A 6.5kb intergenic structural variation enhances P450-mediated resistance to pyrethroids in malaria vectors lowering bed net efficacy

Elucidating the complex evolutionary armory that mosquitoes deploy against insecticides is crucial to maintain the effectiveness of insecticide-based interventions. Here, we deciphered the role of a 6.5kb structural variation (SV) in driving cytochrome P450-mediated pyrethroid resistance in the malaria vector, Anopheles funestus. Whole genome pooled sequencing detected an intergenic 6.5kb SV between duplicated CYP6P9a/b P450s in pyrethroid resistant mosquitoes through a translocation event. Promoter analysis revealed a 17.5-fold higher activity (P<0.0001) for the SV-carrying fragment than the SV-free one. qRT-PCR expression profiling of CYP6P9a/b for each SV genotype supported its role as an enhancer since SV+/SV+ homozygote mosquitoes had significantly greater expression for both genes than heterozygotes SV+/SV- (1.7-2-fold) and homozygotes SV-/SV- (4-5-fold). Designing a PCR assay revealed a strong association between this SV and pyrethroid resistance (SV+/SV+ vs SV-/SV-; OR=2079.4, P=<0.001). The 6.5kb SV is present at high frequency in southern Africa (80-100%) but absent in East/Central/West Africa. Experimental hut trials revealed that homozygote SV mosquitoes had significantly greater chance to survive exposure to pyrethroid-treated Nets (OR 27.7; P < 0.0001) and to blood feed than susceptible. Furthermore, triple homozygote resistant (SV+/CYP6P9a_R/CYP6P9b_R) exhibit a higher resistance level leading to a far superior ability to survive exposure to nets than triple susceptible mosquitoes, revealing a strong additive effect. This study highlights the important role of structural variations in the development of insecticide resistance in malaria vectors and their detrimental impact on the effectiveness of pyrethroid-based nets.


9 8
genes is also in line with previous reports stating that enhancers can regulate the expression 3 9 9 of multiple genes [22]. It will be good to establish whether this 6.5kb regulates the expression  Thirdly, the geographical distribution of the 6.5kb SV tightly correlated with the high Accord retrotransposon in the regulatory region of this gene [24,25]. However, contrary to 4 1 0 other insertion elements linked with over-expression identified so far, this 6.5kb insertion 4 1 1 does not contain a transposable element but mainly putative cis-regulatory elements [13]. The design a simple PCR diagnostic assay to genotype the 6.5kb SV has allowed to 4 1 4 establish its contribution to the resistance phenotype. First, it has been shown that the 6.5kb 4 1 5 SV segregates independently from CYP6P9a and CYP6P9b and thus that it is an additional 4 1 6 genetic factor driving pyrethroid resistance beside the allelic variation of both genes 4 1 7 previously reported [11]. The independent segregation of the 6.5kb SV is also shown by the 4 1 8 increased pyrethroid resistance that it confers either when using WHO bioassays or cone insecticide-treated nets [26]. The near fixation of the 6.5kb SV seen here in South indication that this SV is strongly associated with resistance exacerbation. The fixation of the 4 2 5 6.5kb in highly resistant wild populations also suggests that escalation of pyrethroid 4 2 6 resistance could be driven, among others, by an increased metabolic resistance through 4 2 7 genetic elements such as enhancers. The strong association observed between this structural variation and pyrethroid 4 2 9 resistance either with WHO bioassays or cone assays shows that this SV can be used as a [13] and CYP6P9b [14] markers previously identified in the promoters of these genes. This 4 3 2 novel assay is even simpler than the PCR-RFLPs previously designed for CYP6P9a and 4 3 3 CYP6P9b as it does not require restriction enzymes. The detection of this 6.5kb enhancer in the cis-regulatory region of major resistance 4 3 5 genes further supports that genetic variations in this region play a major role in driving 4 3 6 metabolic resistance as seen for several resistance genes including the P450 CYP9M10 in Drosophila [24]. Therefore, cis-regulatory region of major metabolic resistance genes should 4 3 9 be thoroughly investigated to identify more markers to design simple DNA-based assay to 4 4 0 detect such resistance in mosquitoes. The geographical distribution of the 6.5kb across Africa mirrors closely that of the from other regions [30][31][32]. In contrast, the 6.5kb SV distribution is opposite to that of other 4 5 3 markers in this species notably the L119F-GSTe2 conferring DDT/pyrethroid resistance [9] 4 5 4 and the A296S-RDL dieldrin resistance marker [33]. Because the 6.5kb SV allele was already detectable although at low frequency (10.9 % and 5.2%) it is likely that this SV was 4 5 7 selected later as resistance level increased potential after greater selection pressure. Analysis 4 5 8 of larger temporal samples will further help to track the selection of this SV.

