be cause live bacteria were only a part of t he cumulative genome and th i s led to the conclusion that the data were unstable. Las bacteria were un evenly d istributed throughout the host and this fact should have major impact when the bacterial population was l ow T his reason alone could not be responsible for all the systematically deviated results found in these 4 plant hosts. Therefore PMA qPCR m ethod was not recommended for use with samples with extremely low bacterial population s (i.e. Cq values above 36)
133 Figure 4 1. Psyllid activities observed on alternative hosts. Eggs and nymphs could be easily observed on C alamondin plants. P syllids were found on plants from B) to E) but eggs and nymphs were never observed. A ) Calamondin B ) Citropsis gilletiana C ) Choisya D ) Zanthoxylum fagara E ) Esenbeckia runyonii and F ) Mesh bag method used to force psyll id to feed on alternative host using E. runyonii as an example A B B C D E F
134 Table 4 1 Results of psyllid transmission pathway 1 from citrus to Calamondin Calamondin plant no. Assay 1 2 month PI (Cq) Assay 2 3 month PI (Cq) Assay 3 4 month PI (Cq) Assay 4 5 month PI (Cq) 1 1 0.0 0.0 0.0 ... 1 2 0.0 0.0 0.0 ... 2 1 31.2 30.0 25.0 23.4 2 2 0.0 0.0 34.4 32.1 2 3 0.0 0.0 38.3 35.8 3 1 29.4 27.5 ... ... 3 2 36.6 34.2 ... ... 3 3 0.0 0.0 ... ... 3 4 36.1 33.7 ... ... Table 4 2 Results of psyllid transmission pathway 2 f rom Calamondin to citrus citrus plant no. Assay 1 2 month PI (Cq) Assay 2 3 month PI (Cq) Assay 3 4 month PI (Cq) Assay 4 5 month PI (Cq) 1 1 0.0 39.7 37.7 35.3 1 2 0.0 0.0 37.9 35.4 1 3 0.0 0.0 39.0 36.4 2 1 0.0 39.4 38.4 35.9 2 2 0.0 38.3 35.0 32.7 2 3 0.0 39.7 38.7 36.2 3 1 0.0 0.0 0.0 0.0 3 2 0.0 39.6 38.4 35.9 Table 4 3 Results of psyllid transmission pathway 3 f rom Calamondin to Calamondin Calamondin plant no. Assay 1 2 month PI (Cq) Assay 2 3 month PI (Cq) Assay 3 4 month PI (Cq) Assay 4 5 month PI (Cq) Assay 5 6 month PI (Cq) 1 1 0.0 0.0 39.8 0.0 39.3 1 2 0.0 0.0 33.1 38.2 36.5 1 3 0.0 0.0 0.0 35.8 38.5 2 1 25.1 23.4 ... ... ... 2 2 23.0 21.5 ... ... ... 2 3 25.4 23.8 ... ... ... 3 1 0.0 0.0 ... ... ... 3 2 26.7 24.9 ... ... ... 3 3 25.2 23.5 ... ... ... ; PI means post inoculation; Cq means quantitative cycle.
135 Table 4 4. Results of psyllid transmission pathway 1 f rom citrus to Citropsis gilletiana Assay/Months PI (Cq) C. gilletiana 1/2 2/4 3/6 4/8 5/10 6/12 7/14 1 1 30.9 36.1 36.1 39.3 38.4 35.7 34.0 1 2 35.3 36.4 0.0 39.3 38.1 36.6 39.0 1 3 0.0 0.0 0.0 0.0 0.0 0.0 1 4 0.0 0.0 0.0 0.0 0.0 0.0 1 5 0.0 0.0 0.0 0.0 0.0 0 .0 1 6 36.8 37.4 38.4 0.0 39.6 37.2 38.5 2 1 37.0 37.5 38.2 39.3 39.0 37.4 37.1 2 2 38.1 33.2 39.5 0.0 38.9 38.6 36.7 2 3 0.0 0.0 0.0 0.0 0.0 0.0 3 1 0.0 0.0 35.9 33.6 0.0 0.0 36.5 3 2 0.0 0.0 0.0 0.0 0.0 0.0 3 3 0.0 0.0 0.0 0.0 0.0 0.0 Table 4 5. Results of psyllid transmission pathway 2 f rom C. gilletiana to citrus C itrus plant no. Assay 1 2 month PI (Cq) Assay 2 3 month PI (Cq) Assay 3 4 month PI (Cq) Assay 4 5 month PI (Cq) 1 1 0.0 39.2 39.3 37.8 1 2 0.0 39.2 37.7 38.3 1 3 39.5 0.0 38. 6 38 2 1 0.0 37.6 38.6 37.1 2 2 0.0 39.3 37.7 38.2 3 1 39.0 37.6 38.5 38.9 3 2 0.0 38.6 38.2 37.9 Table 4 6 Results of psyllid transmission pathway 3 f rom C. gilletiana to C. gilletiana C. gilletiana plant no. Assay 1 2 month PI (Cq) Assay 2 3 mo nth PI (Cq) Assay 3 4 month PI (Cq) Assay 4 5 month PI (Cq) 1 1 0.0 38.2 36.8 34.4 1 2 0.0 0.0 38.2 35.7 2 1 0.0 0.0 ... ... 2 2 34.7 34.1 ... ... 2 3 0.0 0.0 ... ... 2 4 0.0 0.0 ... ... 2 5 0.0 0.0 ... ... 2 6 0.0 0.0 ... ... 3 1 0.0 0.0 ... ... 3 2 0.0 0.0 ... ... 3 3 0.0 0.0 ... ... 3 4 0.0 0.0 ... ... 3 5 0.0 0.0 ... ... 3 6 0.0 0.0 ... ... means post inoculation; Cq means quantitative cycle.
136 Table 4 7. Results of psyllid transmission pathway 1 f rom citrus to Choisya spp Choisya spp. plant no. Assay/Months PI (Cq) 1/2 2/4 3/6 4/8 5/10 6/12 7/13 1 1 0.0 1 2 0.0 36.3 0.0 0.0 0. 0 1 3 0.0 36.7 0.0 0.0 0.0 1 4 0.0 36.9 0.0 34.1 0.0 2 1 32.7 36.5 0.0 33.7 38.0 35.8 38.0 2 2 36.9 38.5 2 3 34.3 36.3 38.8 0.0 0.0 0.0 39.3 2 4 34.9 37.0 38.7 3 1 39.7 0.0 0.0 0.0 0.0 ... ... 3 2 38.5 0.0 38.8 38.1 31.9 ... ... 3 3 32.0 0.0 39.8 37.3 39.2 ... ... Table 4 8 Results of psyllid transmission pathway 2 from Choisya spp. to citrus C itrus plant no. Assay 1 2 month PI (Cq) Assay 2 3 month PI (Cq) 1 1 38.6 36.1 1 2 38.5 35.9 2 1 37.6 35.2 2 2 34.8 3 4 .5 2 3 39.9 37. 2 3 1 35.2 3 4 .9 3 2 34.1 3 4 .9 3 3 38.7 36.2 3 4 38.9 36.4 3 5 38.3 35.8 3 6 37.8 35.3 3 7 37.4 35.0 3 8 37.7 35.2 3 9 36.9 34.5 3 10 34.0 3 5 .8 Table 4 9. Results of psyllid transmission pathway 3 f rom Choisya to Choisya Choisya spp. plant no Assay 1 2 month PI (Cq) Assay 2 3 month PI (Cq) 1 1 33.9 3 3 .7 1 2 38.3 36.8 1 3 0.0 37.1 1 4 38.0 37.5 1 5 0.0 0.0 1 6 0.0 36.4 2 1 39.6 37 2 2 37.6 38.1 2 3 0.0 0.0 2 4 38.5 39.1 2 5 36.7 34.3 2 6 0.0 0.0 3 1 36.9 35.8 3 2 0.0 0.0 3 3 39. 0 38.1 3 4 38.3 36.8 3 5 0.0 36.4
137 Table 4 10. Results of psyllid transmission pathway 1 from citrus to Zanthoxylum fagara Z fagara plant no. Assay/Months PI (Cq) 1/2 2/3 3/4 4/6 5/8 6/9 7/10 8/12 9/13 1 1 38.0 ... 32.7 39.7 37.3 39.3 39.2 38.6 0. 0 1 2 38.3 ... 34.0 0.0 0.0 0.0 0.0 0.0 0.0 2 1 38.6 37.4 0.0 0.0 0.0 ... ... ... ... 2 2 0.0 37.3 38.7 38.0 0.0 ... ... ... ... 3 1 38.0 38.9 35.9 ... ... ... ... ... ... 3 2 38.0 35.6 0.0 ... ... ... ... ... ... 3 3 0.0 39.0 0.0 ... ... ... ... ... ... 4 1 38.8 37.2 36.4 ... ... ... ... ... ... 4 2 0.0 34.3 38.6 ... ... ... ... ... ... 4 3 37.9 32.5 39.1 ... ... ... ... ... ... Table 4 11. Results of psyllid transmission pathway 2 f rom Z. fagara to citrus Citrus plant no. Assay 1 2 month PI (C q) Assay 2 3 month PI (Cq) Assay 3 4 month PI (Cq) Assay 4 5 month PI (Cq) 1 1 0.0 39.5 0.0 38.7 1 2 0.0 38.8 0.0 36.6 1 3 39.0 36.1 37.0 34.6 2 1 35.5 0.0 36.3 ... 3 1 37.1 34.7 ... ... 3 2 28.0 26.2 ... ... 3 3 39.4 36.8 ... ... 3 4 39.0 36.4 ... ... 3 5 0.0 0.0 ... ... 3 6 35.5 33.2 ... ... 3 7 37.8 35.3 ... ... 3 8 36.9 34.5 ... ... 3 9 36.7 34.3 ... ... on; Cq means quantitative cycle
138 Table 4 12. Results of psyllid transmission pathway 3 f rom Z. fagara to Z. fagara Zanthoxylum plant no. Assay 1 2 month PI (Cq) Assay 2 3 month PI (Cq) Assay 3 4 month PI (Cq) Assay 4 5 month PI (Cq) Assay 5 7 month PI (Cq) 1 1 0.0 36.5 32.1 38.2 0.0 2 1 0.0 32.4 36.4 35.3 ... 2 2 0.0 39.4 31.6 33.5 ... 3 1 0.0 0.0 ... ... ... 3 2 0.0 0.0 ... ... ... on; C q means quantitative cycle.
