Page 1

Low-Density Macroarray for Rapid Detection and Identification of 1

Crimean-Congo Hemorrhagic Fever Virus 2

3

Crimean-Congo Hemorrhagic Fever Virus Macroarray 4

5

Roman Wölfel*, Bundeswehr Institute of Microbiology, Munich, Germany 6

Janusz T. Paweska, National Institute for Communicable Diseases, Sandringham, South Africa 7

Nadine Petersen, Bernhard Nocht Institute for Tropical Medicine, Hamburg, Germany 8

Antoinette A. Grobbelaar, National Institute for Communicable Diseases, Sandringham, South 9

Africa 10

Patricia A. Leman, National Institute for Communicable Diseases, Sandringham, South Africa 11

Roger Hewson, Health Protection Agency, Porton Down, Salisbury, UK 12

Marie-Claude Georges-Courbot, Unit of Biology of Viral Emerging Infections, Institute Pas-13

teur, Lyon, France 14

Anna Papa, Microbiology Department, Medical School, Aristotle University of Thessaloniki, 15

Greece 16

Volker Heiser, Chipron GmbH, Berlin, Germany 17

Marcus Panning, Bernhard Nocht Institute for Tropical Medicine, Hamburg, Germany 18

Stephan Günther, Bernhard Nocht Institute for Tropical Medicine, Hamburg, Germany 19

Christian Drosten, Bernhard Nocht Institute for Tropical Medicine, Hamburg, Germany 20

21

22

* Dr. Roman Wölfel, Bundeswehr Institute of Microbiology, Dept. for Medical Biological Re-23

connaissance & Verification, Neuherbergstrasse 11, 80937 Munich, Germany. Phone +49-89-24

3168-3894, Fax +49-89-3168-3292, Email: romanwoelfel@bundeswehr.org 25

26

Copyright © 2009, American Society for Microbiology and/or the Listed Authors/Institutions. All Rights Reserved.J. Clin. Microbiol. doi:10.1128/JCM.01920-08 JCM Accepts, published online ahead of print on 18 February 2009

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 2

Abstract: 27

Crimean-Congo hemorrhagic fever (CCHF) is a tick-borne viral zoonosis which occurs through-28

out Africa, Eastern Europe, and Asia and results in an approximately 30% fatality rate. A RT-29

PCR assay including a competitive internal control was developed on the basis of the most up-to-30

date genome information. Biotinylated amplification products were hybridised to DNA macroar-31

rays on the surface of polymer supports and hybridisation events were visualized by incubation 32

with a streptavidin-horse radish peroxidase (HRP) conjugate and the formation of a visible sub-33

strate precipitate. Optimal assay conditions were established for detection of as few as 6.3 ge-34

nome copies per reaction. 18 geographic and historic diverse CCHF virus strains representing all 35

clinical relevant isolates were detected. The feasibility of the assay for clinical diagnosis was 36

validated with acute phase patient samples from South Africa, Iran and Pakistan. The assay pro-37

vides a specific, sensitive, and rapid method for CCHF virus detection without requiring sophisti-38

cated equipments. It has usefulness for clinical diagnosis and surveillance of CCHF infections 39

under limited laboratory conditions in developing countries or in field situations. 40

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 3

Introduction 41

Crimean–Congo haemorrhagic fever (CCHF) is a severe viral infection transmitted by hard-42

ticks (Ixodidae) of the genus Hyalomma. It has a case fatality rate around 30% and can be trans-43

mitted from human to human in the nosocomial setting (14,22). The disease is endemic in large 44

areas of Sub-Saharan Africa and the Middle-/Far East, as well as in Eastern Europe. A significant 45

increase of cases has recently been observed in countries such as Kosovo, Albania, Turkey, Iran 46

and Greece (1,9,15-18,20,21,23). The causative CCHF virus is an enveloped, segmented nega-47

tive-stranded RNA virus family Bunyaviridae, genus Nairovirus. It is classified as a biosafety-48

level four agent. 49

Due to clinical similarity between CCHF and other diseases, proper triage and isolation of 50

patients depends on laboratory confirmation (15). Available diagnostic methods include virus 51

culture, antigen-specific enzyme-linked immunoassay (EIA), antibody-specific EIA, and reverse-52

transcription polymerase chain reaction (RT-PCR) (2,4,5,24). Virus detection is necessary in the 53

acute stage of disease and RT-PCR provides the best sensitivity (8). 54

Conventional RT-PCR protocols take up to 8 h for cDNA synthesis, amplification, gel 55

analysis, and in some instances a second round of nested amplification (3,19). Sequencing of RT-56

