This is an exploration of the potential for a citizen science project, with the goal to get the general public involved in microbial soil biodiversity around Uppsala, Sweden. Biodiversity serves an important role in how an ecosystem performs and functions. A large part of Earth's biodiversity exists below ground in soil, where microorganisms interact with plants. It would be beneficial to analyse the abundance and spread of some microorganisms in order to gain a better understanding of soil biodiversity. We suggest that one species family to study could be Phytophthora. Phytophthora is a genus of oomycetes that often are pathogenic, causing disease in various trees and other plants. It is unknown exactly how widespread the genus is today, making it extra interesting for the proposed study. For the general public to be able to do this a device needs to be developed that is easy to use and preferably could be used directly in the field. An isothermal amplification method is suitable for identifying the microorganism under these conditions. Many isothermal amplification methods are expensive, perhaps too expensive for a citizen science study, but have great potential for easy field testing. We propose a device utilizing RPA and lateral flow strips. RPA - Recombinase Polymerase Amplification is a method for amplification that might be suitable since it is simple, sensitive, and has a short run time. It is however expensive, which is an issue, but isothermal amplifications are expensive across the board. Lateral flow strips can be used to visualize the results. They utilize antibodies to detect the previously amplified amplicons, and give a positive or negative test answer that would be understandable to even untrained study participants. One of the biggest obstacles identified in this project concerns amplifying DNA from a soil sample, because an extraction step is necessary. The methods we have identified for extraction are not performable in the field, since they require centrifugation. In the proposition for a device a possible work-around for this is proposed, but since it has yet to be tested it is not yet known whether it will work or not.

Quantification of pathogen load, although challenging, is of paramount importance for accurate diagnosis and clinical management of a range of infectious diseases in a point-of-need testing (PONT) scenario such as in resource-limited settings. We formulated a quantification approach to test the standard-curve based absolute quantification ability of isothermal recombinase polymerase amplification (RPA) assay. As a test of principle, a 10-fold dilution series of Leishmania donovani (LD) genomic DNA prepared in nuclease-free-water (NFW), and from culture-spiked-blood (CSB) were tested, and a 15 min assay was performed. A modified algorithm was formulated to derive the detection outcome. The threshold-record times (Tr) in seconds thus obtained were plotted against the initial load of parasite genomes for log-linear regression analysis. The quantitative RPA (Q-RPA) assay was further evaluated against a LD quantitative (q)-PCR assay with DNA extracted from visceral and post-Kala-azar dermal leishmaniasis case specimens and stratified into different ranges of threshold cycle (Ct). The best-fitted regression models were found linear with mean r2/root mean square error (RMSE) values of residual points (in seconds) estimated as 0.996/8.063 and 0.992/7.46 for replicated series of NFW and CSB, respectively. In both series, the lower limit of detection reached less than 0.1 parasite genome equivalent DNA. Absolute agreement between Q-RPA and LD-qPCR was found for test positivity, and strong positive correlations were observed between the Tr and Ct values (r = 0.89; p < 0.0001) as well as between the absolute parasite loads (r = 0.87; p < 0.0001) quantified by respective assays. The findings in this very first Q-RPA assay for leishmaniasis are suggestive of its potential in monitoring LD load in clinical specimens, and the development of rapid Q-RPA assays for other infectious diseases.

Yaws is a skin debilitating disease caused by Treponema pallidum subspecies pertenue with most cases reported in children. World Health Organization (WHO) aims at total eradication of this disease through mass treatment of suspected cases followed by an intensive follow-up program. However, effective diagnosis is pivotal in the successful implementation of this control program. Recombinase polymerase amplification (RPA), an isothermal nucleic acid amplification technique offers a wider range of differentiation of pathogens including those isolated from chronic skin ulcers with similar characteristics such as Haemophilus ducreyi (H. ducreyi). We have developed a RPA assay for the simultaneous detection of Treponema pallidum (T. pallidum) and H. ducreyi (TPHD-RPA). The assay demonstrated no cross-reaction with other pathogens and enable detection of T. pallidum and H. ducreyi within 15 min at 42 °C. The RPA assay was validated with 49 clinical samples from individuals confirmed to have yaws by serological tests. Comparing the developed assay with commercial multiplex real-time PCR, the assay demonstrated 94% and 95% sensitivity for T. pallidum and H. ducreyi, respectively and 100% specificity. This simple novel TPHD-RPA assay enables the rapid detection of both T. pallidum and H. ducreyi in yaws-like lesions. This test could support the yaws eradication efforts by ensuring reliable diagnosis, to enable monitoring of program success and planning of follow-up interventions at the community level.