139 Table 4 13. Results of psyllid transmission pathway 1 from citrus to Esenbeckia runyonii E. runyonii plant no. Assay/Months PI (Cq) 1/2 2/4 3/6 4/8 5/10 6/12 7/14 1 1 0.0 0.0 0.0 37.1 0.0 0.0 0.0 1 2 0.0 0.0 0.0 38.1 0.0 0.0 0.0 1 3 0.0 0.0 0.0 0.0 0.0 0.0 1 4 0.0 0.0 0.0 0.0 0.0 0.0 1 5 0.0 0.0 0.0 0.0 0.0 0.0 1 6 0.0 0.0 0.0 0.0 0.0 0.0 2 1 0.0 38.3 37.2 36.6 0.0 0.0 0.0 2 2 37.8 36.7 35.7 0.0 0.0 0.0 0.0 2 3 0.0 0.0 0.0 0.0 0.0 0.0 2 4 0.0 0.0 0.0 0.0 0.0 0 .0 2 5 0.0 0.0 0.0 0.0 0.0 0.0 2 6 0.0 0.0 0.0 0.0 0.0 0.0 3 1 0.0 0.0 0.0 39.3 0.0 0.0 0.0 Table 4 14. Results of psyllid transmission pathway 2 from E. runyonii to citrus Citrus plant no. Assay 1 2 month PI (Cq) Assay 2 3 month PI (Cq) Assay 3 4 month PI (Cq) Assay 4 5 month PI (Cq) 1 1 38.5 0.0 37.7 37.6 1 2 38.8 37.1 0.0 38.2 2 1 37.4 38.2 ... ... 2 2 38.4 39.2 ... ... 2 3 39.2 38.7 ... ... 3 1 39.1 37.6 ... ... 3 2 38.7 39.2 ... ... 3 3 39.5 38.9 ... ... Table 4 15. Results of psylli d transmission pathway 3 from E. runyonii to E. runyonii E. runyonii plant no. Assay 1 2 month PI (Cq) Assay 2 4 month PI (Cq) Assay 3 5 month PI(Cq) 1 1 0.0 38.7 37.9 2 1 0.0 0.0 ... 2 2 0.0 0.0 ... 2 3 0.0 0.0 ... 2 4 0.0 0.0 ... 2 5 0.0 0.0 ... 2 6 35.2 36.5 ... 2 7 0.0 0.0 ... 2 8 0.0 0.0 ... 3 1 0.0 0.0 ... 3 2 0.0 0.0 ... 3 3 0.0 38.2 ... 3 4 0.0 0.0 ... 3 5 0.0 0.0 ... 3 6 0.0 0.0 ... d cycle
140 Figure 4 2. Leaf s ymptoms observed on Las infected plant hosts A) C itrus Typical HLB symptoms were observed on citrus p lants inoculated with Las in this study. B) Citropsis gilletiana C) Choisya D) Zanthoxylum fagara E) Esenbeckia runyonii and F) Amyris texana Symptoms shown in B) to E) were considered suspicious symptoms since they were not consistent with qPCR resul ts. B E C D F A
141 Figure 4 3. Gel images of PCR amplification of Las with different primers A ) 1 st round Candidatus S (75 bp amplicon) and rpoB primers (111 bp amplicon). Number 1 to 6, Citropsis gilletiana Choisya ternata Citrofortunella microcarpa (Calamondin), Severinia buxifolia Zanthoxylum fagara and Esenbeckia runyonii ; M, 1 00 bp DNA marker (Invitrogen) ; negative control (one negative control for each primer set). B ) 2 nd round of PCR amplification of Candidatus S (75 bp amplicon) and rpoB primers (111 bp amplicon). Sample arrangement in each line was the same as in Figure 4 3A. Due to low bacterial population in alternative host samples and the sensitivity limit of conventional PCR, 1 st round of conventional PCR usually resulted in weak bands as shown in Figure 4 3A, so a 2 nd round of PCR (using the PCR products from the 1 st round as template) was done to obtain better visual res ults in gel and for direct sequencing afterward. C ) Colony PCR with a 2588 bp amplicon. Numbers in this figure represented the same host plant as above. Five colonies were chosen from each host plant to run the colony PCR M, 1 K b Plus DNA marker (Invitrog en) ; negative control. A 75 bp 111 bp 1 2 3 4 5 6 M 6 5 4 3 2 1 B 1 2 3 4 5 6 M 6 5 4 3 2 1 75 bp 111 bp
142 Figure 4 3. Continued. C 2588 bp 1 2 3 5 6 M 4
143 Table 4 1 6. Dynamic change of live Las population inside various graft transmitted host plants plant Aug 2010 Oct 2010 Dec 2010 Feb 2011 Mar 2011 Apr 2011 May 2011 Jun 2011 Jul 201 1 Dec 2011 Jan 2012 Feb 2012 Mar 2012 Apr 2012 Calamondin 1 0.0 0.0 0.0 34.1 ... 29.2 ... 24.9 ... 25.9 31.1 28.7 28.8 28.5 Calamondin 2 25 24 29.9 28 ... 25.3 ... 25.4 ... 30.1 37.7 32 Citrus 1 23.6 23.8 23.3 25.4 ... 24.4 ... 22.3 Citrus 2 21.8 21.9 23.9 24.8 ... 23.4 ... 21.9 ... 23.2 22.7 24.1 21.2 21.8 Sb1 ... ... 0.0 38.9 0.0 32.9 29.4 24.4 29.2 28.3 31.9 30.9 26.1 27.1 Sb2 ... ... 0.0 35.1 32.1 29.5 31.4 20.4 33.5 28.9 33.4 32.5 34.1 30.2 Table 4 17. Dynamic change of cumulative La s genome inside various graft transmitted host plants plant Aug 2010 Oct 2010 Dec 2010 Feb 2011 Mar 2011 Apr 2011 May 2011 Jun 2011 Jul 2011 Dec 2011 Jan 2012 Feb 2012 Mar 2012 Apr 2012 Calamondin 1 0.0 0.0 0.0 33 ... 28.2 ... 25.1 ... 24.5 30.3 26.7 27 26.9 Calamondin 2 22.9 23.9 29.4 26.9 ... 24.5 ... 25.7 ... 29.2 36.9 30.1 Citrus 1 21.9 23 22.1 24 ... 22.9 ... 21 Citrus 2 20.6 21.3 22.6 24 ... 22 ... 21.6 ... 22.6 22.2 22.2 19.4 21.1 Sb1 ... ... 0.0 38.4 0.0 29.6 26.6 24.2 28 26.3 30.8 29. 5 24 25.5 Sb2 ... ... 0.0 33.1 32.1 27.9 29.6 25 34.8 25.5 31.4 30.3 22.3 26.7 qPCR assay data point.
144 Table 4 18. Summary of H LB transmission experiments on all the tested citrus relatives Plant Grafting (%) Psyllid transmission Citrus to AH (%) AH to Citrus (%) AH to AH (%) Severinia buxifolia 100 (2/2) a 100 (12/12) 46.4 (13/28) 75 (12/16) Calamondin 22.2 (2/9) 66.7 (6/9 ) 87.5 (7/8) 88.9 (8/9) Citropsis gilletiana 50 (6/12) 100 (7/7) 21.4 (3/14) Choisya spp. 90.9 (10/11) 100 (15/15) 76.4 (13/17) Zanthoxylum fagara 100 (10/10) 92.3 (12/13) 60 (3/5) Esenbeckia runyonii 38.5 (5/13) 100 (8/8) 20 (3/15) Amyris tex ana Table 4 19. Summary of Las persistency and psyllid positive rate on all the tested citrus relatives Plant Las population in plants (Cq) Las duration (months) % Psyllids with Las Las population from psyllids (Cq) Severinia buxifolia 20 39 >3 6 70 100 16 30 Calamondin 21 39 >30 80 100 19 34 Citropsis gilletiana 30 39 >14 80 100 16 34 Choisya spp. 32 39 >13 30 100 20 38 Zanthoxylum fagara 31 39 3 12 20 40 17 32 Esenbeckia runyonii 35 39 <8 Amyris texana Note: a Las positive pla nts out of total plants inoculated; Cq means quantitative cycle ; AH means alternative host.