PCR products is needed for strain identification. Real-time RT-PCR for CCHFV is difficult to 57

develop due to a remarkable genetic variability between virus strains (11,25,26). Current proto-58

cols are often not appropriate for field-based outbreak investigations, and may be difficult to im-59

plement in those countries where CCHF virus is endemic. Simpler, field-compatible assays are 60

required. Such an approach is described here. 61

A robust one-step RT-PCR assay with an internal control was established, using the most 62

recent genome information. Based on our prior experiences (25), the assay was formulated for 63

compatibility with an inexpensive and simple non fluorescent DNA macro array hybridisation 64

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 4

platform. Detection was possible with the naked eye, using simple and robust biotin / strepta-65

vidin-HRP conjugate chemistry in combination with a tetramethyl-benzidine (TMB) substrate 66

resulting in the formation of a clearly visible dark precipitate at array positions where DNA-DNA 67

hybridisations took place. No gel analysis was necessary. The possible patterns of hybridisation 68

spots were sufficiently heterogeneous to facilitate reliable differentiation between virus strains. 69

Validation was done on strain collections from several collaborating biosafety level-four facili-70

ties, essentially covering the full range of global diversity of CCHF virus (Figure 1). Clinical 71

evaluation utilized a comprehensive panel of original clinical samples from confirmed cases of 72

CCHF, collected over almost 20 years by a WHO reference facility. 73

Material and Methods 74

Virus strains 75

Eighteen CCHF virus strains representing all genetic lineages described worldwide (7) were 76

selected from the strain collections of participating laboratories (Table 1). The material was quan-77

tified by real-time RT-PCR (25) and tested by the conventional CCHF virus assay described here. 78

RNA from all strains was successfully amplified. 79

Clinical samples 80

Clinical material was provided by the Special Pathogens Unit of the National Institute for 81

Communicable Diseases, Sandringham, South Africa. A total of 63 serum samples from 31 con-82

firmed CCHF patients, received from 1986 to 2006 and stored at -70°C, were tested. Of the 31 83

confirmed cases of CCHF included in the study, 27 occurred in widely separated locations in 84

South Africa and Namibia, three in Iran, and one in Pakistan. Serum had been collected 1 to 18 85

days (Mean 6 days, SD 2.8 days) after the onset of disease. Viral load of the stored samples was 86

quantified by real-time RT-PCR (25), ranging from 1,6 × 103 to 5 × 10

8 genome copies per mL 87

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 5

(mean 106 genome copies per mL). In addition, a panel of 128 serum samples collected from 88

healthy individuals was also included in the study as negative controls. 89

RNA standards and Internal Control 90

A synthetic RNA standard was generated by amplifying the full S segment of CCHF virus 91

strain BT-958 (EF123122) as described before (13). After TA-cloning in E. coli plasmid pCR2.1 92

(Invitrogen, Germany) and sequencing, a clone with a correct insert sequence was selected and 93

the complete insert including a plasmid-derived T7 promoter was amplified by PCR. RNA was 94

transcribed from the purified PCR product with the MegaScript T7 in vitro transcription kit (Am-95

bion, USA). The DNA template was removed by DNase 1 and RNA was purified by affinity 96

chromatography (RNeasy columns, Qiagen, Germany) before spectrophotometric quantification. 97

A competitive internal control was constructed by overlap-extension PCR as described previously 98

(10). The resulting construct contained a 350 bp fragment of the CCHF virus S segment and 70 99

bp of an unrelated sequence motif. It was cloned back in pCR 2.1 and transferred into RNA as 100

described above. 101

RT-PCR 102

A 50 µL reaction contained 1 X reaction buffer (One-step RT-PCR kit, Qiagen, Germany), 103

200 µmol/L dNTP, 200 nM of each primer as listed in Table 2, 2 µL of one-step RT-PCR kit en-104

zyme mix, and 5 µL RNA extract. For subsequent hybridisation to the macroarrays, a pre-105

formulated biotinylated primer mixture (LCD-Array Kit, Chipron GmbH, Germany) was used 106

instead of the conventional primers. Amplification in a conventional PCR cycler (Primus 25, 107