京都大学 / 新制・課程博士 / 博士(農学) / 甲第25357号 / 農博第2623号 / 新制||農||1109(附属図書館) / 京都大学大学院農学研究科食品生物科学専攻 / (主査)教授 保川 清, 教授 井上 和生, 教授 谷 史人 / 学位規則第4条第1項該当 / Doctor of Agricultural Science / Kyoto University / DFAM

No description available.

Feline coronavirus (FCoV) is endemic in cat populations worldwide. Persistently, subclinically infected cats play a significant role in spreading the infection. Testing fecal samples of cats may facilitate efforts to decrease the viral burden within a population. Real-time RT-PCR is highly sensitive and specific for the detection of FCoV but must be performed in a fully equipped laboratory. A simple and accurate assay is needed to identify FCoV at the point-of-need. The aim of this study was to develop a rapid FCoV detection assay based on isothermal amplification technology, i.e., reverse transcription-recombinase polymerase amplification (RT-RPA). Primers were designed to target the highly conserved 3′ untranslated region of the 7b gene. Running on a constant temperature of 42 °C, reverse transcription as well as DNA amplification and detection was achieved in a maximum of 15 min. A probit analysis revealed a detection limit of 58.5 RNA copies/reaction. For cross-detection, nucleic acids from 19 viruses were tested. Both RT-RPA and real-time RT-PCR showed cross-detection with canine coronavirus and transmissible gastroenteritis virus, but not with other pathogens. To evaluate clinical performance, RNA was extracted from 39 fecal samples from cats. All samples were tested simultaneously with real-time RT-PCR resulting in a RT-RPA sensitivity and specificity of 90.9% and 100%, respectively. RT-RPA can be considered a promising simple method for rapid detection of FCoV.

To detect Post-kala-azar leishmaniasis (PKDL) cases, several molecular methods with promising diagnostic efficacy have been developed that involve complicated and expensive DNA extraction methods, thus limiting their application in resource-poor settings. As an alternative, we evaluated two rapid DNA extraction methods and determined their impact on the detection of the parasite DNA using our newly developed recombinase polymerase amplification (RPA) assay. Skin samples were collected from suspected PKDL cases following their diagnosis through national guidelines. The extracted DNA from three skin biopsy samples using three different extraction methods was subjected to RPA and qPCR. The qPCR and RPA assays exhibited highest sensitivities when reference DNA extraction method using Qiagen (Q) kit was followed. In contrast, the sensitivity of the RPA assay dropped to 76.7% and 63.3%, respectively, when the boil & spin (B&S) and SpeedXtract (SE) rapid extraction methods were performed. Despite this compromised sensitivity, the B&S-RPA technique yielded an excellent agreement with both Q-qPCR (k = 0.828) and Q-RPA (k = 0.831) techniques. As expected, the reference DNA extraction method was found to be superior in terms of diagnostic efficacy. Finally, to apply the rapid DNA extraction methods in resource-constrained settings, further methodological refinement is warranted to improve DNA yield and purity through rigorous experiments.

An unprecedented use of high-throughput sequencing for routine monitoring of SARS-CoV-2 viruses in patient samples has created a dataset of over 6 million SARS-CoV-2 genomes. To monitor genomes, deposited in the GISAID database, and to track the continuous sequence evolution of molecular assay oligonucleotide target sequences. A simple pipeline tool for non-experts was developed to mine this database for nucleotide changes in oligonucleotides and tested with the long oligonucleotides of a Recombinase polymerase amplification (RPA) assay targeting the RNA-dependent RNA polymerase (RdRP) gene of the SARS-CoV-2. Results indicate the emergence of a single nucleotide change in the reverse oligonucleotide from 0.03 to 26.23% (January to May 2021) in Alpha variant genomes, which however reduced to 17.64% by September after which the Alpha variant was completely displaced by the Delta variant. For all other variants, no relevant nucleotide changes were observed. The oligonucleotide screening pipeline allows efficient screening of nucleotide changes in oligonucleotides of all sizes in minutes.