5 9
Analysis of the distribution of the 6.5kb SV and that of CYP6P9a and CYP6P9b in 4 6 0 natural populations revealed that alleles at these markers do segregate independently in the 4 6 1 field as also seen in the hybrid strain FUMOZ-R/FANG at F4 where most of the genotypic 4 6 2 combinations were observed. This suggests that regardless of the close proximity of the 3 loci 4 6 3 (a total span of 12kb), the three loci are not physically linked. The higher linkage frequency 4 6 4 observed in Mozambique and Tanzania can be due to stronger insecticide selection applied [31]. To determine which gene between CYP6P9a and CYP6P9b is more linked to the 6.5kb 4 6 7 SV, the percentage linkage for CYP6P9a and the 6.5kb SV was compared with that for CYP6P9a and the 6.5kb SV had more identical genotypes (12%) than for CYP6P9b and the 4 7 0 6.5kb SV (8%) while in Tanzania, only CYP6P9a and the 6.5kb SV identical genotypes 4 7 1 (16%) were identified and none for CYP6P9b and the 6.5kb SV (0%). Hence this SV 4 7 2 although impacting both genes as shown by the comparative qRT-PCR, appears to have a 4 7 3 greater linkage to CYP6P9a. This could also support a higher fold-change observed for 4 7 4 CYP6P9a in the field in southern Africa [2,13,19]. This could be due to the fact that the 4 7 5 6.5kb is located upstream of the 5' UTR of CYP6P9a but downstream of the 3'UTR of 4 7 6 CYP6P9b. The design of the simple PCR-based assay to genotype the 6.5kb enabled us to assess 4 7 9 the impact of such structural variation on the efficacy of insecticide-treated nets including the 4 8 0 pyrethroid-only and the PBO-synergist nets. The greater reduction of efficacy that the 6.5kb  over-expression of CYP6P9a and CYP6P9b and the fixation of all three resistance alleles 4 9 1 (6.5kb SV, CYP6P9a-R and CYP6P9b-R). The availability of additional DNA-based markers 4 9 2 such as these will now enable control programs to assess how resistance is impacting the 4 9 3 efficacy of the bed nets in their country and decide whether to adopt PBO-based nets or even 4 9 4 new generation nets with another class of insecticide than just pyrethroids. In conclusion, by elucidating the role of a 6.5kb structural variant in the pyrethroid 4 9 6 resistance in An. funestus, this study highlighted the important contribution of structural 4 9 7 variations in cis-regulatory regions in metabolic resistance in mosquitoes. It also highlighted 4 9 8 the role of enhancers in the over-expression of metabolic resistance genes. This study 4 9 9 designed a simple molecular diagnostic assay to easily monitor this P450-based metabolic 5 0 0 resistance in wild populations. The additive resistance confers by this 6.5kb in the presence of  with PBO and more preferably with new another insecticide class.  several countries across Africa were also used for genotyping including Democratic Republic  Mozambique (2016) for Southern Africa. These field samples were collected indoor using 5 1 8 electric aspirators as previously described [13].

1 9
Furthermore, reciprocal crosses between FUMOZ-R and FANG were carried out to assess the 5 2 0 correlation between the 6.5kb SV and pyrethroid resistance phenotype. To perform these 5 2 1 crosses, 30 males FUMOZ-R and 30 virgin females FANG where allowed to mate and the 5 2 2 eggs reared to the next generation and adults of following generations inter-crossed until up 5 2 3 to the F 6 generation. The resistance level of mosquitoes from F 4 to F 6 generation was tested using WHO bioassays 5 2 6 performed according to WHO protocols [35]. Briefly, 2-5 day-old, unfed female mosquitoes 5 2 7 were exposed for 1h to papers impregnated with the pyrethroids permethrin (0.75%) and deltamethrin (0.05%). Moreover, these mosquitoes were also exposed to the same papers for 30 min and 90 min to generate highly susceptible and highly resistant individuals. An 5 3 0 untreated paper was also used as negative control. The efficacy of insecticide-treated net against the strains used was tested using cone assay  The Pool-Seq data from several populations of An. funestus were analysed to confirm the 5 4 7 presence and distribution of the 6.5kb insert Africa-wide following protocol described recently [13]. Initial processing and quality assessment of the sequence data was performed 5 4 9 as described recently [13]. Pool-Seq R1/R2 read pairs and R0 reads were aligned to the 5 5 0 reference sequence using bowtie2 [37]. Variant calling was carried out using SNVer version 5 5 1 0.5.3, with default parameters [38]. SNPs were filtered to remove those with total coverage 5 5 2 depth less than 10 and more the 95 th centile for each sample as the allele frequency estimates 5 5 3 could be inaccurate due to low coverage or misaligned paralogous sequence, respectively. The Livak protocol [39] was used to extract DNA from the collected samples. The extracted 5 5 6 DNA was quantified using NanoDrop lite™ spectrophotometer (Thermo Scientifc,