Peqlab, Germany) comprised 50°C 30 min, 95°C 15 min, 40 cycles of 94°C 30 sec, 55°C 30 sec, 108

72°C 60 sec. 109

Array HybridisationLow Cost, low Density (LCD) DNA macroarrays are based on trans-110

parent, pre-structured polymer supports containing eight identical arrays in well-separated, indi-111

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 6

vidually addressable hybridisation fields (Figure 1 A). The outer dimensions of 50 × 50 mm al-112

lowed the use of economical transmission-light film scanners to generate images with 10 µm 113

resolution for data analysis. Arrays were manufactured by Chipron GmbH using contact-free 114

piezo dispensing technology. Capture probes were spotted in duplicates leading to the formation 115

of a 9 × 9 pattern, with average spot diameters of 325 µm (Figure 1 B). Twenty different CCHF 116

virus-specific capture probes between 16 and 25 nucleotides in length were selected (Table 2). 117

Each probe was designed to detect a broad range of sequence diversity at its binding site, i.e., 118

probes were not designed for specificity but for breadth of detection. Four capture probes for the 119

competitive internal control RT-PCR product were included in each array. Additional functional 120

control probes with an unrelated sequence motif were placed in three corners of each field, in 121

order to visualize successful hybridisation and staining steps, and to provide orientation marks for 122

signal analysis. Hybridisation of biotinylated RT-PCR products, labelling and staining was done 123

with the LCD-Array Detection Kit (Chipron, Germany) according to manufacturers instructions. 124

In brief, 10 µL of the biotinylated RT-PCR products were combined with 24 µL of a formamide-125

based hybridisation buffer. These mixtures were applied to the individual fields of the macroar-126

rays and hybridised for 30 min at 37°C in a standard incubator. Following a 2 min washing step 127

in the supplied wash buffer (low stringency), the arrays were dried for 10 seconds by airstream 128

using a simple compressed-air can. Dried arrays were incubated for 5 minutes (room tempera-129

ture) with the provided labelling solution (HRP-streptavidin conjugate). After a final wash-and-130

dry step (2 min, as above) 30 µl of TMB substrate was added to each field for staining (3 minutes 131

at room temperature). The reaction was stopped by rinsing the chip in the last wash buffer for 10 132

seconds followed by a drying step. 133

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 7

Statistical analysis 134

The Statgraphics Plus 5 Software Package (Umex, Germany) was used for all analysis. Pro-135

bit analysis used as its input dataset the cumulative hit rates in parallel reactions and their respec-136

tive target RNA concentrations. It determined a continuous dose-response relationship (response 137

rate of the assay dependent of the dose of RNA per reaction) with 95% confidence intervals (CI) 138

(12). 139

Results 140

The design of a novel broad-range CCHF virus RT-PCR was based on all 139 full or partial 141

s-gene sequences of CCHF virus as available in GenBank by autumn 2007 (6,7,13). Three addi-142

tional sequences were determined de novo from virus strains kept at several collaborating bio-143

safety-level four laboratories (25). Two conserved binding sites for primers were identified, re-144

sulting in a RT-PCR amplicon of 280 bp. In designing these primers, primer 3´-ends on the third 145

codon position were avoided. Degenerated ("wobble") positions in primers were not used, in or-146

der to guarantee reliable re-synthesis of primers. Instead, mismatched base pairings were adjusted 147

by mixing of defined oligonucleotides. The stable non-Watson-Crick base pairing T:G was not 148

strictly adjusted for. At each of the two binding sites, three differential primers were thus selected 149

which covered in summary the observed range of sequence heterogeneity at the binding sites. 150

The assay was optimized for sensitivity by titration of essential RT-PCR reaction mix com-151

ponents. The assay amplified a broad range of CCHF virus strains (Table 1, data not shown). To 152

determine the analytical sensitivity, limiting log10 dilution series were amplified from represen-153

tative strains Bangui BT-958, Turkey 4348/02, and ArD 39554. Down to 800, 1000 and 780 cop-154

ies per mL, respectively, could be detected. 155

To identify samples with RT-PCR inhibition, a synthetic internal control RNA was de-156

signed. To ensure amplification of the molecule without additional (possibly interfering) primers, 157

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 8

an internal control was chosen that contained the same primer binding sites as CCHF virus (com-158

petitive internal control) (10). To ensure that it did not out-compete even low amounts of virus 159