Nile tilapia, Oreochromis niloticus, is one of the most important farmed fish globally. One of the most serious bacterial diseases constraining global tilapia production is Francisellosis caused by Francisella noatunensis subsp. orientalis (Fno). Although outbreaks of Fno are increasing worldwide, there are no licenced commercial vaccines to prevent the disease for use on tilapia farms. Thus, the current treatment of choice is the use of antibiotics combined with increasing water temperature up to 30°C. Studies investigating the diversity of circulating Fno isolates and the immune response of tilapia elicited by vaccination against piscine francisellosis are lacking. In addition, the current conventional and molecular tools used for detection of Fno have many drawbacks, making detection of Fno a challenging process. In this study, five clinical isolates of Fno from diverse geographical locations (UK, Costa Rica, Mexico, Japan and Austria), previously characterised by morphology, genotype, antimicrobial susceptibility and virulence, were used in a proteomic study. The whole proteomic cell profile of the five isolates were homogenous by one-dimension sodium dodecyl polyacrylamide gel electrophoresis (1D-SDS-PAGE), while minor differences in the intensity of 15 proteins between the strains were observed by two-dimension SDS-PAGE (2DE), including some important virulence related proteins. The UK isolate was the most significantly different isolate when compared to the other Fno isolates in the current study. The Fno UK isolate had significantly higher abundance of 10/15 of the significantly expressed proteins including four of the essential pathogenicity and virulence related proteins (IglC, GroEL, DnaK, ClpB) compared to the other used Fno isolates. The antigenic profiles of the five Fno isolates were studied by 1D western blotting using tilapia hyper immune sera which recognised an immunodominant band of a molecular weight of ~ 17-28 kDa in all tested Fno isolates. Liquid chromatography-electrospray ionization-tandem mass spectrometry (LC/ESI/MS/MS) identified 47 proteins in this antigenic band. Some of the identified proteins are associated with Fno pathogenicity. 2D western blot analysis of the vaccine isolate (Fno UK) revealed differential antigen recognition between sera from vaccinated and non-vaccinated fish following experimental challenge (26 antigenic spots recognised by sera from vaccinated fish; 31 antigenic spots recognised by sera from vaccinated and challenged fish and 30 antigenic spots recognised by non-vaccinated and challenged fish). The identity of these proteins was determined by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) and some of them are known Francisella virulence related proteins. Bioinformatics analyses revealed diverse categories of proteins with high biological functions, however the vast majority of these proteins are involved in energy production and metabolic pathways of the bacteria. This detailed analysis will facilitate the development of cross-strain protective subunit Fno vaccines and antigen-targeted Fno diagnostics. The outer membrane proteins (OMPs) of the same five Fno isolates were extracted using the ionic detergent sarkosyl. The OMP fraction of the different isolates were separated via 1D-SDS PAGE and the digested peptides of the UK isolate were analysed by LC/ESI/MS/MS. High degree of similarity was observed in the OMP profile of the five Fno isolates with an abundant protein band at 17-28 kDa, which was found to be antigenic by 1D western blot using convalescent tilapia sera. LC/ESI/MS/MS analysis of the OMPs of the Fno UK isolate identified 239 proteins, including 44 proteins in the antigenic band (17-28 kDa). Comparison between the proteins identified in the immunogenic band of whole cell lysate and OMP fraction of the Fno UK isolate showed 30 common proteins between the two preparations, 17 proteins were identified only in the whole cell extract and 14 were identified only in OMP fraction. Outer membrane proteins (e.g. Omp-A), virulence related proteins such (e.g. IglC) and other stress related proteins (e.g. AhpC/TSA family peroxiredoxin) were more abundant in the OMP fraction than the whole cell lysate. In silico analysis enabled prediction of the function and location of the OMPs identified by Mass-spectrometry. The findings of this study provide preliminary data