145 CHAPTER 5 SUMMARY AND FUTURE P ERSPECTIVES In this study of Quantitative determination of selective alternative hosts of Las and potential for transmission to 8 citrus relatives in the Rutaceae family, i.e. Severinia buxifolia x Citrofortunella microcarpa Citropsis gilletiana Esenbeckia runyonii Zanthoxylum fagara Choisya aztec Pearl Choisya ternata Sundance and Amyris texana were tested in the greenhouse as alternative host s for Las With repeated psyllid and graft inoculation trials, 7 species were experimentally demonstrated as alternative hosts for Las of which 6 are reported for the 1 st time However, the Las transmissibility back to c itrus varied due to the Las persistenc e and psyllid activities observed o n the plants, which meant different levels of risk for transmission of Las could be assigned to them In the future, more work on Las alternative hosts needs to be done in the greenh ouse, because observations in the field are not reliable. For example, two native Zanthoxylum plants including Z. fagara were considered no n (or poor) hosts of the psyllid vector based on field observations in Florida, but from this study, Z. fagara was de monstrated to be an alternative host of Las with medium high transmissibility to citrus. In order to understand the live Las population in the hosts, a novel technique, PMA qPCR, was optimi zed to work with uncultured Las in plant and psyllid samples, a fter which it was used to determine the live bacterial genome in various investigations. First, a standard curve showing the correlation between qPCR results and microscopic counting of Las cells was established, with which the live bacterial genome could be determined Second, after testing a large collection of citrus plants with PMA qPCR, it was demonstrated that there was no strong co rrelation between plant symptoms and
146 internal live bacterial genome because different citrus species and cultivars all a ccumulate d high level of live/ cumulative Las genome Third, live Las population s in symptomatic and asymptomatic leaves were compared, and threshold region of live bacterial population for HLB symptom expression in Hamlin sweet orange was found Lastly, li ve and cumulative bacterial genome dynamics were monitored in citrus and non citrus hosts through a 20 month period, and a seasonal development pattern for live Las was observed in both plant hosts, in contrast to the random fluctuation of cumulative Las g enome s observed from the same plants. In the future, new HLB research concerning the live bacterial genome in the hosts could be conducted with this optimized PMA qPCR method F or example, the effectiveness of any new chemotherapy targeting the systemic ba cteria could be evaluated using this method. T his method thus offers a new way to investigate various aspects of Las and HLB until Las is cultured Regarding the seasonal development pattern of live Las found in both citrus and non citrus hosts, furthe r investigations could be conducted to check the molecular mechanism behind the sudden change that occurred in April. One possibility is the phage elements found in Las genome may enter a ly sogenic cycle around April, which le ads to a dramatic change of th e bacterial population Using phage specific primers to test the DNA samples from those months could be a good way to start. In addition, some HLB samples showing typical HLB symptoms but qPCR negative with the 16S system suggested that there may be an unk nown liberibacter present in Florida. Plants with this situation (typical HLB symptom but negative PCR results) are excellent study materials to identify the new bacterium.
147 APPENDIX SEQUENCES AND BLAST RESULTS A 1 2588 bp S equence from T7 1/Sp6 1 P rimers and BLASTn R esults AGTTTGATCCTGGCTCAGAACGAACGCTGGCGGCAGGCCTAACACATGCACG TCGAGCGCGTATGCGAATA CG AGCGGC AGACGGGTGAGTAACGCG TAGGAAT CTACCTTTTTCTACGGGATAACGC ATGGAAACGTGTGC TAATACCGTATACGCCCTATTGGGGGAAAGATTTTATTGGAGAGAGATGAGCCTGCGTTGGATTAGCTAGT TGGTAGGG TAAGAGCCTACCAAGGCTACGATCTATAGCTGGTCTGAGAGGACGATCAGCCACACTGGGACT GAGACACGGCCCAGACTCCTACGGGAGGCGGCAGTGGGGAATATTGGACAATGGGGGCAACCCTGATCCAG CCATGCTGCGTGAGTGAAGAAGGCCTTAGGGTTGTAAAGCTCTTTCGCCGGAGAAGATAATGACGGTATTC GGAGAAGAAGCCCCGGCTAACTTCGTGCCAGCAGCCGCGGTAATACGAAGG GGGCGAGCGTTGTTCGGAAT AACTGGGCGTAAAGGGCGCGTAGGCGGGCGATTAAGTTAGAGGTGAAATCCCAGGGCTCAACCTTGGAACT GCCTTTAATACTGGTTGTCTAGAGTTAGGAGAGGTGAGTGGAATTCCGAGTGTAGAGGTGAAATTCGTAGA TATTCGGAGGAACACCGGTGGCGAAGGCGGCTCACTGGCCTGATACTGACGCTGAGGCGCGAAAGCGTGGG GAGCAAACAGGATTAGATACCCT GGTAGTCCACGCCGTAAACGATGAGTGTTAGCTGTTGGGTGGTTTACC ATTCAGTGGCGCAGCTGACGCATTAAGCACTCCGCCTGGGGAGTACGGTCGCAAGATTAAAACTCAAAGGA ATTGACGGGGGCCCGCACAAGCGGTGGAGCATGTGGTTTAATTCGATGCAACGCGCAGAACCTTACCAGCC CTTGACATGTATAGGACGATATCAGAGATGGTATTTTCTTTTCGGAGACCTTTACACAGGTGCTGC ATGGC TGTCGTCAGCTCGTGTCGTGAGATGTTGGGTTAAGTCCCGCAACGAGCGCAACCCCTGCCTCTAGTTGCCA TCAAGTTTAGGTTTTTACCTAGATGTTGGGTACTTTATAGGGACTGCCGGTGATAAGCCGGAGGAAGGTGG GGATGACGTCAAGTCCTCATGGCCCTTATGGGCTGGGCTACACACGTGCTACAATGGTGGTTACAATGGGT TGCGAAGTCGCGAGGCGGAGCTAATCCCCAAAAGCCAT CTCAGTTCGGATTGCACTCTGCAACTCGAGTGC ATGAAGTCGGAATCGCTAGTAATCGCGGATCAGCATGCCGCGGTGAATACGTTCTCGGGCCTTGTACACAC CGCCCGTCACACCATGGGAGTTGGTTTTGCCTGAAGACGGTGTGCTAACCGCAAGGAGGCAGCCGGCCACG GTAGGGTCAGCGACTGGGGTGAAGTCGTAACAAGGTAGCCGTAGGGGAACCTGCGGCTGGATCACCTCCTT TCTAAGGAAG ATGTTGAGTATCATTGAATTTATTGAGTGATCTGAACGTTTTTTGAAGATTAAAGCTTTTA ATTAAGCTTGATATAAATTTGCTGTTGTGAAGCAGCGTTTTTAAAAGGATCGCCGTCCATGTTTCTCTTTC TTTTCGGATTTTTTGCGATGATGGGGGGTCGTTAATATTTGGTTTTGAGGGCCCGTAGCTCAGGTGGTTAG AGTGCACCCCTGATAAGGGTGAGGTCGGTAGTTCAAATCTACCTGGGCCCACC ACTTTTTGTTCAGGGGGC CGTAGCTCAGTCGGTAGAGCGCCTGCTTTGCAAGCAGGATGCCAGCGGTTCGATTCCGCTCGGCTCCACCA TTGGCGTAATTATGGAATTTTGTTATGATTTTTTGGAGCAAGAGTTTTTTGAAAATTGAATAGAAGATATA TTTTTTTGTATTTTTTATGTTGGCATTGTATGCGACATATAAGATACCGGCGTTGTTAACCGCACGTTGAG AATTTATCTCAGGAAATTGGTCTAT TGAAGAAGCATAAGGATATTATGTTTTTTTAATTATAAAGAGTTTG CAAAGAACTTTATGACGATTGACAATGAGAGTAATCAAGCGCGAAAGGGCATTTGGTGGATGCCTTGGCAT GCACAGGCGATGAAGGACGTGATATGCTGCGATAAGCTATGGGGAGCTGCAAATAAGCATTGATCCGTAGA TTTCCGAATGGGGTAACCCGCCTTATGTGCCTAGGAAACTGAACTAAGTTGGTTTAATTTTCTAGGTA TTT TGAAGGTATCTTTATCTGAATGAAACAGGGTAAAAGAAGCGAACGCAGGGAACTGAAACATCTAAGTACCT GTAGGAAAGGACATCAATTGAGACTCCGTTAGTAGTGGCGAGCGAACGCGGATCAGGCCAGTGGTAGGAAA GATTTAAGTAGAATTACCTGGGAAGGTAAGCCATAGTGTGTGATAGCCCCGTATACGTAATAATTTTTTCT ATCCTTGAGTAGGGCGGGACACGTGAAATCCTGTCTGAAT ATGGGGCGACCACGCTCCAAGCCTAAGTACT CGTGCATGACCGATAGTGAACCAGTACCGTGA
149 A 2 75 bp S equence from HLBas/r P rimers and BLASTn Re sults TCGAGCGCGTATGCAATACGAGCGGCAGACGGGTGAGTAACGCGTAGGAATCTACCTTTTTCTACGGGATAACGC