RNA in the reaction, the control was used at low concentration (see below) and its length was 160

extended over that of the virus, providing an inherent amplification disadvantage. The construct 161

was 350 bp instead of 280 bp in length, through insertion of a random sequence by overlap-162

extension PCR. 163

Functionality was evaluated in cross-titration experiments containing different amounts of 164

CCHF virus RNA and internal control RNA, respectively. When CCHF virus RNA was absent, 165

the internal control was amplified clearly (Figure 2). In the presence of increasing concentrations 166

of CCHF virus RNA, amplification of the internal control was either lower or absent because of 167

competitive inhibition by amplification of the CCHF virus genomic target (Figure 2). With in-168

creasing concentrations of internal control in reactions containing constantly low levels of CCHF 169

virus RNA (60 copies per reaction) it was demonstrated that no inhibition of virus amplification 170

was imposed by competitive effects of the internal control. A working concentration of 200 cop-171

ies of internal control per reaction was chosen for all further assays. 172

Oligonucleotides were now transformed into ready-made reaction mixtures containing pro-173

prietary modifications which allowed hybridisation of amplification products to LCD arrays. The 174

limit of modified RT-PCR was analysed by testing a series of human sera, spiked with synthetic 175

full-length S segment RNA from 100,000 to 10 copies per mL. On each concentration step, five 176

replicate test reactions were conducted and the results subjected to probit analysis. Statistically 177

constant detection could be achieved with as little as 10 copies of RNA per reaction (data not 178

shown). Sporadic detection was possible down to a single copy per reaction. Probit analysis de-179

termined a 95% detection limit of 540 copies per mL of serum, corresponding to 6.3 genome cop-180

ies per reaction (95% CI: 4.3-14.3 copies per assay, p=0.05). 181

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 9

Hybridisation to LCD arrays was evaluated next. All 18 virus reference strains listed in Table 1 182

were amplified and hybridised (Figure 3). Signals after the hybridisation and staining procedure 183

were clearly visible to the naked eye. Repeated testing of a virus sample showed constantly iden-184

tical hybridisation patterns (data not shown). The whole protocol took less than 4 h and up to 48 185

samples could easily be analysed in parallel. Documentation images as shown in Figure 3 were 186

obtained using a standard slide scanner purchased from a department store. Array hybridisation 187

provided the same sensitivity as gel detection, as shown in Figure 2. However, at virus concentra-188

tions below 1,000 copies per reaction, some of the capture probes which generated only weak 189

byridisation signals at higher template concentrations were not visible. This might limit the abil-190

ity to discriminate certain CCHF virus strains in samples with very low viral load. 191

The specificity of the RT-PCR assay was verified using purified genomic nucleic acids 192

from culture supernatant or high-titered patient containing pathogens that cause diseases resem-193

bling CCHF infections: Bacillus anthracis, Bacillus cereus, Bacillus subtilis, Coxiella burnetii, 194

Cytomegalovirus, Dengue virus types1-4, Dugbe virus, Ebola virus strain Gulu, Epstein Barr 195

virus, Hepatitis C virus, Japanese encephalitis virus, Lassa virus strain AV, Leptospira interro-196

gans, Listeria monocytogenes, Monkeypox virus, Neisseria menigitidis, Plasmodium falciparum, 197

Poliomyelitis virus types 1, 2, 3, Rabies virus, Rickettsia prowazekii, Rickettsia rickettsii, Rift 198

Valley Fever virus, Ross River virus, Sindbis Virus, Venezuelan Equine Encephalitis virus, West 199

Nile virus strain Uganda, Yellow Fever virus. None of these materials, including Dugbe virus 200

(strain AF014014), a non-pathogenic Nairovirus related to CCHF virus, showed any reactivity. 201

None of these reactions showed any random reactivity, i.e. additional non-specific bands in gel 202

electrophoresis or background hybridisation (data not shown). 203

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 10

Clinical application 204

A total of 63 serum samples from 31 confirmed CCHF patients, received from 1986 to 205

2006, were tested after they had been stored at -70°C for up to 6 years. Of the 31 confirmed cases 206

of CCHF included in the study, 27 occurred in widely separated locations in South Africa and 207

Namibia, three in Iran, and one in Pakistan. Serum had been collected 1 to 18 days (Mean 6 days, 208