on bacterial surface proteins that exist in direct contact with the host immune defence during infection and offering an insight into their potential role as novel targets for Fno diagnostics and vaccine development. The efficacy of an injectable whole cell oil-adjuvanted vaccine was evaluated against challenge with heterologous Fno isolates in Nile tilapia, Oreochromis niloticus. Three duplicate groups of 130 healthy Nile tilapia (~15 g) were intraperitoneally (i.p.) injected with the vaccine, adjuvant-alone or PBS followed by an i.p. challenge with three Fno isolates from geographically distinct locations. The vaccine provided significant protection to all immunised tilapia groups with a significantly higher relative percent survival (RPS) of 82.3% against homologous challenge, compared to 69.8% and 65.9% after heterologous challenge. Protection correlated with significantly elevated specific antibody responses and western blot analysis demonstrated cross-isolate antigenicity with sera from fish post-vaccination and post-challenge. Moreover, a significantly lower bacterial burden was detected by quantitative real-time polymerase chain reaction (qPCR) in conjunction with significantly greater expression of IgM, IL-1β, TNF-a and MHCII 72 hours post-vaccination (hpv) in spleen samples from vaccinated tilapia compared to those of adjuvant-alone and control fish. The latter results suggested stimulation of protective immune responses following vaccination. In addition, a whole cell formalin killed autogenous immersion vaccine against Fno was developed using the same isolate used for the injectable vaccine. Duplicate tanks of 35 tilapia fry were immersed in the vaccine or in sterile Modified Muller Hinton broth (MMHB) diluted in tank water (1:10 dilution) for 30 s and at 30 days post-vaccination (dpv), all fish groups were immersion challenged with the homologous Fno isolate and monitored for 21 days. A moderate RPS of 43.7% was provided by the vaccine. Serum IgM levels were below the threshold in 30 % of the vaccinated fry 30 dpv. Also, the IgM levels of the vaccinated fry were not significantly different from control fry 21 days-post challenge. A recombinase polymerase amplification (RPA) assay was developed and validated for rapid detection of Fno. The RPA reaction was performed at a constant temperature of 42°C for 20 min. The RPA assay was performed using a quantitative plasmid standard containing a unique Fno gene sequence. Validation of the assay was performed not only by using DNA from Fno, closely related Francisella species and other common bacterial pathogens in fish farms, but also by screening 78 Nile tilapia and 5 water samples collected from UK and Thailand. All results were compared with those obtained by previously established real-time qPCR. The developed RPA showed high specificity in detection of Fno with no cross-detection of either the closely related Francisella spp. or the other species of bacteria tested. The Fno-RPA performance was highly comparable to the published qPCR with detection limits at 15 and 11 DNA molecules detected, respectively. The Fno-RPA was rapid, giving results in approximately 6 min in contrast to the qPCR that required approximately 90 min to reach the same detection limits. Moreover, the RPA was more tolerant to reaction inhibitors than qPCR when tested with field samples. The fast reaction, simplicity, cost-effectiveness, sensitivity and specificity make the RPA an attractive diagnostic tool that will contribute to control the infection through prompt on-site detection of Fno. The overall results of this study indicated that Fno isolates from different origins share a high degree of homology in their proteomic and antigenic profile. Proteomic characterisation data of Fno isolates has contributed to understanding the diversity of Fno isolates and assisted in identifying suitable candidates for developing an effective Fno vaccine. / Moreover, this study has proven the efficacy of a cross protective Fno injection vaccine in tilapia fingerlings, with further optimisation needed for immersion vaccination of fry, and given insights into the immune response of tilapia to vaccination against francisellosis. In addition, it provided a rapid, sensitive, specific and robust molecular tool for detection of Fno that can assist surveillance and control of piscine francisellosis on tilapia farms.