150 LIST OF REFERENCES Akula, N., Triv edi, P., and Wang, N. 2011a. Identification of small molecule inhibitors against SecA of Candidatus Liberibacter asiaticus by molecular modeling studies. Phytopathology 101:S4. Akula, N., Zheng, H., Han, F., and Wang, N. 2011b. Discovery of novel SecA inhi bitors of Candidatus Liberibacter asiaticus by structure based design. Bioorganic & Medicinal Chemistry Letters 21:4183 4188. Albrecht, U., and Bowman, K. 2008. Gene expression in Citrus sinensis (L.) Osbeck following infection with the bacterial pathogen Candidatus Liberibacter asiaticus causing Huanglongbing in Florida. Plant Science 175:291 306. Albrecht, U., and Bowman, K. 2009. Candidatus Liberibacter asiaticus and Huanglongbing Effects on Citrus Seeds and Seedlings. Hortscience 44:1967 1973. Albrecht, U., and Bowman, K. 2011. Tolerance of the Trifoliate Citrus Hybrid US 897 ( Citrus reticulata Blanco x Poncirus trifoliata L. Raf.) to Huanglongbing. Hortscience 46:16 22. Albrecht, U., and Bowman, K. 2012. Transcriptional response of susceptible and toler ant citrus to infection with Candidatus Liberibacter asiaticus. Plant Science 185:118 130. A ubert B. 1987. T rioza erytreae Del Guercio and Diaphorina citri Kuwayama (Homoptera: Psylloidea), the two vectors of citrus greening disease: Biological aspects an d possible control strategies. Fruits 42:149 162. Bastianel, C., Garnier Semancik, M., Renaudin, J., Bove, J., and Eveillard, S. 2005. Diversity of Candidatus Liberibacter asiaticus," based on the omp gene sequence. Applied and Environmental Microbiology 71:6473 6478. Belasque, J., Bergamin, A., Bassanezi, R., Barbosa, J., Fernandes, N., Yamamoto, P., Lopes, S., Machado, M., Leite, R., Ayres, A., and Massari, C. 2009. Scientific basis for the eradification of syntomatic and asymptomatic plants in Huanglong bing (HLB, Greening) aimed at effective disease control. Tropical Plant Pathology 34:137 145. Belasque, J., Bassanezi, R., Yamamoto, P., Ayres, A., Tachibana, A., Violante, A., Tank, A., Di Giorgi, F., Tersi, F., Menezes, G., Dragone, J., Jank, R., and Bov e, J. 2010. L essons from Huanglongbing management in Sao Paulo State, Brazil. Journal of Plant Pathology 92:285 302.
151 B hagabati K. N. 1993. Survey of greening disease of mandarin orange in the northeastern states of India, pp. 441 442 In P. Moreno, J. V. d a Graa, and L. W. Timmer [eds.], Proc. 12th Conference of the International Organization of Citrus Virologists. University of California, Riverside. Bove, J. 2006. Huanglongbing: A destructive, newly emerging, century old disease of citrus. Journal of Pla nt Pathology 88:7 37. Bove, J., and Ayres, A. 2007. Etiology of three recent diseases of citrus in Sao Paulo State: Sudden death, variegated chlorosis and huanglongbing. I UBMB Life 59:346 354. Bowman, K., McCollum, G., Niedz, R., Stover, E., Albrecht, U., and Stange, R. 2009. Development of Citrus Cultivars Resistant to Huanglongbing Disease by Transformation with AMP Genes. Hortscience 44:1012. Boyd, E., and Brussow, H. 2002. Common themes among bacteriophage encoded virulence factors and diversity among t he bacteriophages involved. Trends in Microbiology 10:521 529. Boyd, E., Davis, B., and Hochhut, B. 2001. Bacteriophage bacteriophage interactions in the evolution of pathogenic bacteria. Trends in Microbiology 9:137 144. Brlansky, R. H., and Rogers, M. E. 2007. Citrus Huanglongbing: Understanding the vector pathogen interaction for disease management. Plant Health Progr. doi: 10.1094/APSnetFeature 2007 1207. Published online. Casjens, S. 2003. Prophages and bacterial genomics: what have we learned so far? Molecular Microbiology 49:277 300. Cawthorn, D., and Witthuhn, R. 2008. Selective PCR detection of viable Enterobacter sakazakii cells utilizing propidium monoazide or ethidium bromide monoazide. Journal of Applied Microbiology 105:1178 1185. Chamberlain, H. and Irey, M 2008. Comparison of a starch based field test for Huanglongbing to results from real time PCR testing of field samples from symptomatic trees in Florida. Page 130 in Proceeding of the International Research Conference on Huanglongbing, Orl ando, Florida. Chen, J., Deng, X., Liu, S., Pu, X., Li, H., and Civerolo, E. 2008. Detection of phytoplasma and Candidatus Liberibacter asiaticus in citrus showing Huanglongbing (yellow shoot disease) symptoms in Guangdong, P.R. China. Phytopathology 98:S 35.
152 Chen, J., Pu, X., Deng, X., Liu, S., Li, H., and Civerolo, E. 2009. A Phytoplasma Related to Candidatus Phytoplasma asteri' Detected in Citrus Showing Huanglongbing (Yellow Shoot Disease) Symptoms in Guangdong, P. R. China. Phytopathology 99:236 242. Coletta Filho, H. D., Targon, M. L. P. N., Takita, M. A., De Negri, J. D., Pompeu, Jr., and Machado, M. A. 2004. First report of the causal agent of huanglongbing ( Candidatus Liberibacter asiaticus ) in Brazil. Plant Disease 88(12):1382. Coletta Filho, H. D., Takita, M.A., Targon, M. L. P. N., and Machado, M.A. 2005. Analysis of 16S rDNA sequences from citrus Huanglongbing bacteria reveal a Ca citrus disease in So Paulo Plant Dis ease 89:848 852. Contreras, P., Urrutia, H., Sossa, K., and Nocker, A. 2011. Effect of PCR amplicon length on suppressing signals from membrane compromised cells by propidium monoazide treatment. Journal of Microbiological Methods 87:89 95. Cubero, J., and Graham, J. H. 2002. Geneti c relationship among worldwide strains of Xanthomonas causing canker in citrus species and design of new primers for their identification by PCR. Appl ied and Environ mental Microbiol ogy 68:1257 1264. Cubero, J., and Graham, J. H. 2005. Quantitative real tim e polymerase chain reaction for bacterial enumeration and allelic discrimination to differentiate Xanthomonas strains on citrus. Phytopathology 95:1333 1340. d a Graca J V 1991. Citrus greening disease. Annual Review of Phytopathology 29 : 109 1 36. da Grac a, J.V., and Korsten, L. 2004. Citrus huanglongbing: review, present status and future strategies. Diseases of Fruits and Vegetables 1:229 245. Damsteegt, V., Postnikova, E., Stone, A., Kuhlmann, M., Wilson, C., Sechler, A., Schaad, N., Brlansky, R., and S chneider, W. 2010. Murraya paniculata and Related Species as Potential Hosts and Inoculum Reservoirs of Candidatus Liberibacter asiaticus', Causal Agent of Huanglongbing. Plant Disease 94:528 533. Davis, M., Mondal, S., Chen, H., Rogers, M., and BrIansky, R. 2008. Co cultivation of 'Candidatus Liberibacter asiaticus' with Actinobacteria from Citrus with Huanglongbing. Plant Disease 92:1547 1550. Deng, C., Chen, C., Zhao, X., Fu, H., Chen, G., Bai, X., Lou, B., and Wu, R. 2010. Nested PCR detection of the p athogen of citrus huanglongbing on Clausena lansium and Murraya paniculata Chinese Agricultural Science Bulletin 26(17):273 276.