SD 2.8 days) after the onset of disease. Viral load of the stored samples was quantified by real-209

time RT-PCR (25), ranging from 1,6 × 103 to 5 × 10

8 genome copies per mL (mean 10

6, SD 20 210

genome copies per mL). All samples were positive in all assays. Serial samples of same patients 211

showed identical hybridisation patterns (Figure 4). All tested clinical samples from different 212

CCHF endemic regions could be distinguished on the basis of their hybridisation pattern (com-213

pare Figure 3). 214

A panel of 128 serum samples collected from healthy individuals was also used as negative 215

controls. All negative samples showed amplification of the internal control, but no CCHF virus-216

specific signal in the hybridisation assay (data not shown). 217

Discussion 218

The remarkable genetic variability of all known CCHF virus isolates involves a risk of hu-219

man infection by divergent strains which current protocols may not detect. At the same time, 220

even very simple versions of current assays may fail in basic laboratory settings typically encoun-221

tered in countries were CCHF is endemic. Compared to current molecular assays such as nested 222

RT-PCR or real-time RT-PCR, the identification of the amplicons by hybridisation to up to 20 223

different capture probes gives additional sequence verification. Compared to real-time RT-PCR it 224

is more reliable for new virus strains, which may not be conserved at the binding site of a single 225

hybridisation probe. The assay does not depend on expensive lab infrastructure and therefore it 226

can be applied more easily in basic laboratory settings or in the field. The use of an array of short 227

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 11

DNA capture probes (16–25 nt in length) gives a basic genotyping capacity, even though it may 228

not be sufficient for final CCHF virus strain characterization especially in samples with low viral 229

load (compare Figure 2). Nevertheless, preliminary genotyping can be very helpful in a diagnos-230

tic situation when a possible PCR contamination has to be ruled out as a real positive reaction in 231

most cases will provide a hybridisation pattern different from the positive control. The assay pro-232

vides an economical, rapid, and convenient method to identify CCHF virus in acute phase human 233

serum samples without requiring sophisticated equipment. It might prove useful for clinical diag-234

nosis and surveillance of CCHF under basic laboratory conditions in developing countries or in 235

outbreak investigations. 236

Acknowledgments 237

The research described herein is part of the Medical Biological Defence Research Program 238

of the Bundeswehr Joint Medical Service. It was sponsored in parts by the European Union 239

Framework Program 6 (EU FP6) “Viral Haemorrhagic Fevers/Variola-PCR”. Opinions, interpre-240

tations, conclusions, and recommendations are those of the authors and are not necessarily en-241

dorsed by any governmental agency, department or other institutions. 242

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 12

References 243

1. Alavi-Naini, R., A. Moghtaderi, H. R. Koohpayeh, B. Sharifi-Mood, M. Naderi, M. 244

Metanat, and M. Izadi. 2006. Crimean-Congo hemorrhagic fever in Southeast of Iran. 245

J.Infect. 52:378-382. 246

2. Burt, F. J., P. A. Leman, J. C. Abbott, and R. Swanepoel. 1994. Serodiagnosis of Cri-247

mean-Congo haemorrhagic fever. Epidemiol.Infect. 113:551-562. 248

3. Burt, F. J., P. A. Leman, J. F. Smith, and R. Swanepoel. 1998. The use of a reverse 249

transcription-polymerase chain reaction for the detection of viral nucleic acid in the diag-250

nosis of Crimean-Congo haemorrhagic fever. J.Virol.Methods 70:129-137. 251

4. Burt, F. J., R. Swanepoel, and L. E. Braack. 1993. Enzyme-linked immunosorbent as-252

says for the detection of antibody to Crimean-Congo haemorrhagic fever virus in the sera 253

of livestock and wild vertebrates. Epidemiol.Infect. 111:547-557. 254

5. Casals, J. and G. H. Tignor. 1974. Neutralization and hemagglutination-inhibition tests 255

with Crimean hemorrhagic fever-Congo virus. Proc.Soc.Exp.Biol.Med. 145:960-966. 256

6. Chamberlain, J., N. Cook, G. Lloyd, V. Mioulet, H. Tolley, and R. Hewson. 2005. Co-257

evolutionary patterns of variation in small and large RNA segments of Crimean-Congo 258

hemorrhagic fever virus. J.Gen.Virol. 86:3337-3341. 259

7. Deyde, V. M., M. L. Khristova, P. E. Rollin, T. G. Ksiazek, and S. T. Nichol. 2006. 260

Crimean-Congo hemorrhagic fever virus genomics and global diversity. J.Virol. 80:8834-261