Das Afrikanische Schweinepest-Virus (ASPV) verursacht eine tödliche Viruserkrankung bei Schweinen. Dieses hat sich weltweit fortlaufend verbreitet und wurde im September 2020 erstmalig in Deutschland nachgewiesen. Der Ausbruch der Seuche kann schwere wirtschaftliche Verluste nach sich ziehen. Bis heute ist kein Impfstoff zugelassen, daher sind Überwachung der epidemiologischen Situation und der frühzeitige Erregernachweis unerlässlich für die Bekämpfung der Afrikanischen Schweinepest als Tierseuche. Die Polymerase-Kettenreaktion (PCR) gilt als Goldstandard für den Nachweis von ASPV und zeichnet sich durch eine hohe Sensitivität und Spezifität aus. Allerdings erfordert die PCR gut ausgestattete Testlabore und ist zeitintensiv. Point-of-Need-Tests können schnelle und zuverlässige direkt vor Ort liefern und stellen somit eine Alternative zum Goldstandard PCR dar. Ziel dieser Studie war es, einen Point-of-Need-Test zum Nachweis von ASPV zu entwickeln. Dieser beruht auf der Grundlage der Rekombinase-Polymerase-Amplifikation (RPA) und sollte vor Ort einsatzfähig sein. Es wurden drei Primersätze und eine Sonde auf der Grundlage des B646L-Gens, welches für das virale Kapsidprotein p72 vom ASP-Virus kodiert, entwickelt. Alle möglichen Kombinationen wurden getestet. Die analytische Sensitivität wurde mit acht Wiederholungen von Verdünnungsreihen des molekularen Standards (102-100 DNA-Kopien pro µl) ermittelt. Die Nachweisgrenze wurde anhand einer Probit-Analyse dieser Durchläufe berechnet. Die Spezifität wurde mit verschiedenen viralen Nukleinsäuren von anderen das Schwein infizierenden Erregern überprüft. Um den Test im Feld einsatzfähig zu gestalten, wurden mittels ASPV-RPA zwei verschiedene Extraktionsansätze mit allen 73 verfügbaren Schweineblutproben getestet: eine schnelle Hitze/Lysepuffer-Extraktionsmethode und ein standardisiertes Extraktionsverfahren auf Spin-Säule-Basis. Die diagnostische Sensitivität und Spezifität wurde für beide Testverfahren berechnet. Alle Ergebnisse wurden mit einer etablierten real-time PCR für ASPV verglichen. Eine kleine Pilotstudie zum Feldeinsatz des ASPV-RPA-Tests wurde in Uganda mit 20 Blutproben unter Verwendung des Kofferlabors durchgeführt. Die berechnete Nachweisgrenze von ASPV-RPA lag bei 3,5 DNA-Kopien pro µl. Alle untersuchten ASPV-Genotypen wurden detektiert, aber keine anderen viralen Nukleinsäuren. Bei Verwendung der standardisierten DNA-Extraktionsmethode mit anschließender Durchführung der ASPV-RPA lag die diagnostische Sensitivität und Spezifität bei 100%, wie auch mittels der real-time PCR. Auch das schnelle Hitze-/Lysepuffer Protokoll zeigte vielversprechende Ergebnisse und erreichte eine Positivrate von 97% mittels ASPV-RPA im Vergleich zu 38% bei der PCR. In Uganda wurden elf ASPV-RPA-Proben als positiv erkannt, darunter zwei fieberfreie asymptomatische Tiere. Der schnelle Erregernachweis stellt einen essenziellen Aspekt der ASP Seuchenbekämpfung dar. Die ASPV-RPA erwies sich als genauso empfindlich und spezifisch wie die Goldstandard-PCR zur Erregeridentifizierung. In Kombination mit dem Schritt der DNA-Extraktion durch Hitze/Lysepuffer benötigt der entwickelte Test etwa 25 Minuten von der Probenentnahme bis zum Ergebnis. Die Positivrate ist mit 97% vielversprechend, wobei die ASPV-RPA im Vergleich zur PCR eine höhere Toleranz gegenüber Inhibitoren aufwies. Wie die Pilot-Feldstudie in Uganda mit dem Kofferlabor zeigt, ist ASPV-RPA eine im Feld einsatzfähige Nachweismethode. Das Kofferlabor bedarf lediglich einer Grundausstattung und einer Solarbatterie. Somit stellt das Kofferlabor eine vielversprechende Diagnostikmethode dar, welche vor Ort in ressourcenarmen Umgebungen zum Nachweis des ASPV eingesetzt werden kann.:1. Introduction 2. Literature overview 2.1 African swine fever 2.1.1 Aetiology 2.1.1.1 Classification and taxonomy 2.1.1.2 Viral structure and genome 2.1.1.3 Genetic typing and antigenic variability 2.1.2 Epidemiology 2.1.2.1 Disease distribution 2.1.2.2 Host range and epidemiological cycles 2.1.2.2.1 Warthog-tick cycle 2.1.2.2.2 Domestic pig-tick cycle 2.1.2.2.3 Domestic pig cycle 2.1.2.2.4 Wild boar-environment cycle 2.1.2.3 Tenacity, transmission, and infectivity 2.1.3 Pathophysiology 2.1.3.1 Pathogenesis 2.1.3.2 Clinical signs and pathological