153 Deng, X., Lou, Z., Feng, Z., Li, H., Chen, J., and Civerolo, E. 2008. First report of Candidatus Liberibacter asiaticus from Atalantia buxifolia in Guangdong, China. Plant Disease 92(2):314. Ding, F., Wang, G., Yi, G., Zhong, Y., Zeng, J., and Zhou, B. 2005. Infection of wampee and lemon by the Citrus Huanglongbing pathogen ( Candidatus Liberibacter asiaticus) in China. Journal of Plant Pathology 87:207 212. Ding, F., Deng, X., Hong, N., Zhong, Y., Wang, G., and Yi, G. 2009. Phylogenetic analysis of the citrus Huanglongbing (HLB) bacterium based on the sequences of 16S rDNA and 16S/23S rDNA intergenic regions among isolates in Ch ina. European Journal of Plant Pathology 124:495 503. Ding, F., Jin, S., Hong, N., Zhong, Y., Cao, Q., Yi, G., and Wang, G. 2008. Vitrification cryopreservation, an efficient method for eliminating Candidatus Liberobacter asiaticus, the citrus Huanglongbin g pathogen, from in vitro adult shoot tips. Plant Cell Reports 27:241 250. Duan, Y., Gottwald, T., Zhou, L., and Gabriel, D. 2008. First Report of Dodder transmission of Candidatus Liberibacter asiaticus' to Tomato ( Lycopersicon esculentum ). Plant Disease 92:831. Duan, Y., Zhou, L., Hall, D., Li, W., Doddapaneni, H., Lin, H., Liu, L., Vahling, C., Gabriel, D., Williams, K., Dickerman, A., Sun, Y., and Gottwald, T. 2009. Complete Genome Sequence of Citrus Huanglongbing Bacterium, Candidatus Liberibacter as iaticus' Obtained Through Metagenomics. Molecular Plant Microbe Interactions 22:1011 1020. Dutt, M., Ananthakrishnan, G., Jaromin, M., Brlansky, R., and Grosser, J. 2012. Evaluation of four phloem specific promoters in vegetative tissues of transgenic citr us plants. Tree Physiology 32:83 93. Etxeberria, E., Gonzalez, P., Achor, D., and Albrigo, G. 2009. Anatomical distribution of abnormally high levels of starch in HLB affected Valencia orange trees. Physiological and Molecular Plant Pathology 74:76 83. Fan J., Chen, C., Brlansky, R., Gmitter, F., and Li, Z. 2010. Changes in carbohydrate metabolism in Citrus sinensis infected with Candidatus Liberibacter asiaticus'. Plant Pathology 59:1037 1043. Fan, J., Chen, C., Yu, Q., Brlansky, R., Li, Z., and Gmitter, F. 2011. Comparative iTRAQ proteome and transcriptome analyses of sweet orange infected by Candidatus Liberibacter asiaticus". Physiologia Plantarum 143:235 245. Flekna, G., Stefanic, P., Wagner, M., Smulders, F., Mozina, S., and Hein, I. 2007. Insuffici ent differentiation of live and dead Campylobacter jejuni and Listeria monocytogenes cells by ethidium monoazide (EMA) compromises EMA/real time PCR. Research in Microbiology 158:405 412.
154 Folimonova, S., and Achor, D. 2010. Early Events of Citrus Greening (Huanglongbing) Disease Development at the Ultrastructural Level. Phytopathology 100:949 958. Folimonova, S., Robertson, C., Garnsey, S., Gowda, S., and Dawson, W. 2009. Examination of the Responses of Different Genotypes of Citrus to Huanglongbing (Citrus Greening) Under Different Conditions. Phytopathology 99:1346 1354. F rench J. V., K ahlke C. J. A nd da G raca, J. V. 2001. First record of the Asian citrus psylla, Diaphorina citri Kuwayama (Homoptera: Psyllidae), in Texas. Subtro pical Plant Science 53: 14 15. Furuya, N., Matsukura, K., Tomimura, K., Okuda, M., Miyata, S., and Iwanami, T. 2010. Sequence homogeneity of the psi serA trmU tufB secE nusG rplKAJL rpoB gene cluster and the flanking regions of Candidatus Liberibacter asiaticus' isolates around O kinawa Main Island in Japan. Journal of General Plant Pathology 76:122 131. Garnier, M, and Bov e J. M. 1983. Transmission of organism associated with citrus greening disease from sweet orange to periwinkle by dodder. Phytopathology 73:1358 1363. G arnier M., A nd B ove, J. M. 1993. Citrus greening disease, pp. 212 219 In P. Moreno, J. V. da Graa, and L. W. Timmer [eds.], Proc. 12th Conference of the International Organization of Citrus Virologists. University of California, Riverside. Garnier, M., Jagoueix Eveillard, S., Cronje, P., Le Roux, H., and Bove, J. 2000. Genomic characterization of a liberibacter present in an ornamental rutaceous tree, Calodendrum capense in the Western Cape province of South Africa. Proposal of Candidatus Liberibacter africanus subsp capensis'. International Journal of Systematic and Evolutionary Microbiology 50:2119 2125. Gonzalez Segnana, L., Villalba, N., Armadans, A., and Shohara, K. 1997. Incidence of Tristeza and other citrus diseases in Paraguay. Proceedings of Florida St ate Horticultural Society 110:43 46. Gottwald, T 2010. Current epidemiological understanding of citrus huanglongbing. Annu al Review of Phytopathology 48:119 39. G ottwald, T., Aubert, B., and Zhao, X. 1989. Preliminary analysis of citrus greening (huanglu ngbin) epidemics in the People s Republic of China and French Reunion Island. Phytopathology 79:687 693. Graham, J. H., Irey, M. S. Dawson, W. O. Hall, D. and Duan Y. 2008. Assessment of transmission of Liberibacter asiaticus from seed to seedlings of orange and Carrizo citrange, p. 174. In: Proc. Intl. Res. Conf. on Huanglongbing, Orlando, FL.
155 Halbert, S. E. 2005. The discovery of Huanglongbing in Florida. Page 50 in: Proc. 2nd Int. Citrus Canker and Huanglongbing Workshop, Orlando, F L U.S.A. Halbert, S., and Manjunath, K. 2004. Asian citrus psyllids (Sternorrhyncha : Psyllidae) and greening disease of citrus: A literature review and assessment of risk in Florida. Florida Entomologist 87:330 353. Halbert, S., Manjunath, K. and Brodie M. 2008. Large scale distribution of Diaphorina citri Kuwayama and citrus huanglongbing in Florida. Page 107 in Proceeding of the International Research Conference on Huanglongbing, Orlando, Florida U.S.A. H albert, S. N iblett, C. M anjunath K., L ee R ., A nd B rown, L 2002. Establishment of two new vectors of citrus pathogens in Florida. Proc. International Soc. Citriculture IX Congress, ASHS Press, Alexandria, VA. pp. 1016 1017. Halbert, S., Manjunath, K., Ramadugu, C., and Lee, R. 2010a. Spread of hu anglongbing through citrus and citrus relatives in retail nurseries and garden centers. Phytopathology 100:S46. Halbert, S., Manjunath, K., Ramadugu, C., Brodie, M., Webb, S., and Lee, R. 2010b. T railers transporting oranges to processing plants move Asian citrus psyllids. Florida Entomologist 93:33 38. Halbert, S., and Nunez, C. 2004. Distribution of the Asian citrus psyllid, Diaphorina citri Kuwayama (Rhynchota : Psyllidae) in the Caribbean Basin. Florida Entomologist 87:401 402. Hartung, J., Shao, J., an d Kuykendall, L. 2011. Comparison of the Ca Liberibacter asiaticus' Genome Adapted for an Intracellular Lifestyle with Other Members of the Rhizobiales. Plos One 6 (8): e23289. doi:10.1371/journal.pone.0023289 Hartung, J., Paul, C., Brlansky, R., and Acho r, D. 2009. Transmission of the huanglongbing pathogen Ca liberibacter spp. from Citrus by dodder, Cuscuta indecora Choisy, to periwinkle, Catharanthus roseus G. Don. Phytopathology 99:S52. Hartung, J., Paul, C., Achor, D., and Brlansky, R. 2010a. Coloniz ation of Dodder, Cuscuta indecora by Candidatus Liberibacter asiaticus' and Ca L. americanus'. Phytopathology 100:756 762. Hartung, J., Halbert, S., Pelz Stelinski, K., Brlansky, R., Chen, C., and Gmitter, F. 2010b. Lack of Evidence for Transmission of Candidatus Liberibacter asiaticus' Through Citrus Seed Taken from Affected Fruit. Plant Disease 94:1200 1205. Hilf, M Candidatus Implications for seed transmission of the bacterium. Ph ytopathology 101:1242 1250.
156 Hocquellet, A., Bove, J., and Garnier, M. 1997. Production and evaluation of non radioactive probes for the detection of the two Candidatus Liberobacter' species associated with citrus huanglongbing (greening). Molecular and Ce llular Probes 11:433 438. Hocquellet, A., Toorawa, P., Bove, J., and Garnier, M. 1999. Detection and identification of the two Candidatus Liberobacter species associated with citrus huanglongbing by PCR amplification of ribosomal protein genes of the beta operon. Molecular and Cellular Probes 13:373 379. Hu, H., Schuenzel, E., Sechler, A., Wang, Z., and Schaad, N. 2008. Multilocus sequence typing system for Candidatus Liberibacter asiaticus, a causal agent of Huanglongbing. Phytopathology 98:S215. Hu, H., Y in, Y., Zhang, L., Zhao, Y., Xia, Y., Wang, Z., and Qin, J. 2006. Detection of citrus huanglongbing by conventional and two fluorescence quantitative PCR assays. Scientia Agricultura Sinica 39(12):2491 2497. Hung, T., Wu, M., and Su, H. 1999. Development o f a rapid method for the diagnosis of citrus greening disease using the polymerase chain reaction. Journal of Phytopathology Phytopathologische Zeitschrift 147:599 604. Hung, T., Wu, M., and Su, H. 2000. Identification of alternative hosts of the fastidiou s bacterium causing citrus greening disease. Journal of Phytopathology Phytopathologische Zeitschrift 148:321 326. Hung, T., Wu, M., and Su, H. 2001. Identification of the chinese box orange ( Severinia buxifolia ) as an alternative host of the bacterium cau sing citrus huanglongbing. European Journal of Plant Pathology 107:183 189. Hung, T., Hung, S., Chen, C., Hsu, M., and Su, H. 2004. Detection by PCR of Candidatus Liberibacter asiaticus, the bacterium causing citrus huanglongbing in vector psyllids: applic ation to the study of vector pathogen relationships. Plant Pathology 53:96 102. Inoue, H., Ohnishi, J., Ito, T., Tomimura, K., Miyata, S., Iwanami, T., and Ashihara, W. 2009. Enhanced proliferation and efficient transmission of Candidatus Liberibacter asia ticus by adult Diaphorina citri after acquisition feeding in the nymphal stage. Annals of Applied Biology 155:29 36. J agoueix, S., Bove, J., and Garnier, M. 1994. The phloem limited bacterium of greening disease of citrus is a member of the alpha subdivisi on of the Proteobacteria. International Journal of Systematic Bacteriology 44:379 386. Jagoueix, S., Bove, J., and Garnier, M. 1996. PCR detection of the two Candidatus liberobacter species associated with greening disease of citrus. Molecular and Cellul ar Probes 10:43 50.