8842. 262

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 13

8. Drosten, C., S. Gottig, S. Schilling, M. Asper, M. Panning, H. Schmitz, and S. Gun-263

ther. 2002. Rapid detection and quantification of RNA of Ebola and Marburg viruses, 264

Lassa virus, Crimean-Congo hemorrhagic fever virus, Rift Valley fever virus, dengue vi-265

rus, and yellow fever virus by real-time reverse transcription-PCR. J.Clin.Microbiol. 266

40:2323-2330. 267

9. Drosten, C., D. Minnak, P. Emmerich, H. Schmitz, and T. Reinicke. 2002. Crimean-268

Congo hemorrhagic fever in Kosovo. J.Clin.Microbiol. 40:1122-1123. 269

10. Drosten, C., M. Weber, E. Seifried, and W. K. Roth. 2000. Evaluation of a new PCR 270

assay with competitive internal control sequence for blood donor screening. Transfusion 271

40:718-724. 272

11. Duh, D., A. Saksida, M. Petrovec, I. Dedushaj, and T. vsic-Zupanc. 2006. Novel one-273

step real-time RT-PCR assay for rapid and specific diagnosis of Crimean-Congo hemor-274

rhagic fever encountered in the Balkans. J.Virol.Methods 133:175-179. 275

12. Fink, H. and G. Hund. 1965. [Probit analysis with programmed computers]. Arzneimit-276

telforschung. 15:624-630. 277

13. Hewson, R., J. Chamberlain, V. Mioulet, G. Lloyd, B. Jamil, R. Hasan, A. Gmyl, L. 278

Gmyl, S. E. Smirnova, A. Lukashev, G. Karganova, and C. Clegg. 2004. Crimean-279

Congo haemorrhagic fever virus: sequence analysis of the small RNA segments from a 280

collection of viruses world wide. Virus Res. 102:185-189. 281

14. Hoogstraal, H. 1979. The epidemiology of tick-borne Crimean-Congo hemorrhagic fever 282

in Asia, Europe, and Africa. J.Med.Entomol. 15:307-417. 283

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 14

15. Jamil, B., R. S. Hasan, A. R. Sarwari, J. Burton, R. Hewson, and C. Clegg. 2005. 284

Crimean-Congo hemorrhagic fever: experience at a tertiary care hospital in Karachi, Paki-285

stan. Trans.R.Soc.Trop.Med.Hyg. 99:577-584. 286

16. Karti, S. S., Z. Odabasi, V. Korten, M. Yilmaz, M. Sonmez, R. Caylan, E. Akdogan, 287

N. Eren, I. Koksal, E. Ovali, B. R. Erickson, M. J. Vincent, S. T. Nichol, J. A. Comer, 288

P. E. Rollin, and T. G. Ksiazek. 2004. Crimean-Congo hemorrhagic fever in Turkey. 289

Emerg.Infect.Dis. 10:1379-1384. 290

17. Papa, A., S. Bino, A. Llagami, B. Brahimaj, E. Papadimitriou, V. Pavlidou, E. Velo, 291

G. Cahani, M. Hajdini, A. Pilaca, A. Harxhi, and A. Antoniadis. 2002. Crimean-292

Congo hemorrhagic fever in Albania, 2001. Eur.J.Clin.Microbiol.Infect.Dis. 21:603-606. 293

18. Papa, A., I. Christova, E. Papadimitriou, and A. Antoniadis. 2004. Crimean-Congo 294

hemorrhagic fever in Bulgaria. Emerg.Infect.Dis. 10:1465-1467. 295

19. Rodriguez, L. L., G. O. Maupin, T. G. Ksiazek, P. E. Rollin, A. S. Khan, T. F. 296

Schwarz, R. S. Lofts, J. F. Smith, A. M. Noor, C. J. Peters, and S. T. Nichol. 1997. 297