7
Wilmington, USA). The intergenic region between CYP6P9a and CYP6P9b was amplified lab strain. The intergenic region for the FANG was amplified using 6P9a5F and GAP3R (S1) µL PCR reaction mix using the following conditions, initial denaturation step of 3 min at 95 (S1) primers using Phusion high-fidelity DNA polymerase (Thermos Scientific, Waltham, Massachusetts, United States). The mix was made using the GC buffer, 3% DMSO, 25mM were as follows 10 minutes at 98°C pre-denaturation, 35 cycles 10 seconds at 98°C  Cloning of the intergenic CYP6P9a-CYP6P9b for promoter activity assay The 8.2 kb intergenic region of CYP6P9a and CYP6P9b of the FUMOZ-R colony was Phusion high-fidelity DNA polymerase (Thermo Scientific) with the GC buffer, 3% DMSO, 25mM dNTPs, 10 mM of each primer and DNA from FUMOZ-R as template. The PCR 5 8 0 conditions were as follows; 10 minutes at 98°C pre-denaturation, 35 cycles 10 seconds at 5 8 1 98°C denaturation, 30 seconds at 62°C annealing, 4 minutes at 72°C extension and a final 5 8 2 extension at 72°C for 10 minutes. The PCR product was run on a 1% gel and visualized on 5 8 3 UV transilluminator. The band at 8.2 kb was gel purified and ligated to pjet1.2 vector ( Figure   5 8 4 S3a). The recombinant plasmid was digested with Sac1 and Nde1 restriction enzyme and sub-  The transfection and the luciferase assay was done as previously described [13] using An.  construct of FUMOZ-R contained the core promoter of CYP6P9a before the insertion point. Briefly, 600 ng of each promoter construct were transfected using the effectene transfection 5 9 5 reagent (Qiagen, Hilden, Germany). Two additional plasmids were used, the LRIM promoter construct which served as a positive control and the actin-Renilla internal control which is 5 9 7 used for normalization. A luminometer (EG & G Berthold, Wildbad, Germany) was used to 5 9 8 measure luciferase activity two days post transfection.

9 9
Development of a PCR for genotyping the 6.5 kb structural variant between CYP6P9a 6 0 0 and CYP6P9b.