157 Jagoueix, S., Bove, J., and Garnier, M. 1997. Comparison of the 16S/23S ribosomal intergenic regions of '' Candidatus Liberobacter asiaticum'' and '' Candidatus Liberobacter africanum,'' the two species associated with citrus huanglongbin g (greening) disease. International Journal of Systematic Bacteriology 47:224 227. Jiang, H., Wei, W., Saiki, T., Kawakita, H., Watanabe, K., and Sato M 2004. Distribution patterns of mulberry dwarf phytoplasma in reproductive organs, winter buds, and ro ots of mulberry trees. J ournal of General Plant Pathology 70 :168 173. J osephson, K., Gerba, C., and Pepper, I. 1993. Polymerase chain reaction detection of nonviable bacterial pathogens. Applied and Environmental Microbiology 59:3513 3515. Katoh, H., Suban diyah, S., Tomimura, K., Okuda, M., Su, H., and Iwanami, T. 2011. Differentiation of Candidatus Liberibacter asiaticus" Isolates by Variable Number Tandem Repeat Analysis. Applied and Environmental Microbiology 77:1910 1917. Khalaf, A., Gmitter, F., Brlan sky, R., Fan, J., and Moore, G. 2010. Gene expression profiling of two Citrus cultivars in response to huanglongbing (HLB) using the Agilent Citrus custom microarray Chip. Phytopathology 100:S61. Kim, J., Sagaram, U., Burns, J., Li, J., and Wang, N. 2009. Response of Sweet Orange ( Citrus sinensis ) to Candidatus Liberibacter asiaticus' Infection: Microscopy and Microarray Analyses. Phytopathology 99:50 57. Kobayashi, H., Oethinger, M., Tuohy, M., Hall, G., and Bauer, T. 2009a. Unsuitable distinction between viable and dead Staphylococcus aureus and Staphylococcus epidermidis by ethidium bromide monoazide. Letters in Applied Microbiology 48:633 638. Kobayashi, H., Oethinger, M., Tuohy, M., Hall, G., and Bauer, T. 2009b. Improving Clinical Significance of PCR: Use of Propidium Monoazide to Distinguish Viable from Dead Staphylococcus aureus and Staphylococcus epidermidis Journal of Orthopaedic Research 27:1243 1247. Kobayashi, H., Oethinger, M., Tuohy, M., Hall, G., and Bauer, T. 2010. Distinction between Intac t and Antibiotic Inactivated Bacteria by Real Time PCR after Treatment with Propidium Monoazide. Journal of Orthopaedic Research 28:1245 1251. K oizumi M., P rommintara M., L inwattana G., and K aisuwan, T. 1993. Field evaluation of citrus cultivars for gre ening resistance in Thailand, pp. 274 279 In P. Moreno, J. V. da Graa, and L. W. Timmer [eds.], Proc. 12th Conference of the International Organization of Citrus Virologists. University of California, Riverside.
158 K oizumi M., P rommintara M., and O htsu, Y 1996. Wood apple, Limonia acidissima : A new host for the huanglongbing (greening) vector, Diaphorina citri, pp. 271 275 In J. V. da Graa, P. Moreno, and R. K. Yokomi [eds.], Proc. 13th Conference of the International Organization of Citrus Virologists (I OCV). University of California, Riverside. K orsten L. J agoueix S. B ove J., and G arnier, M 1996. Huanglongbing (greening) detection in South Africa, pp. 395 398 In J. V. da Graa, P. Moreno, and R. K. Yokomi [eds.], Proc. 13th Conference of the Intern ational Organization of Citrus Virologists (IOCV). University of California, Riverside. Kralik, P., Nocker, A., and Pavlik, I. 2010. Mycobacterium avium subsp paratuberculosis viability determination using F57 quantitative PCR in combination with propidium monoazide treatment. International Journal of Food Microbiology 141:S80 S86. Kunta, M., da Graca, J., Setamou, M., and Skaria, M. 2010. Towards solving 'inconclusive' quantitative PCR for the presence of Huanglongbing (HLB) in orange jasmine leaf samples in Texas. Phytopathology 100:S66. Li, W., Hartung, J. and Levy L., 2006. Quantitative real time PCR for detection and identification of Candidatus Liberibacter species associated with citrus huanglongbing. Journal of Microbiological Methods 66:104 115. Li W., L i, D., Twieg, E., Hartung, J. and Levy, L., 2008. Optimized quantification of unculturable Candidatus Liberibacter spp. in host plants using real time PCR. Plant Disease 92(6):854 861. Liang, N., Dong, J., Luo, L., an d Li, Y. 2011. Detection of liv e Salmonella in Lettuce by propidium monoazide real time PCR. Journal of Food Science 76(4):M234 237. Lin, H., Chen, C., Doddapaneni, H., Duan, Y., Civerolo, E., Bai, X., and Zhao, X. 2010. A new diagnostic system for ultra sensitive and specific detection and quantification of Candidatus Liberibacter asiaticus, the bacterium associated with citrus Huanglongbing. Journal of Microbiological Methods 81:17 25. Lin K 1 956. Study on citrus yellow shooting (huanglongbing) disease. Acta Phytopathologia Sinica 2 : 1 4. Liu, Y., and Tsai, J. 2000. Effects of temperature on biology and life table parameters of the Asian citrus psyllid, Diaphorina citri Kuwayama (Homoptera : Psyllidae). Annals of Applied Biology 137:201 206. Lopes, S., and Frare, G. 2008. Graft transmi ssion and cultivar reaction of citrus to Candidatus Liberibacter americanus'. Plant Disease 92:21 24. Lopes, S., Frare, G., Yamamoto, P., Ayres, A., and Barbosa, J. 2007. Ineffectiveness of pruning to control citrus huanglongbing caused by Candidatus Libe ribacter americanus. European Journal of Plant Pathology 119:463 468.
159 Lopes, S., Bertolini, E., Frare, G., Martins, E., Wulff, N., Teixeira, D., Fernandes, N., and Cambra, M. 2009a. Graft Transmission Efficiencies and Multiplication of Candidatus Liberiba cter americanus' and Ca Liberibacter asiaticus' in Citrus Plants. Phytopathology 99:301 306. Lopes, S., Frare, G., Bertolini, E., Cambra, M., Fernandes, N., Ayres, A., Marin, D., and Bove, J. 2009b. Liberibacters Associated with Citrus Huanglongbing in B razil: Candidatus Liberibacter asiaticus' Is Heat Tolerant, Ca L. americanus' Is Heat Sensitive. Plant Disease 93:257 262. Lopes, S., Frare, G., Camargo, L., Wulff, N., Teixeira, D., Bassanezi, R., Beattie, G., and Ayres, A. 2010. Liberibacters associat ed with orange jasmine in Brazil: incidence in urban areas and relatedness to citrus liberibacters. Plant Pathology 59:1044 1053. Luis, M., Collazo, C., Llauger, R., Blanco, E., Pena, I., Lopez, D., Gonzalez, C., Casin, J., Batista, L., Kitajima, E., Tanak a, F., Salaroli, R., Teixeita, D., Martins, E., and Bove, J. 2009. O ccurrence of citrus huanglongbing in Cuba and association of the disease with Candidatus Liberibacter asiaticus. Journal of Plant Pathology 91:709 712. Manjunath, K. 2009. Huanglongbing in India. Phytopathology 99:S157. Manjunath, K., Halbert, S., Ramadugu, C., Webb, S., and Lee, R. 2007a. Use of psyllids in early identification of huanglongbing disease of citrus. Phytopathology 97:S70. Manjunath, K., Halbert, S., Ramadugu, C., Webb, S., an d Lee, R. 2008a. Detection of Candidatus Liberibacter asiaticus' in Diaphorina citri and its importance in the management of Citrus huanglongbing in Florida. Phytopathology 98:387 396. Manjunath, K., Harakava, R., Ramadugu, C., Halbert, S., Yamamoto, P., and Lee, R. 2007b. Psyllids as a tool in evaluating the efficiency of different management practices for control of citrus huanglongbing disease. Phytopathology 97:S70. Manjunath, K., Harakava, R., Ramadugu, C., Yamamoto, P., Halbert, S., and Lee, R. 2008b Monitoring management of huanglongbing disease of citrus in Brazil. Phytopathology 98:S97. Manjunath, K., Ramadugu, C., Majil, V., Williams, S., Irey, M., and Lee, R. 2010. First Report of the Citrus Huanglongbing Associated Bacterium Candidatus Liberib acter asiaticus' from Sweet Orange, Mexican Lime, and Asian Citrus Psyllid in Belize. Plant Disease 94:781. Marques, R., Teixeira, D., Yamamoto, P., and Lopes, J. 2012. Weedy Hosts and Prevalence of Potential Leafhopper Vectors (Hemiptera: Cicadellidae) of a Phytoplasma (16SrIX group) Associated With Huanglongbing Symptoms in Citrus Groves. Journal of Economic Entomology 105:329 337.