Molecular investigation of a multisource outbreak of Crimean-Congo hemorrhagic fever 298

in the United Arab Emirates. Am.J.Trop.Med.Hyg. 57:512-518. 299

20. Schwarz, T. F., H. Nitschko, G. Jager, H. Nsanze, M. Longson, R. N. Pugh, and A. K. 300

Abraham. 1995. Crimean-Congo haemorrhagic fever in Oman. Lancet 346:1230. 301

21. Schwarz, T. F., H. Nsanze, and A. M. Ameen. 1997. Clinical features of Crimean-302

Congo haemorrhagic fever in the United Arab Emirates. Infection 25:364-367. 303

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 15

22. Swanepoel, R., D. E. Gill, A. J. Shepherd, P. A. Leman, J. H. Mynhardt, and S. Har-304

vey. 1989. The clinical pathology of Crimean-Congo hemorrhagic fever. Rev.Infect.Dis. 305

11 Suppl 4:S794-S800. 306

23. Swanepoel, R., J. K. Struthers, A. J. Shepherd, G. M. McGillivray, M. J. Nel, and P. 307

G. Jupp. 1983. Crimean-congo hemorrhagic fever in South Africa. Am.J.Trop.Med.Hyg. 308

32:1407-1415. 309

24. Whitehouse, C. A. 2004. Crimean-Congo hemorrhagic fever. Antiviral Res. 64:145-160. 310

25. Wolfel, R., J. T. Paweska, N. Petersen, A. A. Grobbelaar, P. A. Leman, R. Hewson, 311

M. C. Georges-Courbot, A. Papa, S. Gunther, and C. Drosten. 2007. Virus detection 312

and monitoring of viral load in Crimean-Congo hemorrhagic fever virus patients. 313

Emerg.Infect.Dis. 13:1097-1100. 314

26. Yapar, M., H. Aydogan, A. Pahsa, B. A. Besirbellioglu, H. Bodur, A. C. Basustaoglu, 315

C. Guney, I. Y. Avci, K. Sener, M. H. Setteh, and A. Kubar. 2005. Rapid and quantita-316

tive detection of Crimean-Congo hemorrhagic fever virus by one-step real-time reverse 317

transcriptase-PCR. Jpn.J.Infect.Dis. 58:358-362. 318

319

320

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 16

Figures: 321

322

FIG. 1. Schematic diagram of microarray. A: Illustration of the 50 × 50 mm polymer support 323

with the eight identical, individually addressable array fields. B: Spotting pattern of one array 324

field. 20 CCHF virus-specific capture probes were spotted as vertical duplicates in a 9 × 9 pattern 325

with average spot diameters of 325 µm (positions 1-20). Four capture probes for the competitive 326

internal control RT-PCR product are included at the bottom of each array (positions 21). Addi-327

tional functional control probes to visualize successful hybridisation and staining are immobilized 328

in three angles of each field (positions ‘C’). 329

330

FIG. 2. RT-PCR amplification with the CCHF virus-specific assay (280-bp amplicon) and 331

the internal control (350-bp amplicon). A: Conventional 1.5% agarose gel analysis of the RT-332

PCR products. B: Specific hybridisation pattern of the same RT-PCR products on the macroarray. 333

Lane M: 100-bp molecular size ladder; lanes/fields 1-3: CCHF virus strain BT-958 in-vitro tran-334

scribed RNA, lane/field 1: 6 × 106 copies per reaction (cps/rx), lane/field 2: 6 × 10

4 cps/rx, 335

lane/field 3: 6 cps/rx; lane/field 4: internal control RNA only; lane/field 5: control to which no 336

RNA was added. Note that both analysis methods visualize the suppression of the internal control 337

amplification by increasing concentrations of CCHF virus target RNA. 338

339

340

341

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 17

FIG. 3. A, Global distribution and phylogenetic relationships of CCHF virus strains se-342

lected for the design and validation of the assay: Phylogenetic analysis was based on NCBI-343

available 450bp sequences of CCHF virus small (S) segment and generated by the neighbour-344

joining method with TreeCon for Windows Version 1.3b (Yves van de Peer, University Kon-345

stanz, Germany). * These CCHF virus strains were not available for testing with the novel uni-346

versal CCHF virus qRT-PCR assay, but genetically closely related isolates have been tested. ** 347

Strain AP92 has also not been available for testing. It was isolated from a Rhipicephalus bursa 348

tick and has never been associated with human disease. 349

B, Representative hybridisation patterns of CCHF virus strains listed in table 2: Different 350

strains of CCHF virus show individual hybridisation patterns on the macroarray. However these 351

patterns are based only on sequence variability within an approximately 25-nt region of the 352