findings 2.1.3.3 Differential diagnosis 2.2 Available diagnostic tools for ASFV 2.1.4 Diagnosis based on immune response 2.1.5 Diagnosis based on agent identification 2.3 Gaps in African swine fever diagnostics 3. Publication 3.1 Statement of contribution 3.1.1 Publication 4. Discussion 5. Summary 6. Zusammenfassung 7. References 8. Appendix 9. Acknowledgements / African swine fever virus (ASFV) causes a deadly viral disease in pigs. The virus has gradually spread throughout the world and was reported in Germany in September 2020. ASF outbreak can lead to huge economical loss. No vaccine is commercially available and thus, surveillance and early detection play a pivotal role to control an ASF outbreak. Polymerase Chain Reaction (PCR) is considered the gold standard for ASFV detection due to its superior sensitivity and specificity. However, it is time-consuming and requires well-equipped laboratories. Point-of-need tests can offer an alternative, delivering fast and reliable results directly in the field. The aim of this study was to establish a field-deployable point-of-need test based on Recombinase Polymerase Amplification (RPA) to detect ASFV. Material and Methods: Three sets of primers and one probe based on the B646L gene which encodes for the viral capsid protein p72 were designed. All possible combinations were screened. Analytical sensitivity was tested with eight replicates of serial dilutions of the molecular standard (102-10° DNA copies per µl). The limit of detection was calculated using probit analysis. ASFV-RPA’s specificity was tested using various viral nucleic acids of pathogens infecting pigs. To allow the deployment at point of need, two different extraction approaches were tested in ASFV-RPA with all 73 pig blood samples included in this study: a rapid heat/lysis buffer extraction method and a standardized spin-column based extraction kit. Diagnostic sensitivity and specificity were calculated for both test approaches. All results were compared to an established real-time PCR for ASFV. A small pilot study for ASFV-RPA assay deployment was done in Uganda with 20 blood samples of a suspected outbreak using the field-deployable suitcaselab. The calculated limit of detection of ASFV-RPA was 3.5 DNA copies per µl. All screened ASFV genotypes were detected while no other viral nucleic acids were identified. Using the standardized DNA extraction method in ASFV-RPA, and compared to real-time PCR, diagnostic sensitivity and specificity were 100%. The rapid heat/lysis buffer protocol showed very promising results, achieving 97% of positivity rate compared to a 38% of the real-time PCR. In Uganda, ASFV-RPA detected 11 samples as positive, including two known afebrile animals. Immediate agent detection is a key aspect of ASF outbreak control. ASFV-RPA is as sensitive and specific as a gold standard PCR for ASFV identification. Combined with the heat/lysis buffer DNA isolation step, the duration of the assay is around 25 minutes from sample collection to result readout, with a promising positivity rate of 97% which indicates tolerance against inhibitors. ASFV-RPA is a portable detection method, as revealed during the pilot field study in Uganda. Only requiring basic equipment and solar batteries, the suitcase lab is a promising tool for on-site diagnostics in resource limited settings to detect ASFV.:1. Introduction 2. Literature overview 2.1 African swine fever 2.1.1 Aetiology 2.1.1.1 Classification and taxonomy 2.1.1.2 Viral structure and genome 2.1.1.3 Genetic typing and antigenic variability 2.1.2 Epidemiology 2.1.2.1 Disease distribution 2.1.2.2 Host range and epidemiological cycles 2.1.2.2.1 Warthog-tick cycle 2.1.2.2.2 Domestic pig-tick cycle 2.1.2.2.3 Domestic pig cycle 2.1.2.2.4 Wild boar-environment cycle 2.1.2.3 Tenacity, transmission, and infectivity 2.1.3 Pathophysiology 2.1.3.1 Pathogenesis 2.1.3.2 Clinical signs and pathological findings 2.1.3.3 Differential diagnosis 2.2 Available diagnostic tools for ASFV 2.1.4 Diagnosis based on immune response 2.1.5 Diagnosis based on agent identification 2.3 Gaps in African swine fever diagnostics 3. Publication 3.1 Statement of contribution 3.1.1 Publication 4. Discussion 5. Summary 6. Zusammenfassung 7. References 8. Appendix 9. Acknowledgements