160 Martinez, Y., Llauger, R., Batista, L., Luis, M., Iglesia, A., Collazo, C., Pena, I., Casin, J., Cueto, J., and Tablada, L. 2 009. First report of Candidatus Liberibacter asiaticus' associated with Huanglongbing in Cuba. Plant Pathology 58:389 389. M asters, C., Shallcross, J., and Mackey, B. 1994. Effect of stress treatments on the detection of Listeria monocytogenes and enterot oxigenic Escherichia coli by the polymerase chain reaction. Journal of Applied Bacteriology 77:73 79. Matos, L., Hilf M. and Comejo, J. 2009. First report of Candidatus Liberibacter asiaticus infecting citrus in the Dominican Republic Plant Disease 93:6 68. Mavrodieva, V., Levy, L., and Gabriel, D. W. 2004. Improved sampling methods for Real Time polymerase chain reaction diagnosis of citrus canker from field samples. Phytopathology 94:61 68. M ead F. 1977. The Asiatic citrus psyllid, Diaphorina citri Ku wayama (Homoptera: Psyllidae). Entomology Circular 180. Florida Department of Agriculture and Consumer Services, Division of Plant Industry. 4 pp. Michaud, J. 2001. Numerical response of Olla v nigrum (Coleoptera : Coccinellidae) to infestations of Asian citrus psyllid, (Hemiptera : Psyllidae) in Florida. Florida Entomologist 84:608 612. Michaud, J. 2002a. Biological control of Asian citrus psyllid, Diaphorina citri (Hemiptera : Psyllidae) in Florida: A preliminary report. Entomological News 113:216 222. M ichaud, J. 2002b. Classical biological control: A critical review of recent programs against citrus pests in Florida. Annals of the Entomological Society of America 95:531 540. Michaud, J. 2004. Natural mortality of Asian citrus psyllid (Homoptera : Psylli dae) in central Florida. Biological Control 29:260 269. Michaud, J., and Olsen, L. 2004. Suitability of Asian citrus psyllid, Diaphorina citri, as prey for ladybeetles. Biocontrol 49:417 431. M oll J. and M artin, M. 1973. Electron microscope evidence that citrus psylla ( Trioza erytreae ) is a vector of greening disease in South Africa. Phytophylactica 5: 41 44. Nebe von Caron, G., Stephens, P., and Badley, R. 1998. Assessment of bacterial viability status by flow cytometry and single cell sorting. Journal o f Applied Microbiology 84:988 998. Nebe von Caron, G., Stephens, P., Hewitt, C., Powell, J., and Badley, R. 2000. Analysis of bacterial function by multi color fluorescence flow cytometry and single cell sorting. Journal of Microbiological Methods 42:97 11 4.
161 Nocker, A., and Camper, A. 2006. Selective removal of DNA from dead cells of mixed bacterial communities by use of ethidium monoazide. Applied and Environmental Microbiology 72:1997 2004. Nocker, A., and Camper, A. 2009. Novel approaches toward prefere ntial detection of viable cells using nucleic acid amplification techniques. Fems Microbiology Letters 291:137 142. Nocker, A., Cheung, C., and Camper, A. 2006. Comparison of propidium monoazide with ethidium monoazide for differentiation of live vs. dead bacteria by selective removal of DNA from dead cells. Journal of Microbiological Methods 67:310 320. Nocker, A., Sossa, K., and Camper, A. 2007a. Molecular monitoring of disinfection efficacy using propidium monoazide in combination with quantitative PCR. Journal of Microbiological Methods 70:252 260. Nocker, A., Sossa Fernandez, P., Burr, M., and Camper, A. 2007b. Use of propidium monoazide for live/dead distinction in microbial ecology. Applied and Environmental Microbiology 73:5111 5117. Nocker, A., Mazz a, A., Masson, L., Camper, A., and Brousseau, R. 2009. Selective detection of live bacteria combining propidium monoazide sample treatment with microarray technology. Journal of Microbiological Methods 76:253 261. Nocker, A., Richter Heitmann, T., Montijn, R., Schuren, F., and Kort, R. 2010. Discrimination between live and dead cells in bacterial communities from environmental water samples analyzed by 454 pyrosequencing. International Microbiology 13:59 65. Nogva, H., Dromtorp, S., Nissen, H., and Rudi, K. 2003. Ethidium monoazide for DNA based differentiation of viable and dead bacteria by 5 nuclease. PCR BioTechniques 810:812 813. Pelz Stelinski, K., Brlansky, R., Ebert, T., and Rogers, M. 2010. Transmission Parameters for Candidatus Liberibacter asiatic us by Asian Citrus Psyllid (Hemiptera: Psyllidae). Journal of Economic Entomology 103:1531 1541. Pietersen, G., Arrebola, E., Breytenbach, J., Korsten, L., le Roux, H., la Grange, H., Lopes, S., Meyer, J., Pretorius, M., Schwerdtfeger, M., van Vuuren, S., and Yamamoto, P. 2010. A Survey for Candidatus Liberibacter' Species in South Africa Confirms the Presence of Only Ca L. africanus' in Commercial Citrus. Plant Disease 94:244 249. P lanet, P., Jagoueix, S., Bove, J., and Garnier, M. 1995. Detection and c haracterization of the African citrus greening liberobacter by amplification, cloning, and sequencing of the rplKAJL rpoBC operon. Current Microbiology 30:137 141.
162 Pluke, R., Escribano, A., Michaud, J., and Stansly, P. 2005. Potential impact of lady beetle s on Diaphorina citri (Homoptera : Psyllidae) in Puerto Rico. Florida Entomologist 88:123 128. Ramadugu, C., Manjunath, K., Ramos, C., Halbert, S., and Lee, R. 2007. Incidence of huanglongbing in psyllids under natural conditions in Florida. Phytopathology 97:S96. Ramadugu, C., Manjunath, K., Ramos, C., Halbert, S., Webb, S., and Lee, R. 2008. Role of garden centers and retail nurseries in spreading citrus huanglongbing disease. Phytopathology 98:S129 S130. Ramadugu, C., Manjunath, K., Halbert, S., Brlansky R., Roose, M., and Lee, R. 2010. Characterization of Huanglongbing associated Candidatus Liberibacter asiaticus' from citrus relatives. Phytopathology 100:S107. Ramadugu, C., Manjunath, K., Razi, M., Khan, I., Roose, M., and Lee, R. 2011. Analysis of Ci trus Huanglongbing associated Candidatus Liberibacter strains from Pakistan. Phytopathology 101:S150. Rudi, K., Moen, B., Dromtorp, S., and Holck, A. 2005a. Use of ethidium monoazide and PCR in combination for quantification of viable and dead cells in com plex samples. Applied and Environmental Microbiology 71:1018 1024. Rudi, K., Naterstad, K., Dromtorp, S., and Holo, H. 2005b. Detection of viable and dead Listeria monocytogenes on gouda like cheeses by real time PCR. Letters in Applied Microbiology 40:301 306. Sagaram, U., Tatineni, S., Kim, J., and Wang, N. 2008. In planta distribution and quantification of Asiatic strain of citrus Huanglongbing pathogen. Phytopathology 98:S138. Sankaran, S., Ehsani, R., and Etxeberria, E. 2010. Mid infrared spectroscopy for detection of Huanglongbing (greening) in citrus leaves. Talanta 83:574 581. Saunt, J. 2000. Citrus Varieties of the World. An Illustrated Guide, second edition. Sinclair International Ltd., Norwich, England. Schuenzel, E., Sechler, A., Hu, H., Cooke, P ., Postnikova, E., Stone, A., Schneider, W., Damsteegt, V., and Schaad, N. 2008. Limited cultivation of Candidatus Liberibacter asiaticus, suspected causal agent of Huanglongbing of citrus. Phytopathology 98:S142. Sechler, A., Schuenzel, E., Cooke, P., Don nua, S., Thaveechai, N., Postnikova, E., Stone, A., Schneider, W., Damsteegt, V., and Schaad, N. 2009. Cultivation of Candidatus Liberibacter asiaticus', Ca L. africanus', and Ca L. americanus' Associated with Huanglongbing. Phytopathology 99:480 486.