0 1
A PCR was designed to discriminate between mosquitoes with the 8.2kb (FUMOZ-R) and Stain (Nippon genetics Europe GmbH) and revealed on UV transilluminator. To assess the ability of the assay to discriminate between mosquitoes with the structural 6 1 3 variant and those without, the assay was tried on genomic DNA extracted from 48 FUMOZ-6 1 4 R and 50 FANG female mosquitoes. And later on, mosquitoes for 8 th generation of crosses 6 1 5 between FUMOZ-R and FANG which had been exposed to 0.75% permethrin for 180 min 6 1 6 (alive providing the highly resistant) and 30 min (the dead being the highly susceptible) were 6 1 7 genotyped for this SV to assess the correlation between the insertion genotype and the 6 1 8 insecticide susceptibility phenotype. Correlation between the 6.5kb PCR assay and the CYP6P9a/b PCR-RFLP: The efficacy 6 2 0 of the newly designed PCR for the 6.5kb SV was compared with the PCR-RFLP methods for 6 2 1 detecting CYP6P9a and CYP6P9b resistant allele to check the association between the 6.5k 6 2 2 SV and each of the P450s and also for the combined effect of these three markers as 6 2 3 described in previously [13,14]. DNA samples from previous studies were used as test 6 2 4 materials to compare the three assays. Total RNA was extracted from F 2 generation of FUMOZ-R and FANG crosses which had 6 2 7 been genotyped using the newly designed 6.5k SV detection PCR and grouped in three sets: 6 2 8 homozygous for 6.5kb SV (SV+/SV+), heterozygous (SV+/SV-) and homozygous without 6 2 9 the SV (SV-/SV-). DNA was extracted from the legs. Briefly 4 to 6 legs were pulled from 6 3 0 each mosquito and placed in 1.5 ml tube. 25μl of 1X PCR buffer B (Kapa Biosystems, 6 3 1 Boston, MA, USA) pre-warmed at 65°C was added then the legs were ground. After the 6 3 2 grinding the tubes were centrifuged at 13000rpm for 2 minutes. The samples were then Technologies Carlsbad, California, United States). cDNA were synthesised from each RNA 6 3 7 pool using the Superscript III (Invitrogen, Carlsbad, California, United States) as previously quantitative real time PCR (qRT-PCR) (Agilent MX3005) to assess the correlation between 6 4 0 the presence of the 6.5kb SV and the expression pattern of these two resistance genes. To assess the association between the 6.5kb SV and a potential reduction in the efficacy of   B  h  a  t  t  S  ,  W  e  i  s  s  D  ,  C  a  m  e  r  o  n  E  ,  B  i  s  a  n  z  i  o  D  ,  M  a  p  p  i  n  B  ,  D  a  l  r  y  m  p  l  e  U  ,  e  t  a  l  .  T  h  e  e  f  f  e  c  t  o  f  m  a  l  a  r  i  a  6  6  2  c  o  n  t  r  o  l  o  n  P  l  a  s  m  o  d  i  u  m  f  a  l  c  i  p  a  r  u  m  i  n  A  f  r  i  c  a  b  e  t  w  e  e  n  2  0  0  0  a  n  d  2  0  1  5  .  N  a  t  u  r  e  .  2  0  1  5  ;  5  2  6  (  7  5  7  2  ) :   r  o  n  J  M  ,  I  r  v  i  n  g  H  ,  N  d  u  l  a  M  ,  B  a  r  n  e  s  K  G  ,  I  b  r  a  h  i  m  S  S  ,  P  a  i  n  e  M  J  ,  e  t  a  l  .  D  i  r  e  c  t  i  o  n  a  l  l  y  s  e  l  e  c  t  e  d  7  0  9  c  y  t  o  c  h  r  o  m  e  P  4  5  0  a  l  l  e  l  e  s  a  r  e  d  r  i  v  i  n  g  t  h  e  s  p  r  e  a  d  o  f  p  y  r  e  t  h  r  o  i  d  r  e  s  i  s  t  a  n  c  e  i  n  t  h  e  m  a  j  o  r  m  a  l  a  r  i  a  v  e  c  t  o  r  7  1  0  A  n  o  p  h  e  l  e  s  f  u  n  e  s  t  u  s  .  P  r  o  c  e  e  d  i  n  g  s  o  f  t  h  e  N  a  t  i  o  n  a  l  A  c  a  d  e  m  y  o  f  S  c  i  e  n  c  e  s  .  2  0  1  3  ;  1  1  0  (  1  )  :  2  5  2  -7  .  7  1  1  2  0  .  D  i  c  k  e  l  D  E  ,  V  i  s  e  l  A  ,  P  e  n  n  a  c  c  h  i  o  L  A  .  F  u  n  c  t  i  o  n  a  l  a  n  a  t  o  m  y  o  f  d  i  s  t  a  n  t  -a  c  t  i  n  g  m  a  m  m  a  l  i  a  n  7  1  2  e  n  h  a  n  c  e  r  s  .  P  h  i  l  o  s  T  r  a  n  s  R  S  o  c  L  o  n  d  B  B  i  o  l  S  c  i  .  2  0  1  3  ;  3  6  8  (  1  6  2   representation of the design of the PCR assay to detect the 6.5 kb insertion using three SV) and one located within the 6.5kb SV allowing to detect its presence (expected band the combined genotypes of 6.5 kb SV and that of CYP6P9a after WHO bioassays with 0.75% 8 1 5 permethrin showing that genotypes of both loci combined to increase the pyrethroid 8 1 6 resistance. (F) Distribution of the combined triple genotypes of 6.5kb besides that of both 8 1 7 P450s CYP6P9a and CYP6P9b after WHO bioassays with 0.75% permethrin showing that 8 1 8 the triple resistance genotypes combined to further increase the pyrethroid resistance vicinity of the 6.5 kb SV. Error bars represent standard deviation (n = 3). P<0.05 is represented by *, ** for p< 0.01, *** for p<0.001. showing the genotypic distribution of the 6.5kb SV across the continent which correlated showing and extensive segregation of the genotypes to these loci in the field. Correlation between frequency of 6.5 kb SV alleles and ability to survive exposure to allele increases the ability to take a blood meal even for PBO-based nets. PermaNet 2.0 for the various combined genotypes for the 6.5 kb SV CYP6P9a and CYP6P9b. Mozambique on selection of the 6.5 kb SV, CYP6P9a and CYP6P9b genotypes with no triple  Tables  8  6  7   Table 1: Correlation between genotypes of the 6.5kb SV and ability to survive exposure to 8 6 8 PermaNet 2.0 in experimental huts using unfed samples.  Table 2: Correlation between genotypes of the 6.5kb SV and ability to blood-feed when 8 7 1 exposed to PermaNet 3.0 in experimental huts.   . . .