CCHF virus S segment. Therefore they cannot be considered unique for a specific CCHF virus 353

strain as shown in patterns 5 and 6. Dugbe virus, a non-pathogenic Nairovirus closely related to 354

CCHF virus is not detected by the CCHF virus-specific array. Note that the internal control spots 355

are not visible in the CCHF virus patterns as the amplification of the internal control RNA is sup-356

pressed in the presence of high concentrations of CCHF virus RNA (also compare Fig. 2). 357

358

FIG. 4. Serial samples of an acute case of CCHF. This example shows hybridisation patterns 359

obtained from serum samples on days 1, 5 and 9 days after onset of the disease. Viral load of the 360

samples (genome copies per reaction, cps/rx) was quantified by real-time RT-PCR as described 361

before (25). 362

363

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 18

364

365

TABLE 1. CCHF virus strains used for validation of the RT-PCR assay

CCHF virus isolate Origin NCBI-Genbank

Accession No. Log RNA copies per mL

7803 XinJiang China AF354296 5.2

Ug3010 Congo U88416 7.1

ArB 604 "1976" Congo U15092 5.4

SPU 190/00/18 Iran AY905654 5.8

SPU9/00/5 Iran AY905653 6.3

Baghdad-12 Iraq AJ538196 4.5

ArMg951 Madagascar U15024 9.3

ArD39554 Mauritania U15089 7.9

IbAr10200 Nigeria U88410 8.9

SPU 280/02/10 Pakistan AY905663 7.1

SPU 68/98 South Africa AY905639 6.5

SPU 60/89 South Africa AY905636 5.9

SPU 70/01 South Africa AY905650 6.2

SPU 51/01 South Africa AY905649 6.6

4348/02 Turkey DQ211649 4.0

SPU128/81/7 Uganda DQ076415 5.5

Hoti Yugoslavia (Kosovo) DQ133507 4.3

Bangui BT-958 Centralafric. Republic EF123122 6.9

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

Page 19

366

367

TABLE 2. Primer and Probe sequences used for RT-PCR and macroarray hybridi-sation

Name Sequence Position on CCHF virus

strain 10200 (U88410)

CC1a_for 5’-GTGCCACTGATGATGCACAAAAGGATTCCATCT 210-242

CC1b_for 5’-GTGCCACTGATGATGCACAAAAGGATTCTATCT 210-242

CC1c_for 5’-GTGCCACTGATGATGCACAAAAGGACTCCATCT 210-242

CC1a_rev 5’-GTGTTTGCATTGACACGGAAACCTATGTC 489-461

CC1b_rev 5’-GTGTTTGCATTGACACGGAAGCCTATGTC 489-461

CC1c_rev 5’-GTGTTTGCATTGACACGGAAACCTATATC 489-461

CCHF-01 5’-CAACAGGCTGCTCTCAAGTGGAG

CCHF-02 5’-CCAGCAGGCTGCTCTCAAGTGG

CCHF-03 5’-CCAACAAGCTGCCTTGAAATGG

CCHF-04 5’-CCAACAGGCTGCCTTGAAATGG

CCHF-05 5’-CCAACAGGCTGCTCTAAAGTGGAG

CCHF-06 5’-CCAACAAGCTGCCTTGAAGTGG

CCHF-07 5’-CAGCAGGCTGCTCTCAAGTGG

CCHF-08 5’-CAGCAGGCCGCTCTCAAGTG

CCHF-09 5’-CAACAGGCTGCTCTCAAATGGAG

CCHF-10 5’-CCAACATGCTGCTCTCAAGTGGA

CCHF-11 5’-AGCAAGCTGCCCTCAAGTGGA

CCHF-12 5’-TCAACAGGCTGCTCTAAAGTGGAGA

CCHF-13 5’-AGCAGGCAGCCCTCAAGTGG

CCHF-14 5’-CAACAAGCCGCCTTAAAGTGGAG

CCHF-15 5’-CAACAAGCCGCCTTGAAGTGG

CCHF-16 5’-CAACAGGCTGCCTTGAAGTGGA

CCHF-17 5’-CAACAGGCTGCTTTGAAATGGAG

CCHF-18 5’-CCAGCAGGCTGCTCTGAAGTG

CCHF-19 5’-CCAGCAGGCTGCTCTAAAGTGG

CCHF-20 5’-GCAGGCCGCCCTCAAGTG

434-456

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

on March 2, 2020 by guest

http://jcm.asm

.org/D

ownloaded from