163 Takushi, T., Toyozato, T., Kawano, S., Taba, S., Taba, K., Ooshiro, A., Numazawa, M., and Tokeshi, M. 2007. Scratch method for simple, rapid diagnosis of citrus huanglongbing using iodine to detect high accumulation of starch in the citrus leaves. J apanes e J ournal of Phytopathol ogy 71:3 8. Tatineni, S., Sagaram, U., Gowda, S., Robertson, C., Dawson, W., Iwanami, T., and Wang, N. 2008. In planta distribution of Candidatus Liberibacter asiaticus' as revealed by polymerase chain reaction (PCR) and real time PCR. Phytopathology 98:592 599. Teixeira, D., Ayres, J., Kitajima, E.W., Danet, J.L., Jagoueix Eveillard, S., Saillard, C., Bove, J.M. 2005a. First report of a huanglongbing like disease of citrus in Sao Paulo State, Brazil and association of a new Liberib acter species, Candidatus Liberibacter americanus with the disease. Plant Disease. 89(1): 107. Teixeira, D., Danet, J., Eveillard, S., Martins, E., Junior, W., Yamamoto, P., Lopes, S., Bassanezi, R., Ayres, A., Saillard, C., and Bove, J. 2005 b Citrus hu anglongbing in Sao Paulo State, Brazil: PCR detection of the Candidatus Liberibacter species associated with the disease. Molecular and Cellular Probes 19:173 179. Teixeira, D., Eveillard, S., Sirand Pugnet, P., Wulff, A., Saillard, C., Ayres, A., and Bo ve, J. 2008a. The tufB secE nusG rplKAJL rpoB gene cluster of the liberibacters: sequence comparisons, phylogeny and speciation. International Journal of Systematic and Evolutionary Microbiology 58:1414 1421. Teixeira, D., Saillard, C., Couture, C., Martin s, E., Wulff, N., Eveillard Jagoueix, S., Yamamoto, P., Ayres, A., and Bove, J. 2008b. Distribution and quantification of Candidatus Liberibacter americanus, agent of huanglongbing disease of citrus in Sao Paulo State, Brasil, in leaves of an affected swee t orange tree as determined by PCR. Molecular and Cellular Probes 22:139 150. Teixeira, D., Wulff, N., Martins, E., Kitajima, E., Bassanezi, R., Ayres, A., Eveillard, S., Saillard, C., and Bove, J. 2008c. A p hytoplasma closely related to the pigeon pea witches' broom phytoplasma (16Sr IX) is associated with citrus huanglongbing symptoms in the state of Sao Paulo, Brazil. Phytopathology 98:977 984. Tirtawidjaja, S. 1981. Insect, dodder and seed transmissions of citrus vein phloem degeneration (CVPD). Proceedings of the International Society of Citriculture 1:469 471. Tomimura, K., Miyata, S., Furuya, N., Kubota, K., Okuda, M., Subandiyah, S., Hung, T., Su, H., and Iwanami, T. 2009. Evaluation of Genetic Diversit y Among Candidatus Liberibacter asiaticus' Isolates Collected in Southeast Asia. Phytopathology 99:1062 1069.
164 Tomimura, K., Furuya, N., Miyata, S., Hamashima, A., Torigoe, H., Murayama, Y., Kawano, S., Okuda, M., Subandiyah, S., Su, H., and Iwanami, T. 20 10. Distribution of Two Distinct Genotypes of Citrus Greening Organism in the Ryukyu Islands of Japan. Jarq Japan Agricultural Research Quarterly 44:151 158. Trivedi, P., Sagaram, U., Kim, J., Brlansky, R., Rogers, M., Stelinski, L., Oswalt, C., and Wang, N. 2009. Quantification of viable Candidatus Liberibacter asiaticus in hosts using quantitative PCR with the aid of ethidium monoazide (EMA). European Journal of Plant Pathology 124:553 563. Tsai, C., Hung, T., and Su, H. 2008. Strain identification and di stribution of citrus Huangiongbing bacteria in Taiwan. Botanical Studies 49:49 56. Tsai, C., Su, H., Liao, Y., and Hung, T. 2006. First report of the causal agent of Huanglongbing (" Candidatus Liberibacter asiaticus") infecting kumquat in Taiwan. Plant Dis ease 90:1360. Tsai, J., and Liu, Y. 2000. Biology of Diaphorina citri (Homoptera : Psyllidae) on four host plants. Journal of Economic Entomology 93:1721 1725. Tsai, J., Wang, J., and Liu, Y. 2000. Sampling of Diaphorina citri (Homoptera : Psyllidae) on or ange jessamine in southern Florida. Florida Entomologist 83:446 459. Tsai, J., Wang, J., and Liu, Y. 2002. Seasonal abundance of the Asian citrus psyllid, Diaphorina citri (Homoptera : Psyllidae) in southern Florida. Florida Entomologist 85:446 451. Villec hanoux, S., Garnier, M., Laigret, F., Renaudin, J., and Bove, J. 1993. The genome of the non cultured, bacterial like organism associated with citrus greening disease contains the nusG rplKAJL rpoBC gene cluster and the gene for a bacteriophage type DNA po lymerase. Current Microbiology 26:161 166. Villechanoux, S., Garnier, M., Renaudin, J., and Bove, J. 1992. Detection of several strains of the bacterium like organism of citrus greening disease by DNA probes. Current Microbiology 24:89 95. Wagner, P., and Waldor, M. 2002. Bacteriophage control of bacterial virulence. Infection and Immunity 70:3985 3993. Walter, A.J., Hall, D.G., and Duan, Y.P. 2012. Low incidence of Candidatus Liberibacter asiaticus in Murraya paniculata and associated Diaphorina citri P lant Disease 96(6):827 832.
165 Wang, Q., Panis, B., Engelmann, F., Lambardi, M., and Valkonen, J. 2009. Cryotherapy of shoot tips: a technique for pathogen eradication to produce healthy planting materials and prepare healthy plant genetic resources for cryop reservation. Annals of Applied Biology 154:351 363. Wang, Z., Yin, Y., Hu, H., Yuan, Q., Peng, G., and Xia, Y. 2006. Development and application of molecular based diagnosis for Candidatus Liberibacter asiaticus', the causal pathogen of citrus huanglongbi ng. Plant Pathology 55:630 638. Westbrook, C., Hall, D., Stover, E., Duan, Y., and Lee, R. 2011. Colonization of Citrus and Citrus related Germplasm by Diaphorina citri (Hemiptera: Psyllidae). Hortscience 46:997 1005. Wulff, N., Eveillard, S., Foissac, X., Ayres, A., and Bove, J. 2009. rRNA operons and genome size of Candidatus Liberibacter americanus', a bacterium associated with citrus huanglongbing in Brazil. International Journal of Systematic and Evolutionary Microbiology 59:1984 1991. Yanez, M., Nock er, A., Soria Soria, E., Murtula, R., Martinez, L., and Catalan, V. 2011. Quantification of viable Legionella pneumophila cells using propidium monoazide combined with quantitative PCR. Journal of Microbiological Methods 85:124 130. Zhang, M., Duan, Y., an d Powell, C. 2010a. Bactericidal activities of antimicrobial molecules against huanglongbing associated Candidatus Liberibacter asiaticus' in the diseased periwinkle. Phytopathology 100:S145. Zhang, M., Duan, Y., Zhou, L., Turechek, W., Stover, E., and Po well, C. 2010b. Screening Molecules for Control of Citrus Huanglongbing Using an Optimized Regeneration System for Candidatus Liberibacter asiaticus' Infected Periwinkle ( Catharanthus roseus ) Cuttings. Phytopathology 100:239 245. Zhang, M., Powell, C., Gu o, Y., Doud, M. and Duan, Y. 2012. A graft based chemotherapy method for screening effective molecules and rescuing huanglongbing affected citrus plants. Phytopathology 102:567 574. Zhang, M., Powell, C., Zhou, L., He, Z., Stover, E., and Duan, Y. 2011a. Chemical Compounds Effective Against the Citrus Huanglongbing Bacterium Candidatus Liberibacter asiaticus' In Planta. Phytopathology 101:1097 1103. Zhang, S., Flores Cruz, Z., Zhou, L., Kang, B., Fleites, L., Gooch, M., Wulff, N., Davis, M., Duan, Y., and Gabriel, D. 2011b. Ca Liberibacter asiaticus' Carries an Excision Plasmid Prophage and a Chromosomally Integrated Prophage That Becomes Lytic in Plant Infections. Molecular Plant Microbe Interactions 24:458 468. Zhou, L., Gabriel, D., Duan, Y., Halbert, S., and Dixon, W. 2007. First report of dodder transmission of Huanglongbing from naturally infected Murraya paniculata to citrus. Plant Disease 91:227.
166 BIOGRAPHICAL SKETCH Hao Hu was born and raised in Huangshi Hu bei China. H e received h is degree majoring in b ioengineering from Chongqing University Chongqing, China H e received h is master s degree majoring in m icrobiology from Chongqing Uni versity. In 200 8 Hu joined the Department of Plant Pathology at the University of Florida for h is Ph. D and successfully completed h is program in August 201 2